c/EPA
United states March 2011
Environments Protection rn\/^r\r\m -if\if\^^\
Agency EPA/600/R-10/076 A
Integrated Science Assessment for Ozone and
Related Photochemical Oxidants
National Center for Environmental Assessment-RTF Division
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC
-------
Disclaimer
This document is the first external review draft for review purposes only and does not constitute
U.S. Environmental Protection Agency policy. Mention of trade names or commercial products does
not constitute endorsement or recommendation for use.
March 2011
-------
Table of Contents
LIST OF TABLES xm
LIST OF FIGURES xvn
OZONE PROJECT TEAM xxvm
AUTHORS, CONTRIBUTORS, AND REVIEWERS xxxi
CLEAN AIR SCIENTIFIC ADVISORY COMMITTEE OZONE NAAQS REVIEW PANEL xxxvi
ACRONYMS AND ABBREVIATIONS xxxvm
CHAPTER 1. INTRODUCTION 1-1
1.1. Legislative Requirements 1-2
1.2. History of the NAAQS for Ozone
1.3. ISA Development
1.4. Document Organization
1.5. Document Scope
1.6. EPA Framework for Causal Determination
1 .6. 1 . Scientific Evidence Used in Establishing Causality
1.6.2. Association and Causation
1.6.3. Evaluating Evidence for Inferring Causation
1.6.4. Application of Framework for Causal Determination
1.6.5. Determination of Causality
1.6.5.1. Effects on Human Populations
1.6.5.2. Effects on Ecosystems or Public Welfare
1.6.6. Concepts in Evaluating Adversity of Health Effects
1.7. Summary
1-4
1-7
1-11
1-12
1-12
1-13
1-14
1-14
1-18
1-20
1-22
1-24
1-24
1-25
References 1-26
CHAPTER 2. INTEGRATIVE HEALTH AND WELFARE EFFECTS OVERVIEW 2-1
2.1. Atmospheric Chemistry and Ambient Concentrations 2-2
2.1.1. Physical and Chemical Processes 2-3
2.1.1.1. Gas Phase Reactions Leading to Ozone Formation and Loss 2-3
2.1.1.2. Sources of Precursors Involved in Ozone Formation 2-4
2.1.2. Atmospheric Modeling 2-4
2.1.3. Policy Relevant Background Concentrations 2-5
2.1.3.1. Contributions from anthropogenic emissions outside North America 2-6
2.1.3.2. Contributions from the stratosphere 2-6
2.1.3.3. Natural sources of precursors to PRB Ozone formation 2-6
2.1.3.4. Estimating PRB Concentrations 2-7
2.1.4. Monitoring 2-8
2.1.4.1. Routine Monitoring Techniques 2-8
2.1.4.2. Ambient Ozone Network Design 2-8
2.1.5. Ambient Concentrations 2-9
March 2011 iii DRAFT - DO NOT CITE OR QUOTE
-------
2.1.5.1. Urban-Focused Spatial Variability 2-9
2.1.5.2. Rural-Focused Spatial Variability 2-10
2.1.5.3. National Trends 2-11
2.1.5.4. Hourly Variations 2-11
2.1.5.5. Associations with Co-pollutants 2-12
2.2. Human Exposure 2-12
2.2.1. Exposure Measurement 2-12
2.2.1.1. Measurement of Ozone Exposure 2-12
2.2.1.2. Co-Exposure to Ozone and Other Pollutants 2-14
2.2.2. Exposure Modeling 2-14
2.2.3. Implications for Epidemiologic Studies 2-15
2.3. Dosimetry and Mode of Action 2-16
2.3.1. Human and Animal Ozone Dosimetry 2-16
2.3.2. Possible Pathways/Modes of Action 2-17
2.4. Health Effects 2-21
2.4.1. Effects of Short-Term Exposure to Ozone 2-21
2.4.1.1. Respiratory Effects 2-21
2.4.1.2. Cardiovascular Effects 2-25
2.4.1.3. Central Nervous System Effects 2-25
2.4.1.4. Mortality 2-26
2.4.2. Effects of Long-Term Exposure to Ozone 2-28
2.4.2.1. Respiratory Effects 2-28
2.4.2.2. Cardiovascular Effects 2-29
2.4.2.3. Reproductive and Developmental Effects 2-29
2.4.2.4. Central Nervous System Effects 2-30
2.4.2.5. Cancer, Mutagenicity and Genotoxicity 2-30
2.4.2.6. Mortality 2-31
2.5. Policy Relevant Considerations 2-31
2.5.1. Potentially Susceptible Populations 2-31
2.5.2. Lag Structure of Ozone-Morbidity and Ozone-Mortality Associations 2-32
2.5.2.1. Ozone-Respiratory Effect Associations 2-33
2.5.2.2. Ozone-Mortality Associations 2-33
2.5.3. Ozone Concentration-Response Relationship 2-34
2.5.3.1. Concentration-Response Relationship Characterized by Controlled
Human Exposure Studies 2-34
2.5.3.2. Concentration-Response Relationship Characterized by Epidemiologic
Studies 2-34
2.6. Integration of Ozone Health Effects 2-35
2.6.1. Respiratory Health Effects 2-38
2.6.2. Mortality Effects 2-45
2.6.3. Cardiovascular Health Effects 2-46
2.6.4. Central Nervous System Effects 2-46
2.6.5. Reproductive and Developmental Effects 2-46
2.6.6. Cancer and Mutagenicity and Genotoxicity 2-47
2.7. Effects on Vegetation and Ecosystems 2-47
2.7.1. Mechanisms Governing Response 2-49
2.7.2. Nature of Effects on Vegetation 2-52
2.7.2.1. Effects on Woody and Herbaceous Vegetation 2-52
2.7.2.2. Agricultural Crops 2-55
2.7.2.3. Factors That Modify Functional and Growth Response 2-55
2.7.3. Ecosystems and Services 2-56
2.7.3.1. Productivity and Carbon Sequestration 2-56
2.7.3.2. Water Cycling 2-57
2.7.3.3. Below-Ground Processes 2-57
2.7.3.4. Community Composition 2-58
March 2011 iv DRAFT - DO NOT CITE OR QUOTE
-------
2.7.4. Air Quality Indices 2-59
2.7.4.1. Modeled Ozone Deposition or "Flux" 2-60
2.7.4.2. Night-Time Exposures 2-60
2.7.5. Exposure-Response 2-60
2.8. The Role of Tropospheric Ozone in Climate Change and UV-B Effects 2-62
2.8.1. Tropospheric Ozone as a Greenhouse Gas 2-62
2.8.2. Tropospheric Ozone and UV-B related effects 2-63
2.9. Summary of Causal Determinations for Health Effects and Welfare Effects 2-64
References 2-66
CHAPTER 3. ATMOSPHERIC CHEMISTRY AND AMBIENT CONCENTRATIONS 3-1
3.1. Introduction 3-1
3.2. Physical and Chemical Processes 3-1
3.2.1. Sources of Precursors Involved in Ozone Formation 3-4
3.2.2. Gas Phase Reactions Leading to Ozone Formation 3-8
3.2.3. Multiphase Processes 3-12
3.2.4. Temperature and Chemical Precursor Relationships 3-13
3.3. Atmospheric Modeling 3-17
3.3.1. Global Scale CTMs 3-23
3.4. Policy Relevant Background Concentrations 3-25
3.4.1. Contributions from Anthropogenic Emissions Outside North America 3-26
3.4.2. Contributions from the Stratosphere 3-29
3.4.2.1. Other Natural Sources of Precursors to PRB Ozone Formation 3-30
3.4.3. Estimating PRB Concentrations 3-31
3.5. Monitoring 3-40
3.5.1. Routine Monitoring Techniques 3-40
3.5.2. Precision and Bias 3-42
3.5.2.1. Precision from Co-located UV Ozone Monitors in Missouri 3-44
3.5.3. Performance Specifications 3-46
3.5.4. Monitor Calibration 3-46
3.5.5. Other Monitoring Techniques 3-47
3.5.5.1. Portable UV Ozone Monitors 3-47
3.5.5.2. Teledyne Advanced Pollution Instrumentation Model 265E CLM 3-48
3.5.5.3. Passive Air Sampling Devices and Sensors 3-48
3.5.5.4. Differential Optical Absorption Spectrometry 3-49
3.5.5.5. Satellite Remote Sensing 3-50
3.5.6. Ambient Ozone Network Design 3-50
3.5.6.1. Monitor Siting Requirements 3-50
3.5.6.2. Probe/Inlet Siting Requirements 3-54
3.6. Ambient Concentrations 3-55
3.6.1. Measurement Units, Metrics, and Averaging Times 3-55
3.6.2. Spatial Variability 3-57
3.6.2.1. Urban-Focused Variability 3-57
3.6.2.2. Rural-Focused Variability and Ground-Level Vertical Gradients 3-86
3.6.3. Temporal Variability 3-93
3.6.3.1. Multiyear Trends 3-93
3.6.3.2. Hourly Variations 3-95
3.6.4. Associations with Co-pollutants 3-97
3.7. Chapter 3 References 3-100
3.8. Chapter 3 Appendix - Supplemental Figures and Tables 3-111
3.8.1. Time Series of GEOS-Chem Model Predictions and Observations at Selected
CASTNET Sites 3-111
March 2011 v DRAFT - DO NOT CITE OR QUOTE
-------
3.8.2. Ozone Monitor Maps for the Urban Focus Cities 3-119
3.8.3. Ozone Concentration Box Plots for the Urban Focus Cities 3-139
3.8.4. Ozone Concentration Relationships for the Urban Focus Cities 3-152
3.8.5. Hourly Variations in Ozone for the Urban Focus Cities 3-193
CHAPTER 4. EXPOSURE TO AMBIENT OZONE 4-1
4.1. Introduction 4-1
4.2. General Exposure Concepts 4-1
4.3. Exposure Measurement 4-3
4.3.1. Personal Monitoring Techniques 4-3
4.3.2. Indoor-Outdoor Concentration Relationships 4-4
4.3.3. Personal-Ambient Concentration Relationships 4-5
4.3.3.1. Personal-Ambient Correlations 4-5
4.3.3.2. Personal-Ambient Ratios 4-7
4.3.4. Co-Exposure to Other Pollutants and Environmental Stressors 4-8
4.3.4.1. Personal Exposure to Ozone and Co-pollutants 4-8
4.3.4.2. Near-Road Exposure to Ozone and Co-pollutants 4-9
4.3.4.3. Indoor Exposure to Ozone and Co-pollutants 4-10
4.3.5. Population Proximity to Fixed-Site Ozone Monitors 4-10
4.4. Exposure Modeling 4-12
4.4.1. Concentration Surface Modeling 4-12
4.4.2. Microenvironmental Models 4-14
4.4.3. Hybrid Model Approaches 4-17
4.5. Implications for Epidemiologic Studies 4-17
4.5.1. Exposure Measurement Error 4-17
4.5.2. Nonambient Ozone Exposure 4-18
4.5.3. Spatiotemporal Variability 4-18
4.5.3.1. Spatial Variability 4-18
4.5.3.2. Seasonality 4-19
4.5.4. Exposure to Co-pollutants and Ozone Reaction Products 4-20
4.5.5. Exposure Estimation Methods in Epidemiologic Studies 4-20
4.6. Summary and Conclusions 4-21
4.6.1. Exposure Measurement 4-21
4.6.1.1. Measurement of Ozone Exposure 4-21
4.6.1.2. Co-Exposure to Ozone and Other Pollutants 4-23
4.6.2. Exposure Modeling 4-23
4.6.3. Implications for Epidemiologic Studies 4-24
References 4-26
CHAPTER 5. DOSIMETRY AND MODE OF ACTION 5-1
5.1. Human and Animal Ozone Dosimetry 5-1
5.1.1. Introduction 5-1
5.1.2. Ozone Reactions and Reaction Products 5-2
5.1.2.1. Summary of Findings from 2006 Ozone AQCD 5-2
5.1.2.2. Recent Publications 5-6
5.1.3. Ozone Uptake 5-7
5.1.3.1. Summary of Findings from the 2006 Ozone AQCD 5-7
5.1.3.2. Recent Publications 5-15
5.1.4. Species Homology, Sensitivity, and Animal-to-Human Dose Extrapolation 5-16
5.1.4.1. Summary of Findings from 2006 Ozone AQCD 5-16
5.1.4.2. Recent Publications 5-19
5.2. Possible Pathways/Modes of Action 5-21
5.2.1. Introduction 5-21
March 2011 vi DRAFT - DO NOT CITE OR QUOTE
-------
5.2.2. Formation of Secondary Oxidation Products in the Respiratory Tract
5.2.2.1. Summary
5.2.3. Activation of Neural Reflexes
5.2.3.1. New Cellular and Molecular Insights
5.2.3.2. Summary
5.2.4. Respiratory Tract Injury and Inflammation
5.2.4.1. New Cellular and Molecular Insights
5.2.4.2. Summary
5.2.5. Increased Bronchial Reactivity
5.2.5.1. New Cellular and Molecular Insights
5.2.5.2. Summary
5.2.6. Exacerbation and Induction of Asthma and Allergic Responses
5.2.6.1. New Cellular and Molecular Insights
5.2.6.2. Summary
5.2.7. Impaired Host Defense
5.2.7.1. New Cellular and Molecular Insights
5.2.7.2. Summary
5.2.8. Extrapulmonary Effects
5.2.8.1. Cardiovascular Effects
5.2.8.2. Hepatic Effects
5.2.8.3. Summary
5.2.9. Factors Affecting Responses to Ozone
5.2.9.1. Gene-Environment Interactions
5.2.9.2. Preexisting Diseases and Conditions
5.2.9.3. Lifestage: Postnatal development
5.2.9.4. Lifestage: Aging
5.2.9.5. Adaptation
5.2.9.6. Co-Exposures with Particulate Matter
5.2.9.7. Summary
5.2.10. Overall Summary
5.2.11. Gaps in Knowledge
References
5-22
5-28
5-29
5-32
5-33
5-33
5-39
5-40
5-41
5-42
5-44
5-44
5-47
5-49
5-50
5-51
5-51
5-52
5-52
5-53
5-54
5-54
5-54
5-56
5-57
5-59
5-59
5-59
5-60
5-60
5-61
5-62
CHAPTER 6. INTEGRATED HEALTH EFFECTS OF SHORT-TERM OZONE EXPOSURE 6-1
6.1. Introduction 6-1
6.2. Respiratory Effects 6-1
6.2.1. Lung Function 6-3
6.2.1.1. Controlled Human Exposure 6-3
6.2.1.2. Epidemiology 6-16
6.2.1.3. Toxicology 6-39
6.2.2. Airway Hyperresponsiveness 6-40
6.2.2.1. Controlled Human Exposures 6-40
6.2.2.2. Toxicology 6-41
6.2.3. Pulmonary Inflammation, Injury and Oxidative Stress 6-43
6.2.3.1. Controlled Human Exposures 6-43
6.2.3.2. Epidemiology 6-46
6.2.3.3. Toxicology 6-56
6.2.4. Respiratory Symptoms and Medication Use 6-59
6.2.4.1. Epidemiology 6-59
6.2.4.2. Summary of Epidemiologic Studies of Respiratory Symptoms and
Asthma Medication Use 6-69
6.2.5. Lung Host Defenses 6-70
6.2.5.1. Mucociliary Clearance 6-71
6.2.5.2. Alveolobronchiolar Transport Mechanism 6-71
6.2.5.3. Alveolar Macrophages 6-72
6.2.5.4. Infection and Adaptive Immunity 6-73
March 2011 vii DRAFT - DO NOT CITE OR QUOTE
-------
6.2.6. Allergic and Asthma-Related Responses
6.2.7. Hospital Admissions, Emergency Department Visits, and Physicians Visits
6.2.7.1. Summary of Findings from 2006 Ozone AQCD
6.2.7.2. Hospital Admission Studies
6.2.7.3. Emergency Department Visit Studies
6.2.7.4. Outpatient and Physician Visit Studies
6.2.7.5. Summary
6.2.8. Respiratory Mortality
6.2.9. Summary and Causal Determination
6.3. Cardiovascular Effects
6.3.1. Controlled Human Exposure
6.3.2. Epidemiology
6.3.2.1. Arrhythmia
6.3.2.2. Heart Rate/Heart Rate Variability
6.3.2.3. Stroke
6.3.2.4. Biomarkers
6.3.2.5. Mvocardial Infarction (Ml)
6.3.2.6. Blood Pressure
6.3.2.7. Hospital Admissions and Emergency Department Visits
6.3.2.8. Cardiovascular Mortality
6.3.2.9. Summary of Epidemiologic Studies
6.3.3. Toxicology
6.3.3.1. Summary of Findings from Previous Ozone AQCDs
6.3.3.2. Recent Ozone-induced Cardiovascular Effects
6.3.4. Summary and Causal Determination
6.4. Central Nervous System Effects
6.4.1. Neuroendocrine Effects
6.4.2. Summary and Causal Determination
6.5. Effects on Other Organ Systems
6.5.1. Effects on the Liver and Xenobiotic Metabolism
6.5.2. Effects on Cutaneous and Ocular Tissues
6.6. Mortality
6.6. 1. Summary of Findings from 2006 Ozone AQCD
6.6.2. Associations of Mortality and Short-Term Ozone Exposure
6.6.2.1. Confounding
6.6.2.2. Effect Modification
6.6.2.3. Interaction
6.6.2.4. Evaluation of the Ozone-Mortality C-R Relationship and Related
6.6.2.5. Associations of Cause-Specific Mortality and Short-term Ozone
Exposure
6.6.3. Summary and Causal Determination
6.7. Overall Summary
References
6-74
6-75
6-75
6-78
6-87
6-92
6-92
6-96
6-97
6-101
6-101
6-102
6-102
6-104
6-107
6-109
6-113
6-113
6-114
6-127
6-128
6-128
6-128
6-129
6-133
6-134
6-137
6-137
6-138
6-138
6-139
6-139
6-139
6-140
6-144
6-152
6-160
lssues_6-161
6-170
6-174
6-176
6-177
CHAPTER 7. INTEGRATED HEALTH EFFECTS OF LONG-TERM OZONE EXPOSURE 7-1
7.1. Introduction 7-1
7.2. Respiratory Effects 7-1
7.2.1. New Onset Asthma 7-2
7.2.2. Asthma Hospital Admissions and ED Visits 7-9
7.2.3. Pulmonary Structure and Function 7-11
7.2.3.1. Evidence from Toxicological Studies 7-14
7.2.4. Pulmonary Inflammation, Injury, and Oxidative Stress 7-17
7.2.5. Allergic Responses 7-19
7.2.6. Host Defense 7-20
March 2011 viii DRAFT - DO NOT CITE OR QUOTE
-------
7.2.7. Respiratory Mortality 7-20
7.2.8. Summary and Causal Determination 7-20
7.3. Cardiovascular Effects 7-23
7.3.1. Cardiovascular Disease 7-23
7.3.1.1. Cardiovascular Epidemiology 7-23
7.3.1.2. Long-Term Cardiovascular Toxicology 7-24
7.3.2. Cardiac Mortality 7-26
7.3.3. Summary and Causal Determination 7-26
7.4. Reproductive and Developmental Effects 7-26
7.4.1. Effects on Sperm 7-27
7.4.2. Effects on Reproduction 7-28
7.4.3. Birth Weight 7-29
7.4.4. Preterm Birth 7-33
7.4.5. Fetal Growth 7-37
7.4.6. Birth Defects 7-40
7.4.7. Developmental Respiratory Effects 7-42
7.4.8. Developmental Central Nervous System Effects 7-45
7.4.8.1. Laterality 7-45
7.4.8.2. Brain Morphology and Neurochemical Changes 7-45
7.4.8.3. Neurobehavioral Outcomes 7-46
7.4.8.4. Sleep Aberrations after Developmental Ozone Exposure 7-46
7.4.9. Early Life Mortality 7-47
7.4.9.1. Stillbirth 7-47
7.4.9.2. Infant Mortality, Less than 1 Year 7-48
7.4.9.3. Neonatal Mortality, Less than 1 Month 7-48
7.4.9.4. Postneonatal Mortality, 1 Month to 1 Year 7-48
7.4.9.5. Sudden Infant Death Syndrome 7-50
7.4.10. Summary and Causal Determination 7-52
7.5. Central Nervous System Effects 7-53
7.5.1. Effects on the Brain and Behavior 7-53
7.5.2. Summary and Causal Determination 7-54
7.6. Carcinogenic and Genotoxic Potential of Ozone 7-54
7.6.1. Introduction 7-54
7.6.2. Lung Cancer Incidence and Mortality 7-56
7.6.3. DNA Damage 7-57
7.6.4. Summary and Causal Determination 7-59
7.7. Mortality 7-59
7.7.1. Summary and Causal Determination 7-62
References 7-63
CHAPTER 8. POPULATIONS SUSCEPTIBLE TO OZONE-RELATED HEALTH EFFECTS 8-1
8.1. Pre-existing Disease/Conditions 8-2
8.1.1. Influenza/Infections 8-3
8.1.2. Asthma/Corticosteroid Use 8-3
8.1.3. Chronic obstructive pulmonary disease (COPD) 8-5
8.1.4. Cardiovascular Disease 8-6
8.1.5. Diabetes 8-7
8.2. Lifestage 8-7
8.2.1. Children 8-7
8.2.2. Older Adults 8-10
8.3. Sex 8-12
8.4. Genetics 8-14
March 2011 ix DRAFT - DO NOT CITE OR QUOTE
-------
8.5. Diet
8.6. Body Mass Index
8.7. Socioeconomic Status
8.8. Air Conditioning Use
8.9. Involvement in Outdoor Activities
8.10. Race/Ethnicity
8.11. Physical Conditioning
8.12. Smoking
8.13. Hyperthyroidism
8.14. Summary
References
CHAPTER 9. ENVIRONMENTAL EFFECTS: OZONE EFFECTS ON VEGETATION AND
ECOSYSTEMS
9.1. Introduction
9.2. Summary and Integration
9.2.1. Introduction
9.2.2. Mechanisms Governing Response
9.2.3. Nature of Effects on Vegetation
9.2.3.1. Effects on Woody and Herbaceous Vegetation
9.2.3.2. Agricultural Crops
9.2.3.3. Factors That Modify Functional and Growth Response
9.2.4. Ecosystems and Services
9.2.4.1. Productivity and Carbon Sequestration
9.2.4.2. Water Cycling
9.2.4.3. Below-Ground Processes
9.2.4.4. Community Composition
9.2.5. Air Quality Indices
9.2.5.1. Modeled Ozone Deposition or "Flux"
9.2.5.2. Night-Time Exposures
9.2.6. Exposure-Response
9.3. Experimental Exposure Methodologies
9.3.1. Introduction
9.3.2. "Indoor," Controlled Environment, and Greenhouse Chambers
9.3.3. Field Chambers
9.3.4. Plume and FACE-Type Systems
9.3.5. Ambient Gradients
9.3.6. Comparative Studies
9.4. Mechanisms Governing Vegetation Response to Ozone
9.4.1. Introduction
9.4.2. Ozone Uptake into the Leaf
9.4.3. Cellular to Systemic Responses
9.4.3.1. Ozone Sensing and Signal Transduction
9.4.3.2. Gene Expression Changes in Response to Ozone
9.4.3.3. Role of Phytohormones in Plant Response to Ozone
9.4.4. Detoxification
9.4.4.1. Overview of Ozone-Induced Defense Mechanisms
9.4.4.2. Role of Antioxidants in Plant Defense Responses
9.4.5. Effects on Primary and Secondary Metabolism
9.4.5.1. Light and Dark Reactions of Photosynthesis
8-16
8-17
8-18
8-20
8-20
8-21
8-22
8-22
8-23
8-23
8-25
9-1
9-1
9-2
9-2
9-5
9-8
9-8
9-11
9-12
9-12
9-12
9-13
9-14
9-14
9-15
9-16
9-17
9-17
9-18
9-18
9-18
9-19
9-21
9-22
9-23
9-24
9-24
9-26
9-29
9-29
9-31
9-34
9-37
9-37
9-38
9-40
9-40
March 2011 x DRAFT - DO NOT CITE OR QUOTE
-------
9.4.5.2. Respiration and Dark Respiration 9-41
9.4.5.3. Secondary Metabolism 9-42
9.4.6. Changes in Stomatal Function 9-45
9.5. Nature of Effects on Vegetation 9-46
9.5.1. Introduction 9-46
9.5.2. Effects on Woody and Herbaceous Vegetation 9-47
9.5.2.1. Growth and Biomass Allocation 9-47
9.5.2.2. Reproduction 9-49
9.5.2.3. Visible Foliar Injury 9-51
9.5.2.4. Leaf Gas Exchange 9-54
9.5.3. Agricultural Crops 9-54
9.5.3.1. Yield 9-55
9.5.3.2. Crop Quality 9-58
9.5.4. Factors that Modify Functional and Growth Response 9-62
9.5.4.1. Genetics 9-62
9.5.4.2. Environmental Biological Factors 9-63
9.5.4.3. Physical Factors 9-63
9.5.4.4. Interactions with other Pollutants 9-64
9.6. Effects of Ozone on Ecosystems and Services 9-68
9.6.1. Ecosystem Scale, Function, and Structure 9-68
9.6.1.1. Ecosystem Services 9-69
9.6.1.2. Assessing Ozone Effects at Larger Spatial Scales 9-69
9.6.2. Productivity and Carbon Sequestration 9-70
9.6.2.1. Stand Scale 9-70
9.6.2.2. Regional and Global Scales 9-71
9.6.3. Water Cycling 9-74
9.6.4. Below-Ground Processes 9-76
9.6.4.1. Litter Carbon Chemistry, Litter Nutrient and Their Ecosystem Budgets 9-77
9.6.4.2. Decomposer Metabolism and Litter Decomposition 9-78
9.6.4.3. Soil respiration and carbon formation 9-79
9.6.4.4. Nutrient cycling 9-82
9.6.4.5. Dissolved Organic Carbon and Biogenic Trace Gases Emission 9-82
9.6.5. Competition and biodiversity 9-83
9.6.5.1. Forest 9-83
9.6.5.2. Grassland and Agricultural Land 9-84
9.6.5.3. Microbes 9-85
9.6.6. Insects and Wildlife 9-86
9.6.6.1. Insects 9-86
9.6.6.2. Wildlife 9-88
9.6.6.3. Indirect Effects on Wildlife 9-89
9.7. Effects-Based Air Quality Exposure Indices and Dose Modeling 9-92
9.7.1. Introduction 9-92
9.7.2. Description of Exposure Indices Available in the Literature 9-93
9.7.3. Important Components of Exposure Indices 9-98
9.7.3.1. Role of Concentration 9-98
9.7.3.2. Diurnal and Seasonal Exposure 9-102
9.7.4. Ozone Uptake/Dose Modeling for Vegetation 9-108
9.7.4.1. Canopy Structure 9-110
9.7.4.2. Site and Climate Factors 9-110
9.7.4.3. Plant Defense Mechanism - Detoxification 9-110
9.8. Ozone Exposure-Plant Response Relationships 9-111
9.8.1. Introduction 9-111
9.8.2. 1996 and 2006 Ozone AQCDs Estimates Of Crop Yield Loss And Tree Seedling
Biomass Loss 9-114
9.8.3. Validation of 1996 and 2006 Ozone AQCD Models and Methodology Using the
90 day 12-h W126 and Current FACE Data. 9-120
March 2011 xi DRAFT - DO NOT CITE OR QUOTE
-------
9.8.3.1.
9.8.3.2.
9.8.3.3.
9.8.3.4.
9.8.3.5.
Comparison of NCLAN-Based Prediction and SoyFACE Data.
Comparison of NHEERL/WED-Based Prediction of Tree Biomass
Response and Aspen FACE Data
Exposure-Response in a Gradient Study
Meta-analyses of growth and yield studies
Additional exposure-response data
9-122
9-125
9-128
9-130
9-131
9-137
References
CHAPTER 10. THE ROLE OF TROPOSPHERIC OZONE IN CLIMATE CHANGE AND
UV-B EFFECTS 10-1
10.1. Introduction 10-1
10.2. Effects of Tropospheric Ozone on Climate 10-1
10.2.1. Background 10-1
10.2.2. Physics and Chemistry of Climate Change and Radiative Forcing 10-2
10.2.2.1. Physics of Greenhouse Gases 10-2
10.2.2.2. Climate Change in the Recent Past 10-2
10.2.2.3. Projections of Future Climate Change 10-3
10.2.2.4. Metrics of Potential Climate Change 10-4
10.2.2.5. Tropospheric Ozone as a Greenhouse Gas 10-4
10.2.3. Factors that Influence the Effect of Tropospheric Ozone on Climate 10-6
10.2.3.1. Trends in the Burden of Tropospheric Ozone 10-7
10.2.3.2. The Effect of Surface Albedo on Ozone Forcing 10-9
10.2.3.3. The Effect of Vertical Distribution on Ozone Forcing 10-9
10.2.3.4. Feedback Factors that Alter the Climate Response to Changes in
Ozone Forcing 10-9
10.2.3.5. Indirect Effects of Tropospheric Ozone on the Carbon Cycle 10-11
10.2.4. Competing Effects of Ozone Precursors on Climate 10-11
10.2.5. Calculating Radiative Forcing and Climate Response to Past Trends in
Tropospheric Ozone 10-11
10.2.6. Calculating the Radiative Forcing and Climate Response to Future Trends in
Tropospheric Ozone 10-13
10.2.6.1. Emissions of Anthropogenic Ozone Precursors across the 21st Century 10-14
10.2.6.2. Impact of 21st Century Trends in Emissions on Tropospheric Ozone 10-15
10.2.6.3. Impact of 21st Century Climate on Tropospheric Ozone 10-16
10.2.6.4. Radiative Forcing and Climate Response from 21st Century Trends in
Tropospheric Ozone 10-16
10.2.7. Summary of the Effects of Tropospheric Ozone on Climate 10-18
10.3. UV-B Related Effects and Tropospheric Ozone 10-18
10.3.1. Background 10-18
10.3.2. Physics of UV Radiation and Flux 10-19
10.3.3. Human Exposure and Susceptibility to Ultraviolet Radiation 10-22
10.3.4. Human Health Effects due to UV-B Radiation 10-24
10.3.5. Ecosystem and Materials Damage Effects Due to UV-B Radiation 10-25
10.3.6. UV-B Related Effects Associated with Changes in Tropospheric Ozone
Concentrations 10-26
References 10-29
March 2011 xii DRAFT - DO NOT CITE OR QUOTE
-------
List of Tables
Table 1-1. Summary of primary and secondary NAAQS promulgated for ozone during the period 1971-2008 1-4
Table 1-2. Aspects to aid in judging causality 1-19
Table 1-3. Weight of evidence for causal determination 1-21
Table 2-1. Summary of causal determinations for short-term exposure to ozone 2-21
Table 2-2. Summary of causal determinations for long-term exposure to ozone 2-28
Table 2-3. Summary of evidence from epidemiologic, controlled human exposure, and animal toxicological studies on the health
effects associated with short- and long-term exposure to ozone 2-36
Table 2-4. Summary of ozone causal determinations for vegetation and ecosystem effects 2-49
Table 2-5. Summary of ozone causal determinations by exposure duration and health outcome 2-64
Table 2-6. Summary of ozone causal determination for welfare effects 2-65
Table 2-7. Summary of ozone causal determination for climate change and UV-B effects 2-65
Table 3-1. Seasonal means of the daily max 8-h avg ozone concentrations in spring and summer at selected CASTNET and other
National Park Service monitoring sites in the continental U.S. and in the U.S. Virgin Islands, in 2001. GEOS-Chem
and PRB are included for comparison 3-38
Table 3-2. Seasonal maximums of the daily max 8-h avg ozone concentrations in spring and summer at selected CASTNET and
other National Park Service monitoring sites in the continental U.S. and in the U.S. Virgin Islands, in 2001. GEOS-
Chem and PRB are included for comparison 3-39
Table 3-3. Summary of monitors meeting 40 CFR Part 58, Appendix A Precision and Bias Goals 3-43
Table 3-4. Performance specifications for ozone based in 40 CFR Part 53 3-46
Table 3-5. Summary of ozone data sets originating from AQS 3-59
Table 3-6. Nationwide distributions of ozone concentrations (ppb) from the year-round data set 3-61
Table 3-7. Nationwide distributions of ozone concentrations (ppb) from the warm-season data set 3-62
Table 3-8. Seasonally stratified distributions of 8-h daily max ozone concentrations (ppb) from the year-round data set (2007-2009)3-64
Table 3-9. Focus cities used in this and previous assessments 3-68
Table 3-10. City-specific distributions of 8-h daily max ozone concentrations (ppb) from the warm-season data set (2007-2009) 3-69
Table 3-11. Rural focus areas 3-87
Table 4-1. Fraction of the 2009 population living within a specified distance of an ozone monitor in selected U.S. cities 4-11
Table 5-1. Human respiratory tract uptake efficiency data 5-8
Table 6-1. Mean and upper percentile concentrations of ozone in epidemiologic studies examining lung function in populations with
increased outdoor exposures 6-18
Table 6-2. Additional characteristics and quantitative data for studies represented in Figure 6-3 6-21
Table 6-3. Additional characteristics and quantitative data for studies represented in Figure 6-4 6-23
Table 6-4. Additional characteristics and quantitative data for studies represented in Figure 6-5 6-24
Table 6-5. Mean and Upper Percentile Concentrations of Ozone in Epidemiologic Studies Examining Lung Function in Asthmatic
Children 6-26
March 2011 xiii DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-6. Additional characteristics and quantitative data for studies represented in Figure 6-6 6-27
Table 6-7. Additional characteristics and quantitative data for studies represented in Figure 6-7 6-29
Table 6-8. Mean and upper percentile concentrations of ozone in epidemiologic studies examining lung function in asthmatic adults6-32
Table 6-9. Mean and upper percentile concentrations of ozone in epidemiologic studies examining lung function in populations not
restricted to asthmatic subjects 6-33
Table 6-10. Additional characteristics and quantitative data for studies represented in Figure 6-8 6-35
Table 6-11. Additional characteristics and quantitative data for studies presented in Figure 6-9 6-39
Table 6-12. Mean and upper percentile ozone concentrations in studies examining biological markers of airway inflammation and
oxidative stress 6-48
Table 6-13. Additional characteristics and quantitative data for studies presented in Figure 6-10 6-49
Table 6-14. Additional characteristics and quantitative data for studies presented in Figure 6-11 6-51
Table 6-15. Mean and upper percentile ozone concentrations in studies examining respiratory symptoms, asthma medication use,
and activity levels 6-60
Table 6-16. Additional characteristics and quantitative data for studies presented in Figure 6-12 6-63
Table 6-17. Additional characteristics and quantitative data for studies presented in Figure 6-13. 6-67
Table 6-18. Additional characteristics and quantitative data for studies presented in Figure 6-14. 6-69
Table 6-19. Mean and upper percentile concentrations of respiratory-related hospital admission and emergency department visit
studies evaluated 6-77
Table 6-20. Corresponding effect estimates for Figure 6-15 6-82
Table 6-21. Corresponding Effect Estimates for Figure 6-19 6-94
Table 6-22. Corresponding effect estimates for Figure 6-20 6-96
Table 6-23. Characterization of ozone concentrations (in ppb) from studies of arrhythmias 6-102
Table 6-24. Characterization of ozone concentrations (in ppb) from studies of heart rate variability 6-105
Table 6-25. Characterization of ozone concentrations (in ppb) from studies of biomarkers 6-109
Table 6-26. Characterization of ozone concentrations (in ppb) from studies of blood pressure 6-113
Table 6-27. Characterization of ozone concentrations (in ppb) from studies of HAs and ED visits 6-115
Table 6-28. Odds ratio (95% Cl) per increment ppb increase in ozone for overall cardiovascular ED visits or HAs in studies
presented in Figure 6-22. 6-120
Table 6-29. Odds Ratio (95% Cl) per increment ppb increase in ozone for congestive heart failure ED visits or HAs for studies
presented in Figure 6-23 6-122
Table 6-30. Odds Ratio (95% Cl) per increment ppb increase in ozone for myocardial infarction, angina, ischemic heart disease, and
coronary heart disease ED visits or HAs for studies presented in Figure 6-24 6-124
Table 6-31. Odds Ratio (95% Cl) per increment ppb increase in ozone for stroke ED visits or HAs for studies presented in Figure 6-
25 6-126
Table 6-32. Odds Ratio (95% Cl) per increment ppb* increase in ozone for arrhythmia and dysrhythmia ED visits or HAs for studies
presented in Figure 6-26 6-127
Table 6-33. Corresponding effect estimates for Figure 6-27 6-142
Table 6-34. Range of mean and upper percentile ozone concentrations in previous and recent multicity studies 6-143
Table 6-35. Correlations between PM and ozone by season and region 6-145
March 2011 xiv DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-36. Corresponding Effect Estimates for Figure 6-30 6-150
Table 6-37. Sensitivity of ozone risk estimates per 10 |jg/m3 increase in 24-h avg ozone concentrations at lag 0-1 to alternative
methods for adjustment of seasonal trend, for all-cause mortality using Berkey MLE and TLNSE Hierarchical Models6-152
Table 6-38. Additional percent change in ozone-related mortality for individual-level susceptibility factors 6-154
Table 6-39. Percent change in all-cause mortality, for all ages, associated with a 40-ppb increase in 1-h max ozone concentrations
at Lag 0-1 at the 25th and 75th percentile of the center-specific distribution of selected effect modifiers 6-157
Table 6-40. Percentage increase in daily mortality for a 10-ppb increase in 24-h avg ozone concentrations during the previous week
by geographic region in the U.S., 1987-2000 6-158
Table 6-41. Estimated effect of a 10-ppb increase in 8-h max ozone concentrations on mortality during the summer months for
single-day and distributed lag models 6-163
Table 6-42. Estimated percent increase in cause-specific mortality (and 95% CIs) for a 10-|jg/m3 increase in maximum 8-h ozone
during June-August, for the same day (lag 0), the average of the same and previous day (lag 0-1), the unconstrained
distributed lag model for the sum of 0-20 days and the penalized distributed lag model (lag 0-20) 6-165
Table. 6-43. Percent excess all-cause mortality per 10-ppb increase in daily 8-h max ozone on the same day, by season, month,
and age groups 6-168
Table 6-44. Corresponding effect estimates for Figure 6-37 6-173
Table 6-45. Summary of causal determinations for short-term exposures to ozone 6-176
Table 7-1. Summary of selected key new studies examining annual ozone exposure and respiratory health effects 7-21
Table 7-2. Brief summary of epidemiologic studies of birth weight 7-32
Table 7-3. Brief summary of epidemiologic studies of PTB 7-36
Table 7-4. Brief summary of epidemiologic studies of fetal growth 7-40
Table 7-5. Brief summary of epidemiologic studies of birth defects 7-42
Table 7-6. Brief summary of infant mortality studies 7-51
Table 7-7. Relative risk (and 95% Cl) of death attributable to a 10-ppb change in the ambient ozone concentration 7-62
Table 8-1. Prevalence of respiratory diseases, cardiovascular diseases, and diabetes by age and region among individuals 18
years and older in the U.S. 8-3
Table 9-1 Summary of ozone causal determinations for vegetation and ecosystem effects 9-5
Table 9-2. Ozone effects on plant reproductive processes (derived from Table AX9-22 of the 2006 ozone AQCD) 9-51
Table 9-3. Summary of recent studies of ozone effects on crops (exclusive of growth and yield) 9-60
Table 9-4. Response of plants to the interactive effects of elevated ozone exposure and N enrichment 9-66
Table 9-5. The effects of ozone on primary production, C exchange, C sequestration and yield loss 9-74
Table 9-6. The effect of elevated ozone on leaf/litter nutrient concentrations 9-78
Table 9-7. The temporal variation of ecosystem responses to ozone exposure at AspenFACE site 9-80
Table 9-8. Ozone exposures at which 10 and 20% yield loss is predicted for 50 and 75% of crop species, based on composite
functions for the 50th and 75th percentiles of 34 Weibull curves for relative yield loss data from 34 non-draughted
NCLAN studies of 12 crop species; curves were standardized to 90-day W126 9-119
Table 9-9. Ozone exposures at which 10 and 20% yield loss is predicted for 50 and 75% of crop species under drought conditions
and adequate moisture, based on composite functions for the 50th and 75th percentiles of 16 Weibull curves for
relative yield loss data from 8 NCLAN studies that paired draughted and watered conditions for the same genotype;
curves were standardized to 90-day W126 9-119
March 2011 xv DRAFT - DO NOT CITE OR QUOTE
-------
Table 9-10. Ozone exposures at which 10 and 20% biomass loss is predicted for 50 and 75 % of tree species, based on composite
functions for the 50th and 75th percentiles of 49 Weibull curves for relative above-ground biomass loss data from 49
studies of 11 tree species grown under well-watered conditions for 1 or 2 year; curves were standardized to 90-day
W1269-120
Table 9-11. Comparison between relative yield observed in the SoyFACE experiment, and relative yield predicted at the same
values of ozone by the median composite function for NCLAN (two-parameter relative yield model) 9-123
Table 9-12. Comparison between yield observed in the SoyFACE experiment and yield predicted at the same values of ozone by
the median composite function for NCLAN (three-parameter absolute yield model), using two scaling methods to
calculate the intercept 9-123
Table 9-13. Comparison between above-ground biomass observed under elevated ozone in Aspen FACE experiment in 6 year,
relative to above-ground biomass observed under ambient ozone and relative above-ground biomass above-ground
biomass at the same values of ozone predicted by the median composite function for NHEERL/WED (two-parameter
relative biomass model) 9-126
Table 9-14. Comparison between above-ground biomass observed in Aspen FACE experiment in 6 year and biomass predicted by
the median composite function based on NHEERL/WED (three-parameter absolute biomass model), using 2 scaling
methods to calculate the intercept 9-127
Table 9-15. Meta-analyses of growth or yield studies published since 2005 9-130
Table 9-16. Summary of studies of effects of ozone exposure on growth and yield of agricultural crops 9-132
Table 9-17. Summary of studies of effects of ozone exposure on growth of natural vegetation 9-135
Table 10-1.2000-2030 changes in anthropogenic emissions, and CH4 and tropospheric ozone burdens, and the associated
tropospheric ozone forcing for three scenarios; values are ensemble means 10-15
March 2011 xvi DRAFT - DO NOT CITE OR QUOTE
-------
List of Figures
Figure 1 -1. Identification of studies for inclusion in the ISA. 1 -9
Figure 2-1. Snapshot of evidence for the spectrum of respiratory health effects associated with short- and long-term exposure to
ozone. 2-39
Figure 2-2. Schematic depicting key events in the ozone toxicity pathway. 2-40
Figure 2-3. The effects of ozone at leaf, plant and ecosystem scales. 2-48
Figure 3-1. Schematic overview of photochemical processes influencing stratospheric and tropospheric ozone. 3-2
Figure 3-2. Estimated anthropogenic emissions of ozone precursors for 2005 including NOX (top), VOCs (middle), and CO (bottom)
in the U.S. in million metric tons (MT) per year. 3-5
Figure 3-3. Measured concentrations of ozone and NOZ (NOy-NOx) during the afternoon at rural sites in the eastern U.S. (grey
circles) and in urban areas and urban plumes associated with Nashville, TN (gray dashes); Paris, France (black
diamonds); and Los Angeles, CA (Xs). 3-16
Figure 3-4. Sample CMAQ modeling domains: 36 km-grid-spacing; outer parent domain in black; 12 km western U.S. (WUS)
domain in red; 12 km eastern U.S. (EUS) domain in blue. 3-18
Figure 3-5. Main components of a comprehensive atmospheric chemistry modeling system, such as the U.S. EPA's Community
Model for Air Quality (CMAQ) System. 3-19
Figure 3-6. Comparison of global CTM predictions of maximum daily 8-h avg ozone concentrations and multi-model mean with
monthly averaged CASTNET observations in the Mountain West and Southeast regions of the U.S. 3-24
Figure 3-7. Time series of daily maximum 8-h avg ozone concentrations (ppm) measured at Trinidad Head, CA, from April 18,2002
through December 31,2009. 3-28
Figure 3-8. Time series of measured ozone and model calculated ozone at Trinidad Head, CA, during April and May, 2002. _ 3-29
Figure 3-9. Mean daily 8-h max ozone concentrations in surface air for the base case (top) and PRB case (bottom) in April-May,
2001.3-34
Figure 3-10. Mean daily 8-h max ozone concentrations in surface air for base case (top) and PRB case (bottom) in June-August,
2001.3-35
Figure 3-11. Distribution of ozone measured at CASTNET sites (blue diamonds), ozone calculated by the GEOS-Chem base model
(red squares) and PRB ozone (green triangles) at (a) sites <1500 m elevation and (b) at sites >1500 m elevation for
April - May; and (c) at sites <1500 m elevation and (d) at sites >1500 m elevation for June-July-August. 3-36
Figure 3-12. Box plots of precision data by year (2005-2009) for all ozone monitors reporting single-point QC check data to AQS.3-43
Figure 3-13. Box plots of percent-difference data by year (2005-2009) for all ozone monitors reporting single-point QC check data to
AQS. 3-44
Figure 3-14. Box plots of RPD data by year for the two co-located sites in Missouri from 2006-2009. 3-45
Figure 3-15. Box plots of RPD data by year for all U.S. ozone sites reporting single-point QC check data to AQS from 2005-2009.3-45
Figure3-16. U.S. ozone sites reporting data to AQS as of 2009. 3-53
Figure 3-17. U.S. Rural NCore, CASTNET and NPS POMS current and proposed sites as of October, 2010. 3-54
Figure 3-18. Distribution in nation-wide year-round site-level correlations between daily ozone metrics including 24-h avg, 1-h daily
max and 8-h daily max using AQS data, 2007-2009. 3-57
Figure 3-19. Required ozone monitoring time periods (ozone season) identified by monitoring site. 3-58
March 2011 xvii DRAFT - DO NOT CITE OR QUOTE
-------
Figure 3-20. Location of the 458 ozone monitors meeting the year-round completeness criterion for all 3 years between 2007 and
2009.3-59
Figure 3-21. Location of the 1,064 ozone monitors meeting the warm-season completeness criteria for all 3 years between 2007
and 2009. 3-60
Figure 3-22. Highest monitor (by county) 3-year avg (2007-2009) of the 8-h daily max ozone concentration based on the year-round
data set (top map) with seasonal stratification (bottom 4 maps). 3-65
Figure 3-23. Highest monitor (by county) 3-year avg (2007-2009) of the 8-h daily max ozone concentration based on the warm-
season data set (top map) with annual stratification (bottom 3 maps). 3-66
Figure 3-24. Map of the Atlanta CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-70
Figure 3-25. Map of the Boston CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-71
Figure 3-26. Map of the Los Angeles CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-72
Figure 3-27. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data set
inclusion criteria within the Atlanta CSA. 3-74
Figure 3-28. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data set
inclusion criteria within the Boston CSA. 3-74
Figure 3-29. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data set
inclusion criteria within the Los Angeles CSA. 3-75
Figure 3-30. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Atlanta CSA. 3-77
Figure 3-31. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Boston CSA. 3-78
Figure 3-32. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Los Angeles CSA. 3-79
Figure 3-33. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Atlanta CSA. 3-80
Figure 3-34. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Boston CSA. 3-81
Figure 3-35. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Los Angeles CSA. 3-82
Figure 3-36. Terrain map showing the location of two nearby AQS ozone monitoring sites (red dots) along the western edge of the
Los Angeles CSA. Site AL is near shore, 3 m above sea level. 3-84
Figure 3-37. Rural focus area site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the
warm-season data set inclusion criteria 3-87
Figure 3-38. Terrain map showing the location of five AQS ozone monitoring sites (green/black stars) in Great Smoky Mountain
National Park, NC-TN (SMNP). 3-89
Figure 3-39. Pair-wise monitor correlations (left) and coefficients of divergence 3-90
Figure 3-40. Terrain map showing the location of the AQS ozone monitoring site in Rocky Mountain National Park, CO (black/green
star) and the Denver CSA (red dots) 3-91
Figure 3-41. Terrain map showing the location of two AQS ozone monitoring sites (black/green stars) in Sequoia National Park, CA.3-93
Figure 3-42. National 8-h ozone trends, 2001-2008 (average of the annual fourth highest 8-h daily max concentrations in ppm). 3-94
March 2011 xviii DRAFT - DO NOT CITE OR QUOTE
-------
Figure 3-43. Diel patterns in 1-h avg ozone for Atlanta, Boston and Los Angeles between 2007 and 2009 using the year-round data
set for the cold month/warm month comparison (left half) and the warm-season data set for the weekday/weekend
comparison (right half). 3-96
Figure 3-44. Distribution of Pearson correlation coefficients for comparison of 8-h daily max ozone from the year-round data set with
co-located 24-h avg CO, S02, N02, PM10 and PM2.5 from AQS, 2007-2009 3-98
Figure 3-45. Distribution of Pearson correlation coefficients for comparison of 8-h daily max ozone from the warm-season (May-
Sept) data set with co-located 24-h avg CO, S02, N02, PM10 and PM2.5 from AQS, 2007-2009. 3-99
Figure 3A-1. Time series of observed and GEOS-Chem base model and PRB model 8-h daily max ozone concentrations (ppb) with
PRB estimates for Acadia NP, ME. 3-112
Figure 3A-2. Time series of observed and GEOS-Chem base model and PRB model 8-h daily max ozone concentrations (ppb) with
PRB estimates for Everglades NP, FL. 3-112
Figure 3A-3. Time series of observed and GEOS-Chem base model and PRB model 8-h daily max ozone concentrations (ppb) with
PRB estimates for Crockett, KY. 3-113
Figure 3A-4. Time series of observed and GEOS-Chem base model and PRB model 8-h daily max ozone concentrations (ppb) with
PRB estimates for Coffeeville, MS. 3-113
Figure 3A-5. Time series of observed and GEOS-Chem base model and PRB model 8-h daily max ozone concentrations (ppb) with
PRB estimates for VoyageursNP, MN. 3-114
Figure 3A-6. Time series of observed and GEOS-Chem base model and PRB model 8-h daily max ozone concentrations (ppb) with
PRB estimates for Big Bend NP, TX. 3-114
Figure 3A-7. Time series of observed and GEOS-Chem base model and PRB model 8-h daily max ozone concentrations (ppb) with
PRB estimates for Rocky Mountain NP, CO. 3-115
Figure 3A-8. Time series of observed and GEOS-Chem base model and PRB model 8-h daily max ozone concentrations (ppb) with
PRB estimates for Mesa Verde NP, CO. 3-115
Figure 3A-9. Time series of observed and GEOS-Chem base model and PRB model 8-h daily max ozone concentrations (ppb) with
PRB estimates for Yellowstone NP, WY. 3-116
Figure 3A-10. Time series of observed and GEOS-Chem base model and PRB model 8-h daily max ozone concentrations (ppb)
with PRB estimates for Centennial, WY. 3-116
Figure 3A-11. Time series of observed and GEOS-Chem base model and PRB model 8-h daily max ozone concentrations (ppb)
with PRB estimates for Canyonlands NP, UT. 3-117
Figure 3A-12. Time series of observed and GEOS-Chem base model and PRB model 8-h daily max ozone concentrations (ppb)
with PRB estimates for Glacier NP, MT. 3-117
Figure 3A-13. Time series of observed and GEOS-Chem base model and PRB model 8-h daily max ozone concentrations (ppb)
with PRB estimates for Mt. Rainier NP, WA. 3-118
Figure 3A-14. Time series of observed and GEOS-Chem base model and PRB model 8-h daily max ozone concentrations (ppb)
with PRB estimates for Lassen Volcanic NP, CA. 3-118
Figure 3A-15. Time series of observed and GEOS-Chem base model and PRB model 8-h daily max ozone concentrations (ppb)
with PRB estimates for Pinnacles NM, CA. 3-119
Figure 3A-16. Map of the Atlanta CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-120
Figure 3A-17. Map of the Baltimore CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-121
Figure 3A-18. Map of the Birmingham CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-122
March 2011 xix DRAFT - DO NOT CITE OR QUOTE
-------
Figure 3A-19. Map of the Boston CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-123
Figure 3A-20. Map of the Chicago CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-124
Figure 3A-21. Map of the Dallas CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-125
Figure 3A-22. Map of the Denver CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-126
Figure 3A-23. Map of the Detroit CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-127
Figure 3A-24. Map of the Houston CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-128
Figure 3A-25. Map of the Los Angeles CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-129
Figure 3A-26. Map of the Minneapolis CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-130
Figure 3A-27. Map of the New York CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-131
Figure 3A-28. Map of the Philadelphia CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-132
Figure 3A-29. Map of the Phoenix CBSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-133
Figure 3A-30. Map of the Pittsburgh CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-134
Figure 3A-31. Map of the Salt Lake City CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-135
Figure 3A-32. Map of the San Antonio CBSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-136
Figure 3A-33. Map of the San Francisco CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-137
Figure 3A-34. Map of the Seattle CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-138
Figure 3A-35. Map of the St. Louis CSA including ozone monitor locations, population gravity centers, urban areas, and major
roadways. 3-139
Figure 3A-36. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the Atlanta CSA. 3-140
Figure 3A-37. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the Baltimore CSA. 3-140
Figure 3A-38. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the Birmingham CSA. 3-141
Figure 3A-39. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the Boston CSA. 3-141
March 2011 xx DRAFT - DO NOT CITE OR QUOTE
-------
Figure 3A-40. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the Chicago CSA. 3-142
Figure 3A-41. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the Dallas CSA. 3-142
Figure 3A-42. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the Denver CSA. 3-143
Figure 3A-43. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the Detroit CSA. 3-143
Figure 3A-44. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the Houston CSA. 3-144
Figure 3A-45. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the Los Angeles CSA. 3-145
Figure 3A-46. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the Minneapolis CSA. 3-146
Figure 3A-47. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the New York CSA. 3-146
Figure 3A-48. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the Philadelphia CSA. 3-147
Figure 3A-49. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the Phoenix CBSA. 3-148
Figure 3A-50. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the Pittsburgh CSA. 3-149
Figure 3A-51. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the Salt Lake City CSA. 3-149
Figure 3A-52. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the San Antonio CBSA. 3-150
Figure 3A-53. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the San Francisco CSA. 3-150
Figure 3A-54. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the Seattle CSA. 3-151
Figure 3A-55. Site information, statistics and box plots for 8-h daily max ozone from AQS monitors meeting the warm-season data
set inclusion criteria within the St. Louis CSA. 3-151
Figure 3A-56. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the Atlanta CSA. 3-153
Figure 3A-57. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the Baltimore CSA. 3-154
Figure 3A-58. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the Birmingham CSA. 3-155
Figure 3A-59. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the Boston CSA. 3-156
Figure 3A-60. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the Chicago CSA. 3-157
March 2011 xxi DRAFT - DO NOT CITE OR QUOTE
-------
Figure 3A-61. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the Dallas CSA. 3-158
Figure 3A-62. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the Denver CSA. 3-159
Figure 3A-63. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the Detroit CSA. 3-160
Figure 3A-64. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the Houston CSA. 3-161
Figure 3A-65. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the Los Angeles CSA. 3-162
Figure 3A-66. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the Minneapolis CSA. 3-163
Figure 3A-67. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the New York CSA. 3-164
Figure 3A-68. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the Philadelphia CSA. 3-165
Figure 3A-69. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the Phoenix CBSA. 3-166
Figure 3A-70. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the Pittsburgh CSA. 3-167
Figure 3A-71. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the Salt Lake City CSA. 3-168
Figure 3A-72. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the San Antonio CBSA. 3-169
Figure 3A-73. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the San Francisco CSA. 3-170
Figure 3A-74. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the Seattle CSA. 3-171
Figure 3A-75. Pair-wise monitor correlations expressed as a histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for the St. Louis CSA. 3-172
Figure 3A-76. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Atlanta CSA. 3-173
Figure 3A-77. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Baltimore CSA. 3-174
Figure 3A-78. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Birmingham CSA. 3-175
Figure 3A-79. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Boston CSA. 3-176
Figure 3A-80. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Chicago CSA. 3-177
Figure 3A-81. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Dallas CSA. 3-178
March 2011 xxii DRAFT - DO NOT CITE OR QUOTE
-------
Figure 3A-82. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Denver CSA. 3-179
Figure 3A-83. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Detroit CSA. 3-180
Figure 3A-84. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Houston CSA. The colors in the histogram bins correspond to the levels of the
contour matrix. 3-181
Figure 3A-85. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Los Angeles CSA. 3-182
Figure 3A-86. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Minneapolis CSA. 3-183
Figure 3A-87. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the New York CSA. 3-184
Figure 3A-88. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Philadelphia CSA. 3-185
Figure 3A-89. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Phoenix CBSA. 3-186
Figure 3A-90. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Pittsburgh CSA. 3-187
Figure 3A-91. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Salt Lake City CSA. 3-188
Figure 3A-92. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the San Antonio CBSA. 3-189
Figure 3A-93. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the San Francisco CSA. 3-190
Figure 3A-94. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the Seattle CSA. 3-191
Figure 3A-95. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for the St. Louis CSA. 3-192
Figure 3A-96. Diel patterns in 1 -h avg ozone for the Atlanta CSA between 2007 and 2009 using the year-round data set for the cold
month/warm month comparison (left half) and the warm-season data set for the weekday/weekend comparison (right
half). No year-round monitors were available for the cold month/warm month comparison in this CSA. 3-193
Figure 3A-97. Diel patterns in 1 -h avg ozone for the Baltimore CSA between 2007 and 2009 using the year-round data set for the
cold month/warm month comparison (left half) and the warm-season data set for the weekday/weekend comparison
(right half). 3-194
Figure 3A-98. Diel patterns in 1-h avg ozone for the Birmingham CSA between 2007 and 2009 using the year-round data set for the
cold month/warm month comparison (left half) and the warm-season data set for the weekday/weekend comparison
(right half). 3-194
Figure 3A-99. Diel patterns in 1-h avg ozone for the Boston CSA between 2007 and 2009 using the year-round data set for the cold
month/warm month comparison (left half) and the warm-season data set for the weekday/weekend comparison (right
half). 3-195
Figure 3A-100. Diel patterns in 1 -h avg ozone for the Chicago CSA between 2007 and 2009 using the year-round data set for the
cold month/warm month comparison (left half) and the warm-season data set for the weekday/weekend comparison
(right half). 3-195
March 2011 xxiii DRAFT - DO NOT CITE OR QUOTE
-------
Figure 3A-101. Diel patterns in 1 -h avg ozone for the Dallas CSA between 2007 and 2009 using the year-round data set for the cold
month/warm month comparison (left half) and the warm-season data set for the weekday/weekend comparison (right
half). 3-196
Figure 3A-102. Diel patterns in 1 -h avg ozone for the Denver CSA between 2007 and 2009 using the year-round data set for the
cold month/warm month comparison (left half) and the warm-season data set for the weekday/weekend comparison
(right half). 3-196
Figure 3A-103. Diel patterns in 1 -h avg ozone for the Detroit CSA between 2007 and 2009 using the year-round data set for the
cold month/warm month comparison (left half) and the warm-season data set for the weekday/weekend comparison
(right half). 3-197
Figure 3A-104. Diel patterns in 1-h avg ozone for the Houston CSA between 2007 and 2009 using the year-round data set for the
cold month/warm month comparison (left half) and the warm-season data set for the weekday/weekend comparison
(right half). 3-197
Figure 3A-105. Diel patterns in 1 -h avg ozone for the Los Angeles CSA between 2007 and 2009 using the year-round data set for
the cold month/warm month comparison (left half) and the warm-season data set for the weekday/weekend
comparison (right half). 3-198
Figure 3A-106. Diel patterns in 1-h avg ozone for the Minneapolis CSA between 2007 and 2009 using the year-round data set for
the cold month/warm month comparison (left half) and the warm-season data set for the weekday/weekend
comparison (right half). 3-198
Figure 3A-107. Diel patterns in 1 -h avg ozone for the New York CSA between 2007 and 2009 using the year-round data set for the
cold month/warm month comparison (left half) and the warm-season data set for the weekday/weekend comparison
(right half). 3-199
Figure 3A-108. Diel patterns in 1-h avg ozone for the Philadelphia CSA between 2007 and 2009 using the year-round data set for
the cold month/warm month comparison (left half) and the warm-season data set for the weekday/weekend
comparison (right half). 3-199
Figure 3A-109. Diel patterns in 1 -h avg ozone for the Phoenix CBSA between 2007 and 2009 using the year-round data set for the
cold month/warm month comparison (left half) and the warm-season data set for the weekday/weekend comparison
(right half). 3-200
Figure 3A-110. Diel patterns in 1 -h avg ozone for the Pittsburgh CSA between 2007 and 2009 using the year-round data set for the
cold month/warm month comparison (left half) and the warm-season data set for the weekday/weekend comparison
(right half). 3-200
Figure 3A-111. Diel patterns in 1 -h avg ozone for the Salt Lake City CSA between 2007 and 2009 using the year-round data set for
the cold month/warm month comparison (left half) and the warm-season data set for the weekday/weekend
comparison (right half). 3-201
Figure 3A-112. Diel patterns in 1 -h avg ozone for the San Antonio CBSA between 2007 and 2009 using the year-round data set for
the cold month/warm month comparison (left half) and the warm-season data set for the weekday/weekend
comparison (right half). 3-201
Figure 3A-113. Diel patterns in 1 -h avg ozone for the San Francisco CSA between 2007 and 2009 using the year-round data set for
the cold month/warm month comparison (left half) and the warm-season data set for the weekday/weekend
comparison (right half). 3-202
Figure 3A-114. Diel patterns in 1 -h avg ozone for the Seattle CSA between 2007 and 2009 using the year-round data set for the
cold month/warm month comparison (left half) and the warm-season data set for the weekday/weekend comparison
(right half). 3-202
Figure 3A-115. Diel patterns in 1 -h avg ozone for the St. Louis CSA between 2007 and 2009 using the year-round data set for the
cold month/warm month comparison (left half) and the warm-season data set for the weekday/weekend comparison
(right half). 3-203
Figure 4-1. Correlations between 1 -week ozone concentrations and co-pollutants. 4-9
March 2011 xxiv DRAFT - DO NOT CITE OR QUOTE
-------
Figure 4-2. Distribution of time that NHAPS respondents spent in ten microenvironments based on smoothed 1-min diary data.4-16
Figure 5-1. Structure of lower airways with progression from the large airways to the alveolus. 5-3
Figure 5-2. Schematic overview of ozone interaction with PUFA in ELF and lung cells. It should be noted that not all secondary
reaction products are shown. 5-4
Figure 5-3. Total ozone uptake efficiency as a function of breathing frequency at a minute ventilation of 30 L/min. Subjects breathed
0.25 ppm ozone oronasally through a breathing mask. 5-10
Figure 5-4. Ozone uptake fraction as a function of volumetric penetration (VP) in a representative subject. 5-11
Figure 5-5. Species comparison of antioxidant / protein ratios of: (a) nasal lavage fluid and, (b) bronchoalveolar lavage fluid. _5-18
Figure 5-6. Schematic depicting key events in ozone's toxicity pathway. Solid arrows denote pathways for which there is greater
certainty. Broken arrows represent pathways of emerging interest. 5-60
Figure 6-1. Cross-study comparison of mean ozone-induced FEVi decrements following 6.6 hours of constant, square-wave
exposure to ozone. 6-5
Figure 6-2. Frequency distributions of FE\A decrements observed by Schelegle et al. (2009) in young healthy adults (16 F, 15 M)
following 6.6-h exposures to ozone or filtered air. 6-10
Figure 6-3. Changes in FE\A (ml) or PEF (mL/sec) in association with ambient ozone exposure in studies of children attending
summer camp. 6-20
Figure 6-4. Changes in FE\A (ml or percent change) or PEF (mL/sec) in association with ambient ozone exposures of adults and
children during outdoor exercise. 6-22
Figure 6-5. Changes in lung function parameters in association with ambient ozone exposures among outdoor workers. 6-24
Figure 6-6. Changes in lung function parameters (percent-predicted or %change) in association with ambient ozone exposures
among asthmatic children. 6-27
Figure 6-7. Changes in lung function parameters (L/min or mL/10) in association with ambient ozone exposures among asthmatic
children. 6-28
Figure 6-8. Changes in lung function parameters in association with ambient ozone exposures in studies not restricted to asthmatic
populations. 6-34
Figure 6-9. Comparison of ozone-lung function effect estimates in single- and co-pollutant models. 6-38
Figure 6-10. Associations of ambient ozone exposure with changes in concentrations of exhaled nitric oxide (eNO). 6-49
Figure 6-11. Associations of ambient ozone exposure with biological markers of airway oxidative stress and airway inflammation.6-50
Figure 6-12. Associations of ambient ozone exposure with respiratory symptoms in asthmatic subjects. 6-62
Figure 6-13. Associations of ambient ozone exposure with asthma medication use. 6-66
Figure 6-14. Associations of ambient ozone exposure with respiratory symptoms in studies not restricted to asthmatic populations.6-69
Figure 6-15. Percent increase in respiratory hospital admissions from natural spline models for a 40-ppb increase in 1 -h max ozone
concentrations for each location of the APHENA study. 6-81
Figure 6-16. Estimated relative risks (RRs) of ozone-related asthma hospital admissions allowing for possible nonlinear relationships
using natural splines. 6-87
Figure 6-17. Risk ratio for respiratory ED visits and different ozone exposure metrics in Atlanta from 1993-2004. 6-89
Figure 6-18. Loess dose-response estimates and twice-standard error estimates from generalized additive models for associations
between 3-day avg ozone concentrations and ED visits for pediatric asthma. 6-91
Figure 6-19. Percent increase in respiratory-related hospital admission and ED visits in studies that presented all-year and
seasonal analyses. 6-93
March 2011 xxv DRAFT - DO NOT CITE OR QUOTE
-------
Figure 6-20. Percent increase in respiratory-related hospital admissions and ED visits for studies that presented single and
co-pollutant model results. 6-95
Figure 6-21. Odds ratio (95% confidence interval) for stroke by quintiles of ozone 6-108
Figure 6-22. Odds ratio (95% Cl) per increment ppb increase in ozone for over all cardiovascular ED visits or HAs. 6-119
Figure 6-23. Odds Ratio (95% Cl) per increment ppb increase in ozone for congestive heart failure ED visits or HAs. 6-121
Figure 6-24. Odds Ratio (95% confidence interval) per increment ppb increase in ozone for myocardial infarction, angina, ischemic
heart disease, and coronary heart disease ED visits or HAs. 6-123
Figure 6-25. Odds Ratio (95% confidence interval) per increment ppb increase in ozone for stroke ED visits or HAs. 6-125
Figure 6-26. Odds Ratio (95% confidence interval) per increment ppb* increase in ozone for arrhythmia and dysrhythmia ED visits or
HAs. 6-126
Figure 6-27. Summary of mortality risk estimates for short-term ozone exposure and all-cause (nonaccidental) mortality from all-year
and summer season analyses. 6-141
Figure 6-28. Scatter plots of ozone mortality risk estimates with versus without adjustment for PM10 in NMMAPS cities. 6-146
Figure 6-29. Community-specific ozone-mortality risk estimates for nonaccidental mortality per 10-ppb increase in same-day 24-h
avg summertime ozone concentrations in single-pollutant models and co-pollutant models with sulfate. 6-147
Figure 6-30. Percent increase in all-cause (nonaccidental) and cause-specific mortality from the APHENA study for single- and co-
pollutant models. 6-149
Figure 6-31. Ozone mortality risk estimates and community-specific characteristics, U.S., 1987-2000. 6-156
Figure 6-32. Community-specific Bayesian ozone-mortality risk estimates in 98 U.S. communities. 6-159
Figure 6-33. Map of spatially dependent ozone-mortality coefficients for 8-h max ozone concentrations using summer data. 6-159
Figure 6-34. Estimated combined smooth distributed lag for 48 U.S. cities during the summer months. 6-164
Figure 6-35. Estimated combined smooth distributed lag in 21 European cities during the summer (June-August) months. 6-166
Figure 6-36. Estimated combined C-R curve for ozone and nonaccidental mortality using the nonlinear (spline) model. 6-169
Figure 6-37. Percent increase in cause-specific mortality. 6-172
Figure 7-1. Interaction of gene presence and ozone level on the Hazard Ratio (HR) of new-onset asthma in the 12 Children's Health
Study communities. 7-5
Figure 7-2. Ozone modifies the effect of TNF G-308A genotype on bronchitic symptoms among children with asthma in the CHS.7-7
Figure 7-3. Ozone-asthma concentration-response relationship using the mean concentration during the entire follow-up period
adjusted for child's sex, age, birth weight, and gestational age; maternal race, ethnicity, age, education, insurance,
and smoking status during pregnancy; and regional poverty level and temperature. 7-11
Figure 7-4. Birthweight deficit by decile of 24-h avg ozone concentration averaged over the entire pregnancy compared with the
decile group with the lowest ozone exposure. 7-30
Figure 7-5. Adjusted ozone-mortality relative risk estimates (95% Cl) by time period of analysis per subject-weighted mean ozone
concentration in the Cancer Prevention Study II by the American Cancer Society. 7-60
Figure 9-1. The effects of ozone at leaf, plant and ecosystem scales. 9-4
Figure 9-2. The microarchitecture of a dicot leaf. 9-27
Figure 9-3. Possible reactions of ozone within water. 9-27
Figure 9-4. The Crigee mechanism of ozone attack of a double bond. 9-28
Figure 9-5 Composite diagram of major themes in the temporal evolution of the genetic response to ozone stress. 9-34
Figure 9-6. The oxidative cell death cycle. 9-37
March 2011 xxvi DRAFT - DO NOT CITE OR QUOTE
-------
Figure 9-7. Conceptual diagram showing where ozone disrupts C flow in a tree-soil system, including transfer between biotic and
abiotic components below ground that influence soil physical and chemical properties. 9-77
Figure 9-8. Diagrammatic representation of several exposure indices, illustrating how they weight concentration and accumulate
exposure. 9-94
Figure 9-9. Trends in May to September 12-h SUM06, peak 1 -h ozone concentration and number of daily exceedances of 95 ppb for
the Crestline site in 1963 to 1999 in relation to trends in mean daily maximum temperature for Crestline and daily
reactive organic gases (ROG) and oxides of nitrogen (NOX) for San Bernardino County. 9-101
Figure 9-10. The number of hourly average concentrations between 50 and 89 ppb for the period 1980-2000 for the Crestline, San
Bernardino County, CA, monitoring site. 9-102
Figure 9-11. Mean diurnal. 9-105
Figure 9-12. Maximum 3-month, 12-h W126 plotted against maximum 6-month, 12-hW126. 9-108
Figure 9-13. Quantiles of predicted relative yield loss for 34 NCLAN crop experiments. 9-115
Figure 9-14. Quantiles of predicted relative yield loss for 4 crop species in NCLAN experiments. 9-116
Figure 9-15. Quantiles of predicted relative biomass loss for 49 tree species in NHEERL/WED experiments. 9-117
Figure 9-16. Quantiles of predicted relative biomass loss for 4 tree species in NHEERL/WED experiments. 9-118
Figure 9-17. Comparison of yield observed in Soy FACE experiment in a given year with yield predicted by the median composite
function based on NCLAN. 9-124
Figure 9-18. Comparison of composite functions for the quartiles of 7 curves for 7 genotypes of soybean grown in the SoyFACE
experiment, and for the quartiles of 11 curves for 5 genotypes of soybean grown in the NCLAN project. 9-125
Figure 9-19. Comparison between above-ground biomass observed in Aspen FACE experiment in 6 year and biomass predicted by
the median composite function based on NHEERL/WED. 9-127
Figure 9-20. Above-ground biomass for one genotype of cottonwood grown in seven locations for one season in 3 years. 9-129
Figure 10-1. Flow chart for the effects of tropospheric ozone on climate. 10-5
Figure 10-2. Global average radiative forcing (RF) estimates and ranges in 2005 for anthropogenic C02, CH4, ozone and other
important agents and mechanisms, together with the typical geographical extent (spatial scale) of the forcing and the
assessed level of scientific understanding (LOSU). 10-6
Figure 10-3. Ensemble average 1900-2000 surface temperature trends (°C per century) in response to tropospheric ozone changes
and the input radiative forcing (W/m2), as computed by the NASA GISS chemistry-climate model. 10-13
Figure 10-4. Global mean radiative forcing estimates calculated by a set of models for the 2000-2100 change in tropospheric ozone.10-17
Figure 10-5. Diagram ofthe factors that determine human exposure to ultraviolet radiation. 10-19
Figure 10-6. Monthly stratospheric ozone number density (scaled by 1 x 1 o12 molecules/cm3), in 3 latitude belts as a function of time
(August 2002 - December 2008) and altitude (15-40 km) from the Global Ozone Monitoring by Occultation of Stars
(GOMOS) instrument onboard the European Space Agency's ENVISAT satellite. 10-21
March 2011 xxvii DRAFT - DO NOT CITE OR QUOTE
-------
Ozone Project Team
Executive Direction
Dr. John Vandenberg (Director)—National Center for Environmental Assessment-RTF Division,
Office of Research and Development, U.S. Environmental Protection Agency, Research Triangle
Park, NC
Ms. Debra Walsh (Deputy Director)—National Center for Environmental Assessment-RTP Division,
Office of Research and Development, U.S. Environmental Protection Agency, Research Triangle
Park, NC
Dr. Mary Ross (Branch Chief)—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Doug Johns (Acting Branch Chief)—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Scientific Staff
Dr. James Brown (O3 Team Leader)—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Christal Bowman—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Barbara Buckley—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Ye Cao—Oak Ridge Institute for Science and Education, National Center for Environmental
Assessment, Office of Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Mr. Allen Davis—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Jean-Jacques Dubois—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Steven J. Dutton—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Jeffrey Herrick—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Erin Hines—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Dennis Kotchmar—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
March 2011 xxviii DRAFT - DO NOT CITE OR QUOTE
-------
Dr. Meredith Lassiter—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Lingli Liu— Oak Ridge Institute for Science and Education, National Center for Environmental
Assessment, Office of Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Thomas Long—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Thomas Luben—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Qingyu Meng— Oak Ridge Institute for Science and Education, National Center for
Environmental Assessment, Office of Research and Development, U.S. Environmental Protection
Agency, Research Triangle Park, NC
Dr. Kristopher Novak—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Elizabeth Oesterling Owens—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Molini Patel—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Joseph P. Pinto—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Joann Rice—on detail to the National Center for Environmental Assessment, Office of Research
and Development, from the Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. Jason Sacks—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Lisa Vinikoor-Imler—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Technical Support Staff
Mr. Kenneth J. Breito-Senior Environmental Employment Program, National Center for
Environmental Assessment, Office of Research and Development, U.S. Environmental Protection
Agency, Research Triangle Park, NC
Ms. Ellen Lorang—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. J. Sawyer Lucy-Student Services Authority, National Center for Environmental Assessment,
Office of Research and Development, U.S. Environmental Protection Agency, Research Triangle
Park, NC
Ms. Deborah Wales—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
March 2011 xxix DRAFT - DO NOT CITE OR QUOTE
-------
Mr. Richard N. Wilson-National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Barbara Wright—Senior Environmental Employment Program, National Center for
Environmental Assessment, Office of Research and Development, U.S. Environmental Protection
Agency, Research Triangle Park, NC
March 2011 xxx DRAFT - DO NOT CITE OR QUOTE
-------
Authors, Contributors, and Reviewers
Authors
Dr. James Brown (O3 Team Leader)—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Christal Bowman—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Barbara Buckley—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Maggie Clark—Department of Environmental and Radiological Health Sciences, Colorado State
University, Fort Collins, CO
Dr. Jean-Jacques Dubois—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Steven J. Dutton—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Kelly Gillespie— Donald Danforth Plant Science Center, St. Louis, MO
Dr. Terry Gordon—Department of Environmental Medicine, New York University School of
Medicine, Tuxedo, NY
Dr. Jeffrey Herrick—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Erin Hines—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Kazuhiko Ito—Department of Environmental Medicine, New York University School of
Medicine, Tuxedo, NY
Dr. Dennis Kotchmar—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Meredith Lassiter—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Lingli Liu— Oak Ridge Institute for Science and Education, Postdoctoral Research Fellow to
National Center for Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Thomas Long—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Thomas Luben—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
March 2011 xxxi DRAFT - DO NOT CITE OR QUOTE
-------
Dr. Loretta J. Mickley—School of Engineering & Applied Sciences, Harvard University, Cambridge,
MA
Dr. Kristopher Novak—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Elizabeth Oesterling Owens—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Molini Patel—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Jennifer Peel—Department of Environmental and Radiological Health Sciences, Colorado State
University, Fort Collins, CO
Dr. Joseph Pinto—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Joann Rice—on detail to the National Center for Environmental Assessment, Office of Research
and Development, from the Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. Jason Sacks—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. George Thurston—Department of Environmental Medicine, New York University School of
Medicine, Tuxedo, NY
Dr. Lisa Vinikoor-Imler—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Cosima Wiese—Department of Biology, Misericordia University, Dallas, PA
Contributors
Mr. Brian Adams—Oak Ridge Institute for Science and Education, National Center for
Environmental Assessment, Office of Research and Development, U.S. Environmental Protection
Agency, Research Triangle Park, NC
Dr. Halil Cakir—Oak Ridge Institute for Science and Education, National Center for Environmental
Assessment, Office of Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Ms. Ye Cao—Oak Ridge Institute for Science and Education, National Center for Environmental
Assessment, Office of Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Mr. Allen Davis—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. Mark Evangelista—Office of Air Quality Planning and Standards, Office of Air and Radiation,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. E. Henry Lee—National Health and Environmental Effects Research Laboratory, U.S.
Environmental Protection Agency, Corvallis, OR
March 2011 xxxii DRAFT - DO NOT CITE OR QUOTE
-------
Dr. Qingyu Meng—Oak Ridge Institute for Science and Education, Postdoctoral Research Fellow to
National Center for Environmental Assessment, Office of Research and Development, U.S.
Environmental Protection Agency, Research Triangle Park, NC
Mr. David Mintz—Office of Air Quality Planning and Standards, Office of Air and Radiation, U.S.
Environmental Protection Agency, Research Triangle Park, NC
Mr. Mark Schmidt—Office of Air Quality Planning and Standards, Office of Air and Radiation, U.S.
Environmental Protection Agency, Research Triangle Park, NC
Dr. Huiquin Wang, School of Engineering and Applied Science, Harvard University, Cambridge, MA
Mr. Benjamin Wells—Office of Air Quality Planning and Standards, Office of Air and Radiation,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Reviewers
Dr. Christian Andersen—National Health and Environmental Effects Research Laboratory, U.S.
Environmental Protection Agency, Corvallis, OR
Ms. Lea Anderson—Office of General Counsel, U.S. Environmental Protection Agency, Washington,
D.C.
Dr. Robert Arnts—National Exposure Research Laboratory, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. John Balmes— Department of Medicine, University of California, San Francisco and School of
Public Health, University of California, Berkeley, CA
Dr. Souad Benromdhane—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Fitzgerald Booker—USDA-ARS Plant Science Research Unit, Raleigh, NC
Dr. Philip Bromberg—School of Medicine, University of North Carolina, Chapel Hill, NC
Dr. Kent Burkey—USDA-ARS Plant Science Research Unit, Raleigh, NC
Dr. David DeMarini—National Health and Environmental Effects Research Laboratory, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Russ Dickerson—Department of Atmospheric and Oceanic Science, University of Maryland,
College Park, MD
Mr. Patrick Dolwick—Office of Air Quality Planning and Standards, Office of Air and Radiation,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Aimen Farraj—National Health and Environmental Effects Research Laboratory, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Arlene Fiore—NOAA/Geophysical Dynamics Laboratory, Princeton, NJ
Dr. Ian Gilmour—National Health and Environmental Effects Research Laboratory, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
March 2011 xxxiii DRAFT - DO NOT CITE OR QUOTE
-------
Dr. Tara Greaver—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Gary Hatch—National Health and Environmental Effects Research Laboratory, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Bryan Hubbel—Office of Air Quality Planning and Standards, Office of Air and Radiation,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Karl Jensen—National Health and Environmental Effects Research Laboratory, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Urmila Kodavanti—National Health and Environmental Effects Research Laboratory, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Petros Koutrakis—Department of Environmental Health, Harvard School of Public Health,
Boston, MA
Mr. John Langstaff—Office of Air Quality Planning and Standards, Office of Air and Radiation,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Christopher Lau—National Health and Environmental Effects Research Laboratory, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. Gary Lear—Office of Air and Radiation, U.S. Environmental Protection Agency, Office of
Administration and Policy, Washington, DC
Dr. Morton Lippmann—Nelson Institute of Environmental Medicine, New York University, Tuxedo,
NY
Dr. Karen Martin—Office of Air Quality Planning and Standards, Office of Air and Radiation,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Connie Meacham—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. David Mintz—Office of Air Quality Planning and Standards, Office of Air and Radiation, U.S.
Environmental Protection Agency, Research Triangle Park, NC
Dr. Pradeep Rajan—Office of Air Quality Planning and Standards, Office of Air and Radiation,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. John Rogers—National Health and Environmental Effects Research Laboratory, Office of
Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Vicki Sandiford—Office of Air Quality Planning and Standards, Office of Air and Radiation,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Susan Stone—Office of Air Quality Planning and Standards, Office of Air and Radiation,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. John Vandenberg—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. James G Wagner—Department of Pathobiology and Diagnostic Investigation, Michigan State
University, East Lansing, MI
March 2011 xxxiv DRAFT - DO NOT CITE OR QUOTE
-------
Ms. Debra Walsh—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Jason West—Department of Environmental Sciences & Engineering, University of North
Carolina, Chapel Hill, NC
March 2011 xxxv DRAFT - DO NOT CITE OR QUOTE
-------
Clean Air Scientific Advisory Committee
Ozone NAAQS Review Panel
Chair of the Environmental Protection Agency's Clean Air Scientific Advisory Committee
Dr. Jonathan M. Samet*, Department of Preventive Medicine at the Keck School of Medicine, and
Director of the Institute for Global Health at the University of Southern California, Los Angeles, CA
Chair of the Ozone Review Panel
Dr. Jonathan M. Samet*, Department of Preventive Medicine at the Keck School of Medicine, and
Director of the Institute for Global Health at the University of Southern California, Los Angeles, CA
Members
Dr. George A. Allen*, Northeast States for Coordinated Air Use Management (NESCAUM), Boston,
MA
Professor Ed Avol, Department of Preventive Medicine, Keck School of Medicine, University of
Southern California, Los Angeles, CA
Dr. John Bailar, The National Academies, Washington, D.C.
Dr. Michelle Bell, School of Forestry & Environmental Studies, Yale University, New Haven, CT
Dr. Joseph Brain*, Department of Environmental Health, Harvard School of Public Health, Harvard
University, Boston, MA
Dr. David Chock, Independent Consultant, Bloomfield Hills, MI
Dr. William Michael Foster, Division of Pulmonary, Allergy, and Critical Care Medicine, Duke
University Medical Center, Durham, NC
Dr. H. Christopher Frey*, Department of Civil, Construction and Environmental Engineering,
College of Engineering, North Carolina State University, Raleigh, NC
Dr. Judith Graham, Independent Consultant, Pittsboro, NC
Dr. David Grantz, College of Natural and Agricultural Sciences, Air Pollution Research Center,
University of California Riverside, Parlier, CA
Dr. Jack Harkema, Center for Integrated Toxicology, Michigan State University, East Lansing, MI
Dr. Daniel Jacob, Atmospheric Chemistry and Environmental Engineering, Harvard University,
Cambridge, MA
Dr. Steven Kleeberger, National Institute of Environmental Health Sciences, National Institutes of
Health, Research Triangle Park, NC
March 2011 xxxvi DRAFT - DO NOT CITE OR QUOTE
-------
Dr. Frederick J. Miller, Independent Consultant, Gary, NC
Dr. Howard Neufeld, Department of Biology, Appalachian State University, Boone, NC
Dr. Armistead (Ted) Russell*, Department of Civil and Environmental Engineering, Georgia Institute
of Technology, Atlanta, GA
Dr. Helen Suh Macintosh*, Environmental Health, NORC at the University of Chicago, and the
School of Public Health, Harvard University, Boston, MA
Dr. James Ultman, Department of Chemical Engineering, Pennsylvania State University, University
Park, PA
Dr. Sverre Vedal, Department of Environmental and Occupational Health Sciences, School of Public
Health and Community Medicine, University of Washington, Seattle, WA
Dr. Kathleen Weathers*, Gary Institute of Ecosystem Studies, Millbrook, NY
Dr. Peter Woodbury, Department of Crop and Soil Sciences, Cornell University, Ithaca, NY
* Members of the statutory Clean Air Scientific Advisory Committee (CAS AC) appointed by the
EPA Administrator
Science Advisory Board Staff
Dr. Holly Stallworth, Designated Federal Officer, Environmental Protection Agency, Mail Code
1400R, 1300 Pennsylvania Avenue, NW, Washington, DC, 20004, Phone: 202-564-2073, Email:
stallworth.holly@epa.gov
March 2011 xxxvii DRAFT - DO NOT CITE OR QUOTE
-------
Acronyms and Abbreviations
a
a-TOH
A, A
AA
AADT
AAS
ABA
ABI2
AC
ACC
ACE
ACGIH
ACh
AChE
ACS
ACS-CPS-II
ADC
ADSS
AED
AER
AEROCE
AF
AGL
AH2
AHCs
A horizon
AHR
AhR
AHSMOG
AirPEx
AirQUIS
AIRS
alpha, ambient exposure factor
Alpha tocopherol
Angstrom, angstrom (10"1U meter)
arachidonic acid; ambient air; atomic absorption; ascorbic acid
annual average daily traffic
atomic absorption (spectrophotometry, spectrometry,
spectroscopy)
abscisic acid
phospho-tyrosine-specific protein phosphatase
air conditioning
1 -aminocyclopropane-1 -carboxylate
angiotensin converting enzyme
American Conference of Governmental Industrial Hygienists
acetylcholine
acetylcholinesterase
American Cancer Society; 1-aminocyclopropane-1-carboxylase
synthase
ACS Cancer Prevention Study II
arginine decarboxylase
aged and diluted side stream cigarette smoke
aerodynamic equivalent diameter
air exchange rate
Atmospheric/Ocean Chemistry Experiment
atrial fibrillation; absorption fraction; adsorbed fraction
above ground level
ascorbic acid
aromatic hydrocarbons
uppermost layer of soil (litter and humus)
airway(s) hyperresponsiveness, airway(s) hyperreactivity
aryl hydrocarbon receptor
(California Seventh Day) Adventist Heath and Smog (Study)
Air Pollution Exposure (model)
Air Quality Information System (model, Norwegian Institute for
Air Research [NILU])
Aerometric Information Retrieval System; Atmospheric Infrared
Sounder (instrument)
March 2011
XXXVIII
DRAFT - DO NOT CITE OR QUOTE
-------
ALI
AM
•
-------
A/V
AVG
avg
AZO
P
P2-AR
BNP
BAL
BALF
BALT
B[a]P
BC
BCB
BEIS
BERLIOZ
BHC
BLD
BME
BMI
BMZ
BP
bpm
BrdU
BS
BSA
Bscatter
bw
C
C3
C3a
C4
CA
Ca
Caz+
CAA
surface-to-volume ratio
1 -aminoethoxy vinyl-glycine
average
azoxystrobin
beta, beta coefficient, slope; log relative risk
beta-2-adrenergic receptor
P -type natriuretic peptide
bronchoalveolar lavage; British anti-Lewisite (AKA
dimercaprol)
bronchoalveolar lavage fluid
bronchus-associated lymphoid tissue(s)
benzo[a]pyrene
black carbon
blue copper binding protein
Biogenic Emissions Inventory System
Berlin Ozone Experiment
biogenic hydrocarbons
below limit of detection
Bayesian Maximum Entropy (framework)
body mass index
basement membrane zone
blood pressure
breaths per minute
bromodeoxyuridine
black smoke
bovine serum albumin; body surface area
back scatter
body weight
carbon; concentration
plants that use only the Calvin cycle for fixing the carbon
dioxide from the air
complement protein fragment
plants that use the Hatch-Slack cycle for fixing the carbon
dioxide from the air
Conducting airways
calcium
calcium ion
Clean Air Act
March 2011
DRAFT - DO NOT CITE OR QUOTE
-------
CAAA
Cab
CADS
CAM
CAP(s)
CAPMoN
CAR
CARS
CASAC
CASTNET
CAT
CB
CBL
CBU
CBVD
% c/c
CC16
CCh
CCSP
CDC
cDNA
CDPHE
CDT
Cdyn, Cdyn
CE
CEC
CEPEX
CF
CFA
CFCs
CFD
CFI
CFR
CG
CGRP
CH3
Amendments to the Clean Air Act (1990)
chlorophyll a/b binding protein
Cincinnati Activity Diary Study
plants that use crassulacean acid metabolism for fixing the
carbon dioxide from the air
concentrated ambient particles
Canadian Air and Precipitation Monitoring Network
centriacinar region
California Air Resources Board
Clean Air Scientific Advisory Committee
Clean Air Status and Trends Network
catalase; computer-aided tomography
carbon black
convective boundary layer
cumulative breath units
cerebrovascular disease
percent carbon of total carbon
Clara cell protein, Clara cell 16 protein
carbachol
Climate Change Science Program; Clara cell secretory protein
Centers for Disease Control and Prevention
complementary DNA
Colorado Department of Public Health and Environment
Central Daylight Time
dynamic lung compliance
continuous exercise
controlled environment chambers
Central Equatorial Pacific Experiment
charcoal-filtered
charcoal/Purafil-filtered air
chlorinated fluorocarbons
computational fluid dynamics (modeling)
continuous forest inventory
Code of Federal Regulations; reference method
cloud-to-ground (lightning flash)
Calcitonin gene-related peptide
methyl group
March 2011
xli
DRAFT - DO NOT CITE OR QUOTE
-------
CH4
C2H4
C5H8
CHAD
ChAT
CH3Br
CH2=C(CH3)-CHO
CH3CC13
CH3CHO
CH3CH(ONO2)CHO
CHC13
CH3C1
CH3CN
CH3-CO
CH3-C(O)-CH=CH2
CH3C(O)CH2ONO2
CH3CO3NO2
CH3-C(O)O2, CH3-C(O)OO
C2Hs— H; C2He
CH2O
CH30
CH302'
CH3OH
CH3-0(0)CH3
CH3 OOH
CHD
CHF
CH3I
CHIP
CHO
CI
CIE
CIMS
methane
ethene
isoprene
benzene
terpene
Consolidated Human Activity Database
choline acetyl-transferase
methyl bromide
methacrolein
Methyl chloroform
toluene
acetaldehyde
2-nitratopropanol
chloroform
Methyl chloride
acetonitrile
acetyl
methyl vinyl ketone
1 -nitratopropanone
PAN, peroxyacetyl nitrate
acetyl peroxy radical; peroxyacetyl
ethane
formaldehyde
methoxy
methyl peroxy (radical)
methanol
acetone
acetic acid; methyl hydroperoxide
coronary heart disease
congestive heart failure
methyl iodide
Effects of Elevated Carbon Dioxide and Ozone on Potato Tuber
Quality in the European Multiple Site Experiment
Chinese hamster ovary cells
confidence interval(s)
Commission Internationale de 1'Eclaiarage (International
Commission on Illumination)
chemical ionization mass spectroscopy
March 2011
xlii
DRAFT - DO NOT CITE OR QUOTE
-------
CINC
CIU
CL
Cl
CLM
CMAQ
CMBO
CMD
CMSA
CN
CO
C02
COD
ConA
COP
COPD
C-R
CRKs
CRP
CS
CSA
CSTR
CTL
CTM
CU
CUOt
CV; c.v.; cv
CV
C.V.
CVD
CYP
CYP 1A1
CyS
cyt
A, 6
cytokine-induced neutrophil chemoattractant
cumulative inhalation units
chemiluminescence
chlorine
chemiluminescence method
Community Multi-scale Air Quality modeling system;
Congestion Mitigation and Air Quality
chloromethylbutenone
count median diameter
consolidated metropolitan statistical area
condensation nuclei
carbon monoxide; Cardiac output
carbon dioxide
coefficient of divergence; coefficient of determination
concanavalin A
Conference of Parties
chronic obstructive pulmonary disease
concentration-response
cysteine-rich RLKs, which are part of the receptor-like/Pelle
kinase (RLKs) group
C-reactive protein
cortico steroid
Combined Statistical Area
continuous stirred tank reactor
cytotoxic T lymphocyte
chemical transport model
cumulative uptake (coefficient of variation)
The cumulative stomatal uptake of Os, using a constant Os
uptake rate threshold oft nmol/m2/s
Cultivar
cardiovascular
coefficient of variation
cardiovascular disease
cytochrome (e.g., CYP1A, CYP-2A6, CYP3A4, CYP450)
cytochrome P450 1 Al
Protein cysteines
cytochrome
delta, difference; change
March 2011
xliii
DRAFT - DO NOT CITE OR QUOTE
-------
2-D
3-D
dso
Da
Da
DA
DAMPS
DBF
DD
df
DG
DGDG
DHA
DHAR
DHBA
DI
DIAL
DLEM
DMPO
DNA
DOAS
DOC
DOE
DOPAC
DPCC
DR
DTPA
DU
8
EEC
EC
ECo.05%
ECG
ECM
ECOPHYS
EC-SOD
two-dimensional
three-dimensional
50 percent cut point or 50 percent diameter
aerodynamic diameter
Dalton
dry airstream; dopamine; dopaminergic
3-deoxy-D-arabino-heptulosonat-7-phosphate synthase
diastolic blood pressure
doubling dose
degrees of freedom
diacylglycerol
digalactosyldiacylglycerol
dehydroascorbate; docosahexaenoic acid
dehydroascorbate reductase
2,3-dihydroxybenzoic acid
dry intrusion
differential absorption lidar (system)
Dynamic Land Ecosystem Model
dimethylphrrolise 1-oxide; 5,5-dimethyl-1-pyrrolineN-oxide
deoxyribonucleic acid
differential optical absorption spectroscopy
dissolved organic carbon
U.S. Department of Energy
3,4-dihydroxyphenylacetic acid
l,2-dipalmitoyl-SN-glycero-3-phosphocholine
disulfide reductase
diethylene triamine pentaacetic acid
Dobson units
epsilon; convergence precision
exhaled breath condensate (fluid)
elemental carbon
0.05% excess risk in mortality
effect concentration for 50% of test population
electrocardiography; electrocardiogram
ectomycorrhizal fungi
whole-tree ecophysiological growth process model
extracellular superoxide dismutase
March 2011
xliv
DRAFT - DO NOT CITE OR QUOTE
-------
ED
EDMAS
EDTA
EDU
EE
EEA(s)
EEG
eGPx
EGTA
EGU(s)
EKG, ECG
ELF
ELISA
EM
ENA-78
eNO
eNOS
ENSO
EOFs
EOTCP
EPA
EPEM
EPO
EPR
EPRI
ER
ERAQS
ERD1
ESPACE-wheat
ESR
EST
ET
ETS
EU
EVR
F
emergency department
Exposure and Dose Modeling and Analysis System
ethylenediaminetetraacetic acid
ethylenediurea
energy expenditure (average EE rate)
Essential Ecological Attribute(s)
electroencephalogram; electroencephalographic
extracellular glutathione peroxidase
ethyleneglycoltetraacetic acid
electricity generating unit(s)
electrocardiogram
epithelial lining fluid; extracellular lining fluid
enzyme-linked immunosorbent assay
electron microscopy
epithelial cell-derived neutrophil-activating peptide 78
exhaled nitric oxide
endothelial nitric oxide synthase
El Nino-Southern Oscillation
empirical orthogonal functions
European Open Top Chamber Programme
U.S. Environmental Protection Agency
Event Probability Exposure Model
epoxyconazole
Electron Paramagnetic Resonance
Electric Power Research Institute
emergency room; excess risk
Eastern Regional Air Quality Study
ethylene response
European Stress Physiology and Climate Experiment on the
Effects of Carbon Dioxide and Oxygen on Spring Wheat
electron spin resonance (spectroscopy); EPR
Eastern Standard Time
ethylene; endotracheal
environmental tobacco smoke
endotoxin units; European Union
equivalent ventilation rate
female
March 2011
xlv
DRAFT - DO NOT CITE OR QUOTE
-------
f, /, fB
F344
FA
FAA
FACE
fB
FDA
FEF
FEF25-75
FEFX
FEM
FEVi
FFAs
FGF
FGFR
FHM
FIA
Finf
FIVC
FLAG
FN;Fn
FP
FPM
FR
FRAP
FRC
FRM
FS
FTIR
FVC
Fv/Fm
FVI
G
GAM
GBS
frequency of breathing
Fischer 344 strain of rats
filtered air; fatty acid; fractional absorption; absorbed fraction
Federal Aviation Administration
free-air CO2 enrichment (system)
breathing frequency
Food and Drug Administration
forced expiratory flow
forced expiratory flow between the times at which 25% and 75%
of the vital capacity is reached
forced expiratory flow after (x)% vital capacity (e.g., after 25,
50, or 75% vital capacity)
Federal equivalent method
forced expiratory volume in 1 second
free fatty acids
fibroblast growth factor
fibroblast growth factor receptor
Forest Health Monitoring
USDA Forest Inventory and Analysis Program
infiltration factor
forced inspiratory vital capacity
Federal Land Managers' Air Quality Related Values Workgroup
fibronectin
fluticasone propionate
Forest Pest Management
Federal Register; fixed-ratio operant conditioning; fixed ratio
schedule
ferric reducing ability of plasma
functional residual capacity
Federal reference method
field stimulation
Fourier Transform Infrared Spectroscopy
forced volume vital capacity
a measure of the maximum efficiency of Photosystem II
fruit and vegetable index
plants rooted in ground
generalized additive model(s)
group B streptococcus
March 2011
xlvi
DRAFT - DO NOT CITE OR QUOTE
-------
GC
GCE
GC-FID
GCM(s)
GC/MS
GD
GDP
GDI
GEE
GEOS
GEOS-1DAS
GEOS-Chem
GHG
GLM(s)
GLMM(s)
GLRAG
GLU
GM-CSF
GMD
GMT
G6P
G6PD
GPP
GPx
GR
GRSM
GSFC
GSH
GSHPx, GPx
GSMNP
GSSG
GST
GSTM1
GSTMlnull
GSTP1
GSx
H; FT: FT
gas chromatography
Goddard Cumulus Ensemble (model)
gas chromatography-flame ionization detection
general circulation model(s), global climate model
gas chromatography/mass spectrometry
gestational day
guanosine diphosphate
glutathione-disulfide transhydrogenase
generalized estimating equations
Goddard Earth Observing System
Goddard Earth Observing System Data Assimilation System
Goddard Earth Observing System-Chemistry (model)
greenhouse gas
generalized linear model(s)
generalized linear mixed model(s)
Great Lakes Regional Assessment Group
glutamate
granulocyte macrophage colony-stimulating factor
Global Monitoring Division
Greenwich mean time
glucose-6-phosphate
glucose-6-phosphate dehydrogenase
Gross Primary Production
glutathione peroxidase
glutathione reductase
Great Smoky Mountains National Park
NASA Goddard Space Flight Center
glutathione; reduced glutathione
glutathione peroxidase
Great Smoky Mountains National Park
oxidized glutathione; glutathione disulfide
glutathione transferase; glutathione ^-transferase
glutathione S-transferase polymorphism Ml
glutathione S-transferase u-1 null (genotype)
glutathione S-transferase P 1
glutathione; glutathione-containing material
atomic hydrogen, hydrogen ion; hydrogen radical
March 2011
xlvii
DRAFT - DO NOT CITE OR QUOTE
-------
h
H2
HA
ha
HC(s)
HCFC(s)
HCHO
HCO'
H2CO, HCHO
HDM
2HDM, 2ndHDM
HDMA
HEPA
HERO
HF
HFCs
Hg
HHP-C9
HIST
HLA
HMOX
HMOX1
HNE
HNO2, HONO
HNO3
radiolabeled hydrogen; tritium
hour
molecular hydrogen
hospital admission
hectare
hydrocarbon(s)
hydrochlorofluorocarbon(s)
formaldehyde
formyl (radical)
formaldehyde
house dust mite
second-highest daily maximum 1-h concentration
house dust mite allergen
high efficiency particle air (filter)
Health and Environmental Research Online, NCEA Database
System
hydrogen fluoride; hydrofluoride; Howland Forest site;
(HRV signal) high-frequency power
hydrofluorocarbons
mercury
1 -hydroxy-1 -hydroperoxynonane
histamine
human leukocyte antigen
Heme oxygenase
heme-oxygenase (decycling)-l
4-hydroxynonenal
nitrous acid
nitric acid
HNO4
HO
HO'
HO2'
HO3'
H2O
H3O+
H2O2
HOCH2OOH
pernitric acid
hydroxyl; heme oxygenase
hydroxyl radical
hydroperoxyl; hydroperoxy radical; protonated superoxide
protonated ozone radical
water
Hydronium ion
hydrogen peroxide
hydroxymethylhydroperoxide
March 2011
xlviii
DRAFT - DO NOT CITE OR QUOTE
-------
HONO
HO2NO2
HOONO
HOX
HPLC
HPOT
HR
HRmax
HRP
HRV
hs-CRP
H2SO4
5-HT
hv
HVAC
Hz
I
I/O
IARC
IAS
IBM
1C
ICAM-1
ICAS
ICD
ICD-9
ICD-10
ICEM
ICNIRP
ICP Forests
ICS
ID#
IDW
IE
IFN
IFN-y
nitrous acid
peroxynitric acid
pernitrous acid
hydrogen radical(s)
high-pressure liquid chromatography
13-hydroperoxide linolenic acid
heart rate, hazard ratio; hypersensitive response
maximum heart rate
horseradish peroxidase
heart rate variability
high-sensitivity C-reactive protein
sulfuric acid
5 -hy droxytryptamine
Energy per photon of electromagnetic energy at frequency v
heating, ventilation, and air conditioning
hertz
iodine
indoor-outdoor ratio
International Agency for Research on Cancer
interalveolar septum
individual-based model or modeling
inspiratory capacity; intracloud (lightning flash)
intercellular adhesion molecule
Inner City Asthma Study
implantable cardioverter defibrillator(s); International
Classification of Diseases
International Classification of Disease 9th revision
International Classification of Disease 10th revision
Indoor Chemistry and Exposure Model
International Commission on Non-Ionizing Radiation Protection
International Cooperative Programme on Assessment of Air
Pollution Effects on Forests
inhaled steroids
identification number
inverse-distance-weighted
intermittent exercise
interferon (e.g., IFN-Q)
interferon-gamma
March 2011
xlix
DRAFT - DO NOT CITE OR QUOTE
-------
Ig
IgA
IgE
IGFi
IgG
IgM
IHD
IL
He
IL-x
i.m.
IMPACT
IMPROVE
IN
INF
inh
iNOS
INTRASTAND
IL-lp
IOM
i.p.
IPCC
IPCC-AR4
IPCC-TAR
IPMMI
IQR
IR
ISA
ISCCP
ISO
8-iso-PGF
IT
IU
IUGR
immunoglobulin (e.g., IgE)
immunoglobin A
immunoglobin E
insulin-like growth factor 1
immunoglobin G
Immunoglobulin M fraction
ischemic heart disease
interleukin
isoleucine
interleukin-6, 8, etc.
intramuscular (route)
Interactive Modeling Project for Atmospheric Chemistry and
Transport
Interagency Monitoring of Protected Visual Environment
intranasal
interferon
inhalation
inducible nitric oxide synthase
a stand-level model designed for hourly, daily and annual
integration of forest carbon and water cycle fluxes
Interleukin-1B eta
Institute of Medicine
intraperitoneal (route)
Intergovernmental Panel on Climate Change
Intergovernmental Panel on Climate Change 4th Assessment
Report
Intergovernmental Panel on Climate Change Third Assessment
Report
International Photolysis Frequency Measurement and Modeling
Inter-comparison
interquartile range
infrared
Integrated Science Assessment
International Satellite Cloud Climatology Project
International Standards Organization
8-isoprostane
intratracheal, intratracheally
International Units
intrauterine growth restriction
March 2011
DRAFT - DO NOT CITE OR QUOTE
-------
l.V.
IVF
JA
Jmax
J(N02)
J(03)
JPL
Jsat
KB
k
K+
Ka
Kg
kHz
kJ
Ki
km
KM
KO
Kr
KROFEX
KTB
L, dL, mL, uL
LAI
LEW
LCso
LCL
LDH
LDL
LF
LFHFR
LFT
LI
LIDAR
LIF
LINKAGES
LIS
intravenous (route)
in vitro fertilization
jasmonic acid
maximum rate of electron transport (for regeneration of RuBP)
photolysis rate coefficient for NO2
photolysis rate coefficient for 63
Jet Propulsion Laboratory
saturating light
kappa B
dissociation rate; root:shoot allometric coefficient
potassium ion
intrinsic mass transfer coefficient/parameter
mass transfer coefficient for gas phase
kilohertz
kilojoules
mass transfer coefficient for liquid phase
kilometer
particle optical reflectance
knockout
reaction rate constant
Krauzberg Ozone Fumigation Experiment
tracheobronchial region overall mass transfer coefficient
Liter, deciLiter, milliLiter, microLiter
leaf area index
low birth weight
median lethal concentration
lower 95th% confidence limit
lactate dehydrogenase, lactic acid dehydrogenase
low-density lipoprotein
(HRV signal) low-frequency power
low frequency/high frequency (ratio)
lower free troposphere
labeling index
Light Detection and Ranging (remote sensing system)
laser-induced fluorescence
individual-based model of forest succession
lateral intercellular space
March 2011
DRAFT - DO NOT CITE OR QUOTE
-------
LLJ
LM
Ln
LnRMSSD
InSDNN
LOAEL
LOD
LOEL
LOESS
LOP
LOWESS
LOX-1
LPS
LRS
LRT
LSI
LT
LTA
LT-a
LUR
LWC
(^
ueq
ug
ug/mj
urn
M
M7
M12
M, mM, uM, nM, pM
m, cm, um, nm
ma
mAOT
MAP
MAPK
MAQSIP
low-level jet
light microscopy
Natural logarithm
natural log of RMSSD; measure of HRV
natural log of the standard deviation of NN intervals in an EKG
lowest observed adverse effect level
limit of detection
lowest-observed-effect level
locally weighted scatterplot smoothing
lipid ozonation products
locally weighted scatter plot smoother
Lipoxygenase; lectin-like oxidized low density lipoprotein
receptor-1
lipopolysaccharide
lower respiratory symptoms
lower respiratory tract; lower airways; Long range transport
local standard time
leukotriene (e.g., LTB4 , LTC 4, LTD4 , LTE4); local time
lymphotoxin-alpha
lymphotoxin-a
land use regression
liquid water content
mu, micro
microequivalent
microgram
micrograms per cubic meter
micrometer, micron
male; maximum number of iterations; air molecule
7-hour seasonal mean
12-hour seasonal mean of Os
Molar, milliMolar, microMolar, nanoMolar, picoMolar
meter(s), centimeter(s), micrometer/[micron](s), nanometer(s)
moving average
modified accumulated exposure over threshold
mitogen-activated protein; mean arterial pressure
mitogen-activated protein kinase(s), MAP kinase
Multiscale Air Quality Simulation Platform (model)
March 2011
DRAFT - DO NOT CITE OR QUOTE
-------
MARAT
MARCO
max
MBL
MCCP
Mch; MCh
MCM
MCP
MDA
MDAR
MDI
MDL
MED
MEF50o/0
MeJA
MENTOR
METs
MGDG
MHC
MI
MIESR
min
MIP
mL
MLN
Mm
MMMt.
MM5
MMAD
MMEF
mmHg
MMMD
MnSOD
mo
MOA(s)
Mid-Atlantic Regional Assessment Team
Macrophage receptor with collagenous structure
maximum
marine boundary layer
Mountain Cloud Chemistry Program
methacholine
master chemical mechanism
monocyte chemotactic protein
malondialdehyde
monodehydroascorbate reductase
Mediterranean diet index
minimum detection level
minimal erythema dose
maximal midexpiratory flow at 50% of forced vital capacity
methyl jasmonate
Modeling Environment for Total Risk Studies
metabolic equivalent unit(s) [of work]
monogalactosyldiacylglycerol
major histocompatibility complex
myocardial infarction, "heart attack"
matrix isolation electron spin resonance (spectroscopy)
minute; minimum
macrophage inflammatory protein
milliliter
mediastinal lymph node
megameter
Mt. Mitchell site
National Center for Atmospheric Research/Penn State
Mesoscale Model, version 5
mass median aerodynamic diameter; mass median aerodynamic
density
maximal midexpiratory flow
millimeters of mercury
mean maximum mixing height depth
Manganese superoxide dismutase
month
mode(s) of Action
March 2011
DRAFT - DO NOT CITE OR QUOTE
-------
MONICA
MoOx
MOSES
MOZAIC
MPAN
MPO
MQL
MRI
mRNA
MS
ms
MS/MS
MSA
MSL
MT
MT, Mt
MT1
Mtn
MV
MW
N
°N
n, N
N2
N100
NA
NA; N/A
Na
Na+
NAAQS
NAD
NADH
NADP
NADPH; NAD(P)H
Monitoring of Trends and Determinants in Cardiovascular
Disease
molybdenum oxides
Met Office Surface Exchange Scheme
Measurement of Ozone and Water Vapor by Airbus In-Service
Aircraft
peroxymethacryloyl nitrate; peroxy-methacrylic nitric anhydride
myeloperoxidase
Minimum quantification limit
magnetic resonance imaging; Midwest Research Institute;
Meteorological Research Institute
messenger RNA
mass spectrometry; Mt. Moosilauke site
millisecond
tandem mass spectrometry
Metropolitan Statistical Area; methane sulfonic acid
mean sea level
million tons; metric ton
metallothionein
mitochondria
mountain
methyl viologen
molecular weight
nitrogen; North
nitrogen-15, stable isotope of nitrogen
number; number of observations
molecular nitrogen; nonreactive nitrogen
number of hours > 0.10 ppm
noradrenaline
not available
sodium
sodium ion
National Ambient Air Quality Standards
nicotinamide adenine nucleotide
reduced nicotinamide adenine dinucleotide; nicotinamide
adenine dinucleotide dehydrogenase
National Atmospheric Deposition Program
reduced nicotinamide adenine dinucleotide phosphate
March 2011
liv
DRAFT - DO NOT CITE OR QUOTE
-------
NADPH-CR
NaE
Na-K-ATPase
NAMS
NAMS/SLAMS
NAPAP
NAPBN
NARE
NARSTO
NAS
NASA
NBS
NBTH
NCEA
NCEA-RTP
NCHS
NCICAS
NCLAN
NCore
NC-R
NC-S
ND; n.d.
NDF
NEE
NEI
NEM
NEP
+NERAG
NERL
NESCAUM
NF
NF-KB
NH
NH3
NH4+
NHANES
reduced nicotinamide adenine dinucleotide
phosphatecytochrome c reductase
sodium erythorbate
sodium-potassium-dependent adenosine triphosphatase
National Ambient Monitoring Stations
National Ambient Monitoring Stations and State and Local Air
Monitoring Stations
National Acid Precipitation Assessment Program
National Air Pollution Background Network
North Atlantic Regional Experiment
North American Regional Strategy for Atmospheric Ozone
National Academy of Sciences; Normative Aging Study
National Aeronautics and Space Administration
National Bureau of Standards
3-methyl-2-benzothiazolinone acetone azine
National Center for Environmental Assessment
NCEA Division in Research Triangle Park, NC
National Center for Health Statistics
National Cooperative Inner-City Asthma Study
National Crop Loss Assessment Network
National Core multi-pollutant monitoring network
resistant clones of white clover
sensitive clones of white clover
not detectable; not detected; no data
neutral detergent fiber
net ecosystem CC>2 exchange
National Emissions Inventory
National Ambient Air Quality Standards Exposure Model
Net Ecosystem Production
New England Regional Assessment Group
National Exposure Research Laboratory
Northeast States for Coordinated Air Use Management
National Forest; non-filtered
nuclear factor kappa B
northern hemisphere
ammonia
ammonium ion
National Health and Nutrition Examination Survey
March 2011
DRAFT - DO NOT CITE OR QUOTE
-------
NHAPS
NHEERL/WED
NH4HSO4
(NH4)2SO4; (NH4)2HSO4
NHIS
NIH
NIST
NK
NKT
NL
NLF
NM
NMHC(s)
NMMAPS
NMOC(s)
NMVOCs
NN
NNK
nNOS
NO
NO2
N03; N03'
N03
N2O
N205
NOAA
NOAEL
NOS
NOS-1
NOS-2
NOS-3
NOX
NOY
NO/
National Human Activity Pattern Survey
EPA National Health and Environmental Effects Research
Laboratory, Western Ecology Division
Ammonium bisulfate
ammonium sulfate
National Health Interview Survey
National Institutes of Health
National Institute of Standards and Technology
natural killer cells; neurokinin
Natural killer T cells
nasal lavage
nasal lavage fluid
National Monument
nonmethane hydrocarbon(s)
National Morbidity, Mortality, and Air Pollution Study
nonmethane organic compound(s)
nonmethane volatile organic compounds
normal-to-normal (NN or RR) time interval between each QRS
complex in the EKG
4-(N-nitrosomethylamino)-1 -(3 -pyridyl)-1 -butanone
neuronal nitric oxide synthase (NOS)
nitric oxide
nitrogen dioxide
nitrate, nitrate radical
nitrate, nitrate ion
nitrous oxide
dinitrogen pentoxide
National Oceanic and Atmospheric Administration
no observed adverse effect level
nitric oxide synthase
neuronal nitric oxide synthase
inducible nitric oxide synthase; iNOS
endothelial nitric oxide synthase
nitrogen oxides, oxides of nitrogen (NO + NO2)
sum of NOx and NOz; odd nitrogen species; total oxidized
nitrogen
sum of all inorganic and organic reaction products of NOx
(HONO, HNO3, HNO4, organic nitrates, particulate nitrate,
nitro-PAHs, etc.)
March 2011
Ivi
DRAFT - DO NOT CITE OR QUOTE
-------
NP
NPP
NFS
NQO1
NQOlwt
NR
Nr
NRC
Nrf-2
NS; n.s.
NSAID
NSBR
NSF
NTE
NTN
NTP
NTRMs
NTS
NWR
NWS
NZW
O
U
15o
02
02
02'
03
1503
03*
OAQPS
OAR
OEMs
OC
OD
OCD)
OH, OH'
National Park
net primary production
National Park Service, U.S. Department of the Interior
NAD(P)H-quinone oxidoreductase (genotype)
NAD(P)H-quinone oxidoreductase wild type (genotype)
not reported
reactive nitrogen
National Research Council
nuclear factor erythroid 2-related factor 2
nonsignificant; non-smoker; national seashore; natural spline
non-steroidal anti-inflammatory agent
nonspecific bronchial responsiveness
National Science Foundation
nasal turbinate epithelial (cells)
National Trends Network
National Toxicology Program
NIST Traceable Reference Materials
nucleus of the solitary tract (in brainstem)
national wildlife refuge
National Weather Service
New Zealand white (rabbit)
oxygen; horizon forest floor
singlet oxygen
oxygen-18, stable isotope of oxygen
molecular oxygen
superoxide
superoxide radical
ozone
(oxygen-18 labeled) ozone
electronically excited ozone
Office of Air Quality Planning and Standards
Office of Air and Radiation
observationally based methods
organic carbon
outer diameter; optical density
electronically excited oxygen atom
hydroxyl group, hydroxyl radical
March 2011
DRAFT - DO NOT CITE OR QUOTE
-------
8-OHdG
OLS
OMI
0(JP)
OPE
OPECs
OR
ORD
OSHA
OTC
OVA
Ox
OxComp
O
OPSII-max
P
P
P9o
P450
PAD
PAF
PAH(s)
PAL
PAMS
PAN
PaO2
PAR
Patm
/7-ATP
PEL
PBM
PEN
PBPK
PBS
PC
PC20
8-hydroxy-2 '-deoxyguanosine
ordinary least squares
Ozone Monitoring Instrument
ground-state oxygen atom
ozone production efficiency
Outdoor Plant Environment Chambers
odds ratio
Office of Research and Development
Occupational Safety and Health Administration
open-top chamber
ovalbumin
odd oxygen species; total oxidants
oxidative capacity of the atmosphere
Phi; calculated efficiency
maximum photochemical effective quantum yield of PSII
pressure in atmospheres; plants grown in pots
probability value
90th percentile of the absolute difference in concentrations
cytochrome P450
peripheral arterial disease; pollutant applied dose
platelet-activating factor; paroxysmal atrial fibrillation
polycyclic aromatic hydrocarbon(s)
phenylalanine ammonia lyase
Photochemical Assessment Monitoring Stations network
peroxyacetyl nitrate; peroxyacetic nitric anhydride
arterial oxygen pressure
photosynthetically active radiation; proximal alveolar region
Pressure in atmospheres
/>ara-acetamidophenol
planetary boundary layer; peripheral blood lymphocytes
population-based model or modeling
C-phenyl N-tert-butyl nitrone
physiologically based pharmacokinetic (model)
phosphate buffered saline
phosphatidylchloline
provocative concentration that produces a 20% decrease in
forced expiratory volume in 1 second
March 2011
DRAFT - DO NOT CITE OR QUOTE
-------
PC20FEVi
PC50
PCA
PC-ALF
PCD
PCI
pCNEM
PC02
pC02
PCR
PCR-DGGE
PDiooSRaw
PD20; PD20FEVi
PE
PEF
PEFo.75
PEFR
PEFT
PEG-CAT
PEG-SOD
PEM(s)
Penh
PEPc
PFD
PG
6PGD
PGHS-2
PGP
PGSM
PH
PHA
PI
PIF
PK
provovative concentration that produces a 20% decrease in
FEV1
provocative concentration that produces a 50% decrease in
forced expiratory volume in 1 second
principal component analysis
1 -palmitoyl-2-(9-oxonononoyl)-5w-glycero-3 -phosphocholine
programmed cell death
picryl chloride
Canadian version of National Ambient Air Quality Standards
Exposure Model
Average partial pressure of O2 in lung capillaries
partial pressure of carbon dioxide
polymerase chain reaction
PCR-denaturing gradient gel electrophoresis
provocative dose that produces a 100% increase in sRAW
provocative dose that produces a 100% increase in SRaw
provocative dose that produces a 20% decrease in FEVi
post exposure, phosphatidylethanolamine
peak expiratory flow
peak expiratory flow in 0.75 second
peak expiratory flow rate
time to peak flow
polyethylene glycol-catalase
polyethylene glycol-superoxide dismutase
personal exposure monitor(s)
enhanced pause
phosphoenolpyruvate carboxylase
photo synthetic flux density
prostaglandin (e.g., PGE2 ,PGF2); phosphatidylglycerol
6-phosphogluconate dehydrogenase
prostaglandin endoperoxide G/H synthase 2
protein gene product (e.g., PGP9.5)
Plant Growth Stress Model
relative acidity; Log of the reciprocal of the hydrogen ion
concentration
phytohemagglutinin A
phosphatidylinositol; probability interval; posterior interval
peak inspiratory flow
pharmacokinetic s
March 2011
lix
DRAFT - DO NOT CITE OR QUOTE
-------
pKa
PLFA
PM
PMo.i
PMi
PM2.5
PMiO-2.5
PMx.y
PM15
PMX
PM-CAMx
PMN(s)
PMT
PND
pNEM
PnET
PNN
PNN
50
POC
POD
polyADPR
dissociation constant
Phospholipid fatty acid
particulate matter
particulate matter with a nominal mobility diameter less than or
equal to 0.1 um (referred to as ultrafme PM)
particulate matter with a nominal aerodynamic diameter less
than or equal to 1 um
particulate matter with a nominal aerodynamic diameter less
than or equal to 2.5 um (a surrogate for fine PM)
particulate matter with a nominal aerodynamic diameter greater
than 2.5 um and less than or equal to 10 um (a surrogate for
thoracic coarse particulate matter or the coarse fraction of PMio).
Concentration may be measured or calculated as the difference
between measured PMio and measured PM2.5 concentrations.
particulate matter with a nominal diameter greater than x um and
less than y um where x and y are the numeric mean aerodynamic
or mobility diameters (um).
particulate matter with a nominal aerodynamic diameter less
than or equal to 10 um
particulate matter with a nominal aerodynamic diameter less
than or equal to 13 um
particulate matter with a nominal aerodynamic diameter less
than or equal to 15 um
particulate matter of a specific size range. X refers to the
diameter at which the sampler collects 50% of the particles and
rejects 50% of the particles. Collection efficiency increases for
particles with smaller diameters and decreases for particles with
larger diameters. The variation of collection efficiency with size
is given by a collection efficiency curve. The definition of PMx
is frequently abbreviated as "particles with a nominal
aerodynamic diameter less than or equal to x um. See 40 CFR
58.1 for a full definition.
Comprehensive Air Quality Model with extensions and with
particulate matter chemistry
polymorphonuclear leukocyte(s)
photomultiplier tube
post natal day
probabilistic National Exposure Model
Photo synthetic EvapoTranspiration model
proportion of interval differences of successive normal-beat
intervals in EKG
proportion of interval differences of successive normal-beat
intervals greater than 50 ms in EKG
particulate organic carbon
peroxidase
poly(adenosinediphosphate-ribose)
March 2011
DRAFT - DO NOT CITE OR QUOTE
-------
ppb
ppbv
pphm
ppm
ppm-h
ppmv
PPN
PPPs
ppt
pptv
PQH2
PR
PR-1
PRB
PRYL
PS
PS II
PSA
PSC
PTB
PTR-MS
PU, PUL
PUFA(s)
PV
PVCD
PVD
PVOCs
PWM
Pxase
QCE
C[NP
qP
QRS
QT
parts per billion
parts per billion by volume
parts per hundred million
parts per million
parts per million hours; weighted concentration values based on
hourly concentrations: usually summed over a certain number of
hours, day(s), months, and/or season.
parts per million by volume
peroxypropionyl nitrate; peroxypropionic nitric anhydride
power plant plumes
parts per trillion
parts per trillion by volume
plastoquinone
pathogenesis-related (protein)
promoter region 1
policy-relevant background
predicted relative yield (biomass) loss
penalized spline
Photosystem II: enzyme that uses light to obtain electrons from
water (for photosynthesis).
picryl sulfonic acid
polar stratospheric clouds
preterm birth
proton-transfer-reaction mass spectroscopy
pulmonary
polyunsaturated fatty acid(s)
potential vorticity
peripheral vascular and cerebrovascular disease
peripheral vascular disease
photochemical volatile organic compounds
pokeweed mitogen
peroxidase
quasi continuous exercise
non-photochemical quenching
photochemical quenching
A complex of three distinct electrocardiogram waves which
represent the beginning of ventricular contraction
interval measure of the time interval between the start of the Q
wave and the end of the T wave in the heart's electrical cycle
March 2011
Ixi
DRAFT - DO NOT CITE OR QUOTE
-------
r
R,r
RACM
RADM
rALP
RAMS
RANTES
Raw, Raw
RB
RBC(s)
rbcL
rbcS
rcdl
R'C(0)-02
RDBMS
Re
REHEX
RER
RF
RGR
RH
RIOPA
RL
RLKs
RMR
rMSSD
Rn
RNA
RO2
ROG
ROI
RONO2
ROOH
Pearson correlation coefficient
correlation coefficient
correlation coefficient
multiple regression correlation coefficient
coefficient of determination
Regional Atmospheric Chemistry Mechanism
Regional Acid Deposition Model
recombinant antileukoprotease
Regional Atmospheric Modeling System
regulated upon activation, normal T cell expressed and secreted
(cells)
airway resistance
respiratory bronchiole
red blood cell(s); erythrocyte(s)
Rubisco large subunit
Rubisco small subunit
Arabidopsis mutant radical induced cell death
acyl peroxy
Relational Database Management Systems
Reynolds number
Regional Human Exposure Model
rough endoplasmic reticulum; Respiratory exchange ratio
radiative forcing
Relative growth rate
relative humidity
Relationship of Indoor, Outdoor, and Personal Air (study)
total pulmonary resistance
receptor-like/Pelle kinase group
resting metabolic rate
root mean squared differences between adjacent normal-to-
normal heartbeat intervals
nasal resistance
ribonucleic acid
organic peroxyl; organic peroxy
reactive organic gases
reactive oxygen intermediate/superoxide anion
organic nitrate
organic peroxides
March 2011
Ixii
DRAFT - DO NOT CITE OR QUOTE
-------
ROONO2, RO2NO2
ROS
RR
RRMS
RT
RT
RTLF
RuBisCO; Rubisco
RuBP
G
°g
s
s
SA
SAB
SAC
SAG21
SAI
SAMD
Sa02
SAPALDIA
SAR
SAROAD
SAWgrp
SBP
SBUV
SC
Sc
s.c.
SCAQS
SCE(s)
SD
SDNN
SE
peroxy nitrate
reactive oxygen species
normal-to-normal (NN or RR) time interval between each QRS
complex in the EKG; ribonucleotide reductase; risk ratio;
relative risk
relatively remote monitoring sites
respiratory tract
transepithelial resistance
respiratory tract lining fluid
ribulose-l,5-bisphosphate carboxylase/oxygenase
ribulose bisphosphate
sigma, standard deviation
sigma-g; geometric standard deviation
smoker; sulfur; South
second
salicylic acid
Science Advisory Board
Staphylococcus aureus Cowan 1 strain
senescence
Systems Applications International
S-adenosyl methionine decarboxylase
oxygen saturation of arterial blood
Study of Air Pollution and Lung Diseases in Adults
systemic acquired resistance
Storage and Retrieval of Aerometric Data (U.S. Environmental
Protection Agency centralized database; superseded by
Aerometric Information Retrieval System [AIRS])
small airway function group
systolic blood pressure
Solar Backscatter Ultraviolet Spectrometer
stratum corneum
scandium
subcutaneous
Southern California Air Quality Study
sister chromatid exchange(s)
standard deviation; Sprague-Dawley rat
standard deviation normal-to-normal (NN or RR) time interval
between each QRS complex in the EKG
standard error
March 2011
Ixiii
DRAFT - DO NOT CITE OR QUOTE
-------
SEE AS
SEM
SES
SF6
SGA
SGaw
SH
SHEDS
SHEN
sIC AM-1
SIDS
SIGMOID
SINIC
SIPK
SK
SLA
SLAMS
SMD
SME
SNAAQS
SNP(s)
SO2
SO4Z
SOA
SOD
SOS
SOX
SoyFACE
SP
SPF
SRaw, sRaw,
SRBC
SRES
SRM
SSCP
STE
Social Environment and Biomarkers of Aging Study
simultaneously extracted metal; standard error of the mean;
scanning electron microscopy
socioeconomic status
sulfur hexafluoride (tracer gas)
small for gestational age
specific airway conductance
Shenandoah National Park site
Stochastic Human Exposure and Dose Simulation
Shenandoah National Park
soluble intercellular adhesion molecule
sudden infant death syndrome
sigmoid weighted summed concentration
Simple Nitrogen Cycle model
salicylic acid (SA) induced protein kinase
shikimate kinase
specific leaf area
State and Local Air Monitoring Stations
soil moisture deficit
soybean oil methyl ester
Secondary National Ambient Air Quality Standards
single-nucleotide polymorphism
sulfur dioxide
sulfate
secondary organic aerosol
superoxide dismutase
Southern Oxidant Study
sulfur oxides
Soybean Free Air gas Concentration Enrichment Facility
surfactant protein (e.g., SPA, SPD); substance P
specific pathogen free
specific airway resistance
sheep red blood cell
Special Report on Emissions Scenarios
standard reference method
single-strand conformation polymorphism
strato sphere-tropo sphere exchange
March 2011
Ixiv
DRAFT - DO NOT CITE OR QUOTE
-------
STEP
STN
STP
STPD
STRF
SUMOO
SUM06
SUM07
SUM08
SURE
SZA
T
T
t
Tcell
T lymphocytes
T3
T4
TAR
TARWGI
TB
TEA
TEARS
TC
yymTc
yymTc-DTPA
T-cells
T
J^CO
TCTL
TOLAS
Te
TEM
TexAQS
Tg
TGF
TGFP
Th
Stratospheric-Tropospheric-Exchange Project
speciation trends network
standard temperature and pressure
standard temperature and pressure, dry
Spatio-Temporal Random Field (theory)
sum of all hourly average concentrations
seasonal sum of all hourly average concentrations^ 0.06 ppm
seasonal sum of all hourly average concentrations^ 0.07 ppm
seasonal sum of all hourly average concentrations^ 0.08 ppm
Sulfate Regional Experiment Program
solar zenith angle
tau, photochemical lifetime; atmospheric lifetime
time; duration of exposure
Mest statistical value; t statistic
T lymphocyte
thymus-dependent lymphocytes
triiodothyronine
thyroxine
IPCC Third Assessment Report
IPCC Third Assessment Report of Working Group I
tracheobronchial; terminal bronchioles; tuberculosis
thiobarbituric acid
thiobarbituric acid reactive substances
total carbon
Technetium-99m
99 mTc-diethylenetriaminepentaacetic acid
Thymus-derived lymphocytes
core temperature
cytotoxic T-lymphocytes
Tunable Diode Laser Absorption Spectrometer
expiratory time
transmission electron microscopy; Terrestrial Ecosystem Model
Texas Air Quality Field Study
teragram(s)
transforming growth factor
P transforming growth factor
T helper type
March 2011
Ixv
DRAFT - DO NOT CITE OR QUOTE
-------
tHcy
Ti
Ti
TIA
TiO2
TLC
TLNISE
TLR
TMPC
TNC
TNF
TNFR
TNF-a
TOMS
TOPSE
TPLIF
TREGRO
TRIFFID
TRIM
TRIM.Expo
TSH
TSP
TTFMS
TVA
TWA
TX
U.K.
U.S.
U.S.C.; USC
UA
UAM
UCL
UDGT
UDP
ULLI
total homocysteine
inspiratory time
titanium
transient ischemic attack
titanium dioxide
total lung capacity
two-level normal independent sampling estimation
Toll-like receptor
tetramethylphrrolise 1-oxide
total nonstructural carbohydrate
tumor necrosis factor (e.g., TNF-a)
tumor necrosis factor receptor
tissue necrosis factor alpha
Total Ozone Mapping/Monitoring Satellite; total ozone mapping
spectrometer
Tropospheric Ozone Production About the Spring Equinox
two-photon laser-induced fluorescence
Tree Growth Model
Top-down Representation of Interactive Foliage and Flora
Including Dynamics
Total Risk Integrated Methodology (model)
Total Risk Integrated Methodology Exposure Event (model)
thyroid stimulating hormone
total suspended particles
two-tone frequency-modulated spectroscopy
Tennessee Valley Authority
time-weighted average
tromboxane (e.g., TXB2)
United Kingdom
United States of America
U.S. Code
uric acid
Urban Airshed Model
upper 95th% confidence limit
UDP -galactose-l,2,-diacylglycerol galactosyltransferase
uridine diphosphate
unit length labeling index
March 2011
Ixvi
DRAFT - DO NOT CITE OR QUOTE
-------
UN ECEICP- Vegetation
UNECE
UNEP
UNFCCC
URI
URS
URT
USDA
USFS
USGCRP
USGS
UT
UTC
UV
UV-A
UV-B
UV-C
UV-DIAL
V
V, mV, uV
VA
Val
vc
VCAM
Vd
VD
VE
VE
Vfimax
'max
Vmax25%
Vmax50%
VMD
Vn
United Nations Economic Commission for Europe International
Cooperative Programme on effects of air pollution and other
stresses on crops and non-woody plants (UN/ECE-Vegetation;
formerly -Crops)
United Nations Economic Commission for Europe
United Nations Environmental Programme
United Nations Framework Convention on Climate Change
upper respiratory infection
upper respiratory symptoms
upper respiratory tract; upper airways
U.S. Department of Agriculture
U.S. Forest Service
U.S. Global Change Research Program
U.S. Geological Survey
Universal Time
Coordinated Universal Time
ultraviolet radiation
ultraviolet radiation at wavelengths of 320 to 400 nm
ultraviolet radiation at wavelengths of 280 to 320 nm
ultraviolet radiation at wavelengths of 200 to 280 nm
Ultraviolet Differential Absorption Lidar
vanadium
volt, millivolt, microvolt
alveolar ventilation
valine
vital capacity
vascular cell adhesion molecule
deposition rate, deposition velocity (cm/s)
volume of the anatomic or physiological dead space
ventilatory volume
ventilation rate; minute ventilation
maximum minute ventilation
maximum velocity
maximum expiratory flow at 25% of the vital capacity
maximum expiratory flow at 50% of the vital capacity
maximum expiratory flow at 75% of the vital capacity
volume median diameter
Nasal volume
March 2011
Ixvii
DRAFT - DO NOT CITE OR QUOTE
-------
VO2
V02max
VOC(s)
VP
Vpso%
VPD
VT
VTB
Vlmax
VuA
vWF
W
W/mz, W m'z
w/v
W126
W95
WBC
WBGT
WCB
WF, WFM
WHI
WHO
wk
WMO
WMO/UNEP
Ws
WS
WT
wt%
yr
ZAPS
ZELIG
Zn
oxygen consumption
maximum volume per time, of oxygen (maximal oxygen
consumption, maximal oxygen uptake or aerobic capacity)
volatile organic compound(s)
volumetric penetration
volume at which 50% of an inhaled bolus is absorbed
vapor pressure deficit; Vehicles per day; Ventricular premature
depolarization
tidal volume
terminal bronchiole region volume
maximum tidal volume
volume of the upper airways
von Willebrand factor
width; wilderness
watts per square meter
weight per volume
cumulative integrated exposure index with a sigmoidal
weighting function
cumulative integrated exposure index with a sigmoidal
weighting function
white blood cell
wet bulb globe temperature
warm conveyor belt
White Face Mountain site
Women's Health Initiative
World Health Organization
week(s)
World Meteorological Organization
World Meteorological Organization/United Nations
Environment Program
Wassilewskija Arabidopsis ecotype
wood smoke
wild type; White Top Mountain site
percent by weight
year
Zonal Air Pollution System
a forest succession simulation model
zinc
March 2011
Ixviii
DRAFT - DO NOT CITE OR QUOTE
-------
Chapter 1. Introduction
1 The Integrated Science Assessment (ISA) is a concise evaluation and synthesis of the most
2 policy-relevant science for reviewing the national ambient air quality standards (NAAQS). Because
3 the ISA communicates critical science judgments relevant to the NAAQS review, it forms the
4 scientific foundation for the review of the NAAQS for ozone (O3). The existing primary O3 standard
5 includes an 8-hour (h) average (avg) standard set at 75 parts per billion (ppb) and not to be exceeded
6 more than once per year. The secondary standard for O3 is set equal to the primary standard.
7 The ISA accurately reflects "the latest scientific knowledge useful in indicating the kind and
8 extent of identifiable effects on public health which may be expected from the presence of [a]
9 pollutant in ambient air" (1990, 080701). Key information and judgments formerly contained in the
10 Air Quality Criteria Document (AQCD) for O3 are incorporated in this assessment. Additional
11 details of the pertinent scientific literature published since the last review, as well as selected older
12 studies of particular interest, are included. This ISA thus serves to update and revise the evaluation of
13 the scientific evidence available at the time of the completion of the 2006 O3 AQCD (U.S. EPA,
14 2006, 088089).
15 The draft Integrated Review Plan for the Ozone National Ambient Air Quality Standards (IRP)
16 (U.S. EPA, 2009, 684024} identified key policy-relevant questions that provided a framework for
17 this assessment of the scientific evidence. These questions frame the entire review of the NAAQS for
18 O3 and thus are informed by both science and policy considerations. The ISA organizes, presents,
19 and integrates the scientific evidence which is considered along with findings from risk analyses and
20 policy considerations to help the U.S. Environmental Protection Agency (EPA) address these
21 questions during the NAAQS review. In evaluating the health evidence, the focus of this assessment
22 is on scientific evidence that is most relevant to the following questions taken directly from the
23 Integrated Review Plan:
24 • To what extent has new scientific information become available that alters or
25 substantiates our understanding of the health effects associated with various time periods
26 of exposure to ambient O3, including short-term (1-3 hours), prolonged (6-8 hours), and
27 chronic (months to years) exposures?
28 • To what extent has new scientific information become available that alters or
29 substantiates our understanding of the health effects of O3 on at-risk populations,
30 including those with potentially increased susceptibility such as children and
31 disadvantaged populations?
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments such as the Integrated Science Assessments (ISA) and the Integrated Risk Information System (IRIS).
March 2011 1-1 DRAFT - DO NOT CITE OR QUOTE
-------
1 • To what extent has new scientific information become available that alters or
2 substantiates conclusions from previous reviews regarding the plausibility of adverse
3 health effects caused by O3 exposure?
4 "At what levels of O3 exposure are health effects observed? Is there evidence of effects at
5 exposure levels lower than those previously observed, and what are the important
6 uncertainties associated with that evidence? What is the nature of the exposure-response
7 relationships of O3 for the various health effects evaluated?
8 "To what extent has new scientific information become available that alters or
9 substantiates our understanding of non-O3-exposure factors that might influence the
10 associations between O3 levels and health effects being considered (e.g., weather-related
11 factors; behavioral factors such as heating/air conditioning use; driving patterns; and
12 time-activity patterns)?
13 "To what extent do risk and/or exposure analyses suggest that exposures of concern for
14 O3-related health effects are likely to occur with current ambient levels of O3 or with
15 levels that just meet the O3 standard? Are these risks/exposures of sufficient magnitude
16 such that the health effects might reasonably be judged to be important from a public
17 health perspective? What are the important uncertainties associated with these
18 risk/exposure estimates?
19 "To what extent have important uncertainties identified in the last rulemaking been
20 addressed and/or have new uncertainties emerged?
1.1. Legislative Requirements
21 Two sections of the Clean Air Act (CAA) (2011, 013410) govern the establishment and
22 revision of the NAAQS. Section 108 (42 USC §7408) directs the Administrator to identify and list
23 certain air pollutants and then to issue air quality criteria for those pollutants. The Administrator is to
24 list those air pollutants that in her "judgement; cause or contribute to air pollution which may
25 reasonably be anticipated to endanger public health or welfare" and whose "presence... in the
26 ambient air results from numerous or diverse moble or stationary sources."(1990, 080701). Air
27 quality criteria are intended to "accurately reflect the latest scientific knowledge useful in indicating
28 the kind and extent of identifiable effects on public health or welfare which may be expected from
29 the presence of [a] pollutant in ambient air ... (42 USC §7408(b)).
30 Section 109 (1990, 037658) directs the Administrator to propose and promulgate "primary"
31 and "secondary" NAAQS for pollutants for which air quality criteria have been issued. Section
March 2011 1 -2 DRAFT - DO NOT CITE OR QUOTE
-------
1 109(b)(l) defines a primary standard as one "the attainment and maintenance of which in the
2 judgment of the Administrator, based on such criteria and allowing an adequate margin of safety, are
3 requisite to protect the public health.".1 A secondary standard, as defined in section 109(b)(2), must
4 "specify a level of air quality the attainment and maintenance of which, in the judgment of the
5 Administrator, based on such criteria, is required to protect the public welfare from any known or
6 anticipated adverse effects associated with the presence of [the] pollutant in the ambient air.".2
7 The requirement that primary standards include an adequate margin of safety was intended to
8 address uncertainties associated with inconclusive scientific and technical information available at
9 the time of standard setting. It was also intended to provide a reasonable degree of protection against
10 hazards that research has not yet identified. See Lead Industries Association v. EPA, 647 F.2d 1130,
11 1154 (D.C. Cir 1980) (1980, 090977). cert, denied, 449 U.S. 1042 (1980); American Petroleum
12 Institute v. Costle, 665 F.2d 1176, 1186 (D.C. Cir. 1981) (1981, 090978). cert, denied, 455 U.S. 1034
13 (1982). Both kinds of uncertainties are components of the risk associated with pollution at levels
14 below those at which human health effects can be said to occur with reasonable scientific certainty.
15 Thus, in selecting primary standards that include an adequate margin of safety, the Administrator is
16 seeking not only to prevent pollution levels that have been demonstrated to be harmful but also to
17 prevent lower pollutant levels that may pose an unacceptable risk of harm, even if the risk is not
18 precisely identified as to nature or degree.
19 In selecting a margin of safety, the EPA considers such factors as the nature and severity of the
20 health effects involved, the size of the sensitive population(s) at risk, and the kind and degree of the
21 uncertainties that must be addressed. The selection of any particular approach to providing an
22 adequate margin of safety is a policy choice left specifically to the Administrator's judgment. See
23 Lead Industries Association v. EPA, supra, 647 F.2d at 1161-1162 (1980, 090977).
24 In setting standards that are "requisite" to protect public health and welfare, as provided in
25 Section 109(b), EPA's task is to establish standards that are neither more nor less stringent than
26 necessary. In so doing, EPA may not consider the costs of implementing the standards. See generally
27 Whitman v. American Trucking Associations, 531 U.S. 457, 465-472, 475-76 (2001, 043004).
28 Section 109(d)(l) requires that "not later than December 31, 1980, and at 5-year
29 intervals thereafter, the Administrator shall complete a thorough review of the criteria
30 published under section 108 and the national ambient air quality standards ... and shall make such
31 revisions in such criteria and standards and promulgate such new standards as may be
32 appropriate...". Section 109(d)(2) requires that an independent scientific review committee "shall
33 complete a review of the criteria ... and the national primary and secondary ambient air quality
34 standards ... and shall recommend to the Administrator any new . . . standards and revisions of
1 The legislative history of section 109 indicates that a primary standard is to be set at "the maximum permissible ambient air level . . .
which will protect the health of any [sensitive] group of the population," and that for this purpose "reference should be made to a
representative sample of persons comprising the sensitive group rather than to a single person in such a group" [S. Rep. No. 91-1196, 91S|
Cong., 2d Sess. 10 (1970)].
Welfare effects as defined in section 302(h) include, but are not limited to, "effects on soils, water, crops, vegetation, man-made
materials, animals, wildlife, weather, visibility and climate, damage to and deterioration of property, and hazards to transportation, as
well as effects on economic values and on personal comfort and well-being." (2005, 090976).
March 2011 1 -3 DRAFT - DO NOT CITE OR QUOTE
-------
1 existing criteria and standards as may be appropriate ...". Since the early 1980's, this independent
2 review function has been performed by CAS AC.
1.2. History of the NAAQS for Ozone
3 Tropospheric (ground-level) O3 is the indicator for the mix of photochemical oxidants (e.g.,
4 peroxyacetyl nitrate, hydrogen peroxide) formed from biogenic and anthropogenic precursor
5 emissions. Naturally occurring O3 in the troposphere can result from biogenic organic precursors
6 reacting with naturally occurring nitrogen oxides (NOX) and by stratospheric O3 intrusion into the
7 troposphere. Anthropogenic precursors of O3, especially NOX, and volatile organic compounds
8 (VOCs), originate from a wide variety of stationary and mobile sources. Ambient O3 concentrations
9 produced by these emissions are directly affected by temperature, solar radiation, wind speed, and
10 other meteorological factors.
11 NAAQS are comprised of four basic elements: indicator, averaging time, level, and form. The
12 indicator defines the pollutant to be measured in the ambient air for the purpose of determining
13 compliance with the standard. The averaging time defines the time period over which air quality
14 measurements are to be obtained and averaged or cumulated, considering evidence of effects
15 associated with various time periods of exposure. The level of a standard defines the air quality
16 concentration used (i.e., an ambient concentration of the indicator pollutant) in determining whether
17 the standard is achieved. The form of the standard specifies the air quality measurements that are to
18 be used for compliance purposes (e.g., the annual fourth-highest daily maximum 8-hour
19 concentration, averaged over 3 years), and whether the statistic is to be averaged across multiple
20 years. These four elements taken together determine the degree of public health and welfare
21 protection afforded by the NAAQ S.
Table 1-1. Summary of primary and secondary NAAQS promulgated for ozone during the period
1971-2008
Final Rule
1971 (36 FR 81 86)
(1971.039176)
1 979 (44 FR 8202)
(1979.039177)
1 993 (58 FR 13008)
(1993.043977)
1 997 (62 FR 38856)
(1997.083356)
2008 (73 FR 16483)
(2008, 684051)
Indicator Avg Time
Total
photochemical 1-h
oxidants
03 1-h
Level (ppm)
0.08
0.12
Form
Not to be exceeded more than 1 hour per year
Attainment is defined when the expected number of
days per calendar year, with maximum hourly average
concentration greater than 0.12 ppm, is< 1
EPA decided that revisions to the standards were not warranted at the time.
03 8-h
03 8-h
0.08
0.075
Annual fourth-highest daily maximum 8-h concentration
averaged over 3 years
Form of the standards remained unchanged relative to
the 1 997 standard
March 2011 1 -4 DRAFT - DO NOT CITE OR QUOTE
-------
1 Table 1-1 summarizes the O3 NAAQS that have been promulgated to date. In each review, the
2 secondary standard has been set to be identical to the primary standard. These reviews are briefly
3 described below.
4 EPA first established primary and secondary NAAQS for photochemical oxidants in 1971
5 (U.S. EPA, 1971, 039176). Both primary and secondary standards were set at a level of 0.08 parts
6 per million (ppm), 1-h avg, total photochemical oxidants, not to be exceeded more than 1 hour per
7 year. The standards were based on scientific information contained in the 1970 AQCD (U.S. EPA,
8 1970.014681).
9 In 1977, EPA announced the first periodic review of the 1970 AQCD (U.S. EPA, 1970,
10 014681) in accordance with Section 109(d)(l) of the Clean Air Act. In 1978, EPA published an
11 AQCD (U.S. EPA, 1978, 040586). Based on the 1978 AQCD, EPA published proposed revisions to
12 the original NAAQS in (1978, 684050) and final revisions in 1979 (1979, 039177). The level of the
13 primary and secondary standards was revised from 0.08 to 0.12 ppm; the indicator was revised from
14 photochemical oxidants to O3; and the form of the standards was revised from a deterministic to a
15 statistical form, which defined attainment of the standards as occurring when the expected number of
16 days per calendar year with maximum hourly average concentration greater than 0.12 ppm is equal
17 to or less than one.
18 In 1982 (Air quality criteria document for ozone and other photochemical oxidants, 1982,
19 043971). EPA announced plans to revise the 1978 AQCD (U.S. EPA, 1978, 040586). In 1983, EPA
20 announced that the second periodic review of the primary and secondary standards for O3 had been
21 initiated (Review of the national ambient air quality standards for ozone, 1983, 043972). EPA
22 subsequently published the 1986 O3 AQCD (U.S. EPA, 1986, 017607) and 1989 Staff Paper
23 (U.S. EPA, 1989, 041919). Following publication of the 1986 O3 AQCD, a number of scientific
24 abstracts and articles were published that appeared to be of sufficient importance concerning
25 potential health and welfare effects of O3 to warrant preparation of a Supplement to the 1986 O3
26 AQCD (U.S. EPA, 1992, 042599). Under the terms of a court order, on August 10, 1992, EPA
27 published a proposed decision (1992, 043976) stating that revisions to the existing primary and
28 secondary standards were not appropriate at the time (1992, 043976). This notice explained that the
29 proposed decision would complete EPA's review of information on health and welfare effects of O3
30 assembled over a 7-year period and contained in the 1986 O3 AQCD (U.S. EPA, 1986, 017607) and
31 its Supplement to the 1986 O3 AQCD (U.S. EPA, 1992, 042599). The proposal also announced
32 EPA's intention to proceed as rapidly as possible with the next review of the air quality criteria and
33 standards for O3 in light of emerging evidence of health effects related to 6- to 8-hour O3 exposures.
34 On March 9, 1993, EPA concluded the review by deciding that revisions to the standards were not
35 warranted at that time (1993, 043977).
36 In August 1992, EPA announced plans to initiate the third periodic review of the air quality
37 criteria and O3 NAAQS (1992, 043976). On the basis of the scientific evidence contained in the
38 1996 O3 AQCD (U.S. EPA, 1996, 017831) and the 1996 Staff Paper (U.S. EPA, 1996, 039046). and
39 related technical support documents, linking exposures to ambient O3 to adverse health and welfare
March 2011 1 -5 DRAFT - DO NOT CITE OR QUOTE
-------
1 effects at levels allowed by the then existing standards, EPA proposed to revise the primary and
2 secondary O3 standards on December 13, 1996 (U.S. EPA, 1996, 031951). The EPA proposed to
3 replace the then existing 1-hour primary and secondary standards with 8-h avg O3 standards set at a
4 level of 0.08 ppm (equivalent to 0.084 ppm using standard rounding conventions). The EPA also
5 proposed, in the alternative, to establish a new distinct secondary standard using a biologically based
6 cumulative seasonal form. The EPA completed the review on July 18, 1997 by setting the primary
7 standard at a level of 0.08 ppm, based on the annual fourth-highest daily maximum 8-h avg
8 concentration, averaged over 3 years, and setting the secondary standard identical to the revised
9 primary standard (U.S. EPA, 1997, 083356V
10 On May 14, 1999, in response to challenges to EPA's 1997 decision by industry and others, the
11 U.S. Court of Appeals for the District of Columbia Circuit (D.C. Circuit Court) remanded the O3
12 NAAQS to EPA, finding that Section 109 of the CAA, as interpreted by EPA, effected an
13 unconstitutional delegation of legislative authority. In addition, the D.C. Circuit Court directed that,
14 in responding to the remand, EPA should consider the potential beneficial health effects of O3
15 pollution in shielding the public from the effects of solar ultraviolet (UV) radiation, as well as
16 adverse health effects. On January 27, 2000, EPA petitioned the U.S. Supreme Court for certiorari on
17 the constitutional issue (and two other issues) but did not request review of the D.C. Circuit Court
18 ruling regarding the potential beneficial health effects of O3. On February 27, 2001, the U.S.
19 Supreme Court unanimously reversed the judgment of the D.C. Circuit Court on the constitutional
20 issue, holding that Section 109 of the CAA does not delegate legislative power to the EPA in
21 contravention of the Constitution, and remanded the case to the D.C. Circuit Court to consider
22 challenges to the O3 NAAQS that had not been addressed by that Court's earlier decisions. On
23 March 26, 2002, the D.C. Circuit Court issued its final decision, finding the 1997 O3 NAAQS to be
24 "neither arbitrary nor capricious," and denied the remaining petitions for review. In response to the
25 D.C. Circuit Court remand to consider the potential beneficial health effects of O3 pollution in
26 shielding the public from effects of solar (UV) radiation, on November 14, 2001, EPA proposed to
27 leave the 1997 8-h O3 NAAQS unchanged (2001, 684026). After considering public comment on the
28 proposed decision, EPA published its final response to this remand on January 6, 2003, reaffirming
29 the 8-h O3 NAAQS set in 1997 (National ambient air quality standards for ozone: Final response to
30 remand; final rule, 2003, 015702). Finally, on April 30, 2004, EPA announced the decision to make
31 the 1-h O3 NAAQS no longer applicable to areas 1 year after the effective date of the designation of
32 those areas for the 8-h NAAQS (2004, 684054). For most areas, the date that the 1-h NAAQS no
33 longer applied was June 15, 2005.
34 EPA initiated the next periodic review if the air quality criteria and O3 standards in September
35 2000 with a call for information (2000, 080678). The schedule for completion of that rulemaking
36 later became governed by a consent decree resolving a lawsuit filed in March 2003 by a group of
37 plaintiffs representing national environmental and public health organizations. Based on the 2006 O3
38 AQCD (U.S. EPA, 2006, 088089) published in March 2006, and the Staff Paper (U.S. EPA, 2007,
39 090207) and related technical support documents, the proposed decision was published in the
March 2011 1 -6 DRAFT - DO NOT CITE OR QUOTE
-------
1 Federal Register on July 11, 2007 (2007, 684055). The EPA proposed to revise the level of the
2 primary standard to a level within the range of 0.075 to 0.070 ppm. Two options were proposed for
3 the secondary standard: (1) replacing the current standard with a cumulative, seasonal standard,
4 expressed as an index of the annual sum of weighted hourly concentrations cumulated over
5 12 daylight hours during the consecutive 3-month period within the O3 season with the maximum
6 index value, set at a level within the range of 7 to 21 ppm-h; and (2) setting the secondary standard
7 identical to the revised primary standard. The EPA completed the rulemaking with publication of a
8 final decision on March 27, 2008 (2008, 684051). revising the level of the 8-hour primary O3
9 standard from 0.08 ppm to 0.075 ppm and revising the secondary standard to be identical to the
10 primary standard.
11 On September 16, 2009, the EPA Administrator announced her decision to reconsider the
12 March 2008 decisions on revisions to the primary and secondary O3 NAAQS.
1.3. ISADevelopment
13 EPA initiated the current review of the NAAQS for O3 on September 29, 2008, with a call for
14 information from the public (2008, 684057). In addition to the call for information, publications
15 were identified through an ongoing literature search process that includes extensive computer
16 database mining on specific topics. Literature searches were conducted routinely to identify studies
17 published since the last review, focusing on publications since 2005. Search strategies were
18 iteratively modified to optimize identification of pertinent publications. Additional papers are
19 identified for inclusion in several ways: review of pre-publication tables of contents for journals in
20 which relevant papers may be published; independent identification of relevant literature by expert
21 authors; and identification by the public and CAS AC during the external review process.
22 Publications considered for inclusion in the ISA were added to the Health and Environmental
23 Research Online (HERO) database recently developed by EPA (http://hero.epa.gov/): note that the
24 references in the ISA include a HERO ID that provides a link to the database. All references that are
25 considered for inclusion in each chapter, organized by discipline, will be found through the HERO
26 links provided at the beginning of the individual chapter reference sections. The HERO link provides
27 the list of references that are included, as well as those that are considered and not included in the
28 ISA, with bibliographic information and abstracts.
29 Typically, only information that had undergone scientific peer review and had been published
30 or accepted for publication was considered for inclusion, along with analyses conducted by EPA
31 using publicly available data. This review has attempted to evaluate all relevant data published since
32 the last review pertaining to the atmospheric science of O3, human exposure to ambient O3, and
33 epidemiologic, controlled human exposure, and animal toxicological studies on O3, including those
34 related to exposure-response relationships, mode(s) of action (MOA), or susceptible populations, and
35 literature on the ecological or welfare effects of ambient O3. Added to the body of research on O3
March 2011 1 -7 DRAFT - DO NOT CITE OR QUOTE
-------
1 effects were EPA's analyses of air quality and emissions data, studies on atmospheric chemistry,
2 transport, and fate of these emissions, as well as issues related to exposure to O3.
3 In general, in assessing the scientific quality and relevance of health and environmental effects
4 studies, the following considerations have been taken into account when selecting studies for
5 inclusion in the ISA. The selection process for studies included in this ISA is shown in Figure 1-1.
6 • Are the study populations, subjects, or animal models adequately selected, and are they
7 sufficiently well defined to allow for meaningful comparisons between study or exposure
8 groups?
9 • Are the statistical analyses appropriate, properly performed, and properly interpreted?
10 Are likely covariates adequately controlled or taken into account in the study design and
11 statistical analysis?
12 • Are the air quality data, exposure, or dose metrics of adequate quality and sufficiently
13 representative of information regarding ambient O3?
14 • Are the health, ecological or welfare effect measurements meaningful and reliable?
15 In selecting epidemiologic studies, EPA considered whether a given study presented
16 information on associations with short- or long-term O3 exposures at or near ambient levels of O3;
17 considered approaches to evaluate issues related to potential confounding by other pollutants;
18 assessed potential effect modifiers; addressed health endpoints and populations not previously
19 extensively researched; and evaluated important methodological issues (e.g., lag or time period
20 between exposure and effects, model specifications, thresholds, mortality displacement) related to
21 interpretation of the health evidence. Among the epidemiologic studies selected, particular emphasis
22 was placed on those studies most relevant to the review of the NAAQS. Specifically, studies
23 conducted in the United States (U.S.) or Canada were discussed in more detail than those from other
24 geographical regions. Particular emphasis was placed on: (1) recent multicity studies that employ
25 standardized analysis methods for evaluating effects of O3 and that provide overall estimates for
26 effects, based on combined analyses of information pooled across multiple cities; (2) studies that
27 help understand quantitative relationships between exposure concentrations and effects; (3) new
28 studies that provide evidence on effects in susceptible populations; and (4) studies that consider and
29 report O3 as a component of a complex mixture of air pollutants.
March 2011 1 -8 DRAFT - DO NOT CITE OR QUOTE
-------
Informative
studies
are identified
KEY DEFINITIONS
INFORMATIVE studies are well-designed,
properly implemented, thoroughly described,
HIGHLY INFORMATIVE studies reduce
uncertainty on critical issues, may include
analyses of confounding or effect modification
by copollutants or other variables, analyses of
concentration-response or dose-response
relationships, analyses related to time
between exposure and response, and offer
innovation in method or design.
POLICY-RELEVANT studies may include
those conducted at or near ambient concen-
trations and studies conducted in U.S. and
Canadian airsheds.
V J
ISA
Policy relevant and highly informative studies discussed in the ISA text include
those that provide a basis for or describe the association between the criteria
pollutant and effects. Studies summarized in tables and figures are included
because they are sufficiently comparable to be displayed together. A study
highlighted in the ISA text does not necessarily appear in a summary table or
figure.
Studies are
evaluated for
inclusion
in the ISA
Figure 1-1. Identification of studies for inclusion in the ISA.
1 Criteria for the selection of research evaluating controlled human exposure or animal
2 toxicological studies included a focus on studies conducted using relevant pollutant exposures. For
3 both types of studies, relevant pollutant exposures are considered to be those generally within one or
4 two orders of magnitude of ambient O3 concentrations. Studies in which higher doses were used may
5 also be considered if they provide information relevant to understanding modes of action or
6 mechanisms, as noted below.
7 Evaluation of controlled human exposure studies focused on those that approximated expected
8 human exposure conditions in terms of concentration and duration. In the selection of controlled
9 human exposure studies, emphasis is placed on studies that: (1) investigate potentially susceptible
March 2011
1-9
DRAFT - DO NOT CITE OR QUOTE
-------
1 populations such as people with cardiovascular diseases; (2) address issues such as concentration-
2 response or time-course of responses; (3) include control exposures to filtered air; and (4) have
3 sufficient statistical power to assess findings.
4 Review of the animal toxicological evidence focused on studies that approximate expected
5 human dose conditions, which will vary depending on the toxicokinetics and biological sensitivity of
6 the particular laboratory animal species or strains studied. Due to resource constraints on exposure
7 duration and numbers of animals tested, animal studies typically utilize high-concentration
8 exposures to acquire data relating to mechanisms and assure a measurable response. Such studies
9 were considered to the extent that they provided useful information to inform our understanding of
10 interspecies differences and potential sensitivity differences between healthy and susceptible human
11 populations.
12 Evaluation of the ecological impact of O3 focused on studies of vegetation and effects on
13 ecosystems that occur in the U.S. and report endpoints or processes most relevant to the review of
14 the secondary standard. Many studies have been published about vegetation and ecosystems outside
15 of U.S. and North America, largely in Europe and Asia. This document includes discussion of studies
16 of vegetation and ecosystems outside of North America if those studies contribute to the
17 understanding of O3 effects across species and ecosystems. For example, studies outside North
18 America that deal with physiological and biochemical processes that contribute to the effects of O3
19 across species are discussed. Also, ecosystem studies outside of North America that contribute to the
20 understanding of O3 effects on ecosystem processes are discussed. In addition, the specific role that
21 tropospheric O3 plays in the earth's radiation budget and how perturbations in tropospheric O3 might
22 affect (1) climate through its role as a greenhouse gas and (2) health, ecology and welfare through its
23 role in shielding ultraviolet radiation are assessed.
24 These criteria provide benchmarks for evaluating various studies and for focusing on the
25 policy-relevant studies in assessing the body of health and welfare effects evidence. Detailed critical
26 analysis of all O3 health, ecological and welfare effects studies, especially in relation to the above
27 considerations, is beyond the scope of this document. Of most relevance for evaluation of studies is
28 whether they provide useful qualitative or quantitative information on exposure-effect or
29 exposure-response relationships for effects associated with current ambient air concentrations of O3
30 that can inform decisions on whether to retain or revise the standards.
31 In developing the O3 ISA, EPA began by reviewing and summarizing the evidence on
32 atmospheric sciences and exposure and the health effects evidence from in vivo and in vitro
33 toxicological studies, controlled human exposure studies, and epidemiologic studies. In August 2010,
34 EPA held a public workshop, in which EPA experts and several non-EPA experts were asked to
35 review the scientific content of preliminary draft materials for the draft ISA. The purpose of the
36 initial review workshop was to ensure that the ISA is up to date and focused on the most policy-
37 relevant findings, and to assist EPA with integration of evidence within and across disciplines.
38 Subsequently, EPA addressed comments and completed the initial integration and synthesis of the
39 evidence.
March 2011 1-10 DRAFT - DO NOT CITE OR QUOTE
-------
1 The integration of evidence on health, and ecological or welfare effects, involves collaboration
2 between scientists from various disciplines. As described in the section below, the ISA organization
3 is based on health and ecological effect categories. As an example, an evaluation of health effects
4 evidence would include summaries of findings from epidemiologic, controlled human exposure, and
5 toxicological studies, and integration of the results to draw conclusions - based on the causal
6 framework described below. Using the causal framework described in Section 1.6, EPA scientists
7 consider aspects such as strength, consistency, coherence, and biological plausibility of the evidence,
8 and develop draft causality judgments on the nature of the relationships. The draft integrative
9 synthesis sections and conclusions are reviewed by EPA internal experts and, as appropriate, by
10 outside expert authors. In practice, causality determinations often entail an iterative process of
11 review and evaluation of the evidence. The draft ISA is released for review by the CASAC and the
12 public, and comments received on the characterization of the science as well as the implementation
13 of the causal framework are carefully considered in revising and completing the ISA.
1.4. Document Organization
14 The ISA is composed of 10 chapters. This introductory chapter presents background
15 information and provides an overview of EPA's framework for making causal judgments. Chapter 2
16 is an integrated summary of key findings and conclusions regarding the source to dose paradigm,
17 MO A, important health effects of O3, including respiratory, cardiovascular, nervous system,
18 perinatal/developmental, and mortality outcomes, and ecological or welfare effects. Chapter 3
19 highlights key concepts and evidence relevant to understanding the sources, ambient concentrations,
20 and atmospheric behavior of ambient O3. Chapter 4 evaluates the evidence on human exposure to
21 ambient O3. Chapter 5 describes the dosimetry of O3 as well as a discussion of the MO A of O3.
22 Chapter 6 reviews, evaluates and integrates epidemiologic, human clinical, and animal toxicological
23 information on health effects related to short-term exposures (i.e., hours, days, or weeks) to O3,
24 including respiratory effects, cardiovascular and systemic effects, central nervous system (CNS)
25 effects, effects on the liver and cutaneous/ocular tissues, and mortality. Chapter 7 is similar to
26 Chapter 6, but focuses on health effects related to long-term exposures (i.e., months or years) to O3.
27 Chapter 8 summarizes the evidence on potentially susceptible populations for health effects of O3
28 exposure. Chapter 9 reviews, evaluates and integrates evidence on the ecological effects of ambient
29 O3, including crop and forest productivity, visible foliar injury, and ecosystem processes. Chapter 10
30 presents relevant evidence on the welfare effects of O3, focusing primarily on the effects on climate
31 and UV exposure. As noted above, these discussions focus on the most policy-relevant studies, and
32 the broader body of literature considered is included in HERO; and additional HERO links are
33 provided at the beginning of the individual chapter reference sections.
March 2011 1-11 DRAFT-DO NOT CITE OR QUOTE
-------
1.5. Document Scope
1 For the current review of the primary O3 standard, relevant scientific information on human
2 exposures and health effects associated with exposure to ambient O3 has been assessed. Previous
3 reviews have included an extensive body of evidence from all three major health disciplines -
4 toxicology, controlled human exposure studies and epidemiology - on the health effects of O3
5 exposure (U.S. EPA, 2006, 088089). In this ISA, the conclusions from previous reviews are
6 summarized at the beginning of each health outcome discussion to provide the foundation for
7 consideration of evidence from recent studies. Results of key studies from previous reviews are
8 included in discussions or tables and figures, as appropriate, and conclusions are drawn based on the
9 synthesis of evidence from recent studies with the extensive literature summarized in previous
10 reviews.
11 The ISA also includes the assessment of scientific information associated with known or
12 anticipated ecological and public welfare effects that is relevant to the review of the secondary O3
13 standard. Research on the ecological effects of O3, including impacts on vegetation, have been
14 discussed extensively in previous AQCDs (U.S. EPA, 2006, 088089). The welfare effects of O3,
15 particularly focusing on climate forcing effects and shielding of UV light, are discussed. The current
16 document incorporates findings of recent studies, building upon previous evaluations and
17 conclusions.
1.6. EPAFramework for Causal Determination
18 The EPA has developed a consistent and transparent basis to evaluate the causal nature of air
19 pollution-induced health or environmental effects. The framework described below establishes
20 uniform language concerning causality and brings more specificity to the findings. This standardized
21 language was drawn from across the federal government and wider scientific community, especially
22 from the recent National Academy of Sciences (NAS) Institute of Medicine (IOM) document,
23 Improving the Presumptive Disability Decision-Making Process for Veterans, (2008, 156586) the
24 most recent comprehensive work on evaluating causality.
25 The introductory portion of this section focuses on the evaluation of health effects evidence.
26 While focusing on human health outcomes, the concepts are also generally relevant to causality
27 determination for welfare effects. This section:
28 • describes the kinds of scientific evidence used in establishing a general causal
29 relationship between exposure and health effects;
30 • defines cause, in contrast to statistical association;
31 • discusses the sources of evidence necessary to reach a conclusion about the existence of
32 a causal relationship;
March 2011 1-12 DRAFT - DO NOT CITE OR QUOTE
-------
1 • highlights the issue of multifactorial causation;
2 • identifies issues and approaches related to uncertainty; and
3 • provides a framework for classifying and characterizing the weight of evidence in
4 support of a general causal relationship.
5 Approaches to assessing the separate and combined lines of evidence (e.g., epidemiologic,
6 human clinical, and animal toxicological studies) have been formulated by a number of regulatory
7 and science agencies, including the IOM of the NAS (2008, 156586). International Agency for
8 Research on Cancer (2006, 093206). EPA Guidelines for Carcinogen Risk Assessment (2005,
9 086237). and Centers for Disease Control and Prevention (2004, 056384). These formalized
10 approaches offer guidance for assessing causality. The frameworks are similar in nature, although
11 adapted to different purposes, and have proven effective in providing a uniform structure and
12 language for causal determinations. Moreover, these frameworks have supported decision-making
13 under conditions of uncertainty.
1.6.1. Scientific Evidence Used in Establishing Causality
14 Causality determinations are based on the evaluation and synthesis of evidence from across
15 scientific disciplines; the type of evidence that is most important for such determinations will vary
16 by pollutant or assessment. The most compelling evidence of a causal relationship between pollutant
17 exposures and human health effects comes from human clinical studies. This type of study
18 experimentally evaluates the health effects of administered exposures in human volunteers under
19 highly controlled laboratory conditions.
20 In epidemiologic or observational studies of humans, the investigator does not control
21 exposures or intervene with the study population. Broadly, observational studies can describe
22 associations between exposures and effects. These studies fall into several categories:
23 cross-sectional, prospective cohort, and time-series studies. "Natural experiments" offer the
24 opportunity to investigate changes in health with a change in exposure; these include comparisons of
25 health effects before and after a change in population exposures, such as closure of a pollution
26 source.
27 Experimental animal data can help characterize effects of concern, exposure-response
28 relationships, susceptible populations and MOAs. In the absence of controlled human exposure or
29 epidemiologic data, animal data alone may be sufficient to support a likely causal determination,
30 assuming that humans respond similarly to the experimental species.
March 2011 1-13 DRAFT - DO NOT CITE OR QUOTE
-------
1.6.2. Association and Causation
1 "Cause" is a significant, effectual relationship between an agent and an effect on health or
2 public welfare. "Association" is the statistical dependence among events, characteristics, or other
3 variables. An association is prima facie evidence for causation; alone, however, it is insufficient
4 proof of a causal relationship between exposure and disease. Unlike an association, a causal claim
5 supports the creation of counterfactual claims; that is, a claim about what the world would have been
6 like under different or changed circumstances (IOM, 2008, 156586). Much of the newly available
7 health information evaluated in this ISA comes from epidemiologic studies that report a statistical
8 association between ambient exposure and health outcome.
9 Many of the health and environmental outcomes reported in these studies have complex
10 etiologies. Diseases such as asthma, coronary heart disease (CHD) or cancer are typically initiated
11 by multiple agents. Outcomes depend on a variety of factors, such as age, genetic susceptibility,
12 nutritional status, immune competence, and social factors (Gee and Payne-Sturges, 2004, 093070;
13 IOM, 2008, 156586). Effects on ecosystems are often also multifactorial with a complex web of
14 causation. Further, exposure to a combination of agents could cause synergistic or antagonistic
15 effects. Thus, the observed risk represents the net effect of many actions and counteractions.
1.6.3. Evaluating Evidence for Inferring Causation
16 Moving from association to causation involves the elimination of alternative explanations for
17 the association. In estimating the causal influence of an exposure on health or environmental effects,
18 it is recognized that scientific findings incorporate uncertainty. "Uncertainty" can be defined as a
19 state of having limited knowledge where it is impossible to exactly describe an existing state or
20 future outcome, e.g., the lack of knowledge about the correct value for a specific measure or
21 estimate. Uncertainty characterization and uncertainty assessment are two activities that lead to
22 different degrees of sophistication in describing uncertainty. Uncertainty characterization generally
23 involves a qualitative discussion of the thought processes that lead to the selection and rejection of
24 specific data, estimates, scenarios, etc. Uncertainty assessment is more quantitative. The process
25 begins with simpler measures (e.g., ranges) and simpler analytical techniques and progresses, to the
26 extent needed to support the decision for which the assessment is conducted, to more complex
27 measures and techniques. Data may not be available for all aspects of an assessment and those data
28 that are available may be of questionable or unknown quality. In these situations, evaluation of
29 uncertainty can include professional judgment or inferences based on analogy with similar situations.
30 The net result is that the assessment will be based on a number of assumptions with varying degrees
31 of uncertainty. Uncertainties commonly encountered in evaluating health evidence for the criteria air
32 pollutants are outlined below for epidemiologic and experimental studies. Various approaches to
33 evaluating uncertainty include classical statistical methods, sensitivity analysis, or probabilistic
34 uncertainty analysis, in order of increasing complexity and data requirements. The ISA generally
March 2011 1-14 DRAFT - DO NOT CITE OR QUOTE
-------
1 evaluates uncertainties qualitatively in assessing the evidence from across studies; in some situations
2 quantitative analysis approaches, such as metaregression, may be used.
3 Meta-analysis may be a valuable tool for evaluating evidence by combining results from a
4 body of studies. Blair et al. (1995, 079190) observed that meta-analysis can enhance understanding
5 of associations between exposures and effects that are not readily apparent in examination of
6 individual study results and can be particularly useful for formally examining sources of
7 heterogeneity. However, these authors noted that meta-analysis may not be useful when the
8 relationship between the exposure and outcome is obvious, when only a few studies are available for
9 a particular exposure-outcome relationship, where there is limited access to data of sufficient quality,
10 or where there is substantial variation in study design or population. In addition, important
11 differences in effect estimates, exposure metrics, or other factors may limit or even preclude
12 quantitative statistical combination of multiple studies.
13 Controlled human exposure studies evaluate the effects of exposures to a variety of pollutants
14 in a highly controlled laboratory setting. Also referred to as human clinical studies, these
15 experiments allow investigators to expose subjects to known concentrations of air pollutants under
16 carefully regulated environmental conditions and activity levels. In some instances, controlled
17 human exposure studies can also be used to characterize concentration-response relationships at
18 pollutant concentrations relevant to ambient conditions. Controlled human exposures are typically
19 conducted using a randomized crossover design, with subjects exposed both to O3 and a clean air
20 control. In this way, subjects serve as their own controls, effectively controlling for many potential
21 confounders. However, human clinical studies are limited by a number of factors, including a small
22 sample size and short exposure times. The repetitive nature of ambient O3 exposures at levels that
23 can vary widely may lead to cumulative health effects, but this type of exposure is not practical to
24 replicate in a laboratory setting. In addition, although subjects do serve as their own controls,
25 personal exposure to pollutants in the hours and days preceding the controlled exposures may vary
26 significantly between and within individuals. Finally, human clinical studies require investigators to
27 adhere to stringent health criteria for a subject to be included in the study, and therefore the results
28 cannot necessarily be generalized to an entire population. Although some human clinical studies
29 have included health-compromised individuals such as those with respiratory or cardiovascular
30 disease, these individuals must also be relatively healthy and do not represent the most sensitive
31 individuals in the population. Thus, a lack of observation of effects from human clinical studies does
32 not necessarily mean that a causal relationship does not exist. While human clinical studies provide
33 important information on the biological plausibility of associations observed between air pollutant
34 exposure and health outcomes in epidemiologic studies, observed effects in these studies may
35 underestimate the response in certain populations.
36 Epidemiologic studies provide important information on the associations between health
37 effects and exposure of human populations to ambient air pollution. In the evaluation of
38 epidemiologic evidence, one important consideration is potential confounding. Confounding is "... a
39 confusion of effects. Specifically, the apparent effect of the exposure of interest is distorted because
March 2011 1-15 DRAFT - DO NOT CITE OR QUOTE
-------
1 the effect of an extraneous factor is mistaken for or mixed with the actual exposure effect (which
2 may be null)" (Rothman and Greenland, 1998, 086599). One approach to remove spurious
3 associations due to possible confounders is to control for characteristics that may differ between
4 exposed and unexposed persons; this is frequently termed "adjustment." Scientific judgment is
5 needed regarding likely sources and magnitude of confounding, together with consideration of how
6 well the existing constellation of study designs, results, and analyses address this potential threat to
7 inferential validity.
8 One key consideration in this review is evaluation of the potential contribution of O3 to health
9 effects when it is a component of a complex air pollutant mixture. Reported O3 effect estimates in
10 epidemiologic studies may reflect independent O3 effects on health outcomes. Ambient O3 may also
11 be serving as an indicator of complex ambient air pollution mixtures, particularly the photochemical
12 oxidant mixture. Alternatively, co-pollutants may mediate the effects of O3, or O3 may influence the
13 toxicity of co-pollutants.
14 Another important consideration in the evaluation of epidemiologic evidence is effect
15 modification. "Effect-measure modification differs from confounding in several ways. The main
16 difference is that, whereas confounding is a bias that the investigator hopes to prevent or remove
17 from the effect estimate, effect-measure modification is a property of the effect under study ... In
18 epidemiologic analysis one tries to eliminate confounding but one tries to detect and estimate effect-
19 measure modification" (Rothman and Greenland, 1998, 086599). When a risk factor is a confounder,
20 it is the true cause of the association observed between the exposure and the outcome; when a risk
21 factor is an effect modifier, it changes the magnitude of the association between the exposure and the
22 outcome in stratified analyses. Examples of potential effect modifiers in some of the studies
23 evaluated in this ISA may include environmental variables, such as temperature or humidity,
24 individual risk factors, such as education, cigarette smoking status, age in a prospective cohort study,
25 and community factors, such as percent of population >65 years old. It is often possible to stratify
26 the relationship between health outcome and exposure by one or more of these potential effect
27 modifiers. For variables that modify the association, effect estimates in each stratum will be different
28 from one another and different from the overall estimate, indicating a different exposure-response
29 relationship may exist in populations represented by these variables. Effect modifiers may be
30 encountered (1) within single-city time-series studies; or (2) across cities in a two-stage hierarchical
31 model or meta-analysis.
32 Several statistical methods are available to detect and control for potential confounders, with
33 none of them being completely satisfactory. Multivariable regression models constitute one tool for
34 estimating the association between exposure and outcome after adjusting for characteristics of
35 participants that might confound the results. The use of multi-pollutant regression models has been
36 the prevailing approach for controlling potential confounding by co-pollutants in air pollution health
37 effects studies. Finding the likely causal pollutant from multi-pollutant regression models is made
38 difficult by the possibility that one or more air pollutants may be acting as a surrogate for an
39 unmeasured or poorly measured pollutant or for a particular mixture of pollutants. In addition, more
March 2011 1-16 DRAFT - DO NOT CITE OR QUOTE
-------
1 than one pollutant may exert similar health effects, resulting in independently observed associations
2 for multiple pollutants. For example, O3, NO2 and SO2 have each been linked to respiratory effects in
3 epidemiologic studies. Correlation between O3 concentrations and various co-pollutants makes it
4 difficult to quantitatively interpret associations between different pollutant exposures and health
5 effects. Thus, results of models that attempt to distinguish O3 effects from those of co-pollutants
6 must be interpreted with caution. The number and degree of diversity of covariates, as well as their
7 relevance to the potential confounders, remain matters of scientific judgment. Despite these
8 limitations, the use of multi-pollutant models is still the prevailing approach employed in most air
9 pollution epidemiologic studies and provides some insight into the potential for confounding or
10 interaction among pollutants.
11 Another way to adjust for potential confounding is through stratified analysis, i.e., examining
12 the association within homogeneous groups with respect to the confounding variable. The use of
13 stratified analyses has an additional benefit: it allows examination of effect modification through
14 comparison of the effect estimates across different groups. If investigators successfully measured
15 characteristics that distort the results, adjustment of these factors help separate a spurious from a true
16 causal association. Appropriate statistical adjustment for confounders requires identifying and
17 measuring all reasonably expected confounders. Deciding which variables to control for in a
18 statistical analysis of the association between exposure and disease or health outcome depends on
19 knowledge about possible mechanisms and the distributions of these factors in the population under
20 study. Identifying these mechanisms makes it possible to control for potential sources that may result
21 in a spurious association.
22 Adjustment for potential confounders can be influenced by differential exposure measurement
23 error. There are several components that contribute to exposure measurement error in epidemiologic
24 studies, including the difference between true and measured ambient concentrations, the difference
25 between average personal exposure to ambient pollutants and ambient concentrations at central
26 monitoring sites, and the use of average population exposure rather than individual exposure
27 estimates. Consideration of issues important for evaluation of exposure to ambient O3 include the
28 relationship between O3 measured at central site monitors with exposure to ambient O3 in indoor
29 environments, since penetration of O3 into buildings may be limited. Previous AQCDs have
30 examined the role of measurement error for non-reactive pollutants in time-series epidemiologic
31 studies using simulated data and mathematical analyses and suggested that transfer of effects from
32 the "causal" variable to the confounder would only occur under unusual circumstances (i.e., "true"
33 predictors having high positive or negative correlation; substantial measurement error; or extremely
34 negatively correlated measurement errors) (U.S. EPA, 2004, 056905).
35 Confidence that unmeasured confounders are not producing the findings is increased when
36 multiple studies are conducted in various settings using different subjects or exposures, each of
37 which might eliminate another source of confounding from consideration. Thus, multicity studies
38 which use a consistent method to analyze data from across locations with different levels of
March 2011 1-17 DRAFT - DO NOT CITE OR QUOTE
-------
1 covariates can provide insight on potential confounding in associations. Intervention studies, because
2 of their quasi-experimental nature, can be particularly useful in characterizing causation.
3 In addition to clinical and epidemiologic studies, the tools of experimental biology have been
4 valuable for developing insights into human physiology and pathology. Laboratory tools have been
5 extended to explore the effects of putative toxicants on human health, especially through the study of
6 model systems in other species. These studies evaluate the effects of exposures to a variety of
7 pollutants in a highly controlled laboratory setting and allow exploration of MOAs or mechanisms
8 by which a pollutant may cause effects. Understanding the biological mechanisms underlying
9 various health outcomes can prove crucial in establishing or negating causality. There are, however,
10 uncertainties associated with quantitative extrapolations between laboratory animals and humans on
11 the pathophysiological effects of any pollutant. Animal species can differ from each other in
12 fundamental aspects of physiology and anatomy (e.g., metabolism, airway branching, hormonal
13 regulation) that may limit extrapolation.
14 Interpretations of experimental studies of air pollution effects in laboratory animals, as in the
15 case of environmental comparative toxicology studies, are affected by limitations associated with
16 extrapolation models. The differences between humans and other species with regard to pollutant
17 absorption and distribution profiles based on metabolism, hormonal regulation, breathing pattern,
18 exposure dose, and differences in lung structure and anatomy, all have to be taken into consideration.
19 Also, in spite of a high degree of homology and the existence of a high percentage of orthologous
20 genes across humans and rodents (particularly mice), extrapolation of molecular alterations at the
21 gene level is complicated by species-specific differences in transcriptional regulation. Given these
22 molecular differences, at this time there are uncertainties associated with quantitative extrapolations
23 between laboratory animals and humans of observed pollutant-induced pathophysiological
24 alterations under the control of widely varying biochemical, endocrine, and neuronal factors.
1.6.4. Application of Framework for Causal Determination
25 EPA uses a two-step approach to evaluate the scientific evidence on health or ecological
26 effects of criteria pollutants. The first step determines the weight of evidence in support of causation
27 and characterizes the strength of any resulting causal classification. The second step includes further
28 evaluation of the quantitative evidence regarding the concentration-response relationships and the
29 loads or levels, duration and pattern of exposures at which effects are observed.
March 2011 1-18 DRAFT - DO NOT CITE OR QUOTE
-------
Table 1-2. Aspects to aid in judging causality
Consistency of the
observed association
An inference of causality is strengthened when a pattern of elevated risks is observed across several
independent studies. The reproducibility of findings constitutes one of the strongest arguments for
causality. If there are discordant results among investigations, possible reasons such as differences in
exposure, confounding factors, and the power of the study are considered.
An inference of causality from epidemiologic associations may be strengthened by other lines of
evidence (e.g., clinical and animal studies) that support a cause-and-effect interpretation of the
association. Evidence on ecological or welfare effects may be drawn from a variety of experimental
Coherence approaches (e.g., greenhouse, laboratory, and field) and subdisciplines of ecology (e.g., community
ecology, biogeochemistry and paleological/historical reconstructions). The coherence of evidence from
various fields greatly adds to the strength of an inference of causality. The absence of other lines of
evidence, however, is not a reason to reject causality.
An inference of causality tends to be strengthened by consistency with data from experimental studies or
other sources demonstrating plausible biological mechanisms. A proposed mechanistic linking between
an effect and exposure to the agent is an important source of support for causality, especially when data
establishing the existence and functioning of those mechanistic links are available. A lack of biologic
understanding, however, is not a reason to reject causality.
A well-characterized exposure-response relationship (e.g., increasing effects associated with greater
exposure) strongly suggests cause and effect, especially when such relationships are also observed for
duration of exposure (e.g., increasing effects observed following longer exposure times). There are,
however, many possible reasons that a study may fail to detect an exposure-response relationship.
Thus, although the presence of a biologic gradient may support causality, the absence of an
exposure-response relationship does not exclude a causal relationship.
Biological plausibility.
Biological gradient
(exposure-response
relationship)
The finding of large, precise risks increases confidence that the association is not likely due to chance,
bias, or other factors. However, given a truly causal agent, a small magnitude in the effect could follow
from a lower level of exposure, a lower potency, or the prevalence of other agents causing similar
effects. While large effects support causality, modest effects therefore do not preclude it.
Strength of the observed
association
The strongest evidence for causality can be provided when a change in exposure brings about a change
in occurrence or frequency of health or welfare effects.
Evidence of a temporal sequence between the introduction of an agent, and appearance of the effect,
constitutes another argument in favor of causality.
As originally intended, this refers to increased inference of causality if one cause is associated with a
single effect or disease (Hill, 1965, 0716641 Based on our current understanding, this is now considered
one of the weaker guidelines for causality; for example, many agents cause respiratory disease and
respiratory disease has multiple causes. At the scale of ecosystems, as in epidemiology, complexity is
such that single agents causing single effects, and single effects following single causes, are extremely
unlikely. The ability to demonstrate specificity under certain conditions remains, however, a powerful
attribute of experimental studies. Thus, although the presence of specificity may support causality, its
absence does not exclude it.
Structure activity relationships and information on the agent's structural analogs can provide insight into
Analogy whether an association is causal. Similarly, information on mode of action for a chemical, as one of
many structural analogs, can inform decisions regarding likely causality.
Experimental evidence
Temporal relationship of
the observed association
Specificity of the
observed association
1
2
3
4
5
6
7
To aid judgment, various "aspects"1 of causality have been discussed by many philosophers
and scientists. The most widely cited aspects of causality in epidemiology, and public health, in
general, were articulated by Sir Austin Bradford Hill (1965, 071664) and have been widely used
(CDC, 2004, 056384: IARC, 2006, 093206: IOM, 2008, 156586: U.S. EPA, 2005, 086237).These
aspects (Hill, 1965, 071664) have been modified (Table 1-2) for use in causal determinations
specific to health and welfare effects or pollutant exposures (U.S. EPA, 2009, 179916).2 Some
aspects are more likely than others to be relevant for evaluating evidence on the health or ecological
effects of criteria air pollutants. For example, the "analogy" aspect does not always apply, especially
for the gaseous criteria pollutants, and specificity would not be expected for multi-etiologic health
1 The "aspects" described by Hill (1965, 071664) have become, in the subsequent literature, more commonly described as "criteria." The
original term "aspects" is used here to avoid confusion with 'criteria' as it is used, with different meaning, in the Clean Air Act.
2 The Hill aspects were developed for interpretation of epidemiologic results. They have been modified here for use with a broader array of
data, i.e., epidemiologic, controlled human exposure, ecological, and animal toxicological studies, as well as in vitro data, and to be more
consistent with EPA's Guidelines for Carcinogen Risk Assessment.
March 2011
1-19
DRAFT - DO NOT CITE OR QUOTE
-------
1 outcomes, such as asthma or cardiovascular disease, or ecological effects related to acidification.
2 Aspects that usually play a larger role in determination of causality are consistency of results across
3 studies, coherence of effects observed in different study types or disciplines, biological plausibility,
4 exposure-response relationship, and evidence from "natural" experiments.
5 Although these aspects provide a framework for assessing the evidence, they do not lend
6 themselves to being considered in terms of simple formulas or fixed rules of evidence leading to
7 conclusions about causality (Hill, 1965, 071664). For example, one cannot simply count the number
8 of studies reporting statistically significant results or statistically nonsignificant results and reach
9 credible conclusions about the relative weight of the evidence and the likelihood of causality. Rather,
10 these important considerations are taken into account with the goal of producing an objective
11 appraisal of the evidence, informed by peer and public comment and advice, which includes
12 weighing alternative views on controversial issues. In addition, it is important to note that the aspects
13 in Table 1-2 cannot be used as a strict checklist, but rather to determine the weight of the evidence
14 for inferring causality. In particular, not meeting one or more of the principles does not automatically
15 preclude a determination of causality (See discussion in CDC, 2004, 056384).
1.6.5. Determination of Causality
16 In the ISA, EPA assesses the results of recent relevant publications, building upon evidence
17 available during the previous NAAQS review, to draw conclusions on the causal relationships
18 between relevant pollutant exposures and health or environmental effects. This ISA uses a five-level
19 hierarchy that classifies the weight of evidence for causation, not just association1; that is, whether
20 the weight of scientific evidence makes causation at least as likely as not, in the judgment of the
21 reviewing group. In developing this hierarchy, EPA has drawn on the work of previous evaluations,
22 most prominently the lOM's Improving the Presumptive Disability Decision-Making Process for
23 Veterans (2008, 156586). EPA's Guidelines for Carcinogen Risk Assessment (2005, 086237). and the
24 U.S. Surgeon General's smoking report (CDC, 2004, 056384). In the ISA, EPA uses a series of five
25 descriptors to characterize the weight of evidence for causality. This weight of evidence evaluation is
26 based on various lines of evidence from across the health and environmental effects disciplines.
27 These separate judgments are integrated into a qualitative statement about the overall weight of the
28 evidence and causality. The five descriptors for causal determination are described in Table 1-3.
1 It should be noted that the CDC and IOM frameworks use a four-category hierarchy for the strength of the evidence. A five-level
hierarchy is used here to be consistent with the EPA Guidelines for Carcinogen Risk Assessment and to provide a more nuanced set of
categories.
March 2011 1 -20 DRAFT - DO NOT CITE OR QUOTE
-------
Table 1-3. Weight of evidence for causal determination
Health Effects
Ecological and Welfare Effects
Causal
relationship
Evidence is sufficient to conclude that there is a causal
relationship with relevant pollutant exposures. That is, the
pollutant has been shown to result in health effects in
studies in which chance, bias, and confounding could be
ruled out with reasonable confidence. For example: a)
controlled human exposure studies that demonstrate
consistent effects; or b) observational studies that cannot
be explained by plausible alternatives or are supported by
other lines of evidence (e.g., animal studies or mode of
action information). Evidence includes replicated and
consistent high-quality studies by multiple investigators.
Evidence is sufficient to conclude that there is a causal
relationship with relevant pollutant exposures. That is, the
pollutant has been shown to result in effects in studies in
which chance, bias, and confounding could be ruled out
with reasonable confidence. Controlled exposure studies
(laboratory or small- to medium-scale field studies)
provide the strongest evidence for causality, but the
scope of inference may be limited. Generally,
determination is based on multiple studies conducted by
multiple research groups, and evidence that is considered
sufficient to infer a causal relationship is usually obtained
from the joint consideration of many lines of evidence that
reinforce each other.
Evidence is sufficient to conclude that a causal
relationship is likely to exist with relevant pollutant
exposures, but important uncertainties remain. That is,
the pollutant has been shown to result in health effects in
studies in which chance and bias can be ruled out with
reasonable confidence but potential issues remain. For
Likely to be a example: a) observational studies show an association,
causal but co-pollutant exposures are difficult to address and/or
relationship other lines of evidence (controlled human exposure,
animal, or mode of action information) are limited or
inconsistent; or b) animal toxicological evidence from
multiple studies from different laboratories that
demonstrate effects, but limited or no human data are
available. Evidence generally includes replicated and
high-quality studies by multiple investigators.
Evidence is sufficient to conclude that there is a likely
causal association with relevant pollutant exposures. That
is, an association has been observed between the
pollutant and the outcome in studies in which chance,
bias and confounding are minimized, but uncertainties
remain. For example, field studies show a relationship,
but suspected interacting factors cannot be controlled,
and other lines of evidence are limited or inconsistent.
Generally, determination is based on multiple studies in
multiple research groups.
Suggestive of
a causal
relationship
Evidence is suggestive of a causal relationship with
relevant pollutant exposures, but is limited because
chance, bias and confounding cannot be ruled out. For
example, at least one high-quality epidemiologic study
shows an association with a given health outcome but the
results of other studies are inconsistent.
Evidence is suggestive of a causal relationship with
relevant pollutant exposures, but chance, bias and
confounding cannot be ruled out. For example, at least
one high-quality study shows an effect, but the results of
other studies are inconsistent.
Evidence is inadequate to determine that a causal
Inadequate to relationship exists with relevant pollutant exposures. The
infer a causal available studies are of insufficient quantity, quality,
relationship consistency or statistical power to permit a conclusion
regarding the presence or absence of an effect.
The available studies are of insufficient quality,
consistency or statistical power to permit a conclusion
regarding the presence or absence of an effect.
Not likely to
be a causal
relationship
Evidence is suggestive of no causal relationship with
relevant pollutant exposures. Several adequate studies,
covering the full range of levels of exposure that human
beings are known to encounter and considering
susceptible populations, are mutually consistent in not
showing an effect at any level of exposure.
Several adequate studies, examining relationships with
relevant exposures, are consistent in failing to show an
effect at any level of exposure.
1 For the O3 ISA, determination of causality involved the evaluation of evidence for different
2 types of health effects associated with short- and long-term exposure periods. In making
3 determinations of causality, evidence was evaluated for health outcome categories, such as
4 respiratory effects, and then conclusions were drawn based upon the integration of evidence from
5 across disciplines (e.g., epidemiology, clinical studies and toxicology) and also across the suite of
6 related individual health outcomes. To accomplish this integration, evidence from multiple and
7 various types of studies was considered. Response was evaluated over a range of observations which
8 was determined by the type of study, methods of exposure or dose, and response measurements.
9 Results from different protocols were compared and contrasted. EPA focuses on health outcome
10 categories, rather than very specific endpoints, since the coherence of evidence across a spectrum of
11 related endpoints (e.g., effects ranging from inflammatory effects to respiratory mortality) is an
12 important aspect for drawing conclusions regarding causality.
March 2011
1-21
DRAFT - DO NOT CITE OR QUOTE
-------
1 In drawing judgments regarding causality for the criteria air pollutants, EPA focuses on
2 evidence of effects at relevant pollutant exposures. To best inform reviews of the NAAQS, these
3 evaluations go beyond a determination of causality at any dose or concentration to emphasize the
4 relationship apparent at relevant pollutant exposures. Concentrations generally within an order of
5 magnitude or two of ambient pollutant measurements are considered to be relevant for this
6 determination. Building upon the determination of causality are questions relevant to quantifying
7 health or environmental risks based on our understanding of the quantitative relationships between
8 pollutant exposures and health or welfare effects. While the causality determination is based
9 primarily on evaluation of health or environmental effects evidence, EPA also evaluates evidence
10 related to the doses or levels at which effects are observed. Considerations relevant to evaluation of
11 quantitative relationships for health and environmental effects are summarized below.
1.6.5.1. Effects on Human Populations
12 Once a determination is made regarding the causal relationship between the pollutant and
13 outcome category, important questions regarding quantitative relationships include:
14 • What is the concentration-response, exposure-response, or dose-response relationship in
15 the human population?
16 • What is the interrelationship between incidence and severity of effect?
17 • What exposure conditions (dose or exposure, duration and pattern) are important?
18 • What populations appear to be differentially affected (i.e., more susceptible to effects)?
19 To address these questions, the entirety of policy-relevant quantitative evidence is evaluated to
20 best quantify those concentration-response relationships that exist. This requires evaluation of
21 pollutant concentrations and exposure durations at which effects were observed for exposed
22 populations, including potentially susceptible populations. This integration of evidence resulted in
23 identification of a study or set of studies that best approximated the concentration-response
24 relationships between health outcomes and O3, given the current state of knowledge and the
25 uncertainties that surrounded these estimates. To accomplish this, evidence is considered from
26 multiple and diverse types of studies. To the extent available, the ISA evaluates results from across
27 epidemiologic studies that use various methods to evaluate the form of relationships between O3 and
28 health outcomes and draws conclusions on the most well-supported shape of these relationships.
29 Animal data may also inform evaluation of concentration-response relationships, particularly relative
30 to MOAs and characteristics of susceptible populations. Controlled human exposure studies have
31 provided the strongest and most quantifiable exposure-response data on the human health effects of
32 O3. Chapter 2 presents the integrated findings informative for evaluation of population risks.
March 2011 1 -22 DRAFT - DO NOT CITE OR QUOTE
-------
1 An important consideration in characterizing the public health impacts associated with
2 exposure to a pollutant is whether the concentration-response relationship is linear across the full
3 concentration range encountered or if nonlinear relationships exist along any part of this range. Of
4 particular interest is the shape of the concentration-response curve at and below the level of the
5 current standards. The shape of the concentration-response curve varies, depending on the type of
6 health outcome, underlying biological mechanisms and dose. At the human population level,
7 however, various sources of variability and uncertainty, such as the low data density in the lower
8 concentration range, possible influence of exposure measurement error, and individual differences in
9 susceptibility to air pollution health effects, tend to smooth and "linearize" the concentration-
10 response function. In addition, many chemicals and agents may act by perturbing naturally occurring
11 background processes that lead to disease, which also linearizes population concentration-response
12 relationships (Clewell and Crump, 2005, 156359: Crump et al, 1976, 003192: Hoel, 1980, 156555V
13 These attributes of population dose-response may explain why the available human data at ambient
14 concentrations for some environmental pollutants (e.g., PM, O3, lead [Pb], environmental tobacco
15 smoke [ETS], radiation) do not exhibit evident thresholds for cancer or noncancer health effects,
16 even though likely mechanisms include nonlinear processes for some key events. These attributes of
17 human population dose-response relationships have been extensively discussed in the broader
18 epidemiologic literature (Rothman and Greenland, 1998, 086599).
19 Publication bias is a source of uncertainty regarding the magnitude of health risk estimates. It
20 is well understood that studies reporting non-null findings are more likely to be published than
21 reports of null findings, and publication bias can also result in overestimation of effect estimate sizes
22 (loannidis, 2008, 188317). For example, effect estimates from single-city epidemiologic studies have
23 been found to be generally larger than those from multicity studies (Anderson et al., 2005, 087916)
24 Although publication bias commonly exists for many research areas, it may be present to a lesser
25 degree for epidemiologic studies on O3. Many epidemiologic studies have focused on the effects of
26 PM, and O3 was largely considered as a potentially confounding co-pollutant of PM. Thus, O3-effect
27 estimates may have been presented in these studies regardless of the statistical significance of the
28 results.
29 Finally, identification of the susceptible population groups contributes to an understanding of
30 the public health impact of pollutant exposures. In this ISA, the term "susceptible population" will
31 be used as an overarching concept to encompass populations variously described as susceptible,
32 vulnerable, or sensitive. "Susceptible populations" is defined here as those populations that have a
33 greater likelihood of experiencing health effects related to exposure to an air pollutant (e.g., O3) due
34 to a variety of factors including but not limited to: genetic or developmental factors, race, gender,
35 lifestage, lifestyle (e.g., smoking status and nutrition) or preexisting disease; as well as population-
36 level factors that can increase an individual's exposure to an air pollutant (e.g., O3) such as
37 socioeconomic status [SES], which encompasses reduced access to health care, low educational
38 attainment, residential location, and other factors. Epidemiologic studies can help identify
39 susceptible populations by evaluating health responses in the study population. Examples include
March 2011 1 -23 DRAFT - DO NOT CITE OR QUOTE
-------
1 stratified analyses for subsets of the population under study or testing for interactions or effect
2 modification by factors such as gender, age group, or health status. Experimental studies using
3 animal models of susceptibility or disease can also inform the extent to which health risks are likely
4 greater in specific population groups. Further discussion of these groups is presented in Section 5.7.
1.6.5.2. Effects on Ecosystems or Public Welfare
Key questions for understanding the quantitative relationships between exposure (or
ntration or deposition) to a pollutant and risk to ecosystems or the public welfare include:
5
6 concentration or
7 • What elements of the ecosystem (e.g., types, regions, taxonomic groups, populations,
8 functions, etc.) appear to be affected, or are more sensitive to effects?
9 • Under what exposure conditions (amount deposited or concentration, duration and
10 pattern) are effects seen?
11 • What is the shape of the concentration-response or exposure-response relationship?
12 Evaluations of causality generally consider the probability of quantitative changes in
13 ecological and welfare effects in response to exposure. A challenge to the quantification of exposure-
14 response relationships for ecological effects is the great regional and local variability in ecosystems.
15 Thus, exposure-response relationships are often determined for a specific ecological system and
16 scale, rather than at the national or even regional scale. Quantitative relationships therefore are
17 available site by site. For example, an ecological response to deposition of a given pollutant can
18 differ greatly between ecosystems. Where results from greenhouse or animal ecotoxicological
19 studies are available, they may be used to aid in characterizing exposure-response relations,
20 particularly relative to mechanisms of action, and characteristics of sensitive biota.
1.6.6. Concepts in Evaluating Adversity of Health Effects
21 In evaluating the health evidence, a number of factors can be considered in determining the
22 extent to which health effects are "adverse" for health outcomes such as changes in lung function or
23 in cardiovascular health measures. Some health outcome events, such as hospitalization for
24 respiratory or cardiovascular diseases, are clearly considered adverse; what is more difficult is
25 determining the extent of change in the more subtle health measures that is adverse. What constitutes
26 an adverse health effect may vary between populations. Some changes in healthy individuals may
27 not be considered adverse; while those of a similar type and magnitude are potentially adverse in
28 more susceptible individuals.
29 For example, the extent to which changes in lung function are adverse has been discussed by
30 the American Thoracic Society (ATS) in an official statement titled What Constitutes an Adverse
31 Health Effect of Air Pollution? (2000, 011738). This statement updated the guidance for defining
March 2011 1 -24 DRAFT - DO NOT CITE OR QUOTE
-------
1 adverse respiratory health effects that had been published 15 years earlier (ATS, 1985, 006522).
2 taking into account new investigative approaches used to identify the effects of air pollution and
3 reflecting concern for impacts of air pollution on specific susceptible groups. In the 2000 update,
4 there was an increased focus on quality of life measures as indicators of adversity and a more
5 specific consideration of population risk. Exposure to air pollution that increases the risk of an
6 adverse effect to the entire population is viewed as adverse, even though it may not increase the risk
7 of any identifiable individual to an unacceptable level. For example, a population of asthmatics
8 could have a distribution of lung function such that no identifiable individual has a level associated
9 with significant impairment. Exposure to air pollution could shift the distribution such that no
10 identifiable individual experiences clinically relevant effects. This shift toward decreased lung
11 function, however, would be considered adverse because individuals within the population would
12 have diminished reserve function and therefore would be at increased risk to further environmental
13 insult.
14 It is important to recognize that the more subtle health outcomes may be linked to health
15 events that are clearly adverse. For example, air pollution has been shown to affect markers of
16 transient myocardial ischemia such as ST-segment abnormalities and onset of exertional angina. In
17 some cases, these effects are silent yet may still increase the risk of a number of cardiac events,
18 including MI and sudden death.
1.7. Summary
19 This draft ISA is a concise evaluation and synthesis of the most policy-relevant science for
20 reviewing the NAAQS for O3, and it is the chief means for communicating the critical science
21 judgments relevant to that NAAQS review. It reviews the most policy-relevant evidence from
22 atmospheric science, exposure, health, and ecological and welfare effects studies; and includes
23 mechanistic evidence from basic biological science. A framework for making critical judgments
24 concerning causality was presented in this chapter. It relies on a widely accepted set of principles and
25 standardized language to express evaluation of the evidence. This approach can bring rigor and
26 clarity to current and future assessments. Once complete, the ISA should assist EPA and others, now
27 and in the future, to accurately represent what is presently known and what remains unknown
28 concerning the effects of O3 on human health and public welfare.
March 2011 1 -25 DRAFT - DO NOT CITE OR QUOTE
-------
References
Air quality criteria and control techniques, Section 108 of the Clean Air Act. 42 USC A§ 7408. (1990).
http://www.epa.gov/air/caa/titlel .html#ia. 080701
Air quality criteria document for ozone and other photochemical oxidants. 47 Fed. Reg. 11561 (1982). 043971
Air quality criteria for ozone and related photochemical oxidants; notice; call for information. 65 Fed. Reg. 57810 (2000).
080678
American Petroleum Institute v. Costle (D.C. Cir. 1981). http://cases.justia.com/us-court-of-appeals/F2/665/1176/408302/
090978
Anderson, H. R.; Atkinson, R. W.; Peacock, J. L.; Sweeting, M. J.; Marston, L. (2005). Ambient particulate matter and
health effects: Publication bias in studies of short-term associations. Epidemiology, 16: 155-163. 087916
ATS (1985). American Thoracic Society: Guidelines as to what constitutes an adverse respiratory health effect, with special
reference to epidemiologic studies of air pollution. Am Rev Respir Dis, 131: 666-668. 006522
ATS (2000). What constitutes an adverse health effect of air pollution? Official statement of the American Thoracic
Society. Am J Respir Crit Care Med, 161: 665-673. 011738
Blair, A.; Burg, J.; Foran, J.; Gibb, H.; Greenland, S.; Morris, R.; Raabe, G; Savitz, D.; Teta, J.; Wartenberg, D.; Wong, O.;
Zimmerman, R. (1995). Guidelines for application of meta-analysis in environmental epidemiology. Regul Toxicol
Pharmacol, 22: 189-197. 079190
CDC (2004). The health consequences of smoking: A report of the Surgeon General. Washington, DC: Centers for Disease
Control and Prevention, U.S. Department of Health and Human Services. 056384
Clewell, H. J.; Crump, K. S. (2005). Quantitative estimates of risk for noncancer endpoints. Risk Anal, 25: 285-289.
http://dx.doi.0rg/10.llll/i.1539-6924.2005.00589.xl56359
Crump, K. S.; Hoel, D. G; Langley, C. H.; Peto, R. (1976). Fundamental carcinogenic processes and their implications for
low dose risk assessment. Cancer Res, 36: 2973-2979. 003192
Definitions, Section 302 of the Clean Air Act. 42 USC A§ 7602. (2005).
http://www.law.cornell.edu/uscode/42/7602.htmltfg. 090976
Final rule to implement the 8-hour ozone national ambient air quality standard-phase 1. 69 Fed. Reg. 23951-24000 (2004).
684054
Gee, G. C.; Payne-Sturges, D. C. (2004). Environmental health disparities: A framework integrating psychosocial and
environmental concepts. Environ Health Perspect, 112: 1645-1653. 093070
Hill, A. B. (1965). The environment and disease: Association or causation? Proc R Soc Med, 58: 295-300.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1898525/pdf/procrsmed00196-0010.pdf 071664
Hoel, D. G. (1980). Incorporation of background in dose-response models. Fed Proc, 39: 73-75. 156555
IARC (2006). IARC monographs on the evaluation of carcinogenic risks to humans: Preamble. Lyon, France: International
Agency for Research on Cancer. http://monographs.iarc.fr/ENG/Preamble/CurrentPreamble.pdf 093206
loannidis, J. P. A. (2008). Why most discovered true associations are inflated. Epidemiology, 19: 640-648. 188317
IOM (2008). Improving the Presumptive Disability Decision-Making Process for Veterans; Committee on Evaluation of the
Presumptive Disability Decision-Making Process for Veterans, Board on Military and Veterans Health. Washington,
DC: Institute of Medicine of the National Academies, National Academies Press.
http://www.nap.edu/openbook.php?record_id=l 1908 156586
Lead Industries v. U.S. Environmental Protection Agency (D.C. Cir. 1980).
http://en.wikisource.org/wiki/Lead_Industries_Association_Inc_v._Environmental_Protection_Agency 090977
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments such as the Integrated Science Assessments (ISAs) and the Integrated Risk Information System (IRIS).
March 2011 1 -26 DRAFT - DO NOT CITE OR QUOTE
-------
National ambient air quality standards for ozone (final rule). 73 Fed. Reg. 16436-16514 (2008). 684051
National ambient air quality standards for ozone - final decision. 58 Fed. Reg. 13008-13019 (1993). 043977
National ambient air quality standards for ozone - Proposed rule. 72 Fed. Reg. 37818 (2007). 684055
National ambient air quality standards for ozone: Final response to remand; final rule. 68 Fed. Reg. 614-645 (2003).
015702
National ambient air quality standards for ozone: Proposed decision. 61 Fed. Reg. 65716-65750 (1996). 031951
National ambient air quality standards for ozone: Proposed response to remand. 66 Fed. Reg. 57268-57292 (2001). 684026
National ambient air quality standards for ozone; final rule. 62 Fed. Reg. 38856-38896 (1997). 083356
National ambient air quality standards for ozone; proposed decision. 57 Fed. Reg. 35542-35557 (1992). 043976
National primary and secondary ambient air quality standards, Section 109 of the Clean Air Act. 42 USC A§ 7409. (1990).
http://www.epa.gov/air/caa/titlel.html#ia. 037658
National primary and secondary ambient air quality standards. 36 Fed. Reg. 8186-8201 (1971). 039176
National primary and secondary ambient air quality standards: revisions to the national ambient air quality standards for
photochemical oxidants. 44 Fed. Reg. 8202-8237 (1979). 039177
Notice of workshop and call for information on integrated science assessment for ozone. 73 Fed. Reg. 56581-56582 (2008).
684057
Photochemical oxidants: Proposed revisions to the national ambient air quality standards. 43 Fed. Reg. 26962-26971
(1978). 684050
Review of the national ambient air quality standards for ozone. 48 Fed. Reg. 38009 (1983). 043972
Rothman, K. J.; Greenland, S. (1998). Modern epidemiology. Philadelphia, PA: Lippincott-Raven Publishers. 086599
U.S. EPA (1970). Air quality criteria for photochemical oxidants. Washington, DC: U.S. Environmental Protection Agency.
014681
U.S. EPA (1978). Air quality criteria for ozone and other photochemical oxidants (Report No. EPA/600/8-78/004).
Washington, DC: U.S. Environmental Protection Agency, Office of Research and Development. 040586
U.S. EPA (1986). Air quality criteria for ozone and other photochemical oxidants (Report No. EPA-600/8-84-020aF - EPA-
600/8-84-020eF). Research Triangle Park, NC: U.S. Environmental Protection Agency. 017607
U. S. EPA (1989). Review of the national ambient air quality standards for ozone: Assessment of scientific and technical
information: OAQPS staff report (Report No. EPA/450/2-92-001). Research Triangle Park, NC: U.S.
Environmental Protection Agency, Office of Air Quality Planning and Standards.
http://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=2000LOW6.txt. 041919
U. S. EPA (1992). Summary of selected new information on effects of ozone on health and vegetation: Supplement to 1986
air quality criteria for ozone and other photochemical oxidants (Report No. EPA/600/8-88/105F). Research Triangle
Park, NC: U.S. Environmental Protection Agency, Office of Health and Environmental Assessment.
http://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=30001HVP.txt. 042599
U.S. EPA (1996). Air quality criteria for ozone and related photochemical oxidants (Report No. EPA/600/P-93/004AF).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development. (NTIS
No. PB94173127) 017831
U.S. EPA (1996). Review of national ambient air quality standards for ozone: assessment of scientific and technical
information OAQPS staff paper (Report No. EPA/452/R-96/007). Research Triangle Park, NC: U.S. Environmental
Protection Agency. (NTIS No. PB96-203435) 039046
U.S. EPA (2004). Air quality criteria for particulate matter (Report No. EPA/600/P-99/002AF). Research Triangle Park,
NC: U.S. Environmental Protection Agency, Office of Research and Development.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=87903. 056905
U.S. EPA (2005). Guidelines for carcinogen risk assessment, final report (Report No. EPA/630/P-03/001F). Washington,
DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid= 116283. 086237
March 2011 1-27 DRAFT - DO NOT CITE OR QUOTE
-------
U.S. EPA (2006). Air quality criteria for ozone and related photochemical oxidants (Report No. EPA/600/R-05/004AF).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid= 149923. 088089
U.S. EPA (2007). Review of the national ambient air quality standards for ozone: policy assessment of scientific and
technical information OAQPS staff paper (Report No. EPA/452/R-07/003). Research Triangle Park, NC: U.S.
Environmental Protection Agency, Office of Air Quality Planning and Standards. 090207
U.S. EPA (2009). Integrated review plan for the ozone national ambient air quality standards review (external review draft)
(Report No. EPA452/D-09-001). Washington, DC: U.S. Environmental Protection Agency.
http://www.epa.gov/ttnnaaqs/standards/ozone/data/externalreviewdraftO3IRP093009.pdf 684024
U.S. EPA (2009). Integrated science assessment for particulate matter (Report No. EPA/600/R-08/139F). Research Triangle
Park, NC: U.S. Environmental Protection Agency, Office of Research and Development.
http://cfpub.epa. gov/ncea/cfm/recordisplay.cfm?deid=216546. 179916
US, E. P. A. (2011). Clean Air Act. Retrieved January 25, 2011 from http://epa.gov/oar/caa/index.html. 013410
Whitman v. American Trucking Associations, 531 U.S.457 (U.S. Supreme Court 2001).
http://supreme.justia.com/us/531/457/ 043004
March 2011 1 -28 DRAFT - DO NOT CITE OR QUOTE
-------
Chapter 2. Integrative Health and
Welfare Effects Overview
1 The subsequent chapters of this ISA will present the most policy-relevant information related
2 to this review of the NAAQS for O3. This chapter integrates the key findings from the disciplines
3 evaluated in this current assessment of the O3 scientific literature, which includes the atmospheric
4 sciences, ambient air data analyses, exposure assessment, dosimetry, health studies
5 (e.g., toxicological, controlled human exposure, and epidemiologic), and welfare effects. The EPA
6 framework for causal determinations described in Chapter 1 has been applied to the body of
7 scientific evidence in order to collectively examine the health or welfare effects attributed to O3
8 exposure in a two-step process.
9 As described in Chapter 1, EPA assesses the results of recent relevant publications, building
10 upon evidence available during the previous NAAQS review, to draw conclusions on the causal
11 relationships between relevant pollutant exposures and health or environmental effects. This ISA
12 uses a five-level hierarchy that classifies the weight of evidence for causation:
13 • Causal relationship
14 • Likely to be a causal relationship
15 • Suggestive of a causal relationship
16 • Inadequate to infer a causal relationship
17 • Not likely to be a causal relationship
18 Beyond judgments regarding causality are questions relevant to quantifying health or
19 environmental risks based on our understanding of the quantitative relationships between pollutant
20 exposures and health or welfare effects. Once a determination is made regarding the causal
21 relationship between the pollutant and outcome category, important questions regarding quantitative
22 relationships include:
23 • What is the concentration-response or dose-response relationship?
24 • Under what exposure conditions (amount deposited, dose or concentration, duration and
25 pattern) are effects observed?
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments sucn as tne Integrated Science Assessments (ISA) and the Integrated Risk Information System (IRIS).
March 2011 2-1 DRAFT - DO NOT CITE OR QUOTE
-------
1 • What populations appear to be differentially affected i.e., more susceptible to effects?
2 • What elements of the ecosystem (e.g., types, regions, taxonomic groups, populations,
3 functions, etc.) appear to be affected or are more sensitive to effects?
4 To address these questions, in the second step of the EPA framework, the entirety of
5 quantitative evidence is evaluated to identify and characterize potential concentration-response
6 relationships. This requires evaluation of levels of pollutant and exposure durations at which effects
7 were observed for exposed populations including potentially susceptible populations.
8 This chapter summarizes and integrates the newly available scientific evidence that best
9 informs consideration of the policy-relevant questions that frame this assessment, presented in
10 Chapter 1. Section 2.1 discusses the trends in ambient concentrations and sources of O3 and provides
11 a brief summary of ambient air quality for short- and long-term exposure durations. Section 2.2
12 presents the evidence regarding personal exposure to ambient O3 in outdoor and indoor
13 microenvironments, and it discusses the relationship between ambient O3 concentrations and
14 exposure to O3 from ambient sources. Section 2.3 provides a discussion of the dosimetry and mode
15 of action evidence for O3 exposure. Section 2.4 integrates the evidence for studies that examine the
16 health effects associated with short- and long-term exposure to O3 and discusses important
17 uncertainties identified in the interpretation of the scientific evidence. Section 2.5 provides a
18 discussion of policy-relevant considerations, such as potentially susceptible populations, lag
19 structure, and the O3 concentration-response relationship. Section 2.6 integrates the health evidence
20 from the different scientific disciplines and exposure durations. Finally, Section 2.7 summarizes the
21 evidence for welfare effects related to O3 exposure, and Section 2.8 reviews the literature on climate
22 and UV-B.
2.1. Atmospheric Chemistry and Ambient Concentrations
23 In the stratosphere, O3 serves the beneficial role of blocking the Sun's harmful ultraviolet
24 radiation and preventing the majority of it from reaching the Earth's surface. In the troposphere,
25 however, O3 and other photochemical oxidants are air pollutants with potentially harmful effects on
26 living organisms and materials. Chapter 3 of this review addresses the atmospheric chemistry
27 associated with tropospheric O3 and other related photochemical oxidants and provides a detailed
28 analysis of recent surface-level concentrations. This material builds on information reported in the
29 2006 O3 AQCD (U.S. EPA, 2006, 088089). Topics covered below and addressed in further detail in
30 Chapter 3 include: (1) physical and chemical processes of O3 formation and removal;
31 (2) atmospheric modeling; (3) policy relevant background concentrations; (4) monitoring techniques
32 and networks; and (5) ambient concentrations.
March 2011 2-2 DRAFT - DO NOT CITE OR QUOTE
-------
2.1.1. Physical and Chemical Processes
1 Ozone in the troposphere is a secondary pollutant; it is formed by photochemical reactions of
2 precursor gasses and is not directly emitted from specific sources. Ozone and other oxidants, such as
3 peroxyacetyl nitrate (PAN) and hydrogen peroxide (H2O2) form in polluted areas by atmospheric
4 reactions involving two main classes of precursor pollutants: VOCs and NOX. Carbon monoxide
5 (CO) is also important for O3 formation in polluted areas and in the remote troposphere. The
6 formation of O3, other oxidants and oxidation products from these precursors is a complex, nonlinear
7 function of many factors: (1) the intensity and spectral distribution of sunlight; (2) atmospheric
8 mixing; (3) concentrations of precursors in the ambient air and the rates of chemical reactions of
9 these precursors; and (4) processing on cloud and aerosol particles.
10 Ozone is present not only in polluted urban atmospheres but throughout the troposphere, even
11 in remote areas of the globe. The same basic processes involving sunlight-driven reactions of NOX,
12 VOCs and CO contribute to O3 formation throughout the troposphere. These processes also lead to
13 the formation of other photochemical products, such as PAN, nitric acid (HNO3), and sulfuric acid
14 (H2SO4), and to other compounds, such as formaldehyde (HCHO) and other carbonyl compounds.
2.1.1.1. Gas Phase Reactions Leading to Ozone Formation and Loss
15 Photochemical processes involved in O3 formation are relatively well understood and were
16 reviewed in detail in the 2006 O3 AQCD (U.S. EPA, 2006, 088089). The photochemical formation of
17 O3 in the troposphere proceeds through the photolysis of nitrogen dioxide (NO2) to yield nitric oxide
18 (NO) and a ground-state oxygen atom, O(3P), which then reacts with molecular oxygen (O2) to form
19 O3. Free radicals formed in the atmosphere through the oxidation of VOCs and CO proceed to
20 oxidize NO back to NO2, hence perpetuating the O3 forming cycle. In urban areas, VOCs and CO are
21 both important for O3 formation. In nonurban vegetated areas, biogenic VOCs emitted from
22 vegetation tend to be the most important. In the remote troposphere, methane (CH4) - structurally the
23 simplest VOC - and CO are the main carbon-containing precursors to O3 formation. A schematic
24 overview of the major photochemical cycles influencing O3 in the troposphere and the stratosphere is
25 given in Figure 3-1.
26 Ozone is lost through a number of gas phase reactions and deposition to surfaces. The reaction
27 of O3 with NO to produce NO2 mainly results in the recycling of O3 downwind via the
28 recombination of O(3P) with O2 to reform O3. By itself, this reaction does not lead to a net change in
29 O3 unless the NO2 is converted to a stable end product such as HNO3 or a temporary reservoir
30 product such as PAN. Ozone also reacts with unsaturated hydrocarbons and with hydrogen
31 containing free radicals (OH, HO2). Recent field studies aimed at obtaining a better understanding of
32 atmospheric chemical processes involved in O3 formation are discussed in Section 3.2.
33 Convective processes and small scale turbulence transport O3 and other pollutants both upward
34 and downward throughout the planetary boundary layer and the free troposphere. In many areas of
35 the U.S., O3 and its precursors can be transported over long distances, aided by vertical mixing. The
March 2011 2-3 DRAFT - DO NOT CITE OR QUOTE
-------
1 transport of pollutants downwind of major urban centers is characterized by the development of
2 urban plumes. Meteorological conditions, small-scale circulation patterns, localized chemistry, and
3 mountain barriers can influence mixing on a smaller scale, resulting in frequent heterogeneous O3
4 concentrations across an individual urban area. More details and observations of these processes are
5 included in Section 3.2 and Section 3.6.
2.1.1.2. Sources ofPrecursors Involved in Ozone Formation
6 Emissions of O3 precursor compounds (NOX, VOCs, and CO) can be divided into
7 anthropogenic and natural source categories. Natural sources can be further divided into biogenic
8 from vegetation, microbes, and animals, and abiotic from biomass burning, lightning, and geogenic
9 sources. However, the distinction between natural sources and anthropogenic sources is often
10 difficult to make in practice, as human activities affect directly or indirectly emissions from what
11 would have been considered natural sources during the preindustrial era. The magnitudes of O3
12 precursor sources are strongly location- and time-dependent and so average emission estimates
13 should not be used to apportion sources of exposure. More details on O3 precursor emission
14 inventories are included in Section 3.2.
2.1.2. Atmospheric Modeling
15 Chemistry-transport models (CTMs) have been widely used to compute the interactions
16 among atmospheric pollutants and their transformation products, and the transport and deposition of
17 pollutants. They have also been widely used to improve our basic understanding of atmospheric
18 chemical processes and to develop control strategies. The main components of a comprehensive
19 atmospheric chemistry modeling system are shown in Figure 3-5 and are discussed in more detail in
20 Section 3.3.
21 The domains of CTMs extend from a few hundred kilometers on a side to the entire globe.
22 Most major regional (i.e., sub-continental) scale air-related modeling efforts at EPA rely on the
23 Community Multi-scale Air Quality modeling system (CMAQ). CMAQ's horizontal domain
24 typically extends over North America with efforts underway to extend it over the entire Northern
25 Hemisphere. The upper boundary for CMAQ is typically set at 100 hPa, which is located on average
26 at about 16-km altitude. CMAQ is most often driven by the MM5 mesoscale meteorological model,
27 though it may be driven by other meteorological models including the Weather Research Forecasting
28 (WRF) model and the Regional Atmospheric Modeling System (RAMS). Other major air quality
29 systems used for regional scale applications include The Comprehensive Air Quality Model with
30 extensions (CAMx) and the Weather Research and Forecast model with Chemistry (WRF/Chem).
31 Fine scale resolution is necessary to resolve features which can affect pollutant concentrations
32 such as urban heat island circulation; sea breezes; mountain and valley breezes; and the nocturnal
33 low-level jet. Horizontal domains are typically modeled by nesting a finer grid model within a larger
34 domain model of coarser resolution. Caution must be exercised in using nested models because
March 2011 2-4 DRAFT - DO NOT CITE OR QUOTE
-------
1 certain parameterizations like those for convection might be valid on a relatively coarse grid scale
2 but may not be valid on finer scales and because incompatibilities can occur at the model boundaries.
3 The use of finer resolution in CTMs will require advanced parameterizations of meteorological
4 processes such as boundary layer fluxes, deep convection, and clouds, and necessitate finer-scale
5 inventories of land use, source locations, and emission inventories.
6 Because of the large number of chemical species and reactions that are involved in the
7 oxidation of realistic mixtures of anthropogenic and biogenic hydrocarbons, condensed mechanisms
8 must be used to simplify atmospheric models. These mechanisms can be tested by comparison with
9 smog chamber data. However, the existing chemical mechanisms often neglect many important
10 processes such as the formation and subsequent reactions of long-lived carbonyl compounds, the
11 incorporation of the most recent information about intermediate compounds, and heterogeneous
12 reactions involving cloud droplets and aerosol particles. As a result, models such as CMAQ have had
13 difficulties with capturing the regional nature of O3 episodes, in part because of uncertainty in the
14 chemical pathways converting NOX to HNO3 and recycling of NOX.
15 Each of the model components shown in Figure 3-5 has associated uncertainties and the
16 relative importance of these uncertainties varies with the modeling application. The largest errors in
17 photochemical modeling are still thought to arise from the meteorological and emissions inputs to
18 the model. Algorithms must be used for simulating meteorological processes that occur on spatial
19 scales smaller than the model's grid spacing and for calculating the dependence of emissions on
20 meteorology and time. Significant errors in emissions can occur if inappropriate assumptions are
21 used in these parameterizations.
22 The performance of CTMs must be evaluated by comparison with field data as part of a cycle
23 of model evaluations and subsequent improvements. Discrepancies between model predictions and
24 observations can be used to point out gaps in current understanding of atmospheric chemistry and to
25 spur improvements in parameterizations of atmospheric chemical and physical processes.
2.1.3. Policy Relevant Background Concentrations
26 The background concentrations of O3 that are useful for risk and policy assessments informing
27 decisions about the NAAQS are referred to as policy-relevant background (PRB) concentrations.
28 PRB concentrations have historically been defined by EPA as those concentrations that would occur
29 in the U.S. in the absence of anthropogenic emissions in continental North America (CNA) defined
30 here as the U.S., Canada, and Mexico. For this document, PRB concentrations include contributions
31 from natural sources everywhere in the world and from anthropogenic sources outside CNA.
32 Contributions to PRB O3 include photochemical reactions involving natural emissions of
33 VOCs, NOX, and CO as well as the long-range transport of O3 and its precursors from outside CNA
34 and the stratospheric-tropospheric exchange (STE) of O3. Natural sources of O3 precursors include
35 biogenic emissions, wildfires, and lightning. Biogenic emissions from agricultural activities in CNA
36 are not considered in the formation of PRB O3. PRB O3 sources and concentrations are summarized
37 here with further details in Section 3.4.
March 2011 2-5 DRAFT - DO NOT CITE OR QUOTE
-------
2.1.3.1. Contributions from anthropogenic emissions outside North America
1 Because the mean tropospheric lifetime of O3 is 30-35 days, O3 can be transported from
2 continent to continent and around the globe in the Northern Hemisphere and O3 produced by U.S.
3 emissions can be recirculated around northern mid-latitudes back to the U.S. High elevation sites are
4 most susceptible to the intercontinental transport of pollution, particularly during spring. Surface
5 PRB O3 contributions are much smaller than those derived in the free troposphere because of
6 dilution and chemical destruction during downward transport to the surface. There are no instances
7 where direct observation of PRB contributions from anthropogenic emissions outside North America
8 are directly observable; careful screening of observations and application of photochemical models
9 must be used instead to estimate anthropogenic contributions to PRB from sources outside North
10 America.
2.1.3.2. Contributions from the stratosphere
11 Ozone is produced naturally by photochemical reactions in the stratosphere and some of this
12 O3 is transported downward into the troposphere throughout the year in a process known as
13 tropopause folding. Maximum stratospheric contributions occur during late winter and early spring,
14 particularly behind cold fronts that mix tropospheric and stratospheric air. Stratospheric intrusions
15 that reach the surface are rare. Much more common are intrusions which penetrate only to the middle
16 and upper troposphere. However, O3 transported to the upper and middle troposphere can still affect
17 surface concentrations through various exchange mechanisms that mix air from the free troposphere
18 with air in the planetary boundary layer. There is considerable uncertainty in the magnitude and
19 distribution of this potentially important source of tropospheric O3.
2.1.3.3. Natural sources of precursors to PRB Ozone formation
20 Biogenic sources of VOC and CO emissions contribute to precursors to PRB O3 formation.
21 These sources were discussed above in Section 2.1.1.1 with further details in Section 3.2.
22 Biomass burning in the form of wildfires and prescribed fires contribute to NOX, CO and
23 VOCs, precursors to PRB O3 formation. Biomass burning exhibits strong seasonality and interannual
24 variability, with most biomass burned during the local dry season. There is considerable uncertainty
25 in attributing the fraction of wildfire emissions to human activities because the emissions from
26 naturally occurring fires that would have been present in the absence of fire suppression practices are
27 not known.
28 Lightning is also a source for NOX production. Although total column estimates of lightning
29 produced NOX are substantial, this source does not contribute substantially to the NOX burden in the
30 continental boundary layer. This is because only 2% of NOX production by lightning occurs within
31 the boundary layer and most occurs in the free troposphere where much of the NOX produced is
32 converted to more oxidized nitrogen species during downward transport.
March 2011 2-6 DRAFT - DO NOT CITE OR QUOTE
-------
2.1.3.4. Estimating PRB Concentrations
1 There are two approaches to estimating PRB concentrations that have been considered thus
2 far. The first involves using measurements and the second the use of chemistry-transport models.
3 The 2006 O3 AQCD (Section 3.9) (U.S. EPA, 2006, 088089) noted that estimates of PRB
4 concentrations cannot be obtained solely by examining measurements of O3 obtained at relatively
5 remote monitoring sites in the U.S. because of the long-range transport from anthropogenic source
6 regions within North America. The 2006 AQCD also noted that it is impossible to determine sources
7 of O3 without ancillary data that could be used as tracers of sources or to calculate photochemical
8 production and loss rates. Furthermore, the use of monitoring data is limited to the edges of the
9 domain of interest because PRB O3 entering from outside North America is destroyed over North
10 America either through chemical reactions or by deposition. Within North America, PRB O3 is only
11 produced by natural sources. Therefore, the current definition of PRB implies that only CTMs can be
12 used to estimate the range of PRB values. A further advantage to using models is that the entire range
13 of O3 concentrations in different environments can be used to evaluate model performance.
14 However, there may be specific instances such as stratospheric intrusions that occur on spatial scales
15 too fine to be resolved by the current generation of global CTMs.
16 Estimates of PRB concentrations for April-May, 2001 (Figure 3-9) and June-August, 2001
17 (Figure 3-10) from the GEOS-Chem model used in the 2006 O3 AQCD (U.S. EPA, 2006, 088089)
18 are described in Section 3.4. These estimates indicated that PRB O3 concentrations in the U.S.
19 surface air were generally 15-35 ppb from June through August. Concentrations decline from spring
20 to summer and are generally <25 ppb under conditions conducive to high O3 episodes. PRB O3
21 concentrations may be higher, especially at high altitude sites during the spring, due to enhanced
22 contributions from (1) pollution sources outside North America; and (2) stratospheric O3 exchange.
23 Simulated monthly mean concentrations in different quadrants of the U.S. are typically within
24 5 ppbv of observations at remote CASTNET sites, with no significant bias, except in the Southeast
25 in summer when the model is 8-12 ppbv too high. This bias might be due to excessive background
26 O3 transported in from the Gulf of Mexico and the tropical Atlantic Ocean or to inaccuracies in
27 emissions inventories within the U.S. The model reproduced the occurrences of relatively high O3 at
28 remote sites, and shows that these can generally be explained by North American pollution.
29 Although many of the features of the day-to-day variability of O3 at relatively remote
30 monitoring sites in the U.S. are simulated reasonably well, uncertainties in the calculation of the
31 temporal variability of O3 originating from different sources on shorter time scales must be
32 recognized. The uncertainties stem in part from an underestimate in the seasonal variability in the
33 STE of O3, the geographical variability of this exchange, and the variability in the exchange between
34 the free troposphere and the planetary boundary layer in the model. In addition, the relatively coarse
35 spatial resolution in the model (20*2.5°) limited the ability to provide separate estimates for cities
36 located close to each other, and so only regional estimates were provided for the 2006 O3 AQCD
37 (U.S. EPA, 2006, 088089V
March 2011 2-7 DRAFT - DO NOT CITE OR QUOTE
-------
2.1.4. Monitoring
2.1.4.1. Routine Monitoring Techniques
1 The Federal Reference Method (FRM) for O3 measurement is called the Chemiluminescence
2 Method (CLM) and is based on the detection of chemiluminescence resulting from the reaction of O3
3 with ethylene gas. The first ultraviolet (UV) absorption photometric analyzers were approved as
4 Federal Equivalent Methods (FEMs) in 1977 and gained rapid acceptance for NAAQS compliance
5 purposes due to ease of operation, relatively low cost, and reliability. Almost all of the state or local
6 air monitoring stations (SLAMS) that reported data to EPA Air Quality System (AQS) from 2005 to
7 2009 used UV absorption photometer FEMs and greater than 96% of O3 monitors met precision and
8 bias goals during this period. The rationale, history, and calibration of O3 measurements were
9 summarized in the 1996 O3 AQCD (U.S. EPA, 1996, 017831) and the 2006 O3 AQCD (U.S. EPA,
10 2006, 088089) and focused on the state of ambient O3 measurements at that time as well as
11 evaluation of interferences and new developments. Section 3.5 in this review includes the current
12 state of O3 measurements, interferences, and new developments for the period 2005-2009.
13 Satellite observations for O3 are growing as a resource for many purposes, including model
14 evaluation, assessing emissions reductions, pollutant transport, and air quality management. Satellite
15 remote sensing instruments do not directly measure the composition of the atmosphere. Satellite
16 retrievals are conducted using the solar backscatter or thermal infrared emission spectra and a variety
17 of algorithms. Most satellite measurement systems have been developed for stratospheric
18 measurement of the total O3 column. Mathematical techniques have been developed and must be
19 applied to derive information from these systems about tropospheric O3.
2.1.4.2. Ambient Ozone Network Design
20 To support the NAAQS, state and local monitoring agencies must operate O3 monitors at
21 various locations depending on the area size and typical peak concentrations (expressed in
22 percentages below, or near the O3 NAAQS). SLAMS make up the ambient air quality monitoring
23 sites that are primarily needed for NAAQS comparisons and include Photochemical Assessment
24 Monitoring Stations (PAMS), National Core (NCore), and all other State or locally-operated stations
25 except for the monitors designated as special purpose monitors (SPMs).
26 In 2009, there were 1208 SLAMS O3 monitors reporting values to the EPA AQS database
27 (Figure 3-16). Since O3 levels decrease significantly in the colder parts of the year in many areas, O3
28 is required to be monitored at SLAMS monitoring sites only during the "O3 season." PAMS provides
29 more comprehensive data on O3 in areas classified as serious, severe, or extreme nonattainment for
30 O3. There were a total of 119 PAMS reporting values to the EPA AQS database in 2009. NCore is a
31 new multi-pollutant monitoring network currently being implemented to meet multiple monitoring
32 objectives. Each state is required to operate at least one NCore site and the network will consist of
33 about 60 urban and 20 rural sites nationwide.
March 2011 2-8 DRAFT - DO NOT CITE OR QUOTE
-------
1 The Clean Air Status and Trends Network (CASTNET) is a regional monitoring network
2 established to assess trends in acidic deposition and also provides concentration measurements of O3.
3 CASTNET O3 monitors operate year round and are primarily located in rural areas. At the beginning
4 of 2010, there were 80 CASTNET sites located in, or near, rural areas. The NFS also operates a
5 Portable Ozone Monitoring Systems (POMS) network. The POMS couples the small, low-power O3
6 monitor with a data logger, meteorological measurements, and solar power in a self contained system
7 for monitoring in remote locations. Twenty NPS POMS reported O3 data to AQS in 2010. A map of
8 the current and proposed rural NCore sites, along with the CASTNET, and the NPS POMS sites is
9 shown in Figure 3-17.
2.1.5. Ambient Concentrations
10 Ozone is the only photochemical oxidant other than NO2 that is routinely monitored and for
11 which a comprehensive database exists. Data for other photochemical oxidants typically have been
12 obtained only as part of special field studies. Most continuous O3 monitors report hourly average
13 concentrations. This data can be used as reported 1-h avg, or reported as a daily metric such as: (1)
14 the average of the hourly observations over a 24-h period (24-h avg); (2) the maximum hourly
15 observation occurring in a 24-h period (1-h daily max); and (3) the maximum 8-h running average of
16 the hourly observations occurring in a 24-h period (8-h daily max).
17 Section 3.6.1 includes an analysis of U.S. O3 data reported to AQS between 2007 and 2009.
18 The median 1-h daily max, 8-h daily max, and 24-h avg O3 concentrations across all sites were 44,
19 40, and 29 ppb, respectively. The 98th percentiles of these same metrics across all sites were 86, 74,
20 and 55 ppb, respectively. The 8-h daily max and 1-h daily max metrics were highly correlated
21 (median r = 0.97, IQR = 0.96-0.98) while comparisons with the 24-h avg metric were lower (e.g.,
22 median r = 0.83, IQR = 0.78-0.88 for comparison between the 24-h avg and the 1-h daily max). The
23 ratio and correlation between these metrics, however, can be very site-specific.
2.1.5.1. Urban-Focused Spatial Variability
24 AQS O3 concentrations were used to investigate urban-focused spatial variability in
25 Section 3.6.2. Figure 3-22 contains the county-scale 8-h daily max O3 concentrations from the
26 highest monitor within each U.S. county for 2007-2009 (top map) with seasonal stratification
27 (bottom 4 maps). This map is only meant to illustrate the general spatial and temporal distribution in
28 nationwide O3 concentrations, and is limited by monitor availability, resulting in the majority of U.S.
29 counties not having available data. Furthermore, this map is not representative of O3 concentrations
30 at all locations or times within the counties shown; considerable spatial variability can and does exist
31 within a county.
32 The highest 3-yr avg (2007-2009) 8-h daily max O3 concentrations £ 50 ppb), shown in
33 Figure 3-22, occur in counties in southern California, Arizona, Colorado and Tennessee. The lowest
34 monitored 3-yr avg 8-h daily max O3 concentrations (<30 ppb) occur in Pacific Coast counties in
March 2011 2-9 DRAFT - DO NOT CITE OR QUOTE
-------
1 northern California and Washington, as well as in two northeastern counties in Pennsylvania and
2 Massachusetts. The seasonally-stratified county-scale maps in the lower half of Figure 3-22 illustrate
3 the strong seasonality in 8-h daily max O3 concentrations.
4 To investigate urban-scale variability, 20 focus cities were selected for closer analysis here and
5 in Section 3.6.2; these cities were selected based on their importance in O3 epidemiology studies and
6 on their geographic distribution across the U.S. The warm season (May-September) distribution of
7 the 8-h daily max O3 concentrations from 2007-2009 for the 20 focus cities is included in Table 3-10.
8 Box plots of the distribution of 2007-2009 warm-season 8-h daily max O3 data from each
9 individual monitor in the 20 focus cities were used in Section 3.6.2 to investigate individual city
10 variability in O3 concentrations. Several cities had relatively little spatial variability in 8-h daily max
11 O3 concentrations (e.g., correlations ranging from 0.61 to 0.96 in Atlanta) while other cities exhibited
12 considerably more variability in O3 concentrations (e.g., correlations ranging from -0.06 to 0.97 for
13 Los Angeles). The negative and near-zero correlations in Los Angeles were between monitors with a
14 relatively large separation distance (>150 km), but even some of the closer monitor pairs were not
15 very highly correlated. Similar to the correlation, the coefficient of divergence (COD) was found to
16 be highly dependent on the urban area under investigation. As a result, caution should be observed in
17 using data from a sparse network of ambient O3 monitors to approximate community-scale
18 exposures.
2.1.5.2. Rural-Focused Spatial Variability
19 AQS O3 data for monitors located within six rural monitoring sites were used in Section 3.6.2
20 to investigate rural-focused O3 concentration variability. These rural monitoring sites tend to be less
21 directly affected by obvious anthropogenic pollution sources than urban sites. However, they can be
22 regularly affected by transport of O3 or O3 precursors from upwind urban areas, or by local
23 anthropogenic emissions within the rural areas such as emissions from motor vehicles, power
24 generation, biomass combustion, or oil and gas operations. As a result, monitoring data from these
25 rural locations are not unaffected by anthropogenic emissions.
26 Box plots of 8-h daily max O3 concentrations measured at the 6 rural monitoring sites during
27 the warm season (May-September) between 2007 and 2009 are shown in Figure 3-37. The sites
28 include one in Adirondack State Park (ADSP) on Whiteface Mountain in Upstate NY, one in Mount
29 Mitchell State Park (MMSP) in NC, five in Great Smoky Mountain National Park (SMNP) in NC
30 and TN, one in Rocky Mountain National Park (RMNP) in CO, one in San Bernardino National
31 Forest (SBNF), CA, and two in Sequoia National Park (SENP), CA. Within SMNP, the median
32 warm-season 8-h daily max O3 concentration ranged from 47 ppb at the lowest elevation site
33 (elevation = 564 m; site ID = 470090102) to 60 ppb at the highest elevation site (elevation = 2021 m;
34 site ID = 471550102), with correlations between the 5 sites ranging from 0.78 to 0.92 and CODs
35 ranging from 0.04 to 0.16. The correlation between the 2 sites in SENP was 0.86 and the COD was
36 0.09. A host of factors may contribute to variations observed at these rural sites, including proximity
37 to local O3 precursor emissions, variations in boundary-layer influences, meteorology and
March 2011 2-10 DRAFT - DO NOT CITE OR QUOTE
-------
1 stratospheric intrusion as a function of elevation, and differences in wind patterns and transport
2 behavior due to local topography. Expanded analyses of O3 concentrations measured using the more
3 rural-focused CASTNET monitoring network are included in Chapter 9.
4 Since O3 produced from emissions in urban areas is transported to more rural downwind
5 locations, elevated O3 concentrations can occur at considerable distances from urban centers. In
6 addition, major sources of O3 precursors such as highways, power plants, biomass combustion, and
7 oil and gas operations are commonly found in rural areas, adding to the O3 in these areas. Due to
8 lower chemical scavenging in nonurban areas, O3 tends to persist longer in rural than in urban areas
9 which tends to lead to higher cumulative exposures in rural areas influenced by anthropogenic
10 precursor emissions. The persistently high O3 concentrations observed at many of these rural sites
11 investigated here indicate that cumulative exposures for humans and vegetation in rural areas can be
12 substantial and often higher than cumulative exposures in urban areas.
2.1.5.3. National Trends
13 Nationally, O3 concentrations have declined over the last decade, as shown in Figure 3-41
14 from the 2010 National Air Quality Status and Trends report (U.S. EPA, 2010, 647278). The
15 majority of this decline occurred before 2004 with national average concentrations remaining
16 relatively flat between 2004 and 2008. The large decreases in 2003 and 2004 coincides with NOX
17 emissions reductions resulting from implementation of the NOX State Implementation Plan (SIP)
18 Call rule, which began in 2003 and was fully implemented in 2004. This rule was designed to reduce
19 NOX emissions from power plants and other large combustion sources in the eastern U.S.
20 As noted in the 2006 O3 AQCD (U.S. EPA, 2006, 088089). trends in national parks and rural
21 areas are similar to nearby urban areas, reflecting the regional nature of O3 pollution. However,
22 caution should be exercised in using trends calculated at national parks to infer contributions from
23 distant sources either inside or outside of North America because of the influence of regional
24 pollution.
2.1.5.4. Hourly Variations
25 Ozone concentrations show a strong degree of diel variability resulting from daily patterns in
26 temperature, sunlight, and precursor emissions. Other factors, such as the relative importance of
27 transport versus local photochemical production and loss rates, the timing for entrainment of air from
28 the nocturnal residual boundary layer, and the diurnal variability in mixing layer height also play a
29 role in daily O3 patterns. Urban diel variations investigated in Section 3.6.3.2 of this assessment
30 show no substantial change in patterns since the 2006 O3 AQCD (U.S. EPA, 2006, 088089). The 1-h
31 max concentrations tend to occur in mid-afternoon and 1 -h min concentrations tend to occur in early
32 morning, with more pronounced peaks in the warm months relative to the cold months. Diel patterns
33 in O3 have remained stable over the last 20 years, with times of occurrence of the daily maxima
34 varying by no more than an hour from year to year. There is city-to-city variability in these times,
March 2011 2-11 DRAFT-DO NOT CITE OR QUOTE
-------
1 however, and caution is raised in extrapolating results from one city to another in determining the
2 time of day for O3 maxima and minima.
2.1.5.5. Associations with Co-pollutants
3 Since O3 is a secondary pollutant formed in the atmosphere from precursor emissions, it is not
4 expected to be highly correlated with primary pollutants such as CO and NOX. Furthermore, O3
5 formation is strongly influenced by meteorology, entrainment, and transport of both O3 and O3
6 precursors, resulting in a broad range in correlations with other pollutants which can vary
7 substantially with season. To investigate correlations with co-pollutants, 8-h daily max O3 was
8 compared with co-located 24-h avg CO, SO2, NO2, PM2 5 and PMi0 obtained from AQS for
9 2007-2009. Figure 3-43 contains co-pollutant box plots of the correlation between co-located
10 monitors for the year-round data set and broken down by season.
11 The year-round 8-h daily max O3 data exhibited a very wide range in correlations with all the
12 24-h avg co-pollutants. A clearer pattern emerged when the data are stratified by season with mostly
13 negative correlations in the winter and mostly positive correlations in the summer for all co-
14 pollutants. The median seasonal correlations are modest at best with the highest positive correlation
15 at 0.52 for PM2 5 in the summer and the highest negative correlation at -0.38 for PM2 5 in the winter.
16 Spring and fall lie in between with spring having a slightly narrower distribution than fall for all co-
17 pollutants. Expanded discussion of co-pollutant correlation can be found in Section 3.6.4.
2.2. Human Exposure
2.2.1. Exposure Measurement
2.2.1.1. Measurement of Ozone Exposure
18 Passive badge samplers are the most widely used technique for measuring personal O3
19 exposure. They operate on the nitrite-nitrate conversion principle, and are convenient since they
20 require no pumps or wet chemistry in the field. They represent a cumulative (rather than continuous)
21 sample, and their detection limit makes them suitable for monitoring periods of 24 hours or greater.
22 This limits their applicability in measuring short-term daily fluctuations in personal exposure. Over a
23 24-h period, the detection limit of the badges is approximately 5-10 ppb, which may result in an
24 appreciable fraction of the samples being below the detection limit. An active sampler based on the
25 nitrite-nitrate conversion reaction is also available, with a reported detection limit of 10 ppb-h,
26 enabling measurement of sub-daily O3 concentrations. A portable continuous O3 monitor based on a
27 different principle, UV absorption, has recently become available. Its size and weight make it
28 suitable for use in a backpack configuration, although its use for personal exposure measurements
29 has been limited.
March 2011 2-12 DRAFT - DO NOT CITE OR QUOTE
-------
1 Several studies described in the 2006 O3 AQCD, along with a few new studies published since,
2 describe the relationship between indoor O3 concentration and the O3 concentration immediately
3 outside the indoor microenvironment. These studies show that the indoor concentration is often
4 substantially lower than the outdoor concentration unless indoor sources are present. Low indoor O3
5 concentrations can be explained by reaction of O3 with surfaces and airborne constituents. However,
6 the indoor-outdoor relationship is greatly affected by the air exchange rate; under conditions of high
7 air exchange rate, such as open windows, the indoor O3 concentration may approach the outdoor
8 concentration. In residential microenvironments, studies report indoor-outdoor ratios ranging from
9 approximately 0.1-0.4, with the highest ratios observed in the summer O3 season and for homes with
10 increased window ventilation. A correlation of 0.58 was reported between indoor and outdoor O3
11 concentrations, indicating that variations in outdoor concentration may be reflected indoors, though
12 the magnitude of the concentration is lower. Indoor-outdoor ratios at schools were similar, with
13 higher ratios observed during the school day when opening doors and windows may lead to
14 increased air exchange rates. In vehicles, high air exchange rates that would normally lead to high
15 interior-exterior concentration ratios are offset by O3 scavenging through vehicle-emitted NO,
16 resulting in reported in-vehicle concentrations that were approximately 50% of those measured at the
17 roadside.
18 The relationship between personal exposure and ambient O3 concentrations has been evaluated
19 in several research studies, many of which were conducted prior to 2005 and are discussed in the
20 2006 O3 AQCD. The results of these studies indicate that personal exposures are moderately well
21 correlated with ambient concentrations, and that the ratio of personal exposure to ambient
22 concentration is higher in outdoor microenvironments and during the summer season. In situations
23 where a lack of correlation was observed, this may be due in part to a high proportion of personal
24 measurements below the detection limit. Correlations reported for daily or multi-day measurements
25 range from approximately 0.3-0.8, with the upper end of the range reflecting longer-duration (4-day)
26 community average measurements that may limit the influence of inter-individual variability in
27 exposure. Hourly measurements in specific microenvironments show greater variability in
28 correlations between personal exposure and ambient concentration, with residential indoor
29 correlations <0.1 and outdoor correlations of 0.7-0.9. Slopes from regression analyses of personal
30 exposure on ambient concentration generally ranged from approximately 0.1-0.3. Higher slopes were
31 observed in studies that either adjusted for activity pattern and air exchange rate (0.54) or focused on
32 outdoor shoe cleaners (0.56), who may have increased exposure due to spending a substantial
33 fraction of the day outdoors. Ratios of personal exposure to ambient concentration showed similar
34 results, with a ratio of 0.3 reported for a year-round study in southern California, while ratios ranged
35 from 0.28-0.96 for outdoor workers, increasing with time spent outdoors.
36 Taken together, results from previous and recently published studies indicate that while the
37 relationship between personal exposures and ambient concentrations varies due to a number of
38 factors, such as activity patterns, housing characteristics, and season, O3 concentrations measured at
39 central-site monitors are representative of day-to-day changes in average personal O3 exposure,
March 2011 2-13 DRAFT - DO NOT CITE OR QUOTE
-------
1 which is the important parameter for time-series epidemiologic studies. Another important finding is
2 that the magnitude of personal exposures is smaller than concentrations reported at fixed-site
3 monitors due to time spent indoors and the low indoor penetration of O3.
2.2.1.2. Co-Exposure to Ozone and Other Pollutants
4 Individuals may be exposed to other pollutants in conjunction with exposure to O3. Personal
5 exposure to O3 shows variable association with personal exposure to other pollutants, with
6 differences in association depending on factors such as season, city-specific characteristics, and
7 spatial variability of the co-pollutant. For PM2 5, a rank correlation of 0.14 was reported between
8 daily O3 and PM2 5 exposures during spring and fall in Atlanta. Positive slopes were reported during
9 summer in both Baltimore and Boston, although the slopes were somewhat different (0.21 and 0.72,
10 respectively). The summertime slope in Baltimore was higher for children (0.37) than for adults
11 (0.07), which may be the result of different activity patterns and time spent outdoors. Additional
12 evidence of variation by season and city is provided by the differing signs of the wintertime slopes,
13 with Baltimore showing a negative slope and Boston showing a positive slope. Interindividual
14 variability likely played a role as well, since both cities showed a wide range (including both
15 negative and positive values) for individual-specific personal O3- PM25 slopes. For EC and NO2,
16 near-zero correlations were reported with O3 during spring and fall in Atlanta. These extremely low
17 correlations for the traffic-related and spatially variable pollutants EC and NO2 contrast with the
18 higher correlation observed for PM25, a regional pollutant.
19 In near-road and on-road microenvironments, correlations between O3 and traffic-related
20 pollutants are moderately to strongly negative, with the most strongly negative correlations observed
21 for NO2 (-0.8 to -0.9). This is consistent with the chemistry of NO oxidation, in which O3 is
22 consumed to form NO2. The more moderate negative correlations observed for PM2 5, UFP, and VOC
23 may reflect reduced concentrations of O3 in more polluted environments due to other scavenging
24 reactions. A similar process occurs indoors, where infiltrated O3 reacts with airborne or surface-
25 associated materials to form secondary compounds, such as formaldehyde. Although such reactions
26 decrease indoor O3 exposure, they result in increasing exposure to other species which may
27 themselves have health effects.
2.2.2. Exposure Modeling
28 Exposures estimates in urban areas may be improved by constructing a concentration surface
29 over a geographic domain using a model to compensate for missing data. The calculated
30 concentration surface can then be used to estimate exposures outside residences, schools,
31 workplaces, roadways, or other locations of interest. This technique does not estimate exposure
32 directly because it does not account for activity patterns or concentrations in different
33 microenvironments. Most such modeling efforts have focused on the less-reactive pollutants PM or
34 NO2. In a study that extended CALINE4 NOX modeling results to evaluate the impact on residential
March 2011 2-14 DRAFT - DO NOT CITE OR QUOTE
-------
1 O3 concentrations, O3 concentrations were reduced by 0.51 ppb O3 per 1 ppb NOX. This intra-urban
2 traffic-related variability in O3 concentrations suggests that differences in traffic density between the
3 central site monitor and individual homes could result in either an overestimate or underestimate of
4 residential O3.
5 A separate class of models, known as microenvironmental models, estimate time-weighted
6 exposure for modeled individuals by summing exposure in each microenvironment visited during the
7 exposure period. Stochastic microenvironmental models, such as APEX and SHEDS, utilize
8 distributions of pollutant-related and individual-level variables, such as ambient and local O3
9 concentration contributions and breathing rate respectively, to compute the distribution of individual
10 exposures across the modeled population. The models also have the capability to estimate received
11 dose through a dosimetry model. Using distributions of input parameters in the model framework
12 rather than point estimates allows the models to incorporate uncertainty and variability explicitly into
13 exposure estimates. For the APEX model, an analysis has been conducted indicating that the
14 uncertainty in model exposure estimates for asthmatic children during moderate exercise is small to
15 moderate; however, APEX appears to substantially underestimate the frequency of multiple high-
16 exposure events for a single individual. Microenvironmental models, such as EMI, are also being
17 developed to use individual-specific information derived from measurements or questionnaires,
18 rather than population distributions, to estimate exposures. This approach is particularly suitable for
19 panel health studies where information is available for each participant, and may reduce uncertainty
20 in health effect estimates by improving exposure estimates.
2.2.3. Implications for Epidemiologic Studies
21 Exposure error can be an important contributor to variability in epidemiologic study results,
22 although this may be less of an issue for O3 because it is a secondary pollutant with relatively low
23 spatial variability across an urban area. For example, an epidemiologic study in Atlanta observed
24 similar associations between heart rate variability (HRV) parameters and either ambient
25 concentrations or personal exposures of O3 and PM2 5, another regional pollutant. The importance of
26 exposure error varies with study design and is dependent on the spatial and temporal aspects of the
27 design. Several factors that could influence exposure estimates include nonambient exposure, spatial
28 and temporal variability, and the presence of O3 in a mixture of pollutants. Nonambient exposure is
29 unlikely to influence health effect estimates because of the lack of indoor O3 sources and because
30 indoor-generated O3 exposures are unlikely to be correlated with ambient O3 exposure. Compared
31 with directly emitted pollutants such as CO and NOX, O3 exhibits relatively low spatial variability
32 across urban areas, as discussed in Chapter 3. Averaging data from a large number of samplers will
33 dampen intersampler variability, and use of multiple monitors over smaller land areas may allow for
34 more variability to be incorporated into an epidemiologic analysis. Evidence from a study comparing
35 the effect of spatial variability on effect estimates for O3, PM2 5, NO2, and CO suggests that choice of
36 monitor for spatially homogenous pollutants such as O3 may have little impact on the results of
37 epidemiologic studies. Season, however, may have a substantial effect due to much lower O3
March 2011 2-15 DRAFT - DO NOT CITE OR QUOTE
-------
1 concentrations during the winter, along with the higher correlations between ambient concentrations
2 and personal exposures observed during the summer. Studies conducted during the O3 season or in
3 periods when communities are likely to have high air exchange rates (e.g., during mild weather) may
4 be less prone to exposure error than studies conducted only during winter. Year-round studies that
5 include both the O3 and non-O3 seasons may have an intermediate level of exposure error. Exposure
6 to mixtures of pollutants containing O3 also complicates interpretation of epidemiologic results.
7 Moderate to strong negative correlations between O3 and traffic-related pollutants, particularly NO2,
8 make it difficult to determine to what extent O3-based effect estimates quantitatively reflect the
9 independent effect of O3 itself, or the effect of another pollutant or pollutants in the mixture.
10 Interpretation of O3 effects in the presence of PM is additionally complicated by the highly variable
11 correlations observed, which differ by city, season, and population characteristics (e.g., children
12 versus adults). Although these sources of exposure error should be considered in evaluating
13 epidemiologic results, previous and recently published exposure research indicate that O3
14 concentrations measured at central-site monitors are indicative of day-to-day changes in average
15 personal O3 exposure, making ambient concentrations a useful parameter for epidemiologic studies.
2.3. Dos imetry and Mode of Action
16 Ozone is a highly reactive and poorly water soluble gas allowing it to penetrate into targets in
17 the lower respiratory tract. The fact that it is so chemically reactive suggests that the effective dose at
18 target sites exists in the form of secondary oxidation products such as aldehydes and peroxides.
19 Reaction products are formed when O3 interacts with components of the extracellular lining fluid
20 (ELF) such as lipids, proteins, and antioxidants. Ozone uptake relates directly to these ELF substrate
21 reactions and is termed 'reactive absorption'. The level and type of antioxidants varies between
22 species, regions of the respiratory tract itself, and can be altered by O3 exposure. ELF constituents
23 appear in most cases to limit interaction of O3 with underlying tissues and to prevent penetration of
24 O3 deeper into the lung. However, in some cases, the antioxidants and secondary oxidation products
25 formed in the aqueous phase might penetrate into the cells and cause injury. Ozone toxicity is
26 observed to some extent in the nasal cavity, however further toxicity exists in the deep lung where
27 the ELF thickness narrows allowing O3 to react directly with the epithelial cells and surface
28 macrophages.
2.3.1. Human and Animal Ozone Dosimetry
29 O3 uptake efficiency is chemical-reaction dependent, driven by the conversion of O3 to
30 reaction products. The primary site of O3 uptake and greatest O3 dose in the lungs is the centriacinar
31 region (CAR), containing the respiratory bronchioles. Recent studies have provided evidence for hot
32 spots of O3 flux around bifurcations in the airways. Ozone uptake is 80-95% efficient in humans and
33 approximately 54% efficient in rats. The nasopharyngeal region provides defense against O3 entering
34 the lungs and removes -50% of the absorbed O3 in both species. Ozone uptake efficiency is sensitive
March 2011 2-16 DRAFT - DO NOT CITE OR QUOTE
-------
1 to a number of factors. Fractional absorption will decrease with increased flow and increase
2 proportional to tidal volume (VT). Increased breathing frequency (fB) and oronasal breathing, as
3 occurs during exercise, will shift the O3 dose distribution deeper and lead to a greater dose to the
4 pulmonary region, increasing the potential of damage to bronchiolar and alveolar tissues. Individual
5 total airway O3 uptake efficiency is also sensitive to large changes in O3 concentration, exposure
6 time, and minute ventilation (VE). Major sources of variability in absorption of O3 include O3
7 concentration, exposure time, breathing frequency, minute volume, and tidal volume, but the
8 interindividual variation is the greatest source of variability uptake efficiency. However, to date, no
9 studies have shown that the large differences in biological response between subjects (forced
10 expiratory volume in 1 sec [FEVi], bronchoalveolar lavage fluid [BAL], cell inflammatory response,
11 etc.) are explainable by the differences in O3 uptake.
12 Interspecies differences limit quantitative comparison between species; however, the acute and
13 chronic functional responses of laboratory animals to O3 appear qualitatively homologous to that of
14 the human making them a useful tool in determining mechanistic and cause-effect relationships with
15 O3 exposure. Recent studies have shown that varied O3 response in different mouse strains was not
16 due to differences in delivered dose of O3 to the lung but more likely genetic sensitivity. Dose
17 comparison between humans and rats shows that exercising humans accumulated 4-5 times higher
18 O3 reactants in BAL compared to similarly exposed resting rats and it was necessary to expose
19 resting rats to 2 ppm O3 to achieve increases in BAL protein and polymorphonuclear cells [PMNs]
20 similar to those of the 0.4 ppm exposed humans.
2.3.2. Possible Pathways/Modes of Action
21 Three distinct short-term responses have been well-characterized in humans challenged with
22 O3: decreased pulmonary function, airways inflammation, and increased bronchial reactivity. In
23 addition, evidence has been accumulating that O3 exposure exacerbates, and possibly causes, asthma
24 and allergic airways disease in humans. Effects on the nasal airways and distal lung of humans,
25 including inflammation and injury, have also been described. Animal studies have demonstrated a
26 wide range of respiratory system effects. While the respiratory tract is the primary target tissue,
27 cardiovascular and other organ effects occur following short- and long-term exposures of animals to
28 O3. Mechanisms responsible for these effects are incompletely understood.
29 The initial key event in the O3 toxicity pathway is the formation of secondary oxidation
30 products in the respiratory tract. Pathways for the removal of those products are also of great
31 importance. Due to the highly reactive nature of O3, direct reactions most likely involve components
32 of the ELF and/or plasma membranes of surface macrophages which extend beyond the ELF.
33 Reaction products likely mediate O3 effects on respiratory tract epithelium.
34 Another key event in the O3 toxicity pathway is the activation of neural reflexes which leads to
35 decrements in pulmonary function. Evidence is accumulating that secondary oxidation products are
36 responsible for this effect. Eicosanoids have been implicated in humans while eicosanoids and
37 aldehydes are effective in animal models. Different receptors on bronchial C-fibers have been shown
March 2011 2-17 DRAFT - DO NOT CITE OR QUOTE
-------
1 to mediate separate effects of O3 on pulmonary function. Nociceptor sensory nerves are involved in
2 the involuntary truncation of respiration which results in decreases in forced vital capacity (FVC),
3 FEVi, tidal volume and an increase in respiratory frequency and pain upon deep inspiration. Opioids
4 block these responses while atropine does not. New evidence in an animal model suggests that
5 TRPA1 receptors on bronchial C-fibers mediate this pathway. Ozone exposure also results in
6 activation of vagal sensory nerves and a mild increase in airways obstruction measured as increased
7 specific airway resistance (sRaw). Atropine and beta-adrenergic agonists blocked this response in
8 one study indicating that the airway obstruction was due to bronchoconstriction. Other studies in
9 humans implicated SP release from bronchial C-fibers resulting in airway narrowing due to either
10 neurogenic edema or bronchoconstriction. New evidence in an animal model suggests that the SP-
11 NK receptor pathway caused bronchoconstriction following O3 exposure. Considerable inter-
12 individual variability exists in O3 responsiveness measured by decrements in pulmonary function.
13 Further, attenuation of these pulmonary function decrements occurs following O3 exposure for
14 several consecutive days. Mechanisms responsible for these effects are not known but may be related
15 to inherent differences in neural sensitivity.
16 Injury and inflammation are additional key events in the O3 toxicity pathway. Secondary
17 oxidation products have been implicated in a number of these processes. Although there may be
18 inter-species differences with respect to specific mediators, mechanisms involved in the acute
19 responses to O3 include epithelial injury and airways neutrophilia. Longer-term exposures may result
20 in mucus cell metaplasia of nasal epithelium or airways remodeling and fibrosis. Work from several
21 laboratories in humans and animal models suggest that O3 triggers the release of tachykinins such as
22 SP from airway sensory nerves which could contribute to downstream effects including injury and
23 inflammation. New investigations show that O3 exposure leads to the generation of hyaluronan
24 fragments which activate TLR4 and CD44-dependent signaling pathways in macrophages and result
25 in a greater turnover of macrophage populations in the lung. Activation of these pathways occurs
26 later than the acute neutrophilic response suggesting that they may contribute to longer-term effects
27 of O3. The mechanisms involved in clearing O3-provoked inflammation remain to be clarified.
28 Similar to the pulmonary function responses, considerable inter-individual variability exists in O3
29 responsiveness as measured by airways neutrophilia. Further, attenuation of the inflammatory
30 response occurs following O3 exposure for several consecutive days. However evidence suggests
31 that injury may continue despite the dampening of the inflammatory response during repeated
32 exposures. Mechanisms responsible for inter-individual variability and response attenuation, or the
33 lack thereof, are not known. It should be noted that inflammation, as measured by airways
34 neutrophilia, is not correlated with decrements in pulmonary function as measured by spirometry.
35 Consequently, spirometric measures are not a good surrogate for the degree of inflammation in any
36 given individual following O3 exposure. Furthermore, airways neutrophilia may not be a good
37 indicator of O3-mediated lung injury.
38 Increased bronchial reactivity is a key event in the toxicity pathway of O3. It can be both a
39 rapidly occurring and persistent response, although adaptation can also occur during multi-day
March 2011 2-18 DRAFT - DO NOT CITE OR QUOTE
-------
1 exposures. Both direct effects on smooth muscle and neurally-mediated effects on smooth muscle
2 have been proposed to contribute to airway hyperresponsiveness (AHR) following O3 exposure.
3 Currently, more evidence has accumulated for the latter mechanism. In humans exposed to O3,
4 atropine was found to block the early AHR response indicating the involvement of cholinergic
5 postganglionic pathways. Inhibition of arachidonic acid metabolism was ineffective in blocking this
6 response in humans while mixed results were found in animal models. Studies in O3-exposed
7 animals have demonstrated a role for SP release from bronchial C fibers in mediating neurally-
8 mediated effects on smooth muscle. Later phases of increased bronchial reactivity may involve the
9 induction of interleukin (IL)-lbeta which in turn upregulates SP production. In guinea pigs,
10 eosinophil-derived major basic protein contributed to the stimulation of cholinergic postganglionic
11 pathways. A novel role for hyaluronan in mediating the later phase effects of O3 has recently been
12 demonstrated. High molecular weight polymers of hyaluronan normally found in the ELF were
13 degraded following O3 exposure in mice. The resulting hyaluronan fragments stimulated AHR in a
14 toll-like receptor (TLR4) and CD44 receptor-dependent manner. Previous work has shown that O3-
15 mediated increases in lung permeability required a functioning TLR4 suggesting a possible
16 relationship between increased epithelial permeability and AHR in this model. Other cytokines and
17 chemokines have been implicated in the AHR response to O3 in animal models.
18 Both older and more recent studies provide insight into the ability of O3 to provoke asthma
19 exacerbations in humans. Greater airways inflammation and/or greater bronchial reactivity have
20 been demonstrated in asthmatics compared to non-asthmatics. This pre-existing inflammation and
21 altered baseline bronchial reactivity may contribute to the enhanced bronchoconstriction seen in
22 asthmatics exposed to O3. Furthermore inflammation may contribute to O3-mediated AHR. Animal
23 studies have demonstrated a role for eosinophil-derived proteins in mediating these effects. Since
24 airways eosinophilia occurs in both allergic humans and allergic animal models, this pathway may
25 underlie the exacerbation of allergic asthma by O3. In addition, differences have been noted in
26 epithelial cytokine expression in bronchial biopsy samples of healthy and asthmatic subjects. ATh2
27 phenotype, indicative of adaptive immune system activation and enhanced allergic responses, was
28 observed before O3 exposure and was increased by O3 exposure in asthmatics. Since eosinophilia is a
29 hallmark of a Th2 phenotype, these findings support links between allergic asthma, sensitivity to O3
30 and adaptive immunity. Studies in humans and animal models also provide evidence for activation of
31 innate immunity by O3. In humans, O3 exposure resulted in increased numbers of airways monocytes
32 and dendritic-like cells. Altered expression of cell surface markers characteristic of innate immunity
33 and antigen presentation was observed on monocytes and macrophages. Recruitment of these
34 activated immune cells could lead to activation of allergen-specific memory T cells in allergic
35 individuals and result in the exacerbation of existing asthma in response to an allergen trigger. In
36 animal studies, O3 exposure primed the innate immune system and led to increased endotoxin-
37 induced AHR by a mechanism involving hyaluronan and TLR4. The exaggerated immune response
38 to O3 + endotoxin could lead to a more pronounced lung injury response to a bacterial trigger.
39 Enhanced bronchial reactivity, airways eosinophilia, Th2 phenotype, recruitment of activated innate
March 2011 2-19 DRAFT - DO NOT CITE OR QUOTE
-------
1 immune cells, and enhanced responsiveness to endotoxin all provide biological plausibility for
2 epidemiologic evidence of asthma exacerbations associated with exposure to O3. Thus, the influx of
3 immunomodulatory cells and the activation of innate and adaptive immunity leads to the
4 exacerbation of asthma and allergic responses which is emerging as a key event in the toxicity
5 path way of O3.
6 Recent studies in humans and animal models also provide evidence that O3 exposure causes
7 induction of AHR and allergic responses. Both activation of innate immunity and promotion of
8 adaptive immunity have been implicated. In humans, O3 exposure resulted in increased numbers of
9 dendritic-like cells and levels of a cytokine associated with dendritic cell activation in the sputum,
10 suggesting the presence of a population of activated dendritic cells which could stimulate naive
11 T-cells to promote the development of asthma. Evidence for activated dendritic cells was also found
12 in glutathione S transferase Ml (GSTM1) null human subjects (Section 5.2.9.1) and in allergen-
13 sensitized animals exposed to O3. In the latter study, O3 acted as an adjuvant for allergic sensitization
14 and the development of AHR by a mechanism involving TLR4. In a different animal model, O3-
15 induced AHR required the presence of NKT cells and IL-17, both of which indicate innate immune
16 system activation. Ozone-induced goblet cell metaplasia has also been demonstrated. These findings
17 suggest that O3 may be capable of causing new onset asthma and allergic responses in humans. Thus,
18 promotion of adaptive immunity and activation of innate immunity leads to the induction of AHR
19 and allergic responses which is emerging as a key event in the toxicity pathway of O3.
20 Both older and more recent studies in animal models provide several mechanisms by which O3
21 exposure could enhance susceptibility to lung infections. Both decreased mucociliary particle
22 clearance and decreased numbers and function of alveolar macrophage have been implicated. Recent
23 studies suggest that O3-mediated oxidation of SP-A oxidation and priming of the innate immune
24 system may contribute to decreased pathogen clearance. Immune dysfunction outside of the lung has
25 also been demonstrated. Thus, immune system modulation is emerging as a key event in the O3
26 toxicity pathway.
27 Studies in animals provide evidence for extrapulmonary effects of O3. Although it was
28 suggested that these effects are directly mediated by secondary oxidation products formed in the lung
29 as a result of O3 exposure, there is no evidence that these species enter the circulation. Alternatively,
30 extrapulmonary effects may be due to activation of neural reflexes or to release of diffusible
31 mediators which may initiate or propagate inflammatory responses in the vascular or systemic
32 compartments. Recent studies suggest that oxidative/nitrosative stress contributes to O3-induced
33 cardiovascular effects. Thus, systemic inflammation and vascular oxidative/nitrosative stress are
34 emerging as key events in the toxicity pathway of O3.
35 Collectively, older and more recent studies provide evidence for mechanisms which may
36 underlie the variability in responsiveness seen among individuals. Certain functional genetic
37 polymorphisms, pre-existing conditions and diseases, lifestages and co-exposures contribute to
38 enhanced susceptibility to O3. Adaptation may also be important, but it is incompletely understood -
39 both in terms of the pathways involved and the resulting consequences.
March 2011 2-20 DRAFT - DO NOT CITE OR QUOTE
-------
1 Overall, key events in the toxicity pathway of O3 have been identified in humans and animal
2 models. They include the formation of secondary oxidation products in the lung, activation of neural
3 reflexes, pulmonary injury and inflammation and increased bronchial reactivity. In addition,
4 evidence is accumulating that immune system modulation may lead to impaired host defense and the
5 exacerbation and/or induction of asthma and allergic responses. Systemic inflammation and vascular
6 oxidative/nitrosative stress may be critical to the extrapulmonary effects of O3.
2.4. Health Effects
7 This section evaluates the evidence from toxicological, controlled human exposure, and
8 epidemiologic studies that examined the health effects associated with short- and long-term exposure
9 to O3. The results from the health studies evaluated in combination with the evidence from
10 atmospheric chemistry and exposure assessment studies contribute to the causal determinations made
11 for the health outcomes discussed in this assessment (Section 1.6.4). In the following sections a
12 discussion of the causal determinations will be presented by exposure duration (i.e., short- or long-
13 term exposure) for the health effects for which sufficient evidence was available to conclude a
14 causal, likely to be causal or suggestive relationship. Although not presented in depth in this chapter,
15 a detailed discussion of the underlying evidence used to formulate each causal determination can be
16 found in Chapters 6 and 7.
2.4.1. Effects of Short-Term Exposure to Ozone
Table 2-1. Summary of causal determinations for short-term exposure to ozone
Outcome Causality Determination
Respiratory Effects Causal Relationship
Cardiovascular Effects Suggestive of a Causal Relationship
Central Nervous System Effects Suggestive of a Causal Relationship
Mortality Likely to be a Causal Relationship
2.4.1.1. Respiratory Effects
17 The 2006 O3 AQCD concluded that there was clear, consistent evidence of a causal
18 relationship between short-term exposure to O3 and respiratory health effects (U.S. EPA, 2006,
19 088089). This causal association was substantiated by the coherence of effects observed across
20 controlled human exposure, epidemiologic, and toxicological studies indicating associations of
21 short-term O3 exposures with a range of respiratory health endpoints from respiratory tract
22 inflammation to respiratory hospital admissions (HA) and ED visits. Across disciplines, acute O3
23 exposures induced or were associated with statistically significant declines in lung function. An
March 2011 2-21 DRAFT-DO NOT CITE OR QUOTE
-------
1 equally strong body of evidence from controlled human exposure and toxicological studies
2 demonstrated O3-induced inflammatory responses, increased epithelial permeability, and airway
3 hyperresponsiveness (both specific and nonspecific). Toxicological studies provided additional
4 evidence for O3-induced impairment of host defenses. Coherent with inflammation and airway
5 hyperresponsiveness, epidemiologic studies consistently demonstrated positive associations of
6 increases in ambient O3 concentrations with increases in respiratory symptoms and asthma
7 medication use in asthmatic children and with respiratory-related hospital admissions and asthma-
8 related emergency department (ED) visits. Although O3 was consistently associated with
9 nonaccidental and cardiopulmonary mortality, the contribution of respiratory causes to these findings
10 was uncertain.
11 Building on the strong body of evidence presented in the 2006 AQCD, recent studies continue
12 to support associations between short-term O3 exposure and respiratory effects. In young healthy
13 adults exposed to O3 for 6.6 hours, studies demonstrate mean FEVi decrements of about 3% at
14 60 ppb, 5% at 70 ppb, and 6-8% at 80 ppb (Section 6.2.1.2). These studies also show considerable
15 intersubject variability in responsiveness to O3, with the percentage of subjects with >10%
16 decrement in FEVi increasing with increasing concentration of O3 exposure. The proportion
17 (uncorrected for filtered air [FA] responses) of individuals with >10% FEVi decrements ranges from
18 3 to 20% at an average O3 exposure level of 60 ppb and from 17 to 29% at 80 ppb.
19 The collective body of epidemiologic evidence demonstrates associations between ambient O3
20 and decrements in lung function, although recent studies contributed more mixed evidence. A
21 notable difference among newer studies is the limited investigation of populations engaged in
22 outdoor recreation, exercise, or work, which contributed to the strength of evidence in previous
23 AQCDs (U.S. EPA, 1996, 017831: U.S. EPA, 2006, 088089). Some recent evidence suggests that
24 public attention to daily AQI may be reducing exposures of some groups. Recent epidemiologic
25 studies contributed insight into susceptibility factors for O3-associated respiratory morbidity. Among
26 subjects with atopy, asthmatics with concurrent respiratory infection, elderly with AHR or obesity, or
27 groups with diminished antioxidant enzyme activity, lung function responses to ambient O3
28 exposures generally were exacerbated. The susceptibility of these populations is supported by
29 extensive laboratory evidence (human and animal) for O3-induced exacerbation of allergic
30 inflammation, increased susceptibility to bacterial and viral infections, exacerbation of O3-induced
31 AHR by obesity, and modulation of O3 effects by the oxidative stress/antioxidant balance. In recent
32 controlled human exposure studies, lung function responses to O3 are enhanced in subjects with
33 higher body mass index (BMI).
34 As with lung function, recent controlled human exposure studies demonstrate increases in
35 respiratory symptoms in healthy, young adults following 5.6- to 6.6-h exposure to O3 at levels
36 <80 ppb. The collective body of epidemiologic studies strongly demonstrates positive associations of
37 ambient O3 exposure with respiratory symptoms and asthma medication use among asthmatic
38 subjects, especially in populations with additional susceptibility factors such as asthmatics with
39 atopy, asthmatics with diminished antioxidant enzyme activity, or infants with asthmatic mothers.
March 2011 2-22 DRAFT - DO NOT CITE OR QUOTE
-------
1 Recent studies in animals and in vitro models also continue to demonstrate O3-induced lung
2 injury and inflammatory responses. Building on the extensive experimental evidence, new
3 epidemiologic evidence emerged for ambient O3-associated increases in mediators of inflammation
4 measured in upper and lower airway samples, including eNO, cytokines such as IL-6 or IL-8, and
5 inflammatory cells such as eosinophils. Epidemiologic studies also report associations of increases in
6 ambient O3 with decreased levels of glutathione and increased levels of malondialdehyde in airways.
7 At the time of the 2006 O3 AQCD, controlled human studies of dietary antioxidant supplementation
8 had shown some protective effects of alpha-tocopherol and ascorbate on lung function from O3
9 exposure, but not on the intensity of subjective symptoms and inflammatory response. More recent
10 evidence indicates that diminished activity of oxidant metabolizing enzymes (e.g., GSTM1, GSTP1)
11 or intake of antioxidant vitamins influences inflammatory responses to O3 exposure. Across all three
12 disciplines, evidence suggests a role antioxidant defenses in modulating responses to O3.
13 Recent epidemiologic studies build upon the strong body of evidence that demonstrated
14 consistent positive associations between daily changes in O3 exposure and respiratory-related
15 hospital admissions and ED visits by demonstrating associations in diverse populations across the
16 U.S., Canada, and Europe. In all-year analyses, recent multicity studies and a multicontinent study
17 found an approximate 1.6-5.4% increase in all respiratory-related hospital admissions and ED visits
18 for standardized increases in ambient O3 concentrations1. Positive associations persisted in analyses
19 restricted to the summer season, but the magnitude varied depending on the study location.
20 Compared with studies reviewed in the 2006 O3 AQCD, more recent studies examine associations
21 between short-term O3 exposure and specific respiratory outcomes. Although still limited in number,
22 both single- and multicity studies found consistent, positive associations of daily changes in O3
23 concentrations with asthma and chronic obstructive pulmonary disease (COPD) hospital admissions
24 and ED visits. Evidence was more limited for pneumonia. Consistent with the conclusions of the
25 2006 O3 AQCD (U.S. EPA, 2006, 088089). in studies that conducted seasonal analyses, larger effects
26 were estimated for the warm season or summer months than for the cold season or for all seasons,
27 particularly for asthma and COPD. Although the current body of evidence did not include detailed
28 age-stratified results, the increased risk of asthma hospital admissions observed for children provided
29 additional support for the conclusion from the 2006 O3 AQCD that children are particularly
30 susceptible to O3-induced respiratory effects (U.S. EPA, 2006, 088089). Among studies that
31 evaluated the potential confounding effects of co-pollutants, O3 effect estimates for respiratory-
32 related hospital admissions and ED visits remained relatively robust upon the inclusion of PM and
33 gaseous pollutants in two-pollutant models. Although the concentration-response relationship
34 between short-term O3 exposure and respiratory-related hospital admissions and ED visits has not
35 been extensively examined, preliminary examinations found no evidence of a threshold between
36 short-term O3 exposure and asthma hospital admissions and pediatric asthma ED visits.
1 Effect estimates were standardized to a 20-ppb increase for 24-h avg O3, a 30-ppb increase for 8-h max O3, and a 40 ppb increase for 1-h
max O3.
March 2011 2-23 DRAFT - DO NOT CITE OR QUOTE
-------
1 New evidence extends the potential continuum of well-established O3-associated respiratory
2 effects (e.g., airway inflammation; impaired host defense; lung function decrements; and respiratory
3 symptoms, ED visits, and hospital admissions) by demonstrating associations between ambient O3
4 exposure and respiratory-related mortality. The multicontinent APHENA study reported primarily
5 positive associations with respiratory mortality in all-year analyses, with stronger associations
6 observed in analyses restricted to the summer season. These findings were supported by U.S. and
7 European multicity studies, in which a majority of respiratory mortality effect estimates ranged from
8 a 2.3 to 6.8% increase per standardized increase in ambient O3 concentrations. Although co-pollutant
9 confounding was not extensively examined, the O3-respiratory mortality relationship was moderately
10 to substantially sensitive (e.g., increased or attenuated) to inclusion of PMi0 in co-pollutant models.
11 However, interpretation of these results requires caution due to the limited PM datasets used in these
12 studies.
13 In summary, new studies evaluated in the current review support or expand upon the strong
14 body of evidence presented in the 2006 O3 AQCD that short-term O3 exposure is causally associated
15 with respiratory health effects. Recent controlled human exposure studies demonstrate decreases in
16 FEVi in the range of 2.8 to 3.6% with prolonged O3 exposures (6.6 hours) as low as 60 ppb in
17 concentration. By demonstrating O3-induced airway hyperresponsiveness, activation of neural
18 reflexes, allergic responses, lung injury, impaired host defense, and airway inflammation,
19 toxicological studies have characterized O3 modes of action and provided biological plausibility for
20 epidemiologic observations of associations of ambient O3 exposure with decreases in lung function
21 and increases in respiratory symptoms. The coherence of results across studies for O3-associated
22 changes in lung function, airway inflammation, and respiratory symptoms, in turn, provides the
23 biological plausibility for epidemiologic findings of consistently positive associations of ambient O3
24 exposure with respiratory hospital admissions and ED visits in diverse populations across the U.S.,
25 Europe, and Canada. Additionally, a multicontinent study and several multicity studies reported
26 positive associations between ambient O3 exposures and respiratory mortality. New epidemiologic
27 studies provide evidence for associations of ambient O3 exposure with biological markers of airway
28 inflammation and oxidative stress and indicated that groups with diminished antioxidant capacity or
29 comorbidities such as atopy, AHR, or obesity may have increased susceptibility to respiratory
30 morbidity associated with O3 exposure. This new information is consistent with previously available
31 toxicological and clinical evidence as well as current information on modes of action. A common
32 observation among epidemiologic studies of respiratory morbidity and mortality was stronger
33 associations in analyses restricted to warm seasons compared to cold seasons. Additionally, although
34 co-pollutant confounding was evaluated infrequently, O3 effect estimates generally remained
35 statistically significant in co-pollutant models with PM2 5, PMi0, or NO2. Collectively, the evidence
36 integrated across controlled human exposure, epidemiologic, and toxicological studies as well as
37 across the spectrum of respiratory health endpoints continues to demonstrate that there is a causal
38 relationship between short-term Os exposure and respiratory health effects.
March 2011 2-24 DRAFT - DO NOT CITE OR QUOTE
-------
2.4.1.2. Cardiovascular Effects
1 In past O3 AQCDs, the effects of O3 to the cardiovascular system could not be thoroughly
2 evaluated due to the paucity of information available. However, in recent years, investigation of O3-
3 induced cardiovascular events has advanced. In general, compared with the epidemiologic evidence,
4 the toxicological evidence is more supportive of an O3-induced cardiovascular effect. Epidemiologic
5 evidence does not consistently demonstrate a positive relationship between short-term O3 exposure
6 and cardiovascular morbidity. However, most epidemiologic studies have not extensively
7 investigated the cardiovascular effects of O3 exposure in susceptible populations, which may further
8 support the toxicological findings. Although the epidemiologic evidence is limited, single-city
9 studies reviewed in the 2006 O3 AQCD (U.S. EPA, 2006, 088089). recent multicity studies, and the
10 multicontinent APHENA study provide evidence of consistent positive associations between short-
11 term O3 exposure and cardiovascular mortality. However, in contrast with respiratory effects, there is
12 weak coherence between the associations for cardiovascular morbidity and mortality. Further, there
13 is no apparent biological mechanism to explain the association observed between short-term O3
14 exposure and cardiovascular mortality and not for cardiovascular morbidity.
15 Animal toxicological studies (O3 concentration 0.5-0.8 ppm) provide evidence for O3-induced
16 cardiovascular effects, specifically enhanced ischemia/reperfusion injury, disrupted NO-induced
17 vascular reactivity, decreased cardiac function, and increased HRV. The observed increase in HRV is
18 supported by a recent controlled human exposure study that also finds increased high frequency
19 HRV, but not altered blood pressure, following O3 exposure. Toxicological studies investigating the
20 role of O3 in heart rate regulation are mixed with both bradycardic and tachycardic responses
21 observed. These changes in cardiac function provide evidence for O3-induced alterations in the
22 autonomic nervous system leading to cardiovascular complications. Epidemiologic studies showing
23 positive associations between O3 and arrhythmias confirm the development of autonomic
24 dysfunction following O3 exposure. It is still uncertain how O3 inhalation may cause systemic
25 toxicity, however the cardiovascular effects of O3 found in animals correspond to the development
26 and maintenance of a extrapulmonary oxidative, proinflammatory environment.
27 Overall, animal toxicological studies provide stronger evidence for O3 exposure leading to
28 cardiovascular morbidity than epidemiologic evidence which observed a lack of coherent evidence.
29 Based on the relatively strong body of toxicological evidence, and the consistent evidence of an
30 association between O3 and cardiovascular mortality, but weak coherence and biological plausibility
31 for O3-induced cardiovascular morbidity, the generally limited body of evidence is suggestive of a
32 causal relationship between relevant short-term exposures to O3 and cardiovascular effects.
2.4.1.3. Central Nervous System Effects
33 In rodents, O3 exposure has been shown to cause physicochemical changes in the brain
34 indicative of oxidative stress and inflammation. Recent toxicological studies add to earlier evidence
35 that acute exposures to O3 can produce a range of effects on the central nervous system (CNS) and
March 2011 2-25 DRAFT - DO NOT CITE OR QUOTE
-------
1 behavior. Previously observed effects, including neurodegeneration, alterations in neurotransmitters,
2 short- and long-term memory, and sleep patterns, have been further supported by recent studies. In
3 instances where pathology and behavior are both examined, animals exhibit decrements in behaviors
4 tied to the brain regions or chemicals found to be affected or damaged. For example, damage in the
5 hippocampus, which is important for memory acquisition, was correlated with impaired performance
6 in tests designed to assess memory. Thus the brain is functionally affected by O3 exposure. The
7 single epidemiologic study conducted showed that O3 affects memory in humans as well, albeit on a
8 long-term exposure basis. Notably, exposure to O3 levels as low as 0.25 ppm has resulted in
9 progressive neurodegeneration and deficits in both short- and long-term memory in rodents.
10 Additionally, changes in the CNS, including biochemical, cellular, and behavioral effects, have been
J ? O "O "" "
11 observed in animals whose sole exposure occurred in utero, at levels as a low as 0.3 ppm. Although
12 evidence from epidemiologic and controlled human exposure studies is lacking, the toxicological
13 evidence for the impact of O3 on the brain and behavior is strong, and is suggestive of a causal
14 relationship between O3 exposure and adverse CNS effects.
2.4.1.4. Mortality
15 The 2006 O3 AQCD concluded that the overall body of evidence was highly suggestive that O3
16 directly or indirectly contributes to non-accidental and cardiopulmonary-related mortality, but
17 additional research was needed to more fully establish underlying mechanisms by which such effects
18 occur. The evaluation of new multicity studies that examined the association between short-term O3
19 exposure and mortality found evidence which supports the conclusions of the 2006 O3 AQCD. These
20 new studies reported consistent positive associations between short-term O3 exposure and total
21 (nonaccidental) mortality, with associations being stronger during the warm season, as well as
22 additional support for associations between O3 exposure and cardiovascular mortality being similar
23 or larger in magnitude compared to respiratory mortality. Additionally, these new studies examined
24 previously identified areas of uncertainty in the O3-mortality relationship.
25 Recent studies further examined potential confounders (i.e., co-pollutants and seasonality) of
26 the O3-mortality relationship (Section 6.6.2.1). Unlike previous studies that were limited to primarily
27 examining the confounding effects of PMi0, these studies expanded their analyses to include multiple
28 PM indices (i.e., PMi0, PM2 5, and PM components). Co-pollutant models found evidence that
29 associations between O3 and total mortality were robust to the inclusion of PMi0 or PM2s, while
30 other studies found evidence for a modest reduction (-20-30%) when examining PMi0 and by age
31 group or cause-specific mortality (i.e., cardiovascular). Additional reductions in O3-mortality risk
32 estimates were also observed when examining PM components, specifically sulfate, in co-pollutant
33 models. Overall, the impact of PM indices on O3-mortality risk estimates was found to be much
34 smaller than the variation in O3-mortality risk estimates across cities. Although some studies suggest
35 that O3-mortality risk estimates may be confounded by PM or its chemical components, the
36 interpretation of these results requires caution due to the limited datasets used in these studies. When
37 examining the potential for seasonal confounding of the O3-mortality relationship it was observed
March 2011 2-26 DRAFT - DO NOT CITE OR QUOTE
-------
1 that the extent of smoothing or the methods used for adjustment can influence O3 risk estimates
2 because of the opposing seasonal trends of O3 and mortality when not instituting enough degrees of
3 freedom (df) to control for temporal/seasonal trends.
4 The multicity studies evaluated in this review also examined the potential regional
5 heterogeneity in O3-mortality risk estimates (Section 6.6.2.2). These studies provide evidence which
6 suggests generally higher O3-mortality risk estimates in northeastern U.S. cities with some regions
7 showing no associations between O3 exposure and mortality (e.g., Southwest, urban Midwest).
8 Multicity studies that examined individual- and community-level characteristics identified factors
9 that may explain the observed regional heterogeneity in O3-mortality risk estimates as well as
10 characteristics of populations potentially susceptible to O3-related health effects. An examination of
11 community-level characteristics found an increase in the O3-mortality risk estimates in cities with
12 higher unemployment, percentage of the population Black/African-American, percentage of the
13 working population that uses public transportation, lower temperatures, and lower prevalence of
14 central air conditioning.
15 Additional studies were evaluated that examined factors, such as multi-day effects, mortality
16 displacement, adaptation, and whether a threshold exists in the O3-mortality relationship, which may
17 influence the shape of the O3-mortality concentration-response (C-R) curve. An examination of
18 multiday effects in a U.S. and European multicity study found conflicting evidence for mortality
19 displacement, but the evidence suggests that the positive associations between O3 and mortality are
20 observed mainly in the first few days after exposure. Additionally, a U.S. multicity study found
21 evidence of an adaptive response to O3 exposure, with the highest risk estimates earlier in the O3
22 season (i.e., July) and diminished effects later (i.e., August). However, the evidence of adaptive
23 effects has an implication for the interpretation of multi-day effects, and requires further analysis.
24 Analyses that specifically focused on the O3-mortality C-R relationship found no evidence of a
25 threshold, but did observe evidence for potential differences in the C-R relationship across cities.
26 Overall, this evidence supports the 2006 O3 AQCD which concluded that "if a population threshold
27 level exists in O3 health effects, it is likely near the lower limit of ambient O3 concentrations in the
28 U.S." (U.S. EPA, 2006, 088089). Taken together, the body of evidence indicates that there is likely
29 to be a causal relationship between short-term exposures to O3 and all-cause mortality.
March 2011 2-27 DRAFT - DO NOT CITE OR QUOTE
-------
2.4.2. Effects of Long-Term Exposure to Ozone
Table 2-2. Summary of causal determinations for long-term exposure to ozone
Outcome Causality Determination
Respiratory Effects Likely to be a Causal Relationship
Cardiovascular Effects Suggestive of a Causal Relationship
Mortality Suggestive of a Causal Relationship
Reproductive and Developmental Suggestive of a Causal Relationship
Central Nervous System Effects Suggestive of a Causal Relationship
Cancer, Mutagenicity, and Genotoxicity Inadequate to Infer a Causal Relationship
2.4.2.1. Respiratory Effects
1 The epidemiologic studies reviewed in the 2006 O3 AQCD detected no associations between
2 long-term O3 exposures and asthma-related symptoms, asthma prevalence or allergy to common
3 aeroallergens among children after controlling for covariates. Little evidence was available to relate
4 long-term exposure to current ambient O3 concentrations to deficits in the growth of lung function in
5 children. Additionally, limited evidence was available evaluating the relationship between long-term
6 O3 levels and pulmonary inflammation and other endpoints.
7 Recent studies examine the relationship between long-term O3 exposure and new onset asthma
8 in children (Section 7.2.1). Studies have provided evidence for a relationship between different
9 genetic variants (e.g., heme oxygenase [HMOX], GSTs, arginase [ARG]) that, in combination with
10 O3 exposure, are related to new onset asthma. Studies using a cross-sectional design provide support
11 for a relationship between long-term O3 exposure and health effects in asthmatics, including
12 bronchitic symptoms and respiratory-related school absences. Additionally, chronic O3 exposure was
13 related to childhood asthma hospital admissions.
14 Studies of long-term exposure to O3 and pulmonary function effects are inconclusive, with
15 some new epidemiologic studies relating effects at higher exposure levels. Information from
16 toxicological studies in adult and infant non-human primates indicates that long-term exposure to O3
17 during development can result in irreversible morphological changes in the lung along with changes
18 in pulmonary function.
19 The strongest evidence for a relationship between long-term O3 exposure and respiratory
20 morbidity in recent studies demonstrates associations between long-term measures of O3 exposure
21 and new-onset asthma in children and increased respiratory symptom effects in asthmatics. While the
22 evidence may be limited, these U.S. multi-community prospective cohort studies demonstrate that
23 asthma risk is associated with the important relationships between genetic variability, environmental
24 O3 exposure, and behavior. These relationships are complex. The genes evaluated in these studies are
25 both key candidates in the oxidative stress pathway and have been shown to play an important role in
March 2011 2-28 DRAFT - DO NOT CITE OR QUOTE
-------
1 asthma development. Reduced risk for asthma development is reported in some studies in children
2 living in low-O3 communities. Other recent studies provide coherent evidence for long-term O3
3 exposure and respiratory morbidity effects such as first asthma hospitalization and respiratory
4 symptoms in asthmatics. Studies considering other pollutants provide data supporting the notion that
5 the effects related to O3 are independent from potential effects of the other pollutants. Some studies
6 provide evidence for a positive concentration-response relationship. Generally, the epidemiologic
7 and toxicological evidence provides a compelling case that supports the hypothesis that a
8 relationship exists between long-term exposure to ambient O3 and measures of respiratory morbidity.
9 The 2006 O3 AQCD concluded the evidence was suggestive but inconclusive at that time. Building
10 upon that evidence, the more recent epidemiologic evidence, combined with toxicological studies in
11 rodents and non-human primates, provides biologically plausible evidence that there is likely to be
12 a causal relationship between long-term exposure to O3 and respiratory morbidity.
2.4.2.2. Cardiovascular Effects
13 Previous AQCDs did not address the cardiovascular effects of long-term O3 exposure. The
14 evidence remains limited; however the emerging data is supportive of a role for O3 in chronic
15 cardiovascular diseases. Two epidemiologic studies have investigated cardiovascular morbidity after
16 long-term O3 exposure and both assessed cardiovascular disease related biomarkers. A study of the
17 relationship between O3 and cardiovascular mortality reported no association after adjustment for
18 PM25 levels. Additional epidemiologic studies on cardiovascular morbidity and mortality after long-
19 term exposure have not been published.
20 Toxicological evidence of long-term O3 exposure is also limited but three strong toxicological
21 studies have been published since the previous AQCD. These studies provide evidence for O3
22 enhanced atherosclerosis and ischemia/reperfusion injury, corresponding with development of a
23 systemic oxidative, proinflammatory environment (Section 7.3.1.2). Although questions exist for
24 how O3 inhalation causes systemic effects, a recent study proposes a mechanism for development of
25 vascular pathology that involves activation of LOX-1 by O3 oxidized lipids. This activation may also
26 be responsible for O3 induced changes in genes involved in proteolysis, thrombosis, and
27 vasoconstriction. Taking into consideration the positive toxicological studies reported, the generally
28 limited body of evidence is suggestive of a causal relationship between relevant long-term
29 exposures to O3 and cardiovascular effects.
2.4.2.3. Reproductive and Developmental Effects
30 The 2006 O3 AQCD concluded that the limited number of studies that investigated O3
31 demonstrated no associations between O3 and birth outcomes, with the possible exception of birth
32 defects. The current review included an expanded body of evidence of the associations between O3
33 and reproductive and developmental effects. Recent toxicological studies provide evidence for an
34 effect of prenatal exposure to O3 on ultrastructural changes in bronchiole development, alterations in
35 placental and offspring cytokines, and increased offspring airway hyper-reactivity (Section 7.4.7).
March 2011 2-29 DRAFT - DO NOT CITE OR QUOTE
-------
1 Also, there is limited toxicological evidence for an effect of prenatal and early life exposure on
2 central nervous system effects, including laterality, brain morphology, neurobehavioral
3 abnormalities, and sleep aberration (Section 7.4.8). Recent epidemiologic studies have begun to
4 explore the effects of O3 on sperm quality, and provide limited evidence for decrements in sperm
5 concentration, while there is limited toxicological evidence for testicular degeneration caused by O3
6 (Section 7.4.2). There is no evidence that prenatal or early life O3 concentrations are associated with
7 infant mortality.
8 Collectively, there is limited though positive toxicological evidence for O3-induced
9 developmental effects, including effects on pulmonary structure and function and central nervous
10 system effects. Limited epidemiologic evidence exists for an association with O3 concentration and
11 decreased sperm concentration. A recent toxicological study provides limited evidence for a possible
12 biological mechanism (histopathology showing impaired spermatogenesis) for such an association.
13 Additionally, though the evidence for an association between O3 concentrations and adverse birth
14 outcomes is generally inconsistent, there are several influential studies that indicate an association
15 with reduced birth weight and restricted fetal growth. Taking into consideration the positive evidence
16 for developmental and reproductive outcomes from toxicological and epidemiological studies, and
17 the few influential birth outcome studies, the evidence is suggestive of a causal relationship
18 between long-term exposures to Os and reproductive and developmental effects.
2.4.2.4. Central Nervous System Effects
19 Recent research in the area of O3-induced neurotoxicity has included several long-term
20 exposure studies. Notably, the first epidemiologic study to examine the relationship between O3
21 exposure and neurobehavioral effects observed an association between annual O3 levels and an
22 aging-related cognitive performance decline in tests measuring coding ability and attention/short-
23 term memory. This observation is supported by studies in rodents which demonstrate oxidative stress
24 in the brain and associated decrements in behavioral tests, including those measuring memory, after
25 subchronic exposure to 0.25 ppm O3. Additionally, neurobehavioral changes are evident in animals
26 whose only exposure to O3 occurred in utero. Collectively, the limited epidemiologic and
27 toxicological evidence is coherent and is suggestive of a causal relationship between Os exposure
28 and adverse CNS effects.
2.4.2.5. Cancer, Mutagenicity and Genotoxicity
29 The 2006 O3 AQCD (U.S. EPA, 2006, 088089) reported that evidence did not support ambient
30 O3 as a pulmonary carcinogen. Since the 2006 AQCD, very few epidemiologic and toxicological
31 studies have been published that examine O3 as a carcinogen, but collectively, study results indicate
32 that O3 may contribute to DNA damage. Overall, the evidence is inadequate to determine if a
33 causal relationship exists between ambient Os exposures and cancer.
March 2011 2-30 DRAFT - DO NOT CITE OR QUOTE
-------
2.4.2.6. Mortality
1 The 2006 O3 AQCD concluded that an insufficient amount of evidence existed "to suggest a
2 causal relationship between chronic O3 exposure and increased risk for mortality in humans"
3 (U.S. EPA, 2006, 088089). Two additional studies have been conducted since the last review, an
4 ecologic study that finds no association between mortality and O3, and a reanalysis of the ACS
5 cohort that specifically points to a relationship between long-term O3 exposure and an increased risk
6 of respiratory mortality (Section 7.7.1). The findings from the reanalysis of the ACS cohort are
7 consistent and coherent with the evidence from epidemiologic, controlled human exposure, and
8 animal toxicological studies for the effects of short- and long-term exposure to O3 on respiratory
9 effects. Additionally, the evidence for short- and long-term respiratory morbidity provides biological
10 plausibility for mortality due to respiratory disease. Collectively, the evidence is suggestive of a
11 causal relationship between long-term Os exposures and all-cause mortality.
2.5. Policy Relevant Considerations
9
.5.1. Potentially Susceptible Populations
12 Upon evaluating the association between short- and long-term exposure to O3 and various
13 health outcomes, studies also attempted to identify populations that are more susceptible1 to O3.
14 These studies did so by conducting stratified epidemiologic analyses; by examining individuals with
15 an underlying health condition in controlled human exposure studies; or by developing animal
16 models that mimic the pathophysiological conditions associated with an adverse health effect. These
17 studies identified a multitude of factors that could potentially contribute to whether an individual is
18 susceptible to O3 (Table 8-1). The examination of susceptible populations to O3 exposure allows for
19 the NAAQS to provide an adequate margin of safety for both the general population and for sensitive
20 populations.
21 The populations identified in Chapter 8 that are most susceptible to O3-related health effects
22 are individuals with influenza/infection, individuals with asthma, and older age groups. There were a
23 small number of studies on influenza/infection but both reported influenza/infection to modify the
24 association between O3 exposure and respiratory effects, with individuals having influenza or an
25 infection being at increased susceptibility. Asthma as a susceptibility factor was supported by
26 controlled human exposure and toxicological studies, as well as some evidence from epidemiologic
27 studies. Most studies comparing age groups reported greater effects of short-term O3 exposure on
28 mortality among older adults. Diet and obesity are also both likely susceptibility factors. Multiple
29 epidemiologic, controlled human exposure, and toxicological studies reported that diets deficient in
1 Populations that have a greater likelihood of experiencing health effects related to exposure to an air pollutant (e.g., O3) due to a variety
of factors including, but not limited to: genetic background, birth outcomes (e.g., low birth weight, birth defects), race, sex, lifestage,
lifestyle (e.g., smoking status, nutrition), preexisting disease, SES (e.g., educational attainment, reduced access to health care), and
characteristics that may modify exposure to O3 (e.g., time spent outdoors)
March 2011 2-31 DRAFT - DO NOT CITE OR QUOTE
-------
1 Vitamins E and C are associated with susceptibility to O3 -related health effects. Similarly, studies of
2 effect measure modification by BMI observed greater O3 -related respiratory decrements for
3 individuals who were obese.
4 Other potential factors [pre-existing conditions (such as COPD and CVD) young age, sex, and
5 multiple genes (such as GSTM1, GSTP1, HMOX-1, NQO1, and TNF-a)] provided some evidence
6 of susceptibility, but further evidence is needed. In addition, examination of modification of the
7 associations between O3 exposure and health effects by SES and race were available in a limited
8 number of studies, and demonstrated possible increased odds of health effects related to O3 exposure
9 among those with low SES and Blacks.
10 Individuals involved in outdoor activities were examined in a recent study but no effect
11 modification was observed. However, previous evidence along with biological plausibility from
12 toxicological and controlled human studies has shown this population to be susceptible to O3-related
13 health effects. The only studies examining effect measure modification by diabetes examined O3
14 exposure and cardiovascular outcomes, but none of the studies reported any change in the
15 association by diabetes.
16 Studies of air conditioning use, physical conditioning, and smoking were conducted but not
17 much evidence was available to determine whether susceptibility to O3-related health effects is
18 present for these factors. Toxicological studies also identified hyperthyroidism and the lifestage of
19 gestation to be factors warranting further examination. Future research on these will provide
20 additional insight into whether these factors affect susceptibility to O3-related health effects.
2.5.2. Lag Structure of Ozone-Morbidity and Ozone-Mortality Associations
21 Epidemiologic studies have attempted to identify the time-frame in which exposure to O3 can
22 impart a health effect. Although O3 exposure-response relationships have traditionally been
23 examined using air quality data for a defined lag period (e.g., 1 day or average of 0-1 days), the
24 relationship can potentially be influenced by a multitude of factors, such as the underlying
25 susceptibility of an individual (e.g., age, pre-existing diseases), which could increase or decrease the
26 lag times observed.
27 An attempt has been made to identify whether certain lag periods are more strongly associated
28 with specific health outcomes. The epidemiologic evidence evaluated in the 2006 O3 AQCD
29 indicated that one of the remaining uncertainties in characterizing the O3-mortality relationship was
30 identifying the appropriate lag structure (e.g., single-day lags versus distributed lag model).
31 Currently, many investigators have chosen to examine the lag structure of associations between O3
32 concentration and health outcome instead of focusing on a priori lag times. This approach is
33 informative because if effects are cumulative, higher overall risks may exist than would be observed
34 for any given single day lag. An examination of lag times used in the epidemiologic studies
35 evaluated in this assessment can provide further insight on the relationship between O3 exposure and
36 morbidity and mortality outcomes.
March 2011 2-32 DRAFT - DO NOT CITE OR QUOTE
-------
2.5.2.1. Ozone-Respiratory Effect Associations
1 Collectively, recent epidemiologic studies of lung function, respiratory symptoms, and
2 biological markers of airway inflammation and oxidative stress examined associations with single-
3 day ambient O3 exposures lagged from 0 to 7 days as well as concentrations averaged over 2 to 19
4 days. Lag 0 and 1 ambient O3 exposures were associated with decreases in lung function and
5 increases in respiratory symptoms, airway inflammation, and oxidative stress. In several studies,
6 multiday averages of O3 exposure were associated with these endpoints, indicating that exposures
7 accumulated over several days may be important or may be subject to less measurement error.
8 Studies have suggested that O3-related respiratory morbidity may occur via multiple
9 mechanisms with varying time courses of action. Many epidemiologic studies simultaneously
10 examined associations of short-term ambient O3 exposure with lung function, respiratory symptoms,
11 and biological markers of airway inflammation and oxidative stress and found inconsistent
12 associations among endpoints whether evaluated at the same or different lags of O3 exposure. In
13 most cases, investigators examined a limited number of O3 exposure lags and did not assign different
14 O3 exposure lags for each endpoint based on hypothesized mechanisms of action. These limitations
15 may explain some of the inconsistencies in associations of O3 with different respiratory health
16 endpoints and may contribute to uncertainty over the important lags of ambient O3 exposure for
17 particular respiratory endpoints.
18 In studies of respiratory hospital admissions and ED visits, results were somewhat sensitive to
19 the lag day selected (i.e., reduced when using a single-day lag and increased when using a distributed
20 lag, up to 0-3 days), though when analyses were restricted to summer months only, the lag period
21 tended to influence the results much less. Overall, among studies that examined a range of single-day
22 lags and multiday averages, evidence did not overwhelmingly point to stronger immediate, delayed,
23 or cumulative effects of O3 exposure on respiratory effects.
2.5.2.2. Ozone-Mortality Associations
24 Epidemiologic studies that focused on the association between short-term O3 exposure and
25 mortality (i.e., all-cause, respiratory and cardiovascular) mostly examined a priori lag structures of
26 either 1 or 0-1 days. Although mortality studies do not often examine alternative lag structures,
27 several recent studies have conducted more extensive analysis of lag structure to investigate
28 "mortality displacement" (i.e., deaths are occurring in frail individuals and exposure is only moving
29 the day of death to a day slightly earlier), with varying results (Section 6.6.2.4). An examination of
30 multi-day effects in a U.S. and European multicity study found conflicting evidence for mortality
31 displacement, but the evidence suggests that the positive associations between O3 and mortality are
32 observed mainly in the first few days after exposure. A study conducted in 48 U.S. cities during the
33 warm season (i.e., June-August) for the years 1989-2000 found that larger risk estimates were
34 observed for distributed lag models compared to the lag 0 day estimates; however, larger risk
35 estimates at lag 0-3 compared to 0-20 provide additional support for O3-induced mortality effects
March 2011 2-33 DRAFT - DO NOT CITE OR QUOTE
-------
1 occurring within the first few days after exposure. Although not a traditional mortality displacement
2 study, the APHENA study found no indication that a distributed lag model with up to a 2-day lag
3 yielded meaningfully larger O3 mortality risk estimates than the lag 0-1 and lag 1 results. Finally, a
4 study conducted in 21 European cities reports that using single-day exposures may overestimate the
5 effects on all-cause and cardiovascular mortality, but underestimated the effects on respiratory
6 mortality. Thus, the results in part suggest evidence of mortality displacement for all-cause and
7 cardiovascular mortality. However, it should be noted that the difference in results observed across
8 studies may be due to the different model specifications used. Collectively, these studies suggest that
9 the positive associations between O3 and mortality are observed mainly in the first few days after
10 exposure. Overall, the evidence suggests that estimating the mortality risk using a single day of O3
11 exposure may underestimate the public health impact, but the extent of multi-day effects appear to be
12 limited to a few days.
2.5.3. Ozone Concentration-Response Relationship
13 An important consideration in characterizing the O3-morbidity and mortality association is
14 whether the C-R relationship is linear across the full concentration range that is encountered or if
15 there are concentration ranges where there are departures from linearity (i.e., nonlinearity). In this
16 ISA studies have been identified that attempt to characterize the shape of the O3 C-R curve along
17 with possible O3 "thresholds" (i.e., O3 levels which must be exceeded in order to elicit a health
18 response). The controlled human exposure and epidemiologic studies that examined the shape of the
19 C-R curve and the potential presence of a threshold have indicated a generally linear C-R function
20 with no indication of a threshold for O3 concentrations greater than 30 or 40 ppb, which corresponds
21 with PRB and the lower bound of O3 concentrations included in the C-R functions.
2.5.3.1. Concentration-Response Relationship Characterized by Controlled Human
Exposure Studies
22 Controlled human exposure studies have provided strong and quantifiable C-R data on the
23 human health effects of O3. The magnitude of respiratory effects in these studies is generally a
24 function of O3 exposure, i.e., the product of concentration (C), minute ventilation (VE), and exposure
25 duration. Recent studies provide evidence for a smooth C-R curve without indication of a threshold
26 in young healthy adults, exposed during moderate exercise for 6.6 hours to O3 concentrations of
27 between 40 and 120 ppb (Figure 6-1).
2.5.3.2. Concentration-Response Relationship Characterized by Epidemiologic
Studies
28 A study examining the C-R relationship found no evidence of a threshold between short-term
29 O3 exposure and pediatric asthma ED visits. One study reports that both quintile and loess dose-
30 response analyses (Figure 6-11) suggest that there are elevated associations with O3 at relatively low
March 2011 2-34 DRAFT - DO NOT CITE OR QUOTE
-------
1 concentrations, between 30 and 40 ppb, with stronger evidence at concentrations of 40 ppb and
2 above. In an additional analysis, using a smooth function the authors examined whether the shape of
3 the C-R curve for short-term exposure to O3 and asthma hospital admissions (i.e., both general and
4 ICU for all ages) is linear. When comparing the curve to a linear fit line the authors found that the
5 linear fit is a reasonable approximation of the C-R relationship between O3 and asthma hospital
6 admissions around and below the current NAAQS (Figure 6-9). Although the C-R relationship
7 between short-term O3 exposure and respiratory-related hospital admissions and ED visits has not
8 been extensively examined, preliminary examinations found no evidence of a threshold between
9 short-term O3 exposure and pediatric asthma ED visits.
10 Evidence associating long-term O3 exposure to first asthma hospital admission in a C-R
11 relationship is provided in a retrospective cohort study (Figure 7.3).
12 Evaluation of the short-term exposure to O3-mortality C-R relationship is difficult due to the
13 highly heterogeneous O3-mortality associations among regions in multicity studies (using log-linear
14 models). In addition, there are numerous issues that may influence the shape of the O3-mortality C-R
15 relationship that warrant examination including: multi-day effects (distributed lags), potential
16 adaptation and mortality displacement (i.e., hastening of death by a short period). Several recent
17 studies applied a variety of statistical approaches to examine the shape of the O3-mortality C-R
18 relationship and whether a threshold exists. These studies did not find any evidence that supports a
19 threshold for the association between short-term exposure to O3 and mortality. It was also
20 demonstrated that the heterogeneity in the O3-mortality relationship across cities (or regions)
21 complicates the interpretation of a combined C-R curve and threshold analysis. Additionally, given
22 the effect modifiers identified in the mortality analyses that are also expected to vary regionally (e.g.,
23 temperature, air conditioning prevalence), a national or combined analysis may not be appropriate to
24 identify whether a threshold exists in the O3-mortality C-R relationship.
25 Additionally, several studies of long-term exposure to O3 and birth outcomes have
26 characterized the C-R relationship. Evidence from the southern California Children's Health Study
27 identified a C-R relationship of birth weight with 24-h avg O3 concentrations averaged over the
28 entire pregnancy that was clearest above the 30 ppb level (Figure 7-4). Relative to the lowest decile
29 of 24-h avg O3, estimates for the next 5 lowest deciles were approximately -40 g to -50 g, with no
30 clear trend and with 95% confidence bounds that included zero. The highest four deciles of O3
31 exposure showed an approximately linear decrease in birth weight, and all four 95% CIs excluded
32 zero, and ranged from mean decreases of 74 grams to decreases of 148 grams. Another study
33 conducted in southern California reported increased risks for cardiac birth defects in a dose-response
34 manner with second-month O3 exposure.
2.6. Integration of Ozone Health Effects
35 This section summarizes the main conclusions of this assessment regarding the health effects
36 of O3 and the concentrations at which those effects are observed. The conclusions from the previous
March 2011 2-35 DRAFT - DO NOT CITE OR QUOTE
-------
1 NAAQS review and the causality determinations from this review are summarized in Table 2-3. This
2 section also integrates the evidence from short- and long-term exposure studies across scientific
3 disciplines (i.e., controlled human exposure studies, toxicology, and epidemiology) in interpreting
4 the health effects evidence that spans from prenatal development to death. The clearest evidence for
5 health effects associated with short-term exposure to O3 is provided by studies of respiratory effects.
6 The combined health effects evidence supports a causal relationship for this outcome. The evidence
7 is also sufficient to infer a relationship that is likely to be causal for short-term exposure to O3 and
8 mortality and long-term exposure to O3 and respiratory effects.
Table 2-3. Summary of evidence from epidemiologic, controlled human exposure, and animal
toxicological studies on the health effects associated with short- and long-term exposure to
ozone
Health Outcome
Conclusions from 2006 O3AQCD
Conclusions from 2011 1st Draft ISA
Short-Term Exposure to O3
Respiratory effects
The overall evidence supports a causal
relationship between acute ambient O3
exposures and increased respiratory morbidity
outcomes.
Causal relationship
Lung function
Results from controlled human exposure
studies and animal toxicological studies
provide clear evidence of causality for the
associations observed between acutes; 24 h)
O3 exposure and relatively small, but
statistically significant declines in lung function
observed in numerous recent epidemiologic
studies. Declines in lung function are
particularly noted in children, asthmatics, and
adults who work or exercise outdoors.
Recent controlled human exposure studies
demonstrated decreases in FE\A| in the range of
2.8 to 3.6% with O3 exposures 6.6 h in duration
and as low as 60 ppb in concentration. The
collective body of epidemiologic evidence
demonstrates associations between acute
ambient O3 exposure and decrements in lung
function, particularly in asthmatics, children, and
adults who work or exercise outdoors.
Airway
hyperresponsiveness
Evidence from human clinical and animal
toxicological studies clearly indicate that acute
exposure to O3 can induce airway
hyperreactivity, thus likely placing atopic
asthmatics at greater risk for more prolonged
bouts of breathing difficulties due to airway
constriction in response to various airborne
allergens or other triggering stimuli.
A limited number of studies have observed
airway hyperresponsiveness in rodents and
guinea pigs after exposure to less than 300 ppb
O3. As previously reported in the 2006 O3 AQCD,
50 ppb O3 induced airway hyperresponsiveness
in certain strains of rats, suggesting a genetic
component.
Pulmonary
inflammation, injury
and oxidative stress
The extensive human clinical and animal
toxicological evidence, together with the limited
available epidemiologic evidence, is clearly
indicative of a causal role for O3 in
inflammatory responses in the airways.
Epidemiologic studies provided new evidence for
associations of ambient O3 with mediators of
airway inflammation and oxidative stress and
indicated that groups with diminished antioxidant
capacity or comorbidities such as atopy, AHR, or
obesity may have increased susceptibility to
respiratory morbidity associated with O3
exposure. Generally, these studies were
conducted in locations where the 8-h max O3
concentration ranged from 31 to 66 ppb.
Respiratory symptoms
and medication use
Young healthy adult subjects exposed in
clinical studies to O3 concentrations a 80 ppb
for 6 to 8 h during moderate exercise exhibit
symptoms of cough and pain on deep
inspiration. The epidemiologic evidence shows
significant associations between acute
exposure to ambient O3 and increases in a
wide variety of respiratory symptoms (e.g.,
cough, wheeze, production of phlegm, and
shortness of breath) and medication use in
asthmatic children.
The collective body of epidemiologic evidence
demonstrates positive associations between
acute exposure to ambient O3 and respiratory
symptoms (e.g., cough, wheeze, production of
phlegm, and shortness of breath) in asthmatic
children. Generally, these studies were
conducted in locations where the 8-h max O3
concentration ranged from 17 to 66 ppb.
March 2011
2-36
DRAFT - DO NOT CITE OR QUOTE
-------
Health Outcome
Conclusions from 2006 O3AQCD
Conclusions from 2011 1st Draft ISA
Lung host defenses
Toxicological studies provided extensive
evidence that acute O3 exposures as low as 80
to 500 ppb can cause increases in
susceptibility to infectious diseases due to
modulation of lung host defenses. A single
controlled human exposure study found
decrements in the ability of alveolar
macrophages to phagocytose microorganisms
upon exposure to 80 to 100 ppb O3.
Recent studies build on prior evidence that O3
can affect multiple aspects of innate and
acquired immunity, including macrophage
function, immune cell proliferation, and natural
killer cell function with acute O3 exposures as
low as 80 ppb.
Allergic and asthma
related responses
Previous toxicological evidence indicated that
O3 exposure skews immune responses toward
an allergic phenotype, and enhances the
development and severity of asthma-related
responses such as AHR.
Further evidence for O3-induced allergic skewing
is provided by a few recent studies in rodents
using exposure concentrations as low as
200 ppb.
Hospital admissions,
ED visits, and
physician visits
Aggregate population time-series studies
observed that ambient O3 concentrations are
positively and robustly associated with
respiratory-related hospitalizations and asthma
ED visits during the warm season.
Strong evidence demonstrated associations of
ambient O3 with respiratory hospital admissions
and ED visits in diverse populations across the
U.S., Europe, and Canada. Generally, these
studies were conducted in locations where the
8-h max O3 concentration ranged from 18 to
60 ppb.
Respiratory Mortality
Aggregate population time-series studies
specifically examining mortality from
respiratory causes were limited in number and
showed inconsistent associations between
acute exposure to ambient O3 exposure and
respiratory mortality.
Recent multicity time-series studies and a
multicontinent study consistently demonstrated
associations between ambient O3 and
respiratory-related mortality visits in diverse
populations across the U.S., Europe, and
Canada. Generally, these studies were
conducted in locations where the 8-h max O3
concentration ranged from 20 to 63 ppb.
Cardiovascular effects
The limited evidence is highly suggestive that
O3 directly and/or indirectly contributes to
cardiovascular-related morbidity, but much
remains to be done to more fully substantiate
the association.
Suggestive of a Causal Relationship
Central nervous system
effects
Toxicological studies report that acute
exposures to O3 are associated with alterations
in neurotransmitters, motor activity, short- and
long-term memory, sleep patterns, and
histological signs of neurodegeneration.
Suggestive of a Causal Relationship
All-cause mortality
The evidence is highly suggestive that O3
directly or indirectly contributes to non-
accidental and cardiopulmonary-related
mortality.
Likely to be a Causal Relationship
Long-term Exposure to O3
Respiratory effects
The current evidence is suggestive but
inconclusive for respiratory health effects from
long-term O3 exposure.
Likely to be a Causal Relationship
New onset asthma
No Studies
Evidence for a relationship between different
genetic variants (HMOX, GST, ARC) that, in
combination with O3 exposure, are related to
new onset asthma. These results were observed
when subjects living in areas where the mean
annual 8-h max O3 concentration was 55.2 ppb,
compared to those who lived where it was
38.4 ppb.
Asthma hospital
admissions
No Studies
Chronic O3 exposure was related to first
childhood asthma hospital admissions in a
positive concentration-response relationship.
Generally, these studies were conducted in
locations where the 8-h max O3 concentration
ranged from 30 to 41 ppb.
March 2011
2-37
DRAFT - DO NOT CITE OR QUOTE
-------
Health Outcome
Conclusions from 2006 O3AQCD
Conclusions from 2011 1st Draft ISA
Pulmonary structure
and function
Epidemiologic studies observed that reduced
lung function growth in children was associated
with seasonal exposure to O3; however, cohort
studies of annual or multiyear O3 exposure
observed little clear evidence for impacts of
longer-term, relatively low-level O3 exposure on
lung function development in children. Animal
toxicological studies reported chronic O3-
induced structural alterations in several regions
of the respiratory tract including the
centriacinar region. Morphologic evidence from
studies using exposure regimens that mimic
seasonal exposure patterns report increased
lung injury compared to conventional chronic
stable exposures.
Evidence for pulmonary function effects is
inconclusive, with some new epidemiologic
studies relating effects at exposure levels from
30 to 65 ppb. Information from toxicological
studies indicates that long-term exposure
(500 ppb) during gestation or development can
result in irreversible morphological changes in
the lung, which in turn can influence pulmonary
function.
Pulmonary
inflammation, injury
and oxidative stress
Extensive human clinical and animal
toxicological evidence, together with limited
epidemiologic evidence available, suggests a
causal role for O3 in inflammatory responses in
the airways
Several epidemiologic and toxicology studies (as
low as 500 ppb) add to observations of
O3-induced inflammation and injury.
Allergic responses
Limited epidemiologic evidence supported an
association between ambient O3 and allergic
symptoms. Little if any information was
available from toxicological studies.
Evidence relates positive outcomes of allergic
response and O3 exposure but with variable
strength for the effect estimates; exposure to O3
may increase total IgE in adult asthmatics.
Allergic antibody levels in rodents were
increased by exposure to O3 concentrations as
low as 200 ppb
Respiratory mortality
Studies of cardio-pulmonary mortality were
insufficient to suggest a causal relationship
between chronic O3 exposure and increased
risk for mortality in humans
A single study demonstrated that exposure to O3
(1-h max 45 to 60 ppb) elevated the risk of
death from respiratory causes and this effect
was robust to the inclusion of PM2.5
Cardiovascular Effects
No studies from previous review
Suggestive of a Causal Relationship
Reproductive and
developmental effects
Limited evidence for a relationship between air
pollution and birth-related health outcomes,
including mortality, premature births, low birth
weights, and birth defects, with little evidence
being found for O3 effects.
Suggestive of a Causal Relationship
Central nervous system
effects
Toxicological studies report that acute
exposures to O3 are associated with alterations
in neurotransmitters, motor activity, short and
long term memory, sleep patterns, and
histological signs of neurodegeneration.
Suggestive of a Causal Relationship
Cancer
Little evidence for a relationship between
chronic O3 exposure and increased risk of lung
cancer.
Inadequate to infer a Causal Relationship
All-cause mortality
There is little evidence to suggest a causal
relationship between chronic O3 exposure and
increased risk for mortality in humans.
Suggestive of a Causal Relationship
2.6.1. Respiratory Health Effects
1 Collectively, there is a vast amount of evidence spanning several decades that supports a
2 causal association between exposure to O3 and a spectrum of respiratory effects, including
3 development of the respiratory system; pulmonary structure, inflammation, injury and function;
4 changes in host defense; new onset asthma and asthma exacerbations; respiratory symptoms; ED
5 visits and hospital admissions for respiratory diseases; and even death attributable to respiratory
6 disease (Figure 2-1). The majority of this evidence is derived from studies investigating short-term
7 exposure (i.e., days to weeks) to O3, although more recent evidence demonstrates that long-term
8 exposure (i.e., months to years) may also be detrimental to the respiratory system.
March 2011
2-38
DRAFT - DO NOT CITE OR QUOTE
-------
Short- and long-term
exposure
6-8 h exposure, moderate
exercise, 0,08 0,18 ppm
D.G fl KA^U.Wtfc, lft.chi.fl |/
adults witli enercise, 0,06-
0.08 ppm
Healthy adults; 6.6 n
exposure to O.OB ppn
Short-term exposure in
children (1-h max ~0.1 ppi
Respiratory
Mortality
ED Visits and
Hospital Admissions
Respiratory Symptoms
and Asthma
Exacerbations
Airways
Hyperresponsiveness
Pulmonary Function
New Onset Asthma
Host Defenses
Short- and long-term exposure;
mean 1-h max O3 concentrations
between 11 and 60 ppb
Short- and long-term exposure,
asthmatic children and healthy
adults
Allergic asthmatic
<;jDjects respond to
allergen challenge 0,25-
0.4 ppm
Children at summer camp
0.12 ppm
Mobile labs with ambient
™ 03 0.14-0.17 ppm
Long-term exposure and new
onset asthma modified by
genetic variants
Pulmonary Inflammation
Development and
Structure of Respiratory
Tract
Figure 2-1. Snapshot of evidence for the spectrum of respiratory health effects associated with
short- and long-term exposure to ozone.
Green=Animal Toxicological Studies; Blue=Controlled Human Exposure Studies;
Purple=Epidemiologic Studies; AM=Alveolar Macrophage.D
March 2011
2-39
DRAFT - DO NOT CITE OR QUOTE
-------
1 Mechanistic evidence for the effect of O3 on the respiratory system was initially characterized
2 in the 1996 O3 AQCD, and identified a variety of lung lipid changes which may be further
3 metabolized to produce numerous biologically active mediators that can affect host defenses, lung
4 function, the immune system and other functions. As summarized in Section 2.3 and fully
5 characterized in Chapter 5 key events in the toxicity pathway of O3 have been identified in humans
6 and animal models. They include the formation of secondary oxidation products in the lung,
7 activation of neural reflexes, pulmonary injury and inflammation and increased bronchial reactivity.
8 In addition, evidence is accumulating that influx of immunomodulatory cells, activation of innate
9 and adaptive immunity, induction of AHR and allergic responses, impairment of host defense,
10 systemic inflammation and vascular oxidative/nitrosative stress may also be critical to the O3
11 toxicity pathway (Figure 2-2).
Mode of Action/Possible Pathways: Respiratory System
Ozone + Respiratory Tract
v
Formation of secondary oxidation products
/ I \
Increased «
-^^ bronchial <
reactivity
(AHR)
Activation of
neural reflexes
'-t Epithelial
permeability
Inflammation
and injury
4/lnspiratory Mild
Capacity bronchoconstriction
\ /
Airways
neutrophilia
Repair or
remodeling
Decrements in pulmonary function
\
Exacerbation/induction of
asthma and allergic responses
Immune system
modulation
Allergic priming
& sensitization
Decreased
pathogen clearance
Respiratory Mortality
uA/rrMf •* Impaired host defense/
HA/ED Visits <- ^piratory tract infections
Figure 2-2. Schematic depicting key events in the ozone toxicity pathway. Solid arrows denote
pathways for which there is greater certainty. Broken arrows represent pathways of
emerging interest.
March 2011
2-40
DRAFT - DO NOT CITE OR QUOTE
-------
1 Recent toxicological studies of long-term exposure to O3 occurring throughout various
2 lifestages, beginning with prenatal and early life exposures, provide novel evidence for effects on
3 development of the respiratory system, including ultrastructural changes in bronchiole development,
4 alterations in placental and offspring cytokines, and increased offspring airway hyper-reactivity
5 (Section 7.4.7). The strongest evidence for O3-induced effects on the developing lung comes from a
6 series of experiments using infant rhesus monkeys repeatedly exposed to 0.5 ppm O3 starting at
7 one month of age. Functional changes in the conducting airways of infant rhesus monkeys exposed
8 to either O3 alone or O3 + antigen were accompanied by a number of cellular and morphological
9 changes. In addition to these functional and cellular changes, significant structural changes in the
10 respiratory tract were observed. Importantly, the O3-induced structural pathway changes persisted
11 after recovery in filtered air for six months after cessation of the O3 exposures. Exposure to O3 has
12 also been associated with similar types of alterations in pulmonary structure in all adult laboratory
13 animal species studied, from rats to monkeys (U.S. EPA, 1996, 017831).
14 In addition to effects on the development and structure of the respiratory tract, there is
15 extensive evidence for the effect of short-term exposure to O3 on pulmonary inflammation. Previous
16 evidence from controlled human exposure studies indicated that O3 causes an inflammatory response
17 in the lungs (U.S. EPA, 1996, 017831). This inflammatory response to O3 was detected after a single
18 1-h exposure with exercise to O3 concentrations of 0.3 ppm; the increased levels of some
19 inflammatory cells and mediators persisted for at least 18 hours. Toxicological studies provided
20 additional evidence for increases in permeability and inflammation in rabbits at levels as low as
21 0.1 ppm O3. Evidence summarized in the 2006 O3 AQCD demonstrated that inflammatory responses
22 were observed subsequent to 6.6 h O3 exposure to the lowest tested level of 0.08 ppm in healthy
23 human adults, while animal toxicological studies provided extensive evidence that short-term
24 (1-3 hours) O3 exposure as low as 0.1-0.5 ppm could cause lung inflammatory responses. The
25 limited epidemiologic evidence demonstrated an association between short-term ambient O3
26 exposure and airway inflammation in children (1-h max O3 of approximately 0.1 ppm). The most
27 recent epidemiologic studies provide additional supporting evidence by demonstrating associations
28 of ambient O3 with mediators of airway inflammation and indicating that groups with diminished
29 antioxidant capacity or comorbidities such as atopy, AHR, or obesity may have increased
30 susceptibility to respiratory morbidity associated with O3 exposure (Sections 6.2.4 and 8.1).
31 The normal inflammatory response in lung tissue is part of host defense that aids in removing
32 microorganisms or particles that have reached the distal airways and alveolar surface. The 1996 O3
33 AQCD concluded that short-term exposure to elevated concentrations of O3 resulted in alterations in
34 these host defense mechanisms in the respiratory system. Specifically, toxicological studies of short-
35 term exposures as low as 0.1 ppm O3 were shown to decrease the ability of alveolar macrophages to
36 ingest particles, and short-term exposures as low as 0.08 ppm for 3 hours prevented mice from
37 resisting infection with streptococcal bacteria, resulting in mortality. Similarly, alveolar macrophages
38 removed from the lungs of human subjects after 6.6 hours of exposure to 0.08 and 0.10 ppm O3
39 resulted in a decreased ability to ingest microorganisms, indicating some impairment of host defense
March 2011 2-41 DRAFT - DO NOT CITE OR QUOTE
-------
1 capability. These altered host defense mechanisms can lead to susceptibility to respiratory infections,
2 which are associated with increased risk of asthma when occurring in early life.
3 In addition to pulmonary inflammation and host defenses, recent epidemiologic evidence has
4 revealed an association between long-term exposure to O3 and new onset asthma (Section 7.2.1).
5 Studies have provided evidence for a relationship between different genetic variants (e.g., HMOX,
6 GST, ARG) that, in combination with O3 exposure, are related to new onset asthma. This is the first
7 time that evidence has extended beyond the association of exposure to O3 and asthma exacerbations
8 to suggest that long-term exposure to O3 may play a role in the development of the disease and
9 contribute to incident cases of asthma.
10 The most commonly observed and strongest evidence for respiratory effects associated with
11 short-term exposure to O3 are increased frequency of breathing and decreased tidal volume (i.e.,
12 rapid, shallow breathing). Previous controlled human exposure studies demonstrated O3-induced
13 decrements in pulmonary function, characterized by alterations in lung volumes and flow and airway
14 resistance and responsiveness for multihour exposures (up to 7 hours) to O3 concentrations as low as
15 0.08 ppm (U.S. EPA, 1996, 017831). A series of mobile laboratory studies of lung function and
16 respiratory symptoms reported pulmonary function decrements at mean ambient O3 concentrations
17 of 0.14 ppm in exercising healthy adolescents and increased respiratory symptoms and pulmonary
18 function decrements at 0.15 ppm in heavily exercising athletes and at 0.17 ppm in lightly exercising
19 healthy and asthmatic subjects. Epidemiologic and animal toxicological evidence is coherent with
20 the results of the controlled human exposure studies, both indicating decrements in lung function
21 upon O3 exposure. Combined statistical analysis of six epidemiologic studies in children at summer
22 camp demonstrated decrements in FEVi of 0.50 mL/ppb with previous hour O3 concentration. For
23 preadolescent children exposed to 120 ppb (0.12 ppm) ambient O3, this amounted to an average
24 decrement of 2.4-3.0% in FEVi. Two key studies of lung function measurements before and after
25 well-defined outdoor exercise events in adults yielded exposure-response slopes of 0.40 and
26 1.35 mL/ppb. Animal toxicological studies reported similar respiratory effects in rats at exposures as
27 low as 0.2 ppm O3 for 3 hours. The 2006 O3 AQCD characterized the controlled human exposure
28 and animal toxicological studies as providing clear evidence of causality for the associations
29 observed between acute (< 24 hours) O3 exposure and relatively small, but statistically significant
30 declines in lung function observed in numerous recent epidemiologic studies. Declines in lung
31 function were particularly noted in children, asthmatics, and adults who work or exercise outdoors.
32 Recent studies in animals and in vitro models described inflammatory and injury responses mediated
33 by toll-like receptors (e.g., TLR4, TL2), receptors for TNF or IL-1, multiple signaling pathways
34 (e.g., p38, JNK, NFKB, MAPK/AP-1), and oxidative stress (Section 6.2.3.3). Recent controlled
35 human exposure studies examined lower concentration O3 exposures (40-80 ppb) and demonstrated
36 that FEVi, respiratory symptoms, and inflammatory responses were affected by O3 exposures of
37 6.6 hours and in the range of 60 to 80 ppb (Section 6.2.1.2). These studies demonstrated decreases in
38 FEVi in the range of 2.8 to 3.6% with O3 exposures 6.6 h in duration and as low as 60 ppb in
39 concentration. Recent epidemiologic studies provide greater insight into subject factors that may
March 2011 2-42 DRAFT - DO NOT CITE OR QUOTE
-------
1 increase susceptibility for O3-associated respiratory morbidity. It was in these potentially susceptible
2 populations (e.g., asthmatics with atopy or concurrent respiratory infection, infants with asthmatic
3 mothers, elderly with AHR or obesity, or groups with diminished antioxidant capacity) that O3-
4 associated decreases in lung function tended to be observed.
5 Ozone exposure has been shown to result in airway hyperresponsiveness (both specific and
6 non-specific), epithelial permeability, and respiratory tract inflammation. Increased airway
7 responsiveness is an important consequence of exposure to O3 because its presence means a change
8 in airway smooth muscle reactivity and implies that the airways are predisposed to narrowing on
9 inhalation of a variety of stimuli (e.g., specific allergens, SO2, cold air). Specifically, acute (2 or
10 3 hours) exposure to 0.25 or 0.4 ppm O3 was found to cause increases in airway responsiveness in
11 response to allergen challenges among allergic asthmatic subjects who characteristically already had
12 somewhat increased airway responsiveness at baseline.
13 In addition to alterations in lung volumes and flow, changes in pulmonary function due to
14 exposure to O3 may be elicited as respiratory symptoms (e.g., coughing, wheezing, shortness of
15 breath). The 1996 O3 AQCD identified an association between respiratory symptoms and increasing
16 ambient O3, particularly among asthmatic children. In the 2006 O3 AQCD, the evidence was
17 extended to include young healthy adult subjects that exhibited symptoms of cough and pain on deep
18 inspiration after exposure to 0.08 ppm O3 for 6-8 hours during moderate exercise. The increase in the
19 incidence of cough was found in controlled human exposure studies as low as 0.12 ppm in healthy
20 adults during 1-3 hours with very heavy exercise and other respiratory symptoms, such as pain on
21 deep inspiration and shortness of breath, were observed at 0.16-0.18 ppm with heavy and very heavy
22 exercise. Previous epidemiologic evidence showed significant associations between acute exposure
23 to ambient O3 and increases in a wide variety of respiratory symptoms (e.g., cough, wheeze,
24 production of phlegm, and shortness of breath) in asthmatic children (U.S. EPA, 2006, 088089).
25 Epidemiologic studies also indicated that acute O3 exposure is likely associated with increased
26 asthma medication use in asthmatic children. Similar to what was observed for pulmonary function,
27 recent epidemiologic studies provided greater insight into subject factors that may increase
28 susceptibility for O3-associated respiratory symptoms. It was in these potentially susceptible
29 populations (e.g., asthmatics with atopy or concurrent respiratory infection, infants with asthmatic
30 mothers, elderly with AHR or obesity, or groups with diminished antioxidant capacity) that O3-
31 associated increases in respiratory symptoms tended to be observed. Additionally, recent evidence
32 provides evidence for an association between long-term exposure to O3 and respiratory symptoms
33 (Section 7.2.2).
34 When respiratory symptoms, asthma exacerbations, or other respiratory diseases become too
35 serious to be cared for at home, they can result in visits to hospital emergency departments (ED) or
36 hospital admissions. The frequency of these types of ED visits and hospital admissions is associated
37 with ambient O3 concentrations. Summertime daily hospital admissions for respiratory causes in
38 various locations of eastern North America consistently reported a relationship with ambient levels
39 of O3 in studies reviewed in the 1996 O3 AQCD. This association remained even when considering
March 2011 2-43 DRAFT - DO NOT CITE OR QUOTE
-------
1 only concentrations below 0.12 ppm O3. The 2006 O3 AQCD concluded that aggregate population
2 time-series studies demonstrate a positive and robust association between ambient O3 concentrations
3 and respiratory-related hospitalizations and asthma ED visits during the warm season. Recent
4 epidemiologic time-series studies included additional multicity and multicontinent studies in which
5 short-term increases in ambient O3 concentrations were consistently associated with increases in
6 respiratory hospital admissions and ED visits across diverse populations, geographic locations, and
7 range of O3 concentrations (Section 6.2.7). There is also recent evidence for an association between
8 respiratory hospital admissions and long-term exposure to O3 (Section 7.2.2).
9 Finally, in very serious cases, O3 exposure may contribute to death from respiratory causes.
10 Recent evidence from several multicity and multicontinent studies demonstrated associations
11 between increases in short-term exposure to ambient O3 concentrations and increases in respiratory
12 mortality (Section 6.6.2.5). Similarly, a study of long-term exposure to ambient O3 concentrations
13 also demonstrated an association between O3 and increases in respiratory mortality (Section 7.7.1).
14 Evidence from these new mortality studies is consistent and coherent with the evidence from
15 epidemiologic, controlled human exposure, and animal toxicological studies for the effects of short-
16 and long-term exposure to O3 on respiratory effects. Additionally, the evidence for short- and long-
17 term respiratory morbidity provides biological plausibility for mortality due to respiratory disease.
18 In summary, recent studies support or build upon the strong body of evidence presented in the
19 1996 and 2006 O3 AQCDs that short-term O3 exposure is causally associated with adverse
20 respiratory health effects. Recent controlled human exposure studies demonstrated decreases in
21 FEVi in the range of 2.8 to 3.6% with O3 exposures 6.6 hours in duration and as low as 60 ppb in
22 concentration. Equally strong evidence demonstrated associations of ambient O3 with respiratory
23 hospital admissions and ED visits in diverse populations across the U.S., Europe, and Canada. Most
24 effect estimates ranged from a 1.4 to 2.9% increase in daily ED visits or hospital admissions and
25 were observed in locations with mean 1-h max O3 concentrations between 11 and 60 ppb. Several
26 multicity and multicontinent studies reported associations between increases in ambient O3
27 concentrations and increases in respiratory mortality. Individual-level epidemiologic studies
28 provided new evidence for associations of ambient O3 with mediators of airway inflammation and
29 oxidative stress and indicated that groups with diminished antioxidant capacity or comorbidities such
30 as atopy, AHR, or obesity may have increased susceptibility to respiratory morbidity associated with
31 O3 exposure. These recent epidemiologic findings provided support for ambient O3 exposure having
32 similar effects and modes of action as those observed in studies with experimental exposures. By
33 demonstrating O3-induced airway hyperresponsiveness, activation of neural reflexes (indicative of
34 decreased pulmonary function), allergic responses, lung injury, impaired host defense, and airway
35 inflammation, toxicological studies have characterized O3 modes of action and have provided
36 biological plausibility for epidemiologic associations of ambient O3 exposure with lung function and
37 respiratory symptoms, hospital admissions, ED visits, and mortality. Together, the evidence
38 integrated across controlled human exposure, epidemiologic, and toxicological studies and across the
March 2011 2-44 DRAFT - DO NOT CITE OR QUOTE
-------
1 spectrum of respiratory health endpoints continues to demonstrate that there is a causal
2 relationship between short-term O3 exposure and respiratory health effects.
3 The strongest evidence for a relationship between long-term O3 exposure and respiratory
4 morbidity in recent studies demonstrates associations between long-term measures of O3 exposure
5 and new-onset asthma in children and increased respiratory symptom effects in asthmatics. While the
6 evidence may be limited, these U.S. multi-community prospective cohort studies demonstrate that
7 asthma risk is associated with the important relationships between genetic variability, environmental
8 O3 exposure, and behavior. Other recent studies provide coherent evidence for long-term O3
9 exposure and respiratory morbidity effects such as first asthma hospitalization and respiratory
10 symptoms in asthmatics. Generally, the epidemiologic and toxicological evidence provides a
11 compelling case that supports the hypothesis that a relationship exists between long-term exposure to
12 ambient O3 and measures of respiratory morbidity. Building upon that evidence, the more recent
13 epidemiologic evidence, combined with toxicological studies in rodents and non-human primates,
14 provides biologically plausible evidence that there is likely to be a causal relationship between
15 long-term exposure to O3 and respiratory morbidity.
2.6.2. Mortality Effects
16 The 2006 O3 AQCD concluded that the overall body of evidence was highly suggestive that
17 short-term exposure to O3 directly or indirectly contributes to non-accidental and cardiopulmonary-
18 related mortality, but additional research was needed to more fully establish underlying mechanisms
19 by which such effects occur. The evaluation of new multicity studies that examined the association
20 between short-term O3 exposure and mortality found evidence which supports the conclusions of the
21 2006 O3 AQCD. These new studies reported consistent positive associations between short-term O3
22 exposure and total (nonaccidental) mortality, with associations being stronger during the warm
23 season, as well as additional support for associations between O3 exposure and cardiovascular
24 mortality being similar or larger in magnitude compared to respiratory mortality. Additionally, these
25 new studies examined previously identified areas of uncertainty in the O3-mortality relationship.
26 Taken together, the body of evidence indicates that there is likely to be a causal relationship
27 between short-term exposures to Os and all-cause mortality.
28 The 2006 O3 AQCD concluded that an insufficient amount of evidence existed "to suggest a
29 causal relationship between chronic O3 exposure and increased risk for mortality in humans"
30 (U.S. EPA, 2006, 088089). Two additional studies have been conducted since the last review, an
31 ecologic study that finds no association between mortality and O3, and a reanalysis of the ACS
32 cohort that specifically points to a relationship between long-term O3 exposure and an increased risk
33 of respiratory mortality. The findings from the reanalysis of the ACS study are consistent and
34 coherent with the evidence from epidemiologic, controlled human exposure, and animal
35 toxicological studies for the effects of short- and long-term exposure to O3 on respiratory effects.
36 Additionally, the evidence for short- and long-term respiratory morbidity provides biological
March 2011 2-45 DRAFT - DO NOT CITE OR QUOTE
-------
1 plausibility for mortality due to respiratory disease. Collectively, the evidence is suggestive of a
2 causal relationship between long-term O3 exposures and mortality.
2.6.3. Cardiovascular Health Effects
3 In past O3 AQCDs the effects of short- and long-term exposure to O3 on the cardiovascular
4 system could not be thoroughly evaluated due to the paucity of information available. However,
5 studies investigating O3-induced cardiovascular events have advanced in the last two decades.
6 Overall, there is limited, mixed evidence for cardiovascular effects in epidemiologic studies. Animal
7 toxicological studies provide more evidence for O3 exposure leading to cardiovascular morbidity.
8 The toxicological studies demonstrate O3-induced cardiovascular effects, specifically enhanced
9 ischemia/reperfusion injury with or without the corresponding development of a systemic oxidative,
10 proinflammatory environment, disrupted NO-induced vascular reactivity, decreased cardiac function,
11 and increased HRV. Taking into consideration the positive toxicological studies reported, the
12 generally limited body of evidence is suggestive of a causal relationship for both relevant short-
13 and long-term exposures to Os and cardiovascular effects.
2.6.4. Central Nervous System Effects
14 In rodents, O3 exposure has been shown to cause physicochemical changes in the brain
15 indicative of oxidative stress and inflammation. Recent toxicological studies add to earlier evidence
16 that short- and long-term exposures to O3 can produce a range of effects on the central nervous
17 system and behavior. Previously observed effects, including neurodegeneration, alterations in
18 neurotransmitters, short- and long-term memory, and sleep patterns, have been further supported by
19 recent studies. In instances where pathology and behavior are both examined, animals exhibit
20 decrements in behaviors tied to the brain regions or chemicals found to be affected or damaged. The
21 single epidemiologic study conducted showed that long-term exposure to O3 affects memory in
22 humans as well. Notably, exposure to O3 levels as low as 0.25 ppm has resulted in progressive
23 neurodegeneration and deficits in both short- and long-term memory in rodents. Additionally,
24 changes in the CNS, including biochemical, cellular, and behavioral effects, have been observed in
25 animals whose sole exposure occurred in utero, at levels as a low as 0.3 ppm. Together the evidence
26 from studies of short- and long-term exposure to O3 is suggestive of a causal relationship between
27 Os exposure and adverse CNS effects.
2.6.5. Reproductive and Developmental Effects
28 There is limited though positive toxicological evidence for O3-induced developmental effects,
29 including effects on pulmonary structure and function and central nervous system effects. Limited
30 epidemiologic evidence exists for an association with O3 concentration and decreased sperm
31 concentration. A recent toxicological study provides limited evidence for a possible biological
32 mechanism (histopathology showing impaired spermatogenesis) for such an association.
March 2011 2-46 DRAFT - DO NOT CITE OR QUOTE
-------
1 Additionally, though the evidence for an association between O3 concentrations and adverse birth
2 outcomes is generally inconsistent, there are several influential studies that indicate an association
3 with reduced birth weight and restricted fetal growth. Overall, the evidence is suggestive of a causal
4 relationship between long-term exposures to Os and reproductive and developmental effects.
2.6.6. Cancer and Mutagenicity and Genotoxicity
5 The 2006 O3 AQCD (U.S. EPA, 2006, 088089) reported that evidence did not support ambient
6 O3 as a pulmonary carcinogen. Since the 2006 AQCD, very few epidemiologic and toxicological
7 studies have been published that examine O3 as a carcinogen, but collectively, study results indicate
8 that O3 may contribute to DNA damage. Overall, the evidence is inadequate to determine if a
9 causal relationship exists between ambient Os exposures and cancer.
2.7. Effects on Vegetation and Ecosystems
10 Chapter 9 presents the most policy-relevant information related to this review of the NAAQS
11 for the effects of O3 on vegetation and ecosystems. This section integrates the key findings from the
12 disciplines evaluated in this current assessment of the O3 scientific literature, which includes plant
13 physiology, biochemistry, whole plant biology, ecosystems and exposure-response.
14 Ozone effects at small scales, such as the leaf of an individual plant, can result in effects at a
15 continuum of larger scales. Figure 2-3 is a simplified diagram of the major pathway through which
16 O3 enters plants and the major endpoints O3 may affect from small to large scales. The sections of
17 Chapter 9 are organized around this paradigm of effects at the cellular and subcellular level followed
18 by consideration of the whole plant and finally, O3 impacts on ecosystem-level processes. Ozone
19 enters leaves through stomata, and can alter stomatal conductance and disrupt CO2 fixation (Section
20 9.4). These effects can change rates of leaf gas exchange, growth and reproduction at the individual
21 plant level (Section 9.5). Those O3-induced effects can translate from the individual plant level to the
22 ecosystem level, and cause changes in ecosystem services, such as C storage, water production,
23 nutrient cycling, and community composition (Section 9.6). The EPA framework for causal
24 determinations described in Chapter 1 has been applied to the body of scientific evidence to
25 collectively examine effects attributed to O3 exposure (Table 2-4).
March 2011 2-47 DRAFT - DO NOT CITE OR QUOTE
-------
Leaf
\7
Plant
\7
Ecosystem
Aboveground process
Belowground process
Water cycling
Light
Cuticle
Epidermis
Vascular
System
Epidermis
Cuticle
Photosynthesis
transpiration
Aboveground
growth
Reproduction;
Foliar Injury
Crop yield
and quality
^-^S^
respiration
Root
growth
Microbial activity
decomposition
Ecosystem productivity ^,
*^s&
r
Insects and wildlife
Competition and biodiversity
Water production
Soil C storage
Nutrient cycling
Figure 2-3. The effects of ozone at leaf, plant and ecosystem scales.
March 2011
2-48
DRAFT - DO NOT CITE OR QUOTE
-------
Table 24. Summary of ozone causal determinations for vegetation
Vegetation and Ecosystem Effects
Reduced Vegetation Growth
Alteration of Vegetation Reproduction
Visible Foliar Injury Effects on Vegetation
Alteration of Leaf Gas Exchange in Vegetation
Reduced Yield and Quality of Agricultural Crops
Reduced Productivity in Terrestrial Ecosystems
Reduced Carbon (C) Sequestration in Terrestrial Ecosystems
Alteration of Terrestrial Ecosystem Water Cycling
Alteration of Below-ground Biogeochemical Cycles
Alteration of Terrestrial Community Composition
and ecosystem effects
Causality Determination
Causal
Causal
Causal
Causal
Causal
Causal
Likely Causal
Likely Causal
Causal
Likely Causal
2.7.1. Mechanisms Governing Response
1 Section 9.4 focuses on the effects of O3 stress on plants and their responses to that stress on the
2 molecular, biochemical and physiological levels. Many of the studies focus on the molecular
3 mechanisms that underlie the observed biochemical and physiological changes observed in many
4 plant species in response to O3 exposure. The results support and strengthen those reported in the
5 2006 O3 AQCD. The most significant change in this section from the 2006 O3 AQCD is the emphasis
6 on molecular mechanisms as new techniques, such as those used in evaluating transcriptomes (total
7 set of RNA transcripts in a particular cell at a particular time) and proteomes (total set of proteins
8 expressed in a particular cell at a particular time), have been utilized to perform very comprehensive
9 analyses of changes in gene transcription and protein expression in plants exposed to O3. These
10 newer molecular studies not only provide very important and wide-ranging information regarding
11 the many mechanisms of plant responses to O3, they also allow for the analysis of interactions
12 between various biochemical pathways which are induced in response to O3. However, many of
13 these studies are conducted in artificial conditions with model plants which are typically exposed to
14 very high, short doses of O3. Therefore, additional work remains to elucidate whether these plant
15 responses are transferable to other plant species exposed to more realistic ambient conditions.
16 Ozone is taken up into leaves through open stomata. Once inside the substomatal cavity, O3 is
17 thought to rapidly react with the aqueous layer surrounding the cell (apoplast) to form breakdown
18 products such as H2O2, superoxide (O2), hydroxyl radicals (HO') and peroxy radicals (HO2'). These
19 radicals may play a role in signaling processes and may also interact with sensitive molecules both
20 outside and inside the cell to cause damage. This process was very comprehensively described in the
21 2006 O3 AQCD and is only summarized in this document in Section 9.4.2.
22 As plants have been shown to respond to O3 exposure very rapidly, this response must result
23 from a plant's ability to sense the presence of O3 and/or its breakdown products and then
24 communicate that information to the nucleus to initiate changes in gene expression. While it remains
March 2011 2-49 DRAFT - DO NOT CITE OR QUOTE
-------
1 unclear what the exact mechanism is by which the plant senses the presence of O3, whether there are
2 multiple simultaneous mechanisms by which O3 can be sensed, and how much variation exists in O3
3 sensing between species and exposure conditions, some progress has been made in the understanding
4 of this process since the 2006 O3 AQCD. Experimental evidence described in Section 9.4.3.1
5 suggests that O3 and/or its breakdown products may be directly sensed by apoplastic receptor
6 proteins (although they have not yet been identified). Additionally, a change in cellular redox state
7 due to plant exposure to O3 could be the manner in which plants sense the presence of the pollutant.
8 Once the plant has sensed the presence of the pollutant, there is much evidence to suggest that
9 mitogen-activated protein kinases (MAPK) play an important role in communicating signals to the
10 nucleus that result in gene expression changes in response to O3. Calcium has also been implicated to
11 play a role in the signal transduction processes. To summarize, the evidence to date suggests there
12 may be several mechanisms by which plants sense the presence of O3 and then communicate this
13 signal to the nucleus to induce changes in gene expression.
14 New technologies have allowed for the evaluation of changes in the entire transcriptome and
15 proteome, rather than analyzing the modification of the expression of individual genes and proteins;
16 the results of these studies are presented in Section 9.4.3.2. While transcriptome and proteome
17 analyses per se were not previously addressed, the 2006 O3 AQCD did provide much information
18 regarding changes in gene expression and protein quantity of individual genes and proteins in
19 O3-treated plants. In the transcriptome and proteome studies described here, O3 exposure conditions
20 (concentration, duration of exposure), plant species and sampling times vary significantly; however,
21 functional classification of the genes and proteins that are either up- or down-regulated by plant
22 exposure to O3 exhibit common trends. In summary, genes involved in plant defense, signaling, and
23 those associated with the synthesis of plant hormones and secondary metabolism are generally up-
24 regulated in plants exposed to O3, while those related to photosynthesis and general metabolism are
25 typically down-regulated. Proteome studies support these results by demonstrating concomitant
26 increases or decreases in the proteins encoded by these genes. The transcriptome and proteome
27 results support and enhance the findings of the 2006 O3 AQCD.
28 The 2006 O3 AQCD included a discussion on the role of phytohormones, including salicylic
29 acid, ethylene and jasmonic acid, in plant response to O3. Many additional studies using microarray
30 technology (used to determine changes in the transcriptome) and a variety ofArabidopsis mutants
31 are described in Section 9.4.3.3 and support the conclusions from the 2006 O3 AQCD. Transcriptome
32 analysis has also illuminated the complex interactions that exist between these hormones to better
33 define plant response to O3. To summarize, the results indicate that while ethylene and salicylic acid
34 are needed to develop O3-induced leaf lesions, jasmonic acid acts antagonistically to ethylene and
35 salicylic acid to limit the spread of the lesions. Abscisic acid, in addition to its role in regulating
36 stomatal aperture, may also act antagonistically to the jasmonic acid signaling pathway. Nitric oxide
37 has also been proposed to play a role in regulating O3-induced changes in gene expression; however,
38 its role is not yet well defined. Changes in phytohormones and the interactions between them reveal
39 some of the complexity of plant responses to an oxidative stressor such as O3.
March 2011 2-50 DRAFT - DO NOT CITE OR QUOTE
-------
1 Antioxidant metabolites, such as ascorbate and glutathione, and the enzymes that regenerate
2 them are a critical part of plant defense responses to oxidative stress. The role of ascorbate, which is
3 located in several cellular compartments and also in the apoplast, was comprehensively evaluated in
4 the 2006 O3 AQCD as a first line of defense (due to its location in the apoplast) against oxidative
5 stress. Ascorbate has also been the focus of studies investigating differences in O3 tolerance between
6 plant species or cultivars/genotypes within one species. While the studies evaluated for the current
7 document support the important role of ascorbate, several studies suggest that ascorbate quantity,
8 especially in the apoplast, is not the primary factor in determining plant tolerance to O3. In summary,
9 antioxidant metabolites and enzymes increase in quantity in plants exposed to O3. In most cases,
10 there is a correlation between the degree to which these defensive systems are induced and the
11 ability of the plant to tolerate exposure to O3. This up-regulation of antioxidant defenses and the
12 need to keep antioxidant metabolites in a reduced state requires a significant shift in C metabolism
13 away from growth and reproduction to sustain the energy needs of the plant for defense.
14 While declines in C fixation as a result of plant exposure to O3 were extensively described in
15 the 2006 O3 AQCD, some recent studies (described in Section 9.4.5.1) of O3-induced declines in
16 photosynthesis have focused also on O3 effects on the light reactions. Declines in the Fv/Fm ratio (a
17 measure of the maximum efficiency of the light reactions of photosynthesis) were observed in
18 several studies using a variety of plant species and exposure conditions. Additionally, O3 increased
19 the coefficient of non-photochemical quenching in several species, an indication that defense and
20 repair mechanisms of a non-photochemical nature are activated in these plants while less absorbed
21 light is being used to drive photosynthesis. This indicates a shift away from photosynthesis to
22 defense, resulting in negative impacts on growth and reproduction.
23 Section 9.4.5.2 evaluates the effects of O3 on respiration. While C assimilation declines in O3
24 exposed plants, respiration is generally up-regulated. These increases in respiration are thought to
25 result from a plant's greater energy needs for defense (maintaining its antioxidant metabolites in a
26 reduced state) and repair. The increased energy needs will negatively impact plant growth and
27 reproduction.
28 Secondary metabolism is most often up-regulated in a variety of species exposed to either
29 acute or chronic O3 exposures as a part of a generalized plant defense mechanism. Changes in gene
30 expression, quantity and activity of enzymes associated with secondary metabolism and alterations
31 in secondary metabolite quantity have been documented in plants exposed to O3. Some secondary
32 metabolites, such as flavonoids and polyamines, are of particular interest as they are known to have
33 antioxidant properties. Investigations on the importance of isoprenes in plant response to O3 have
34 revealed conflicting results; however, there is some evidence to suggest that they may play a
35 protective role. In summary, secondary metabolites increase in quantity in O3-treated plants as part of
36 a generalized plant defense response. Some secondary metabolites are of particular importance in
37 O3-treated plants as they may have antioxidant functions. Increased synthesis of secondary
38 metabolites represents a large energy investment of the plant into defense responses and away from
39 growth and reproduction.
March 2011 2-51 DRAFT - DO NOT CITE OR QUOTE
-------
1 Section 9.4.6 focuses on O3-induced changes in stomatal function. Stomata play a critical role
2 in limiting O3 uptake into the plant by reducing stomatal aperture. Declines in stomatal conductance
3 in response to O3 have been documented for many plant species, and much evidence suggests that
4 this results from increases in intercellular CO2 concentration due to reductions in C fixation.
5 Additionally, sensitivity of some plants to O3 has been related to a sluggish stomatal response, in
6 which plants are unable to close their stomata rapidly in response to O3. To summarize, stomatal
7 response to O3 can help to determine plant sensitivity to the pollutant, and the decreases in stomatal
8 conductance are thought to be related to declines in C fixation rates. Reduced stomatal conductance
9 will decrease rates of C assimilation and lead to diminished growth and reproduction in plants.
2.7.2. Nature of Effects on Vegetation
10 Ambient O3 concentrations have long been known to cause visible foliar injury, decreases in
11 photosynthetic rate, decreases in growth, and decreases in the quality and yield of some plant species
12 (U.S. EPA, 1978, 040586: U.S. EPA, 1984, 029711: U.S. EPA, 1996, 080827: U.S. EPA, 2006,
13 088089). Numerous studies have related O3 exposure to plant responses, with most research effort
14 focused on the growth of tree seedlings and the yield of crops as endpoints. The response of a plant
15 species to O3 exposure depends upon many factors, including genetic characteristics, biochemical
16 and physiological status, and previous and current exposure to other stressors. The associated
17 sections in Section 9.5 focus mainly on studies published since the release of the 2006 O3 AQCD
18 (U.S. EPA, 2006, 088089). However, because much O3 research was conducted prior to the 2006 O3
19 AQCD, the conclusions presented below are collectively based on this ISA as well as the 1978,
20 1986, 1996, and 2006 AQCDs (U.S. EPA, 1978, 040586: U.S. EPA, 1984, 029711: U.S. EPA, 1996,
21 080827: U.S. EPA, 2006, 088089).
2.7.2.1. Effects on Woody and Herbaceous Vegetation
Growth and Biomass Allocation
22 The previous O3 AQCDs concluded that there is strong and consistent evidence that ambient
23 concentrations of O3 decrease growth in numerous plant species across the U.S. Studies published
24 since the last review continue to support that conclusion (Section 9.5.2.1).
25 A recently published meta-analysis of 263 studies reported that current ambient O3
26 concentrations (~40 ppb) significantly decreased annual total biomass growth of forest species by an
27 average of 7%, with potentially greater decreases (11 to 17%) in areas that have higher O3
28 concentrations and as background O3 increases in the future. This meta-analysis demonstrates the
29 coherence of O3 effects across numerous studies and species using a variety of experimental
30 techniques. A study conducted on mature forest trees reported that the cumulative effects of ambient
31 levels of O3 decreased seasonal stem growth by 30-50% for most of the species in a high O3-year in
32 comparison to a low O3-year.
March 2011 2-52 DRAFT - DO NOT CITE OR QUOTE
-------
1 Since the 2006 O3 AQCD, several studies were published based on the Aspen free-air carbon-
2 dioxide/O3 enrichment (FACE) experiment using "free air," O3, and CO2 exposures in a forest in
3 Wisconsin. It was found that O3 caused reductions in total biomass relative to the control in aspen,
4 paper birch, and sugar maple communities during the first seven years of stand development.
5 Overall, the studies at the Aspen FACE experiment were consistent with many of the open-top
6 chamber (OTC) studies that were the foundation of previous O3 NAAQS reviews. These results
7 strengthen our understanding of O3 effects on forests and demonstrate the relevance of the
8 knowledge gained from trees grown in open-top chamber studies.
9 In recent studies, O3 was shown to have either negative, non-significant, or positive effects on
10 root biomass and root:shoot ratio. While the findings of individual studies were mixed, recent meta-
11 analyses have generally indicated that O3 reduced C allocated to roots.
12 For some annual species, particularly crops, the endpoint for an assessment of the risk of O3
13 exposure can be defined as yield or growth, e.g., production of grain. For plants grown in mixtures
14 such as hayfields, and natural or semi-natural grasslands (including native nonagricultural species),
15 endpoints other than production of biomass may be important. Such endpoints include biodiversity
16 or species composition, and measures of plant quality. Effects may also result from competitive
17 interactions among plants in mixed-species communities. Most of the available data on non-crop
18 herbaceous species are for grasslands with many of the recent studies conducted in Europe.
19 Evidence is sufficient to conclude that there is a causal relationship between Os exposure
20 and reduced growth of woody and herbaceous vegetation.
Reproduction
21 Studies during recent decades have demonstrated O3 effects on different stages of plant
22 reproduction (Section 9.5.2.2). Several recent studies published since the 2006 O3 AQCD further
23 demonstrate the effects of O3 on reproductive processes in herbaceous and woody plant species.
24 The impacts of O3 on reproductive development can occur by influencing (1) age at time of
25 initial flowering, particularly in long-lived trees that often have long juvenile periods of early growth
26 without flower and seed production; (2) flower bud initiation and development; (3) pollen
27 germination and pollen tube growth; and (4) seed, fruit, or cone yields and seed quality.
28 Evidence is sufficient to conclude that there is a causal relationship between Os exposure
29 and changes in reproduction of vegetation.
Visible Foliar Injury
30 Visible foliar injury resulting from exposure to O3 has been well characterized and
31 documented over several decades of research on many tree, shrub, herbaceous, and crop species
32 (U.S. EPA, 1978, 040586: U.S. EPA, 1984, 029711: U.S. EPA, 1996, 080827: U.S. EPA, 2006,
33 088089) (Section 9.5.2.3). Ozone-induced visible foliar injury symptoms on certain bioindicator
34 plant species are considered diagnostic as they have been verified experimentally in exposure-
35 response studies, using exposure methodologies such as continuous stirred tank reactors (CSTRs),
March 2011 2-53 DRAFT - DO NOT CITE OR QUOTE
-------
1 OTCs, and free-air fumigation. Experimental evidence has clearly established a consistent
2 association of visible injury with O3 exposure, with greater exposure often resulting in greater and
3 more prevalent injury. Since the 2006 O3 AQCD, several multiple-year field surveys of O3-induced
4 visible foliar injury have been conducted at National Wildlife Refuges in Maine, Michigan, New
5 Jersey, and South Carolina. New sensitive species showing visible foliar injury continue to be
6 identified from field surveys and verified in controlled exposure studies.
7 The use of biological indicators in field surveys to detect phytotoxic levels of O3 is a
8 longstanding and effective methodology. The USDA Forest Service through the Forest Health
9 Monitoring (FHM) Program (1990-2001) and currently the Forest Inventory and Analysis (FIA)
10 Program has been collecting data regarding the incidence and severity of visible foliar injury on a
11 variety of O3 sensitive plant species throughout the U.S. The network has provided evidence that O3
12 concentrations were high enough to induce visible symptoms on sensitive vegetation. From repeated
13 observations and measurements made over a number of years, specific patterns of areas experiencing
14 visible O3 injury symptoms can be identified.
15 In addition, a study assessed the risk of O3-induced visible foliar injury on bioindicator plants
16 in 244 national parks in support of the National Park Service's Vital Signs Monitoring Network. The
17 results of the study demonstrated that the risk of visible foliar injury was high in 65 parks (27%),
18 moderate in 46 parks (19%), and low in 131 parks (54%). Some of the well-known parks with a high
19 risk of O3-induced visible foliar injury include: Gettysburg, Valley Forge, Delaware Water Gap, Cape
20 Cod, Fire Island, Antietam, Harpers Ferry, Manassas, Wolf Trap Farm Park, Mammoth Cave, Shiloh,
21 Sleeping Bear Dunes, Great Smoky Mountains, Joshua Tree, Sequoia and Kings Canyon, and
22 Yosemite.
23 Evidence is sufficient to conclude that there is a causal relationship between ambient Os
24 exposure and the occurrence of O3-induced visible foliar injury on sensitive vegetation across
25 the U.S.
Gas Exchange
26 There is strong experimental evidence over several decades of research that exposure to O3
27 reduces photosynthesis and alters stomatal conductance in a wide variety of plant species. The mode
28 of action, as characterized in Section 9.4 and in previous reviews, provides biological plausibility for
29 O3 effects on leaf gas exchange.
30 In compiling more than 55 studies, a meta-analysis reported that current O3 concentrations in
31 the northern hemisphere are decreasing photosynthesis (11%) and stomatal conductance (13%)
32 across tree species. It was also found that younger trees less than four years old) were affected less
33 by O3 than older trees. Further, the authors also found that decreases in photosynthesis are consistent
34 with the cumulative uptake of O3 into the leaf. In contrast, several studies reported that O3 exposure
35 may result in loss of stomatal control, incomplete stomatal closure at night and a decoupling of
36 photosynthesis and stomatal conductance, which may have implications for whole-plant water use
37 (Section 9.6.3).
March 2011 2-54 DRAFT - DO NOT CITE OR QUOTE
-------
1 Evidence is sufficient to conclude that there is a causal relationship between Os exposure
2 and the alteration of leaf gas exchange in vegetation.
2.7.2.2. Agricultural Crops
Yield and Crop Quality
3 The detrimental effect of O3 on crop production has been recognized since the 1960s and a
4 large body of research has subsequently stemmed from those initial findings. Previous O3 AQCDs
5 have extensively reviewed this body of literature (U.S. EPA, 2006, 088089). Recent experimental
6 studies of O3 effects on crops are discussed in Section 9.5.3 and summarized in Tables 9-3 and 9-16.
7 Current O3 concentrations across the U.S. are high enough to cause yield loss for a variety of
8 agricultural crops including, but not limited to, soybean, wheat, cotton, potato, watermelon, beans,
9 turnip, onion, lettuce, and tomato. Continued increases in O3 concentration may further decrease
10 yield in these sensitive crops while also initiating yield losses in less sensitive crops. Despite the
11 well-documented yield losses due to increasing O3 concentration, there is still a knowledge gap
12 pertaining to the exact mechanism of O3-induced yield loss. Research has linked increasing O3
13 concentration to decreased photo synthetic rates and accelerated senescence, which are related to
14 yield.
15 Recent modeling research has correlated satellite air-column observations with direct air-
16 sampling O3 data and modeled the yield-loss due to O3 over the continuous tri-state area of Illinois,
17 Iowa and Wisconsin. This modeling data correlates well with the previous results from FACE-type
18 experiments and OTC experiments.
19 New research is beginning to consider the mechanism of damage caused by long, lower O3
20 concentration (so-called chronic exposure) compared to short, very high O3 concentration (so-called
21 acute exposure). Both types of O3 exposure cause damage to agricultural crops, but through very
22 different mechanisms. Until recently, most research on the mechanism of O3 damage has used acute
23 exposure studies. It has become clear that the same cellular and biochemical processes involved in
24 the response to acute O3 exposure are not involved in response to chronic O3 exposure, yet both
25 cause yield-loss in agriculturally important crops.
26 In addition, new research has highlighted the effects of O3 on crop quality. Increasing O3
27 concentration decreases nutritive quality of grasses, decreases macro- and micro-nutrient
28 concentrations in fruits and vegetable crops, and decreases cotton fiber quality. These areas of
29 research require further investigation to determine the mechanism and dose-responses.
30 Evidence is sufficient to conclude that there is a causal relationship between Os exposure
31 and reduced yield and quality of agricultural crops.
2.7.2.3. Factors That Modify Functional and Growth Response
32 Many biotic and abiotic factors, including insects, pathogens, root microbes and fungi,
33 temperature, water and nutrient availability, and other air pollutants, as well as elevated CO2,
March 2011 2-55 DRAFT - DO NOT CITE OR QUOTE
-------
1 influence or alter plant response to O3. These modifying factors were comprehensively reviewed in
2 the 2006 O3 AQCD. A limited number of studies published since 2006 provide further support for
3 our understanding of the role of these interactions in modifying O3-induced plant responses and are
4 discussed in Section 9.5.4.
2.7.3. Ecosystems and Services
5 Ozone has been found to alter plant physiological processes such as growth, biomass
6 allocation, reproduction and gas exchange (Section 9.5). Those O3-induced effects at the individual
7 plant scale have the potential to translate to effects at the ecosystem level, and cause changes in
8 biogeochemical cycling and community composition. Information presented in the associated section
9 (Section 9.6) was collected at multiple scales, ranging from responses at the population level to the
10 ecosystem level. The effects of O3 on ecosystem productivity, C sequestration, water cycling,
11 nutrient cycling, and community composition are reviewed.
2.7.3.1. Productivity and Carbon Sequestration
12 During the previous NAAQS reviews, there were very few studies that investigated the effect
13 of O3 exposure on ecosystem productivity and C sequestration. Recent studies from long-term FACE
14 experiments provided evidence of the association of O3 exposure and reduced productivity at the
15 ecosystem level. Elevated O3 reduced stand-level biomass by 13-23% at Aspen FACE after 7 years
16 of O3 exposure, and annual volume growth by 9.5 nvVha at the Kranzberg Forest (Germany) FACE.
17 Studies at the leaf and plant scales showed that O3 reduced photosynthesis and plant growth, which
18 provided coherence and biological plausibility for the decrease in ecosystem productivity. Results
19 across different ecosystem models were consistent with the FACE experimental evidence, which
20 showed that O3 reduced ecosystem productivity.
21 Although O3 generally causes negative effects on plant growth, the magnitude of the response
22 varies among plant communities. For example, O3 had little impact on white fir, but greatly reduced
23 growth of ponderosa pine in southern California. Ozone decreased net primary production (NPP) of
24 most forest types by 7-8% in Mid-Atlantic region, but had small impacts on spruce-fir forest, which
25 was decreased by only 1%. Among crop species, the estimated yield loss for wheat (7-12%) and
26 soybean (6-16%) were higher than rice (3-4%) and maize (3-5%).
27 In addition to plant growth, other indicators that are typically estimated by model studies
28 include net ecosystem CO2 exchange (NEE), C sequestration, and crop yield. Model simulations
29 consistently found that O3 exposure caused negative impacts on those indicators (Section 9.6.2,
30 Table 9-5), but the severity of these impacts was influenced by multiple interactions of biological
31 and environmental factors. For example, the largest O3-induced crop yield losses occurred in high-
32 production areas exposed to high O3 concentrations, such the Midwest and the Mississippi Valley
33 regions of the U.S.
March 2011 2-56 DRAFT - DO NOT CITE OR QUOTE
-------
1 The suppression of ecosystem C sinks results in more CO2 accumulation in the atmosphere.
2 Globally, the indirect radiative forcing, reported in Watts/square meter (W/m2), caused by O3
3 exposure through lowering ecosystem C sink (0.62-1.09 W/m2) could have an even greater impact
4 on global warming than the direct radiative forcing of O3 (0.89 W/m2). Ozone could also affect
5 regional C budgets through interacting with multiple factors, such as N deposition, elevated CO2 and
6 land use history. Model simulations suggested that O3 partially offset the growth stimulation caused
7 by elevated CO2 and N deposition in both Northeast- and Mid Atlantic-region forest ecosystems of
8 the U.S.
9 The evidence is sufficient to infer that there is a causal relationship between Os exposure
10 and reduced productivity, and there is likely to be causal relationship between Os exposure and
11 reduced carbon sequestration in terrestrial ecosystems.
2.7.3.2. Water Cycling
12 Although the evidence was from a limited number of field and modeling studies, these
13 findings showed an association of O3 exposure and the alteration of water cycle at the ecosystem
14 level. Field studies suggested that peak hourly O3 exposure increased the rate of water loss from
15 several tree species, and led to a reduction in the late-season modeled stream flow in three forested
16 watersheds in eastern Tennessee. Evidence of sluggish stomatal responses during O3 exposure was
17 found in their study and several other studies (Section 9.6.3), which provided biological plausibility
18 for the observed higher water loss at the ecosystem level. However, many experiments, mostly based
19 on short-term O3 exposure, found that O3 generally reduced stomatal conductance. The O3-induced
20 reduction in stomatal aperture is the biological assumption for most process-based models.
21 Therefore, results of those models normally found that O3 reduced water loss. For example, one
22 study found that O3 damage and N limitation together reduced evapotranspiration and increase
23 runoff.
24 Although the direction of the response differed among studies, the evidence is sufficient to
25 conclude that there is likely to be a causal relationship between Os exposure and the alteration
26 of ecosystem water production.
2.7.3.3. Below-Ground Processes
27 Since the 2006 O3 AQCD, more evidence has shown that although the responses are often
28 species specific, O3 altered the quality and quantity of C input to soil, microbial community
29 composition, and C and nutrient cycling. Biogeochemical cycling of below-ground processes is
30 driven by C input from plants. Studies at the leaf and plant level have provided biologically plausible
31 mechanisms, such as reduced photosynthetic rates, increased metabolic cost, and reduced root C
32 allocation (Section 9.6.4) for the association of O3 exposure and the alteration of below-ground
33 processes.
March 2011 2-57 DRAFT - DO NOT CITE OR QUOTE
-------
1 Results from Aspen FACE and other experimental studies consistently found that O3 reduced
2 litter production and altered C chemistry, such as soluble sugars, soluble phenolics, condensed
3 tannins, lignin, and macro/micro nutrient concentration in litter. The changes in substrate quality and
4 quantity could alter microbial metabolism under elevated O3, and therefore soil C and nutrient
5 cycling. Several studies indicated that O3 generally suppressed soil enzyme activities. However, the
6 impact of O3 on litter decomposition was inconsistent and varied among species, sites and exposure
7 length. Ozone had small impact on dynamics of micro and macro nutrients, except for N. Ozone was
8 found to reduce N release from leaf litter and decrease gross N mineralization, which could
9 potentially decrease N availability to plants.
10 Studies from the Aspen FACE experiment suggested that the response of below-ground
11 C cycle to O3 exposure, such as litter decomposition, soil respiration and soil C content, changed
12 over time. For example, in the early part of the experiment (1998-2003), O3 had no impact on soil
13 respiration but reduced the formation rates of total soil C under elevated CO2. However, after 10-
14 11 years of exposure, O3 was found to increase soil respiration but have no significant impact on soil
15 C formation under elevated CO2 (Section 9.6.4.1).
16 The evidence is sufficient to infer that there is a causal relationship between Os exposure
17 and the alteration of below-ground biogeochemical cycles.
2.7.3.4. Community Composition
18 In the 2006 O3 AQCD, the impact of O3 exposure on species competition and community
19 composition was assessed. Ozone was found to cause a significant decline in ponderosa and Jeffrey
20 pine in the San Bernardino Mountains in southern California. Ozone-exposure also tended to shift
21 the grass-legume mixtures in favor of grass species (U.S. EPA, 2006, 088089). Since the 2006 O3
22 AQCD, more evidence has shown that O3 exposure changed the competitive interactions and led to
23 loss of O3 sensitive species or genotypes. Studies at plant level found that the severity of O3 damage
24 on growth, reproduction and foliar injury varied among species (Section 9.6.5), which provided the
25 biological plausibility for the alteration of community composition. Additionally, research since the
26 last review has shown that O3 can alter community composition and diversity of soil microbial
27 communities.
28 The decline of conifer forests under O3 exposure was continually observed in several regions.
29 Ozone damage was believed to be an important causal factor in the dramatic decline of sacred fir in
30 the valley of Mexico, as well as cembran pine in southern France and Carpathian Mountains. Results
31 from the Aspen FACE site indicated that O3 could alter community composition of broadleaf forests
32 as well. At the Aspen FACE site, O3 reduced growth and increased mortality of a sensitive aspen
33 clone, while the O3 tolerant clone emerged as the dominant clone in the pure aspen community. In
34 the mixed aspen-birch and aspen-maple communities, O3 reduced the competitive capacity of aspen
35 compared to birch and maple.
36 The tendency for O3-exposure to shift the biomass of grass-legume mixtures in favor of grass
37 species, was reported in the 2006 O3 AQCD (U.S. EPA, 2006, 088089) and has been generally
March 2011 2-58 DRAFT - DO NOT CITE OR QUOTE
-------
1 confirmed by recent studies. However, in a high elevation mature/species-rich grass-legume pasture,
2 O3 fumigation showed no significant impact on community composition.
3 Ozone exposure not only altered community composition of plant species, but also
4 microorganisms. The shift in community composition of bacteria and fungi has been observed in
5 both natural and agricultural ecosystems, although no general patterns could be identified.
6 The evidence is sufficient to conclude that there is likely to be a causal relationship between
7 Os exposure and the alteration of community composition.
2.7.4. Air Quality Indices
8 Exposure indices are metrics that quantify exposure as it relates to measured plant damage
9 (i.e., reduced growth). They are summary measures of monitored ambient O3 concentrations over
10 time intended to provide a consistent metric for reviewing and comparing exposure-response effects
11 obtained from various studies. No new information is available since 2006 that alters the basic
12 conclusions put forth in the 2006 and 1996 O3 AQCDs (U.S. EPA, 1996, 080827: U.S. EPA, 2006,
13 088089). These AQCDs focused on the research used to develop various exposure indices to help
14 quantify effects on growth and yield in crops, perennials, and trees (primarily seedlings). The
15 performance of indices was compared through regression analyses of earlier studies designed to
16 support the estimation of predictive O3 exposure-response models for growth and/or yield of crops
17 and tree (seedling) species.
18 The main conclusions from the 1996 and 2006 O3 AQCDs (U.S. EPA, 1996, 080827:
19 U.S. EPA, 2006, 088089) regarding an index based on ambient exposure are still valid. These key
20 conclusions can be restated as follows:
21 • O3 effects in plants are cumulative;
22 • higher O3 concentrations appear to be more important than lower concentrations in
23 eliciting a response;
24 • plant sensitivity to O3 varies with time of day and plant development stage; and
25 • exposure indices that cumulate hourly O3 concentrations and preferentially weight the
26 higher concentrations have better statistical fits to growth/yield response data than do the
27 mean and peak indices.
28 Various weighting functions have been used, including threshold-weighted (e.g., SUM06) and
29 continuous sigmoid-weighted (e.g., W126) functions. Based on statistical goodness-of-fit tests, these
30 cumulative, concentration-weighted indices could not be differentiated from one another using data
31 from previous exposure studies. Additional statistical forms for O3 exposure indices are summarized
32 in Section 9.7 of this ISA. The majority of studies published since the 2006 O3 AQCD (2006,
March 2011 2-59 DRAFT - DO NOT CITE OR QUOTE
-------
1 088089) do not change earlier conclusions, including the importance of peak concentrations, and the
2 duration and occurrence of O3 exposures in altering plant growth and yield.
3 Given the current state of knowledge and the best available data, exposure indices that
4 cumulate and differentially weight the higher hourly average concentrations and also include the
5 mid-level values continue to offer the most defensible approach for use in developing response
6 functions and comparing studies, as well as for defining future indices for vegetation protection.
2.7.4.1. Modeled Ozone Deposition or "Flux"
7 Another approach for improving risk assessment of vegetation response to ambient O3 is based
8 on determining the O3 concentration from the atmosphere that enters the leaf (i.e., flux or
9 deposition). Interest has been increasing in recent years, particularly in Europe, in using
10 mathematically tractable flux models for O3 assessments at the regional, national, and European
11 scale. While some efforts have been made in the U.S. to calculate O3 flux into leaves and canopies,
12 little information has been published relating these fluxes to effects on vegetation. There is also
13 concern that not all O3 stomatal uptake results in a yield reduction, which depends to some degree on
14 the amount of internal detoxification occurring with each particular species. Those species having
15 high amounts of detoxification potential may, in fact, show little relationship between O3 stomatal
16 uptake and plant response. The lack of data in the U.S. and the lack of understanding of
17 detoxification processes have made this technique less viable for vulnerability and risk assessments
18 in the U.S.
2.7.4.2. Night-Time Exposures
19 A 12-h (8:00 a.m. to 8:00 p.m.) daylight period for a W126 cumulating exposure was proposed
20 in 2007 and 2009 following the release of the 2006 O3 AQCD and was based primarily on evidence
21 that the conditions for uptake of O3 into the plant occur mainly during the daytime hours. Plants have
22 the highest stomatal conductance during the daytime and atmospheric turbulent mixing is greatest
23 then as well (U.S. EPA, 2006, 088089). Recent reviews of the literature reported that a large number
24 of species had varying degrees of nocturnal stomatal conductance. In general, stomatal conductance
25 at night is at a much lower rate compared to daytime conductance. For significant nocturnal stomatal
26 flux and O3 effects to occur, specific conditions must exist. A susceptible plant with nocturnal
27 stomatal conductance and low defense must be growing in an area with relatively high night-time O3
28 and appreciable nocturnal turbulence. It is unclear how many areas there are in the U.S. where these
29 conditions occur. More information is needed in these locations in order to assess the local O3
30 patterns, micrometeorology and responses of potentially vulnerable plant species.
2.7.5. Exposure-Response
31 None of the information on effects of O3 on vegetation published since the 2006 O3 AQCD has
32 modified the assessment of quantitative exposure-response relationships that was presented in that
March 2011 2-60 DRAFT - DO NOT CITE OR QUOTE
-------
1 document (U.S. EPA, 2006, 088089). This assessment updates the 2006 exposure-response models
2 by computing them using the W126 metric, cumulated over 90 days. Almost all of the experimental
3 research on the effects of O3 on growth or yield of plants published since 2006 used only two levels
4 of exposure. In addition, hourly O3 concentration data that would allow calculations of exposure
5 using the W126 scale are generally unavailable. However, two long-term experiments, one with a
6 crop species (soybean), one with a tree species (aspen), have produced data that can be used to
7 validate the exposure-response models presented in the 2006 O3 AQCD, and methodology used to
8 derive them.
9 Quantitative characterization of exposure-response in the 2006 O3 AQCD was based on
10 experimental data generated for that purpose by the National Crop Loss Assessment Network
11 (NCLAN) and EPA National Health and Environmental Effects Research Laboratory, Western
12 Ecology Division (NHEERL-WED) projects, using OTCs to expose crops and trees seedling to O3.
13 In recent years, yield and growth results for two of the species that had provided extensive exposure-
14 response information in those projects have become available from studies that used FACE
15 technology, which is intended to provide conditions much closer to natural environments. The robust
16 methods that were used previously with exposure measured as SUM06 were applied to the NCLAN
17 and NHEERL-WED data with exposure measured as W126, in order to derive single-species median
18 models for soybean and aspen from studies involving different genotypes, years, and locations. The
19 resulting models were used to predict the change in yield of soybean and biomass of aspen between
20 the two levels of exposure reported in current FACE experiments. Results from these new
21 experiments were exceptionally close to predictions from the models. The accuracy of model
22 predictions for two widely different plant species provides support for the validity of the
23 corresponding multiple-species models for crops and trees in the NCLAN and NHEERL-WED
24 projects. However, variability among species in those projects indicates that the range of sensitivity
25 is likely quite wide. This was confirmed by a recent experiment with cottonwood in a naturally
26 occurring gradient of exposure, which established the occurrence of species with responses
27 substantially more severe under currently existing conditions than are predicted by the median model
28 for multiple species.
29 Results from several meta-analyses have provided approximate values for responses of yield
30 of soybean, wheat, rice and other crops under broad categories of exposure, relative to charcoal -
31 filtered air. Additional reports have summarized yield data for six crop species under various broad
32 comparative exposure categories, and reviewed 263 studies that reported effects on tree biomass.
33 However, these analyses have proved difficult to compare with exposure-response models, especially
34 given that exposure was not expressed on the same W126 scale.
March 2011 2-61 DRAFT - DO NOT CITE OR QUOTE
-------
2.8. The Role of Tropospheric Ozone in Climate Change
and UV-B Effects
1 Atmospheric O3 plays an important role in the Earth's energy budget by interacting with
2 incoming solar radiation and outgoing infrared radiation. Tropospheric O3 makes up only a small
3 portion of the total column of O3, but it has important incremental effects on the overall radiation
4 budget. Chapter 10 assesses the specific role of tropospheric O3 in the earth's radiation budget and
5 how perturbations in tropospheric O3 might affect climate through its role as a greenhouse gas
6 (Section 10.2), and health, ecology and welfare through its role in shielding the earth's surface from
7 solar ultraviolet radiation (Section 10.3).
2.8.1. Tropospheric Ozone as a Greenhouse Gas
8 Tropospheric O3 is a major greenhouse gas, and increases in its abundance may contribute to
9 climate change according to the 2007 climate assessment by the Intergovernmental Panel on Climate
10 Change (IPCC). Models calculate that the global burden of tropospheric O3 has doubled since the
11 preindustrial era, while observations indicate that in some regions O3 may have increased by factors
12 as great as 4 or 5. These increases are tied to the rise in emissions of O3 precursors from human
13 activity, mainly fossil fuel consumption and agricultural processes. The impact on climate of the O3
14 change since preindustrial times has been estimated to be about 25-40% of anthropogenic CO2
15 impact and about 75% of anthropogenic CH4 impact according to the IPCC, ranking it third in
16 importance among the greenhouse gases.
17 The metric frequently used to estimate the potential climate impact of O3 is called radiative
18 forcing (RF). RF is a change in the radiative balance at the tropopause or at the top of the
19 atmosphere when a perturbation is introduced in the earth-atmosphere-ocean system. The units of RF
20 are energy flux per unit area, or W/m2, and positive values indicate warming while negative values
21 indicate cooling. The IPCC estimates a radiative forcing of 0.35 W/m2 for the change in tropospheric
22 O3 since the preindustrial era, compared to 1.66 W/m2 for CO2 and 0.48 W/m2 for CH4. The error
23 bars encompassing the tropospheric O3 radiative forcing estimate range from 0.25 to 0.65 W/m2,
24 making it relatively more uncertain than the long-lived greenhouse gases. Despite these
25 uncertainties, there is a causal relationship between tropospheric Os and radiative forcing.
26 RF does not take into account the climate feedbacks that could amplify or dampen the actual
27 surface temperature response. Quantifying the change in surface temperature requires a complex
28 climate simulation in which all important feedbacks are accounted for. As these processes are not
29 well understood or easily modeled, the surface temperature response to a given RF is highly
30 uncertain and can vary greatly among models and from region to region within the same model.
31 Despite these uncertainties, there is likely to be a causal relationship between tropospheric O3
32 and climate change.
March 2011 2-62 DRAFT - DO NOT CITE OR QUOTE
-------
2.8.2. Tropospheric Ozone and UV-B related effects
1 Ultraviolet (UV) radiation emitted from the Sun contains sufficient energy when it reaches the
2 Earth to break (photolyze) chemical bonds in molecules, thereby leading to damaging effects on
3 living organisms and materials. Atmospheric O3 plays a crucial role in reducing exposure to solar
4 UV radiation at the Earth's surface. Ozone in the stratosphere is responsible for the majority of this
5 shielding effect, as approximately 90% of total atmospheric O3 is located there over mid-latitudes.
6 Ozone in the troposphere provides supplemental shielding of radiation in the wavelength band from
7 280-315 nm, UV-B radiation. UV-B radiation has important effects on human health and ecosystems,
8 and is associated with materials damage.
9 Adverse human health effects associated with solar UV-B radiation exposure include
10 erythema, skin cancer, ocular damage, and immune system suppression. A potential human health
11 benefit of increased UV-B exposure involves the UV-induced production of vitamin D which may
12 help reduce the risk of metabolic bone disease, type I diabetes, mellitus, and rheumatoid arthritis,
13 and may provide beneficial immunomodulatory effects on multiple sclerosis, insulin-dependent
14 diabetes mellitus, and rheumatoid arthritis.
15 Adverse ecosystem and materials damage effects are also associated with solar UV-B radiation
16 exposure. Terrestrial ecosystem effects from increased UV-B radiation include reduced plant
17 productivity and plant cover, changes in biodiversity, susceptibility to infection, and increases in
18 natural UV protective responses. In general, however, these effects are small for moderate UV-B
19 increases at mid-latitudes. Aquatic ecosystem effects from increased UV-B radiation include
20 sensitivity in growth, immune response, and behavioral patterns of aquatic organisms.
21 Biogeochemical cycles, particularly the carbon cycle, can also be influenced by increased UV-B
22 radiation with effects ranging from UV-induced increases in CO2 uptake through soil respiration to
23 UV-induced release of CO2 through photodegradation of above-ground plant litter. Changes in solar
24 UV radiation may also have effects on carbon cycling and CO2 uptake in the oceans as well as
25 release of dissolved organic matter from sediment and algae. Finally, materials damage from
26 increased UV-B radiation include UV-induced photodegradation of wood and plastics.
27 There is a lack of published studies that critically examine the incremental health or welfare
28 effects (adverse or beneficial) attributable specifically to changes in UV-B exposure resulting from
29 perturbations in tropospheric O3 concentrations. While the effects are expected to be small, they
30 cannot yet be critically assessed within reasonable uncertainty. Overall, the evidence is inadequate
31 to determine if a causal relationship exists between tropospheric Os and UV-B related health
32 and welfare effects.
March 2011 2-63 DRAFT - DO NOT CITE OR QUOTE
-------
2.9. Summary of Causal Determinations for Health Effects
and Welfare Effects
1 This chapter has provided an overview of the underlying evidence used in making the causal
2 determinations for the health and welfare effects of O3. This review builds upon the conclusions of
3 the previous AQCDs for O3 (U.S. EPA, 1978, 040586: U.S. EPA, 1984, 029711: U.S. EPA, 1996,
4 017831: U.S. EPA, 1996, 080827: U.S. EPA, 2006, 088089V
5 The evaluation of the epidemiologic, toxicological, and controlled human exposure studies
6 published since the completion of the 2006 O3 AQCD have provided additional evidence for O3-
7 related health outcomes. Table 2-5 provides an overview of the causal determinations for all of the
8 health outcomes evaluated. Causal determinations for O3 and welfare effects are included in
9 Table 2-6, while causal determinations for climate change and UV-B effects are in Table 2-7.
10 Detailed discussions of the scientific evidence and rationale for these causal determinations are
11 provided in subsequent chapters of this ISA.
Table 2-5. Summary of ozone causal determinations by exposure duration and health outcome
Health Outcome
Conclusions from 2006 03 AQCD
Conclusions from
201 11st Draft ISA
Short-Term Exposure to 03
Respiratory effects
Cardiovascular effects
Central nervous system
effects
All-cause mortality
The overall evidence supports a causal relationship between acute ambient 03 exposures and
increased respiratory morbidity outcomes.
The limited evidence is highly suggestive that 03 directly and/or indirectly contributes to
cardiovascular-related morbidity, but much remains to be done to more fully substantiate the
association.
Toxicological studies report that acute exposures to 03 are associated with alterations in
neurotransmitters, motor activity, short and long term memory, sleep patterns, and histological signs of
neurodegeneration.
The evidence is highly suggestive that 03 directly or indirectly contributes to non-accidental and
cardiopulmonary-related mortality.
Causal Relationship
Suggestive of a Causal
Relationship
Suggestive of a Causal
Relationship
Likely to be a Causal
Relationship
Long-term Exposure to 03
Respiratory effects
Cardiovascular Effects
Reproductive and
developmental effects
Central nervous system
effects
Cancer
Mortality
The current evidence is suggestive but inconclusive for respiratory health effects from long-term 03
exposure.
No studies from previous review
Limited evidence for a relationship between air pollution and birth-related health outcomes, including
mortality, premature births, low birth weights, and birth defects, with little evidence being found for 03
effects.
Toxicological studies report that acute exposures to 03 are associated with alterations in
neurotransmitters, motor activity, short and long term memory, sleep patterns, and histological signs of
neurodegeneration.
Little evidence for a relationship between chronic 03 exposure and increased risk of lung cancer.
There is little evidence to suggest a causal relationship between chronic 03 exposure and
increased risk for mortality in humans.
Likely to be a Causal
Relationship
Suggestive of a Causal
Relationship
Suggestive of a Causal
Relationship
Suggestive of a Causal
Relationship
Inadequate to infer a
Causal Relationship
Suggestive of a Causal
Relationship
March 2011
2-64
DRAFT - DO NOT CITE OR QUOTE
-------
Table 2-6. Summary of ozone causal determination for welfare effects
Vegetation and Ecosystem
Effects
Conclusions from 2006 03 AQCD
Conclusions from 2011 1st
Draft ISA
Reduced Vegetation Growth
Alteration of Vegetation
Reproduction
Visible Foliar Injury Effects on
Vegetation
Alteration of Leaf Gas Exchange in
Vegetation
Reduced Yield and Quality of
Agricultural Crops
Reduced Productivity in Terrestrial
Ecosystems
Reduced Carbon (C)
Sequestration in Terrestrial
Ecosystems
Alteration of Terrestrial Ecosystem
Water Cycling
Alteration of Below-ground
Biogeochemical Cycles
Alteration of Terrestrial Community
Composition
Data published since the 1996 AQCD strengthen previous conclusions that there is
strong evidence that current ambient 03 concentrations cause decreased growth and
biomass accumulation in annual, perennial and woody plants, including agronomic
crops, annuals, shrubs, grasses, and trees.
For several decades, studies have demonstrated 03 effects on different stages of
reproduction.
Data published since the 1996 AQCD strengthen previous conclusions that there is
strong evidence that current ambient 03 concentrations cause impaired aesthetic
quality of many native plants and trees by increasing foliar injury.
Ozone exposure reduces photosynthesis, and the mechanisms of this reduction
are better understood as a result of the research since the 1996 AQCD.
Data published since the 1996 AQCD strengthen previous conclusions that there is
strong evidence that current ambient 03 concentrations cause decreased yield
and/or nutritive quality in a large number of agronomic and forage crops.
There is evidence that 03 is an important stressor of ecosystems and that the
effects of 03 on individual plants and processes are scaled up through the
ecosystem, affecting net primary productivity.
Limited studies from previous review
Ecosystem water quantity may be affected by 03 exposure at the landscape level.
Ozone-sensitive species have well known responses to 03 exposure, including
altered C allocation to below-ground tissues, and altered rates of leaf and root
production, turnover, and decomposition. These shifts can affect overall C and N loss
from the ecosystem in terms of respired C, and leached aqueous dissolved organic
and inorganic C and N.
Ozone may be affecting above- and below -ground community composition through
impacts on both growth and reproduction. Significant changes in plant
community composition resulting directly from 03 exposure have been
demonstrated.
Causal Relationship
Causal Relationship
Causal Relationship
Causal Relationship
Causal Relationship
Causal Relationship
Likely to be a Causal
Relationship
Likely to be a Causal
Relationship
Causal Relationship
Likely to be a Causal
Relationship
Table 2-7. Summary of ozone causal determination for climate change and UV-B effects
Effects
Radiative Forcing
Climate Change
UV-B Related Health and
Welfare Effects
Conclusions from 2006 03 AQCD
Climate forcing by 03 at the regional scale may be its most important impact on climate.
While more certain estimates of the overall importance of global-scale forcing due to
tropospheric 03 await further advances in monitoring and chemical transport modeling, the
overall body of scientific evidence suggests that high concentrations of 03 on the regional
scale could have a discernable influence on climate, leading to surface temperature and
hydrological cycle changes.
UV-B has not been studied in sufficient detail to allow for a credible health benefits
assessment. In conclusion, the effect of changes in surface-level 03 concentrations on UV-
induced health outcomes cannot yet be critically assessed within reasonable uncertainty.
Conclusions from 2011 1st
Draft ISA
Causal Relationship
Likely to be a Causal
Relationship
Inadequate to Determine if a
Causal Relationship Exists
March 2011
2-65
DRAFT - DO NOT CITE OR QUOTE
-------
References
U.S. EPA (1978). Air quality criteria for ozone and other photochemical oxidants (Report No. EPA/600/8-78/004).
Washington, DC: U.S. Environmental Protection Agency, Office of Research and Development. 040586
U.S. EPA (1984). Air quality criteria for ozone and other photochemical oxidants, v 3 (Report No. EPA/600/8-84/020 A).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development.
http://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=2000AVEV.txt. 029711
U.S. EPA (1996). Air quality criteria for ozone and related photochemical oxidants (Report No. EPA/600/P-93/004AF).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development. (NTIS
No. PB94173127) 017831
U.S. EPA (1996). Air quality criteria for ozone and related photochemical oxidants Volume II of III (Report No.
EPA/600/P-93/004BF). Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research
and Development. 080827
U.S. EPA (2006). Air quality criteria for ozone and related photochemical oxidants (Report No. EPA/600/R-05/004AF).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid= 149923. 088089
U.S. EPA (2010). Our nation's air: Status and trends through 2008 (Report No. EPA/454/R-09/002). Research Triangle
Park, NC: U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards.
http://www.epa.gov/airtrends/2010/report/fullreport.pdf 647278
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments such as the Integrated Science Assessments (ISAs) and the Integrated Risk Information System (IRIS).
March 2011 2-66 DRAFT - DO NOT CITE OR QUOTE
-------
Chapter 3. Atmospheric Chemistry and
Ambient Concentrations
3.1. Introduction
1 In the stratosphere, O3 serves the beneficial role of blocking the Sun's harmful ultraviolet
2 radiation and preventing the majority of this radiation from reaching the Earth's surface. In the
3 troposphere, however, O3 and other photochemical oxidants are air pollutants with potentially
4 harmful effects on living organisms. This chapter discusses the atmospheric chemistry associated
5 with tropospheric O3 and other related photochemical oxidants and provides a detailed description of
6 their surface-level concentrations. The focus of this chapter is on O3 since it is the NAAQS indicator
7 for all photochemical oxidants. To the extent possible, other photochemical oxidants are discussed,
8 but limited information is currently available. Although O3 is involved in reactions in indoor air, the
9 focus in this chapter will be on chemistry occurring in outdoor, ambient air.
10 The material in this chapter is organized as follows. Section 3.2 outlines the physical and
11 chemical processes involved in O3 formation and removal. Section 3.3 describes the latest methods
12 used to model global O3 concentrations, and Section 3.4 describes the application of some of those
13 methods for estimating background concentrations of O3. Section 3.5 includes a comprehensive
14 description of available O3 monitoring techniques and monitoring networks, while Section 3.6
15 presents information on the spatial and temporal variability of O3 concentrations across the U.S. and
16 their associations with other pollutants using available monitoring data.
3.2. Physical and Chemical Processes
17 O3 in the troposphere is a secondary pollutant formed by photochemical reactions of precursor
18 gases and is not directly emitted from specific sources. Ozone and other oxidants, such as PAN and
19 H2O2 form in polluted areas by atmospheric reactions involving two main classes of precursor
20 pollutants: VOCs and NOX. Carbon monoxide (CO) is also important for O3 formation in polluted
21 areas and in the remote troposphere. The formation of O3, other oxidants and oxidation products
22 from these precursors is a complex, nonlinear function of many factors including (1) the intensity
23 and spectral distribution of sunlight; (2) atmospheric mixing; (3) concentrations of precursors in the
24 ambient air and the rates of chemical reactions of these precursors; and (4) processing on cloud and
25 aerosol particles. Information contained in this chapter briefly describes these processes and
26 numerical models that incorporate these processes to calculate O3 concentrations.
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments such as the Integrated Science Assessments (ISAs) and the Integrated Risk Information System (IRIS).
March 2011 3-1 DRAFT - DO NOT CITE OR QUOTE
-------
1 Ozone is present not only in polluted urban atmospheres, but throughout the troposphere, even
2 in remote areas of the globe. The same basic processes involving sunlight-driven reactions of NOX,
3 VOCs and CO contribute to O3 formation throughout the troposphere. These processes also lead to
4 the formation of other photochemical products, such as PAN, HNO3, and H2SO4, and to other
5 compounds, such as HCHO and other carbonyl compounds.
6 The processes responsible for producing summertime O3 episodes are fairly well understood,
7 and were covered in detail in the previous O3 AQCD (U.S. EPA, 2006, 088089). This section focuses
8 on topics that form the basis for discussions in later chapters and for which there is substantial new
9 information since the previous AQCD. A schematic overview of the major photochemical cycles
10 influencing O3 in the troposphere and the stratosphere is given in Figure 3-1.
Oxygen
St rat os
Non-Polar
v Regions
Polar
Regions
(Ozone Hole) §>
X
Figure 3-1. Schematic overview of photochemical processes influencing stratospheric and
tropospheric ozone.
March 2011
3-2
DRAFT - DO NOT CITE OR QUOTE
-------
1 Major episodes of high O3 concentrations in the eastern U.S. and in Europe are associated with
2 slow moving high pressure systems. High pressure systems during the warmer seasons are associated
3 with the sinking of air, resulting in warm, generally cloudless skies, with light winds. The sinking of
4 air results in the development of stable conditions near the surface which inhibit or reduce the
5 vertical mixing of O3 precursors. The combination of inhibited vertical mixing and light winds
6 minimizes the dispersal of pollutants emitted in urban areas, allowing their concentrations to build
7 up. Photochemical activity involving these precursors is enhanced because of higher temperatures
8 and the availability of sunlight. In the eastern U.S., O3 and other secondary pollutants are determined
9 by meteorological and chemical processes extending typically over spatial scales of several hundred
10 thousand square kilometers (e.g., Civerolo et al., 2003, 053985; Rao et al., 2003, 054094). Ozone
11 episodes are thus best regarded as regional in nature. The conditions conducive to formation of high
12 O3 can persist for several days. These conditions have been described in greater detail in the 1996
13 and 2006 O3 AQCDs (U.S. EPA, 1996, 017831: U.S. EPA, 2006, 088089V The transport of
14 pollutants downwind of major urban centers is characterized by the development of urban plumes.
15 Mountain barriers limit mixing (as in Los Angeles and Mexico City) and result in a higher frequency
16 and duration of days with high O3 concentrations. However, orographic lifting over the San Gabriel
17 Mountains results in O3 transport from Los Angeles to areas hundreds of kilometers downwind (e.g.,
18 in Colorado and Utah) (Langford et al., 2009, 491703). Ozone concentrations in southern urban
19 areas (such as Houston, TX and Atlanta, GA) tend to decrease with increasing wind speed. In
20 northern U.S. cities (such as Chicago, IL; New York, NY; Boston, MA; and Portland, ME), the
21 average O3 concentrations over the metropolitan areas increase with wind speed, indicating that
22 transport of O3 and its precursors from upwind areas is important (Husar and Renard, 1998, 052413;
23 Schichtel and Husar, 2001, 016669).
24 Aircraft observations indicate that there can be substantial differences in mixing ratios of key
25 species between the surface and the overlying atmosphere (Berkowitz and Shaw, 1997, 047593;
26 Fehsenfeld et al., 1996, 047803). In particular, mixing ratios of O3 can be higher in the lower free
27 troposphere (aloft) than in the planetary boundary layer (PEL) during multiday O3 episodes
28 (Taubman et al., 2004, 052228; Taubman et al., 2006, 087582). Convective processes and small scale
29 turbulence transport O3 and other pollutants both upward and downward throughout the planetary
30 boundary layer and the free troposphere. During the day, convection driven by heating of the earth's
31 surface results in a deeper planetary boundary layer (PEL) with vertically well mixed O3 and
32 precursors. As solar heating of the surface decreases going into night, the daytime boundary layer
33 collapses leaving behind O3 and its precursors in a residual layer above a shallow nighttime
34 boundary layer. Pollutants in the residual layer have now become essentially part of the free
35 troposphere, as shown in AX2.3.2 of the 2006 O3 AQCD (U.S. EPA, 2006, 088089). Winds in the
36 free troposphere tend to be stronger than those closer to the surface and so are capable of
37 transporting pollutants over long distances. Thus, O3 and its precursors can be transported vertically
38 by convection into the upper part of the mixed layer on one day, then transported overnight as a layer
March 2011 3-3 DRAFT - DO NOT CITE OR QUOTE
-------
1 of elevated mixing ratios, and then entrained into a growing convective boundary layer downwind
2 and brought back down to the surface.
3 High O3 concentrations showing large diurnal variations at the surface in southern New
4 England were associated with the presence of such layers (Berkowitz et al, 1998, 081467). Winds
5 several hundred meters above the ground can bring pollutants from the west, even though surface
6 winds are from the southwest during periods of high O3 in the eastern U.S. (Blumenthal et al., 1997,
7 052278). These considerations suggest that in many areas of the U.S., O3 and its precursors can be
8 transported over hundreds if not thousands of square kilometers.
9 Nocturnal low level jets (LLJs) are an efficient means for transporting pollutants that have
10 been entrained into the residual boundary layer over hundreds of kilometers (U.S. EPA, 2006,
11 088089. Annex AX2.3.3). LLJs are most prevalent in the central U.S. extending northward from
12 eastern Texas, and along the Atlantic states extending southwest to northeast. LLJs have also been
13 observed off the coast of California. Turbulence associated with LLJs brings pollutants to the surface
14 and results in secondary O3 maxima during the early morning in many locations (Corsmeier et al.,
15 1997, 047620). Stratospheric intrusions and intercontinental transport of O3 are also important and
16 are covered in Section 3.4 in relation to policy relevant background concentrations.
3.2.1. Sources ofPrecursors Involved in Ozone Formation
17 Emissions of O3 precursor compounds (NOX, VOCs, and CO) can be divided into natural and
18 anthropogenic source categories. Natural sources can be further divided into biogenic from
19 vegetation, microbes, and animals, and abiotic from biomass burning, lightning, and geogenic
20 sources. However, the distinction between natural and anthropogenic sources is often difficult to
21 make in practice, as human activities directly or indirectly affect emissions from what would have
22 been considered natural sources during the preindustrial era. Thus, emissions from plants and
23 animals used in agriculture have been referred to as anthropogenic or biogenic in different
24 applications. Wildfire emissions can be considered natural, except that forest management practices
25 can lead to buildup of fuels on the forest floor, thereby altering the frequency and severity of forest
26 fires.
27 Estimates of emissions forNOx, VOCs, and CO (U.S. EPA, 2008, 665052) are shown in
28 Figure 3-2 to provide a general indication of the relative importance of the different sources in the
29 U.S. as a whole. The magnitudes of the sources are strongly location and time dependent and so
30 should not be used to apportion sources of exposure. Shown in Figure 3-2 are Tier 1 categories. The
31 miscellaneous category can be quite large compared to total emissions, especially for CO and VOCs.
32 The miscellaneous category includes agriculture and forestry, wildfires, prescribed burns, and a
33 much more modest contribution from structural fires.
March 2011 3-4 DRAFT - DO NOT CITE OR QUOTE
-------
Nitrogen Oxides (NOJ
Total Emissions = 17.4 MT
HIGHWAY VEHICLES [
OFF-HIGHWAY
FUEL COMB. ELEC. UTIL.
FUEL COMB. INDUSTRIAL
FUEL COMB. OTHER
OTHER INDUSTRIAL PROCESSES
PETROLEUM & RELATED INDUSTRIES
WASTE DISPOSAL & RECYCLING nO.13
METALS PROCESSING D O.O6
CHEMICAL & ALLIED PRODUCT MFG I O.O5
STORAGE & TRANSPORT IO.O15
SOLVENT UTILIZATION O.OO4
14.4
13.4
1 1.fa
IO.66
I I 0.44
CZIO.25
no. 13
DO.O6
DO. 05
O.O15
O.OO4
4567
Emissions (Millions Tons/Year)
Volatile Organic Compounds (VOC)
Total Emissions = 16.7 MT
HIGHWAY VEHICLES
OFF-HIGHWAY
FUEL COMB. ELEC. UTIL. QO.O4
FUEL COMB. INDUSTRIAL CHO.12
FUEL COMB. OTHER
OTHER INDUSTRIAL PROCESSES
PETROLEUM & RELATED INDUSTRIES
MISCELLANEOUS
WASTE DISPOSAL & RECYCLING
METALS PROCESSING BO.O4
CHEMICAL & ALLIED PRODUCT MFG
STORAGE & TRANSPORT
SOLVENT UTILIZATION
O
QO.O4
00.12
I 0 S3
IO.41
l0-51
I 0.36
QO.O4
I 2. 6
I
1 1.:
^.9B
5
Carbon Monoxide (CO)
Total Emissions = 84.6 MT
HIGHWAY VEHICLES
OFF-HIGHWAY
FUEL COMB. ELEC. UTIL. 00.58
FUEL COMB. INDUSTRIAL D 1.04
FUEL COMB. OTHER
OTHER INDUSTRIAL PROCESSES D O.48
PETROLEUM & RELATED INDUSTRIES DO.32
MISCELLANEOUS
WASTE DISPOSAL & RECYCLING LJ 1 .41
METALS PROCESSING DO.75
CHEMICAL & ALLIED PRODUCT MFG IO.19
STORAGE & TRANSPORT I O.1
SOLVENT UTILIZATION O.OO2
13.85
3 4
Emissions (Millions Tons/Year)
144.04
I 1
8.7
DO. 58
n 1.04
DO. 48
IO.32
I 1
LJ 1.41
• O.75
1O.19
10.1
3.7
10
15
20
25
30 35 40 45 SO
Emissions (Millions Tons/Year)
Source: U.S. EPA (2008, 665052)
Figure 3-2. Estimated anthropogenic emissions of ozone precursors for 2005 including NOx
(top), VOCs (middle), and CO (bottom) in the U.S. in million metric tons (MT) per
year.
March 2011
3-5
DRAFT - DO NOT CITE OR QUOTE
-------
1 Anthropogenic NOX emissions are associated with combustion processes. Most emissions are
2 in the form of NO, which is formed at high combustion temperatures from atmospheric nitrogen (N2)
3 and oxygen (O2) and from fuel nitrogen (N). According to the 2005 National Emissions Inventory
4 (2005 NEI)(U.S. EPA, 2008, 665052). the largest sources of NOX are on- and off-road mobile
5 sources and electric power generation plants. Emissions of NOX therefore are highest in areas having
6 a high density of power plants and in urban regions having high traffic density. Dallman and Harley
7 (2010, 665390) compared NOX emissions estimates from the National Emissions Inventory, mobile
8 sector data (U.S. EPA, 2008, 665052) with an alternative method based on fuel consumption and
9 found reasonable agreement in total U.S. anthropogenic emissions between the two techniques (to
10 within about 5%). However, emissions from on-road diesel engines in the fuel based inventory
11 constituted 46% of total mobile source NOX compared to 35% in the EPA inventory. As a result,
12 emissions from on-road diesel engines in the fuel based approach are even larger than electric power
13 generation as estimated in the 2005 NEI, and on-road diesel engines might represent the largest
14 single NOX source category. Differences between the two techniques are largely compensated by
15 differences in emissions from on-road gasoline engines. Uncertainties in the fuel consumption
16 inventory ranged from 3% for on-road gasoline engines to 20% for marine sources, and in the EPA
17 inventory uncertainties ranged from 16% for locomotives to 30% for off-road diesel engines. It
18 should be noted that the on-road diesel engine emissions estimate by Dallman and Harley (2010,
19 665390) is still within the uncertainty of the EPA estimate (22%).
20 Major natural sources of NOX in the U.S. include lightning, soils, and wildfires; stratospheric
21 intrusions can also be important under certain conditions in many locations. Uncertainties in natural
22 NOX emissions are much larger than for anthropogenic NOX emissions. Fang et al. (2010, 665391)
23 estimated lightning generated NOX of-0.6 MT for July 2004. This value is -40% of the
24 anthropogenic emissions for the same period, but Fang et al. estimated that -98% is formed in the
25 free troposphere and so contributions to the surface NOX burden are low because most of this NOX is
26 oxidized to NOZ species during downward transport into the planetary boundary layer. The
27 remaining 2% is formed within the planetary boundary layer. Both nitrifying and denitrifying
28 organisms in the soil can produce NOX, mainly in the form of NO. Emission rates depend mainly on
29 fertilization amount and soil temperature and moisture. Nationwide, about 60% of the total NOX
30 emitted by soils is estimated to occur in the central corn belt of the U.S. Spatial and temporal
31 variability in soil NOX emissions leads to considerable uncertainty in emissions estimates. However,
32 these emissions are relatively low, only -0.97 MT/year, or about 6% of anthropogenic NOX
33 emissions. The oxidation of ammonia (NH3) emitted mainly by livestock and soils, leads to the
34 formation of a small amount of NO.
35 Hundreds of VOCs, containing mainly 2 to -12 carbon (C) atoms, are emitted by evaporation
36 and combustion processes from a large number of anthropogenic sources. The two largest
37 anthropogenic source categories in the U.S. EPA's emissions inventories are industrial processes and
38 transportation. Emissions of VOCs from highway vehicles account for roughly two-thirds of the
39 transportation-related emissions. The accuracy of VOC emission estimates is difficult to determine,
March 2011 3-6 DRAFT - DO NOT CITE OR QUOTE
-------
1 both for stationary and mobile sources. Evaporative emissions, which depend on temperature and
2 other environmental factors, compound the difficulties of assigning accurate emission factors. In
3 assigning VOC emission estimates to the mobile source category, models are used that incorporate
4 numerous input parameters (e.g., type of fuel used, type of emission controls, age of vehicle), each
5 of which has some degree of uncertainty.
6 On the U.S. and global scales, emissions of VOCs from vegetation are much larger than those
7 from anthropogenic sources. Emissions of VOCs from anthropogenic sources in the 2005 NEI were
8 -17 MT/year (including wildfires, which constitute -1/6 of that total), but were 29 MT/year from
9 biogenic sources. Uncertainties in both biogenic and anthropogenic VOC emission inventories
10 prevent determination of the relative contributions of these two categories, at least in many areas.
11 Vegetation emits significant quantities of VOCs, such as terpenoid compounds (isoprene, 2-methyl-
12 3-buten-2-ol, monoterpenes), compounds in the hexanal family, alkenes, aldehydes, organic acids,
13 alcohols, ketones, and alkanes. The major chemicals emitted by plants are isoprene (40%), other
14 terpenoid and sesqui-terpenoid compounds (25%) and the remainder consists of assorted oxygenated
15 compounds and hydrocarbons according to the 2005 NEI. Coniferous forests represent the largest
16 source on a nationwide basis because of their extensive land coverage. Most biogenic emissions
17 occur during the summer because of their dependence on temperature and incident sunlight.
18 Biogenic emissions are also higher in southern states than in northern states for these reasons and
19 because of species variations. The uncertainty in natural emissions is about 50% for isoprene under
20 midday summer conditions and could be as much as a factor often higher for some compounds
21 (Guenther et al., 2000, 025002). In EPA's regional modeling efforts, biogenic emissions of VOCs are
22 estimated using the BEIS model (U.S. EPA, 2010, 677538) with data from the Biogenic Emissions
23 Landcover Database (BELD) and annual meteorological data. However, other emissions models are
24 used such as MEGAN (Model of Emissions of Gases and Aerosols from Nature) (Guenther et al.,
25 2006, 607080). especially in global modeling efforts.
26 Anthropogenic CO is emitted primarily by incomplete combustion of carbon-containing fuels.
27 In general, any increase in fuel O2 content, burn temperature, or mixing time in the combustion zone
28 will tend to decrease production of CO relative to CO2. CO emissions from large fossil-fueled power
29 plants are typically very low since the boilers at these plants are tuned for highly efficient
30 combustion with the lowest possible fuel consumption. Additionally, by allowing time for the
31 furnace flue gases to mix with air and be oxidized by OH to CO2 in the hot gas stream before the OH
32 concentrations drop as the flue gases cool, the CO-to-CO2 ratio in these emissions is shifted toward
33 CO2. Nationally, on-road mobile sources constituted -26% of total CO emissions in the 2005 NEI.
34 When emissions from non-road vehicles are included, it can be seen from Figure 3-2 that all mobile
35 sources accounted for -73% of total anthropogenic CO emissions in the U.S.
36 Analyses by Harley et al. (2005, 088154) and Parrish (2006, 090352) are consistent with the
37 suggestion in Pollack et al. (2004, 184461) that the EPA MOBILE6 vehicle emissions model
38 (U.S. EPA, 2010, 677539) overestimates vehicle CO emissions by a factor of-2. Field
39 measurements by Bishop and Stedman (2008, 194670) were in accord with Parrish's (2006, 090352)
March 2011 3-7 DRAFT - DO NOT CITE OR QUOTE
-------
1 findings that the measured trends of CO and NOX concentrations from mobile sources in the U.S.
2 indicated that modeled CO emission estimates were substantially too high. Hudman et al. (2008,
3 191253) found that the NEI overestimated anthropogenic CO emissions by 60% for the eastern U.S.
4 during the period July 1-August 15, 2004 using aircraft observations of CO from the International
5 Consortium for Atmospheric Research on Transport and Transformation (ICARTT) campaign
6 (Fehsenfeld et al., 2006, 190531) and results from a tropospheric chemistry model (GEOS-Chem).
7 Improvements in emissions technologies not correctly represented in MOBILE emission models
8 have been suggested as one cause for this discrepancy. For example, Pokharel et al. (2002, 052473;
9 2003, 053740) demonstrated substantial decrements in the CO fraction of tailpipe exhaust in several
10 U.S. cities and Burgard et al. (2006, 193222) documented improvements in emission from heavy-
11 duty on-road diesel engines. Some of the largest errors in the MOBILE models are addressed in the
12 successor model, MOVES (U.S. EPA, 2011, 677540).
13 Estimates of biogenic CO emissions in the 2005 NEI are made in a manner similar to that for
14 VOCs. National biogenic emissions, excluding fires, were estimated to contribute -7% and wildfires
15 added another -16% to the national CO emissions total. Photodecomposition of organic matter in
16 oceans, rivers, lakes, and other surface waters, and from soil surfaces also releases CO (Goldstein
17 and Galbally, 2007, 193247). However, soils can act as a CO source or a sink depending on soil
18 moisture, UV flux reaching the soil surface, and soil temperature (Conrad and Seiler, 1985, 029520).
19 Soil uptake of CO is driven by anaerobic bacteria (Inman et al., 1971, 010972). Emissions of CO
20 from soils appear to occur by abiotic processes, such as thermodecomposition or
21 photodecomposition of organic matter. In general, warm and moist conditions found in most soils
22 favor CO uptake, whereas hot and dry conditions found in deserts and some savannas favor the
23 release of CO (King, 1999, 002828).
3.2.2. Gas Phase Reactions Leading to Ozone Formation
24 Photochemical processes involved in O3 formation have been extensively reviewed in a
25 number of books (Finlayson-Pitts and Pitts, 1986, 035054: Jacob, 1999, 091122: Jacobson, 2002,
26 090667: Seinfeld and Pandis, 1998, 018352) and in the previous O3 AQCDs (U.S. EPA, 1996,
27 017831: U.S. EPA, 2006, 088089).
28 The photochemical formation of O3 in the troposphere proceeds through the oxidation of NO
29 to nitrogen dioxide (NO2) by organic (RO2) or hydro-peroxy (HO2) radicals. The photolysis of NO2
30 yields NO and a ground-state oxygen atom, O(3P), which then reacts with molecular oxygen to form
31 O3. Free radicals oxidizing NO to NO2 are formed during the oxidation of VOCs (U.S. EPA, 2006,
32 088089. Annex AX2.2.2).
33 The term VOC refers to all carbon-containing gas-phase compounds in the atmosphere, both
34 biogenic and anthropogenic in origin, excluding CO and CO2. Classes of organic compounds
35 important for the photochemical formation of O3 include alkanes, alkenes, aromatic hydrocarbons,
36 carbonyl compounds (e.g., aldehydes and ketones), alcohols, organic peroxides, and halogenated
37 organic compounds (e.g., alkyl halides). This array of compounds encompasses a wide range of
March 2011 3-8 DRAFT - DO NOT CITE OR QUOTE
-------
1 chemical properties and lifetimes: isoprene has an atmospheric lifetime of approximately an hour,
2 whereas methane has an atmospheric lifetime of about a decade.
3 In urban areas, compounds representing all classes of VOCs and CO are important for O3
4 formation. In nonurban vegetated areas, biogenic VOCs emitted from vegetation tend to be the most
5 important. In the remote troposphere, methane (CH4) and CO are the main carbon-containing
6 precursors to O3 formation. The oxidation of VOCs is initiated mainly by reaction with hydroxyl
7 (OH) radicals. The primary source of OH radicals in the atmosphere is the reaction of electronically
8 excited O atoms, O(:D), with water vapor. O(:D) is produced by the photolysis of O3 in the Hartley
9 bands. In polluted areas, the photolysis of aldehydes (e.g., HCHO), HONO and H2O2 can also be
10 significant sources of OH or HO2 radicals that can rapidly be converted to OH (Eisele et al., 1997,
11 057210). Ozone can oxidize alkenes and, at night, when they are most abundant, NO3 radicals also
12 oxidize alkenes. In coastal environments and other selected environments, atomic Cl and Br radicals
13 can also initiate the oxidation of VOCs (U.S. EPA, 2006, 088089. Annex AX2.2.3). It should also be
14 emphasized that the reactions of oxygenated VOCs are important components of O3 formation
15 (U.S. EPA, 2006, 088089. Annex AX2.2.9). They may be present in ambient air not only as the result
16 of the atmospheric oxidation of hydrocarbons but also by direct emissions. For example, motor
17 vehicles and some industrial processes emit formaldehyde (Rappengliick et al., 2009, 629680) and
18 vegetation emits methanol.
19 There are a large number of oxidized N-containing compounds in the atmosphere including
20 NO, NO2, NO3, HNO2, HNO3, N2O5, HNO4, PAN and its homologues, other organic nitrates, such as
21 alkyl nitrates, isoprene nitrates and particulate nitrate. Collectively these species are referred to as
22 NOY. Oxidized nitrogen compounds are emitted to the atmosphere mainly as NO which rapidly
23 interconverts with NO2 and so NO and NO2 are often "lumped" together into their own group or
24 family, which is referred to as NOX. NOX can be oxidized to reservoir and termination species (PAN
25 and its homologues, organic nitrates, HNO3, HNO4 and particulate nitrate). These reservoir and
26 termination species are referred to as NOZ. The major reactions involving interconversions of
27 oxidized N species were covered in the 2006 O3 AQCD (U.S. EPA, 2006, 088089. Annex AX2.2.4).
28 Mollner et al. (2010, 665393) identified pernitrous acid (HOONO), an unstable isomer of nitric acid
29 as a product of the major gas phase reaction forming HNO3. However, since pernitrous acid is
30 unstable, it is not a reservoir for NOX. This finding stresses the importance of identifying products in
31 addition to measuring the rate of disappearance of reactants in kinetic studies.
32 The photochemical cycles by which the oxidation of hydrocarbons leads to O3 production are
33 best understood by considering the oxidation of methane, structurally the simplest VOC. The CH4
34 oxidation cycle serves as a model for the chemistry of the relatively clean or unpolluted troposphere
35 (although this is a simplification because vegetation releases large quantities of complex VOCs, such
36 as isoprene, into the atmosphere). In the polluted atmosphere, the underlying chemical principles are
37 the same, as discussed in the 2006 O3 AQCD (U.S. EPA, 2006, 088089. Annex AX2.2.5). The
38 conversion of NO to NO2 occurring with the oxidation of VOCs is accompanied by the production of
March 2011 3-9 DRAFT - DO NOT CITE OR QUOTE
-------
1 O3 and the efficient regeneration of the OH radical, which in turn can react with other VOCs as
2 shown in Figure 3-1.
3 The oxidation of alkanes and alkenes in the atmosphere has been treated in depth in the 1996
4 O3 AQCD (U.S. EPA, 1996, 017831) and was updated in the 2006 O3 AQCD (U.S. EPA, 2006,
5 088089, Annex AX2.2.6 and AX2.2.7). In contrast to simple hydrocarbons containing one or two C
6 atoms, detailed kinetic information about the gas phase oxidation pathways of many anthropogenic
7 hydrocarbons (e.g., aromatic compounds such as benzene and toluene), biogenic hydrocarbons (e.g.,
8 isoprene, the monoterpenes), and their intermediate oxidation products (e.g., epoxides, nitrates, and
9 carbonyl compounds) is lacking. Reaction with OH radicals represents the major loss process for
10 alkanes. Reaction with chlorine (Cl) atoms is an additional sink for alkanes. Stable products of
11 alkane photooxidation are known to include carbonyl compounds, alkyl nitrates, and
12 d-hydroxycarbonyls. Major uncertainties in the atmospheric chemistry of the alkanes concern the
13 chemistry of alkyl nitrate formation; these uncertainties affect the amount of NO-to-NO2 conversion
14 occurring and, hence, the amounts of O3 formed during photochemical degradation of the alkanes.
15 The reaction of OH radicals with aldehydes produced during the oxidation of alkanes forms
16 acyl (R'CO) radicals, and acyl peroxy radicals (R'C(O)-O2) are formed by the further addition of O2.
17 As an example, the oxidation of ethane (C2H5-H) yields acetaldehyde (CH3-CHO). The reaction of
18 CH3-CHO with OH radicals yields acetyl radicals (CH3—CO). The acetyl radicals will then
19 participate with O2 in a termolecular recombination reaction to form acetyl peroxy radicals, which
20 can then react with NO to form CH3 + CO2 or they can react with NO2 to form PAN. PAN acts as a
21 temporary reservoir for NO2. Upon the thermal decomposition of PAN, either locally or elsewhere,
22 NO2 is released to participate in the O3 formation process again.
23 Alkenes react in ambient air with OH, NO3, and Cl radicals and with O3. All of these reactions
24 are important atmospheric transformation processes, and all proceed by initial addition to the
25 >C = C< bonds. Major products of alkene photooxidation include carbonyl compounds.
26 Hydroxynitrates and nitratocarbonyls, and decomposition products from the energy-rich biradicals
27 formed in alkene-O3 reactions are also produced. Major uncertainties in the atmospheric chemistry of
28 the alkenes concern the products and mechanisms of their reactions with O3, especially the yields of
29 free radicals that participate in O3 formation. Examples of oxidation mechanisms of complex alkanes
30 and alkenes can be found in comprehensive texts such as Seinfeld and Pandis (1998, 018352). Apart
31 from the effects of the oxidation of isoprene on production of free radicals and O3 formation,
32 isoprene nitrates appear to play an important role as NOX reservoirs over the eastern U.S. (see for
33 example Perring et al. (2009, 616370)). Their decomposition leads to the recycling of NOX, which
34 can participate in the O3 formation process again as was the case with decomposition of PAN and the
35 even more unstable pernitrous acid.
36 The oxidation of aromatic hydrocarbons constitutes an important component of the chemistry
37 of O3 formation in urban atmospheres (U.S. EPA, 2006, 088089. Annex AX2.2.8). Virtually all of the
38 important aromatic hydrocarbon precursors emitted in urban atmospheres are lost through reaction
39 with the hydroxyl radical. Loss rates for these compounds vary from slow (i.e., benzene) to moderate
March 2011 3-10 DRAFT-DO NOT CITE OR QUOTE
-------
1 (e.g., toluene), to very rapid (e.g., xylene and trimethylbenzene isomers). However, the mechanism
2 for the oxidation of aromatic hydrocarbons following reaction with OH is poorly understood, as is
3 evident from the poor mass balance of the reaction products. The mechanism for the oxidation of
4 toluene has been studied most thoroughly, and there is general agreement on the initial steps in the
5 mechanism. However, at present there is no promising approach for resolving the remaining issues
6 concerning the later steps. The oxidation of aromatic hydrocarbons also leads to particle formation
7 that could remove gas-phase constituents that participate in O3 formation.
8 Adequate analytical techniques needed to identify and quantify key intermediate species are
9 not available for many compounds. In addition, methods to synthesize many of the suspected
10 intermediate compounds are not available so that laboratory studies of their reaction kinetics cannot
11 be performed. Similar considerations apply to the oxidation of biogenic hydrocarbons besides
12 isoprene. These considerations are important because oxidants, other than O3, that are formed from
13 the chemistry described above could exert effects on human health and perhaps also on vegetation
14 (Doyle et al, 2004, 088404: Doyle et al, 2007, 596377: Sexton et al., 2004, 087831V Gas phase
15 oxidants include PAN, H2O2, CH3OOH and other organic hydroperoxides.
16 Ozone is lost through a number of gas phase reactions and deposition to surfaces. The reaction
17 of O3 with NO to produce NO2 mainly results in the recycling of O3 downwind via the
18 recombination of O(3P) with O2 to re-form O3. By itself, this reaction does not lead to a net loss of
19 O3 unless the NO2 is converted to stable end products such as HNO3. Ozone reacts with unsaturated
20 hydrocarbons and with hydrogen (H) containing free radicals (OH, HO2).
21 Perhaps the most recent field study aimed at obtaining a better understanding of atmospheric
22 chemical processes was the Second Texas Air Quality Field Study (TexAQS-II) conducted in
23 Houston in August and September 2006 (see overview by Olaguer et al., 2009, 200191). The
24 TexAQS-II Radical and Aerosol Measurement Project (TRAMP) found evidence for the importance
25 of short-lived radical sources such as HCHO and HONO in increasing O3 productivity. During
26 TRAMP, daytime HCHO pulses as large as 32 ppb were observed and attributed to industrial
27 activities upwind in the Houston Ship Channel (HSC) and HCHO peaks as large as 52 ppb were
28 detected by in-situ surface monitors in the HSC. Primary HCHO produced in flares from local
29 refineries and petrochemical facilities could increase peak O3 by ~30 ppb (Webster et al., 2007,
30 104266). Other findings from TexAQS-II included significant concentrations of HONO during the
31 day, with peak concentrations approaching 1 ppb at local noon. These concentrations are well in
32 excess of current air quality model predictions using gas phase mechanisms alone (e.g., Sarwar et al.,
33 2008, 618491) and multiphase processes are needed to account for these observations. Olaguer et al.
34 (2009, 200191) also noted that using measured HONO brings modeled O3 concentrations into much
35 better agreement with observations and could result in the production of an additional 10 ppb O3.
36 Large nocturnal vertical gradients indicating a surface or near-surface source of HONO, and large
37 concentrations of night-time radicals (~30 ppt HO2) were also found during TRAMP.
March 2011 3-11 DRAFT-DO NOT CITE OR QUOTE
-------
3.2.3. Multiphase Processes
1 In addition to reactions occurring in the gas phase, reactions occurring on the surfaces of or
2 within cloud droplets and airborne particles also occur. Their collective surface area is huge,
3 implying that collisions with gas phase species occur on very short time scales. In addition to
4 hydrometeors (e.g., cloud and fog droplets and snow and ice crystals) there are also potential
5 reactions involving atmospheric particles of varying composition (e.g., wet [deliquesced] inorganic
6 particles, mineral dust, carbon chain agglomerates and organic carbon particles) to consider.
7 Multiphase reactions are involved in the formation of a number of species such as particulate nitrate,
8 and gas phase HONO that can act to both increase and reduce the rate of O3 formation in the polluted
9 troposphere. Data collected in Houston as part of TexAQS-II summarized by Olaguer et al. (2009,
10 200191) indicate that concentrations of HONO are much higher than can be explained by gas phase
11 chemistry and by tailpipe emissions; and that the photolysis of HONO formed in multiphase
12 reactions in addition to the other sources can help narrow the discrepancy between observed and
13 predicted production of O3. However, removal of HOx and NOX onto hydrated particles will reduce
14 the production of O3.
15 Multi-phase processes have been associated with the release of gaseous halogen compounds
16 from marine aerosol, mainly in marine and coastal environments. However, Thornton et al., (2010,
17 386870) found production rates of gaseous nitryl chloride near Boulder, CO from reaction of N2O5
18 with particulate Cl", similar to those found in coastal and marine environments. C1NO2 readily
19 photolyzes to yield Cl. They also found that substantial quantities of N2O5 are recycled through
20 C1NO2 back into NOX instead of forming HNO3 (a stable reservoir for reactive nitrogen compounds).
21 The oxidation of hydrocarbons by Cl radicals released from the marine aerosol could lead to the
22 rapid formation of peroxy radicals and higher rates of O3 production in selected coastal environments
23 and in continental environments. It should be noted that in addition to production from marine
24 aerosol, reactive halogen species are also produced by the oxidation of halogenated organic
25 compounds (e.g., CH3C1, CH3Br and CH3I). The atmospheric chemistry of halogens is complex
26 because Cl, Br and I containing species can react among themselves and with hydrocarbons and
27 other species and could also be important for O3 destruction, as has been noted for the lower
28 stratosphere (McElroy et al., 1986, 019501; Yung et al., 1980, 057212). For example, the reactions of
29 Br and Cl containing radicals deplete O3 in selected environments such as the Arctic during the
30 spring (e.g., Barrie et al., 1988, 053377). the tropical marine boundary layer (e.g., Dickerson et al.,
31 1999, 053394). and inland salt flats and salt lakes (e.g., Stutz et al., 2002, 051882V Mahajan et al.
32 (2010, 665392) found that I and Br species acting together resulted in O3 depletion that was much
33 larger than would have been expected if they acted individually and did not interact with each other
34 (see U.S. EPA, 2006, 088089. section AX2.2.10.3 for more detailed descriptions of these processes).
35 It should be stressed that knowledge of multiphase processes is still evolving and there are still many
36 questions that remain to be answered. However, it is becoming clear that multiphase processes are
37 important for O3 chemistry.
March 2011 3-12 DRAFT - DO NOT CITE OR QUOTE
-------
1 Reactions of O3 with monoterpenes have been shown to produce oxidants in the aerosol phase,
2 principally as components of ultrafine particles. These reactions involving ambient O3 and terpene-
3 related compounds from cleaning products, air fresheners and wood products, can also occur in
4 indoor air as was discussed in the previous O3 AQCD (U.S. EPA, 2006, 088089). Docherty et al.
5 (2005, 087613) found evidence for the substantial production of organic hydroperoxides in
6 secondary organic aerosol (SOA) resulting from the reaction of monoterpenes with O3. Analysis of
7 the SOA formed in their environmental chamber indicated that the SOA consisted mainly of organic
8 hydroperoxides. In particular, they obtained yields of 47% and 85% of organic peroxides from the
9 oxidation of a- and (3-pinene. The hydroperoxides then react with aldehydes in particles to form
10 peroxyhemiacetals, which can either rearrange to form other compounds such as alcohols and acids
11 or revert back to the hydroperoxides. The aldehydes are also produced in large measure during the
12 ozonolysis of the monoterpenes. Monoterpenes also react with OH radicals resulting in the
13 production of more lower-molecular-weight products than in the reaction with monoterpenes and O3.
14 Bonn et al. (2004, 053770) estimated that hydroperoxides lead to 63% of global SOA formation from
15 the oxidation of terpenes. The oxidation of anthropogenic aromatic hydrocarbons by OH radicals
16 could also produce organic hydroperoxides in SOA (Johnson et al., 2004, 087659). Recent
17 measurements show that the abundance of oxidized SOA exceeds that of more reduced hydrocarbon
18 like organic aerosol in Pittsburgh (Zhang et al., 2005, 157185) and in about 30 other cities across the
19 Northern Hemisphere (Zhang et al., 2007, 101119). Based on aircraft and ship-based sampling of
20 organic aerosols over coastal waters downwind of northeastern U.S. cities, de Gouw et al. (2008,
21 191757) reported that 40-70% of measured organic mass was water soluble and estimated that
22 approximately 37% of SOA is attributable to aromatic precursors, using PM yields estimated for
23 NOx-limited conditions. Reactions of O3 on the surfaces of particles, in particular those with humic
24 acid like composition, are instrumental in the processing of SOA and the release of
25 low-molecular-weight products such as HCHO (D'Anna et al., 2009, 628847). However, direct
26 reactions of O3 and atmospheric particles appear to be too slow to represent a major O3 sink in the
27 troposphere (D'Anna et al., 2009, 628847).
3.2.4. Temperature and Chemical Precursor Relationships
28 As might be expected based on the temperature dependence of many reactions involved in the
29 production and destruction of O3 and the temperature dependence of emissions processes such as
30 evaporation of hydrocarbon precursors and the emissions of biogenically important precursors such
31 as isoprene, ambient concentrations of O3 also show temperature dependence. Bloomer et al. (2009,
32 628855) determined the sensitivity of O3 to temperature at rural sites in the eastern U.S. They found
33 that O3 increased on average at rural (CASTNET) sites by ~3.2 ppbv/°C before 2002, and after 2002
34 by ~2.2 ppbv/°C. This change in sensitivity was largely the result of reductions in NOX emissions
35 from power plants. These results are in accord with model predictions by Wu et al. (2008, 629684)
36 showing that the sensitivity of O3 to temperature decreases with decreases in precursor emissions.
March 2011 3-13 DRAFT - DO NOT CITE OR QUOTE
-------
1 However, this study was basically confined to the eastern U.S., but results from Phoenix, AZ showed
2 basically no sensitivity of O3 to temperature (U.S. EPA, 2006, 088089).
3 The warmer months of the year are generally regarded as being the most conducive to O3
4 concentrations that are of concern for human health. However, Schnell et al. (2009, 180146) reported
5 observations of high O3 concentrations (maximum 1-h avg of 140 ppb; maximum 8-h avg of
6 120 ppb) in the Jonah-Pinedale gas fields in Wyoming during winter at temperatures of -17°C.
7 Potential factors contributing to these anomalously high concentrations include a highly reflective
8 snow surface, emissions of short-lived radical reservoirs (e.g., HONO and HCHO) and a very
9 shallow, stable boundary layer trapping these emissions (Schnell et al., 2009, 180146). Multiphase
10 processes might also be involved in the production of these short-lived reservoirs. At a temperature
11 of -17°C, the production of hydroxyl radicals (by the photolysis of O3 yielding O:D followed by the
12 reaction, O(:D) + H2O, needed to initiate hydrocarbon oxidation) is severely limited, suggesting that
13 another source of free radicals is needed. Radicals can be produced by the photolysis of molecules
14 such as HONO and HCHO which photolyze in optically thin regions of the solar spectrum. A similar
15 issue, in part due to the under-prediction of free radicals, has arisen in the Houston airshed where
16 chemistry transport models under-predict O3 (Olaguer et al., 2009, 200191).
17 Rather than varying directly with emissions of its precursors, O3 changes in a nonlinear
18 fashion with the concentrations of its precursors. At the low NOX concentrations found in remote
19 continental areas to rural and suburban areas downwind of urban centers (low-NOx regime), the net
20 production of O3 typically increases with increasing NOX. At the high NOX concentrations found in
21 downtown metropolitan areas, especially near busy streets and roadways and in power plant plumes,
22 there is scavenging (titration) of O3 by reaction with NO (high-NOx regime). In between these two
23 regimes, there is a transition stage in which O3 shows only a weak dependence on NOX
24 concentrations.
25 In the low-NOx regime, the overall effect of the oxidation of VOCs is to generate (or at least
26 not consume) free radicals, and O3 production varies directly with NOX. In the high-NOx regime,
27 NO2 scavenges OH radicals which would otherwise oxidize VOCs to produce peroxy radicals, which
28 in turn would oxidize NO to NO2. In this regime, O3 production is limited by the availability of free
29 radicals. The production of free radicals is in turn limited by the availability of solar UV radiation
30 capable of photolyzing O3 (in the Hartley bands) or aldehydes and/or by the abundance of VOCs
31 whose oxidation produce more radicals than they consume. There are a number of ways to refer to
32 the chemistry in these two chemical regimes. Sometimes the terms VOC-limited and NOx-limited
33 are used. However, there are difficulties with this usage because (1) VOC measurements are not as
34 abundant as they are for nitrogen oxides; (2) rate coefficients for reaction of individual VOCs with
35 free radicals vary over an extremely wide range; and (3) consideration is not given to CO nor to
36 reactions that can produce free radicals without involving VOCs. The terms NOx-limited and
37 NOx-saturated (e.g., Jaegle et al., 2001, 057250) will be used wherever possible to more adequately
38 describe these two regimes. However, the terminology used in original articles will also be used
39 here.
March 2011 3-14 DRAFT - DO NOT CITE OR QUOTE
-------
1 The chemistry of OH radicals, which are responsible for initiating the oxidation of
2 hydrocarbons, shows behavior similar to that for O3 with respect to NOX concentrations (e.g.,
3 Hameed et al., 1979, 029458: Poppe et al., 1993, 044229: Zimmermann and Poppe, 1993, 052378).
4 These considerations introduce a high degree of uncertainty into attempts to relate changes in O3
5 concentrations to emissions of precursors. There are no definitive rules governing the concentrations
6 of NOX at which the transition from NOx-limited to NOx-saturated conditions occurs. The transition
7 between these two regimes is highly spatially and temporally dependent and depends also on the
8 nature and abundance of the hydrocarbons that are present.
9 Trainer et al. (1993, 038672) and Olszyna et al. (1994, 038832) have shown that O3 and NOY
10 are highly correlated in rural areas in the eastern U.S. Trainer et al. (1993, 038672) also showed that
11 O3 concentrations correlate even better with NOZ than with NOY, as may be expected because NOZ
12 represents the amount of NOX that has been oxidized, forming O3 in the process. NOZ is equal to the
13 difference between measured total reactive nitrogen (NOY) and NOX and represents the summed
14 products of the oxidation of NOX. NOZ is composed mainly of HNO3, PAN and other organic
15 nitrates, particulate nitrate, and HNO4. Trainer et al. (1993, 038672) also suggested that the slope of
16 the regression line between O3 and NOZ can be used to estimate the rate of O3 production per NOX
17 oxidized (also known as the O3 production efficiency [OPE]). Ryerson et al. (1998, 048310: 2001,
18 016249) used measured correlations between O3 and NOZ to identify different rates of O3 production
19 in plumes from large point sources. A number of studies in the planetary boundary layer over the
20 continental U.S. have found that the OPE ranges typically from 1 to nearly 10. However, it may be
21 higher in the upper troposphere and in certain areas, such as the Houston-Galveston area in Texas.
22 Observations indicate that the OPE depends mainly on the abundance of NOX.
23 Various techniques have been proposed to use ambient NOX and VOC measurements to derive
24 information about the dependence of O3 production on their concentrations. For example, it has been
25 suggested that O3 formation in individual urban areas could be understood in terms of measurements
26 of ambient NOX and VOC concentrations during the early morning (e.g., NRC, 1991, 038041). In
27 this approach, the ratio of summed (unweighted) VOC to NOX is used to determine whether
28 conditions were NOx-limited or VOC-limited. This procedure is inadequate because it omits many
29 factors that are important for O3 production such as the impact of biogenic VOCs (which are
30 typically not present in urban centers during early morning); important differences in the ability of
31 individual VOCs to generate free radicals (rather than just total VOC) and other differences in O3
32 forming potential for individual VOCs (Carter, 1995, 052288): and changes in the VOC to NOX ratio
33 due to photochemical reactions and deposition as air moves downwind from urban areas (Milford et
34 al.. 1994. 038669).
35 Photochemical production of O3 generally occurs simultaneously with the production of
36 various other species such as HNO3, organic nitrates, and other oxidants such as hydrogen peroxide.
37 The relative rate of production of O3 and other species varies depending on photochemical
38 conditions, and can be used to provide information about O3-precursor sensitivity. Sillman (1995,
39 052346) and Sillman and He (2002, 052350) identified several secondary reaction products that
March 2011 3-15 DRAFT - DO NOT CITE OR QUOTE
-------
1 show different correlation patterns for NOx-limited and NOx-saturated conditions. The most
2 important correlations are for O3 versus NOY, O3 versus NOZ, O3 versus HNO3, and H2O2 versus
3 HNO3. The correlations between O3 and NOY, and O3 and NOZ are especially important because
4 measurements of NOY and NOX are more widely available than for VOCs. Measured O3 versus NOZ
5 (Figure 3-3) shows distinctly different patterns in different locations. In rural areas and in urban
6 areas such as Nashville, TN, O3 is highly correlated with NOZ. By contrast, in Los Angeles, CA, O3
7 is not as highly correlated with NOZ, and the rate of increase of O3 with NOZ is lower and the O3
8 concentrations for a given NOZ value are generally lower. The different O3 versus NOZ relations in
9 Nashville, TN and Los Angeles, CA reflects the difference between NOx-limited conditions in
10 Nashville versus an approach to NOx-saturated conditions in Los Angeles.
X
X
X
X
X X
X
X
X
X
X
10
20
NOZ (ppb)
30
40
Source: adapted with permission of American Geophysical Union from Trainer et al. (1993, 0386721, Sillman et al. (1998, 0522231, and Sillman and He (2002,
0523501
Figure 3-3. Measured concentrations of ozone and NOZ (NOY-NOX) during the afternoon at rural
sites in the eastern U.S. (grey circles) and in urban areas and urban plumes
associated with Nashville, TN (gray dashes); Paris, France (black diamonds); and
Los Angeles, CA (Xs).
11 The difference between NOx-limited and NOx-saturated regimes is also reflected in
12 measurements of H2O2. H2O2 production is highly sensitive to the abundance of free radicals and is
13 thus favored in the NOx-limited regime. Measurements in the rural eastern U.S. (Jacob et al., 1995,
14 052308). Nashville, TN (Sillman et al., 1998, 052223). and Los Angeles, CA (Sakugawa and Kaplan,
15 1989, 044129). show large differences in H2O2 concentrations between likely NOx-limited and
16 NOx-saturated locations.
March 2011
3-16
DRAFT - DO NOT CITE OR QUOTE
-------
3.3. Atmospheric Modeling
1 Chemistry-transport models (CTMs) have been widely used to compute the interactions
2 among atmospheric pollutants and their transformation products, and the transport and deposition of
3 pollutants. They have also been widely used to improve our basic understanding of atmospheric
4 chemical processes and to develop control strategies. To do this, CTMs solve a set of coupled, non-
5 linear partial differential equations, or continuity equations, for relevant chemical species. Jacobson
6 (2005, 684174) described the governing partial differential equations, different coordinate systems in
7 use, and the finite difference approximations used to solve the equations numerically. Because of
8 limitations imposed by the complexity and spatial-temporal scales of relevant physical and chemical
9 processes, the CTMs must include parameterizations of these processes, which include atmospheric
10 transport; the transfer of solar radiation through the atmosphere; chemical reactions; and removal to
11 the surface by turbulent motions and precipitation. Development of parameterizations for use in
12 CTMs requires data for three dimensional wind fields, temperatures, humidity, cloudiness, and solar
13 radiation; emissions data for primary (i.e., directly emitted from sources) species such as NOX, SO2,
14 NH3, VOCs, and primary PM; and chemical reactions.
15 The domains of CTMs extend from a few hundred kilometers on a side to the entire globe.
16 Most major regional (i.e., sub-continental) scale air-related modeling efforts at EPA rely on the
17 Community Multi-scale Air Quality modeling system (CMAQ) (Byun and Ching, 1999, 156314;
18 Byun and Schere, 2006, 090560). CMAQ's horizontal domain typically extends over North America
19 with efforts underway to extend it over the entire Northern Hemisphere. Note that CTMs can be
20 'nested' within each other as shown in Figure 3-4 which shows domains for CMAQ (version 4.6.1);
21 additional details on the model configuration and application are found in (U.S. EPA, 2009, 191774).
22 The figure shows the outer domain (36 km horizontal grid spacing) and two 12 km spatial resolution
23 (east and west) sub-domains. The upper boundary for CMAQ is typically set at 100 hPa, which is
24 located on average at about 16 km altitude, although in some recent applications the upper boundary
25 has been set at 50 hPa. These domains and grid spacings are quite common and can also be found in
26 a number of other models.
March 2011 3-17 DRAFT - DO NOT CITE OR QUOTE
-------
Figure 3-4. Sample CMAQ modeling domains: 36 km-grid-spacing; outer parent domain in
black; 12 km western U.S. (WUS) domain in red; 12 km eastern U.S. (EUS) domain in
blue.
1
2
3
4
5
6
1
8
9
10
11
12
13
14
15
16
17
The main components of a CTM such as EPA's CMAQ are summarized in Figure 3-5. The
capabilities of a number of CTMs designed to study local- and regional-scale air pollution problems
were summarized by Russell and Dennis (2000, 035563) and in the 2006 O3 AQCD (U.S. EPA,
2006, 088089). CMAQ is most often driven by the MM5 mesoscale meteorological model (Seaman,
2000, 035562). though it may be driven by other meteorological models including the Weather
Research Forecasting (WRF) model and the Regional Atmospheric Modeling System (RAMS)
(ATMET, 2011, 677541). With the realization of the importance of intercontinental transport and the
need to consider exchange among different media, the domains of models such as RAMS have been
extended and interactions with other geophysical compartments such as land and ocean have been
considered. For example, the Ocean- Land- Atmosphere Model (OLAM) was developed to extend
the capability of RAMS to the global scale (Walko and Avissar, 2008, 665033).
Simulations of pollution episodes over regional domains have been performed with a
horizontal resolution down to 1 km; see the application and general survey results reported in Ching
et al. (2006, 090300). However, simulations at such high resolution require better parameterizations
of meteorological processes such as boundary layer fluxes, deep convection, and clouds (Seaman,
2000, 035562). Finer spatial resolution is necessary to resolve features such as urban heat island
circulation; sea, bay, and land breezes; mountain and valley breezes; and the nocturnal low-level jet,
March 2011
3-18
DRAFT - DO NOT CITE OR QUOTE
-------
1 all of which can affect pollutant concentrations. Other major air quality systems used for regional
2 scale applications include the Comprehensive Air Quality Model with extensions (CAMx)
3 (ENVIRON, 2005, 677542) and the Weather Research and Forecast model with Chemistry
4 (WRF/Chem) (NOAA, 2010, 677543V
Initial/Boundary
Conditions and
Continuous Updates
of Met. Fields
from Observations
Meteorological
Model
Emissions
Model
Anthropogenic
(point, area sources)
Biogenic Emissions
Initial/Boundary
Conditions
Photolysis Rates
Gas-Phase
Chemistry //Aerosol
Chemistry
Deposition V '/ and
Microphysics
Chemistry Transport Model
Visualization of Output
Process Analyses
Figure 3-5. Main components of a comprehensive atmospheric chemistry modeling system,
such as the U.S. EPA's Community Model for Air Quality (CMAQ) System.
5 CMAQ and other grid-based or Eulerian air quality models subdivide the modeling domain
6 into a three-dimensional array of grid cells. Spatial derivatives in the species continuity equations are
7 cast in finite-difference form over this grid and a system of equations for the concentrations of all the
8 chemical species in the model are solved numerically at each grid point. Time-dependent continuity
9 or mass conservation equations are solved for each species in each grid cell including terms for
10 transport, chemical production and destruction, and emissions and deposition (if relevant). Chemical
11 processes are simulated with ordinary differential equations, and transport processes are simulated
12 with partial differential equations. Because of a number of factors such as the different time scales
March 2011
3-19
DRAFT - DO NOT CITE OR QUOTE
-------
1 inherent in different processes, the coupled, nonlinear nature of the chemical process terms, and
2 computer storage limitations, not all of the terms in the equations are solved simultaneously in three
3 dimensions. Instead, operator splitting, in which terms in the continuity equation involving
4 individual processes are solved sequentially, is used.
5 The most common approach to setting up the horizontal domain is to nest a finer grid within a
6 larger domain of coarser resolution. However, there are other strategies such as the stretched grid and
7 the adaptive grid. In a stretched grid, the grid's resolution continuously varies throughout the
8 domain, thereby eliminating any potential problems with the sudden change from one resolution to
9 another at the boundary. Caution should be exercised in using such a formulation because certain
10 parameterizations like those for convection might be valid on a relatively coarse grid scale but may
11 not be valid on finer scales. Adaptive grids are not fixed at the start of the simulation, but instead
12 adapt to the needs of the simulation as it evolves. They have the advantage that they can resolve
13 processes at relevant spatial scales. However, they can be very slow if the situation to be modeled is
14 complex. Additionally, if adaptive grids are used for separate meteorological, emissions, and
15 photochemical models, there is no reason a priori why the resolution of each grid should match, and
16 the gains realized from increased resolution in one model will be wasted in the transition to another
17 model. The use of finer horizontal resolution in CTMs will necessitate finer-scale inventories of land
18 use and better knowledge of the exact paths of roads, locations of factories, and, in general, better
19 methods for locating sources and estimating their emissions.
20 The vertical resolution of these CTMs is variable and usually configured to have more layers
21 in the PEL and fewer higher up. Because the height of the boundary layer is of critical importance in
22 simulations of air quality, improved resolution of the boundary layer height would likely improve air
23 quality simulations. Additionally, current CTMs do not adequately resolve fine-scale features such as
24 the nocturnal low-level jet in part because little is known about the nighttime boundary layer.
25 The meteorological fields are produced either by other numerical prediction models such as
26 those used for weather forecasting (e.g., MM5, WRF), and/or by assimilation of satellite data. The
27 flow of information shown in Figure 3-5 has most often been unidirectional in the sense that
28 information flows into the CTM (large box) from outside; feedbacks on the meteorological fields and
29 on boundary conditions (i.e., out of the box) have not been included. However, CTMs now have the
30 capability to consider these feedbacks as well; see, for example, Binkowski et al. (2007, 090563) and
31 the Weather Research and Forecast model with Chemistry (WRF/Chem).
32 Because of the large number of chemical species and reactions that are involved in the
33 oxidation of realistic mixtures of anthropogenic and biogenic hydrocarbons, condensed mechanisms
34 must be used in atmospheric models. These mechanisms can be tested by comparison with smog
35 chamber data. However, the existing chemical mechanisms often neglect many important processes
36 such as the formation and subsequent reactions of long-lived carbonyl compounds, the incorporation
37 of the most recent information about intermediate compounds, and heterogeneous reactions
38 involving cloud droplets and aerosol particles.
March 2011 3-20 DRAFT - DO NOT CITE OR QUOTE
-------
1 The initial conditions, or starting concentration fields of all species computed by a model, and
2 the boundary conditions, or concentrations of species along the horizontal and upper boundaries of
3 the model domain throughout the simulation, must be specified at the beginning of the simulation.
4 Both initial and boundary conditions can be estimated from models or data or, more generally, model
5 + data hybrids. Because data for vertical profiles of most species of interest are very sparse, results
6 of model simulations over larger, usually global, domains are often used. As might be expected, the
7 influence of boundary conditions depends on the lifetime of the species under consideration and the
8 time scales for transport from the boundaries to the interior of the model.
9 Chemical kinetics mechanisms representing the important reactions occurring in the
10 atmosphere are used in CTMs to estimate the rates of chemical formation and destruction of each
11 pollutant simulated as a function of time. The Master Chemical Mechanism (MCM, 2010, 677544) is
12 viewed as a benchmark database providing as near an explicit treatment of chemical reactions in the
13 troposphere as is possible. The MCM currently includes over 12,600 reactions and 4,500 species.
14 However, mechanisms that are this comprehensive are still computationally too demanding to be
15 incorporated into CTMs for regulatory use. Simpler treatments of tropospheric chemistry have been
16 assembled by combining chemical species into mechanisms that group together compounds with
17 similar chemistry. It should be noted that because of different approaches to the lumping of organic
18 compounds into surrogate groups for computational efficiency, chemical mechanisms can produce
19 different results under similar conditions. Jimenez et al. (2003, 156611) provided brief descriptions
20 of the features of the main mechanisms in use and compared concentrations of several key species
21 predicted by seven chemical mechanisms in a box-model simulation over 24 hours. There are several
22 of these mechanisms (CB04, CB05, SAPRC) that have been incorporated into CMAQ (see for
23 example, Luecken et al., 2008, 190084) and Fuentes et al. (2007, 191251) for RACM2. The CB
24 mechanism is currently undergoing extension (CB06) to include, among other things, longer-lived
25 species to better simulate chemistry in the remote and upper troposphere. These mechanisms were
26 developed primarily for homogeneous gas phase reactions and treat multi-phase chemical reactions
27 in a very cursory manner, if at all. As an example of the effects of their neglect, models such as
28 CMAQ could have difficulties with capturing the regional nature of O3 episodes, in part because of
29 uncertainty in the chemical pathways converting NOX to HNO3 and recycling of NOX (e.g.,
30 Godowitch et al., 2008, 139006; Hains et al., 2008, 137411). Much of this uncertainty also involves
31 multi-phase processes as described by, for example, Thornton et al. (2010, 386870).
32 CMAQ and other CTMs incorporate processes and interactions of aerosol-phase chemistry
33 (Binkowski and Roselle, 2003, 191769: Gaydos et al., 2007, 139738: Zhang and Wexler, 2008,
34 191770). There have also been several attempts to study the feedbacks of chemistry on atmospheric
35 dynamics using meteorological models like MM5 and WRF (Grell et al., 2000, 048047: Liu et al.,
36 2001, 048201: Lu et al., 1997, 048202: Park et al., 2001, 044169). This coupling is necessary to
37 accurately simulate feedbacks which may be caused by the heavy aerosol loading found in forest fire
38 plumes (Lu et al., 1997, 048202: Park et al., 2001, 044169) or in heavily polluted areas. Photolysis
March 2011 3-21 DRAFT-DO NOT CITE OR QUOTE
-------
1 rates in CMAQ can now be calculated interactively with model produced O3, NO2, and aerosol fields
2 (Binkowski et al, 2007, 090563).
3 Spatial and temporal characterizations of anthropogenic and biogenic precursor emissions
4 must be specified as inputs to a CTM. Emissions inventories have been compiled on grids of varying
5 resolution for many hydrocarbons, aldehydes, ketones, CO, NH3, and NOX. Emissions inventories
6 for many species require the application of algorithms for calculating the dependence of emissions
7 on physical variables, such as temperature, and to convert the inventories into formatted emission
8 files which can be used by a CTM. For example, preprocessing of emissions data for CMAQ often is
9 done by the Spare-Matrix Operator Kernel Emissions (SMOKE) system (CEMPD, 2011, 677545).
10 For many species, information concerning the temporal variability of emissions is lacking, so long-
11 term annual averages are used in short-term, episodic simulations. Annual emissions estimates are
12 often modified by the emissions model to produce emissions more characteristic of the time of day
13 and season. Significant errors in emissions can occur if inappropriate time dependence is used.
14 Additional complexity arises in model calculations because different chemical mechanisms can
15 include different species, and inventories constructed for use with one mechanism must be adjusted
16 to reflect these differences in another.
17 Each of the model components described above has associated uncertainties; and the relative
18 importance of these uncertainties varies with the modeling application. The largest errors in
19 photochemical modeling are still thought to arise from the meteorological and emissions inputs to
20 the model (Russell and Dennis, 2000, 035563). While the effects of poorly specified boundary
21 conditions propagate through the model's domain, the effects of these errors remain undetermined.
22 Because many meteorological processes occur on spatial scales smaller than the model's vertical or
23 horizontal grid spacing and thus are not calculated explicitly, parameterizations of these processes
24 must be used. These parameterizations introduce additional uncertainty. Because the chemical
25 production (and loss) terms in the continuity equations for individual species are numerically
26 coupled, the chemical calculations must be performed iteratively until calculated concentrations
27 converge to within some preset criterion. The number of iterations and the convergence criteria
28 chosen also can introduce error.
29 The performance of CTMs must be evaluated by comparison with field data as part of a cycle
30 of model evaluations and subsequent improvements (e.g., NRC Committee on Models in the
31 Regulatory Decision Process, 2007, 632611). However, they are too demanding of computational
32 time to have the full range of their sensitivities examined by using Monte Carlo techniques (NRC
33 Committee on Models in the Regulatory Decision Process, 2007, 632611). Models of this
34 complexity are evaluated by comparison with field observations for O3 and other species.
35 Evaluations of the performance of CMAQ are given in Arnold et al. (2003, 087579). Eder and Yu
36 (2005, 089229). Appel et al. (2005, 089227). and Fuentes and Raftery (2005, 087580). Discrepancies
37 between model predictions and observations can be used to point out gaps in current understanding
38 of atmospheric chemistry and to spur improvements in parameterizations of atmospheric chemical
39 and physical processes. Model evaluation does not merely involve a straightforward comparison
March 2011 3-22 DRAFT - DO NOT CITE OR QUOTE
-------
1 between model predictions and the concentration field of the pollutant of interest. Such comparisons
2 may not be meaningful because it is difficult to determine if agreement between model predictions
3 and observations truly represents an accurate treatment of physical and chemical processes in the
4 CTM or the effects of compensating errors in complex model routines (in other words, a model
5 evaluator would want to know if they had the right answer for the right reasons). Ideally, each of the
6 model components (emissions inventories, chemical mechanism, meteorological driver) should be
7 evaluated individually. However, this is rarely done in practice. In addition to comparisons between
8 concentrations of calculated and measured species, comparisons of correlations between measured
9 primary VOCs and NOX and modeled VOCs and NOX are especially useful for evaluating results
10 from chemistry-transport models. Likewise, comparisons of correlations between measured species
11 and modeled species can be used to provide information about the chemical state of the atmosphere
12 and to evaluate model representations (including: O3 production from NOX; O3-NOX-VOC
13 sensitivity; and the general accuracy of photochemical representations). A CTM that demonstrates
14 the accuracy of both its computed VOC and NOX in comparison with ambient measurements, and
15 the spatial and temporal relations among the critical secondary species associated with O3 has a
16 higher probability of representing O3-precursor relations correctly than one that does not.
17 The above techniques are sometimes referred to as "static" in the sense that individual model
18 variables are compared to observations. It is also crucial to understand the (dynamic) response to
19 changes in inputs and to compare the model responses to those that are observed. These tests might
20 involve changes in some natural forcing or in emissions from an anthropogenic source. As an
21 example, techniques such as the direct decoupled method (DDM) (Bunker, 1981, 040504; Bunker et
22 al, 2002, 665029) could be used in this regard. However, the observational basis for comparing a
23 model's response is largely unavailable for many problems of interest, in large part because
24 meteorological conditions are also changing while the emissions are changing. As a result, methods
25 such as BBM are used mainly to address the effectiveness of emissions controls.
3.3.1. Global Scale CTMs
26 With recognition of the global nature of many air pollution problems, global scale CTMs have
27 been applied to regional scale pollution problems (e.g., NRC Committee on the Significance of
28 International Transport of Air Pollutants, 2009, 202867). Global-scale CTMs are used to address
29 issues associated with global change, to characterize long-range transport of air pollutants, and to
30 provide boundary conditions for the regional-scale models. The upper boundaries of global scale
31 CTMs extend anywhere from the tropopause (~8 km at the poles to ~16 km in the tropics) to the
32 mesopause at -80 km, in order to obtain more realistic boundary conditions for problems involving
33 stratospheric dynamics and chemistry. The global-scale CTMs consider the same processes shown in
34 Figure 3-5 for the regional scale models. In addition, many of the same issues that have arisen for the
35 regional models have also arisen for the global scale models (see Emmerson and Evans, 2009,
36 605119). For example, predictions of HNO3 were found to be too high and predictions of PAN were
37 found to be too low over the U.S. during summer in the MOZART model (Fang et al., 2010,
March 2011 3-23 DRAFT - DO NOT CITE OR QUOTE
-------
1 665391). Similar findings were obtained in a box model of upper tropospheric chemistry (Henderson
2 etal.. 2010. 674771).
3 The GEOS-Chem model is a community global scale CTM that has been widely used to study
4 issues associated with the intra- and inter-hemispheric transport of pollution and global change
5 (Harvard University, 2010, 677546). Comparisons of the capabilities of GEOS-Chem and several
6 other models to simulate intra-hemispheric transport of pollutants are given in a number of articles
7 (e.g., Fiore et al., 2009, 665030: Reidmiller et al, 2009, 644826). Reidmiller et al. (2009, 644826)
8 showed comparisons among 18 global models and their ensemble average to spatially and monthly
9 averaged observations of O3 at CASTNET sites (see Figure 3-6). These results show that the
10 multi-model ensemble agrees much better with the observations than do most of the individual
11 models. The GEOS-Chem model was run for two grid spacings, 4°x4.5° and 2°x2.5° with very
12 similar results that lie close to the ensemble average. In general, the model ensemble and the two
13 GEOS-Chem simulations are much closer to the observations in the Mountain West than in the
14 Southeast. In particular, there are sizable over-predictions by most of the models in the Southeast
15 during summer, the time when major O3 episodes occur.
Mountain West Region
i—i—i—i
CAMCHEM
ECHAM5
- EMEP
B - FRSGCUCI
-a- GEMAQ-EC
-a- GEMAQ-vl pO
GEOSChem-v07
-£— GEOSChem-v4S
GISS-PUCCINI
GM
NCA-vSSz
e- LLNL-IWPACT
e- MOZARTGFDL
MOZECH
OsloCTMZ
a- TM5-JRC
A-DBS
•— Multi-model mean
M A M J
A S O N D
Source: Used with permission from Copernicus Publications, Reidmiller et al. (2009, 6448261
Figure 3-6. Comparison of global CTM predictions of maximum daily 8-h avg ozone
concentrations and multi-model mean with monthly averaged CASTNET
observations in the Mountain West and Southeast regions of the U.S.
March 2011
3-24
DRAFT - DO NOT CITE OR QUOTE
-------
1 The issue of overestimating O3 is not limited to global models. Godowitch et al. (2008,
2 139006). Gilliland et al. (2008, 606585) and Nolte et al. (2008, 679712) found positive O3 biases in
3 regional models over the eastern U.S., as well, which they largely attributed to uncertainties in
4 temperature, relative humidity and planetary boundary layer height. Agreement between monthly
5 average values is expected to be better than with daily values because of a number of factors
6 including the increasing uncertainty of emissions at finer time resolution. Kasibhatla and Chameides
7 (2000, 052237) found that the accuracy of simulations improved in their simulations as the averaging
8 time of both the simulation and the observations increased.
9 Simulations of the effects of long-range transport at particular locations must be able to link
10 multiple horizontal resolutions from the global to the local scale. Because of limitations on
11 computational resources, global simulations are not made at the same horizontal resolutions found in
12 the regional scale models, i.e., down to 1-4 km resolution on a side. They are typically conducted
13 with a horizontal grid spacing of l°-2° of latitude and longitude (or roughly 100-200 km at mid-
14 latitudes). Some models such as GEOS-Chem have the capability to include nested models at a
15 resolution of 0.5°x0.667° (e.g., Wang et al., 2009, 622281) and efforts are underway to achieve even
16 higher spatial resolution. Another approach is to nest regional models within GEOS-Chem. Caution
17 must be exercised with nesting different models because of differences in chemical mechanisms and
18 numerical schemes, and in boundary conditions between the outer and inner models. As an example
19 of these issues, surface O3 concentrations that are too high have been observed in models in which
20 CMAQ was nested inside of GEOS-Chem (see e.g., Lam and Fu, 2010, 665031 for one way to
21 address this issue). The high O3 was the result of stratospheric O3 intruding into the CMAQ domain,
22 followed by too rapid downward mixing of this O3 in CMAQ. Ozone has large vertical gradients in
23 the upper troposphere that must be preserved if its downward transport is to be simulated correctly.
24 Errors in parameterizations of vertical transport (e.g., vertical velocities, diffusivities) leading to too
25 rapid mixing may be involved. It is also highly likely that using a vertical resolution in CMAQ that
26 is too coarse is involved, coupled with using fewer layers in CMAQ than in the driving MM5 or
27 WRF meteorological model. As a result of the above factors, O3 gradients are eliminated and O3 is
28 mixed too rapidly in the upper troposphere. Efforts are also being made to extend the domain of
29 CMAQ over the Northern Hemisphere. In this approach, the same numerical schemes are used for
30 transporting species and the same chemistry is used throughout all spatial scales. Finer resolution in
31 models of any scale can only improve scientific understanding to the extent that the governing
32 processes are accurately described. Consequently, there is a crucial need for observations at the
33 appropriate scales to evaluate the scientific understanding represented by the models.
3.4. Policy Relevant Background Concentrations
34 The background concentrations of O3 that are useful for risk and policy assessments informing
35 decisions about the NAAQS are referred to as PRB concentrations. PRB concentrations have
36 historically been defined by EPA as those concentrations that would occur in the U.S. in the absence
March 2011 3-25 DRAFT - DO NOT CITE OR QUOTE
-------
1 of anthropogenic emissions in continental North America (CNA) defined here as the U.S., Canada,
2 and Mexico. For this document, PRB concentrations include contributions from natural sources
3 everywhere in the world and from anthropogenic sources outside CNA. Background concentrations
4 so defined, facilitate separation of pollution that can be controlled by U.S. regulations or through
5 international agreements with neighboring countries from those that are judged to be generally
6 uncontrollable by the U.S. Over time, consideration of potential broader ranging international
7 agreements may lead to alternative determinations in which O3 precursor source contributions should
8 be considered by EPA as part of PRB.
9 Contributions to PRB O3 include photochemical reactions involving natural emissions of
10 VOCs, NOX, and CO as well as the long-range transport of O3 and its precursors from outside CNA
11 and the STE of O3. These sources have the greatest potential for producing the highest PRB
12 concentrations, and therefore are discussed in greater detail below. Natural sources of O3 precursors
13 include biogenic emissions, wildfires, and lightning. Biogenic emissions from agricultural activities
14 in CNA are not considered in the formation of PRB O3. Definitions of background and approaches to
15 derive background concentrations were reviewed in the 2006 O3 AQCD (U.S. EPA, 2006, 088089)
16 and Reid et al. (2008, 665032).
3.4.1. Contributions from Anthropogenic Emissions Outside North
America
17 In addition to emissions from North America, emissions from Eurasia have contributed to the
18 global burden of O3 in the atmosphere and to the U.S. (e.g., NRC Committee on the Significance of
19 International Transport of Air Pollutants, 2009, 202867 and references therein). Because the mean
20 tropospheric lifetime of O3 is 30-35 days (Hsu and Prather, 2009, 629687). O3 can be transported
21 from continent to continent and around the globe in the Northern Hemisphere and O3 produced by
22 U.S. emissions can be recirculated around northern mid-latitudes back to the U.S. High elevation
23 sites are most susceptible to the intercontinental transport of pollution especially during spring. An
24 O3 concentration of ~85 ppb was observed at Mt. Bachelor Observatory, OR (elevation 2,700 m) on
25 April 22, 2006 with a number of occurrences of O3 >60 ppb from mid-April to mid-May of 2006.
26 Calculations using GEOS-Chem, a global-scale, chemistry-transport model, indicate that Asia
27 contributed 9 ± 3 ppb to a modeled mean concentration of 53 ± 9 ppb O3 at Mt. Bachelor during the
28 same period compared to measured concentrations of 54 ± 10 ppb (Zhang et al., 2008, 624402).
29 Zhang et al. (2008, 624402) also calculated a contribution of 5 to 7 ppb to surface O3 over the
30 western U.S. during that period from Asian anthropogenic emissions. They also estimated an
31 increase in NOX emissions of- 44% from Asia from 2001 to 2006 resulting in an increase of 1-2 ppb
32 in O3 over North America.
33 Cooper et al. (2010, 380093) analyzed all available O3 measurements in the free troposphere
34 above western North America at altitudes of 3-8 km (above sea level) during April and May of 1995
35 to 2008 (i.e., times when intercontinental transport is most prominent). They derived a trend of 0.63
36 ± 0.34 ppb/year in median O3 concentrations with indication of a similar rate of increase since 1984.
March 2011 3-26 DRAFT - DO NOT CITE OR QUOTE
-------
1 Back trajectories that were likely to have been strongly and recently influenced by North American
2 emissions were filtered out, resulting in a trend of 0.71 ± 0.45 ppb/year. Considering only trajectories
3 with an Asian origin resulted in a trend of 0.80 ± 0.34 ppb/year. These results suggest that local
4 North American emissions were not responsible for the measured O3 increases. This O3 could have
5 been produced from natural and anthropogenic precursors in Asia and Europe with some
6 contribution from North American emissions that have circled the globe. Cooper et al. (2010,
7 380093) also found that it is unlikely that the trends in tropospheric O3 are associated with trends in
8 stratospheric intrusions. Note, however, that these results relate to O3 trends above ground level and
9 not to surface O3. Model results (e.g., Zhang et al., 2008, 624402) show that surface O3 contributions
10 from Asia are much smaller than those derived in the free troposphere because of dilution and
11 chemical destruction during downward transport to the surface.
12 There are limited cases where PRB conditions are observable at lower elevations, for example
13 at Trinidad Head, CA at times during spring (Goldstein et al., 2004, 087780: Oltmans et al., 2008,
14 615534). Figure 3-7 shows the time series of daily maximum 8-h avg O3 concentrations measured at
15 Trinidad Head from April 18, 2002 through December 31, 2009. The data show pronounced seasonal
16 variability with spring maxima and summer minima. Springtime concentrations typically range from
17 40 to 50 ppb with a number of occurrences >50 ppb. The two highest daily maxima were 60 and
18 62 ppb. The data also show much lower concentrations during summer, with concentrations typically
19 ranging between 20 and 30 ppb. Oltmans et al. (2008, 615534) examined the time series of O3 and
20 back trajectories reaching Trinidad Head. They found that springtime maxima (April-May) were
21 largely associated with back trajectories passing over the Pacific Ocean and most likely entraining
22 emissions from Asia, with minimal interference from local sources. However, Parrish et al. (2009,
23 616076) noted that only considering trajectories coming from a given direction is not sufficient for
24 ruling out local continental influences, as sea breeze circulations are complex phenomena involving
25 vertical mixing and entrainment of long-shore components. They found that using a wind speed
26 threshold, in addition to a criterion for wind direction, allowed for determination of background
27 trajectories not subject to local influence; as judged by measurements of chemical tracers such as
28 CO2, MTBE and radon. By applying the two criteria for wind speed and direction, they found that
29 Trinidad Head met these criteria only 30% of the time during spring. Goldstein et al. (2004, 087780)
30 used CO2 as an indicator of exchange with the local continental environment and found that O3
31 concentrations were higher by about 2-3 ppb when filtered against local influence indicating higher
32 O3 in air arriving from over the Pacific (see Figure 3-8). At Trinidad Head during spring, O3 is more
33 likely to be titrated by local emissions of NOX than to be photochemically produced (Parrish et al.,
34 2009, 616076). At other times of the year, Trinidad Head is less strongly affected by air passing over
35 Asia and many trajectories have long residence times over the semi-tropical and tropical Pacific
36 Ocean, where O3 concentrations are much lower than they are at mid-latitudes. The application of
37 the Trinidad Head data to PRB conditions would require the use of screening procedures adopted by
38 Parrish et al. (2009, 616076) and the application of photochemical models to determine the extent
March 2011 3-27 DRAFT - DO NOT CITE OR QUOTE
-------
1 either of titration of O3 by fresh NOX emissions and the extent of local production of O3 from these
2 emissions.
0.07
0.06
Trinidad Head
999
a. o •
O) O)
W Q
TO zj
9999
!= Q. O '
TO zj
9999
!= Q. O '
0)
Q
999
Q. O J=
(U o> ™ =i o> o> ™
^>WC|2^>WC|2
999
Q. O J=
a> a> ™
W Q ^
_ CD CD O
9 9 9 V
c Q. O b
=i a>
-------
1
2
3
4
5
6
7
9
10
11
12
13
14
15
16
17
18
19
20
Measured (CO-. < 383 ppm)
M«asur«d (COJ > 383 torn)
A
""• GEOS-CH^M
' MOZART .
QJ O
§ «
110
120
130
140
Day of Year 2002
Observations: 38 ± 7 ppb (unfiltered)
41 ± 5 ppb (filtered against local influence)
GEOS-Chem model: 39 ±5 ppb
MOZART-2 global model: 37 ±9 ppb
Source: Used with permission from Americn Geophysical Union, Goldstein et al. (2004, 0877801
Figure 3-8. Time series of measured ozone and model calculated ozone at Trinidad Head, CA,
during April and May, 2002.
Parrish et al. (2009, 616076) also examined data obtained at other marine boundary layer sites
on the Pacific Coast. These include Olympic NP, Redwood NP, Point Arena, and Point Reyes. Using
data from these sites, they derived trends in O3 of 0.46 ppb/year (with a 95% confidence interval of
0.13 ppb/year) during spring and 0.34 ppb/year (0.09 ppb/year) for the annual mean O3 increase in
air arriving from over the Pacific during the past two decades. Although O3 data are available from
the Channel Islands, Parrish et al. (2009, 616076) noted that these data are not suitable for
determining background influence because of the likelihood of circulating polluted air from the
South Coast Basin.
3.4.2. Contributions from the Stratosphere
As noted in the 2006 O3 AQCD (U.S. EPA, 2006,
, stratospheric air rich in O3 is
transported to the troposphere. Ozone is produced naturally by photochemical reactions in the
stratosphere as shown in Figure 3-1 in Section 3.2. Some of this O3 is transported downward into the
troposphere throughout the year, with maximum contributions during late winter and early spring
mainly in a process known as tropopause folding. These folds occur basically behind every cold
front, bringing stratospheric air with them. The tropopause should not be interpreted as a material
surface through which there is no exchange. Rather these folds should be thought of as regions in
which mixing of tropospheric and stratospheric air is occurring (Shapiro, 1980, 047756). This
imported stratospheric air contributes to the natural background of O3 in the troposphere, especially
in the free troposphere during winter and spring. STE also occurs during other seasons including
summer. Thompson et al., (2007, 090796) found that roughly 20-25% of tropospheric O3 over
northeastern North America during July-August 2004 was of stratospheric origin. This O3 can be
March 2011
3-29
DRAFT - DO NOT CITE OR QUOTE
-------
1 mixed into the PEL where it can either be destroyed or transported to the surface. Yang et al. (2010,
2 628857) estimated that roughly 20% of free tropospheric O3 above coastal California in 2005 and
3 2006 was stratospheric in origin. Some of this O3 could also contribute to O3 at the surface.
4 It should be noted that there is considerable uncertainty in the magnitude and distribution of
5 this potentially important source of tropospheric O3. Stratospheric intrusions that reach the surface
6 are rare. Much more common are intrusions which penetrate only to the middle and upper
7 troposphere. However, O3 transported to the upper and middle troposphere can still affect surface
8 concentrations through various exchange mechanisms that mix air from the free troposphere with air
9 in the PEL. Substantial photochemical production of O3 in the troposphere also begins in late winter
10 and early spring; therefore, it cannot be assumed that O3 present at these times is only stratospheric
11 in origin. The basic atmospheric dynamics and thermodynamics of stratospheric-tropospheric
12 exchange were outlined in the 2006 O3 AQCD (U.S. EPA, 2006, 088089. Annex AX2.3.1).
13 Several instances of STE producing high concentrations of O3 around Denver and Boulder,
14 CO were analyzed by Langford et al. (2009, 491703). Several likely instances of STE, including one
15 of the cases analyzed by Langford et al. (2009, 491703) were also cited in the 2006 O3 AQCD
16 (U.S. EPA, 2006, 088089. Annex AX3, Section AX3.9). Clear examples of STE have also been
17 observed in southern Quebec province by Hocking et al. (2007, 608032). in accord with previous
18 estimates by Wernli et al. (2002, 052425) and James et al. (2003, 043286).
3.4.2.1. Other Natural Sources of Precursors to PRB Ozone Formation
19 Biomass burning consists of wildfires and the intentional burning of vegetation to clear new
20 land for agriculture and for population resettlement; to control the growth of unwanted plants on
21 pasture land; to manage forest resources with prescribed burning; to dispose of agricultural and
22 domestic waste; and as fuel for cooking, heating, and water sterilization. Globally, most wildfires
23 may be ignited directly as the result of human activities, leaving only 10-30% initiated by lightning
24 (Andreae, 1991, 078147). However, because fire management practices suppress natural wildfires,
25 the buildup of fire fuels increases the susceptibility of forests to more severe but less frequent fires in
26 the future. Thus there is considerable uncertainty in attributing the fraction of wildfire emissions to
27 human activities because the emissions from naturally occurring fires that would have been present
28 in the absence of fire suppression practices are not known. Contributions to NOX, CO and VOCs
29 from wild fires and prescribed fires are considered as precursors to PRB O3 formation.
30 Biomass burning also exhibits strong seasonality and interannual variability (van der Werf et
31 al., 2006, 157084). with most biomass burned during the local dry season. This is true for both
32 prescribed burns and wildfires. The unusually warm and dry weather in central Alaska and western
33 Yukon in the summer of 2004, for example, contributed to the burning of 11 million acres there.
34 Subsequent modeling by Pfister et al. (2005, 093009) showed that the CO contribution from these
35 fires in July 2004 was 33.1 (± 5.5) MT that summer, or in the range of the total U.S. anthropogenic
36 CO emissions during the same period. In addition to emissions from forest fires in the U.S.,
37 emissions from forest fires in other countries can be transported to the U.S., for example from boreal
March 2011 3-30 DRAFT - DO NOT CITE OR QUOTE
-------
1 forest fires in Canada (Mathur, 2008, 156742). Siberia (Generoso et al., 2007, 155786) and tropical
2 forest fires in the Yucatan Peninsula and Central America (Wang et al., 2006, 157109).
3 Estimates of biogenic VOC and CO emissions are made using the BEIS model with data from
4 the Biogenic Emissions Landcover Database (BELD) and annual meteorological data. VOC
5 emissions from vegetation were described in Section 3.2. As noted earlier, NOX is produced by
6 lightning. Kaynak et al. (2008, 486686) found contributions of 2 to 3 ppb PRB O3 centered mainly
7 over the southeastern U.S. during summer. Although total column estimates of lightning-produced
8 NOX are large compared to anthropogenic NOX during summer, lightning-generated NOX does not
9 contribute substantially to the NOX burden in the continental boundary layer. This is because only
10 2% of NOX production by lightning occurs within the boundary layer and most occurs in the free
11 troposphere (Fang et al., 2010, 665391). In addition, much of the NOX produced in the free
12 troposphere is converted to more oxidized N species during downward transport. However, Fang et
13 al. (2010, 665391) estimate these NOX oxidation products contribute ~l/3 to wet deposition by total
14 oxidized N species over the U.S.
3.4.3. Estimating PRB Concentrations
15 There are two approaches to estimating PRB concentrations that have been considered in
16 previous assessments. The first involves using measurements and the second the use of chemistry-
17 transport models. Section 3.9 of the 2006 O3 AQCD, (U.S. EPA, 2006, 088089). noted that estimates
18 of PRB concentrations cannot be obtained solely by examining measurements of O3 obtained at
19 relatively remote monitoring sites in the U.S. (AX3.2.3) because of the long-range transport from
20 anthropogenic source regions within North America. The 2006 O3 AQCD also noted that it is
21 impossible to determine sources of O3 without ancillary data that could be used as tracers of sources
22 or to calculate photochemical production and loss rates. As further noted by Reid et al. (2008,
23 665032). the use of monitoring data is essentially limited to the edges of the domain of interest. This
24 is because PRB O3 entering from outside North America can only be destroyed over North America
25 either through chemical reactions or by deposition to the surface. Within North America, PRB O3 is
26 only produced by interactions between natural sources and between North American natural sources
27 and precursors from other continents. PRB O3 as defined above is different from the baseline O3
28 defined by Chan and Vet (2010, 679710). Their baseline O3 refers to "O3 measured at a given site in
29 the absence of strong local influences". The current definition of PRB implies that only CTMs can
30 be used to estimate the range of PRB concentrations. A further advantage to using models is that the
31 entire range of O3 concentrations in different environments can be used to evaluate model
32 performance. However, there may be specific instances such as stratospheric intrusions that occur on
33 spatial scales too fine to be resolved by the current generation of global CTMs.
34 Estimates of PRB concentrations in the 2006 O3 AQCD (U.S. EPA, 2006, 088089) were based
35 on output from the GEOS-Chem model (Fiore et al., 2003, 051226). The GEOS-Chem model
36 estimates indicate that PRB O3 concentrations in eastern U.S. surface air are generally 15-35 ppb
37 from June through August. PRB concentrations decline from spring to summer. PRB O3
March 2011 3-31 DRAFT - DO NOT CITE OR QUOTE
-------
1 concentrations may be higher, especially at high altitude sites during the spring, due to enhanced
2 contributions from (1) pollution sources outside North America; and (2) stratospheric O3 exchange.
3 Only one model (GEOS-Chem (Harvard University, 2010, 677581)) was documented in the
4 literature for calculating PRB O3 concentrations (see Fiore et al, 2003, 051226). The simulated
5 monthly mean concentrations in different quadrants of the U.S. are typically within 5 ppbv of
6 observations at CASTNET sites, with no significant bias, except in the Southeast in summer when
7 the model is 8-12 ppbv too high. This bias might be due to excessive background O3 transported in
8 from the Gulf of Mexico and the tropical Atlantic Ocean or to inaccuracies in emissions inventories
9 within the U.S. The time series comparisons for specific sites show that the model simulates the day-
10 to-day variability of O3 and reveals no further bias. These evaluations focused on the afternoon hours
11 (1:00 p.m. to 5:00 p.m. local time), when surface measurements are representative of a deep mixed
12 layer that can be resolved with the model. At night, surface O3 depletion often takes place by titration
13 or deposition under local, stably stratified conditions, but such conditions cannot be simulated with
14 confidence by a global model. The issue is not only one of vertical resolution (the lowest layers in
15 GEOS-Chem extend to 20, 50, 100, 200, and 400 m above the local surface) but also of horizontal
16 resolution (2°x2.5°).
17 The model reproduced the occurrences of relatively high O3 at remote sites previously
18 reported by Lefohn et al. (2001, 016253). and shows that these can generally be explained by North
19 American pollution. Goldstein et al. (2005, 087880) presented comparisons of GEOS-Chem and
20 MOZART global model results with observations at Trinidad Head, CA, during April-May 2002.
21 The observations, filtered to remove local influence, averaged 41 ± 5 ppbv, as compared to GEOS-
22 Chem (39 ± 5 ppbv) and MOZART (37 ± 9 ppbv). Neither model was successful at reproducing the
23 weak day-to-day structure in the observations, but they showed no bias in the simulation of
24 occasional >50 ppbv days (see Figure 3-8).
25 Although many of the features of the day-to-day variability in O3 at relatively remote
26 monitoring sites in the U.S. are simulated reasonably well by Fiore et al. (2003, 051226).
27 uncertainties in the calculation of the temporal variability of O3 originating from different sources on
28 shorter time scales must be recognized. The uncertainties stem in part from an underestimate in the
29 seasonal variability in the STE of O3 (Fusco and Logan, 2003, 051229). the geographical variability
30 of this exchange, and the variability in the exchange between the free troposphere and the PEL in the
31 model. In addition, the relatively coarse spatial resolution in that version of GEOS-Chem (2°x2.5°)
32 limited the ability to provide separate estimates for cities located close to each other, and so only
33 regional estimates were provided for the 2006 O3 AQCD (U.S. EPA, 2006, 088089) based on the
34 results of Fiore et al. (2003, 051226).
35 Wang et al. (2009, 622281) recomputed PRB concentrations for 2001 using GEOS-Chem at
36 higher spatial resolution (l°x 1°) and not only for afternoon hours but for the daily maximum 8-h O3
37 concentration (the base and PRB results for the 2001 model year simulation are shown in Figure 3-9
38 for spring and Figure 3-10 for summer). These GEOS-Chem calculations represents the latest results
39 documented in the literature. However, all models undergo continuous updating of inputs,
March 2011 3-32 DRAFT - DO NOT CITE OR QUOTE
-------
1 parameterizations of physical and chemical processes, and inputs and improvements in model
2 resolution. Inputs that might be considered most relevant include emissions inventories and
3 meteorological fields. However, the model's results may not be particularly sensitive to changes in
4 model inputs, especially in the current context. For example, as noted above, increases in Asian
5 emissions only accounted for an increase of 1-2 ppb in background O3 even though Asian emissions
6 have increased by about 44% from 2001 to 2006. To the extent that results from an updated model
7 become available, they will be presented and used in the next draft of the ISA. In that case, the
8 results shown here are to be viewed more as illustrating the type of calculations that will ultimately
9 be used for informing NAAQS setting.
10 The base case O3 concentrations show two broad maxima with highest concentrations
11 extending throughout the Southwest, Mountainous West and the East in both spring and summer.
12 These maxima extend over many thousands of kilometers demonstrating that O3 is a regional
13 pollutant. Low-level outflow from the Northeast, the Southwest and over the Gulf of Mexico is also
14 apparent. The PRB O3 concentrations are higher in spring than in summer over the entire U.S.
15 However, highest concentrations are found over the Mountainous West during spring. The resulting
16 PRB concentrations, 26.3 ± 8.3 ppb for summer, are consistent with those reported by Fiore et al.
17 (2003, 051226) of 26 ± 7 ppb, suggesting horizontal resolution was not a significant factor limiting
18 the accuracy of the earlier results. In addition to computing North American PRB contributions,
19 Wang et al. (2009, 622281) also computed U.S. background concentrations (i.e., including
20 anthropogenic contributions from Canada and Mexico) of 29.6 ± 8.3 ppb with higher contributions
21 near the Canadian and Mexican borders.
22 Panels a-d of Figure 3-11 show a comparison of O3 calculated by the base and PRB model
23 cases with measurements at low (<1500 m [1.5 km]) and high (>1500 m [1.5km]) elevation
24 CASTNET sites. Note that all the elevated sites are located in the West. Results are then aggregated
25 within these two sets of sites. In general, the model captures the behavior of O3 observed across the
26 concentration distribution, although there are some differences. In particular, the base model tends to
27 under-predict O3 at elevated sites. The reasons for this are not entirely clear and may be due to the
28 under-predictions of intercontinental transport, downward transport of stratospheric O3, contributions
29 from local and regional pollution, or some combination of these factors. Lower bounds to PRB
30 concentrations tend to be higher by several ppb at high elevations than at low elevations, reflecting
31 the altitude dependence of PRB sources such as stratospheric-tropospheric exchange and
32 intercontinental transport. In addition, PRB concentrations tend to increase with increasing base
33 model (and measured) concentrations at higher elevation sites, particularly during spring. At low
34 elevation sites, there is some indication that PRB concentrations decrease with increasing base
35 model concentrations at the upper end of the concentration range, i.e., during episode conditions.
March 2011 3-33 DRAFT - DO NOT CITE OR QUOTE
-------
April-May
O, (ppb)
I
Source: Adapted with permission from Wang et al. (2009, 6222811
Figure 3-9. Mean daily 8-h max ozone concentrations in surface air for the base case (top) and
PRB case (bottom) in April-May, 2001.
March 2011
3-34
DRAFT - DO NOT CITE OR QUOTE
-------
line-August
O, (ppb)
Source: Adapted with permission from Elsevier Ltd., Wang et al. (2009, 6222811
Figure 3-10. Mean daily 8-h max ozone concentrations in surface air for base case (top) and PRB
case (bottom) in June-August, 2001.
March 2011
3-35
DRAFT - DO NOT CITE OR QUOTE
-------
Surface Sites ( < 1.5 km )
0.1 0.2 0.3 0.4 0.5 0.6 0.7 (
Cumulative probability
» 03 [Obs-CASTNET) • 03 (GEOSChem - Base run) 03 (PRB)
Surface Sites (< 1.5 km)
0.2 0.3 0.4 0.5 0.6 0.7 0.8
Cumulative probability
* O3 (Obs-CASTNET) • 03 (GEOSChem - Base run) O3 (PRB)
Elevated Sites (> 1.5km)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Cumulative probability
* 03 (Obs-CASTNET) • 03 (GEOSChem - Base run) 03 (PRBj
d
Elevated Sites ( > 1.5 km )
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
* 03 (Obs-CASTNET) • 03 (GEOSChem - Base run) O3 (PRB)
Figure 3-11. Distribution of ozone measured at CASTNET sites (blue diamonds), ozone
calculated by the GEOS-Chem base model (red squares) and PRB ozone (green
triangles) at (a) sites <1500 m elevation and (b) at sites >1500 m elevation for April -
May; and (c) at sites <1500 m elevation and (d) at sites >1500 m elevation for June-
July-August.
1 Table 3-1 shows mean concentrations (± SD) of daily max 8-h avg O3 concentrations at these
2 CASTNET sites and GEOS-Chem predictions for the base model and PRB for spring and summer.
3 At eastern sites, seasonal maxima occur during spring or summer. In the East (the first five entries in
4 Table 3-1), the base model mean is generally within a few ppb of measurements during the spring.
5 The largest difference at the sites shown in Table 3-1 occurs at the Everglades site, where mean
6 spring concentrations are over-predicted by 10 ppb. At most western sites, seasonal maxima are
7 observed to occur during spring as opposed to summer. The base model under-predicts mean
8 concentrations during spring at the high-elevation sites: Centennial by ~8 ppb and at Yellowstone by
9 ~5 ppb. However, maximum concentrations at these sites are too high by only 5 and 3 ppb. At the
10 other western sites examined, the model-predicted means do not differ significantly from the
11 measurements during spring, except at Pinnacles, NM.
12 Disagreements between model-predicted versus observed mean concentrations at eastern sites
13 tend to be larger during summer than during spring with over-predictions by the base model
March 2011
3-36
DRAFT - DO NOT CITE OR QUOTE
-------
1 generally higher and at more locations. In particular, summer mean O3 concentrations at the
2 Everglades site are over-predicted by 21 ppb and by 10 ppb at the Virgin Islands site. The Virgin
3 Islands NP site appears to have not been affected by U.S. emissions, as can be seen from the close
4 agreement between the base case and the PRB case and from wind roses calculated for these two
5 sites indicating that flows affecting these sites are mainly easterly/southeasterly in summer. The
6 over-predictions at the Virgin Islands site imply that air affecting this site, i.e., coming from the
7 tropical Atlantic Ocean, is too high. As a result, inflow into the Gulf of Mexico may also be too high.
8 Similar considerations apply to the excess at the Everglades site. However, the Everglades site is
9 often subject to flow from the northeast and production of O3 from emissions from Miami.
10 The base model under-predicts summer mean O3 concentrations by 4 and 7 ppb at the
11 Yellowstone and Centennial sites, with much larger under-predictions at the Lassen and Pinnacles
12 sites. At the other western sites examined, the model either over-predicts or is essentially in
13 agreement with observations. It is not clear why the under-predictions at the California sites are so
14 large during summer. Under-estimation of local emissions may be part of the cause.
15 These model-predicted values can be compared to the baseline O3 concentrations estimated by
16 Chan and Vet (2010, 679710) of 37 ± 9 ppb for the continental eastern U.S., 51 ± 6 ppb for the
17 continental western U.S., 44 ± 10 ppb for the coastal western U.S. from March to May; and 32 ±
18 2 ppb for the continental eastern U.S., 25 ± 10 ppb for the continental western U.S. and 39 ± 12 ppb
19 for the coastal western U.S. from June to August.
20 Table 3-2 shows seasonal maximum concentrations measured at the same sites as in Table 3-1
21 for spring (April-May, 61 days) and summer (June-July-August, 92 days) in 2001 and maximum O3
22 concentrations calculated for the base case and the PRB case for the same time frame by GEOS-
23 Chem at l°x 1° resolution. During the spring, maximum concentrations measured at the sites chosen
24 ranged from 50 to 89 ppb at the CASTNET sites; GEOS-Chem predictions range from 52 to 72 ppb,
25 and from 40 to 57 ppb for PRB. During summer, maximum concentrations measured at the
26 CASTNET sites chosen ranged from 52 to 101 ppb; GEOS-Chem predictions range from 49 to
27 90 ppb, and from 34 to 51 ppb for PRB.
28 The time series of the model predictions for the daily max 8-h avg O3 concentrations, the
29 corresponding PRB concentrations and measurements at the CASTNET sites shown in Table 3-1 and
30 Table 3-2 are given in Chapter 3 Appendix, Figure 3A-1 through Figure 3A-15. In general, model
31 predicted concentrations and observations tend to be slightly better correlated with observations at
32 eastern sites (mean: 0.60, range 0.52-0.74) than at western sites (mean: 0.53, range 0.32-0.69). The
33 lowest model-observed correlation (0.32) is found at Big Bend NP, TX, where rather large over-
34 predictions are also found. Both results may be due in part to errors in Mexican emissions.
35 Correlations between base model and PRB O3 concentrations are very low and slightly negative at
36 most eastern sites (mean: 0.01, range -0.27 to 0.45) and become larger at western sites (mean: 0.56,
37 range (0.33-0.69). The Voyageurs NP site in Minnesota is the only 'eastern' site showing a positive
38 correlation (r = 0.45) between base and PRB model O3 concentrations. The low correlations at the
39 eastern sites chosen arise because base model O3 and PRB concentrations tend to be anti-correlated
March 2011 3-37 DRAFT - DO NOT CITE OR QUOTE
-------
1 at high O3 concentrations, but are positively correlated at the lower end of the concentration range.
2 At high elevation western sites, on the other hand, base model and PRB O3 concentrations tend to be
3 positively correlated throughout much of the concentration range.
Table 3-1. Seasonal means of the daily max 8-h avg ozone concentrations in spring and summer at
selected CASTNET and other National Park Service monitoring sites in the continental U.S.
and in the U.S. Virgin Islands, in 2001. GEOS-Chem and PRB are included for comparison
Spring (April-May)
Sites
Acadia NP, ME
Everglades NP, FL
Crockett, KY
Coffeeville, MS
Voyageurs NP, MN
Big Bend NP, TX
Rocky Mtn. NP, CO
Mesa Verde NP, CO
Yellowstone NP.WY
Centennial, WY
Canyonlands NP, UT
Glacier NP, MT
Mt. Rainier NP, WA
Lassen Volcanic NP, CA
Pinnacles NM, CA
Virgin Islands NP
CASTNET ± SD
47.7 ±9.9
41 .3 ±12.4
57.9 ±9.9
52.9 ±10.9
36.9 ±10.7
39.1 ±5.4
49.7 ±11. 9
56.0 ±4.9
56.1 ±7.4
59.9 ±5.7
54.3 ± 4.2
41 .3 ±4.6
39.7 ±6.9
51 .4 ±7.9
57.1 ± 10.6
21 .6 ±6.4
GEOS-Chem
(PRB)
48.2 ±7.8
(29.8 ± 7.4)
51 .2 ±8.9
(32.9 ± 6.5)
56.8 ±6.3
(27.1 ±6.6)
53.8 ±5.9
(27.5 ±5.1)
42.9 ±5.5
(26.3 ± 6.4)
49.9 ±6.0
(37.0 ± 7.0)
51 .9 ±8.2
(39.3 ± 7.5)
55.9 ±6.1
(40.6 ± 6.3)
51.0 ±7.1
(42.3 ± 5.6)
52.1 ±7.1
(40.7 ±6.1)
53.9 ±6.4
(41.0 ±5.5)
44.9 ± 5.0
(36.3 ± 5.2)
41.1 ±6.4
(32.0 ±7.1)
48.9 ±6.6
(39.6 ± 6.5)
47.6 ±7.7
(26.8 ±9.1)
31 .7 ±7.0
(27.0 ± 5.3)
Summer (June-August)
CASTNET ± SD
50.6 ±19.4
24.3 ±8.5
54.5 ± 10.8
50.8 ±12.8
40.2 ±9.3
43.3 ±9.1
47.7 ±10.1
54.4 ±6.4
50.6 ±5.4
59.5 ±5.5
56.0 ±6.4
35.4 ±7.1
32.8 ±12.7
54.7 ±10.0
55.3 ± 12.9
18.3 ±3.7
GEOS-Chem (PRB)
52.8 ±15.4 (16.8 ±5.6)
45.7 ±6.8
(36.4 ± 4.2)
57.2 ±7.1
(22.7 ± 5.6)
57.4 ±9.4
(22.6 ± 5.0)
35.3 ±9.6
(19.6 ±4.8)
50.9 ±5.6
(36.1 ±4.8)
57.7 ±6.4
(33.2 ± 5.6)
58.2 ±4.8
(33.6 ± 5.8)
46.5 ±5.8
(34.7 ± 5.3)
52.2 ±4.5
(34.2 ±5.1)
53.2 ±5.1
(33.8 ± 5.9)
36.3 ±5.3
(29.6 ± 7.4)
34.6 ± 10.0
(19.6 ±5.6)
43.6 ±7.5
(34.2 ± 7.5)
42.6 ±11. 9
(15.4 ±5.2)
28.2 ±5.3
(25.8 ± 5.5)
March 2011
3-38
DRAFT - DO NOT CITE OR QUOTE
-------
Table 3-2. Seasonal maximums of the daily max 8-h avg ozone concentrations in spring and summer
at selected CASTNET and other National Park Service monitoring sites in the continental
U.S. and in the U.S. Virgin Islands, in 2001. GEOS-Chem and PRB are included for
comparison
Spring (April-May)
Sites
Acadia NP, ME
Everglades NP, FL
Crockett, KY
Coffeeville, MS
Voyageurs NP, MN
Big Bend NP, TX
Rocky Mountain NP, CO
Mesa Verde NP, CO
Yellowstone NP.WY
Centennial, WY
Canyonlands NP, UT
Glacier NP, MT
Mt. Rainier NP, WA
Lassen Volcanic NP, CA
Pinnacles NM, CA
Virgin Islands NP
CASTNET
85.1
65.4
81.5
77.9
61.0
50.5
68.3
65.3
68.9
70.9
64.0
54.4
61.6
71.5
88.8
35
GEOS-Chem
(PRB)
68.1
(42.7)
71.7
(45.7)
71.4
(41.2)
66.6
(36.6)
61.1
(40.1)
62.5
(52.2)
68.4
(56.6)
70.9
(53.8)
65.7
(55.3)
66.0
(52.1)
65.3
(53.2)
52.8
(45.8)
52.3
(44.6)
60.3
(54.0)
68.4
(50.8)
47.1
(38.1)
Summer (June-August)
CASTNET
101.0
61.3
78.0
81.0
70.4
67.5
80.3
72.9
63.9
74.4
76.0
52.2
67.9
79.0
86.5
27
GEOS-Chem
(PRB)
90.3
(33.7)
72.5
(51.1)
72.4
(37.8)
85.1
(37.9)
58.7
(32.5)
64.5
(48.0)
72.8
(46.1)
68.5
(47.1)
59.5
(47.8)
61.7
(45.5)
64.2
(48.4)
48.9
(46.9)
59.6
(40.1)
61.6
(51.5)
73.4
(29.6)
45.9
(45.1)
March 2011
3-39
DRAFT - DO NOT CITE OR QUOTE
-------
3.5. Monitoring
3.5.1. Routine Monitoring Techniques
1 The Federal Reference Method (FRM) for O3 measurement is called the Chemiluminescence
2 Method (CLM) and is based on the detection of chemiluminescence resulting from the reaction of O3
3 with ethylene gas. The first ultraviolet (UV) absorption photometric analyzers were approved as
4 Federal Equivalent Methods (FEMs) in 1977 and gained rapid acceptance forNAAQS compliance
5 purposes due to ease of operation, relatively low cost, and reliability. The UV absorption method is
6 based on the principle that O3 molecules absorb UV radiation at a wavelength of 254 nm from a
7 mercury lamp. The concentration of O3 is computed from Beers law using the radiation absorbed
8 across a fixed path length, the absorption coefficient, and the measured pressure and temperature in
9 the detection cell. UV absorption photometry is the predominant method for assessing compliance
10 with the NAAQS for O3. Almost all of the state or local air monitoring stations (SLAMS) that
11 reported data to EPA Air Quality System (AQS) from 2005 to 2009 used UV absorption photometer
12 FEMs. No CLM monitors, approved as FRMs or FEMs, reported O3 data to AQS from 2005 to 2009
13 and only one monitor reported data using a long-path or open path Differential Optical Absorption
14 Spectrometer (DOAS) FEM during this period.
15 The rationale, history, and calibration of O3 measurements were summarized in the 1996 O3
16 AQCD (U.S. EPA, 1996, 017831) and the 2006 O3 AQCD (U.S. EPA, 2006, 088089) and focused on
17 the state of ambient O3 measurements at that time as well as evaluation of interferences and new
18 developments. This discussion will continue with the current state of O3 measurements,
19 interferences, and new developments for the period 2005 to 2009.
20 UV O3 monitors use mercury lamps as the source of UV radiation and employ an O3 scrubber
21 (typically manganese dioxide) to generate an O3-free air flow to serve as a reference channel for O3
22 measurements. There are known interferences with UV O3 monitors. The 2006 O3 AQCD (U.S. EPA,
23 2006, 088089) reported on the investigation of the effects of water vapor, aromatic compounds,
24 ambient particles, mercury vapor and alternative materials in the instrument's O3 scrubber. The
25 overall conclusions from the 2006 O3 AQCD (U.S. EPA, 2006, 088089) review of the scientific
26 literature are briefly summarized below.
27 Kleindienst et al, (1993, 043956) found water vapor to have no significant impact and
28 aromatic compounds to have a minor impact (as much as 3% higher than the FRM extrapolated to
29 ambient conditions) on UV absorption measurements. UV O3 monitor response evaluated by
30 chamber testing using cigarette smoke, reported an elimination of the O3 monitor response to the
31 smoke when a particle filter was used that filtered out particles less than 0.2 (im in diameter
32 (Arshinov et al., 2002, 080718V One study (Leston et al., 2005, 080717) in Mexico City compared a
33 UV O3 FEM to a CLM FRM. The UV FEM commonly reported consistently higher O3 than the
34 CLM FRM. The typical difference was 20 ppb with a range up to 50 ppb. Leston et al., (2005,
35 080717) also presented smog chamber data which demonstrated that heated metal and heated silver
March 2011 3-40 DRAFT - DO NOT CITE OR QUOTE
-------
1 wool scrubbers perform better in the presence of aromatic hydrocarbon irradiations than manganese
2 dioxide scrubbers when compared to the FRM. They also suggested the use of humidified calibration
3 gas and alternative scrubber materials to improve UV O3 measurements. Some O3 monitor
4 manufacturers now offer heated silver wool scrubbers as an alternative to manganese dioxide.
5 Another possible solution to the O3 scrubber problem may be the use of a gas phase scrubber such as
6 NO. A commercial version of this has recently been introduced by 2B Technologies as an option on
7 their model 202 FEM; however, it has not been field tested or approved for use as an FEM.
8 Subsequent study of UV monitors by Williams et al. (2006, 595152) concluded that well
9 maintained monitors showed no significant interferences when operated in locations with significant
10 concentrations of potentially interfering VOCs including Nashville, Houston, and the Gulf of Maine.
11 Monitors were tested in urban and suburban environments, as well as on board a ship in both
12 polluted and clean marine air. Comparisons of UV measurements to a non-FRM/FEM NO based
13 CLM demonstrated agreement to within 1%. At the Houston location, they did observe a brief period
14 on one day for about 30 minutes where the UV measurements exceeded the CLM by about 8 ppb
15 (max). This was attributed to probable instrument malfunction.
16 Wilson and Birks (2006, 595155) investigated water vapor interference in O3 measurements by
17 four different UV monitors. In extreme cases where a rapid step change in relative humidity between
18 0 and 90% was presented, large transitory responses (tens to hundreds of ppb) were found for all
19 monitors tested. Rapid changes in relative humidity such as this would not be expected during
20 typical ambient O3 measurements and could only be expected during measurement of vertical
21 profiles from balloon or aircraft. The magnitude of the interference and the direction (positive or
22 negative) was dependent on the manufacturer and model. Wilson and Birks (2006, 595155) also
23 hypothesized that water vapor interference is caused by physical interactions of water vapor on the
24 detection cell. The O3 scrubber was also thought to act as a reservoir for water vapor and either
25 added or removed water vapor from the air stream, subsequently affecting the detector signal and
26 producing either a positive or negative response. They demonstrated the use of a Nafion® permeation
27 membrane just before the O3 detection cell to remove water vapor and eliminate this interference.
28 Dunlea et al. (2006, 595108) evaluated multiple UV O3 monitors with two different O3
29 scrubber types (manganese dioxide and heated metal wool) in Mexico City. Large spikes in O3
30 concentrations were observed while measuring diesel exhaust where large increases in particle
31 number density were observed. The interference due to small particles passing through the Teflon®
32 filter and scattering/absorbing light in the detection cell were estimated to cause at most a 3%
33 increase in measurements in typical ambient air environments. This estimate pertains to
34 measurements in the immediate vicinity of fresh diesel emissions and most monitor siting guidelines
35 would not place the monitor close to such sources, so actual interferences are expected to be much
36 less than 3%. Dunlea et al. (2006, 595108) also observed no evidence for either a positive or
37 negative interference or dependence due to variations in aromatics during their field study.
38 Li et al. (2006, 633892) verified early reports of gas phase mercury interference with the UV
39 O3 measurement. They found that 300 ng/m3 of mercury produced an instrument response of about
March 2011 3-41 DRAFT - DO NOT CITE OR QUOTE
-------
1 35 ppb O3. Background concentrations of mercury are around 1-2 ng/m3 and expected to produce an
2 O3 response that would be <1 ppb.
3 Spicer et al. (2010, 670424) examined potential UV O3 monitor interferences by water vapor,
4 mercury, aromatic compounds, and reaction products from smog chamber simulations. Laboratory
5 tests showed little effect of changing humidity on conventional FEM UV O3 monitors with
6 manganese dioxide or heated metal wool scrubbers in the absence of other interferences. Mercury
7 vapor testing produced an O3 response by the UV monitors that was <1 ppb O3 per 1 ppt (about
8 8 ng/m3) mercury vapor. Interference by aromatic compounds at low (3% RH) and high (80% RH)
9 humidity showed some positive responses that varied by UV monitor and ranged from 0 to 2.2 ppb
10 apparent O3 response, per ppb of aromatic compound tested. The authors acknowledged that the
11 aromatic compounds most likely to interfere are rarely measured in the atmosphere and therefore,
12 make it difficult to assess the impact of these compounds during ambient air monitoring.
13 Comparison of UV and CLM responses to photochemical reaction products in smog chamber
14 simulations at 74 to 85% RH showed varied responses under low (0.125 ppmv/0.06 ppmv) to high
15 (0.50 ppmv/0.19 ppmv) hydrocarbon/NOx conditions. The conventional UV monitors were as much
16 as 2 ppb higher than the CLM under low hydrocarbon/NOx conditions and 6 ppb higher under the
17 high hydrocarbon/NOx conditions. Two FEM UV monitors were co-located at six sites in Houston
18 from May to October, 2007 with one UV monitor equipped with Nafion® permeation membrane.
19 The average difference between 8-h daily max O3 concentrations using the UV and the UV with
20 Nafion® permeation membrane ranged from -4.0 to 4.1 ppb.
3.5.2. Precision and Bias
21 In order to provide decision makers with an assessment of data quality, EPA's Quality
22 Assurance (QA) group derives estimates of both precision and bias for O3 and the other gaseous
23 criteria pollutants from the biweekly single point quality control (QC) checks using calibration gas,
24 performed at each site by the monitoring agency. The single point QC checks are typically performed
25 at concentrations around 90 ppb. Annual summary reports of precision and bias can be obtained for
26 each monitoring site at http://www.epa.gov/ttn/amtic/qareport.html (U.S. EPA, 2011, 677486). The
27 assessment of precision and bias are based on the percent-difference values, calculated from single
28 point QC checks. The percent difference is based on the difference between the pollutant
29 concentration indicated by monitoring equipment and the known (actual) concentration of the
30 sample used during the QC check. The monitor precision is estimated from the 90% upper
31 confidence limit of the coefficient of variation (CV) of relative percent difference (RPD) values. The
32 bias is estimated from the 95% upper confidence limit on the mean of the absolute values of percent
33 differences. The data quality goal for O3 precision and bias at the 90 and 95% upper confidence
34 limits is 7% (40 CFR Part 58, Appendix A (1986, 035997)). Table 3-3 presents a summary of the
35 number of monitors that meet the precision and bias goal of 7% for 2005 to 2009. Greater than 96%
36 of O3 monitors met the precision and bias goal between 2005 and 2009.
March 2011 3-42 DRAFT - DO NOT CITE OR QUOTE
-------
Table 3-3. Summary of monitors meeting 40 CFR Part 58, Appendix A Precision and Bias Goals
Year
2005
2006
2007
2008
2009
Number of Monitors
879
881
935
955
958
Monitors with Acceptable Precision (%
96.5
98.1
98.1
97.1
97.4
) Monitors with Acceptable Bias (%)
96.7
97.6
98.1
96.7
97.5
1 Another way to look at the precision and bias information from the monitoring network is to
2 present box plots of the monitors' individual precision and percent-difference data; Figure 3-12 and
3 Figure 3-13 included this information for O3 monitors operating from 2005 to 2009.
National Single Point QC Checks
Min 0.36
Max 14.85
Mean 2.70
N 1151
40 -
35 -
o 30 -
|25-
Coefficient
-i -» N
o in e
5 -
o -
J
f_
T
2005
0.00
38.37
2.55
1159
n
n
D
I
2006
0.00
18.32
2.50
1166
n
I
2007
0.00
21.39
2.60
1178
n
n
2008
0.00
31.82
2.43
1158
n
n
i
T
]
2009
Figure 3-12. Box plots of precision data by year (2005-2009) for all ozone monitors reporting
single-point QC check data to AQS.
March 2011
3-43
DRAFT - DO NOT CITE OR QUOTE
-------
National Single Point QC Checks
Min -28.3
Max 32.26
Mean 0.07
N 52724
100 -
U
C
i
Q 0 -
c
to
Of
Q-
-50 -
-100 -
B
]
2005
-22.1
41.11
-0.02
51814
n
L
I
i
2006
-60.6
47.67
-0.05
53262
i
T
f
D
2007
-49.4
74.44
-0.05
57315
D
n
n
i
2008
-77.8
73.33
0.19
67305
I
I
T
T
n
D
D
D
2009
Figure 3-13. Box plots of percent-difference data by year (2005-2009) for all ozone monitors
reporting single-point QC check data to AQS.
3.5.2.1. Precision from Co-located UVOzone Monitors in Missouri
1 The Missouri Department of Natural Resources (MODNR) maintains a network of co-located
2 UV O3 analyzers. The MODNR provided co-located data from four monitors: two co-located at the
3 same monitoring site in Kansas City (AQS ID 290370003) and two co-located at the same
4 monitoring site in St. Louis (AQS ID 291831002). Hourly observations for the co-located
5 measurements at these two sites during the O3 season (April through October) for 2006-2009 were
6 used to evaluate precision from co-located UV monitors. These data were then compared with the
7 precision obtained by the biweekly single point QC checks for all sites reporting single-point QC
8 check data to AQS between 2005 and 2009; the method normally used for assessing precision. Box
9 plots of the RPD between the primary and co-located hourly O3 measurements in Missouri are
10 shown in Figure 3-14 and box plots of the RPD between the actual and indicated QC check for all
11 U.S. sites are shown in Figure 3-15. As mentioned above, the average concentration of the single-
12 point QC check is 90 ppb, whereas the average ambient O3 concentration measured at the two sites
13 in Missouri was 34 ppb.
March 2011
3-44
DRAFT - DO NOT CITE OR QUOTE
-------
Collocated Missouri Ozone Statistics
Min -16.7
Max 10.00
Mean -0.07
N 10017
30 -
g 20-
c
Ql
£ 10-
b
c.
« n -
u U
Ol
Q.
|-io-
15
* -20^
-30 -
n
n
n
i
2006
-10.0
3.85
-0.48
10133
D
2007
-10.0
2.27
-0.78
9884
I
2008
-16.7
1.11
-0.77
10211
n
n
t
2009
Figure 3-14. Box plots of RPD data by year for the two co-located sites in Missouri from 2006-
2009.
National Single Point QC Checks
Min -6.19
Max 9.62
Mean 0.03
N 52724
30 -
at 20 -
u
c
IB
1 10-
Q
o) n -
u U
Ol
a.
? -10-
'ts
a)
^ -20 -
-30 -
H
I
2005
-4.98
12.94
0.01
51814
n
1
"T"
1
2006
-11.6
15.65
0.00
53262
1
1
T
•
a
i
2007
-9.91
29.65
0.00
57315
n
n
a
i
2008
-14.0
28.95
0.06
67305
D
n
I
T
y
B
n
2009
Figure 3-15. Box plots of RPD data by year for all U.S. ozone sites reporting single-point QC
check data to AQS from 2005-2009.
March 2011
3-45
DRAFT - DO NOT CITE OR QUOTE
-------
3.5.3. Performance Specifications
1 The performance specifications for evaluating and approving new FEMs in accordance with
2 40 CFR Part 53 (1976, 041090) are provided in Table 3-4. These specifications were developed and
3 originally published in the Federal Register in 1975 (1975, 043954). Modern, commercially-
4 available instruments can now perform much better than the requirements specified below. For
5 example, the lower detectable limit (LDL) performance specification is 10 ppb and the typical
6 vendor-stated performance for the LDL is less than 0.60 ppb. The amount of allowable interference
7 equivalent for total interference substances is 60 ppb, and the current NAAQS for O3 is 75 ppb, with
8 an averaging time of 8 hours. Improvements in new measurement technology have occurred since
9 these performance specifications were originally developed. These specifications should be revised
10 to more accurately reflect the necessary performance requirements for O3 monitors used to support
11 the current NAAQ S.
Table 34. Performance specifications for ozone
Parameter
Range
Noise
LDL - defined as two times the noise
based in 40 CFR Part 53
Specification
0-0.5ppm (500 ppb)
0.005 ppm (5 ppb)
0.01 ppm (10 ppb)
Interference equivalent
Each interfering substance
Total interfering substances
+ 0.02 ppm (20 ppb)
0.06 ppm (60 ppb)
Zero drift
12h
24 h
+ 0.02 ppm (20 ppb)
+ 0.02 ppm (20 ppb)
Span Drift, 24 h
20% of upper range limit
80% of upper range limit
Lag time
Rise time
Fall time
± 20.0%
± 5.0%
20min
15min
15min
Precision
20% of upper range limit
80% of upper range limit
0.01 ppm (10 ppb)
0.01 ppm (10 ppb)
3.5.4. Monitor Calibration
12 The calibration of O3 monitors was summarized in detail in the 1996 O3 AQCD (U.S. EPA,
13 1996, 017831). The calibration of O3 monitors is done using an O3 generator and UV photometers.
14 UV photometry is the prescribed procedure for the calibration of reference methods to measure O3 in
15 the atmosphere. Because O3 is unstable and cannot be stored, the O3 calibration procedure
16 specifically allows the use of transfer standards for calibrating ambient O3 monitors. A transfer
17 standard is calibrated against a standard of high authority and traceability and then moved to another
18 location for calibration of O3 monitors. The EPA and the National Institute of Standards and
March 2011
3-46
DRAFT - DO NOT CITE OR QUOTE
-------
1 Technology (NIST) have established a network of standard reference photometers (SRPs) that are
2 used to verify transfer standards. The International Bureau of Weights and Measures (BIPM)
3 maintain one NIST SRP (SRP27) as the World's O3 reference standard. NIST maintains two SRPs
4 (SRPO and SRP2) that are used for comparability to ten other SRPs maintained by the EPA's
5 Regional QA staff.
6 SRPs have been compared to other reference standards. Tanimoto et al. (2006, 595131)
7 compared NIST SRPS5, owned by the National Institute for Environmental Studies in Japan, to gas
8 phase titration (GPT). The SRP was found to be 2% lower than GPT GPT is no longer used as a
9 primary or transfer standard in the U.S. Viallon et al. (2006, 595145) compared SRP27 built at BIPM
10 to four other NIST SRPs maintained by BIPM (SRP28, SRPS 1, SRP32, and SRP33). A minimum
11 bias of+0.5% was found for all SRP measurement results, due to use of the direct cell length
12 measurement for the optical path length; this bias was accounted for by applying the appropriate
13 correction factor. Study of the bias-corrected SRPs showed systematic biases and measurement
14 uncertainties for the BIPM SRPs. A bias of -0.4% in the instrument O3 mole fraction measurement
15 was identified and attributed to non-uniformity of the gas temperature in the instrument gas cells,
16 which was compensated by a bias of+0.5% due to an under-evaluation of the UV light path length in
17 the gas cells. The relative uncertainty of the O3 absorption cross section was 2.1% at 253.65 nm and
18 this was proposed as an internationally accepted consensus value until sufficient experimental data is
19 available to assign a new value.
20 In November, 2010, the EPA revised the Technical Assistance Document for Transfer
21 Standards for Calibration of Air Monitoring Analyzers for Ozone (2010, 677553) that was first
22 finalized in 1979 (U.S. EPA, 1979, 039211). The revision removed methods no longer in use and
23 updated definitions and procedures where appropriate. In the revised document, the discussion of
24 transfer standards for O3 applies to the family of standards that are used beyond SRPs or Level 1
25 standards. To reduce confusion, EPA reduced the number of common terms that were used in the
26 past such as: primary standard, local primary standard, transfer standard, and working standard.
27 Beyond the SRPs, all other standards are considered transfer standards. The revised Ozone Technical
28 Assistance Document is available at
29 http://www.epa.gov/ttn/amtic/files/ambient/qaqc/OzoneTransferStandardGuidance.pdf (U.S. EPA,
30 2010.677553).
3.5.5. Other Monitoring Techniques
3.5.5.1. Portable UV Ozone Monitors
31 Small, lightweight, and portable UV O3 monitors with low power consumption are
32 commercially available. These monitors are based on the same principle of UV absorption by O3 at
33 254 nm. Monitors of this type are typically used for vertical profiling using balloons, kites, or light
34 aircraft where space and weight are limited. They have also been used for monitoring at remote
March 2011 3-47 DRAFT - DO NOT CITE OR QUOTE
-------
1 locations such as National Parks. Burley and Ray (2007, 149069) compared portable O3 monitor
2 measurements to those from a conventional UV monitor in Yosemite National Park. Calibrations of
3 the portable O3 monitors against a transfer standard resulted in an overall precision of ± 4 ppb and
4 accuracy of ± 6%. Field measurement comparisons between the portable and conventional monitor
5 at Turtleback Dome showed the portable monitor to be 3.4 ppb lower on average, with daytime
6 deviation typically on the order of 0-3 ppb. Agreement between the portable and conventional
7 monitor during daylight hours (9:00 a.m. to 5:00 p.m. PST) resulted in an R2 of 0.95, slope of 0.95,
8 and intercept of 0.36 ppb. Significant deviations were observed in the predawn hours where the
9 portable monitor was consistently low. These deviations were attributed to the difference in sampling
10 inlet location. The portable monitor was located at 1.3 m above ground and the conventional monitor
11 was located at 10 m above ground. Agreement between the portable and conventional monitors for
12 all hours sampled resulted in an R2 of 0.88, slope of 1.06, and intercept of -6.8 ppb. Greenberg et al.
13 (2009, 595140) also compared a portable UV O3 monitor to a conventional UV monitor in
14 Mexico City and obtained good agreement for a 14 day period with an R2 of 0.97, slope of 0.97, and
15 intercept of 6 ppb. One portable O3 monitor was recently approved as an FEM (EQOA-0410-190) on
16 April 27, 2010 (75 FR 22126) (2010, 687659).
3.5.5.2. Teledyne Advanced Pollution Instrumentation Model 265E CLM
17 The Teledyne Advanced Pollution Instrument (TAPI) NO-based chemiluminescence
18 instrument is currently undergoing FEM testing. It may also be designated as a second or
19 replacement FRM since the ethene based FRMs are no longer manufactured. Although the TAPI is a
20 relatively new instrument, other NO-based CLM instruments have been custom built for various
21 field studies since the early 1970s. A commercial version that measured both O3 and NOX was
22 offered by Aerochem Research Laboratories (Princeton, NJ) in the early 1970s but failed to gain
23 commercial acceptance. Initial testing with SO2, NO2, C12, C2H2, C2H4 and C3H6 (Stedman et al.,
24 1972, 033730) failed to identify any interferences. In the intervening years, custom built versions
25 have not been found to have any interferences; however, they do experience a slight decrease in
26 response with increasing relative humidity (due to quenching of the excited species by the water
27 molecules). The TAPI instrument solves this problem with the use of a Nafion® membrane dryer. A
28 custom built instrument similar to the 265E was used by Williams et al. (2006, 595152) in Houston,
29 TX; Nashville, TN; and aboard ship along the New England coast. It was found to be in good
30 agreement with a standard UV based FEM and with a custom built Differential Optical Absorption
31 Spectrometer (DOAS).
3.5.5.3. Passive Air Sampling Devices and Sensors
32 A passive O3 sampling device depends on the diffusion of O3 in air to a collecting or indicating
33 medium. In general, passive samplers are not adequate for compliance monitoring because of the
34 limitations in averaging time (typically one week or more), particularly for O3. However, these
35 devices are valuable for personal human exposure estimates and for obtaining long-term data in rural
March 2011 3-48 DRAFT - DO NOT CITE OR QUOTE
-------
1 areas where conventional UV monitors are not practical or feasible to deploy. The 1996 O3 AQCD
2 (U.S. EPA, 1996, 017831) provided a detailed discussion of passive samplers, along with the
3 limitations and uncertainties of the samplers evaluated and published in the literature from 1989 to
4 1995. The 2006 O3 AQCD (U.S. EPA, 2006, 088089) provided a brief update on available passive
5 samplers developed for use in direct measurements of personal exposure published through 2004.
6 The 2006 O3 AQCD (U.S. EPA, 2006, 088089) also noted the sensitivity of these samplers to wind
7 velocity, badge placement, and interference by other co-pollutants that may result in measurement
8 error.
9 Subsequent evaluations of passive diffusion samplers in Europe showed good correlation
10 when compared to conventional UV O3 monitors, but a tendency for the diffusion samplers to
11 overestimate the O3 concentration (Buzica et al., 2008, 595173; Gottardini et al, 2010, 562911;
12 Vardoulakis et al., 2009, 595141). The bias of O3 diffusion tubes were also found to vary with
13 concentration, season, and exposure duration (Vardoulakis et al., 2009, 595141). Development of
14 simple, inexpensive, passive O3 measurement devices that rely on O3 detection papers and a variety
15 of sensors with increased time resolution (sampling for hours instead of weeks) and improved
16 sensitivity have been reported (Ebeling et al., 2009, 595115; Maruo, 2007, 595102; Maruo et al.,
17 2010, 595104; Miwa et al., 2009, 595110; O-Keeffe et al., 2007, 595114; Ohira et al., 2009, 595113;
18 Utembe et al., 2006, 595137). Limitations for some of these sensors and detection papers include air
19 flow dependence and relative humidity interference.
3.5.5.4. Differential Optical Abs orption Spectrometry
20 Optical remote sensing methods can provide direct, sensitive, and specific measurements of O3
21 over a broad area or open path in contrast with conventional single-point UV monitors. The 1996 O3
22 AQCD (U.S. EPA, 1996, 017831) provided a brief discussion of DOAS for O3 measurements and
23 cited references to document the sensitivity (1.5 ppb for a 1-minute averaging time), correlation (r =
24 0.89), and agreement (on the order of 10%) with UV O3 monitors (Stevens et al., 1993, 046404). The
25 2006 O3 AQCD (U.S. EPA, 2006, 088089) provided an update on DOAS where a positive
26 interference due to an unidentified absorber was noted (Reisinger, 2000, 053936).
27 More recent study of the accuracy of UV absorbance monitors by Williams et al. (2006,
28 595152) compared UV and DOAS measurements at two urban locations. In order to compare the
29 open path measurements and UV, the data sets were averaged to 30-minute periods and only data
30 when the boundary layer was expected to be well mixed (between 10:00 a.m. and 6:00 p.m. CST)
31 were evaluated. The comparisons showed variations of no more than ± 7% (based on the slope of the
32 LLS regression over a concentration range from about 20 to 200 ppb) and good correlation (R2 =
33 0.96 and 0.98). Lee et al. (2008, 595175) evaluated DOAS and UV O3 measurements in Korea and
34 found the average DOAS concentration to be 8.6% lower than the UV point measurements with a
35 good correlation (R2 = 0.94).
36 DOAS has also been used for the measurement of HNO2 (or HONO). DOAS was compared to
37 chemical point-measurement methods for HONO. Acker et al. (2006, 595095) obtained good results
March 2011 3-49 DRAFT - DO NOT CITE OR QUOTE
-------
1 when comparing wet chemical and DOAS during well mixed atmospheric conditions (wet chemical
2 = 0.009 + 0.92 x DOAS; r = 0.7). Kleffmann and Wiesen (2008, 488342) noted that interferences
3 with the HONO wet chemical methods can affect results from inter-comparison studies if not
4 addressed. In an earlier study, Kleffman et al. (2006, 130481) demonstrated that when the
5 interferences were addressed, excellent agreement with DOAS can be obtained. Stutz et al. (2009,
6 595124) found good agreement (15% or better) between DOAS and a wet chemical method (Mist
7 Chamber/IC) in Houston, TX except generally during mid-day when the chemical method showed a
8 positive bias that may have been related to concentrations of O3. DOAS remains attractive due to its
9 sensitivity, speed of response, and ability to simultaneously measure multiple pollutants; however,
10 further inter-comparisons and interference testing are recommended.
3.5.5.5. Satellite Remote Sensing
11 Satellite observations for O3 are growing as a resource for many purposes, including model
12 evaluation, assessing emissions reductions, pollutant transport, and air quality management. Satellite
13 remote sensing instruments do not directly measure the composition of the atmosphere. Satellite
14 retrievals are conducted using the solar backscatter or thermal infrared emission spectra and a variety
15 of algorithms. Most satellite measurement systems have been developed for stratospheric
16 measurement of the total O3 column. Mathematical techniques have been developed and must be
17 applied to derive information from these systems about tropospheric O3 (Tarasick and Slater, 2008,
18 596431; Ziemke JR: Chandra et al., 2006, 595159). Direct retrieval of global tropospheric O3
19 distributions from solar backscattered UV spectra have been reported from the Ozone Monitoring
20 Instrument (OMI) and Global Ozone Monitoring Experiment (GOME)(Liu et al., 2006, 093013).
21 Another satellite measurement system, Tropospheric Emission Spectrometer (TES), produces global -
22 scale vertical concentration profiles of tropospheric O3 from measurements of thermal infrared
23 emissions. TES has been designed specifically to focus on mapping the global distribution of
24 tropospheric O3 extending from the surface to about 10-15 km altitude (Beer, 2006, 633893). In
25 order to improve the understanding of the quality and reliability of the data, satellite-based
26 observations of total column and tropospheric O3 have been validated in several studies using a
27 variety of techniques, such as aircraft observations, ozonesondes, CTMs, and ground-based
28 spectroradiometers (Anton et al., 2009, 595098: Richards et al., 2008, 617597: Worden et al., 2007,
29 623018: Zhang et al., 2010, 633894). Satellite observations have also been combined (e.g., OMI and
30 TES) to improve estimates of tropospheric O3 (Worden et al., 2007, 623020).
3.5.6. Ambient Ozone Network Design
3.5.6.1. Monitor Siting Requirements
31 To monitor compliance with the NAAQS, state and local monitoring agencies operate O3
32 monitoring sites at various locations depending on the area size (population and geographic
March 2011 3-50 DRAFT - DO NOT CITE OR QUOTE
-------
1 characteristics1) and typical peak concentrations (expressed in percentages below, or near the O3
2 NAAQS). SLAMS make up the ambient air quality monitoring sites that are primarily needed for
3 NAAQS comparisons, but may also serve some other basic monitoring objectives that include:
4 providing air pollution data to the general public in a timely manner; support for compliance with the
5 NAAQS and emissions strategy development; and support for air pollution research. SLAMS
6 include National Core (NCore), Photochemical Assessment Monitoring Stations (PAMS), and all
7 other State or locally-operated stations except for the monitors designated as SPMs.
8 The SLAMS minimum monitoring requirements to meet the O3 design criteria are specified in
9 40 CFR Part 58, Appendix D. Although NCore and PAMS are a subset of SLAMS, the monitoring
10 requirements for those networks are separate and discussed below. The minimum number of O3
11 monitors required in a Metropolitan Statistical Area (MSA) ranges from zero for areas with a
12 population of at least 50,000 and under 350,000 with no recent history of an O3 design value2 greater
13 than 85 percent of the NAAQS, to four for areas with a population greater than 10 million and an O3
14 design value greater than 85 percent of the NAAQS. Within an O3 network, at least one site for each
15 MSA, or Combined Statistical Area (CSA) if multiple MS As are involved, must be designed to
16 record the maximum concentration for that particular metropolitan area. More than one maximum
17 concentration site may be necessary in some areas. The spatial scales for O3 sites are neighborhood,
18 urban and regional.
19 • Neighborhood scale: represents concentrations within some extended area of the city that
20 has relatively uniform land use with dimensions in the 0.5-4.0 km range. The
21 neighborhood and urban scales listed below have the potential to overlap in applications
22 that concern secondary or homogeneously distributed primary air pollutants.
23 • Urban scale: represents concentrations within an area of city-like dimensions, on the
24 order of 4-50 km. Within a city, the geographic placement of sources may result in there
25 being no single site that can be said to represent air quality on an urban scale.
26 • Regional scale: usually defines a rural area of reasonably homogeneous geography
27 without large sources, and extends from tens to hundreds of kilometers.
28 Since O3 concentrations decrease significantly in the colder parts of the year in many areas, O3
29 is required to be monitored at SLAMS monitoring sites only during the "ozone season." Table D-3 of
30 40 CFR Part 58, Appendix D lists the beginning and ending month of the "ozone season" for each
31 U.S. state or territory. Most operate O3 monitors only during the O3 season. Those that operate some
1 Geographic characteristics such as complexity of terrain, topography, land use, etc.
2 A design value is a statistic that describes the air quality status of a given area relative to the level of the NAAQS. Design values are
typically used to classify nonattainment areas, assess progress towards meeting the NAAQS, and develop control strategies. See
http://epa.gov/airtrends/values.html (U.S. EPA, 2010, 677582) for guidance on how these values are defined.
March 2011 3-51 DRAFT - DO NOT CITE OR QUOTE
-------
1 or all of their O3 monitors on a year-round basis include Arizona, California, Hawaii, Louisiana,
2 Nevada, New Mexico, Puerto Rico, Texas, American Samoa, Guam and the Virgin Islands.
3 The total number of SLAMS O3 sites needed to support the basic monitoring objectives
4 includes more sites than the minimum numbers required in 40 CFR Part 58, Appendix D. In 2009,
5 there were 1208 SLAMS O3 monitors reporting values to the EPA AQS database (Figure 3-16).
6 Monitoring site information for EPA's air quality monitoring networks is available in spreadsheet
7 format (CSV) and keyhole markup language format (KML or KMZ) that is compatible with Google
8 Earth™ and other software applications on the AirExplorer website (U.S. EPA, 2011, 677547).
9 States may operate O3 monitors in non-urban or rural areas to meet other objectives (e.g., support for
10 research studies of atmospheric chemistry or ecosystem impacts). These monitors are often identified
11 as special purpose monitors (SPMs) and can be operated up to 24 months without being considered
12 in NAAQS compliance determinations. The current monitor and probe siting requirements have an
13 urban focus and do not address the siting for SPMs or monitors in non-urban, rural areas to support
14 ecosystem impacts and the secondary standards.
15 NCore is a new multi-pollutant monitoring network implemented to meet multiple monitoring
16 objectives. Those objectives include: timely reporting of data to the public through AirNow
17 (U.S. EPA, 2011, 677548); support for the development of emission reduction strategies; tracking
18 long-term trends of criteria pollutants and precursors; support to ongoing reviews of the NAAQS and
19 NAAQS compliance; model evaluation; support for scientific research studies; and support for
20 ecosystem assessments. Each state is required to operate at least one NCore site and the entire
21 network consists of about 60 urban and 20 rural sites as of January 1, 2011. NCore has leveraged
22 sites in existing networks; for example, some CASTNET and IMPROVE sites serve as rural NCore
23 sites. In addition to O3, other components including CO, NO, NOY, NH3, and HNO3 are also
24 measured at NCore sites. The spatial scale for urban NCore stations is urban or neighborhood;
25 however, a middle-scale1 site may be acceptable in cases where the site can represent many such
26 locations throughout a metropolitan area. Rural NCore sites are located at a regional or larger scale,
27 away from any large local emission sources so that they represent ambient concentrations over an
28 extensive area. Ozone monitors at NCore sites are operated year round.
29 PAMS provides more comprehensive data on O3 in areas classified as serious, severe, or
30 extreme nonattainment for O3. In addition to O3, PAMS provides data for NOX, VOCs, and
31 meteorology. The PAMS network design criteria are based on locations relative to O3 precursor
32 source areas and predominant wind directions associated with high O3 concentrations. The overall
33 network design is location specific and geared toward enabling characterization of precursor
34 emission sources in the area, O3 transport, and photochemical processes related to O3 nonattainment.
35 Minimum monitoring for O3 and its precursors is required annually during the months of June, July,
36 and August when peak O3 concentrations are expected. In 2006, the EPA reduced the minimum
1 Middle scale defines an area up to several city blocks in size with dimensions ranging from about 100 to 500 m.
March 2011 3-52 DRAFT - DO NOT CITE OR QUOTE
-------
1 PAMS monitoring requirements. There were a total of 119 PAMS reporting values to the AQS data
2 base in 2009.
o Urban NCore
o PAMS
• Other Sites Reporting Data to AQS
Figure 3-16. U.S. ozone sites reporting data to AQS as of 2009.
3 The Clean Air Status and Trends Network (CASTNET) is a regional monitoring network
4 established to assess trends in acidic deposition due to emission reduction regulations. CASTNET
5 also provides concentration measurements of air pollutants involved in acidic deposition, such as
6 sulfate and nitrate, in addition to the measurement of O3. CASTNET O3 monitors operate year round
7 and are primarily located in rural areas. At the beginning of 2010, there were 80 CASTNET sites
8 located in, or near, rural areas. As part of CASTNET, the National Park Service (NPS) operates 23
9 sites located in national parks and other Class-I areas. Ozone measurements at the CASTNET sites
10 were not collected with the QA requirements for SLAMS outlined in 40 CFR Part 58, Appendix A,
11 and therefore, the O3 cannot be used for NAAQS compliance purposes. The network is currently
12 implementing the SLAMS QA requirements and procedures. Ozone data collected at the 23 NPS
13 sites is compliant with the SLAMS QA requirements in 40 CFR Part 58.
14 The NPS also operates a Portable Ozone Monitoring Systems (POMS) network. The POMS
15 couples the small, low-power O3 monitor with a data logger, meteorological measurements, and solar
16 power in a self contained system for monitoring in remote locations. Typical uses for the POMS data
17 include research projects, survey monitoring, and assessments of spatial O3 distribution. The portable
March 2011
3-53
DRAFT - DO NOT CITE OR QUOTE
-------
1 O3 monitor in use by the NFS was recently designated as an equivalent method for O3 (75 FR 22126)
2 (2010, 687659). Twenty NFS POMS reported O3 data to AQS in 2010.
3 A map of the current and proposed rural NCore sites, along with the CASTNET, and the NFS
4 POMS sites are shown in Figure 3-17.
~w4*>>
• Rural NCore
o NPS POMS
• CASTNET
Of
Figure 3-17. U.S. Rural NCore, CASTNET and NPS POMS current and proposed sites as of
October, 2010.
3.5.6.2. Probe/Inlet Siting Requirements
5 Probe and monitoring path siting criteria for ambient air quality monitoring are contained in
6 40 CFR Part 58, Appendix E. For O3, the probe must be located between 2 and 15m above ground
7 level and be at least 1 m away (both in the horizontal and vertical directions) from any supporting
8 structure, walls, etc. If it is located on the side of a building, it must be located on the windward side,
9 relative to prevailing wind direction during the season of highest potential O3 concentration. Ozone
10 monitors are placed to determine air quality in larger areas (neighborhood, urban, or regional scales)
11 and therefore, placement of the monitor probe should not be near local, minor sources of NO,
12 O3-scavenging hydrocarbons, or O3 precursors. The probe or inlet must have unrestricted air flow in
13 an arc of at least 180 degrees and be located away from any building or obstacle at a distance of at
14 least twice the height of the obstacle. The arc of unrestricted air flow must include the predominant
15 wind direction for the season of greatest O3 concentrations. Some exceptions can be made for
March 2011
3-54
DRAFT - DO NOT CITE OR QUOTE
-------
1 measurements taken in street canyons or sites where obstruction by buildings or other structures is
2 unavoidable. The scavenging effect of trees on O3 is greater than other pollutants and the probe/inlet
3 must be located at least 10m from the tree drip line to minimize interference with normal air flow.
4 When siting O3 monitors near roadways, it is important to minimize the destructive interferences
5 from sources of NO, since NO reacts readily with O3. For siting neighborhood and urban scale O3
6 monitors, guidance on the minimum distance from the edge of the nearest traffic lane is based on
7 roadway average daily traffic count (40 CFR Part 58, Appendix E, Table E-l). The minimum
8 distance from roadways is 10 m (average daily traffic coung 1 ,000) and increases to a maximum
9 distance of 250 m (average daily traffic count 110,000).
3.6. Ambient Concentrations
10 This section investigates spatiotemporal variability in ambient O3 concentrations and
11 associations between O3 and co-pollutants. To set the stage for the rest of the section, common O3
12 measurement units, metrics, and averaging times are described and compared in Section 3.6.1.
13 Spatial variability is covered in Section 3.6.2 and is divided into urban-focused variability and rural -
14 focused variability. Urban-focused variability is organized by scale, extending from national-scale
15 down to neighborhood-scale and the near-road environment. Rural-focused variability is organized
16 by region and includes observations of ground-level vertical O3 gradients where available. Temporal
17 variability is covered in Section 3.6.3 and is organized by time, extending from multiyear trends
18 down to hourly (diel) variability. In many instances, spatial and temporal variability are inseparable
19 (e.g., seasonal dependence to spatial variability), resulting in some overlap between Sections 3.6.2
20 and 3.6.3. Finally, Section 3.6.4 covers associations between O3 and co-pollutants including CO,
21 SO2,NO2, PM25andPM10.
22 As noted in the 2006 O3 AQCD (U.S. EPA, 2006, 088089). O3 is the only photochemical
23 oxidant other than nitrogen dioxide (NO2) that is routinely monitored and for which a comprehensive
24 database exists. Data for other photochemical oxidants (e.g., PAN, H2O2, etc.) typically have been
25 obtained only as part of special field studies. Consequently, no data on nationwide patterns of
26 occurrence are available for these other oxidants; nor are extensive data available on the
27 relationships of concentrations and patterns of these oxidants to those of O3. As a result, this section
28 focuses solely on O3, the NAAQS indicator for photochemical oxidants. The majority of ambient O3
29 data reported in this section were obtained from AQS, EPA's repository for detailed, hourly data that
30 has been subject to EPA quality control and assurance procedures (see Section 3.5 for a description
31 of the AQS network).
3.6.1. Measurement Units, Metrics, and Averaging Times
32 Several approaches are commonly used for reporting O3 data. In atmospheric sciences and
33 epidemiology, O3 is frequently reported as a concentration, expressed as a volume-to-volume mixing
34 ratio, commonly measured in ppm or ppb. In human exposure, O3 is frequently reported as a
March 2011 3-55 DRAFT - DO NOT CITE OR QUOTE
-------
1 cumulative exposure, expressed as a mixing ratio times time (e.g., ppm-h). In ecology, cumulative
2 exposure indicators are frequently used that extend over longer time periods, such as growing season
3 or year. This section focuses on ambient concentrations derived primarily from hourly average O3
4 measurements and concentrations are reported in ppb wherever possible. Further details on human
5 and ecological exposure metrics can be found in Chapter 4 and Chapter 9, respectively.
6 As discussed in Section 3.5, most continuous O3 monitors report hourly average
7 concentrations. This data can be used as reported (1-h avg), or further summarized in one of several
8 ways to focus on important aspects of the data while simultaneously reducing the volume of
9 information. Three common daily reporting metrics include: (1) the average of the hourly
10 observations over a 24-h period (24-h avg); (2) the maximum hourly observation occurring in a 24-h
11 period (1-h daily max); and (3) the maximum 8-h running average of the hourly observations
12 occurring in a 24-h period (8-h daily max)1. Throughout this ISA and the literature, O3
13 concentrations are reported using different averaging times as appropriate, making it important to
14 recognize the differences between these metrics.
15 Nation-wide, year-round 1-h avg O3 data reported to AQS from 2007-2009 was used to
16 compare these different daily metrics. Correlations between the 24-h avg, 1-h daily max and 8-h
17 daily max metrics were generated on a site-by-site basis. Figure 3-18 contains box plots of the
18 distribution in correlations from all sites. The top comparison in Figure 3-18 is between 8-h daily
19 max and 1-h daily max O3. Not surprisingly, these two metrics are very highly correlated (median r =
20 0.97, IQR = 0.96-0.98). There are a couple outlying sites, with correlations between these two
21 metrics as low as 0.63, but 95% of sites have correlations above 0.93. The middle comparison in
22 Figure 3-18 is between 8-h daily max and 24-h avg O3. Forthese metrics, the distribution in
23 correlations is shifted down and broadened out (median r = 0.89, IQR = 0.86-0.92). Finally, the
24 bottom comparison in Figure 3-18 is between 1-h daily max and 24-h avg O3. Again, for these
25 metrics the distribution in correlations is shifted down and broadened out relative to the other two
26 comparisons (median r = 0.83, IQR = 0.78-0.88). The correlation between the two daily maximum
27 metrics (1-h daily max and 8-h daily max) are quite high for most sites, but correlations between the
28 daily maximum metrics and the daily average metric (24-h avg) are lower. This illustrates the
29 influence of the overnight period on the 24-h avg O3 concentration. In contrast, the 1-h daily max
30 and 8-h daily max are more indicative of the daytime, high O3 periods. The correlation between these
31 metrics, however, can be very site-specific, as is evident from the broad range in correlations in
32 Figure 3-18 for all three comparisons.
1 For O3 regulatory monitoring purposes, the 8-h daily max is calculated by first generating all 8-h running averages and storing these
averages hourly by the first hour in the 8-h period. The 8-h daily max is then set equal to the maximum of the 24 individual 8-h avg
occurring in a given day.
March 2011 3-56 DRAFT - DO NOT CITE OR QUOTE
-------
9
10
11
12
13
8-h daily max
vs.
1-h daily max
8-h daily max
vs.
24-h avg
1-h daily max
vs,
24-h avg
*— 1*-
i Wr
0,0 0,1 0,2 0.3 0,4 0.5 0.6 0.7 0.8 0,9 1.0
Correlation
Figure 3-18. Distribution in nation-wide year-round site-level correlations between daily ozone
metrics including 24-h avg, 1-h daily max and 8-h daily max using AQS data, 2007-
2009.Shown are the median (red line), mean (green star), inner-quartile range (box),
5th and 95th percentiles (whiskers), and extremes (black dots).
1 The median 1-h daily max, 8-h daily max, and 24-h avg O3 concentrations across all sites
2 included in the 3-year nation-wide data set were 44, 40, and 29 ppb, respectively. Representing the
3 upper end of the distribution, the 99th percentiles of these same metrics across all sites were 94, 80,
4 and 60 ppb, respectively. While the ratio of these metrics will vary by location, typically the 1-h
5 daily max will be the highest value representing peak concentrations and the 24-h avg will be
6 considerably lower representing daily average concentrations incorporating the overnight period.
7 The 8-h daily max typically represents the higher mid-day concentrations and will generally lie
8 somewhere between the other two metrics1.
3.6.2. Spatial Variability
3.6.2.1. Urban-Focused Variability
National-Scale Variability
AQS contains a large depository of national O3 data collected to meet the monitoring
objectives described in Section 3.5.6.1. In many areas, O3 concentrations decrease significantly
during months with lower temperatures and decreased sunlight. As a result, year-round O3
monitoring is only required in certain areas. Table D-3 of 40 CFR Part 58, Appendix D lists the
beginning and ending month of the O3 season by geographic area and Figure 3-19 illustrates these
1 The 8-h daily max is not strictly limited to lie between the 1-h daily max and the 24-h avg since the 8-h averaging period used to
calculate the 8-h daily max can extend into the morning hours of the subsequent day. However, the 8-h daily max typically incorporates
the middle of the day when O3 concentrations are at their highest, resulting in an 8-h daily max somewhere between the 1-h daily max
and the 24-h avg.
March 2011
3-57
DRAFT - DO NOT CITE OR QUOTE
-------
1 time periods on a monitor-by-monitor basis. Monitoring is optional outside the "ozone season" and
2 many states elect to operate their monitors year-round or for time periods outside what is strictly
3 mandated.
Required Ozone Monitoring Time Periods
Time Period
• Apr-Sep • Mar-Nov
• Apr-Oct • May-Sep
Apr-Nov • May-Oct
Mar-Sep Jun-Sep
* Mar-Oca • Year round
Puerto Rico
Source: U.S. EPA (2008, 1911901
Figure 3-19. Required ozone monitoring time periods (ozone season) identified by monitoring
site.
4 Hourly FRM and FEM O3 data reported to AQS for the period 2007 - 2009 were used to
5 investigate national-scale spatial variability in O3 concentrations. Given the variability in O3
6 monitoring time periods available in AQS as a result of the regionally-varying "ozone seasons", the
7 analyses in this section were based on two distinct data sets:
8 • year-round data set: data only from monitors reporting year-round;
9 • warm-season data set: data from all monitors reporting May through September.
10 The warm-season data set was used to capture the majority of O3 season data while providing a
11 consistent time-frame for comparison across states. All available monitoring data including data
12 from year-round monitors was included in the warm-season data set after removing observations
13 outside the 5-month window. Data were retrieved from AQS for these two data sets regardless of
March 2011
3-58
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
flags or regional concurrence1. A summary of the two O3 data sets including the applied
completeness criteria is provided in Table 3-5. Figure 3-20 and Figure 3-21 show the location of the
458 year-round and 1,064 warm-season monitors meeting the completeness criteria for all
three years (2007-2009).
Table 3-5. Summary of ozone data sets originating from AQS
Year-Round Data Set
Warm-Season Data Set
Years
2007-2009
2007-2009
Months
January-December (12 mo)
May-September (5 mo)
Completeness Criteria
75% of hours in a day
75% of hours in a day
75% of days in a calendar quarter
75% of days between May - September
all 4 quarters per year
Number of monitors meeting completeness criteria 618 containing at least one valid year in 2007-2009 1,265 containing at least one valid year in 2007-2009
549 containing at least two valid years in 2007-2009 1,168 containing at least two valid years in 2007-2009
458 containing all three valid years in 2007-2009 1,064 containing all three valid years in 2007-2009
Figure 3-20. Location of the 458 ozone monitors meeting the year-round completeness criterion
for all 3 years between 2007 and 2009.
1 Concentrations that might have been affected by exceptional events (and contribute to a violation of the NAAQS) can be flagged in the
Air Quality System (AQS) by the reporting organization. Exceptional events are defined as unusual or naturally occurring events, that
can affect air quality but are not reasonably controllable using techniques that tribal, state or local air agencies may implement in order to
attain and maintain the National Ambient Air Quality Standards (NAAQS). The corresponding EPA Regional Office is responsible for
reviewing the data and evidence of the event, and deciding whether to concur with the flag. Flagged data that has been concurred by the
Regional office is typically excluded for regulatory purposes.
March 2011
3-59
DRAFT - DO NOT CITE OR QUOTE
-------
Figure 3-21. Location of the 1,064 ozone monitors meeting the warm-season completeness
criteria for all 3 years between 2007 and 2009.
1 Tabulated statistics generated from the year-round and warm-season data sets are included in
2 Table 3-6 and Table 3-7, respectively. This information was used to compare (1) the year-round and
3 warm-season data sets; (2) the O3 distribution variability across years (2005-2009); and (3) four
4 different averaging times (1-h avg, 24-h avg, 1-h daily max, and 8-h daily max). Summary statistics
5 for 2005 and 2006 were added to these tables in order to gain a broader view of year-to-year
6 variability, but the year-round and warm-season data sets used in the rest of this section are limited to
7 2007-2009 as described above. The 8-h daily max pooled by site was also included in these tables to
8 show the distribution of the annual and 3-year (2007-2009) site-averages of the 8-h daily max
9 statistic.
March 2011
3-60
DRAFT - DO NOT CITE OR QUOTE
-------
Table 3-6. Nationwide distributions of
Time Period
N Monitors
NObs
ozone concentrations (ppb) from
Mean
SD Min
1 5
10
25
the
50
year-round
75
90
95
data set
96
97 98
99
1 -h avg
2005
2006
2007
2008
2009
2007-2009
499
532
522
520
551
599
4,284,219
4,543,205
4,547,280
4,470,065
4,716,821
13,734,166
29
30
29
30
29
29
18 2
18 2
18 2
17 2
16 2
17 2
2 2
2 2
2 2
2 2
2 2
2 2
2
5
5
6
6
6
15
16
16
17
17
17
28
29
29
29
29
29
41
42
41
41
40
40
53
54
52
52
50
51
61
61
60
59
56
58
64
64
62
61
58
60
67 71
67 71
65 68
64 67
61 64
63 67
78
78
75
74
70
73
24- h avg
2005
2006
2007
2008
2009
2007-2009
504
536
531
528
556
611
183,815
194,884
194,873
191,875
202,147
588,895
29
30
29
30
29
29
13 2
13 2
12 2
12 2
11 2
12 2
4 9
5 10
5 11
5 11
6 11
5 11
13
14
14
14
14
14
20
21
20
21
21
21
28
29
29
29
28
29
37
38
37
38
37
37
46
47
45
46
44
45
51
52
50
50
48
49
52
54
52
52
49
51
54 57
55 58
53 56
54 56
51 53
53 55
61
62
60
61
57
60
1-h daily max
2005
2006
2007
2008
2009
2007-2009
504
536
531
528
556
611
183,815
194,884
194,873
191,875
202,147
588,895
48
48
47
47
45
46
18 2
18 2
17 2
17 2
16 2
16 2
11 21
13 23
14 23
14 23
14 22
14 23
26
28
28
27
27
27
35
36
36
35
35
35
46
46
45
45
44
44
58
58
57
56
54
55
71
71
69
67
64
67
80
80
77
76
72
75
83
82
79
78
75
78
86 91
86 91
82 87
82 87
78 83
81 86
100
100
94
96
91
94
8-h daily max
2005
2006
2007
2008
2009
2007-2009
504
536
528
528
556
608
183,279
194,285
194,266
191,283
201 ,535
587,084
42
42
41
41
40
41
16 2
16 2
15 2
15 2
14 2
15 2
8-h daily max (pooled
2005
2006
2007
2008
2009
2007-2009
508
538
538
529
558
458
508
538
538
529
558
458
42
42
41
41
40
41
6 23
6 12
6 17
6 20
6 20
6 19
7 16
9 18
10 19
11 19
11 18
10 19
by site)
27 32
28 31
27 31
28 31
26 30
29 32
21
23
23
23
23
23
34
34
34
34
33
34
30
31
31
31
30
31
38
38
38
37
36
38
40
41
40
40
39
40
42
43
41
40
39
40
52
52
51
51
49
50
45
46
45
45
44
45
63
63
61
60
57
60
48
50
49
50
48
49
70
70
68
66
63
66
51
52
51
52
50
51
72
72
69
69
65
68
51
53
52
52
51
52
75 78
75 79
72 75
71 75
68 71
70 74
52 53
53 54
53 54
54 55
52 53
52 54
84
85
81
82
77
80
55
55
55
57
54
55
aAQS Site ID corresponding to the observation in the Max column
March 2011
3-61
DRAFT - DO NOT CITE OR QUOTE
-------
Table 3-7. Nationwide distributions of ozone concentrations (ppb) from the warm-season data set
1
2
3
4
5
6
7
Time Period
2005
2006
2007
2008
2009
2007-2009
2005
2006
2007
2008
2009
2007-2009
2005
2006
2007
2008
2009
2007-2009
2005
2006
2007
2008
2009
2007-2009
2005
2006
2007
2008
2009
2007-2009
N Monitors N Obs Mean SD Min 1 5 10 25 50 75 90 95 96 97 98 99
1,023
1,036
1,021
1,034
1,027
1,102
1,103
1,110
1,100
1,120
1,139
1,196
1,103
1,110
1,100
1,120
1,139
1,196
1,104
1,112
1,097
1,120
1,139
1,193
1,141
1,152
1,164
1,163
1,170
1,064
1 -h avg
7,455,018 30 19 2 2 2 5 16 29 43 55 64 66 69 73 79
7,590,796 31 18 2 2 2 6 17 30 43 55 62 65 67 71 77
7,711,463 31 18 2 2 2 6 18 30 43 55 63 65 68 71 77
7,701,597 31 17 2 2 2 7 18 30 42 53 60 62 65 68 74
7,825,513 29 16 2 2 2 7 17 29 40 50 56 58 60 63 69
23,238,573 30 17 2 2 2 7 18 30 42 53 60 62 64 68 74
24-h avg
319,410 30 12 2 5 10 14 22 30 39 46 51 53 55 57 61
324,993 31 12 2 6 12 15 22 30 39 47 52 53 55 58 61
330,197 31 12 2 6 12 16 23 31 39 47 51 53 55 57 61
329,918 31 12 2 6 12 16 22 30 38 46 50 52 53 56 60
334,951 29 11 2 6 12 15 21 29 37 44 48 49 50 53 56
995,066 30 12 2 6 12 16 22 30 38 45 50 51 53 55 59
1-h daily max
319,410 50 18 2 12 23 28 38 49 61 74 81 84 87 91 99
324,993 50 17 2 15 25 29 38 48 60 72 80 82 85 90 98
330,197 50 17 2 16 25 30 38 48 60 72 80 82 85 88 95
329,918 48 16 2 16 25 29 37 47 58 69 76 78 81 86 93
334,951 46 15 2 15 23 28 36 45 54 64 71 73 76 80 87
995,066 48 16 2 16 24 29 37 47 58 68 76 78 81 85 93
8-h daily max
318,771 44 16 2 9 18 23 32 43 55 66 72 74 76 79 85
324,327 44 16 2 11 20 25 33 43 54 64 70 72 75 78 84
329,482 44 15 2 12 20 25 33 43 54 65 71 72 75 78 82
329,223 43 15 2 12 20 25 33 42 52 61 67 69 71 74 80
334,250 40 13 2 12 19 24 31 40 49 57 63 64 66 69 75
992,955 42 15 2 12 20 24 32 42 52 61 67 69 71 75 80
8-h daily max (pooled by site)
1,141 45 6 14 28 34 36 41 46 49 52 54 54 55 56 57
1,152 44 6 12 29 34 37 41 45 48 51 54 54 55 58 59
1,164 45 7 17 28 34 36 40 45 50 54 56 56 57 58 59
1,163 43 6 20 29 33 36 39 44 48 50 53 53 55 56 58
1,170 41 5 20 28 32 35 38 41 44 47 50 51 52 53 55
1,064 43 6 19 29 34 36 39 43 47 50 52 53 54 55 57
The year-round data set includes data from less than half the number of monitors as the warm-
season data set and a larger fraction of the year-round monitors are located in the southern half of the
U.S. due to extended monitoring requirements in these areas. Despite these differences, the mean,
SD and percentiles of the nation-wide O3 concentrations were quite similar for the year-round data
presented in Table 3-6 and the warm-season data presented inTable 3-7. In both data sets, there was
very little variability across years in the central statistics; for example, the median 1-h avg
concentrations between 2005 and 2009 ranged from 28 to 29 ppb for the year-round data and from
March 2011
3-62
DRAFT - DO NOT CITE OR QUOTE
-------
1 29 to 30 ppb for the warm-season data. The 8-h daily max showed similar uniformity in median
2 across the five years, with concentrations ranging from 39 to 41 ppb for the year-round data and
3 from 40 to 43 for the warm-season data. The upper percentiles (95th and above) showed a downward
4 trend from 2005 to 2009 in both nation-wide data sets. For example, the 99th percentile of the 8-h
5 daily max observed in the warm-season data dropped from 85 ppb in 2005 to 75 ppb in 2009. Trends
6 in O3 concentrations investigated over a longer time period are included in Section 3.6.3.1.
7 Given the strong diurnal pattern in O3 concentrations, the selection of averaging time has a
8 substantial effect on the magnitude of concentration reporting. The nation-wide median 1-h avg,
9 24-h avg, 1-h daily max, and 8-h daily max concentrations for the year-round data set in 2009 were
10 29, 28, 44 and 39 ppb, respectively. The median concentrations for the warm-season data set in 2009
11 were: 29, 29, 45 and 40 ppb, respectively. The 1-h avg and 24-h avg both include the lowest
12 concentrations typically observed in the overnight period which lowers their values relative to the
13 daily maximum statistics.
14 A strong seasonal pattern in O3 concentrations can also be seen in the year-round data.
15 Table 3-8 shows the 8-h daily max stratified by season, with the seasons defined as:
16 • winter: December-February;
17 • spring: March-May;
18 • summer: June-August; and
19 • fall: September-November.
20 In addition, warm-season (May-Sept) and cold-season (Oct-Apr) stratifications of the year-round
21 data set are included in the table for comparison with the four seasonal stratifications. Substantial
22 seasonal variability in the 8-h daily max concentration for the period 2007-2009 was evident with
23 lower concentrations present in fall (median = 36 ppb) and winter (median = 32 ppb) and higher
24 concentrations in spring (median = 47 ppb) and summer (median = 46 ppb). The seasonal differences
25 were even more pronounced in the upper percentiles. For example, the 99th percentile in the 8-h
26 daily max over the 2007-09 time period ranged from 52 ppb in winter to 90 ppb in summer. The
27 distribution in 8-h daily max O3 during the warm-season (as defined above) and during summer were
28 very similar, which is not surprising given their close overlap in months. The distribution during the
29 cold-season (as defined above) is shifted toward higher 8-h daily max O3 concentrations compared
30 with the distribution during winter. This is a result of including the four transition months (Oct, Nov,
31 Mar and Apr) in the cold-season when high O3 concentrations can occur. Further investigation of
32 temporal variability including multiyear trends and diel behavior is included in Section 3.6.3.
March 2011 3-63 DRAFT - DO NOT CITE OR QUOTE
-------
Table 3-8. Seasonally stratified distributions of 8-h daily max ozone concentrations (ppb) from the
year-round data set (2007-2009)
Time Period Monitors N Obs Mean SD Min 1 5 10 25 50 75 90 95 96 97 98 "
8-h daily max (2007-2009)
Year-round 608 587,084 41 15 2 10 19 23 31 40 50 60 66 68 70 74 80
8-h daily max by season (2007-2009)
Winter (Dec-Feb) 608 143,847 31 10 2 6 14 18 25 32 38 43 46 47 48 49 52
Spring (Mar-May) 612 148,399 47 12 2 20 28 33 40 47 55 62 67 68 70 72 77
Summer (Jun-Aug) 613 148,280 47 16 2 16 22 26 35 46 57 67 75 77 80 84 90
Fall (Sep-Nov) 608 146,558 37 13 2 10 17 21 28 36 45 54 61 63 65 68 75
Warm-season (May-Sep) 616 246,225 47 16 2 16 22 27 35 46 57 66 73 75 78 81 87
Cold-season (Oct-Apr) 608 340,859 36 12 2 8 16 21 28 36 44 52 57 59 61 63 67
1 A national picture of AQS O3 concentrations was generated from the year-round and warm-
2 season data sets by aggregating the 8-h daily max observations by U.S. county. For this purpose, the
3 8-h daily max concentrations at each site were averaged over one or more calendar years and then
4 the highest site in each county was selected for that county. Figure 3-22 contains the county-scale
5 8-h daily max O3 concentrations from the year-round data set for 2007-2009 (top map) with seasonal
6 stratification (bottom four maps). Figure 3-23 contains the county-scale 8-h daily max O3
7 concentrations from the warm-season data set for 2007-2009 (top map) along with individual maps
8 for each calendar year between 2007 and 2009 (bottom three maps). These maps are meant to
9 illustrate the general national-scale distribution in long-term average 8-h daily max O3
10 concentrations and are not representative of O3 concentrations at all locations or times within the
11 counties shown; considerable spatial variability can exist within a county. This is particularly
12 important in the West where counties are larger on average than in the East. These maps are limited
13 by monitor availability, resulting in the majority of U.S. counties not having available data (the white
14 regions in Figure 3-22 and Figure 3-23).
March 2011 3-64 DRAFT - DO NOT CITE OR QUOTE
-------
O3 > 60 ppb
50
-------
O3 > 60 ppb I
50
-------
1 As shown in the top county-scale map in Figure 3-22, the highest 3-year avg (2007-2009) 8-h
2 daily max O3 concentrations (> 50 ppb) occur in counties in southern California, Arizona, Colorado
3 and high elevation counties in Tennessee. Site #060710005 in San Bernardino County, CA had the
4 highest average of 61 ppb over this period. The lowest 3-year avg 8-h daily max O3 concentrations
5 (<30 ppb) occur in Pacific Coast counties in northern California and Washington as well as in two
6 northeastern counties in Pennsylvania and Massachusetts. The seasonally-stratified county-scale
7 maps in Figure 3-23 reinforce the strong seasonality in 8-h daily max O3 concentrations shown in
8 Table 3-8. The highest wintertime concentrations (> 40 ppb) occur in the West with the highest
9 3-year wintertime avg of 46 ppb calculated for site #080690007 in Larimer County, CO. In spring
10 and summer, the concentrations increase considerably across all counties, with the highest
11 concentrations (> 60 ppb) occurring during the summer in 15 counties in southern California, 3
12 counties in Colorado and 1 county each in Nevada and Arizona. Many counties in rural Wyoming,
13 Montana, North Dakota, Maine, and along the Gulf Coast peak in the spring instead of the summer.
14 In the fall, 8-h daily max O3 concentrations drop back down below their spring and summer
15 concentrations.
16 The top county-scale map in Figure 3-23 based on the 2007-2009 warm-season data set looks
17 similar to the corresponding map in Figure 3-22 based on the year-round data set. The warm-season
18 map, however, incorporates approximately twice as many monitors across the U.S., providing more
19 spatial coverage. Several counties in Utah, New Mexico, Indiana, Ohio, Maryland, North Carolina,
20 and Georgia in addition to California, Arizona, Colorado and Tennessee identified above have 3-year
21 avg (2007-2009) 8-h daily max O3 concentrations > 50 ppb based on the warm-season data set. The
22 individual yearly average county-maximum 8-h daily max O3 concentrations in the lower half of
23 Figure 3-22 show a general decrease in most counties from 2007 to 2009. The number of counties
24 containing a monitor reporting an annual average 8-h daily max O3 concentration above 50 ppb
25 dropped from 231 counties in 2007 to 29 counties in 2009. This is consistent with the general
26 decrease across these years shown in Table 3-6 and Table 3-7 for the upper percentiles of the 8-h
27 daily max O3 concentration.
Urban-Scale Variability
28 Statistical analysis of the human health effects of airborne pollutants based on aggregate
29 population time-series data have often relied on ambient concentrations of pollutants measured at
30 one or more central monitoring sites in a given metropolitan area. The validity of relying on central
31 monitoring sites is strongly dependent on the spatial variability in concentrations within a given
32 metropolitan area. To investigate urban-scale variability, 20 focus cities were selected for closer
33 analysis of O3 concentration variability; these cities are listed in Table 3-9 and were selected based
34 on their importance in O3 epidemiology studies and on their geographic distribution across the U.S.
35 In order to provide a well-defined boundary around each city, the combined statistical area (CSA)
36 encompassing each city was used. If the city was not within a CSA, the smaller core-based statistical
37 area (CBSA) was selected. The CSAs/CBSAs are defined by the U.S. Census Bureau
March 2011 3-67 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
(U.S. Census Bureau, 2011, 677549)1 and have been used to establish analysis regions around cities
in previous IS As for particulate matter (U.S. EPA, 2009, 179916) and carbon monoxide (U.S. EPA,
2010.626035).
Table 3-9. Focus cities used in this and previous assessments
Focus City
Atlanta, GA
Baltimore, MD
Birmingham, AL
Boston, MA
Chicago, IL
Dallas, TX
Denver, CO
Detroit, Ml
Houston, TX
Los Angeles, CA
Minneapolis, MN
New York, NY
Philadelphia, PA
Phoenix, AZ
Pittsburgh, PA
Salt Lake City, UT
San Antonio, TX
San Francisco, CA
Seattle, WA
St Louis, MO
Short Name
Atlanta CSA
Baltimore CSA
Birmingham CSA
Boston CSA
Chicago CSA
Dallas CSA
Denver CSA
Detroit CSA
Houston CSA
Los Angeles CSA
Minneapolis CSA
New York CSA
Philadelphia CSA
Phoenix CBSA
Pittsburgh CSA
Salt Lake City CSA
San Antonio CBSA
San Francisco
CSA
Seattle CSA
St Louis CSA
CSA/CBSA Name3
Atlanta-Sandy Springs-Gainesville
Washington-Baltimore-northern Virginia
Birmingham-Hoover-Cullman
Boston-Worcester-Manchester
Chicago-Naperville-Michigan City
Dallas-Fort Worth
Denver-Aurora-Boulder
Detroit-Warren-Flint
Houston-Baytown-Huntsville
Los Angeles-Long Beach-Riverside
Minneapolis-St. Paul-St. Cloud
New York-Newark-Bridgeport
Philadelphia-Camden-Vineland
Phoenix-Mesa-Scottsdale
Pittsburgh-Newcastle
Salt Lake City-Ogden-Clearfield
San Antonio
San Jose-San Francisco-Oakland
Seattle-Tacoma-Olympia
St. Louis-St. Charles-Farmington
Year-Round 03
Monitoring Sites'1
0
9
1
3
11
19
12
0
21
47
2
20
9
14
2
2
5
25
5
3
Warm-Season 03
Monitoring Sites0
11
19
9
18
15
0
3
9
0
3
6
10
8
17
12
10
0
6
5
13
Included in Prior
ISAsd
CO, PM, SOX,
NOx
PM
CO, PM, NOx
PM, NOx
CO, PM
PM
CO, PM, NOx
CO, PM, SOx,
CO, PM, SOx,
PM, NOx
CO, PM
CO, PM
CO, PM
CO, PM, SOx
NOx
NOx
NOx
'Defined based on 2000 Census data from the U.S. Census Bureau (U.S. Census Bureau, 2011, 677549V
bThe number of sites with AQS monitors meeting the year-round data set inclusion criteria; the year-round data set is limited to these monitors.
cThe number of sites with AQS monitors meeting the warm-season data set inclusion criteria; the warm-season data set includes May - September data
from both the warm-season and year-round monitors.
"Boundaries for CO ISA (U.S. EPA, 2010, 6260351 and PM ISA (U.S. EPA, 2009, 179916) focus cities were based on CSA/CBSA definitions; boundaries
for SOx ISA (U.S. EPA, 2008, 1570751 and NOX ISA (U.S. EPA, 2008, 1570731 focus cities were based on similar metropolitan statistical area (MSA)
definitions from the 1990 U.S. Census.
4 The distribution of the 8-h daily max O3 concentrations from 2007-2009 for each of the 20
5 focus cities is included in Table 3-10. These city-specific distributions were extracted from the
6 warm-season data set and can be compared to the nationwide warm-season 8-h daily max
7 distribution for 2007-2009 in Table 3-7 (and repeated in the first line of Table 3-10). The median 8-h
8 daily max concentration in these focus cities was 41 ppb, similar to the nationwide median of
9 42 ppb. Seattle had the lowest median and Salt Lake City had the highest median of the 20 cities
10 with median 8-h daily max concentrations of 31 and 53 ppb, respectively. The 99th percentile of the
11 8-h daily max concentration in the focus cities was 84 ppb; similar once again to the nationwide 99th
'A CBS A represents a county-based region surrounding an urban center of at least 10,000 people determined using 2000 census data and
replaces the older Metropolitan Statistical Area (MSA) definition from 1990. The CSA represents an aggregate of adjacent CBS As tied
by specific commuting behaviors. The broader CSA definition was used when selecting monitors for the cities listed above with the
exception of Phoenix and San Antonio, which are not contained within a CSA. Therefore, the smaller CBS A definition was used for
these metropolitan areas.
March 2011
3-68
DRAFT - DO NOT CITE OR QUOTE
-------
1 percentile of 80 ppb. Seattle had the lowest 99th percentile and Los Angeles had the highest 99th
2 percentile of the 20 cities with values of 64 and 98 ppb, respectively. In aggregate, the 20 focus cities
3 selected are similar in distribution to the nationwide data set, but there is substantial city-to-city
4 variability in the individual distributions of the 8-h daily max concentrations based on the warm-
5 season data set.
Table 3-10. City-specific distributions of 8-h daily max ozone concentrations (ppb) from the warm-
season data set (2007-2009)
Time Period
N Monitors
NObs
Mean
8-h daily
Nationwide
Atlanta CSA
Baltimore CSA
Birmingham CSA
Boston CSA
Chicago CSA
Dallas CSA
Denver CSA
Detroit CSA
Houston CSA
Los Angeles CSA
Minneapolis CSA
New York CSA
Philadelphia CSA
Phoenix CBSA
Pittsburgh CSA
Salt Lake City CSA
San Antonio CSA
San Francisco CSA
Seattle CSA
St Louis CSA
All CSAs/CBSAs listed
1,193
11
28
10
21
27
19
15
9
21
49
8
21
14
22
13
12
5
31
5
19
360
992,955
8-h daily
7,844
20,999
7,676
12,603
20,764
19,858
12,217
5,016
22,305
49,291
5,285
26,304
12,673
26,129
9,814
5,146
4,701
27,961
6,148
1 1 ,569
314,303
42
max by
47
43
44
41
37
41
44
45
36
47
40
39
41
49
43
51
39
34
31
43
42
SD Min
max
15
1
5
10
25
50 75
90 95
96 97
98 99
(2007-2009)
2
CSA/CBSA
16
16
15
14
14
15
15
14
15
18
12
16
17
12
15
14
13
12
12
15
16
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
12
20
24
32
42 52
61 67
69 71
75 80
(2007-2009)
15
9
14
13
9
11
8
15
8
10
14
6
8
18
12
8
13
8
4
12
9
22
18
21
21
15
20
18
23
15
20
21
15
17
27
19
23
20
16
12
19
18
27
23
25
25
19
24
24
28
19
26
25
20
21
32
24
32
23
20
17
23
22
36
31
34
31
27
31
34
35
25
35
31
28
29
41
32
44
29
27
23
32
31
47 58
43 54
44 54
40 49
37 47
39 50
44 55
44 52
34 46
45 58
40 48
37 47
39 52
50 58
43 53
53 61
37 46
33 41
31 39
43 53
41 52
67 72
64 70
63 68
59 67
57 62
61 67
63 68
62 69
57 64
72 81
54 58
59 68
64 70
65 68
62 68
67 71
56 62
48 55
46 51
61 68
63 69
75 77
72 74
70 73
69 71
64 66
69 71
69 70
72 74
66 68
83 86
59 61
70 73
73 75
69 70
70 72
73 75
63 65
57 59
53 55
69 72
71 74
81 87
78 83
76 83
75 81
69 74
74 79
72 76
77 83
72 78
91 98
63 67
77 83
78 83
72 75
74 78
77 80
67 72
63 68
59 64
76 81
78 84
6 Maps showing the location of central monitoring sites with O3 monitors reporting to AQS for
7 each of the 20 focus cities are included in Chapter 3Appendix, Figure 3A-16 through Figure 3A-35;
8 examples for Atlanta, Boston and Los Angeles are shown in Figure 3-24 through Figure 3-26. The
9 sites are delineated in the maps as year-round or warm-season based on their inclusion in the year-
10 round data set and the warm-season data set (the warm-season data set includes May-September data
11 from both the warm-season monitors and the year-round monitors). The maps also include the
12 CSA/CBSA boundary selected for monitor inclusion, the location of urban areas and water bodies,
13 the major roadway network, as well as the population gravity center based on the entire CSA/CBSA
14 and the individual focus city boundaries. Population gravity center is calculated from the average
March 2011
3-69
DRAFT - DO NOT CITE OR QUOTE
-------
1 longitude and latitude values for the input census tract centroids and represents the mean center of
2 the population in a given area. Census tract centroids are weighted by their population during this
3 calculation.
Legend
Monitor Locations
© Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
(•) CSA-based Population Gravity Center
Interstate Highways
Major Highways
Water Bodies
Urban Areas
Atlanta CSA
30
60 Kilometers
Figure 3-24. Map of the Atlanta CSA including ozone monitor locations, population gravity
centers, urban areas, and major roadways.
March 2011
3-70
DRAFT - DO NOT CITE OR QUOTE
-------
Legend
Monitor Locations
© Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
• CSA-based Population Gravity Center
Interstate Highways
Major Highways
^B Water Bodies
Urban Areas
Boston CSA
100 Kilometers
Figure 3-25. Map of the Boston CSA including ozone monitor locations, population gravity
centers, urban areas, and major roadways.
March 2011
3-71
DRAFT - DO NOT CITE OR QUOTE
-------
Legend
Monitor Locations
O Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
0 CSA-based Population Gravity Center
Interstate Highways
Major Highways
Water Bodies
Urban Areas
Los Angeles CSA
/W
200 Kilometers
Figure 3-26. Map of the Los Angeles CSA including ozone monitor locations, population gravity
centers, urban areas, and major roadways.
1 The Atlanta CSA contains 11 warm-season monitors distributed evenly yet sparsely around the
2 city center (Figure 3-24). The population gravity center for the city and the larger CSA are only
3 separated by 4 km, indicating that the majority of the population lives within or evenly distributed
4 around the city limits. Atlanta is landlocked with a radial network of interstate highways leading to
5 the city center. The Boston CSA contains 3 year-round and 18 warm-season monitors spread evenly
6 throughout the CSA. Boston is a harbor city with the Atlantic Ocean to the east, resulting in the city-
7 based population gravity center being located 17 km east of the CSA-based population gravity
8 center. The Los Angeles CSA contains the largest number of monitors of the 20 CSA/CBSAs
9 investigated with 47 year-round and 3 warm-season monitors. These monitors are primarily
10 concentrated in the Los Angeles urban area with relatively few monitors extending out to the
11 northern and eastern reaches of the CSA. These unmonitored areas are very sparsely populated,
12 resulting in only 15 km separating the city-based and the CSA-based population gravity centers
13 despite the vast area of the Los Angeles CSA.
March 2011
3-72
DRAFT - DO NOT CITE OR QUOTE
-------
1 Other CSAs/CBSAs (see Chapter 3 Appendix) with monitors concentrated within the focus
2 city limits include Birmingham, Chicago, Denver, Houston, Phoenix, San Antonio, and Salt Lake
3 City. The remaining CSAs/CBSAs have monitors distributed more evenly throughout the
4 CSA/CBSA area. Baltimore is contained within the same CSA as Washington DC and suburbs,
5 resulting in a 50-km separation (the largest of the focus cities investigated) between the city-based
6 population gravity center for Baltimore and the CSA-based population gravity center for the
7 Washington-Baltimore-Northern Virginia CSA.
8 Box plots depicting the distribution of 2007-2009 warm-season 8-h daily max O3 data from
9 each individual monitor in the 20 focus cities are included in Chapter 3 Appendix, Figure 3A-36
10 through Figure 3A-55; examples for Atlanta, Boston and Los Angeles are shown in Figure 3-27
11 through Figure 3-29. The Atlanta CSA has very little spatial variability in 8-h daily max O3
12 concentrations with median concentrations ranging from 47 ppb at Sites I and J located far from the
13 city center to 54 ppb at Site A located closest to the city center. The variation in warm-season 8-h
14 daily max concentrations are also relatively uniform across monitors with an IQR ranging from
15 17 ppb at Site J to 23 ppb at Site B. The Boston CSA has more spatial variability in 8-h daily max O3
16 concentrations than the Atlanta CSA with median concentrations ranging from 33 ppb at Site A
17 nearest to the city center to 46 ppb at Site L located 84 km west of the city center. Like the Atlanta
18 CSA, the variation in warm-season 8-h daily max concentrations are relatively uniform across
19 monitors within the Boston CSA with an IQR ranging from 15 ppb at Site U to 21 ppb at Site K. The
20 Los Angeles CSA exhibits the most variability in O3 concentrations between monitors of all the
21 CSAs/CBSAs investigated. The median 8-h daily max O3 concentration in the Los Angeles CSA
22 ranged from 20 ppb at Site AM in the south-central extreme of the CSA to 80 ppb at Site AE near
23 Crestline, CA in the San Bernardino National Forest just north of San Bernardino, CA. These two
24 sites are at approximately the same longitude and are separated by only 85 km, but the Crestline site
25 is downwind of the Los Angeles basin, resulting in substantially higher O3 concentrations. Site AM
26 also contains data for only 2009, which could explain some of the deviation when comparing this
27 site with others in the Los Angeles CSA. Sites AM and AE also had the lowest (8 ppb) and highest
28 (28 ppb) IQR, respectively. The remaining focus cities included in Chapter 3 Appendix exhibited
29 spatial variability ranging from uniform as in the Atlanta CSA to non-uniform as observed in the
30 Los Angeles CSA.
March 2011 3-73 DRAFT - DO NOT CITE OR QUOTE
-------
Atlanta CSA
Site ID
131210055
130890002
131350002
130670003
132470001
130970004
131130001
131510002
130770002
130850001
132230003
Years
07-09
07-09
07-09
07-09
07-09
07-09
07-08
07-09
07-09
07-09
07-09
N
450
452
446
459
450
455
306
459
455
458
455
Mean
53
52
52
51
51
52
52
51
47
47
50
SD
17
18
16
16
18
15
15
17
16
13
14
Median
54
52
52
52
51
53
52
51
47
47
50
IQR
22
23
18
22
22
22
20
22
19
17
21
Site
A-
B-
c-
D-
E-
F-
G-
H-
I-
J-
K-
Key
.C
to
H-
~in
(M
-H
mean
•
median
I
overall
melflan
I
to
h-
\--
In
-------
Los Angeles CSA
Site ID
060371602
060371301
060371302
060371103
060372005
060374002
060595001
060590007
060375005
060371002
060370002
060370113
060370016
060371701
060591003
060371201
060711004
060376012
060650004
060592022
061112002
060658005
060712002
060658001
061110007
060710012
060379033
061110009
060719004
060659001
060710005
060656001
060714003
060714001
060710306
061113001
061111004
061112003
060650009
060650012
060651016
060710001
060655001
060719002
060652002
060651999
060651010
060711234
060650008
060659003
Years
07-09
07-08
09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
08
07-09
07-09
08-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
09
07-09
07-09
07-09
07-09
07-09
07-09
07-08
09
07-09
07,09
07-09
Key
£
I-
N
458
306
152
457
459
459
459
459
459
459
459
459
458
459
459
459
457
457
127
457
455
276
459
440
459
456
452
458
457
453
459
455
459
455
459
453
458
457
153
457
459
455
459
452
448
283
153
453
265
444
c
j= ro
To g
IM E
H •
Mean
48
36
44
46
54
38
50
48
45
56
57
48
64
61
45
61
66
68
69
52
62
65
68
69
54
67
67
58
70
68
79
72
73
68
64
44
57
41
22
73
73
61
69
73
62
49
59
59
58
42
median
I
SD
13
9
10
12
15
10
12
10
9
14
17
10
18
16
9
14
19
18
18
13
12
15
19
16
10
13
13
11
19
16
19
17
18
14
12
9
11
9
8
15
16
11
14
13
13
17
10
10
10
10
overall
"median
I
1
Median
47
34
44
45
53
37
49
47
45
55
56
47
63
60
44
60
66
69
65
50
62
64
67
68
54
67
66
58
70
67
80
73
73
68
64
43
57
40
20
71
73
60
68
73
61
50
59
58
57
42
"to •
\
IOR Site
17
10
12
14
18
11
14
12
12
19
22
13
23
20
12
19
23
27
23
15
16
18
24
18
12
18
19
14
26
21
28
24
25
21
17
11
14
12
8
22
23
15
21
18
18
22
15
13
14
13
!o
H
A-
B-
C-
D-
E-
F-
G-
H-
I -
J-
K-
L-
M-
N-
o-
P-
Q-
R-
S-
T-
U-
V-
w-
X-
Y-
z-
AA-
AB-
AC-
AD-
AE-
AF-
AG-
AH-
Al -
AJ-
AK-
AL-
AN-
AO-
AP-
AQ-
AR-
AS-
AT-
AU-
AV-
AW-
AX-
C
i i i i i i i i 1 i i i i
hgfe....
, i — t — L
r r '
;"a^to---:
"'gjjjb -r H
' 1 . * r '
J ' J L
r i ' T r
• 1, | ™ I '•
^"^SE----; |
>>ijp::-
,--cg3""H
' i 4 i "'
m,
'
;---i--| 1 I H
n 1 '.
•
-------
1 Pair-wise monitor comparisons were used to further evaluate spatial variability between
2 monitors within the 20 focus cities. In the particular case of ground-level O3, central-site monitoring
3 has been justified as a regional measure of exposure mainly on the grounds that correlations between
4 concentrations at neighboring sites measured over time are usually high. In areas with multiple
5 monitoring sites, averages over the monitors have often been used to characterize population
6 exposures. However, substantial differences in concentrations between monitors can exist even
7 though concentrations measured at the monitoring sites are highly correlated, thus leading to the
8 potential for exposure misclassification error. Therefore, both the Pearson correlation coefficient and
9 the coefficient of divergence (COD) were calculated for each monitor pair within the CSA/CBSAs
10 using the 8-h daily max O3 data. The correlation provides an indication of temporal linear
11 dependence across sites while the COD provides an indication of the variability in absolute
12 concentrations across sites. The COD is defined as follows:
COD. =-
Equation 3-1
13 where Xy and Xlk represent observed concentrations averaged over some measurement averaging
14 period i (hourly, daily, etc.) at sites j and k, and p is the number of paired observations. A COD of 0
15 indicates there are no differences between concentrations at paired sites (spatial homogeneity), while
16 a COD approaching 1 indicates extreme spatial heterogeneity. These methods for analysis of spatial
17 variability follow those used in previous IS As for CO, PM, SOX and NOX as well as those used in
18 Pinto et al. (2004, 025033) for PM25.
19 Histograms and contour matrices of the Pearson correlation coefficient between 8-h daily max
20 O3 concentrations from each monitor pair are shown in Chapter 3 Appendix, Figure 3A-56 through
21 Figure 3A-75; examples for Atlanta, Boston and Los Angeles are shown in Figure 3-30 through
22 Figure 3-32. Histograms, contour matrices, and scatter plots of the COD between 8-h daily max O3
23 concentrations from each monitor pair are shown in Figure 3A-76 through Figure 3A-95; examples
24 for Atlanta, Boston and Los Angeles are shown in Figure 3-33 through Figure 3-35. These figures
25 also contain scatter plots of correlation and COD as a function of straight-line distance between
26 monitor pairs.
March 2011 3-76 DRAFT - DO NOT CITE OR QUOTE
-------
Atlanta CSA
20-
§ 15-
0 10-
5 -
B
18
25
6
I
-0.1
0.0
O.I
0.2
0.3
0.4 0.5
Correlation
0.6
0.7
0.8
0.9
1.0
0.69 D.aS 0 3-3 O.B9
0.75 0.76 - A
1.0 -
0.9-
0.8 -
0.7 -
0.6 -
I 0.5H
CO
O n A -
O
0.3-
0.2 -
0.1 -
0.0 -
-0.1
0.86 D.ss : ;: 0.32 o.sa O.SD : :-.- 0.74 0.75 - B
0.79 D.77 0 73 0.75 0.78 0.79 0.6B - C
0.73 0.90 0.82 0.77 0.81 0.81 0.88 • - D
0.85 0.68 0.71 - E
84 0.76 . -: 0.75 ^^^1 - p
- H
- J
K
0 50 100 150 200 250 300 350 400 450
Distance (km)
Figure 3-30. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the Atlanta
CSA. [The colors in the histogram bins correspond to the levels of the contour
matrix. The histogram includes the number of monitor pairs per bin and the contour
matrix includes the numeric values of the correlations.
March 2011
3-77
DRAFT - DO NOT CITE OR QUOTE
-------
Boston CSA
o
40-
20
-C
-> I0
, |
.1 0.0 0.1 0.2 0.3 0.4 0.5 0
Correlation
22
6 0
61
7 0
80
8 0
34
1.0
1.0
0.9
0.8
0.7
0.6
I 0.5
0.4
0.3
0.2
0.1
0.0
-0.1
o
Vjf
•V-r/vS- •
• rktf<•".»..
•*.. f. :*. {,
50 100 150 200 250 300
Distance (krn)
350
400
450
Figure 3-31. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the Boston
CSA. [The colors in the histogram bins correspond to the levels of the contour
matrix. The histogram includes the number of monitor pairs per bin and the contour
matrix includes the numeric values of the correlations.
March 2011
3-78
DRAFT - DO NOT CITE OR QUOTE
-------
o
O
Los Angeles CSA
150
100-
50-
-c
4
.1 0
62
0 0
1 1 n
1 0
1fiQ
2 0
164
3 0
(
148
4 0
^orrelatiot
148
5 0
i
150
6 0
144
7 0
87
8 0.9 1.0
50
100 150 200 250 300 350 400 450
Distance (km)
Figure 3-32. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the
Los Angeles CSA. [The colors in the histogram bins correspond to the levels of the
contour matrix. The histogram includes the number of monitor pairs per bin and the
contour matrix includes the numeric values of the correlations.
March 2011
3-79
DRAFT - DO NOT CITE OR QUOTE
-------
Atlanta CSA
8
0
30-
25-
20-
15'
10-
5-
0.00 0.05 010 0.15
0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40 0.45 0.50
0.55
CD
O
0.55-
0.50-
0.45-
0.40-
o
o
0.35-
D
S 0.30-
-------
Boston CSA
o
o
100-
80 -
60 -
40-
20 -
0.00
I
tt>
0.55-
0.50 -
0.45 -
0.40-
0.35-
Q 0.30 -
"o
£ 0 25 -
i
Q 0.20-
0.15-
0.10-
0.05 -
0.00
0.05 0.10 0.15 0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40
0.45
O a:
0.50
0 50 100 150 200 250 300
Distance (km)
350
400
450
500
0.55
oae 0.11 011 o
Figure 3-34. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and
scatter plot versus distance between monitors (bottom) for the Boston CSA. [The
colors in the histogram bins correspond to the levels of the contour matrix. The
histogram includes the number of monitor pairs per bin and the contour matrix
includes the numeric values of the CODs.
March 2011
3-81
DRAFT - DO NOT CITE OR QUOTE
-------
Los Angeles CSA
400-
_ 300-
c
o 200-
O
100-
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55
Coefficient of Divergence
0.00
0 50 100 150 200 250 300 350 400 450 500
Distance (km)
Figure 3-35. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle) and
scatter plot versus distance between monitors (bottom) for the Los Angeles CSA.D
The colors in the histogram bins correspond to the levels of the contour matrix. The
histogram includes the number of monitor pairs per bin and the contour matrix
includes the numeric values of the CODs.
March 2011
3-82
DRAFT - DO NOT CITE OR QUOTE
-------
1 The monitor pairs within the Atlanta CSA (Figure 3-30) were generally well correlated with
2 correlations between 8-h daily max O3 concentrations ranging from 0.61 to 0.96. The correlations
3 shown in the scatter plot were highest for close monitor pairs and dropped off with distance in a
4 near-linear form. At a monitor separation distance of 50 km or less, the correlations ranged from 0.79
5 to 0.96. The monitor pairs within the Boston CSA (Figure 3-31) were also generally well correlated
6 with correlations ranging from 0.56 to 0.97. Again, the correlations shown in the scatter plot were
7 highest for close monitor pairs, but there was slightly more scatter in correlation as a function of
8 distance in the Boston CSA compared with the Atlanta CSA. At a monitor separation distance of 50
9 km or less, the correlations ranged from 0.74 to 0.97. The monitor pairs within the Los Angeles CSA
10 (Figure 3-32) showed a much broader range in correlations, extending from -0.06 to 0.97. At a
11 monitor separation distance of 50 km or less, the correlations shown in the scatter plot ranged from
12 0.21 to 0.97. The negative and near-zero correlations were between monitors with a relatively large
13 separation distance (>150 km), but even some of the closer monitor pairs were not very highly
14 correlated. For example, Site AL located at Emma Wood State Beach in Ventura and Site AK situated
15 in an agricultural valley surrounded by mountains 20 km inland (see map in Figure 3-36) had a
16 correlation coefficient of only 0.21 over the 2007-2009 warm-season time period. This was slightly
17 lower than the correlation between Site AL and Site AX on the Arizona border, 441 km away (R =
18 0.28). San Francisco (Figure 3A-73) and Seattle (Figure 3A-74) also showed a broad range in pair-
19 wise correlations, likely resulting from their similar geography where background air coming in
20 from the Pacific Ocean rapidly mixes with urban pollutants such as NOX and VOCs from coastal
21 cities and is transported downwind into diversified terrain to create highly spatially varying O3
22 concentrations.
March 2011 3-83 DRAFT - DO NOT CITE OR QUOTE
-------
•*:
Figure 3-36. Terrain map showing the location of two nearby AQS ozone monitoring sites (red
dots) along the western edge of the Los Angeles CSA. Site AL is near shore, 3 m
above sea level.
Site AK is in an agricultural valley surrounded by mountains, 262 m above sea level.
1 The COD between 8-h daily max O3 measured at paired monitors in all CSAs/CBSAs
2 (Figure 3A-76 through Figure 3A-95) were generally low, with values similar to those shown in
3 Figure 3-33 and Figure 3-34 for Atlanta and Boston. This suggests a generally uniform distribution
4 in the 8-h daily max O3 concentration across monitors within these cities and is consistent with the
5 uniformity observed in the box plots (e.g., Figure 3-27, Figure 3-28, Figure 3A-36 through Figure
6 3A-55). Los Angeles (Figure 3-29) and San Francisco (Figure 3A-93), however, had several monitor
7 pairs with COD >0.30 indicating greater spatial heterogeneity. This is consistent with the variability
8 observed in the box plots for these two CSAs (Figure 3-29 and Figure 3A-53). In particular, Site AM
9 in the Los Angeles CSA had consistently lower concentrations (median = 20 ppb, IQR = 17-25 ppb)
10 relative to other sites in the CSA (Figure 3-26), resulting in high CODs across the board in Figure 3-
11 35. The O3 monitor at Site AM is a tribal monitor located on the Pechanga Tribal Government
12 Building in Temecula, CA, and began collecting data on June 9, 2008. It is located in a suburban
March 2011
3-84
DRAFT - DO NOT CITE OR QUOTE
-------
1 setting and is classified as a general background monitor. Another close by tribal monitoring site
2 (site ID = 060731201) located in the Pala Reservation, 9.5 km south of this one (just outside the
3 boundary of the Los Angeles CSA) reported similarly low 2009 8-h daily max O3 concentrations
4 (median = 28 ppb, IQR = 23-32 ppb) between May-June, 2009 (the only warm-season months with
5 available data from this site between 2007 and 2009).
6 Comparison of monitoring data within the selected focus cities has demonstrated considerable
7 variability in the concentration fields. Median O3 concentrations vary considerably within some
8 urban areas and less so in others. Likewise, pair-wise monitor statistics (R and COD) are very
9 dependent on the urban area under investigation. There are instances where sites in an urban area
10 may be moderately correlated, but show substantial differences in absolute concentrations (e.g., Sites
11 A and D in Boston with R = 0.90, COD = 0.16, and an 11 ppb difference in median 8-h daily max O3
12 concentration). These conclusions are consistent with those drawn in the 2006 O3 AQCD (U.S. EPA,
13 2006, 088089) where a subset of these focus cities were investigated using similar statistics. As a
14 result, caution should be observed in using data from a sparse network of ambient O3 monitors to
15 approximate community-scale exposures.
Neighborhood-Scale Variability and the Near-Road Environment
16 O3 is a secondary pollutant formed in the atmosphere from precursor emissions and therefore
17 is generally more regionally homogeneous than primary pollutants emitted from stationary or mobile
18 point sources. However, O3 titration from primary NO emissions does result in substantial localized
19 O3 gradients. This is evident in the near-road environment where fresh NO emissions from motor
20 vehicles titrate O3 present in the urban background air, resulting in an O3 gradient down-wind from
21 the roadway. Ozone titration occurring in street canyons where NO emissions are continuously being
22 generated is more efficient because of inhibited transport away from the source of NO.
23 Several studies have reported O3 concentrations that increase with increasing distance from the
24 roadway, both upwind and downwind of the road. Beckerman et al. (2008, 096484) measured O3
25 profiles in the vicinity of heavily traveled roadways with Annual Average Daily Traffic (AADT)
26 >340,000 vehicles in Toronto, Canada. Ozone was observed to increase with increasing distance
27 from the roadway, both upwind and downwind of the road. This is consistent with scavenging of O3
28 in the near-road environment by reaction with NO to form NO2. Upwind of the road, concentrations
29 were >75% of the maximum observed value at >100 m from the road; downwind, concentrations
30 were approximately 60% of the maximum within 200-400 m of the road. The O3 concentration
31 adjacent to the road on the upwind side was approximately 40% of the maximum value observed at
32 the site. Concentrations measured with Ogawa passive samplers over a 1-week period ranged from
33 7.3-19.4 ppb with the mean at the two sites ranging from 13.0-14.7 ppb. In a study of patrol cars
34 during trooper work shifts, Riediker et al. (2003, 043761) made simultaneous 9-h O3 measurements
35 inside patrol cars, at the roadside, and at a centrally-located ambient monitoring site. The roadside
36 concentrations were approximately 81% of the ambient values (mean of 22.8 ppb versus 28.3 ppb).
37 Wind direction relative to the roadway was not reported.
March 2011 3-85 DRAFT - DO NOT CITE OR QUOTE
-------
1 Johnson (1995, 079215) measured O3, NO, and CO concentrations at upwind and downwind
2 locations near a variety of roadways in Cincinnati, OH. The effects of O3 scavenging by NO were
3 apparent in the O3 reduction in the interval between 9 m upwind and 82 m downwind of the road. A
4 similar effect was observed by Rodes and Holland (1981, 041110) during an earlier study in which
5 outdoor O3 concentrations were monitored downwind of a freeway in Los Angeles, CA. In this study,
6 O3 concentrations measured near the roadway were approximately 20% of the concentrations
7 measured simultaneously at more distant locations judged to be unaffected by the roadway. Minimal
8 separation distances of the samplers from the roadway to eliminate measurable influence were
9 estimated to be approximately 400-500 m for NO, NO2, and O3. Similar results have been observed
10 outside the U.S., e.g., in the city of Daegu, Korea, where the yearly roadside concentrations of CO
11 and SO2 showed a well-defined decreasing trend with distance from the roadway, whereas
12 concentrations of NO2 and O3 exhibited the reverse trend, suggesting that attention should be given
13 to the NO2 and O3 exposures of residents living near roadways (Jo and Park, 2005, 674762) . During
14 the peak O3 month of May, O3 concentrations in a residential neighborhood were approximately 40%
15 higher than concentrations at roadside monitors located 1 m from the edge of multiple-lane freeways.
3.6.2.2. Rural-Focused Variability and Ground-Level Vertical Gradients
16 AQS O3 data for monitors located at several rural monitoring sites (e.g., national parks,
17 national forests, state parks, etc.) were used to investigate rural-focused O3 concentration variability
18 in contrast with the urban-focused variability discussed in Section 3.6.2.1. These rural monitoring
19 sites tend to be less directly affected by obvious anthropogenic pollution sources than urban sites.
20 However, they can be regularly affected by transport of O3 or O3 precursors from upwind urban
21 areas, or by local anthropogenic emissions within the rural areas such as emissions from motor
22 vehicles, power generation, biomass combustion, or oil and gas operations. As a result, monitoring
23 data from these rural locations are not unaffected by anthropogenic emissions.
24 Six rural focus areas were selected for their geographic distribution across the U.S. as well as
25 their unique topography and relevance to the ecological assessment in Chapter 9. Table 3-11 lists the
26 rural focus areas and provides some cursory site information along with the number of available
27 AQS monitors reporting year-round and only during the warm-season. Accompanying box plots
28 depicting the distribution of 2007-2009 warm-season 8-h daily max O3 data from each individual
29 monitor in the six rural focus areas are included in Figure 3-37. This analysis was restricted to AQS
30 monitors meeting the same data completeness criteria outlined in Table 3-5 for a direct comparison
31 with the 20 urban focus areas investigated in Section 3.6.2.1. Given the population-center emphasis
32 of the AQS network, limited monitoring sites (between one and five) were available for each rural
33 focus area. Expanded analyses of O3 concentrations measured using the more rural-focused
34 CASTNET monitoring network are included in Chapter 9.
March 2011 3-86 DRAFT - DO NOT CITE OR QUOTE
-------
Table 3-11. Rural focus areas
Focus Area
Adirondack State Park,
NY
Mount Mitchell State
Park, NC
Great Smoky Mountain
National Park, NC-TN
Rocky Mountain
National Park, CO
San Bernardino National
Forest, CA
Sequoia National Park,
CA
-.^ Year-Round O3
S"°rt Monitoring
Name Sites3
ADSP
MMSP
SMNP
RMNP
SBNFC
SENP
1
0
2
1
1
2
Warm-Season ,,„„=»„,
03 Monitoring ^on-tor „ ^
0 1,483
1 1,982
3 564-2,021
0 2,743
0 1,384
0 560-1,890
Site Descriptions
One site on the summit of Whiteface Mountain in the
Adirondack Mountains
One site near the summit of Mount Mitchell (highest point
in the eastern U.S.) in the Appalachian Mountains
Five different locations within Great Smoky Mountain
National Park in the Appalachian Mountains
One site in a valley at the foot of Longs Peak in the Rocky
Mountains
One site in Lake Gregory Regional Park (near Crestline,
CA) in the San Bernardino Mountains
Two contrasting sites at different elevations within Sequoia
NP in the Sierra Nevada Mountains
'Number of AQS monitors meeting the year-round data set inclusion criteria; the year-round data set is limited to these monitors.
Number of AQS monitors meeting the warm-season data set inclusion criteria; the warm-season data set includes May-September data from both the
warm-season and year-round monitors.
"Same AQS site as Site AE in the Los Angeles CSA shown in Figure 3-26.
Years N
07-09 445
07-09 447
07-09 456
07-09 459
07-09 459
07-09 458
471550102 07-09 457
080690007 07-09 456
060710005
061070009
Site ID
360310002
371990004
370870036
470090102
470090101
471550101
061070006
Rural Focus Areas
Mean SD Median IQR Area Site
07-09 459
07-09 416
07-09 459
50
54
52
47
57
58
60
56
79
76
68
13
11
12
12
13
11
11
9
19
16
15
49
54
51
47
57
58
60
56
80
76
69
16
14
15
16
16
14
13
11
28
21
19
ADSP
MMSP
SMNP
RMNP
SBNF
SENP
Key
to
H--
m
CM
-\
C
CO
0)
•
c
-o
0)
1
C v>
^ o>
1- H
50
100
150
03 (ppb)
Figure 3-37. Rural focus area site information, statistics and box plots for 8-h daily max ozone
from AQS monitors meeting the warm-season data set inclusion criteriar^ithin the
rural focus areaslticluding: Adirondack State Park, NY (ADSP); Mount Mitchell State
Park, NC (MMSP); Great Smoky Mountain National Park, NC-TN (SMNP); Rocky
Mountain National Park, CO (RMNP); San Bernardino National Forest, CA(SBNF);
and Sequoia National Park, CA(SENP).
Eastern Rural Focus Areas
1 In the East, the distribution in warm-season 8-h daily max O3 concentrations from the
2 Adirondack State Park (ADSP) site on Whiteface Mountain in Upstate NY (median = 49 ppb)
3 (Figure 3-37) was among the lowest of the rural focus monitors investigated, but was still higher
March 2011
3-87
DRAFT - DO NOT CITE OR QUOTE
-------
1 than concentration distributions measured in the Boston CSA (medians ranging from 33 to 46 ppb)
2 (Figure 3-28) located 320 km to the southeast. The ADSP AQS site was included in an analysis for
3 the 2006 O3 AQCD (U.S. EPA, 2006, 088089) and had the lowest year-round median hourly O3
4 concentration of the rural forested sites investigated (including Yellowstone NP, the Great Smoky
5 Mountains NP, and Shenandoah NP). For the Appalachian Mountain monitors in Mount Mitchell
6 State Park, NC (MMSP) and Great Smoky Mountain National Park, NC-TN (SMNP), there was a
7 fair amount of variability in concentration distribution. Within SMNP, the median warm-season 8-h
8 daily max O3 concentration ranged from 47 ppb at the lowest elevation site (elevation = 564 m; site
9 ID = 470090102) to 60 ppb at the highest elevation site (elevation = 2021 m; site ID = 471550102);
10 these sites are shown on the terrain map in Figure 3-38. The warm-season median 8-h daily max O3
11 concentration gradient between these two sites located 26.2 km apart in SMNP was 0.9 ppb per
12 100 m elevation gain.
13 Data from the five sites within SMNP allowed for further investigation of spatial variability
14 within the park; Figure 3-39 contains histograms, contour plots and scatter plots as a function of
15 distance for the pair-wise correlation and COD (defined in Equation 3-1) for SMNP. The correlations
16 between the five sites ranged from 0.78 to 0.92 and the CODs ranged from 0.04 to 0.16. The plots of
17 correlation and COD as a function of distance between SMNP monitor pairs in Figure 3-39 show a
18 large degree of spatial variability between monitors over relatively short distances. A host of factors
19 may contribute to these variations, including proximity to local O3 precursor emissions, variations in
20 boundary-layer influences, meteorology and stratospheric intrusion as a function of elevation, and
21 differences in wind patterns and transport behavior due to local topography.
22
March 2011 3-88 DRAFT - DO NOT CITE OR QUOTE
-------
C«y
, Seville
.
©
•:•;;,..
*
r.
.; ••
.Vaifahd .- r ;,>;_, =.
X -< - <~ ,--• j
f
1 0 km
©
Figure 3-38. Terrain map showing the location of five AQS ozone monitoring sites (green/black
stars) in Great Smoky Mountain National Park, NC-TN (SMNP). [The lowest elevation
site (site ID = 470090102) is 564 m above sea level and the highest elevation site
(site ID = 471550102) is 2021 m above sea level.
March 201 1
3-89
DRAFT - DO NOT CITE OR QUOTE
-------
o ;>
o -]
Smokey Mtn NP, NC-TN
0.
-0.1 0.0 0.1 0.2 0.3 C.4 :.5 0.6 0.7 0.3 O.a 1.0
Correlation
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.5D D.55
Coefficient of Divergence
1.0
0.9
0.6
0.7
C 0.6-
g
ro
ffl
t 0.4.
O 0.3-
0.2
D-1
0.0
-0.1
0.55-
0.50-
8 0.45.
C
0 50 1:0 150 200 250 300 350 40: 450
•-1 0.30-
"5
~ o.:s-
1°
"£ 0.15-
O
O 0.10-
o.:s-
o.:o
0.04 006
50 100 150 200 250 300 350 400 450 500
Distance (km)
Distance (km)
Figure 3-39. Pair-wise monitor correlations (left) and coefficients of divergence [{COD, right)
expressed as a histogram (top), contour matrix (middle) and scatter plot vs distance
between monitors (bottom) for Great Smoky Mountain National Park, NC-TN (SMNP).
The colors in the histogram bins correspond to the levels of the contour matrix. The
histograms includes the number of monitor pairs per bin and the contour matrix
includes the numeric values of the correlations and CODs.
Western Rural Focus Areas
1 The Rocky Mountain National Park (RMNP) site in Colorado at 2743 m in elevation had a
2 warm-season 8-h daily max O3 concentration distribution (median = 56 ppb, IQR =11 ppb) (Figure
3 3-37) that is comparable to the distributions at sites in the Denver CSA located 75 km southeast at
4 elevations around 1,600 m (medians ranging from 41 to 59 ppb, IQRs ranging from 10 to 16 ppb; see
5 Figure 3A-27). In nearby Boulder County, CO, a 1-year time-series (Sep 2007 - Aug 2008) of
6 ambient surface-level O3 measurements was collected by Brodin et al. (2010, 663706) along an
7 elevation gradient ranging from 1608 m to 3528 m. The 7 sites used in this study are shown in
8 Figure 3-40 along with the RMNP site and the 15 Denver CSA sites. In fall, winter, and spring, they
9 observed a clear monotonic increase in O3 concentration with elevation, with a rate of increase in the
10 mean O3 concentration of 1.5 ppb per 100 m elevation gain during winter. In summer, the O3
11 gradient was similar in magnitude over the seven-site transect (1.3 ppb per 100 m), but much less
12 monotonic; the majority of the vertical gradient occurred between the lowest two sites (4.5 ppb per
13 100 m) and between the highest two sites (5.5 ppb per 100 m), with the middle five sites all having
14 approximately equal median O3 concentrations. Ozone concentrations at the lowest site in Boulder
March 2011
3-90
DRAFT - DO NOT CITE OR QUOTE
-------
1 were influenced by NO titration as evidenced by traffic-related diel cycles in O3 concentrations, but
2 the remaining six sites were located at elevation in more rural/remote settings and illustrate a
3 positive surface-level O3 elevation gradient similar to that seen in SMNP and typical of areas under
4 less direct influence of boundary layer pollution.
Greeley
10 km
F«Md
fen c.,,1 =<*""*« • Centennial
Figure 3-40. Terrain map showing the location of the AQS ozone monitoring site in Rocky
Mountain National Park, CO (black/green star) and the Denver CSA (red dots) along
with ozone monitoring sites used in the Brodin et al. (2010, 663706) (blue circles)
study. Elevations range from approximately 1600 m above sea level in Denver and
Boulder to 3528 m above sea level at the highest mountainous site.
5 The three sites in California-one in San Bernardino National Forest (SBNF) and two in
6 Sequoia National Park (SENP)-had the highest distribution of 8-h daily max O3 concentrations of
7 the selected rural focus area monitors included in Figure 3-37. The SBNF site had a warm-season 8-
8 h daily max O3 concentration mean of 80 ppb and a maximum of 137 ppb measured on July 1, 2007.
9 This site is located in Crestline, CA, 90 km down-wind of Los Angeles in the San Bernardino
March 2011
3-91
DRAFT - DO NOT CITE OR QUOTE
-------
1 Mountains. This site was included in the Los Angeles CSA shown in Figure 3-26 (Site AE) and had
2 the highest median 8-h daily max O3 concentration of any AQS site in the Los Angeles CSA during
3 this time period (Figure 3-29). This site was also included in an analysis performed for the 2006 O3
4 AQCD (U.S. EPA, 2006, 088089) where similarly high O3 concentrations were observed using
5 2004 year-round hourly observations.
6 The two sites in SENP are located 9.7 km apart at contrasting elevations as is illustrated in the
7 terrain map in Figure 3-41. The correlation in 8-h daily max O3 between these two sites was 0.86 and
8 the COD was 0.09, which are within the range in correlations and CODs for SMNP (Figure 3-39).
9 The distribution of 8-h daily max O3 concentrations at the lower elevation site (elevation = 560 m;
10 site ID = 061070009) is shifted slightly higher with a median of 76 ppb compared to the higher
11 elevation site (elevation = 1890 m; site ID = 061070006) with a median of 69 ppb. The lower
12 elevation site is located at the entrance to the park and is at a low enough elevation to be influenced
13 by boundary layer pollution coming upwind from Fresno and the San Joaquin Valley. The higher
14 elevation site is in the free troposphere above the planetary boundary layer and is less influenced by
15 such pollution. This gives rise to a negative average surface-level elevation gradient of -0.5 ppb per
16 100 m elevation gain in SENP, illustrating the location-specific complexities inherent to high-altitude
17 surface-level O3 concentrations.
18 Since O3 produced from emissions in urban areas is transported to more rural downwind
19 locations, elevated O3 concentrations can occur at considerable distances from urban centers. In
20 addition, major sources of O3 precursors such as highways, power plants, biomass combustion, and
21 oil and gas operations are commonly found in rural areas, adding to the O3 in these areas. Due to
22 lower chemical scavenging in nonurban areas, O3 tends to persist longer in rural than in urban areas
23 which tends to lead to higher cumulative exposures in rural areas influenced by anthropogenic
24 precursor emissions. The persistently high O3 concentrations observed at many of these rural sites
25 investigated here indicate that cumulative exposures for humans and vegetation in rural areas can be
26 substantial and often higher than cumulative exposures in urban areas.
27
March 2011 3-92 DRAFT - DO NOT CITE OR QUOTE
-------
'A
Figure 3-41. Terrain map showing the location of two AQS ozone monitoring sites (black/green
stars) in Sequoia National Park, CAJThe lower site (site ID = 061070009) is 560 m
above sea level and the higher site (site ID = 061070006) is 1890 m above sea level.
3.6.3. Temporal Variability
3.6.3.1. Multiyear Trends
1 Nationally, O3 concentrations have declined over the last decade, as shown in Figure 3-42
2 from the 2010 National Air Quality Status and Trends report (U.S. EPA, 2010, 647278). The
3 majority of this decline occurred before 2004 with national average concentrations remaining
4 relatively flat between 2004 and 2008. The large decreases in 2003 and 2004 coincides with NOX
5 emissions reductions resulting from implementation of the NOX State Implementation Plan (SIP)
6 Call rule, which began in 2003 and was fully implemented in 2004. This rule was designed to reduce
7 NOX emissions from power plants and other large combustion sources in the eastern U.S. The
8 reduction in NOX and O3 during the 2003-2004 timeframe is particularly evident in the eastern U.S.
9 where the NOX SIP Call was focused (U.S. EPA, 2010, 647278).
March 2011
3-93
DRAFT - DO NOT CITE OR QUOTE
-------
0.12
c
o
o
O
0.08
0.06
0.04
0.02
0
Average
1050 sites
90 percent of sites are below this line.
Current National Standard (revised 2008)
10 percent of sites are below this line.
01
02 03 04 05 06 07
2001 to 2008: 10% decrease
08
Source: U.S. EPA (2010, 6472781
Figure 3-42. National 8-h ozone trends, 2001-2008 (average of the annual fourth highest 8-h daily
max concentrations in ppm).
1 Weather can have a strong influence on O3 and O3 trends as well. The number of hot, dry days
2 can significantly alter the number of high-O3 days in any given year, even if O3-forming emissions
3 do not change. To better evaluate the progress and effectiveness of emissions reduction programs,
4 EPA uses a statistical model to estimate the influence of atypical weather on O3 formation
5 (U.S. EPA, 2010, 647278). After adjusting for the influence of weather, the trend in national 8 h O3
6 concentrations between 2001 and 2008 increases slightly from an 8% reduction to an 11% reduction.
7 These trends are region-specific, with lower reductions (3%) in California and higher reductions
8 (15%) in eastern states over this same time period (U.S. EPA, 2010, 647278).
9 Sites that showed the greatest reduction in O3 over this period were in or near the following
10 metropolitan areas: Anderson, IN; Chambersburg, PA; Chicago, IL; Cleveland, OH; Houston, TX;
11 Michigan City, IN; Milwaukee, WI; New York, NY; Racine, WI; Watertown, NY; and parts of
12 Los Angeles, CA. Sites that showed an increase in O3 over this time period and had measured
13 concentrations above the 2008 O3 standard1 during the 2006-2008 time period were located in or
14 near the following metropolitan areas: Atlanta, GA; Baton Rouge, LA; Birmingham, AL; Denver,
15 CO; El Centre, CA; San Diego, CA; Seattle, WA; and parts of Los Angeles, CA.
16 As noted in the 2006 O3 AQCD (U.S. EPA, 2006, 088089). trends in national parks and rural
17 areas are similar to nearby urban areas, reflecting the regional nature of O3 pollution. However,
18 caution should be exercised in using trends calculated at national parks to infer contributions from
1 On September 16, 2009, EPA announced it would reconsider the 2008 O3 NAAQS, which, at the time, included primary and secondary
standards of 0.075 ppm (8-h daily max).
March 2011
3-94
DRAFT - DO NOT CITE OR QUOTE
-------
1 distant sources either inside or outside of North America because of the influence of regional
2 pollution (issues relating to background O3 concentrations are discussed in Section 3.4). Trends in
3 tropospheric O3 on a global scale have been monitored around the world using ozonesondes, remote
4 surface monitors, mountain top monitors, and satellites. Global trends in the burden of tropospheric
5 O3 as they relate to climate change are discussed in Chapter 10, Section 10.2.3.1.
3.6.3.2. Hourly Variations
6 Ozone concentrations show a strong degree of diel variability resulting from daily patterns in
7 temperature, sunlight, and precursor emissions. Other factors, such as the relative importance of
8 transport versus local photochemical production and loss rates, the timing for entrainment of air from
9 the nocturnal residual boundary layer, and the diurnal variability in mixing layer height also play a
10 role in daily O3 patterns. The 2006 O3 AQCD (U.S. EPA, 2006, 088089) looked at composite urban
11 diel variations from April to October 2000 to 2004 and found 1-h maxima to occur in mid-afternoon
12 and 1-h minima to occur in early morning. On a national basis, however, there was a high degree of
13 spread in these times and caution was raised in extrapolating results from one city to another in
14 determining the time of day for O3 maxima and minima.
15 A similar analysis was performed using the 1-h avg O3 data from the 20 focus cities listed in
16 Table 3-9. The year-round data set described in Table 3-5 was used to compare diel patterns during
17 cold months (October - April) and warm months (May - September) between 2007 and 2009. The
18 warm-season data set, also described in Table 3-5, was used to compare weekday and weekend diel
19 patterns. Figure 3A-96 through 3A-115 show these patterns for each of the cities; examples for
20 Atlanta, Boston and Los Angeles are shown in Figure 3-43.
March 2011 3-95 DRAFT - DO NOT CITE OR QUOTE
-------
Cold Months
Warm Months
Weekdays
Weekends
II
100 -
*i 50 -
0 days. 0 year-round sites
— mean
median
( 1 5"-95™
no year-round monitors
0 days. 0 year-round sites
no year-round monitors
327 days, 11 warm-season sites
132 days, 11 warm-season sites
00:00 06:CO 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:CO 12:00 18:00 00:00
O
c
2
S 10D ~
O>
C O
< 50 -
0 -
637 days, 47 year-round sites
— mean
median
'. J 5"-95*
459 days, 47 year-round si:es
327 days, 50 warm-season sites
132 days, 50 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00
hour hour hour hour
Figure 3-43. Diel patterns in 1-h avg ozone for Atlanta, Boston and Los Angeles between 2007
and 2009 using the year-round data set for the cold month/warm month comparison
(left half) and the warm-season data set for the weekday/weekend comparison (right
half). [Atlanta had no year-round monitors available for the cold month/warm month
comparison.
1 In general, all the urban areas showed 1-h daily max concentrations occurring typically in the
2 early afternoon. In all cities, these afternoon peaks were more pronounced in the warm months than
3 in the cold months. However, a small peak was still present during the cold months. During warm
4 months, the difference between the median daily extrema varied considerably by city. For example,
5 in Los Angeles (Figure 3-43), the median 1-h daily min (10 ppb) at -5:00 a.m. was 50 ppb less than
6 the median 1-h daily max (60 ppb) at -2:00 p.m. By contrast, in Boston (Figure 3-43), the median
7 1-h daily min (13 ppb) occurred at the same time, but was only 25 ppb less than the median 1-h daily
8 max (38 ppb). Cities with large daily swings (>40 ppb) in median 1-h O3 concentrations included
9 Atlanta, Birmingham, Los Angeles, Phoenix, Pittsburgh, and Salt Lake City (Figure 3A-96, 3A-98,
10 3A-105, 3A-109, 3A-110, and 3A-111). Cities with small daily swings (<25 ppb) in median 1-h O3
11 concentrations included Boston, Minneapolis, San Francisco and Seattle (Figure 3A-99, 3A-106, 3A-
12 113, and 3A-114). These results are very similar to those found in the 2006 O3 AQCD (U.S. EPA,
March 2011
3-96
DRAFT - DO NOT CITE OR QUOTE
-------
1 2006, 088089) where many of these same urban areas were investigated. This supports the
2 conclusions drawn in the AQCD that diel patterns in O3 have remained stable over the last 20 years,
3 with times of occurrence of the daily maxima varying by no more than an hour from year to year.
4 Using the warm-season data, there was very little difference in the median diel profiles for
5 weekdays compared with weekends across all cities. This result stresses the complexity of O3
6 formation and the importance of meteorology, entrainment, biogenic precursor emissions, and
7 transport in addition to anthropogenic precursor emissions. There was, however, a subtle deviation
8 between weekdays and weekends in the lower percentiles (1st and 5th) of the distribution. The lower
9 end of the distribution tended to be lower on weekdays relative to weekends. This is consistent with
10 analyses in the 2006 O3 AQCD (U.S. EPA, 2006, 088089) and is a result of lower traffic volumes on
11 weekends relative to weekdays, leading to less NO emissions and O3 titration on the weekends.
12 Seasonal and site-to-site variations in diel patterns within a subset of the urban focus areas
13 presented here were investigated in the 2006 O3 AQCD (U.S. EPA, 2006, 088089). In northern cities,
14 there was substantial seasonal variability in the diel patterns with higher extreme values in the O3
15 distribution during the warm season than during the cold season. In southern cities, the seasonal
16 differences in extreme O3 concentrations were much smaller, and some of the highest O3
17 concentrations in the Houston CSA were found outside of summer. The general pattern that emerged
18 from investigating site-to-site variability within the urban areas was that peaks in 1-h avg O3
19 concentrations are higher and tend to occur later in the day at downwind sites relative to sites located
20 in the urban core. Differences between sites were not only related to the distance between them, but
21 also depend on the presence or absence of nearby O3 sources or sinks.
3.6.4. Associations with Co-pollutants
22 Correlations between O3 and other criteria pollutants are discussed in this section. Since O3 is
23 a secondary pollutant formed in the atmosphere from precursor emissions, it is not expected to be
24 highly correlated with primary pollutants such as CO and NOX. Furthermore, O3 formation is
25 strongly influenced by meteorology, entrainment, and transport of both O3 and O3 precursors,
26 resulting in a broad range in correlations with other pollutants which can vary substantially with
27 season.
28 To investigate correlations with co-pollutants, 8-h daily max O3 from the year-round and
29 warm-season data sets (Table 3-6 and Table 3-7) were compared with co-located 24-h avg CO, SO2,
30 NO2, PM2.5 and PM10 obtained from AQS for 2007-2009. Figure 3-44 and Figure 3-45 contain
31 co-pollutant box plots of the correlation between co-located monitors for the year-round data set and
32 the warm-season data set, respectively.
33 The year-round 8-h daily max O3 data (Figure 3-44) had a very wide range in correlations with
34 all the 24-h avg co-pollutants. A more clear pattern emerged when the data were stratified by season
35 (bottom four plots in Figure 3-44) with mostly negative correlations in the winter and mostly
36 positive correlations in the summer for all co-pollutants. In summer, the IQR in correlations is
37 positive for all co-pollutants. However, the median seasonal correlations are still modest at best with
March 2011 3-97 DRAFT - DO NOT CITE OR QUOTE
-------
1 the highest positive correlation at 0.52 for PM2 5 in the summer and the highest negative correlation
2 at -0.38 for PM2 5 in the winter. Spring and fall lie in between with spring having a slightly narrower
3 distribution than fall for all co-pollutants. The warm-season 8-h daily max O3 data (Figure 3-45)
4 shows a very similar distribution to the summer stratification of the year-round data due to their
5 overlap in time periods (May-Sept and Jun-Aug, respectively).
Year-Round
co-
scx-
NG,~
PM;6H
A
-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0,6 0.8 1.0
Correlation with 8-h daily max 0,
W inter
Spring
co-
so.-
NO -
PM,
PM,, •
•1.0 -0.8 -0.6 -04 -0.2 0.0 02 0,4 0,6 0.8 1.0 -10 -0.8 -0.6 -0.4 -02 0.0 0,2 04 06 08 1.0
Summer Fall
CO-
so.-
NO,-
PMn,
PM,.
-1.0 -0.8 -0.6 -0.4 -0 2 0.0 02 0.4 0.6 0.8 1.0 -1.0 -0.8 -0.8 -0.4 -02 0.0 0,2 04 06 08 1.0
Correlation with 8-h daily max 0, Correlation with 8-h daily max 0.,
Figure 3-44. Distribution of Pearson correlation coefficients for comparison of 8-h daily max
ozone from the year-round data set with co-located 24-h avg CO, S02, N02, PMi0 and
PM2.6 from AQS, 2007-2009[(top figure) with seasonal stratification (bottom four
figures). Shown are the median (red line), mean (green star), inner-quartile range
(box), 5th and 95th percentiles (whiskers) and extremes (black dots).
March 2011
3-98
DRAFT - DO NOT CITE OR QUOTE
-------
co-
N02H
PM
10
PM
2,5
-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4
Correlation with 8-h daily max 0
0.6
1.0
Figure 3-45. Distribution of Pearson correlation coefficients for comparison of 8-h daily max
ozone from the warm-season (May-Sept) data set with co-located 24-h avg CO, S02,
N02, PMio and PM2.s from AQS, 2007-2009. Shown are the median (red line), mean
(green star), inner-quartile range (box), 5th and 95th percentiles (whiskers), and
extremes (black dots).
1 The seasonal fluctuations in correlations present in Figure 3-44 result in part from the mixture
2 of primary and secondary sources for the co-pollutants. For example, O3 is a secondary pollutant
3 whereas PM2 5 has both primary and secondary origins and these two pollutants show the largest
4 summertime/wintertime swing in correlation distributions. This situation arises because the
5 secondary component to PM2 5 is larger during the summer and is formed in conditions conducive to
6 secondary O3 formation. The result is positive correlations between O3 and PM2 5 during the summer.
7 During the winter, photochemical production of O3 is much smaller than during summer and O3
8 comes mainly from aloft, i.e., the free troposphere (see Section 3.4 for further details). In addition,
9 concentrations of PM2 5 are much lower aloft. On relatively clean days, this can lead to high
10 concentrations of O3 and lower concentrations of primary pollutants such as PM2 5 or NO. On
11 relatively dirty days with elevated NO and PM2 5, the intruding O3 is readily titrated by NO in the
12 boundary layer. These processes result in negative correlations between O3 and PM2 5 during the
13 winter.
March 2011
3-99
DRAFT - DO NOT CITE OR QUOTE
-------
3.7. Chapter 3 References
A list of all references considered for inclusion in this chapter can be found at
http://hero.epa. gov/index.cfm?action=search.do&submit=Search&group_id=382
Acker, K.; Febo, A.; Trick, S.; Perrino, C.; Bruno, P.; Wiesen, P.; Moller; Wieprecht, W.; Auel, R.; Giusto, M.; Geyer, A.;
Platt, U.; Allegrini, I. (2006). Nitrous acid in the urban area of Rome. Atmos Environ, 40: 3123-3133.
http://dx.doi.0rg/10.1016/i.atmosenv.2006.01.028 595095
Ambient air monitoring reference and equivalent methods. 40 CFR A§ 53. (1976). 041090
Ambient air monitoring reference and equivalent methods. 40 Fed. Reg. 7042-7070 (1975). 043954
Ambient Air Monitoring Reference and Equivalent Methods: Designation of One New Equivalent Method (EPA Office of
Research and Development). 75 Fed. Reg. 22126-22127 (2010). 687659
Andreae, M. O. (1991). Biomass burning: its history, use, and distribution and its impact on environmental quality and
global climate. In Levine JS (Ed.), Global Biomass Burning: Atmospheric, Climatic, and Biospheric Implications
(pp. 1-21). Cambridge, MA: MIT Press. 078147
Anton, M.; Lopez, M.; Vilaplana, J. M.; Kroon, M.; McPeters, R.; Banon, M.; Serrano, A. (2009). Validation of OMI-
TOMS and OMI-DOAS total ozone column using five Brewer spectroradiometers at the Iberian peninsula. J
Geophys Res, 114: D14307. http://dx.doi.org/10.1029/2009JD012003 595098
Appel, K. W.; Gilliland, A.; Eder, B. (2005). An operational evaluation of the 2005 release of models-3 CMAQ version 45.
Washington DC: National Oceanic and Atmospheric Administration-Air Resources. 089227
Arnold, J. R.; Dennis, R. L.; Tonnesen, G S. (2003). Diagnostic evaluation of numerical air quality models with specialized
ambient observations: testing the Community Multiscale Air Quality modeling system (CMAQ) at selected SOS 95
ground sites. Atmos Environ, 37: 1185-1198. 087579
Arshinov, M. Y; Belan, B. D.; Krasnov, O. A.; Kovalevskii, V. K.; Pirogov, V. A.; Plotnikov, A. P.; Tolmachev, G. N.;
Fofonov, A. V. (2002). Comparison of ultraviolet and chemiluminescent ozonometers. Atmos. Ocean , 15: 656-658.
080718
ATMET (2011). Atmospheric, meteorological, and environmental technologies. Retrieved January 28, 2011 from
http://atmet.com/. 677541
Barrie, L. A.; Bottenheim, J. W.; Schnell, R. C.; Crutzen, P. J.; Rasmussen, R. A. (1988). Ozone destruction and
photochemical reactions at polar sunrise in the lower Arctic atmosphere. Nature, 334: 138-141. 053377
Beckerman, B.; Jerrett, M.; Brook, J. R.; Verma, D. K.; Arain, M. A.; Finkelstein, M. M. (2008). Correlation of nitrogen
dioxide with other traffic pollutants near a major expressway. Atmos Environ, 42: 275-290. 096484
Beer, R. (2006). TES on the aura mission: Scientific objectives, measurements, and analysis overview. IEEE Trans Geosci
Remote Sens, 44: 1102-1105. http://dx.doi.org/10.1109/TGRS.2005.863716 633893
Berkowitz, C. M.; Fast, J. D.; Sprinston, S. R.; Larsen, R. J.; Spicer, C. W.; Doskey, P. V; Hubbe, J. M.; Plastridge, R.
(1998). Formation mechanisms and chemical characteristics of elevated photochemical layers over the northeast
United States. J Geophys Res, 103: 10,631-10,647. 081467
Berkowitz, C. M.; Shaw, W. J. (1997). Airborne measurements of boundary layer chemistry during the Southern Oxidant
Study: a case study. J Geophys Res, 102: 12,795-12,804. 047593
Binkowski, F. S.; Arunachalam, S.; Adelman, Z.; Pinto, J. P. (2007). Examining photolysis rates with a prototype online
photolysis module in CMAQ. J Appl Meteor Climatol, 46: 1252-1256. 090563
Binkowski, F.; Roselle, S. (2003). Models-3 Community Multiscale Air Quality(CMAQ) model aerosol component 1.
Model description. J Geophys Res, 108: 4183. http://dx.doi.org/10.1029/2001JD001409 191769
Bishop, G. A.; Stedman, D. H. (2008). A decade of on-road emissions measurements. Environ Sci Technol, 42: 1651-1656.
http://dx.doi.org/10.1021/es702413bl94670
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments such as the Integrated Science Assessments (ISAs) and the Integrated Risk Information System (IRIS).
March 2011 3-100 DRAFT - DO NOT CITE OR QUOTE
-------
Bloomer, B. I; Stehr, J. W.; Piety, C. A.; Salawitch, R. I; Dickerson, R. R. (2009). Observed relationships of ozone air
pollution with temperature and emissions. Geophys Res Lett, 36: L09803. http://dx.doi.org/10.1029/2009GL037308
628855
Blumenthal, D. L.; Lurmann, F. W.; Kumar, N.; Dye, T. S.; Ray, S. E.; Korc, M. E.; Londergan, R.; Moore, G (1997).
Transport and mixing phenomena related to ozone exceedances in the northeast US (analysis based on NARSTO-
northeast data). Santa Rosa, CA: Sonoma Technology. 052278
Bonn, B.; Von Kuhlmann, R.; Lawrence, M. G. (2004). High contribution of biogenic hydroperoxides to secondary organic
aerosol formation. Geophys Res Lett, 31: L10108. http://dx.doi.org/10.1029/2003GL019172 053770
Brodin, M.; Helmig, D.; Oltmans, S. (2010). Seasonal ozone behavior along an elevation gradient in the Colorado Front
Range Mountains. Atmos Environ, 44: 5305-5315. http://dx.doi.Org/10.1016/j.atmosenv.2010.06.033 663706
Burgard, D. A.; Bishop, G. A.; Stedman, D. H.; Gessner, V. H.; Daeschlein, C. (2006). Remote sensing of in-use heavy-duty
diesel trucks. Environ Sci Technol, 40: 6938-6942. 193222
Burley, J. D.; Ray, J. D. (2007). Surface ozone in Yosemite National Park. Atmos Environ, 41: 6048-6062. 149069
Buzica, D.; Gerboles, M.; Plaisance, H. (2008). The equivalence of diffusive samplers to reference methods for monitoring
O3, benzene and NO2 in ambient air. J Environ Monit, 10: 1052-1059. 595173
Byun, D. W.; Ching, J. K. S. (1999). Science algorithms of the EPAmodels-3 community multiscale air quality (CMAQ)
modeling system (Report No. EPA/600-R-99-030). Washington, DC: U.S. Environmental Protection Agency, Office
of Research and Development. http://www.epa.gov/asmdnerl/CMAQ/CMAQscienceDoc.html. 156314
Byun, D.; Schere, K. L. (2006). Review of the governing equations, computational algorithms, and other components of the
models-3 community multiscale air quality (CMAQ) modeling system. Appl. Mech. Rev., 59: 51-77. 090560
Carter, W. P. L. (1995). Computer modeling of environmental chamber studies of maximum incremental reactivities of
volatile organic compounds. Atmos Environ, 29: 2513. 052288
CEMPD (2011). SMOKE. Retrieved January 28, 2011 from http://www.smoke-model.org/index.cfm. 677545
Chan, E.; Vet, R. J. (2010). Baseline levels and trends of ground level ozone in Canada and the United States. Atmos Chem
Phys, 10: 8629-8647. http://dx.doi.org/10.5194/acp-10-8629-2010679710
Ching, J.; Herwehe, J.; Swall, J. (2006). On joint deterministic grid modeling and sub-grid variability conceptual
framework for model evaluation. Atmos Environ, 40: 4935-4945. 090300
Civerolo, K. L.; Mao, H. T; Rao, S. T. (2003). The airshed for ozone and fine particulate pollution in the eastern United
States. Pure Appl Geophys, 160: 81-105. 053985
Conrad, R.; Seiler, W. (1985). Influence of temperature, moisture, and organic carbon on the flux of H2 and CO between
soil and atmosphere: field studies in subtropical regions. J Geophys Res, 90: 5699-5709. 029520
Cooper, O. R.; Parrish, D. D.; Stohl, A.; Trainer, M.; Nedelec, P.; Thouret, V; Cammas, J. P.; Oltmans, S. J.; Johnson, B. J.;
Tarasick, D.; Leblanc, T; McDermid, I. S.; Jaffe, D.; Gao, R.; Stith, J.; Ryerson, T; Aikin, K.; Campos, T;
Weinheimer, A.; Avery, M. A. (2010). Increasing springtime ozone mixing ratios in the free troposphere over
western North America. Nature, 463: 344-348. http://dx.doi.org/10.1038/nature08708 380093
Corsmeier, U.; Kalthhoff, N.; Kolle, O.; Motzian, M.; Fiedler, F. (1997). Ozone concentration jump in the stable nocturnal
boundary layer during a LLJ-event. Atmos Environ, 31: 1977-1989. 047620
DAnna, B.; Jammoul, A.; George, C.; Stemmler, K.; Fahrni, S.; Ammann, M.; Wisthaler, A. (2009). Light-induced ozone
depletion by humic acid films and submicron aerosol particles. J Geophys Res, 114: D12301.
http://dx.doi.org/10.1029/2008JD011237628847
Dallmann, T. R.; Harley, R. A. (2010). Evaluation of mobile source emission trends in the United States. J Geophys Res,
115: D14305. http://dx.doi.org/10.1029/2010JD013862 665390
Dickerson, R. R.; Rhoads, K. P.; Carsey, T. P.; Oltmans, S. J.; Burrows, J. P.; Crutzen, P. J. (1999). Ozone in the remote
marine boundary layer: a possible role for halogens. J Geophys Res, 104: 21,385-21,395. 053394
Docherty, K. S.; Wu, W.; Lim, Y. B.; Ziemann, P. J. (2005). Contributions of organic peroxides to secondary aerosol formed
from reactions of monoterpenes with O3. Environ Sci Technol, 39: 4049-4059. http://dx.doi.org/10.1021/es050228s
087613
March 2011 3-101 DRAFT - DO NOT CITE OR QUOTE
-------
Doyle, M.; Sexton, K. G; Jeffries, H.; Bridge, K.; Jaspers, I. (2004). Effects of 1,3-butadiene, isoprene, and their
photochemical degradation products on human lung cells. Environ Health Perspect, 112: 1488-1495. 088404
Doyle, M.; Sexton, K. G; Jeffries, H.; Jaspers, I. (2007). Atmospheric photochemical transformations enhance 1,3-
butadiene-induced inflammatory responses in human epithelial cells: The role of ozone and other photochemical
degradation products. Chem Biol Interact, 166: 163-169. http://dx.doi.Org/10.1016/i.cbi.2006.05.016596377
Dunker, A. M. (1981). Efficient calculation of sensitivity coefficients for complex atmospheric models. Atmos Environ, 15:
1155-1161.040504
Dunker, A.; Yarwood, G; Ortmann, J.; Wilson, G. (2002). The decoupled direct method for sensitivity analysis in a three-
dimensional air quality model implementation, accuracy, and efficiency. Environ Sci Technol, 36: 2965-2976.
http://dx.doi.org/10.1021/es0112691 665029
Dunlea, E. J.; Herndon, S. C.; Nelson, D. D.; Volkamer, R. M.; Lamb, B. K.; Allwine, E. J.; Grutter, M.; Ramos Villegas, C.
R.; Marquez, C.; Blanco, S.; Cardenas, B.; Kolb, C. E.; Molina, L. T; Molina, M. J. (2006). Technical note:
Evaluation of standard ultraviolet absorption ozone monitors in a polluted urban environment. Atmos Chem Phys
Discuss, 6: 2241-2279. www.atmos-chem-phys.net/6/3163/2006/. 595108
Ebeling, D.; Patel, V; Findlay, M.; Stetter, J. (2009). Electrochemical ozone sensor and instrument with characterization of
the electrode and gas flow effects. Sens Actuators B, 137: 129-133. http://dx.doi.Org/10.1016/j.snb.2008.10.038
595115
Eder, B.; Yu, S. (2005). A performance evaluation of the 2004 release of Models-3 CMAQ. Atmos Environ, 40: 4811-4824.
089229
Eisele, F. L.; Mount, G. H.; Tanner, D.; Jefferson, A.; Shelter, R.; Harder, J. W.; Williams, E. J. (1997). Understanding the
production and intercon version of the hydroxyl radical during the tropospheric OH photochemistry experiment. J
Geophys Res, 102: 6457-6465. 057210
Emmerson, K. M.; Evans, M. J. (2009). Comparison of tropospheric gas-phase chemistry schemes for use within global
models. Atmos Chem Phys, 9: 1831-1845. http://dx.doi.org/10.5194/acpd-8-19957-2008 605119
ENVIRON (2005). CAMx. Retrieved January 28, 2011 from http://www.camx.com/over/. 677542
Fang, Y; Fiore, A. M.; Horowitz, L. W.; Levy II, H.; Hu, Y; Russell, A. G. (2010). Sensitivity of the NOy budget over the
United States to anthropogenic and lightning NOx in summer. J Geophys Res, 115: D18312.
http://dx.doi.org/10.1029/2010JD014079665391
Fehsenfeld, F. C.; Ancellet, G; Bates, T. S.; Goldstein, A. H.; Hardesty, R. M.; Honrath, R.; Law, K. S.; Lewis, A. C.;
Leaitch, R.; McKeen, S.; Meagher, J.; Parrish, D. D.; Pszenny, A. A. P.; Russell, P. B.; Schlager, H.; Seinfeld, J.;
Talbot, R.; Zbinden, R. (2006). International consortium for atmospheric research on transport and transformation
(ICARTT): North America to Europe -overview of the 2004 summer field study. J Geophys Res, 111: D23S01.1-
D23S01.36. http://dx.doi.org/10.1029/2006JD007829 190531
Fehsenfeld, F. C.; Trainer, M.; Parrish, D. D.; Volz-Thomas, A.; Penkett, S. (1996). North Atlantic Regional Experiment
(NARE) 1993 summer intensive: foreword. J Geophys Res, 101: 28,869-28,875. 047803
Finlayson-Pitts, B. J.; Pitts, J. N. Jr (1986). Atmospheric chemistry: Fundamentals and experimental techniques. New York,
NY: John Wiley & Sons. 035054
Fiore, A.; Dentener, F.; Wild, O.; Cuvelier, C.; Schultz, M.; Hess, P.; Textor, C.; Schulz, M.; Doherty, R.; Horowitz, L.;
MacKenzie, L; Sanderson, M.; Shindell, D.; Stevenson, D.; Szopa, S.; Van Dingenen, R.; Zeng, G; Atherton, C.;
Bergmann, D.; Bey, L; Carmichael, G; Collins, W.; Duncan, B.; Faluvegi, G; Folberth, G; Gauss, M.; Gong, S.;
Hauglustaine, D.; Holloway, T; Isaksen, L; Jacob, D.; Jonson, J.; Kaminski, J.; Keating, T; Lupu, A.; Marmer, E.;
Montanaro, V; Park, R.; Pitari, G; Pringle, K.; Pyle, J.; Schroeder, S.; Vivanco, M.; Wind, P.; Wojcik, G; Wu, S.;
Zuber, A. (2009). Multimodel estimates of intercontinental source-receptor relationships for ozone pollution. J
Geophys Res, 114: D04301. http://dx.doi.org/10.1029/2008JD010816 665030
Fiore, A.; Jacob, D. J.; Liu, H.; Yantosca, R. M.; Fairlie, T. D.; Li, Q. (2003). Variability in surface ozone background over
the United States: Implications for air quality policy. J Geophys Res, 108: 4787.
http://dx.doi.org/10.1029/2003JD003855051226
Fuentes, J. D.; Wang, D.; Bowling, D. R.; Potosnak, M.; Monson, R. K.; Goliff, W. S.; Stockwell, W. R. (2007). Biogenic
hydrocarbon chemistry within and above a mixed deciduous forest. J Atmos Chem, 56: 165-185.
http://dx.doi.org/10.1007/sl0874-006-9048-4 191251
March 2011 3-102 DRAFT - DO NOT CITE OR QUOTE
-------
Fuentes, M.; Raftery, A. E. (2005). Model evaluation and spatial interpolation by Bayesian combination of observations
with outputs from numerical models. Biometrics, 61: 36-45. 087580
Fusco, A. C.; Logan, J. A. (2003). Analysis of 1970-1995 trends in tropospheric ozone at Northern Hemisphere
midlatitudes with the GEOS-CHEM model. J Geophys Res, 108: 4449. http://dx.doi.org/10.1029/2002JD002742
051229
Gaydos, T. M.; Finder, R.; Koo, B.; Fahey, K. M.; Yarwood, G; Pandis, S. N. (2007). Development and application of a
three-dimensional aerosol chemical transport model, PMCAMx. Atmos Environ, 41: 2594-2611. 139738
Generoso, S.; Bey, I.; Attie, J.-L.; Breon, F.-M. (2007). A satellite- and model-based assessment of the 2003 Russian fires:
impact on the arctic region. J Geophys Res, 112: 5302. 155786
Gilliland, A. B.; Hogrefe, C.; Finder, R. W; Godowitch, J. M.; Foley, K. L.; Rao, S. T. (2008). Dynamic evaluation of
regional air quality models: Assessing changes in O3 stemming from changes in emissions and meteorology. Atmos
Environ, 42: 5110-5123. 606585
Godowitch, J. M.; Gilliland, A. B.; Draxler, R. R.; Rao, S. T. (2008). Modeling assessment of point source NOx emission
reductions on ozone air quality in the eastern United States. Atmos Environ, 42: 87-100. 139006
Goldstein, A. H.; Millet, D. B.; McKay, M.; Jaegle, L.; Horowitz, L.; Cooper, O.; Hudman, R.; Jacob, D. J.; Oltmans, S.;
Clarke, A. (2004). Impact of Asian emissions on observations at Trinidad Head, California, during ITCT 2K2. J
Geophys Res, 109: D23S17. http://dx.doi.org/10.1029/2003JD004406 087780
Goldstein, A.; Galbally, I. (2007). Known and unexplored organic constituents in the earth's atmosphere. Environ Sci
Technol, 41: 1514-1521. 193247
Gottardini, E.; Cristofori, A.; Cristofolini, F.; Ferretti, M. (2010). Variability of ozone concentration in a montane
environment, northern Italy. Atmos Environ, 44: 147-152. http://dx.doi.Org/10.1016/j.atmosenv.2009.10.017 562911
Gouw Jd; Brock, C.; Atlas, E.; Bates, T.; Fehsenfeld, F.; Goldan, P.; Holloway, J.; Kuster, W.; Lerner, B.; Matthew, B.;
Middlebrook, A.; Onasch, T.; Peltier, R.; Quinn, P.; Senff, C.; Stohl, A.; Sullivan, A.; Trainer, M.; Warneke, C.;
Weber, R.; Williams E (2008). Sources of particulate matter in the northeastern United States in summer: 1. Direct
emissions and secondary formation of organic matter in urban plumes. J Geophys Res, 113: D08301. 191757
Greenberg, J. P.; Guenther, A. B.; Turnipseed, A. (2009). Tethered balloon-based soundings of ozone, aerosols, and solar
radiation near Mexico City during MIRAGE-MEX. Atmos Environ, 43: 2672-2677.
http://dx.doi.0rg/10.1016/i.atmosenv.2009.02.019 595140
Grell, G. A.; Emeis, S.; Stockwell, W. R.; Schoenemeyer, T; Forkel, R.; Michalakes, J.; Knoche, R.; Seidl, W. (2000).
Application of a multiscale, coupled MM5/chemistry model to the complex terrain of the VOTALP valley
campaign. Atmos Environ, 34: 1435-1453. 048047
Guenther, A.; Geron, C.; Pierce, T; Lamb, B.; Harley, P.; Fall, R. (2000). Natural emissions of non-methane volatile
organic compounds, carbon monoxide, and oxides of nitrogen from North America. Atmos Environ, 34: 2205-2230.
025002
Guenther, A.; Karl, T; Harley, P.; Wiedinmyer, C.; Palmer, P. L; Geron, C. (2006). Estimates of global terrestrial isoprene
emissions using MEGAN (Model of Emissions of Gases and Aerosols from Nature). Atmos Chem Phys, 6: 3181-
3210. http://dx.doi.org/10.5194/acp-6-3181-2006 607080
Hains, J. C.; Taubman, B. F.; Thompson, A. M.; Stehr, J. W.; Marufu, L. T; Doddridge, B. G; Dickerson, R. R. (2008).
Origins of chemical pollution derived from Mid-Atlantic aircraft profiles using a clustering technique. Atmos
Environ, 42: 1727-1741. 137411
Hameed, S.; Pinto, J. P.; Stewart, R. W. (1979). Sensitivity of the predicted CO-OH-CH4 perturbation to tropospheric NOx
concentrations. J Geophys Res, 84: 763-768. 029458
Harley, R. A.; Marr, L. C.; Lehner, J. K.; Giddings, S. N. (2005). Changes in motor vehicle emissions on diurnal to decadal
time scales and effects on atmospheric composition. Environ Sci Technol, 39: 5356-5362. 088154
Harvard University (2010). GEOS-Chem Model. Retrieved January 28, 2011 from http://acmg.seas.harvard.edu/geos/.
677546
Harvard University (2010). GEOS-Chem Overview. Retrieved January 28, 2011 from
http://acmg.seas.harvard.edu/geos/geos_overview.html. 677581
March 2011 3-103 DRAFT - DO NOT CITE OR QUOTE
-------
Henderson, B. H.; Finder, R. W.; Crooks, I; Cohen, R. C.; Hutzell, W. T.; Sarwar, G; Goliff, W. S.; Stockwell, W. R.; Fahr,
A.; Mathur, R.; Carlton, A. G; Vizuete, W. (2010). Evaluation of simulated photochemical partitioning of oxidized
nitrogen in the upper troposphere. Atmos Chem Phys, 10: 20125-20165. http://dx.doi.org/10.5194/acpd-10-20125-
2010674771
Hocking, W. K.; Carey-Smith, T.; Tarasick, D. W.; Argall, P. S.; Strong, K.; Rochon, Y; Zawadzki, I.; Taylor, P. A. (2007).
Detection of stratospheric ozone intrusions by windprofiler radars. Nature, 450: 281-284.
http://dx.doi.org/10.1038/nature06312608032
Hsu, J.; Prather, M. J. (2009). Stratospheric variability and tropospheric ozone. J Geophys Res, 114: D06102.
http://dx.doi.org/10.1029/2008JD010942629687
Hudman, R. C.; Murray, L. T.; Jacob, D. J.; Millet, D. B.; Turquety, S.; Wu, S.; Blake, D. R.; Goldstein, A. H.; Holloway,
J.; Sachse, G. W. (2008). Biogenic versus anthropogenic sources of CO in the United States. Geophys Res Lett, 35:
L04801. http://dx.doi.org/10.1029/2007gl032393 191253
Husar, R. B.; Renard, W. P. (1998). Ozone as a function of local wind speed and direction: Evidence of local and regional
transport. Paper presented at 91st annual meeting and exhibition of the Air & Waste Management Association, San
Diego, CA. 052413
Inman, R. E.; Ingersoll, R. B.; Levy, E. A. (1971). Soil: a natural sink for carbon monoxide. Science, 172: 1229-1231.
http://dx.doi.org/10.1126/science.172.3989.1229010972
Jacob, D. J. (1999). Introduction to atmospheric chemistry. New Jersey: Princeton University Press. 091122
Jacob, D. J.; Horowitz, L. W.; Munger, J. W; Heikes, B. G; Dickerson, R. R.; Artz, R. S.; Keene, W. C. (1995). Seasonal
transition from NOx- to hydrocarbon-limited conditions for ozone production over the eastern United States in
September. J Geophys Res, 100: 9315-9324. 052308
Jacobson, M. Z. (2002). Atmospheric pollution: history, science, and regulation. New York: Cambridge University Press.
090667
Jacobson, M. Z. (2005). Fundamentals of Atmospheric Modeling. New York: Cambridge University Press. 684174
Jaegle, L.; Jacob, D. J.; Brune, W. H.; Wennberg, P. O. (2001). Chemistry of HOx radicals in the upper troposphere. Atmos
Environ, 35: 469-489. 057250
James, P.; Stohl, A.; Forster, C.; Eckhardt, S.; Seibert, P.; Frank, A. (2003). A 15-year climatology of stratosphere-
troposphere exchange with a Lagrangian particle dispersion model: 2. Mean climate and seasonal variability. J
Geophys Res, 108: D12. http://dx.doi.org/10.1029/2002JD002639 043286
Jimenez, J. L.; Jayne, J. T.; Shi, Q.; Kolb, C. E.; Worsnop, D. R.; Yourshaw, L; Seinfeld, J. H.; Flagan, R. C.; Zhang, X.;
Smith, K. A. (2003). Ambient aerosol sampling using the Aerodyne Aerosol Mass Spectrometer. J Geophys Res,
108: 8425. 156611
Jo, W. K.; Park, J. H. (2005). Characteristics of roadside air pollution in Korean metropolitan city (Daegu) over last 5 to 6
years: Temporal variations, standard exceedances, and dependence on meteorological conditions. Chemosphere, 59:
1557-1573. http://dx.doi.0rg/10.1016/i.chemosphere.2004.12.021 674762
Johnson, D.; Jenkin, M. E.; Wirtz, K.; Martin-Riviejo, M. (2004). Simulating the formation of secondary organic aerosol
from the photooxidation of toluene. Environ Chem, 1: 150-165. 087659
Johnson, T. R. (1995). Recent advances in the estimation of population exposure to mobile source pollutants. J Expo Sci
Environ Epidemiol, 5: 551-571. 079215
Kasibhatla, P.; Chameides, W. L. (2000). Seasonal modeling of regional ozone pollution in the eastern United States.
Geophys Res Lett, 27: 1415-1418. 052237
Kaynak, B.; Hu, Y; Martin, R. V; Russell, A. G; Choi, Y; Wang, Y. (2008). The effect of lightning NOx production on
surface ozone in the continental United States. Atmos Chem Phys, 8: 5151-5159. 486686
King, G. M. (1999). Characteristics and significance of atmospheric carbon monoxide consumption by soils. Chemosphere,
1:53-63.002828
Kleffmann, J.; Lorzer, J. C.; Wiesen, P.; Kern, C.; Trick, S.; Volkamer, R.; Rodenas, M.; Wirtz, K. (2006). Intercomparison
of the DOAS and LOPAP techniques for the detection of nitrous acid (HONO). Atmos Environ, 40: 3640-3652.
130481
March 2011 3-104 DRAFT - DO NOT CITE OR QUOTE
-------
Kleffmann, I; Wiesen, P. (2008). Technical note: Quantification of interferences of wet chemical HONO LOPAP
measurements under simulated polar conditions. Atmos Chem Phys, 8: 6813-6822. www.atmos-chem-
phys.net/8/6813/20087. 488342
Kleindienst, T. E.; Hudgens, E. E.; Smith, D. R; McElroy, R R; Bufalini, J. J. (1993). Comparison of chemiluminescence
and ultraviolet ozone monitor responses in the presence of humidity and photochemical pollutants. Air and Waste,
43: 213-222. http://secure.awma.org/OnlineLibrary/ProductDetails.aspx?productID=11215. 043956
Kleinman, M.; Sioutas, C.; Stram, D.; Froines, J.; Cho, A.; Chakrabarti, B.; Hamade, A.; Meacher, D.; Oldham, M. (2005).
Inhalation of concentrated ambient particulate matter near a heavily trafficked road stimulates antigen-induced
airway responses in mice. J Air Waste Manag Assoc, 55: 1277-1288. 087880
Lam, Y.; Fu, J. (2010). Corrigendum to" A novel downscaling technique for the linkage of global and regional air quality
modeling" published in Atmos. Chem. Phys., 9, 9169-9185, 2009. Atmos Chem Phys, 10: 4013-4031.
http://dx.doi.org/10.5194/acp-10-4013-2010665031
Langford, A. O.; Aikin, K. C.; Eubank, C. S.; Williams, E. J. (2009). Stratospheric contribution to high surface ozone in
Colorado during springtime. Geophys Res Lett, 36: L12801. http://dx.doi.org/10.1029/2009gl038367 491703
Lee, J.; Kim, K. H.; Kim, Y. J.; Lee, J. (2008). Application of a long-path differential optical absorption spectrometer (LP-
DOAS) on the measurements of NO(2), SO(2), O(3), and HNO(2) in Gwangju, Korea. J Environ Manage, 86: 750-
759. 595175
Lefohn, A. S.; Oltmans, S. J.; Dann, T.; Singh, H. B. (2001). Present-day variability of background ozone in the lower
troposphere. J Geophys Res, 106: 9945-9958. 016253
Leston, A. R.; Ollinson, W. M.; Spicer, C. W; Satola, J. (2005). Potential interference bias in ozone standard compliance
monitoring. J Air Waste Manag Assoc, 55: 1464-1472. http://cat.inist.fr/?aModele=afiicheN&cpsidt=17175252.
080717
Li, Y; Lee, S.-R.; Wu, C.-Y (2006). UV-absorption-based measurements of ozone and mercury: An investigation on their
mutual interferences. Aerosol Air Qual Res, 6: 418-429. 633892
Liu, X. H.; Hegg, D. A.; Stoelinga, M. T. (2001). Numerical simulation of new particle formation over the northwest
Atlantic using the MM5 mesoscale model coupled with sulfur chemistry. J Geophys Res, 106: 9697-9715. 048201
Liu, X.; Chance, K.; Sioris, C. E.; Kurosu, T. P.; Spurr, R. J. D.; Martin, R. V; Fu, T.-M.; Logan, J. A.; Jacob, D. J.; Palmer,
P. I.; Newchurch, M. J.; Megretskaia, I. A.; Chatfield, R. B. (2006). First directly retrieved global distribution of
tropospheric column ozone from GOME: Comparison with the GEOS-CHEM model. J Geophys Res, 111: D02308.
http://dx.doi.org/10.1029/2005JD006564093013
Lu, R.; Turco, R. P.; Jacobson, M. Z. (1997). An integrated air pollution modeling system for urban and regional scales: 1
Structure and performance. J Geophys Res, 102: 6063-6079. 048202
Luecken, D. J.; Phillips, S.; Sarwar, G; Jang, C. (2008). Effects of using the CB05 vs. SAPRC99 vs. CB4 chemical
mechanism on model predictions: Ozone and gas-phase photochemical precursor concentrations. Atmos Environ,
42: 5805-5820. 190084
Mahajan, A. S.; Shaw, M.; Oetjen, H.; Hornsby, K. E.; Carpenter, L. J.; Kaleschke, L.; Tian-Kunze, X.; Lee, J. D.; Moller,
S. J.; Edwards, P. (2010). Evidence of reactive iodine chemistry in the Arctic boundary layer. J Geophys Res, 115:
D20303. http://dx.doi.org/10.1029/2009JD013665 665392
Maruo, Y. Y. (2007). Measurement of ambient ozone using newly developed porous glass sensor. Sens Actuators B, 126:
485-491. http://dx.doi.0rg/10.1016/j.snb.2007.03.041 595102
Maruo, Y. Y; Akaoka, K.; Nakamura, J. (2010). Development and performance evaluation of ozone detection paper using
azo dye orange I: Effect of pH. Sens Actuators B, 143: 487-493. http://dx.doi.Org/10.1016/j.snb.2009.09.042
595104
Mathur, R. (2008). Estimating the impact of the 2004 Alaskan forest fires on episodic particulate matter pollution over the
eastern United States through assimilation of satellite-derived aerosol optical depths in a regional air quality model.
J Geophys Res, 113: D17302. http://dx.doi.org/10.1029/2007JD009767 156742
McElroy, M. B.; Salawitch, R. J.; Wofsy, S. C.; Logan, J. A. (1986). Reductions of Antarctic ozone due to synergistic
interactions of chlorine and bromine. Nature, 321: 759-762. 019501
March 2011 3-105 DRAFT - DO NOT CITE OR QUOTE
-------
MCM (2010). The master chemical mechanism. Retrieved January 28, 2011 from http://mcm.leeds.ac.uk/MCM/home.htt.
677544
Milford, J. B.; Gao, D.; Sillman, S.; Blossey, P.; Russell, A. G. (1994). Total reactive nitrogen (NOy) as an indicator of the
sensitivity of ozone to reductions in hydrocarbon and NOx emissions. J Geophys Res, 99: 3533-3542. 038669
Miwa, T; Maruo, Y. Y; Akaoka, K.; Kunioka, T; Nakamura, J. (2009). Development of colorimetric ozone detection
papers with high ultraviolet resistance using ultraviolet absorbers. J Air Waste Manag Assoc, 59: 801-808.
http://dx.doi.0rg/10.3155/1047-3289.59.7.801 595110
Mollner, A. K.; Valluvadasan, S.; Feng, L.; Sprague, M. K.; Okumura, M.; Milligan, D. B.; Bloss, W. J.; Sander, S. P.;
Martien, P. T; Harley, R. A. (2010). Rate of gas phase association of hydroxyl radical and nitrogen dioxide.
Science, 330: 646-649. http://dx.doi.org/10.1126/science.1193030665393
NOAA (2010). The Rapid Update Cycle (RUC). Retrieved January 28, 2011 from http://ruc.noaa.gov/. 677543
Nolte, C. G; Gilliland, A. M.; Hogrefe, C.; Mickley, L. J. (2008). Linking global to regional models to assess future climate
impacts on surface ozone levels in the United States. J Geophys Res, 113: D14307.
http://dx.doi.org/10.1029/2007JD008497 679712
NRC (1991). Rethinking the ozone problem in urban and regional air pollution. Washington, DC: National Research
Council. 038041
NRC Committee on Models in the Regulatory Decision Process (2007). Models in environmental regulatory decision
making. Washington, DC: National Academies Press . http://www.nap.edu/catalog/11972.html 632611
NRC Committee on the Significance of International Transport of Air Pollutants (2009). Global sources of local pollution:
An assessment of long-range transport of key air pollutants to and from the United States. Washington, DC: The
National Academies Press, http://www.nap.edu/catalog.php?record_id=12743. 202867
O-Keeffe, S.; Fitzpatrick, C.; Lewis, E. (2007). An optical fibre based ultra violet and visible absorption spectroscopy
system for ozone concentration monitoring. Sens Actuators B, 125: 372-378.
http://dx.doi.0rg/10.1016/i.snb.2007.02.023 595114
Ohira, S. L; Dasgupta, P. K.; Schug, K. A. (2009). Fiber optic sensor for simultaneous determination of atmospheric
nitrogen dioxide, ozone, and relative humidity. Anal Chem, 81: 4183-4191. http://dx.doi.org/10.1021/ac801756z
595113
Olaguer, E. P.; Rappengliick, B.; Lefer, B.; Stutz, J.; Dibb, J.; Griffin, R.; Brune, W. H.; Shauck, M.; Buhr, M.; Jeffries, H.;
Vizuete, W; Pinto, J. P. (2009). Deciphering the role of radical precursors during the Second Texas Air Quality
Study. J Air Waste Manag Assoc, 59: 1258-1277. http://dx.doi.org/10.3155/1047-3289.59.ll.1258 200191
Olszyna, K. J.; Bailey, E. M.; Simonaitis, R.; Meagher, J. F. (1994). O3 and NOy relationships at a rural site. J Geophys
Res, 99: 14,557-14,563. 038832
Oltmans, S. J.; Lefohn, A. S.; Harris, J. M.; Shadwick, D. S. (2008). Background ozone levels of air entering the west coast
of the US and assessment of longer-term changes. Atmos Environ, 42: 6020-6038.
http://dx.doi.0rg/10.1016/i.atmosenv.2008.03.034615534
Park, R. J.; Stenchikov, G. L.; Pickering; Dickerson, R. R.; Allen, D. J.; Kondragunta, S. (2001). Regional air pollution and
its radiative forcing: Studies with a single column chemical and radiation transport model. J Geophys Res, 106:
28.751-28.770. 044169
Parrish, D. D. (2006). Critical evaluation of US on-road vehicle emission inventories. Atmos Environ, 40: 2288-2300.
090352
Parrish, D. D.; Millet, D. B.; Goldstein, A. H. (2009). Increasing ozone in marine boundary layer inflow at the west coasts
of North America and Europe. Atmos Chem Phys, 9: 1303-1323. 616076
Part 58 -Ambient air quality surveillance: Appendix A - quality assurance requirements for state and local air monitoring
stations (SLAMS), SPMs, and PSD Air Monitoring. 51 Fed. Reg. 9595 (1986). 035997
Perring, A. E.; Bertram, T. H.; Wooldridge, P. J.; Fried, A.; Heikes, B. G; Dibb, J.; Crounse, J. D.; Wennberg, P. O.; Blake,
N. J.; Blake, D. R.; Brune, W. H.; Singh, H. B.; Cohen, R. C. (2009). Airborne observations of total RONO2: New
constraints on the yield and lifetime of isoprene nitrates. Atmos Chem Phys, 9: 1451-1463. 616370
March 2011 3-106 DRAFT - DO NOT CITE OR QUOTE
-------
Pfister, G; Hess, P. G; Emmons, L. K.; Lamarque, J. R; Wiedinmyer, C.; Edwards, D. P.; Petron, G; Gille, J. C.; Sachese,
G W. (2005). Quantifying CO emissions from the 2004 Alaskan wildfires using MOPITT CO data. Geophys Res
Lett 32: LI 1809. 093009
Pinto, J. P.; Lefohn, A. S.; Shadwick, D. S. (2004). Spatial variability of PM2.5 in urban areas in the United States. J Air
Waste Manag Assoc, 54: 440-449. 025033
Pokharel, S. S.; Bishop, G. A.; Stedman, D. H. (2002). An on-road motor vehicle emissions inventory for Denver: An
efficient alternative to modeling. Atmos Environ, 36: 5177-5184. 052473
Pokharel, S. S.; Bishop, G. A.; Stedman, D. H. (2003). Emissions reductions as a result of automobile improvement.
Environ Sci Technol, 37: 5097-5101. 053740
Pollack, A. K.; Lindhjem, C.; Stoeckenius, T. E.; Iran, C.; Mansell, G; Jimenez, M.; Wilson, G; Coulter-Burke, S. (2004).
Final Report: Evaluation of the US EPA MOBILE6 highway vehicle emission factor model (Report No. CRC
Project E-64). Novato, CA: ENVIRON International Corporation. 184461
Poppe, D.; Wallasch, M.; Zimmermann, J. (1993). The dependence of the concentration of OH on its precursors under
moderately polluted conditions: a model study. J Atmos Chem, 16: 61-78. 044229
Rao, S. T.; Ku, J.-Y; Berman, S.; Zhang, K.; Mao, H. (2003). Summertime characteristics of the atmospheric boundary
layer and relationships to ozone levels over the eastern United States. Pure Appl Geophys, 160: 21-55. 054094
Rappengliick, B.; Dasgupta, P. K.; Leuchner, M.; Li, Q.; Luke, W. (2009). Formaldehyde and its relation to CO, PAN, and
SO2 in the Houston-Galveston airshed. Atmos Chem Phys Discuss, 9: 24193-24223. http://dx.doi.org/10.5194/acp-
10-2413-2010629680
Reid, N.; Yap, D.; Bloxam, R. (2008). The potential role of background ozone on current and emerging air issues: An
overview. Air Qual. Atmos. Health , 1: 19-29. http://dx.doi.org/10.1007/sll869-008-0005-z 665032
Reidmiller, D. R.; and Fiore, A. M.; Jaffe, D. A.; Bergmann, D.; Cuvelier, C.; Dentener, F. J.; Duncan; Bryan, N.; Folberth,
G; Gauss, M.; Gong, S.; Hess, P.; Jonson, J. E.; Keating, T; Lupu, A.; Manner, E.; Park, R.; Schultz, M. G;
Shindell, D. T; Szopa, S.; Vivanco, M. G; Wild, O.; Zuber, A. (2009). The influence of foreign vs. North American
emissions on surface ozone in the US. Atmos Chem Phys, 9: 5027-5042. http://eprints.lancs.ac.uk/27977/. 644826
Reisinger, A. R. (2000). Unidentified interference in DOAS measurements of ozone. Appl Spectros Rev, 54: 72-79. 053936
Richards, N. A. D.; Osterman, G. B.; Browell, E. V; Hair, J. W.; Avery, M.; Qinbin, L. (2008). Validation of tropospheric
emission spectrometer ozone profiles with aircraft observations during the intercontinental chemical transport
experiment-B. J Geophys Res, 113: D16S29. http://dx.doi.org/10.1029/2007jd008815 617597
Riediker, M.; Williams, R.; Devlin, R.; Griggs, T; Bromberg, P. (2003). Exposure to particulate matter, volatile organic
compounds, and other air pollutants inside patrol cars. Environ Sci Technol, 37: 2084-2093. 043761
Rodes, C. E.; Holland, D. M. (1981). Variations of NO, NO2 and O3 concentrations downwind of a Los Angeles freeway.
Atmos Environ, 15: 243-250. 041110
Russell, A.; Dennis, R. (2000). NARSTO critical review of photochemical models and modeling. Atmos Environ, 34:
2283-2324. 035563
Ryerson, T. B.; Buhr, M. P.; Frost, G. J.; Goldan, P. D.; Holloway, J. S.; Hubler, G; Jobson, B. T; Kuster, W. C.; McKeen,
S. A.; Parrish, D. D.; Roberts, J. M.; Sueper, D. T; Trainer, M.; Williams, J.; Fehsenfeld, F. C. (1998). Emissions
lifetimes and ozone formation in power plant plumes. J Geophys Res, 103: 22,569-22,583. 048310
Ryerson, T. B.; Trainer, M.; Holloway, J. S.; Parrish, D. D.; Huey, L. G; Sueper, D. T; Frost, G. J.; Donnelly, S. G;
Schauffler, S.; Atlas, E. L.; Kuster, W. C.; Goldan, P. D.; Hubler, G; Meagher, J. R; Fehsenfeld, F. C. (2001).
Observations of ozone formation in power plant plumes and implications for ozone control strategies. Science, 292:
719-723. 016249
Sakugawa, H.; Kaplan, I. R. (1989). H2O2 and O3 in the atmosphere of Los Angeles and its vicinity: Factors controlling
their formation and their role as oxidants of SO2. J Geophys Res, 94: 12957-12973. 044129
Sarwar, G; Roselle, S. J.; Mathur, R.; Appel, W.; Dennis, R. L.; Vogel, B. (2008). A comparison of CMAQ HONO
predictions with observations from the Northeast Oxidant and Particle Study. Atmos Environ, 42: 5760-5770.
618491
March 2011 3-107 DRAFT - DO NOT CITE OR QUOTE
-------
Schichtel, B. A.; Husar, R. B. (2001). Eastern North American transport climatology during high- and low-ozone days.
Atmos Environ, 35: 1029-1038. 016669
Schnell, R. C.; Oltmans, S. I; Neely, R. R.; Endres, M. S.; Molenar, J. V; White, A. B. (2009). Rapid photochemical
production of ozone at high concentrations in a rural site during winter. Nat Geosci, 2: 120-122.
http://dx.doi.org/10.1038/NGE0415 180146
Seaman, N. L. (2000). Meteorological modeling for air quality assessments. Atmos Environ, 34: 2231-2259. 035562
Seinfeld, J. H.; Pandis, S. N. (1998). Atmospheric chemistry and physics: From air pollution to climate change. New York:
John Wiley & Sons. 018352
Sexton, K. G; Jeffries, H. E.; Jang, M.; Kamens, R. M.; Doyle, M.; Voicu, I.; Jaspers, I. (2004). Photochemical products in
urban mixtures enhance inflammatory responses in lung cells. Inhal Toxicol, 1: 107-114. 087831
Shapiro, M. A. (1980). Turbulent mixing within tropopause folds as a mechanism for the exchange of chemical constituents
between the stratosphere and troposphere. J Atmos Sci, 37: 994-1004. 047756
Sillman, S. (1995). The use of NOy, H2O2 and HNO3 as indicators for ozone-NOx-hydrocarbon sensitivity in urban
locations. J Geophys Res, 100: 14,175-14,188. 052346
Sillman, S.; He, D.-Y. (2002). Some theoretical results concerning O3-NOx-VOC chemistry and NOx-VOC indicators. J
Geophys Res, 107: 4659. http://dx.doi.org/10.1029/2001JD001123 052350
Sillman, S.; He, D.; Pippin, M. R.; Daum, P. H.; Imre, D. G; Kleinman, L. L; Lee, J. H.; Weinstein-Lloyd, J. (1998). Model
correlations for ozone, reactive nitrogen, and peroxides for Nashville in comparison with measurements:
implications for O3-NOx-hydrocarbon chemistry. J Geophys Res, 103: 22,629-22,644. 052223
Spicer, C. W.; Joseph, D. W.; Ollison, W. M. (2010). Are-examination of ambient air ozone monitor interferences. J Air
Waste Manag Assoc, 60: 1353-1364. http://dx.doi.org/10.3155/1047-3289.60.ll.1353 670424
Stedman, D. H.; Daby, E. E.; Stuhl, R; Niki, H. (1972). Analysis of ozone and nitric oxide by a chemiluminescent method
in laboratory and atmospheric studies of photochemical smog. J Air Waste Manag Assoc, 22: 260-263. 033730
Stevens, R.; Pinto, J.; Mamane, Y; Ondov, J.; Abdulraheem, M.; Al-Majed, N.; Sadek, M.; Cofer, W.; Ellenson, W.;
Kellogg, R. (1993). Chemical and physical properties of emissions from Kuwaiti oil fires. Water Sci Technol, 27:
223-233. 046404
Stutz, J.; Ackermann, R.; Fast, J. D.; Barrie, L. (2002). Atmospheric reactive chlorine and bromine at the Great Salt Lake,
Utah. Geophys Res Lett, 29: 1380. http://dx.doi.org/10.1029/2002GL014812 051882
Stutz, J.; Oh, H. J.; Whitlow, S. I.; Anderson, C.; Dibb, J. E.; Flynn, J. H.; Rappengluck, B.; Lefe, B. (2009). Simultaneous
DOAS and mist-chamber 1C measurements of HONO in Houston, TX. Atmos Environ, TBD: TBD.
http://dx.doi.0rg/10.1016/i.atmosenv.2009.02.003595124
Tanimoto, H.; Mukai, H.; Hashimoto, S.; Norris, J. E. (2006). Intercomparison of ultraviolet photometry and gas-phase
titration techniques for ozone reference standards at ambient levels. J Geophys Res, 111: D16313.
http://dx.doi.org/10.1029/2005JD006983595131
Tarasick, D. W.; Slater, R. (2008). Ozone in the troposphere: Measurements, climatology, budget, and trends. Atmos. Ocean
, 46: 93-115. http://dx.doi.org/10.3137/ao.460105 596431
Taubman, B. R; Hains, J. C.; Thompson, A. M.; Marufu, L. T; Doddridge, B. G; Stehr, J. W.; Piety, C. A.; Dickerson, R. R.
(2006). Aircraft vertical profiles of trace gas and aerosol pollution over the mid-Atlantic United States: statistics
and meteorological cluster analysis. J Geophys Res, 111: D10S07. http://dx.doi.org/10.1029/2005JD006196 087582
Taubman, B. R; Marufu, L. T; Piety, C. A.; Doddridge, B. G; Stehr, J. W.; Dickerson, R. R. (2004). Airborne
characterization of the chemical, optical, and meteorological properties, and origins of a combined ozone-haze
episode over the eastern United States. J Atmos Sci, 61: 1781-1793. 052228
Thompson, A. M.; Stone, J. B.; Witte, J. C.; Miller, S. K.; Oltmans, S. J.; Kucsera, T. L.; Ross, K. L.; Pickering, K. E.;
Merrill, J. T; Forbes, G; Tarasick, D. W.; Joseph, E.; Schmidlin, F. J.; McMillan, W. W.; Warner, J.; Hintsa, E. J.;
Johnson, J. E. (2007). Intercontinental Chemical Transport Experiment Ozonesonde Network study (IONS) 2004: 2
Tropospheric ozone budgets and variability over northeastern North America. J Geophys Res, 112: D12S13.
http://dx.doi.org/10.1029/2006JD007670 090796
March 2011 3-108 DRAFT - DO NOT CITE OR QUOTE
-------
Thornton, J. A.; Kercher, J. P.; Riedel, T. P.; Wagner, N. L.; Cozic, I; Holloway, J. S.; Dube, W. P.; Wolfe, G. M.; Quinn, P.
K.; Middlebrook, A. M.; Alexander, B.; Brown, S. S. (2010). A large atomic chlorine source inferred from mid-
continental reactive nitrogen chemistry. Nature, 464: 271-274. http://dx.doi.org/10.1038/nature08905 386870
Trainer, M.; Parrish, D. D.; Buhr, M. P.; Norton, R. B.; Fehsenfeld, F. C.; Anlauf, K. G; Bottenheim, J. W.; Tang, Y. Z.;
Wiebe, H. A.; Roberts, J. M.; Tanner, R. L.; Newman, L.; Bowersox, V. C.; Meagher, J. F.; Olszyna, K. J.; Rodgers,
M. O.; Wang, T; Berresheim, H.; Demerjian, K. L.; Roychowdhury, U. K. (1993). Correlation of ozone with NOy
in photochemically aged air. J Geophys Res, 98: 2917-2925. 038672
U.S. Census Bureau (2011). Census. Retrieved January 28, 2011 from http://www.census.gov/. 677549
U.S. EPA (1979). Transfer standards for the calibration of ambient air monitoring analyzers for ozone: technical assistance
document (Report No. EPA-600/4-79-056). Research Triangle Park, NC: U.S. Environmental Protection Agency.
039211
U.S. EPA (1996). Air quality criteria for ozone and related photochemical oxidants (Report No. EPA/600/P-93/004AF).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development. (NTIS
No. PB94173127) 017831
U.S. EPA (2006). Air quality criteria for ozone and related photochemical oxidants (Report No. EPA/600/R-05/004AF).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid= 149923. 088089
U.S. EPA (2008). 2005 National Emissions Inventory Data & Documentation. Retrieved November 01, 2010 from
http://www.epa.gov/ttn/chief/net/2005inventory.html. 665052
U.S. EPA (2008). Integrated science assessment for oxides of nitrogen - Health criteria (Report No. EPA/600/R-08/071).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development.
http://cfpub.epa. gov/ncea/cfm/recordisplay.cfm?deid= 194645. 157073
U.S. EPA (2008). Integrated science assessment for sulfur oxides - Health criteria (Report No. EPA/600/R-08/047F).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development.
http ://cfpub .epa. go v/ncea/cfm/recordisplay. cfm?deid= 19 8 84 3. 157075
U.S. EPA (2008). National Air Quality Status and Trends Through 2007 (Report No. EPA/454/R-08/006 ). Research
Triangle Park, NC: U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards. 191190
U.S. EPA (2009). Integrated science assessment for particulate matter (Report No. EPA/600/R-08/139F). Research Triangle
Park, NC: U.S. Environmental Protection Agency, Office of Research and Development.
http://cfpub.epa. go v/ncea/cfm/recordisplay. cfm?deid=216546. 179916
U.S. EPA (2009). Risk and Exposure Assessment for Review of the Secondary National Ambient Air Quality Standards for
Oxides of Nitrogen and Oxides of Sulfur Second Draft (Report No. EPA/452/P-09/008A). Research Triangle Park,
NC: U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards.
http://www.epa. go v/ttnnaaqs/standards/no2so2sec/cr_rea.html. 191774
U.S. EPA (2010). Air Trends: Design Values. Retrieved February 01, 2011 from http://epa.gov/airtrends/values.html.
677582
U.S. EPA (2010). Biogenic Emissions Inventory System (BEIS) Modeling. Retrieved January 28, 2011 from
http://www.epa.gov/AMD/biogen.html. 677538
U.S. EPA (2010). Integrated science assessment for carbon monoxide (Report No. EPA/600/R-09/019F). Research Triangle
Park, NC: U.S. Environmental Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=218686. 626035
U.S. EPA (2010). MOBILE6 vehicle emission modeling software. Retrieved January 28, 2011 from
http://www.epa.gov/otaq/m6.htm. 677539
U.S. EPA (2010). Our nation's air: Status and trends through 2008 (Report No. EPA/454/R-09/002). Research Triangle
Park, NC: U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards.
http://www.epa.gov/airtrends/2010/report/fullreport.pdf 647278
U.S. EPA (2010). Transfer standards for calibration of air monitoring analyzers for ozone (Report No. EPA-454/B-10-001).
Research Triangle Park, NC: U.S. Environmental Protection Agency.
http://www.epa.gov/ttn/amtic/files/ambient/qaqc/OzoneTransferStandardGuidance.pdf 677553
March 2011 3-109 DRAFT - DO NOT CITE OR QUOTE
-------
U.S. EPA (2011). AirNow. Retrieved January 28, 2011 from http://www.airnow.gov/. 677548
U.S. EPA (2011). Map monitoring sites. Retrieved January 28, 2011 from
http://www.epa.gov/airexplorer/monitor_kml.htm. 677547
U.S. EPA (2011). MOVES (Motor Vehicle Emission Simulator). Retrieved January 28, 2011 from
http://www.epa.gov/otaq/models/moves/index.htm. 677540
U.S. EPA (2011). Quality indicator assessment reports. Retrieved January 28, 2011 from
http://www.epa.gov/ttn/amtic/qareport.html. 677486
Utembe, S. R.; Hansford, G. M.; Sanderson, M. G; Freshwater, R. A.; Pratt, K. F. E.; Williams, D. E.; Cox, R. A.; Jones, R.
L. (2006). An ozone monitoring instrument based on the tungsten trioxide (WO3) semiconductor. Sens Actuators B,
114: 507-512. http://dx.doi.Org/10.1016/j.snb.2005.04.049 595137
van der Werf, G. R.; Randerson, J. T.; Giglio, L.; Collatz, G. J.; Kasibhatla, P. S.; Arellano Jr, A. F. (2006). Interannual
variability in global biomass burning emissions from 1997 to 2004. Atmos Chem Phys, 6: 3423-3441. 157084
Vardoulakis, S.; Lumbreras, J.; Solazzo, E. (2009). Comparative evaluation of nitrogen oxides and ozone passive diffusion
tubes for exposure studies. Atmos Environ, 43: 2509-2517. http://dx.doi.Org/10.1016/j.atmosenv.2009.02.048
595141
Viallon, J.; Moussay, P.; Norris, J. E.; Guenther, F. R.; Wielgosz, R. I. (2006). A study of systematic biases and
measurement uncertainties in ozone mole fraction measurements with the NIST Standard Reference Photometer.
Metrologia, 43: 441-450. http://dx.doi.Org/10.1088/0026-1394/43/5/016 595145
Walko, R.; Avissar, R. (2008). The ocean-land-atmosphere model (OLAM). Part II: formulation and tests of the
nonhydrostatic dynamic core. Mon Weather Rev, 136: 4045-4062. http://dx.doi.org/10.1175/2008MWR2523.!
665033
Wang, H. Q.; Jacob, D. J.; Le Sager, P.; Streets, D. G; Park, R. J.; Gilliland, A. B.; van Donkelaar, A. (2009). Surface ozone
background in the United States: Canadian and Mexican pollution influences. Atmos Environ, 43: 1310-1319.
http://dx.doi.0rg/10.1016/i.atmosenv.2008.ll.036622281
Wang, J.; Christopher, S. A.; Nair, U. S.; Reid, J. S.; Prins, E. M.; Szykman, J.; Hand, J. L. (2006). Mesoscale modeling of
Central American smoke transport to the United States: 1. "Top-down" assessment of emission strength and diurnal
variation impacts. J Geophys Res, 111: D05S17. http://dx.doi.org/10.1029/2005JD006416 157109
Webster, M.; Nam, J.; Kimura, Y; Jeffries, H.; Vizuete, W.; Allen, D. T. (2007). The effect of variability in industrial
emissions on ozone formation in Houston, Texas. Atmos Environ, 41: 9580-9593.
http://dx.doi.0rg/10.1016/i.atmosenv.2007.08.052 104266
Wernli, H.; Bourqui, M. (2002). A Lagrangian "1-year climatology" of (deep) cross-tropopause exchange in the
extratropical Northern Hemisphere. J Geophys Res, 107: 4021. http://dx.doi.org/10.1029/2001JD000812 052425
Williams, E. J.; Fehsenfeld, F. C.; Jobson, B. T; Kuster, W. C.; Goldan, P. D.; Stutz, J.; McClenny, W. A. (2006).
Comparison of ultraviolet absorbance, chemiluminescence, and DOAS instruments for ambient ozone monitoring.
Environ Sci Technol, 40: 5755-5762. http://dx.doi.org/10.1021/es0523542 595152
Wilson, K. L.; Birks, J. W. (2006). Mechanism and elimination of a water vapor interference in the measurement of ozone
by UV absorbance. Environ Sci Technol, 40: 6361-6367. http://dx.doi.org/10.1021/es052590c 595155
Worden, H. M.; Logan, J. A.; Worden, J. R.; Beer, R.; Bowman, K.; Clough, S. A.; Eldering, A.; Fisher, B. M.; Gunson, M.
R.; Herman, R. L.; Kulawik, S. S.; Lampel, M. C.; Luo, M.; Megretskaia, I. A.; Osterman, G. B.; Shephard, M. W.
(2007). Comparisons of Tropospheric Emission Spectrometer (TES) ozone profiles to ozonesondes: Methods and
initial results. J Geophys Res, 112: [np]. 623018
Worden, J.; Liu, X.; Bowman, K.; Chance, K.; Beer, R.; Eldering, A.; Gunson, M.; Worden, H. (2007). Improved
tropospheric ozone profile retrievals using OMI and TES radiances. Geophys Res Lett, 34: [np]. 623020
Wu, S.; Mickley, L. J.; Leibensperger, E. M.; Jacob, D. J.; Rind, D.; Streets, D. G. (2008). Effects of 2000-2050 global
change on ozone air quality in the United States. J Geophys Res, 113: D06302.
http://dx.doi.org/10.1029/2007JD008917629684
Yang, Q.; Cunnold, D. M.; Choi, Y; Wang, Y; Nam, J.; Wang, H. J.; Froidevaux, L.; Thompson, A. M.; Bhartia, P. K.
(2010). A study of tropospheric ozone column enhancements over North America using satellite data and a global
chemical transport model. J Geophys Res, 115: D08302. http://dx.doi.org/10.1029/2009JD012616 628857
March 2011 3-110 DRAFT-DO NOT CITE OR QUOTE
-------
Yung, Y. L.; Pinto, J. P.; Watson, R. T.; Sander, S. P. (1980). Atmospheric bromine and ozone perturbations in the lower
stratosphere. J Atmos Sci, 37: 339-353. 057212
Zhang, K.; Wexler, A. (2008). Modeling urban and regional aerosols—Development of the UCD Aerosol Module and
implementation in CMAQ model. Atmos Environ, 42: 3166-3178 . 191770
Zhang, L.; Jacob, D. J.; Boersma, K. F.; Jaffe, D. A.; Olson, J. R.; Bowman, K. W.; Worden, J. R.; Thompson, A. M.; Avery,
M. A.; Cohen, R. C.; Dibb, J. E.; Flock, F. M.; Fuelberg, H. E.; Huey, L. G; McMillan, W. W.; Singh, H. B.;
Weinheimer, A. J. (2008). Transpacific transport of ozone pollution and the effect of recent Asian emission
increases on air quality in North America: an integrated analysis using satellite, aircraft, ozonesonde, and surface
observations. Atmos Chem Phys, 8: 6117-6136. 624402
Zhang, L.; Jacob, D. J.; Logan, J. A.; Chance, K.; Eldering, A.; Bojkov, B. R. (2010). Intercomparison methods for satellite
measurements of atmospheric composition: Application to tropospheric ozone from TES and OMI. Atmos Chem
Phys, 10: 4725-4739. http://dx.doi.org/10.5194/acpd-10-1417-2010 633894
Zhang, Q.; Jimenez, J. L.; Canagaratna, M. R.; Jayne, J. T; Worsnop, D. R. (2005). Time- and size-resolved chemical
composition of submicron particles in Pittsburgh: Implications for aerosol sources and processes. J Geophys Res,
110: 1-19. 157185
Zhang, X.; Zhuang, G; Guo, J.; Yin, K.; Zhang, P. (2007). Characterization of aerosol over the Northern South China Sea
during two cruises in 2003. Atmos Environ, 41: 7821-7836. 101119
Ziemke JR: Chandra, S.; Duncan, B. N.; Froidevaux, L.; Bhartia, P. K.; Levelt, P. F.; Waters, J. W. (2006). Tropospheric
ozone determined from Aura OMI and MLS: Evaluation of measurements and comparison with the Global
Modeling Initiative's Chemical Transport Model. J Geophys Res, 111: D19303.
http://dx.doi.org/10.1029/2006JD007089595159
Zimmermann, J.; Poppe, D. (1993). Nonlinear chemical couplings in the tropospheric NOx—HOx gas phase chemistry. J
Atmos Chem, 17: 141-155. 052378
3.8. Chapter 3 Appendix - Supplemental Figures and
Tables
3.8.1. Time Series of GEOS-Chem Model Predictions and Observations at
Selected CASTNET Sites
1 This section contains time series plots of 8-h daily max O3 concentrations observed at 14
2 CASTNET sites during 2001 and corresponding GEOS-Chem predictions for the base model (i.e.,
3 model including all anthropogenic and natural sources) and the PRB model (i.e., model including
4 natural sources everywhere in the world and anthropogenic sources outside the U.S., Canada, and
5 Mexico). Further details on these predictions can be found in Section 3.4.3.
March 2011 3-111 DRAFT-DO NOT CITE OR QUOTE
-------
Acarfia NP. ME
-co
SB
60
70
60
SO
40
30
20
Id
Figure 3A-1. Time series of observed and GEOS-Chem base model and PRB model 8-h daily
max ozone concentrations (ppb) with PRB estimates for Acadia NP, ME.
Obser.cd
-Wadcled-Slaniarc ^
-Modeled-PRB
Everglades NP, FL
Figure 3A-2. Time series of observed and GEOS-Chem base model and PRB model 8-h daily
max ozone concentrations (ppb) with PRB estimates for Everglades NP, FL.
March 2011
3-112
DRAFT - DO NOT CITE OR QUOTE
-------
Crockett. KY
Figure 3A-3. Time series of observed and GEOS-Chem base model and PRB model 8-h daily
max ozone concentrations (ppb) with PRB estimates for Crockett, KY.
': . • vill. . MS
Figure 3A-4. Time series of observed and GEOS-Chem base model and PRB model 8-h daily
max ozone concentrations (ppb) with PRB estimates for Coffeeville, MS.
March 2011
3-113
DRAFT - DO NOT CITE OR QUOTE
-------
100
90
3D
£ "
I"
Cbser.ed
-Modeled- slandarc ^-
Modetel - PRB
0
IJTVC!
Voyageurs NP, MN
Figure 3A-5. Time series of observed and GEOS-Chem base model and PRB model 8-h daily
max ozone concentrations (ppb) with PRB estimates for Voyageurs NP, MN.
Bij Bend NP, TX
Figure 3A-6. Time series of observed and GEOS-Chem base model and PRB model 8-h daily
max ozone concentrations (ppb) with PRB estimates for Big Bend NP, TX.
March 2011
3-114
DRAFT - DO NOT CITE OR QUOTE
-------
1:0
90
aa
I
rf
-Moosea • aval-card Pun
Rooky Mountain NP. CO
Figure 3A-7. Time series of observed and GEOS-Chem base model and PRB model 8-h daily
max ozone concentrations (ppb) with PRB estimates for Rocky Mountain NP, CO.
tm
90
a:
50
1
Oosw.es
- VcaeM - SlsrcorJ 3Ji
- vcaelK • PS=
Mesa Verde NP, CO
Figure 3A-8. Time series of observed and GEOS-Chem base model and PRB model 8-h daily
max ozone concentrations (ppb) with PRB estimates for Mesa Verde NP, CO.
March 2011
3-115
DRAFT - DO NOT CITE OR QUOTE
-------
Yellowstone NP.WY
-------
Canyon lands NP. UT
100
•-t
H
j 50
2 40
«
«
30
Figure 3A-11. Time series of observed and GEOS-Chem base model and PRB model 8-h daily
max ozone concentrations (ppb) with PRB estimates for Canyonlands NP, UT.
j
i
Glacier NP.MT
Figure 3A-12. Time series of observed and GEOS-Chem base model and PRB model 8-h daily
max ozone concentrations (ppb) with PRB estimates for Glacier NP, MT.
March 2011
3-117
DRAFT - DO NOT CITE OR QUOTE
-------
Mount Rainier HP. WA
? ™
j
;
Figure 3A-13. Time series of observed and GEOS-Chem base model and PRB model 8-h daily
max ozone concentrations (ppb) with PRB estimates for Mt. Rainier NP, WA.
5C
j:
j 99
Observed anr-^ax ozc
ir Shi-Wai
Lassen Volcanic NP, CA
Figure 3A-14. Time series of observed and GEOS-Chem base model and PRB model 8-h daily
max ozone concentrations (ppb) with PRB estimates for Lassen Volcanic NP, CA.
March 2011
3-118
DRAFT - DO NOT CITE OR QUOTE
-------
Pinnacles Nil. CA
Figure 3A-15. Time series of observed and GEOS-Chem base model and PRB model 8-h daily
max ozone concentrations (ppb) with PRB estimates for Pinnacles NM, CA.
3.8.2. Ozone Monitor Maps for the Urban Focus Cities
1 This section contains supplemental maps showing the location of O3 monitors reporting to
2 AQS for each of the 20 urban focus cities introduced in Section 3.6.2.1. The monitors are delineated
3 in the maps as year-round or warm-season based on their inclusion in the year-round data set and the
4 warm-season data set discussed in Section 3.6.2.1. The maps also include the CSA/CBSA boundary
5 selected for monitor inclusion, the location of urban areas and water bodies, the major roadway
6 network, as well as the population gravity center based on the entire CSA/CBSA and the individual
7 focus city boundaries. Population gravity center is calculated from the average longitude and latitude
8 values for the input census tract centroids and represents the mean center of the population in a given
9 area. Census tract centroids are weighted by their population during this calculation.
March 2011
3-119
DRAFT - DO NOT CITE OR QUOTE
-------
Legend
Monitor Locations
© Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
• CSA-based Population Gravity Center
Interstate Highways
Major Highways
^B Water Bodies
Urban Areas
Atlanta CSA
60 Kilometers
Figure 3A-16. Map of the Atlanta CSA including ozone monitor locations, population gravity
centers, urban areas, and major roadways.
March 2011
3-120
DRAFT - DO NOT CITE OR QUOTE
-------
Legend
Monitor Locations
O Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
• CSA-based Population Gravity Center
Interstate Highways
Major Highways
Baltimore CSA
Water Bodies
Urban Areas
80 Kilometers
Figure 3A-17. Map of the Baltimore CSA including ozone monitor locations, population gravity
centers, urban areas, and major roadways.
March 2011
3-121
DRAFT - DO NOT CITE OR QUOTE
-------
Legend
Monitor Locations
O Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
• CSA-based Population Gravity Center
Interstate Highways
Major Highways
Water Bodies
Urban Areas
Birmingham CSA
50 Kilometers
Figure 3A-18. Map of the Birmingham CSA including ozone monitor locations, population
gravity centers, urban areas, and major roadways.
March 2011
3-122
DRAFT - DO NOT CITE OR QUOTE
-------
Legend
Monitor Locations
© Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
• CSA-based Population Gravity Center
Interstate Highways
Major Highways
^B Water Bodies
Urban Areas
Boston CSA
100 Kilometers
Figure 3A-19. Map of the Boston CSA including ozone monitor locations, population gravity
centers, urban areas, and major roadways.
March 2011
3-123
DRAFT - DO NOT CITE OR QUOTE
-------
Legend
Monitor Locations
O Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
• CSA-based Population Gravity Center
Interstate Highways
Major Highways
Water Bodies
Urban Areas | ,-
Chicago CSA
0
Figure 3A-20. Map of the Chicago CSA including ozone monitor locations, population gravity
centers, urban areas, and major roadways.
March 2011
3-124
DRAFT - DO NOT CITE OR QUOTE
-------
LS
egend
Monitor Locations
O Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
• CSA-based Population Gravity Center
Interstate Highways
Major Highways
Water Bodies
Urban Areas
Dallas CSA
80 Kilometers
Figure 3A-21. Map of the Dallas CSA including ozone monitor locations, population gravity
centers, urban areas, and major roadways.
March 2011
3-125
DRAFT - DO NOT CITE OR QUOTE
-------
Legend
Monitor Locations
O Warm-season Monitors
• Year-round Monitors
0 City-based Population Gravity Center
• CSA-based Population Gravity Center
Interstate Highways
Major Highways
Water Bodies
Urban Areas
Denver CSA
I I I 'I 'I 'I
0 20 40 80 Kilometers
Figure 3A-22. Map of the Denver CSA including ozone monitor locations, population gravity
centers, urban areas, and major roadways.
March 2011
3-126
DRAFT - DO NOT CITE OR QUOTE
-------
Monitor Locations
0 Warm-season Monitors
• Year-round Monitors
City-based Population Gravity Center
• CSA-based Population Gravity Center
Interstate Highways
Major Highways
Water Bodies
Urban Areas
Detroit CSA
Figure 3A-23. Map of the Detroit CSA including ozone monitor locations, population gravity
centers, urban areas, and major roadways.
March 2011
3-127
DRAFT - DO NOT CITE OR QUOTE
-------
Legend
Monitor Locations
O Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
• CSA-based Population Gravity Center
Interstate Highways
Major Highways
Water Bodies
Urban Areas
Houston CSA
N
50
100 Kilometers
Figure 3A-24. Map of the Houston CSA including ozone monitor locations, population gravity
centers, urban areas, and major roadways.
March 2011
3-128
DRAFT - DO NOT CITE OR QUOTE
-------
Legend
Monitor Locations
O Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
0 CSA-based Population Gravity Center
Interstate Highways
Major Highways
Water Bodies
Urban Areas
Los Angeles CSA
/W
200 Kilometers
Figure 3A-25. Map of the Los Angeles CSA including ozone monitor locations, population
gravity centers, urban areas, and major roadways.
March 2011
3-129
DRAFT - DO NOT CITE OR QUOTE
-------
Legend
Monitor Locations
O Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
• CSA-based Population Gravity Center
Interstate Highways
Major Highways
Water Bodies
Urban Areas
Minneapolis CSA
50
100 Kilometers
Figure 3A-26. Map of the Minneapolis CSA including ozone monitor locations, population
gravity centers, urban areas, and major roadways.
March 2011
3-130
DRAFT - DO NOT CITE OR QUOTE
-------
Monitor Locations
© Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
• CSA-based Population Gravity Center
Interstate Highways
Major Highways
Water Bodies
Urban Areas
New York CSA
Figure 3A-27. Map of the New York CSA including ozone monitor locations, population gravity
centers, urban areas, and major roadways.
March 2011
3-131
DRAFT - DO NOT CITE OR QUOTE
-------
Legend
Monitor Locations
0 Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
• CSA-based Population Gravity Center
Interstate Highways
Major Highways
Water Bodies
Urban Areas
Philadelphia CSA
15
30
60 Kilometers
Figure 3A-28. Map of the Philadelphia CSA including ozone monitor locations, population
gravity centers, urban areas, and major roadways.
March 2011
3-132
DRAFT - DO NOT CITE OR QUOTE
-------
Legend
Monitor Locations
© Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
• CBSA-based Population Gravity Center
Interstate Highways
Major Highways
Water Bodies
Urban Areas
Phoenix CBSA
GO Kilometers
Figure 3A-29. Map of the Phoenix CBSA including ozone monitor locations, population gravity
centers, urban areas, and major roadways.
March 2011
3-133
DRAFT - DO NOT CITE OR QUOTE
-------
Legend
Monitor Locations
O Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
0 CSA-based Population Gravity Center
Interstate Highways
Major Highways
Water Bodies
Urban Areas
Pittsburgh CSA
50 Kilometers
Figure 3A-30. Map of the Pittsburgh CSA including ozone monitor locations, population gravity
centers, urban areas, and major roadways.
March 2011
3-134
DRAFT - DO NOT CITE OR QUOTE
-------
Legend
Monitor Locations
O Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
• CSA-based Population Gravity Center
Interstate Highways
Major Highways
^B Water Bodies
Urban Areas
Salt Lake CSA
30 60
120 Kilometers
Figure 3A-31. Map of the Salt Lake City CSA including ozone monitor locations, population
gravity centers, urban areas, and major roadways.
March 2011
3-135
DRAFT - DO NOT CITE OR QUOTE
-------
Legend
Monitor Locations
O Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
• CBSA-based Population Gravity Center
Interstate Highways
Major Highways
^B Water Bodies
Urban Areas
San Antonio CBSA
12.5 25
50 Kilometers
Figure 3A-32. Map of the San Antonio CBSA including ozone monitor locations, population
gravity centers, urban areas, and major roadways.
March 2011
3-136
DRAFT - DO NOT CITE OR QUOTE
-------
Legend
Monitor Locations
O Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
• CSA-based Population Gravity Center
Interstate Highways
Major Highways
Water Bodies
Urban Areas
San Francisco CSA
25
50
100 Kilometers
Figure 3A-33. Map of the San Francisco CSA including ozone monitor locations, population
gravity centers, urban areas, and major roadways.
March 2011
3-137
DRAFT - DO NOT CITE OR QUOTE
-------
Legend
Monitor Locations
O Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
• CSA-based Population Gravity Center
Interstate Highways
Major Highways
Water Bodies
Urban Areas
Seattle CSA
0 25
100 Kilometers
Figure 3A-34. Map of the Seattle CSA including ozone monitor locations, population gravity
centers, urban areas, and major roadways.
March 2011
3-138
DRAFT - DO NOT CITE OR QUOTE
-------
Monitor Locations
O Warm-season Monitors
• Year-round Monitors
• City-based Population Gravity Center
• CSA-based Population Gravity Center
Interstate Highways
Major Highways
| Water Bodies
Urban Areas
St Louis CSA
0 20 40
80 Kilometers
Figure 3A-35. Map of the St. Louis CSA including ozone monitor locations, population gravity
centers, urban areas, and major roadways.
3.8.3. Ozone Concentration Box Plots for the Urban Focus Cities
1 This section contains box plots depicting the distribution of 2007-2009 warm-season 8-h daily
2 max O3 data from each individual monitor in the 20 urban focus cities introduced in Section 3.6.2.1.
3 Monitor information including the AQS site id, the years containing qualifying data between 2007
4 and 2009, and the number of 8-h daily max O3 observations included in the data set are listed next to
5 the box plot. Statistics including the mean, standard deviation (SD), median and inner quartile range
6 (IQR) are also shown for each monitor with the site letter corresponding to the sites listed in the
7 figures above.
March 2011
3-139
DRAFT - DO NOT CITE OR QUOTE
-------
Atlanta CSA
Site ID
131210055
130890002
131350002
130670003
132470001
130970004
131130001
131510002
130770002
130850001
132230003
Years
07-09
07-09
07-09
07-09
07-09
07-09
07-08
07-09
07-09
07-09
07-09
N
450
452
446
459
450
455
306
459
455
458
455
Mean
53
52
52
51
51
52
52
51
47
47
50
SD
17
18
16
16
18
15
15
17
16
13
14
Median
54
52
52
52
51
53
52
51
47
47
50
IQR
22
23
18
22
22
22
20
22
19
17
21
Key
.C
to
H-
~in
(M
-H
mean
•
median
\
overall
meOian
I
to
1"-
\-
In
os
--1
Site
A
c-
D-
C
G-
\
J-
K-
, , , , \
!-'-"-'&
i J
I 1
' V-vS
,
1
3
i
~~i
j
~i
i i i i i i i i
--:
A
P
-c
-D
C
-G
LJ
I
- J
-K
50 100
03 (ppb)
150
Figure 3A-36. Site information, statistics and box plots for 8-h daily max ozone from AQS
monitors meeting the warm-season data set inclusion criteria within the Atlanta
CSA.
Baltimore CSA
Site ID
245100054
240053001
240051007
240330030
240251001
240030014
240130001
240313001
110010025
110010041
110010043
240259001
240338003
510130020
510595001
515100009
510591005
240210037
510590030
510590018
511071005
240090011
510590005
240170010
511530009
511790001
510690010
510610002
Years
07-09
07-09
07-09
07-09
07-09
07-09
07-09
08-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
Key
~
"to
I
N
454
456
459
445
450
459
459
292
453
459
459
458
452
459
459
456
432
458
459
459
456
439
459
456
453
459
459
456
-------
Birmingham CSA
Site ID
010730023
010731003
010736002
010732006
011170004
010731010
010731005
010735002
010735003
010731009
Years
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
Key
*to
I----
N
450
459
459
455
459
452
459
459
456
457
c
£ «J
V> u
CM E
H *
Mean
47
45
48
48
49
47
47
47
48
46
median
I
SD
15
15
15
16
16
14
16
13
15
15
^il
s|
o;E
,
Median
48
44
49
48
49
46
46
47
48
46
to >
r^
r
IQR Site
21
22
20
23
23
20
21
18
21
21
to
H
A-
B-
c-
D-
E-
F-
G-
H -
I -
J-
C
i i > i
'h-~-"9
•;:1
-a
i i i i
\ -:
^3 -:
|- -:
\ -•
^)---H
| 1,
— 1
i i i i
-A
-B
-C
-D
-E
-F
-G
-H
-1
-J
) 50 100 150
03 (ppb)
Figure 3A-38. Site information, statistics and box plots for 8-h daily max ozone from AQS
monitors meeting the warm-season data set inclusion criteria within the Birmingham
CSA.
Boston CSA
Site ID
250250041
250092006
250213003
250171102
250170009
250095005
330111011
250270024
250094004
440071010
330110020
330150016
330150014
250051002
440030002
440090007
330012004
Years
07-09
08-09
07-09
07-09
07-09
07-09
07-09
07-09
09
07-08
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
Key
to
N
306
459
459
457
439
459
457
153
305
455
458
459
459
459
458
459
459
c
^ ffi
-10 ®
--a
>-'-&
"?
V-?
1 1 1 I! 1 1 1 1 !
r] •!
• } J
^] -;
r~|- H
•r'-nf-}-^""1
}
-A
-B
-C
-D
- E
- F
-G
- H
- I
- J
-K
-L
-M
- N
- P
-Q
- R
-s
-T
- U
,,,,!,,,,,
50 100 150
03 (ppb)
Figure 3A-39. Site information, statistics and box plots for 8-h daily max ozone from AQS
monitors meeting the warm-season data set inclusion criteria within the Boston
CSA.
March 2011
3-141
DRAFT - DO NOT CITE OR QUOTE
-------
Chicago CSA
Site ID
170314002
170311003
170310076
170310042
170310072
170310064
170436001
170310001
170314007
170311601
170310032
170317002
170314201
180890030
180892008
170890005
180890022
171110001
181270024
170971002
170971007
550590019
181270026
171971011
180910005
180910010
Years
07-09
07-09
07-09
07-09
07-09
07-09
07-09
n7-nQ
07-09
07-09
n7-AQ
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
Key
£
m
i —
N
458
452
458
412
459
459
459
AC.J
450
450
£1 1
451
459
455
458
459
453
458
453
456
C
i: CQ
To £
CM E
H •
Mean
39
44
44
45
42
41
39
A&
46
43
45
44
42
43
46
43
42
42
44
median
I
SD
13
13
14
14
12
13
12
13
13
fS
13
13
13
12
13
13
12
12
13
I|
o|E
|
Median
38
43
44
44
42
40
39
45
42
44
42
41
42
46
42
41
41
43
to
l~~
\--~-
IQR
18
17
18
17
17
18
16
1Q
17
17
18
16
15
15
18
18
15
15
17
^
o>
^
Site
A-
B-
c-
D-
E -
F-
G-
J
K-
|
O-
P~
Q-
R-
U-
w-
X-
Y-
Z~
(.
1 1 ! 1 lilt 1 1 1 1
---|_fc_J~H _
' i 1 i I — . '
^ ' j^~L '
;-"ff--V
1 — R — \
rife
Li • '
'r~~^p~:~\
Ifi;1
) 50 100 1£
O3 (ppb)
-A
-B
-c
-D
- E
-F
-G
J
- K
-O
^P
-Q
-R
-U
-w
-X
- Y
-Z
0
Figure 3A-40. Site information, statistics and box plots for 8-h daily max ozone from AQS
monitors meeting the warm-season data set inclusion criteria within the Chicago
CSA
Dallas CSA
Site ID
481130069
481130075
481130087
484393009
484393011
484392003
480850005
481390016
484391002
483970001
481210034
484390075
482570005
481211032
482510003
481391044
482311006
483670081
482210001
Years
07-08
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
08-09
07-09
07-09
07-09
Key
Tn
I
N
279
456
429
459
457
455
456
455
458
449
456
459
459
459
459
306
459
459
459
c
si ra
To |
H •
Mean
41
48
47
48
46
52
52
43
46
47
52
52
47
50
47
47
43
48
44
median
SD
14
15
16
16
15
16
14
14
16
13
15
16
12
13
15
12
12
14
15
iC
m™
1
Median
38
46
44
46
44
50
51
42
44
47
50
50
45
49
45
45
42
47
41
To *
h —
IQR Site
22
21
24
23
22
23
21
22
23
21
22
24
18
19
22
19
18
22
22
-O
A-
B -
c -
D-
E-
G -
H -
I -
J -
K-
L-
M -
N -
O -
Q-
R-
S -
C
I
1
I -:
\ 1
\ H
I;!]""1 -
^J i
.^•'
^| -i
.;
] -:
-A
-B
-C
-D
-E
-F
-G
-H
-I
- J
-K
-L
-M
-N
-O
-P
-Q
-R
-S
) 50 100 150
03 (ppb)
Figure 3A-41. Site information, statistics and box plots for 8-h daily max ozone from AQS
monitors meeting the warm-season data set inclusion criteria within the Dallas CSA.
March 2011
3-142
DRAFT - DO NOT CITE OR QUOTE
-------
Denver CSA
Site ID
noA-j-i nn.4*?^
080310002
080310014
080013001
080590002
080050002
080350004
080590006
080590013
080130011
080137001
080137002
081230009
Years
AQ AQ
07
07-09
07-09
07-09
n7 no
07,09
Q7— Q9
07-09
07-09
09
07-09
07
07
07-09
N
OOQ
153
450
441
459
306
456
457
150
453
152
142
451
Mean
ACt
39
51
55
54
55
54
58
60
50
56
42
56
55
SD
1 O
10
12
11
12
11
1 "?
11
12
9
12
10
11
11
Median
C 'I
41
52
57
56
^fi
55
58
59
50
56
42
56
56
IQR
1 C
13
15
13
16
IK
14
14
15
10
14
12
13
14
Site
B-
c-
D-
E-
G-
I-
J -
K-
L-
M -
N-
o-
I , , , , I
^^^
'•""ffl."\
JP
A
-B
-c
-D
-E
-G
LJ
-J
-K
-L
-M
-N
-o
Key
to
1
to
CM
H
mean
•
c
_ra
'"D
IU
1
_'C
cays
sis
o;E
,
1
£
to
hw.
h
£
»
OS
H
50 100
03 (ppb)
150
Figure 3A-42. Site information, statistics and box plots for 8-h daily max ozone from AQS
monitors meeting the warm-season data set inclusion criteria within the Denver
CSA.
Detroit CSA
Site ID Years N Mean SD Median IQR Site
261250001 07-09 459 46
261630019 07-09 456 47
260991003 07-09 452 47
07-09 459 42
07-09 459 45
260990009 07-09 459 46
260490021 07-09 458 44
261470005 07-09 459 43
07-09 455 45
14
15
15
13
13
15
13
15
Key
c c
c & 76;§
r: TO T3 mi^ -c
= rO £ P O'P !5 ^°
10 CM C b 0|C I*- (j)
1•:f----n
50 100
03 (ppb)
150
Figure 3A-43. Site information, statistics and box plots for 8-h daily max ozone from AQS
monitors meeting the warm-season data set inclusion criteria within the Detroit
CSA.
March 2011
3-143
DRAFT - DO NOT CITE OR QUOTE
-------
Houston CSA
Site ID
482010075
482010070
482010066
482010047
482010055
482010416
482010046
482011035
482010051
482010024
482011034
482010062
480391004
482010026
482011039
482011015
482010029
482011050
483390078
481671034
480391016
Years
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
Key
to
I
N
430
451
433
451
441
451
454
437
445
455
443
450
430
454
442
428
449
444
455
427
455
ra
to £
CM E
-i •
Mean
36
34
37
38
40
38
37
35
38
45
39
34
38
41
41
38
45
40
43
37
34
c
T3
E
I
SD
17
16
17
16
18
17
16
17
17
17
16
16
18
16
18
15
16
17
12
17
16
ll
gfe
§;E
I
1
Median
32
30
32
35
35
34
34
31
33
43
37
28
33
39
36
33
42
35
42
33
28
to
\
IQR Site
25
24
25
22
25
26
23
24
25
24
22
24
26
23
27
21
22
27
16
27
23
s
en
H
A-
p
c -
D-
E -
F-
G —
H -
I-
J -
If
L-
M -
N-
o-
P-
Q-
R-
s-
T-
U -
C
i i . i
:---! p
' ~~^P
i— C
H r*
H
£— -I-
t I"
'i |
• |
•
i f.
» |-
I \~
•
;---! ' |»
'--•\
r--
i i i i i i i i i
^
} H
_!
j • | -;
^Jr-[ '-----.
^J. .;
;--|
•
H |~~» I'-
•
H
-A
~~ R
-c
-D
-E
-F
— G
-H
-1
-J
~~ K
-L
-M
-N
-o
-Q
-R
-s
-T
-u
ii i i ,,,•,,,
) 50 100 150
03 (ppb)
Figure 3A-44. Site information, statistics and box plots for 8-h daily max ozone from AQS
monitors meeting the warm-season data set inclusion criteria within the Houston
CSA.
March 2011
3-144
DRAFT - DO NOT CITE OR QUOTE
-------
Los Angeles CSA
Site ID
060371602
060371301
060371302
060371103
060372005
060374002
060595001
060590007
060375005
060371002
060370002
060370113
060370016
060371701
060591003
060371201
060711004
060376012
060650004
060592022
061112002
060658005
060712002
060658001
061110007
060710012
060379033
061110009
060719004
060659001
060710005
060656001
060714003
060714001
060710306
061113001
061111004
061112003
060650009
060650012
060651016
060710001
060655001
060719002
060652002
060651999
060651010
060711234
060650008
060659003
Years
07-09
07-08
09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
08
07-09
07-09
08-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
09
07-09
07-09
07-09
07-09
07-09
07-09
07-08
09
07-09
07,09
07-09
Key
£
I-
N
458
306
152
457
459
459
459
459
459
459
459
459
458
459
459
459
457
457
127
457
455
276
459
440
459
456
452
458
457
453
459
455
459
455
459
453
458
457
153
457
459
455
459
452
448
283
153
453
265
444
c
j= ro
To g
IM E
H •
Mean
48
36
44
46
54
38
50
48
45
56
57
48
64
61
45
61
66
68
69
52
62
65
68
69
54
67
67
58
70
68
79
72
73
68
64
44
57
41
22
73
73
61
69
73
62
49
59
59
58
42
median
I
SD
13
9
10
12
15
10
12
10
9
14
17
10
18
16
9
14
19
18
18
13
12
15
19
16
10
13
13
11
19
16
19
17
18
14
12
9
11
9
8
15
16
11
14
13
13
17
10
10
10
10
overall
"median
I
1
Median
47
34
44
45
53
37
49
47
45
55
56
47
63
60
44
60
66
69
65
50
62
64
67
68
54
67
66
58
70
67
80
73
73
68
64
43
57
40
20
71
73
60
68
73
61
50
59
58
57
42
"to •
\
IOR Site
17
10
12
14
18
11
14
12
12
19
22
13
23
20
12
19
23
27
23
15
16
18
24
18
12
18
19
14
26
21
28
24
25
21
17
11
14
12
8
22
23
15
21
18
18
22
15
13
14
13
!o
H
A-
B-
C-
D-
E-
F-
G-
H-
I -
J-
K-
L-
M-
N-
o-
P-
Q-
R-
S-
T-
U-
V-
w-
X-
Y-
z-
AA-
AB-
AC-
AD-
AE-
AF-
AG-
AH-
Al -
AJ-
AK-
AL-
AN-
AO-
AP-
AQ-
AR-
AS-
AT-
AU-
AV-
AW-
AX-
C
i i i i i i i i 1 i i i i
hgfe....
, i — t — L
r r '
;"a^to---:
"'gjjjb -r H
' 1 . * r '
J ' J L
r i ' T r
• 1, | ™ I '•
^"^SE----; |
>>ijp::-
,--cg3""H
' i 4 i "'
m,
'
;---i--| 1 I H
n 1 '.
•
-------
Minneapolis CSA
Site ID
Years N Mean SD Median IQR Site
270031002 07-09
271390505 07-09
271636015 07-09
271713201 07-09
270031001 07-09
551091002 07-09
270495302 07-09
271453052 07-09
Key
lo
!____
456
459
439
416
455
457
454
453
c
£ ra
To '3^
o;c
1
41
42
42
42
38
42
44
39
lh ^
\ H
16
14
16
15
18
15
14
15
">
A-
B-
c-
D-
E-
F-
G-
H -
:- — | i
(....r—
,L---Q
t--\
| H
^|___j
> I- !
1 _ '
!--^rfn----:
-A
-B
-c
-D
-E
-F
-G
-H
50
100
03 (ppb)
150
Figure 3A-46. Site information, statistics and box plots for 8-h daily max ozone from AQS
monitors meeting the warm-season data set inclusion criteria within the Minneapolis
CSA.
New York CSA
Site ID
360810124
360610135
360050110
360050133
340030006
340170006
340130003
360850067
361192004
090010017
361030002
340315001
340250005
340230011
340273001
090019003
361030009
340190001
360790005
340210005
090013007
090011123
340290006
360715001
361030004
090090027
360270007
090093002
090050005
361111005
Years
07-09
08-09
07-09
07-09
08-09
07-09
09
07,09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
Key
to
(..--
N
446
298
457
459
300
442
122
298
444
447
454
445
458
459
456
457
890
455
459
456
457
459
456
459
453
456
456
459
446
459
C
CD
-« g
CN £
H •
Mean
43
39
40
41
42
45
36
45
46
49
47
45
47
48
48
47
47
50
44
49
49
46
51
45
48
41
43
47
46
41
c
TO
T3
---GIZ]--1H
,-VS?-T
v5i^
J---I j» 1- ^
^-nn-— ;
-A
-B
-c
-D
-E
— F
-G
-H
-J
— K
-L
-M
-N
-O
— P
-Q
_ C?
-S
-T
-u
-V
-w
-X
-Y
— ~7
-AA
-AB
-AC
-AD
) 50 100 150
O3 (ppb)
Figure 3A-47. Site information, statistics and box plots for 8-h daily max ozone from AQS
monitors meeting the warm-season data set inclusion criteria within the New York
CSA.
March 2011
3-146
DRAFT - DO NOT CITE OR QUOTE
-------
Philadelphia CSA
Site ID
421010014
421010004
340070003
420910013
340150002
421010024
420450002
100031013
100031010
420290100
340110007
100031007
420110006
420110011
Years
07
07-09
07-08
07-09
07-09
07-09
07-09
08-09
07-09
07-08
07-09
07-09
07-09
08-09
08-09
Key
N
153
459
298
433
429
458
455
304
450
457
458
450
306
306
c
£ ra
To $2
CM E
"\ •
Mean
50
39
51
50
49
48
48
53
50
50
49
44
47
median
I
SD
15
13
17
16
16
16
15
16
15
16
15
16
14
15
13
14
overall
mecfian
1
Median
48
38
51
50
50
49
49
47
48
51
50
50
48
43
46
lc '
ID— -
1QR
17
23
20
22
20
21
20
20
21
20
17
19
s
Site
A-
B-
c -
D-
E-
F-
G-
H -
I-
J-
K-
L-
M-
N -
o-
P-
50 100
03 (ppb)
150
Figure 3A-48. Site information, statistics and box plots for 8-h daily max ozone from AQS
monitors meeting the warm-season data set inclusion criteria within the
Philadelphia CSA.
March 2011
3-147
DRAFT - DO NOT CITE OR QUOTE
-------
Phoenix CBSA
Site ID
040133002
040133003
040139997
040131004
040134005
040134003
040130019
040137020
040137024
040137022
040132001
040137021
040134004
040131010
040137003
040132005
040139704
040135100
040134010
040134008
040139702
040139706
040213001
040213010
040213009
040217001
040139508
040134011
040213003
040218001
040213007
Years
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
Key
In
1
N
455
459
455
459
454
459
459
459
459
457
459
457
459
456
455
459
459
453
457
459
451
448
459
458
459
459
459
459
459
459
459
c
i <0
7n ®
c3 E
-\ •
Mean
53
57
56
58
55
55
55
56
56
56
53
59
56
55
52
57
58
55
48
58
53
58
59
45
48
52
57
46
52
59
50
c
ra
T)
1
|
SD
9
10
10
10
10
9
10
9
9
10
10
9
9
9
8
8
9
10
9
9
9
11
9
9
9
9
8
9
9
9
8
iC
2S
o|E
|
Median
53
57
56
58
55
55
55
56
57
56
53
59
56
55
52
57
59
57
48
57
53
58
59
46
47
51
56
46
52
59
50
1 *
t~~
\
IQR Site
11
12
12
13
13
12
12
12
12
13
13
12
12
13
12
11
11
14
13
14
10
14
12
11
12
11
11
12
11
12
10
H
A-
B-
c-
D-
E-
F-
G -
H-
I -
J -
K-
L-
M -
N-
o-
P -
Q-
R-
S-
T-
u-
V-
w-
x-
Y-
z-
AA-
AB-
AC-
AD-
AE-
, , , , ,,,,!,,,
h-
:---
r
•---
!•-
I--
l--
I--
'r-
'---
h-
(•-
;---
;---
i--
;----
;-"C
;---
i--
'r-
I"""L3
r--|_!
;---
I--
;---{
0
u
_ J~~i~
- 1
--C
-[ \
\
g
~ i i
a
-C
3
E
d
L i
6
a
T
Tl
"
50
j--J.
JZ1""":
D — ":
^ I---I
ZI---;
H"";
Zl~";
H~--;
D""":
ZJ"~";
D---;
^ |---;
\~~\---,
— •.
J]---;
j |---;
p 1 1
^ \---i
W \ ':
---;
3"""1
M""":
"}---;
3D""1
-A
-B
-c
-D
-E
__ C
-G
-H
-I
- J
-K
-L
-M
-N
-o
-P
-Q
-R
-s
-T
-U
-V
-w
-X
-Y
-Z
-AA
-AB
-AC
-AD
-AE
100 150
03 (PPb)
Figure 3A-49. Site information, statistics and box plots for 8-h daily max ozone from AQS
monitors meeting the warm-season data set inclusion criteria within the Phoenix
CBSA.
March 2011
3-148
DRAFT - DO NOT CITE OR QUOTE
-------
Pittsburgh CSA
Site ID
/tonn'a.nAAQ
420030010
420030067
421290006
4OAAO -I AAC
421250005
421255001
421250200
421290008
420070005
420070014
420070002
donneLfinni
420730015
Years
A7-.no
07-09
07-09
07-09
A7— AO
07-09
07-09
07-09
07-09
07-09
07-09
07-09
CI7_nQ
07-09
N
A CQ
459
449
459
A&n
459
449
459
456
452
459
452
Af\Q
459
Mean
AO
48
49
45
Kfi
48
48
46
47
47
46
49
^n
45
SD
13
12
13
H C
12
13
12
13
13
13
13
•t I- -:
] -:
^j. --.--;
[ -i
3 -I
3 — -i
3 — -;
3---;
| -;
3 — 1
i i i i
A
-B
-c
-D
^
c.
-F
-G
-H
-i
-J
-K
-L
M
^N
Key
to
h--
"LO
CM
-H
c .1
ra ID
qj o>
• I
overall
mecflan
1
to *„
h- o>
Z}--H
50
100
03 (ppb)
150
Figure 3A-50. Site information, statistics and box plots for 8-h daily max ozone from AQS
monitors meeting the warm-season data set inclusion criteria within the Pittsburgh
CSA.
Salt Lake City CSA
Site ID
490350003
490353007
490353006
490353008
490352004
490110004
490450003
490570007
490570002
490571003
490030003
490037001
Years
07-09
07
07-09
07
07-09
07-09
07-09
07
09
07-09
07-09
07-09
Key
I---H
N
459
153
458
149
450
459
453
153
153
449
454
425
c
ra
•
Mean
56
57
53
59
55
54
54
60
52
57
55
56
fa
0)
I
SD
12
11
11
11
11
12
9
12
9
10
9
8
iC
IS-
oE
,
Median
56
57
53
59
55
53
54
60
53
57
55
56
to -
1--
IQR Site
17
14
16
14
14
16
12
14
12
15
13
11
Lo
H
A-
B-
c -
D-
E-
F-
G-
H -
( -
J-
K-
L-
i i i i
:— C
: — Q
h--
:----[
;"""C!
' 1_
•""C
1--
<---[
:---
0 5
i i i i 1 i i i i
JU-<
np----1
k | — -:
'; t | i
4 | — :
fc |- ;
| 1----:
•| ; i 1 — -•
3D""1
4 |- - -:
4 [ - - -:
f (--•)
-A
-B
-C
-D
-E
-F
-G
-H
- |
-J
-K
-L
0 100 150
03 (ppb)
Figure 3A-51. Site information, statistics and box plots for 8-h daily max ozone from AQS
monitors meeting the warm-season data set inclusion criteria within the Salt Lake
City CSA.
March 2011
3-149
DRAFT - DO NOT CITE OR QUOTE
-------
San Antonio CBSA
Site ID
Years N Mean SD Median IQR Site
480290055 08-09 306 40 14 37
480290032 07-09 454 42 15 39
480290052 07-09 456 43 13 41
480290622 08-09 305 37 13 33
480290059 07-09 450 36 13 33
Key
In
h-
to
CM
-H
mean
*
median
1
overall
mecfian
to ^,
I3--H
50 100
03 (ppb)
150
Figure 3A-52. Site information, statistics and box plots for 8-h daily max ozone from AQS
monitors meeting the warm-season data set inclusion criteria within the San Antonio
CBSA.
San Francisco CSA
Site ID
060010009
OS0750005
060010006
060012004
060012001
060811001
060131004
060011001
060130002
060410001
060950006
060010007
060852007
060950004
060133001
060850005
060851001
060950005
060131002
060550003
060870006
060870003
060953003
060870007
060970003
060852006
060870004
060850002
060971003
060690002
060690003
Years
08-09
07-09
07-08
08-09
07-09
07-09
07-08
07-09
07-09
07-09
07-08
07-09
07-08
07-09
07-08
07-09
07-09
07-09
07-09
07-09
07-08
07-09
07-09
07-09
07-09
07-09
07-09
07-09
08
07-09
07-09
Key
i —
N
306
458
303
306
459
459
306
456
458
458
306
459
306
459
306
456
459
459
459
459
306
456
455
456
459
458
459
459
153
456
457
C
£L ro
K E
H •
Mean
29
28
31
25
35
31
31
34
42
29
40
43
34
35
41
36
39
39
47
37
38
31
44
33
31
44
33
44
39
42
54
c
T3
aj
E
I
SD
9
8
9
7
10
9
8
10
13
8
11
14
10
10
10
10
12
11
12
9
9
8
13
8
8
11
8
11
8
10
12
iC
Ig
IE
Median IQR Site
28
27
30
24
33
29
29
33
40
28
39
41
33
34
41
35
37
37
45
35
37
30
43
32
31
43
32
42
40
40
54
£
10
j-- -
12
10
12
10
12
11
12
12
18
10
13
18
13
12
12
13
16
12
15
10
11
10
17
10
10
14
10
15
13
12
16
CD
A-
B-
c -
D -
E -
F-
G-
H -
I -
J -
K-
L-
M -
N -
o-
p -
Q-
R -
S -
I'-
ll -
V-
w -
X-
Y-
z-
AA-
AB-
AC-
AD -
AE -
i i i i i i < t tiii
H \f fr--H
;--[ ^ |;---:
l--\ ^||--H
:"~CE]~'":
:---|~|j~[---H
r ' H I* il ' ' ~:
|.--| [» |---!
r--| |» |--H
;- - - | ! |» \ -;
[•-•{J1--H
f - - - ^ I ^
I---I; |» 1 H
;- - ! fr ; - - - H
;- - rnp.-- - H
H-t fr | i
:- - - 1 fr \ ----,
:- - '|« ;
;---[Tg} -;
;. _ ,. j |» | .;
:---| |» | -:
'r~\ ||» 1 i
r - -| |» ;|- - - -;
r--| ^ | ^
:---| frit----;
•---CB""H
-ICE--:
;--{jn---;
:--| |> 1 i
h-j; 4 I-H
r - - |» : -;
! 1 f \----i
-A
-B
-C
-D
-E
-F
-G
-H
-1
-J
-K
-L
-M
-N
-O
-P
-Q
-R
-S
-T
-U
-V
-w
-X
-Y
-Z
-AA
-AB
-AC
-AD
-AE
— \ — i 1 1 1 — i 1 1 — i 1 1 1 — i 1 —
0 50 100 150
03 (ppb)
Figure 3A-53. Site information, statistics and box plots for 8-h daily max ozone from AQS
monitors meeting the warm-season data set inclusion criteria within the
San Francisco CSA.
March 2011
3-150
DRAFT - DO NOT CITE OR QUOTE
-------
Seattle CSA
Site ID
530330080
530330010
530330017
con^7nno'5.
DoUoOUU^O
530670005
530531008
530531010
530530012
530570018
530570013
Years
07-09
07-09
07-09
n7— no
u / uy
07-09
07-09
07-09
07,09
07-08
07
N
452
456
432
AAA
*4*f*T
459
443
459
286
279
153
Mean
28
32
36
7P
OO
35
35
31
39
26
30
SD
8
12
12
10
12
11
9
8
10
Median
28
31
34
"3£
OO
34
33
30
38
25
30
IQR
11
16
17
1 7
i i
14
14
16
11
8
12
Site
A-
B-
c -
E-
F-
G-
H -
I -
J-
i i \ i 1 i i i r 1 i i i i
>-CC-i
:----T~|n-----:
!
,
'
--H
-;
;- - - 1 fa | 1'
h - -;|~^~|- - - -;
;- - £j} { - -;
i---rfl----;
-A
-B
-c
— r*
U
-E
-F
-G
-H
- i
-J
Key
^
to
1
i;
If)
CM
-i
C
(0
0)
•
ra
•o
H
50 100
03 (ppb)
150
Figure 3A-54. Site information, statistics and box plots for 8-h daily max ozone from AQS
monitors meeting the warm-season data set inclusion criteria within the Seattle
CSA.
St. Louis CSA
Site ID
oQK-i nnAQc
295100085
OQ1 Qonnn^d
291890014
171630010
290990019
290990012
OCH o-3-i nno
OQ-I o-^-i nn,i
171193007
171 iQnnna
171191009
291890005
i7nR"5inni
291130003
171170002
Years
H7— n&
07-09
H7 HA
07-09
07-09
08-09
07
ri7— nQ
n7— no
07-09
A7 OQ
07-09
07-09
ri7 no
07-09
07-09
N
ono
459
-ICQ
765
444
306
153
AAQ
A£, C!
458
^i^'?
458
755
4^3
457
457
Mean
CA
46
cn
48
43
49
55
51
^n
48
4R
50
46
4fi
50
46
SD
•1 C
14
H C
13
13
12
16
1 ^
14
1^
14
12
1 O
13
11
Median
en
46
C -1
48
44
49
53
ACt
AQ
48
Aft
49
46
AZ,
49
46
IQR
•4 Q
18
1 Q
16
17
16
19
1ft
m
17
1 7
18
16
1K
15
14
Site
B -
D -
E-
F-
G-
|
1
J -
L-
M -
O -
P
, , , , i , , , , i , , , ,
'• 1 L J
1 IF r
^---C33--H '
I----QO-- *
-;
,-on-/"
r,.c^...:,
^-rn--^
A
-B
u
-D
-E
-F
-G
LJ
-J
k"
-L
-M
M
-o
___ p
50
100
03 (ppb)
150
Figure 3A-55. Site information, statistics and box plots for 8-h daily max ozone from AQS
monitors meeting the warm-season data set inclusion criteria within the St. Louis
CSA.
March 2011
3-151
DRAFT - DO NOT CITE OR QUOTE
-------
3.8.4. Ozone Concentration Relationships for the Urban Focus Cities
1 This section contains histograms and contour matrices of the Pearson correlation coefficient
2 and the coefficient of divergence (COD) between 8-h daily max O3 concentrations from each
3 monitor pair within the 20 urban focus cities discussed in Section 3.6.2.1. These figures also contain
4 scatter plots of the correlation and COD as a function of straight-line distance between monitor pairs.
March 2011 3-152 DRAFT - DO NOT CITE OR QUOTE
-------
Atlanta CSA
20-
15-
10-
b-
6
13
25
-0.1
0.0 0.1
0.2 0.3
0.4 0.5
Correlation
0.6 0.7 0.8 0.9
1.0
1.0-
0.9-
0.8-
0.7-
0.6
I °'5
<3
0.3-
0.2-
0.1 -
0.0-
-0.1
0-87 0.89 0.85 0.89 083 0.88 0.75 0.76 -A
0.86 o.85 •: •:: c..?: o.ae 0.90 o.s? 0.74 0.75
80 0.79 0.77 0.73 0.75 0.78 0.79 0.68
0.90 0.82 0.77 0.81 D.61 0.88
- B
-G
0.63 0.70 - H
- I
- J
- K
0 50 100 150 200 250 300 350 400 450
Distance (km)
Figure 3A-56. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the Atlanta
CSA. [The colors in the histogram bins correspond to the levels of the contour
matrix. The histogram includes the number of monitor pairs per bin and the contour
matrix includes the numeric values of the correlations.
March 2011
3-153
DRAFT - DO NOT CITE OR QUOTE
-------
Baltimore CSA
200-
^ 150-
E iooH
o
50 H
2 9
88
209
-0.1
0.0
0.1
0.2
0.3
0.4 0.5
Correlation
0.6
0.7
0.8
70
0.9
1.0
< OQ
-0.1
50 100 150 200 250 300 350 400 450
Distance (km)
Figure 3A-57. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the
Baltimore CSA.[The colors in the histogram bins correspond to the levels of the
contour matrix. The histogram includes the number of monitor pairs per bin and the
contour matrix includes the numeric values of the correlations.
March 2011
3-154
DRAFT - DO NOT CITE OR QUOTE
-------
Birmingham CSA
25-
„ 20-
§ 15-
8 10-
5-
1
27
17
-0.1 0.0 0.1 0.2 0.3 0.4 0.5
Correlation
0.6 0.7 0.8 0.9
1.0
O
1.0-
0.9-
0.8-
0.7-
0.6
0.5 -
m
8
0.3-
0.2-
0.1 -
0.0-
-0.1
D.8B 0.88 O.S6 0.89 0.86 D
-A
-C
-D
-E
-F
-G
-H
-J
0 50 100 150 200 250 300 350 400 450
Distance (km)
Figure 3A-58. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the
Birmingham CSA. [The colors in the histogram bins correspond to the levels of the
contour matrix. The histogram includes the number of monitor pairs per bin and the
contour matrix includes the numeric values of the correlations.
March 2011
3-155
DRAFT - DO NOT CITE OR QUOTE
-------
Boston CSA
40-
20 -
-C
.1 0
2
0 0.1 0.2 0.3 0.4 0
Correlator
10
5 0
l
22
6 0
61
7 0
80
8
ID I _
O cr co
1.0-
0.9-
0.8-
0.7-
0.6
I 0.5 -
0.4
0.3
0.2-
0.1 -
o.o-
-0.1
0.90
0.91
076 0.86 0.69 074
079 0.90 0.78 0.81
077 0.90 0.73 080
0.80 0.33 0.76 0.85
0.86 0.80 0.62 O.BO
077 0.84 0.82 0.59 077
0.88 0.84 0.89 0.63 075
082
0.89 077
• •
• .•
092 052 :!"-:• 091
0.90 0.36 0 90 0 79 0.88
0.86 0 73 0 75 0.89 0. • 0 74 0.85 0.78 0.55 0.73 I
0.87 ^H 0.86 0 73 0.85 0.78 0.73 0.84
0.75 0.79 0.77 I I 0.71 I I 0.65 0 70
0.71 073 0.73 0.72 0.31
0.76 0.88 0.79 0.64 0.83
74 0.87 0.80 0.54 0.69
I 0.67 0.69
0 59 0.70
0.68 0.68
J.78
0.69 0.71
0.81 0.74
0.74 0.74
0.79 0.72
0.65 0.80
0.61 0.83
-B
-c
-D
O.B4
0.71 I I I
0.78,
-o
-Q
0.85 0.60
50 100 150 200 250 300
Distance (km)
350 400
450
Figure 3A-59. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the Boston
CSA. [The colors in the histogram bins correspond to the levels of the contour
matrix. The histogram includes the number of monitor pairs per bin and the contour
matrix includes the numeric values of the correlations.
March 2011
3-156
DRAFT - DO NOT CITE OR QUOTE
-------
Chicago CSA
o
O
150-
100-
50-
-c
5
.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0
Correlation
69
7 0
199
^B
8 0
< m O Q
O i _
X >- N
77 0.86 0.90 0.88JHiO 86 0.87 0.87 0.83 0.86 0.80^10.87 0.83 0.87 0.87 0 80 0.85 0.80 0.80 0.77 0.85 0.83 |- A
0.30 0.91 0 94 0.91 0 92 0 94 0.89^>0.87 0.83 0 86^^^|0 88 0.90^H|0 38 0.88 0.86 0.34 0.79 0.84 0.87 |- B
0 78 0.88g ) 38^ 0.89 C =8 C 3':. 0 91 0 86 O.B7 0.85 0.88 0.81 0.85 0.89 I- C
0.81 0 79 0.76 075 0.80 0.76 0.81 0 80 0.70 074 0.76 0.80 0 74 0.79 0.76 0.69 0.78 0.73 0.76 0.72 077 0.73 L- D
0.80 0.86 0.88 0.85 0 88 0.85 0.89 0 87 0.88 0.87 0.81 0.77 0.83 0.35 [• E
87^Ho.89 0.80 D.9oHlo.B6| J90 O.B9 0.88 0.85 0.78 0.84 0.88 \- F
94 •:• 94 C 51 3.74 0.84 O.SsB i 0 84 0.82 0.81 0.85 0.87 077 0.84 L G
I 0.84 0,87 - H
C
g
"S
-------
Dallas CSA
80-
•£ 60-
o 40-
20
-0.1
0.0
0.1
0.2
0.3
0.4 0.5
Correlation
CL O
CO
0.70
0.84 0.89 H 0.89 0.89 0.85 0.84 0.77
0.88 C80 0.93 0.90 0.89 079 036 0.37 0.71 0.67
0.90 096 090 I:-J4^H 0.89 I 3 091 0.30 0.83 0.78
09C C76 0.95 0.95 065 079 0.85 0.33 0.69 I
Correlation
1.0-
0.9-
0.8-
0.7-
0.6-
0.5-
0.4-
0.3-
0.2-
0.1 -
o.o-
n 1
038 0.85 0.96 058 0.92 097 0.33 0.81 0.63
"i"*£, 0.73 0.82 0.90 1 1 0 85 0.80 1 1 0.67
'..•^•S^ • i 0.89 0.84 0.75 081 0.37 0.65 0
* .*4**r •*•**«. • 3 0.32 0.73 0.87
•«."*•
» !• . • 0.85 0.85 0.89 0.82 0.7c
. '. '.
V- 0 074 0.88 0.71
0.81 0.79 Q.82
0.73 0.60
061
0.79
0.80
0.77
B
C
D
E
- F
-G
- H
J
-K
-L
-M
-N
-O
-P
-Q
-R
-S
50 100 150 200 250 300
Distance (km)
350 400
450
Figure 3A-61. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the Dallas
CSA. [The colors in the histogram bins correspond to the levels of the contour
matrix. The histogram includes the number of monitor pairs per bin and the contour
matrix includes the numeric values of the correlations.
March 2011
3-158
DRAFT - DO NOT CITE OR QUOTE
-------
Denver CSA
50-
„ 40-
c --
= 30 -
8 20-
10-
3
-0.1
0.0
0.1
0.2
0.3
0.4 0.5
Correlation
0.6
0.7
0.3
Q LU
1.0
0.9
0.8
0.7
0.6
| 0.5
p
8 OA
0.3
0.2
0.1
0.0-I
-0.1
•»•
'$£
0.76
0-90 0.39 0.86 0.87 0.85 0.84 076 0.82 0.74 0.84 0.78
0.86 0.80 0.80 0.82 0.81 Q.81 0.66 0.79 0.72 0.89 0.71
O.Sb 09J I E7
0.77 0.88 0.79 0.88 0.7S
: 92 C.39 0.81 0.87 0.77 0.84 0 78
C.31 0.83 0.78 0.66 0.77 0.77
C.93 076 0.89 0.80 0.39 0.76
C.JS 0.89 0.64 07C C.35 0.76
0.77
A
B
C
D
E
F
G
H
J
K
L
M
N
O
0 50 100 150 200 250 300 350 400 450
Distance (km)
Figure 3A-62. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the Denver
CSA. [The colors in the histogram bins correspond to the levels of the contour
matrix. The histogram includes the number of monitor pairs per bin and the contour
matrix includes the numeric values of the correlations.
March 2011
3-159
DRAFT - DO NOT CITE OR QUOTE
-------
Detroit CSA
"r*
0
O
20-
15-
10-
5-
-c
,3
.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Correlation
-------
Houston CSA
80-
•g 60-
o 40-
O
-C
.1 0
2
0 0.1 0.2 0.3 0.4 0
Correlator
m
5 0
1
6 0
33
7 0
88
8 0.
ID I _
O cr co
8
1.0-
0.9-
0.8-
0.7-
0.6
I 0.5 -
0.3
0.2-
0.1 -
o.o-
-0.1
.34 0.83 0.89 0.88 0.83 083 061 0.74
0.83 0.85 062 0.73
::
C..39 C 81 0.87 0.81 0.82 0.87 0 53 0.80 0.33
•ft
0.86 C.79 089 0.87 0.87 0.89 078 070 0.65 0.67
0.86 0.83 0.86 0 53 0.78 0.78
0,87 0.79 0.88 053 0.76 0.76
086 0.86 0.83 080 069 0.65 0.65
0 S? 0 91 0.91 C 55 • 11 0.89 0.81 0.84 0 57 0.76 0.75
: C 83 0.92 0 95 0 78 0 87 Q 5i 0.80 0.
0.76 0.68 0.90 0-82 0.84 0.36 076 0.77 0.57 0.68
.84 0.75 I I 0.88 0. I 0.83 0.80 0.68 0.64 0.63
0 83 I I 0.86 0.76 0.89 0 56 0.80 0.!
0.77 0.88 0.78 0.73 0.86 0 51 0.82 0..
0.80 0.84 068 0.66 0.66
I
0.61 0.81 0.79
0.77 0.84 0.63 0.69 0.70
73 0.77 0.65 0.66
0 57 O.B7
17 0.81
0.46 0.46
A
B
-C
-D
E
-F
-G
-H
50 100 150 200 250 300
Distance (km)
350 400
450
Figure 3A-64. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the
Houston CSA. [The colors in the histogram bins correspond to the levels of the
contour matrix. The histogram includes the number of monitor pairs per bin and the
contour matrix includes the numeric values of the correlations.
March 2011
3-161
DRAFT - DO NOT CITE OR QUOTE
-------
Los Angeles CSA
150
§ 100-
0
0 50-
-c
4
.1 0
62
0 0
iin
1 0
2 0
164
3 0
(
148
4 0
;orrelatior
148
5 0
1
150
6 0
144
7 0
87
8 0.9
29
1.0
1.0-
0.9-
0.8-
0.7-
0.6
I 0.5 -
° 0.4
0.3-
0.2-
0.1 -
o.o-
-0.1
„«,«.
,,.,,-u
o-eo Wfe»o,ao
.»..
iJB^uwu^MiwiBWd avm
: M -"^H «•) u.n - • "• "M u
-------
Minneapolis C5A
15-
I io-
Q 5-
2
~
L
•
-0.1 0.0 0.1 0.2 0.3 0.4 0.5
Correlation
0.6
0.7
0.8 0.9 1.0
< CD O Q LU
o
1.0-
0.9-
0.8-
0.7
0.6
| 0.5 -
8
0.3-
0.2-
0.1 -
o.o-
-0.1
-A
-B
-C
- D
- E
- F
-G
0 50 100 150 200 250 300 350 400 450
Distance (km)
Figure 3A-66. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the
Minneapolis CSA. [The colors in the histogram bins correspond to the levels of the
contour matrix. The histogram includes the number of monitor pairs per bin and the
contour matrix includes the numeric values of the correlations.
March 2011
3-163
DRAFT - DO NOT CITE OR QUOTE
-------
New York CSA
150-
100
50-
-0.1
0.0
0.1
0.2
0.3
0.4 0.5
Correlation
0.6
0.7
< CO O Q
1.0-
0.9-
0.8-
0.7-
0.6
I 0.5 -
J5
ffi
3 °-
0.3
0.2-
0.1 -
o.o-
-0.1
f f ••«• *\
» . v, .' • .' •
..
-A
B
D
E
-F
G
H
I
-J
-K
-L
M
N
rO
P
-Q
-R
S
-T
-U
-V
-w
-X
-Y
-z
•• ** Ir AA
AB
AC
AD
:f
0 50 100 150 200 250 300 350 400 450
Distance (km)
Figure 3A-67. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the
New York CSA. [The colors in the histogram bins correspond to the levels of the
contour matrix. The histogram includes the number of monitor pairs per bin and the
contour matrix includes the numeric values of the correlations.
March 2011
3-164
DRAFT - DO NOT CITE OR QUOTE
-------
Philadelphia C5A
60-
40
20-
pn
15
-0.1
0.0
0.1
0.2
0.3
0.4 0.5
Correlation
0.6
0.7
0.8
_Q
0.9
1.0
0
0.6
I 0.5 -
J5
ffi
8 OA
0.3
0.2-
0.1 -
o.o-
-0.1
0.82 038 0.86 0.82 0.85 0.82
031 C38 D.BG 085 0,89 0,84
03' 09t.l 083 0.81 0.84 0,89
0.89 BH9 0.88 0.79 0.31 C 32 080 0.81 0.82 078
H 0.90 0.89 0.82 0.89 0.88 0.81 0.83 0.87
097 0.92 044 '-1 0.85 0.90 0.36
0.91 0,87 084 090 0.89 0.62 0.83 0.37
035 091 091 064 0.91 033
0.76 089 0.88 0.80 0.87 030
077 0.78 0.86 0.82 0.69
0.78 0.85 0.90
0.60 0.88 0.35
0.86 0,67
0.74
0.84
C 85
091
C 80
0.88
0.87
089
084
0.81
0.71
•
0.86
071
077
I
-A
-B
-C
-D
- E
-F
-G
-H
-I
-J
-K
-L
-M
-N
-o
-P
-Q
50 100 150 200 250 300 350 400 450
Distance (km)
Figure 3A-68. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the
Philadelphia CSA. [The colors in the histogram bins correspond to the levels of the
contour matrix. The histogram includes the number of monitor pairs per bin and the
contour matrix includes the numeric values of the correlations.
March 2011
3-165
DRAFT - DO NOT CITE OR QUOTE
-------
Phoenix CBSA
- ioo H
O
O 50 H
1
30
125
145
122
-0.1
0.0
0.1
0.2
0.3
0.4 0.5
Correlation
0.6
27
0.7
0.8
0.9
1.0
< CO O Q LU
o
O
1.0-
0.9-
0.8
0.7-
0.6-
0.3-
0.2
0.1 -
o.o-
-0.1
...
•-
:::
A
B
C
D
E
F
G
H
-I
J
K
L
M
N
O
P
Q
R
S
T
U
v
w
X
Y
AA
AB
AC
AD
AE
50 100 150 200 250 300
Distance (km)
350 400 450
Figure 3A-69. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the Phoenix
CBSA. [The colors in the histogram bins correspond to the levels of the contour
matrix. The histogram includes the number of monitor pairs per bin and the contour
matrix includes the numeric values of the correlations.
March 2011
3-166
DRAFT - DO NOT CITE OR QUOTE
-------
Pittsburgh CSA
40
1 30-
Q 20-
10
-0.1
0.0
48
42
0.1
0.2
0.3
0.4 0.5
Correlation
0.6
0.7
0.3
0.9
1.0
1.0-
0.9-
0.8-
0.7
0.6
| 0.5 -
8
0.3-
0.2-
0.1 -
o.o-
-0.1
V
0.93 0.90 0.91 0.95 u.«7 0.90 0.88 0.89 0.87
0.81 0.34 0.82 0.89 0.87 0.87 0.80 0.92 0.83.
I
'1 94 C ;•: 053 0.85 0.85 0.84 0.80
C 89 0 91 0.91 0.93 0.88 0.86
A
B
-C
D
-E
-F
-G
- H
-J
-K
L
M
N
0 50 100 150 200 250 300 350 400 450
Distance (km)
Figure 3A-70. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the
Pittsburgh CSA. [The colors in the histogram bins correspond to the levels of the
contour matrix. The histogram includes the number of monitor pairs per bin and the
contour matrix includes the numeric values of the correlations.
March 2011
3-167
DRAFT - DO NOT CITE OR QUOTE
-------
Salt Lake City C5A
25-
c 2°-
g 15-
0 10-
5-
3
21
29
-0.1
0.0
0.1
0.2
0.3
0.4 0.5
Correlation
0.6
0.7
0.3
10
in o a LU
1.0-
0.9-
0.8-
0.7
0.6
| 0.5 -
8
0.3-
0.2-
0.1 -
o.o-
-0.1
t ».-• ' •
!. :
•-. .•••*•
94 0.91 3.S?. 0.91 0.91 0.89
0.79 0.77 0.67
0.83 081 0.7"
0.79 077 0.64
0.76 0.73 0.62
.85 O.B4 0.76 0.77 077 0.72
o.84 : :: 0.81 o.B4 0.82 0.73
0.88 0.74 0.82 080 0.77
0.92 0.85
0.90 086 0.75
0.89 0.86
-A
B
C
-D
E
-F
-G
H
-J
-K
-L
0 50 100 150 200 250 300 350 400 450
Distance (km)
Figure 3A-71. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the
Salt Lake City CSA. [The colors in the histogram bins correspond to the levels of the
contour matrix. The histogram includes the number of monitor pairs per bin and the
contour matrix includes the numeric values of the correlations.
March 2011
3-168
DRAFT - DO NOT CITE OR QUOTE
-------
San Antonio CBSA
5-
3-
2-
1 -
-0
.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0
Correlation
o
1.0-
0.9-
0.8-
0.7
0.6
| 0.5 -
8
0.3-
0.2-
0.1 -
o.o-
-0.1
-A
0 50 100 150 200 250 300 350 400 450
Distance (km)
Figure 3A-72. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the
San Antonio CBSA. [The colors in the histogram bins correspond to the levels of the
contour matrix. The histogram includes the number of monitor pairs per bin and the
contour matrix includes the numeric values of the correlations.
March 2011
3-169
DRAFT - DO NOT CITE OR QUOTE
-------
-0.1
0.0
San Francisco CSA
80-
bU-
4U -
2
21
43
64
92
92
36
0.1
0.2
0.3
0.4 0.5
Correlation
0.6
0.7
0.8
0.9
1.0
< CO O Q LU
-0.1
0 50 100 150 200 250 300 350 400 450
Distance (km)
Figure 3A-73. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the
San Francisco CSA. [The colors in the histogram bins correspond to the levels of
the contour matrix. The histogram includes the number of monitor pairs per bin and
the contour matrix includes the numeric values of the correlations.
March 2011
3-170
DRAFT - DO NOT CITE OR QUOTE
-------
Seattle CSA
8-
1 6"
0
2-
1
4
3
3
7
9
7
9
-0.1 0.0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Correlation
CD o
0.88 0.76 0.81 0.80 0.6
-A
-B
-C
1
p
b
O
1.0-
0.9-
0.8-
0.7-
0.6-
0.5-
0.4-
0.3-
0.2-
0.1 -
o.o-
n 1
^
0.82 0.84 0.61
•
• •* 0.30 0.59
•
-. • ' ^
.' • • .
.
% • • •
.
•
•
•
. • .
•
•
0.26 0.65
0.38 0.64
0.25 0.62
0.59
-J
0 50 100 150 200 250 300 350 400 450
Distance (km)
Figure 3A-74. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the Seattle
CSA. [The colors in the histogram bins correspond to the levels of the contour
matrix. The histogram includes the number of monitor pairs per bin and the contour
matrix includes the numeric values of the correlations.
March 2011
3-171
DRAFT - DO NOT CITE OR QUOTE
-------
St. Louis CSA
50
40
30-
20-
10-
-0.1
0.0
0.1
0.2
0.3
0.4 0.5
Correlation
CQ O Q
1.0-
0.9-
0.8-
0.7-
0.6
I 0.5 -
J5
ffi
8 OA
0.3
0.2-
0.1 -
o.o-
-0.1
•....•
*/.*.
076 0.79 0.78 0.78
0.84 0.86 0.86 0.87 089 0.79 0.85 0.76
0.90 0.88 0.89 0.86 0.89 0.85
0.79
033 0.80 0.86 0.85 0.88 076 0.76 0.76 0.75
0.79 0.77 0.77 0.77 0.87 0.77 0.79 0.72
0.83 0.85 0.86 0.87 089 0.82 0.84 0.79
0.93 0.97 090 C 77
085 C7I3 0.94 0.92 0.83
L'.Bb U.y^ U.ttD
0.91
C 76 0.36 0.83 0.83
0.85 0.87
078 0.32 0.80 0.83
0.72 0.83 0.72
0.82
-B
-C
- D
E
-F
-G
- H
-J
-K
-L
- M
-N
-O
50 100 150 200 250 300 350 400 450
Distance (km)
Figure 3A-75. Pair-wise monitor correlations expressed as a histogram (top), contour matrix
(middle) and scatter plot versus distance between monitors (bottom) for the St.
Louis CSA. [The colors in the histogram bins correspond to the levels of the contour
matrix. The histogram includes the number of monitor pairs per bin and the contour
matrix includes the numeric values of the correlations.
March 2011
3-172
DRAFT - DO NOT CITE OR QUOTE
-------
Atlanta CSA
30-
~ 25-
= 20-
o 15-
0 10-
5-
0.00 0.05 0.10 0.15
0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40 0.45 0.50 0.55
CD
0.55
0.50-
0.45-
0.40-
0)
I °-35 H
P!
m
5 0.30-
2
1 0.25-
'o
3=
-------
Baltimore CSA
200-
150-
100-
50-
0.00 0,05 0.10 0.15 0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40
0.45
0.50
0.55
0.55
0.50-
0.45-
0.40-
a>
I °-35 H
E>
-------
Birmingham CSA
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40
0.45
0.50
0.55
O
0.55
0.50-
0.45-
0.40-
o
5 0.30-
1 0.25-
'o
3=
03
O
0.20-
0.15-
0.10-
0.05-
0.00
0.09 0.09 0.09 c ja : i a.os
0.05 0.08 0.06 0.09 : 1 3.OS
50 100 150 200 250 300
Distance (km)
350 400 450
500
-A
-C
-F
-H
-J
Figure 3A-78. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle)
and scatter plot versus distance between monitors (bottom) for the Birmingham
CSA. [The colors in the histogram bins correspond to the levels of the contour
matrix. The histogram includes the number of monitor pairs per bin and the contour
matrix includes the numeric values of the CODs.
March 2011
3-175
DRAFT - DO NOT CITE OR QUOTE
-------
Boston CSA
100-
~ 80-
g 60-
o 40-
20-
0.
2
DO 0.
75
^^H
D5 0.
0,55-
0.50-
0,45-
0,40-
0)
| 0.35-
i
S 0.30-
| 0,25-
'o
E
-------
Chicago CSA
150
| 100-
o
0 50 H
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40
0.45
0.50
N
10.07 0 07 0.09 0 D7 0 10 0.09 0.08 0.09 0.09 0.10 0.07 0.07 0.09 0.07 0.08 0 09 0.08 0.09 0.10 0.10 0
0.09 0 08 0.09 0 05 0.11 0.080090 100.11 0100.07 0.080 10 0.09006010 0.100.11 C.09 0.10 0
11 0.14 01; t
0 0.11 0.07 0.07 0.09 0.11 0.06 0.09 0.08 0.08 0.080.09 0.09 0.09 0.10 0.10 008 0.09
I CCS 0.100.100.080.090.080.090.090080.100.11 C 100
0.100.120 100.13 :
.100.11 0.10
OB 0 D£
0.55-
0.50-
0.45-
0.40-
5 0.30-
0
+-*
S 0.25-
'o
i
0)
° 0.20-
0.15-
0.10-
0.05-
n nn
U.U' I J 1 1 U.U/ U.'Ut U IU U 1
0.11 013008
1 1 0 C7 0.09 0 09 0.0;'
J11 0.090.100 10 0.100 'DC 10
• *
A •• • •»*
ty*K&W$* ******
V<£*v % **
•"•
^•0.10 0.10 011 0.09 0.12 0 12
0.100.12011 0.12 .
0.05
^
1 0.11 C i 090.09
0.080090.11 0.090090.10
0.090 1C 0.08 0.10 0.09 D.06
OOi 0.11 0.12011
-J
-K
-L
-M
-N
-O
-P
-Q
R
-S
-T
-U
-V
w
X
Y
z
50 100 150 200 250 300
Distance (km)
350 400 450
500
Figure 3A-80. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle)
and scatter plot versus distance between monitors (bottom) for the Chicago CSA. D
The colors in the histogram bins correspond to the levels of the contour matrix. The
histogram includes the number of monitor pairs per bin and the contour matrix
includes the numeric values of the CODs.
March 2011
3-177
DRAFT - DO NOT CITE OR QUOTE
-------
Dallas CSA
80-
0.00
0.55-
0.50-
0.45-
0.40-
s
§ °-35~
C3>
5 0.30-
o
I 0.25-
'o
I
0.20-
0.15-
0.10-
0.05-
0.00
0.05
0.10 0.15 0.20 0.25 0.30 0.35
Coefficient of Divergence
-------
Denver CSA
60-
- 50-
§ 40-
o 30-
0 20-
10-
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40 0.45 0,50 0.55
0.55
0.50-
0.45-
0,40-
01
I °-35 H
E>
0)
5 0.30-
I 0.25
'
Q 0.20-1
0.15-
0.10-
0.05-
0.00
A
B
C
D
E
F
G
H
J
K
L
M
N
O
50
100 150
200 250 300
Distance (km)
350 400 450
500
Figure 3A-82. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle)
and scatter plot versus distance between monitors (bottom) for the Denver CSA. D
The colors in the histogram bins correspond to the levels of the contour matrix. The
histogram includes the number of monitor pairs per bin and the contour matrix
includes the numeric values of the CODs.
March 2011
3-179
DRAFT - DO NOT CITE OR QUOTE
-------
Detroit CSA
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40
0.45
0.50
0.55
m
o
0.55
0.50-
0.45-
0.40-
0)
I °-35 H
P!
o
5 0.30-
1 0.25-
'o
3=
-------
Houston CSA
100-
~ 80-
E 6
u 40-
20-
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40
0.45
0.50
0.55-
0.50-
0.45-
0.40-
a 0.30
o
1 0.25
'o
I
° 0.20
0.15
0.10
0.05
0.00
0 50 100 150 200 250 300 350 400 450 500
Distance (km)
Figure 3A-84. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle)
and scatter plot versus distance between monitors (bottom) for the Houston CSA.
The colors in the histogram bins correspond to the levels of the contour matrix.
The histogram includes the number of monitor pairs per bin and the contour matrix includes
the numeric values of the CODs.
March 2011
3-181
DRAFT - DO NOT CITE OR QUOTE
-------
400
_ 300 H
c
o 200-
O
Los Angeles CSA
16
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55
Coefficient of Divergence
0.00
0 50 100 150 200 250 300 350 400 450
Distance (km)
500
Figure 3A-85. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle)
and scatter plot versus distance between monitors (bottom) for the Los Angeles
CSA. [The colors in the histogram bins correspond to the levels of the contour
matrix. The histogram includes the number of monitor pairs per bin and the contour
matrix includes the numeric values of the CODs.
March 2011
3-182
DRAFT - DO NOT CITE OR QUOTE
-------
Minneapolis C5A
20-
| 15-
3 10-
5-
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40
0.45
0.50
0.55
0.55
0.50-
0.45-
0.40-
0)
I °-35 H
P!
o
5 0.30-
1 0.25-
'o
3=
-------
New York CSA
250-
~ 200-
I 15°-
O 100-
50-
0.00
0.05 0.10 0.15 0.20 0.25 0.30 0.35
Coefficient of Divergence
N < < < <
0.55
0.50-
0.45-
0,40-
0.35-
0.30-
0.25-
0.20-
0.15-
0.10-
0.05-
0.00
50 100 150 200 250 300
Distance (km)
350 400
450
500
Figure 3A-87. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle)
and scatter plot versus distance between monitors (bottom) for the New York CSA.D
The colors in the histogram bins correspond to the levels of the contour matrix. The
histogram includes the number of monitor pairs per bin and the contour matrix
includes the numeric values of the CODs.
March 2011
3-184
DRAFT - DO NOT CITE OR QUOTE
-------
Philadelphia C5A
0.00
0.05
0.10
.15
0.55
0.50-
0.45-
0.40-
01
I °-35 H
s>
0)
5 0.30-
1 0.25-
'o
s=
0)
Q 0.20 H
0.15-
0.10-
0.05-
0.00
0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40
0.45
0,50
a
0.08 0.08 0 0
006 0
COS 008 008 0 '",09 009 0.10 0.1D 012 01
0.05 0.08 0.08 0.10 007 0.08 .' 009 OC
i I5 007 010 006 0.07 0.10 0.09 0.09 00
0.08 0.09 0.07 OQ8 0.10 0.11 0.09 00
0.06 0.10 007 003 0.10 0.07 0.09 00
007 •: uoe 010 oo
0 1C 0.09 0.11 0.13 01
o.oa O.OB oo
D.OS 0.10 00
010 0,13 0
50 100 150 200 250 300 350 400 450 500
Distance (km)
0.55
0,12
A
B
C
D
E
F
G
H
J
K
L
M
N
O
P
Q
Figure 3A-88. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle)
and scatter plot versus distance between monitors (bottom) for the Philadelphia
CSA. [The colors in the histogram bins correspond to the levels of the contour
matrix. The histogram includes the number of monitor pairs per bin and the contour
matrix includes the numeric values of the CODs.
March 2011
3-185
DRAFT - DO NOT CITE OR QUOTE
-------
Phoenix CBSA
0.00 0,05 0.10 0.15 0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40
0.45
0.50
0.55
< CO O Q LU
0.55
0.50-
0.45-
0,40-
01
I °-35 H
E>
-------
Pittsburgh CSA
80 H
0.00 0.05 0.10
0.15
0.55-
0.50-
0.45-
0.40-
s
§ °-35~
g>
5 0.30-
o
I 0.25-
'o
I
0.20-
0.15-
0.10-
0.05-
0.00
..
0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40
0.45 0.50
< CD O O LU
O x
006 0.07 0.08 G.07 0.07 007 0.06 0.07 0.08 0.07 0.07
0.07 0.08 0.06 0.07 C 06 0 06 0 07 0.07 0.07 0.07
0.06 0.07 0.06 007 0.07 0.09 0.06 0.08
0 0.07 0.08 [ '5 0,09 0.08 0.09 0.09
0 10 C- OB 0.09 0.10 0.10 0.07
J, 0.08 0.09
17 G08 007 008 0.06 0.08
0 03 0.37 0.09
o.os o.os
0.07 0.08
0.08 0.08
0 50 100 150 200 250 300 350 400 450 500
Distance (km)
A
B
C
D
E
F
G
H
J
K
L
M
N
Figure 3A-90. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle)
and scatter plot versus distance between monitors (bottom) for the Pittsburgh CSA.D
The colors in the histogram bins correspond to the levels of the contour matrix. The
histogram includes the number of monitor pairs per bin and the contour matrix
includes the numeric values of the CODs.
March 2011
3-187
DRAFT - DO NOT CITE OR QUOTE
-------
Salt Lake City CSA
40-
0.00
0.05
0.10
0.15 0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40 0.45
0.50
TsS
O
LU
0.55-
0.50-
0.45-
0.40-
o
i °-3£H
D)
1
5 0.30H
I 0.25-
0.20-
0.15-
0.10-
0.05-
0.00
i- f*
-A
-B
-C
-D
E
F
G
H
- J
K
-L
0 50 100 150 200 250 300 350 400 450 500
Distance (km)
Figure 3A-91. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle)
and scatter plot versus distance between monitors (bottom) for the Salt Lake City
CSA. [The colors in the histogram bins correspond to the levels of the contour
matrix. The histogram includes the number of monitor pairs per bin and the contour
matrix includes the numeric values of the CODs.
March 2011
3-188
DRAFT - DO NOT CITE OR QUOTE
-------
San Antonio CBSA
5-
t: 4-
•j •
o •
O 2-
1 -
0.00
0.55-
0.50-
0.45-
0.40-
Ol
o
d 0.30-
"o
1 0.25 H
'o
a=
0.20-
0.15-
0.10-
0.05-
0.00
0.05
0.10
0.15
0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40
0.45
0.50
0.55
CD
O
_u
-A
-C
-D
0 50 100 150 200 250 300 350 400 450 500
Distance (km)
Figure 3A-92. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle)
and scatter plot versus distance between monitors (bottom) for the San Antonio
CBSA. [The colors in the histogram bins correspond to the levels of the contour
matrix. The histogram includes the number of monitor pairs per bin and the contour
matrix includes the numeric values of the CODs.
March 2011
3-189
DRAFT - DO NOT CITE OR QUOTE
-------
San Francisco CSA
150-
g 100-1
o
0 50H
1
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50
Coefficient of Divergence
0.55
0.50-
0.45-
0.40-
0.35-
0.30-
0.25
0.20-
0.15-
0.10-
0.05-
0.00
-------
Seattle CSA
0.00
0.05 0.10 0.15
0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40
0.45 0.50 0.55
o
LU
0.55-
0.50-
0.45-
0.40-
D)
I
O 0.30
O
1 0.25
o
° 0.20
0.15
0.10
0.05
0.00
-A
-B
-C
-D
-E
-F
-G
-H
-J
0 50 100 150 200 250 300 350 400 450 500
Distance (km)
Figure 3A-94. Pair-wise monitor COD expressed as a histogram (top), contour matrix (middle)
and scatter plot versus distance between monitors (bottom) for the Seattle CSA. D
The colors in the histogram bins correspond to the levels of the contour matrix. The
histogram includes the number of monitor pairs per bin and the contour matrix
includes the numeric values of the CODs.
March 2011
3-191
DRAFT - DO NOT CITE OR QUOTE
-------
St. Louis CSA
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40
0.45
0,50
0.55
0.55
0.50-
0.45-
0,40-
a>
I °-35 H
E>
-------
3.8.5. Hourly Variations in Ozone for the Urban Focus Cities
1 This section contains diel plots of 1-h avg O3 data to supplement the discussion on hourly
2 variations in O3 concentrations from Section 3.6.3.2 using data from the 20 urban focus cities first
3 introduced in Section 3.6.2.1. Comparisons are made between cold months (October - April) and
4 warm months (May-September), using the year-round data set, and between weekdays and weekends
5 using the warm-season data set.
Cold Months
Warm Months
Weekdays
Weekends
<
5" -95"
: > la-99"
no year-round monitors
0 days, 0 year-round si:es
no year-round monitors
327 days, 11 warm-season sites
]
132 days, 11 warm-season sites
\
]
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 OX):OC
ha.r hour hour ha-r
Figure 3A-96. Diel patterns in 1-h avg ozone for the Atlanta CSA between 2007 and 2009 using
the year-round data set for the cold month/warm month comparison (left half) and
the warm-season data set for the weekday/weekend comparison (right half). No year-
round monitors were available for the cold month/warm month comparison in this
CSA.
March 2011
3-193
DRAFT - DO NOT CITE OR QUOTE
-------
Cold Months
Warm Months
Weekdays
Weekends
< 150 -
O
O
E 6
•5
ra
CO
103 -
50 -
637 days, 9 year-round sites
— mean
median
•• > 5"-95™
459 days, 9 year-round sites
327 days. 28 warm-season sites
132 days. 28 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
ho.r "oir hour hour
Figure 3A-97. Diel patterns in 1-h avg ozone for the Baltimore CSA between 2007 and 2009
using the year-round data set for the cold month/warm month comparison (left half)
and the warm-season data set for the weekday/weekend comparison (right half).
Cold Months
Warm Months
Weekdays
Weekends
CJ
«
•C
CO
150 -
| 100-
o
0 -
637 days, 1 year-round site
mean
median
^^ 5"-95™
; J 1a-99m
451 days, 1 year-round site
327 days, 10 warm-season sites
132 days, 10 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
hour hour hour hour
Figure 3A-98. Diel patterns in 1-h avg ozone for the Birmingham CSA between 2007 and 2009
using the year-round data set for the cold month/warm month comparison (left half)
and the warm-season data set for the weekday/weekend comparison (right half).
March 2011
3-194
DRAFT - DO NOT CITE OR QUOTE
-------
Cold Months
Warm Months
Weekdays
Weekends
15D -
o
a 100 -
o ~
o
DQ
50 -
0 -
637 days, 3 year-round sites
— mean
median
; > 5"-95™
459 days, 3 year-round sites
327 days, 21 warm-season sites
132 days, 21 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 13:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
hour hour hour hour
Figure 3A-99. Diel patterns in 1-h avg ozone for the Boston CSA between 2007 and 2009 using
the year-round data set for the cold month/warm month comparison (left half) and
the warm-season data set for the weekday/weekend comparison (right half).
Cold Months
Warm Months
Weekdays
Weekends
150 -
to
O B- 100 -
O) -3
(0 JT1
o c
£ 50 -
O
0 -
637 days, 11 year-round sites
mean
median
=> 5"-95™
\
\
\
\
459 days, 11 year-round sites
\
\
\
\
327 days, 26 warm-season sites
\
\
\
\
132 days, 26 warm-season sites
\
\
\
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
ho.r "oir hour ho^r
Figure 3A-100. Diel patterns in 1-h avg ozone for the Chicago CSA between 2007 and 2009 using
the year-round data set for the cold month/warm month comparison (left half) and
the warm-season data set for the weekday/weekend comparison (right half).
March 2011
3-195
DRAFT - DO NOT CITE OR QUOTE
-------
Cold Months
Warm Months
Weekdays
Weekends
150 -
0 '£ ioo H
en sS
= <5
Q 50 H
637 days, 19 year-round sites
mean
median
: > 5"-95™
,-—, !«_ 99111
459 days, 19 year-round sites
327 days, 19 warm-season sites
132 days, 19 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
hour hour hour h:u.r
Figure 3A-101. Diel patterns in 1-h avg ozone for the Dallas CSA between 2007 and 2009 using
the year-round data set for the cold month/warm month comparison (left half) and
the warm-season data set for the weekday/weekend comparison (right half).
Cold Months
Warm Months
Weekdays
Weekends
(J
k.
o
150 -
•8- 100 -
£ e
3 5D^
3 -
637 days, 12 year-round sites
mean
median
=> 5"-95™
;—> ia-99"
459 days, 12 year-round si:es
327 days, 15 warm-season sites
132 days, 15 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
hour hour hour hour
Figure 3A-102. Diel patterns in 1-h avg ozone for the Denver CSA between 2007 and 2009 using
the year-round data set for the cold month/warm month comparison (left half) and
the warm-season data set for the weekday/weekend comparison (right half).
March 2011
3-196
DRAFT - DO NOT CITE OR QUOTE
-------
Cold Months
Warm Months
Weekdays
Weekends
^ g 100
'5 "^
is o
3 50
C days. 0 year-round sites
mean
median
> 5" -95™
no year-round monitors
0 days. 0 year-round sites
no year-round monitors
327 days, 9 warm-season sites 132 days. 9 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
hour hour hour ho-r
Figure 3A-103. Diel patterns in 1-h avg ozone for the Detroit CSA between 2007 and 2009 using
the year-round data set for the cold month/warm month comparison (left half) and
the warm-season data set for the weekday/weekend comparison (right half). [Bio
year-round monitors were available for the cold month/warm month comparison in
this CSA.
Cold Months
Warm Months
Weekdays
Weekends
<
tf)
O
150 -
a> o
50H
637 days. 21 year-round sites
— mean
median
=> 5"-95m
\
]
459 days, 21 year-round si:es
]
]
327 days. 21 warm-season sites
]
132 days. 21 warm-season sites
\
]
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
hour hour hour hour
Figure 3A-104. Diel patterns in 1-h avg ozone for the Houston CSA between 2007 and 2009
using the year-round data set for the cold month/warm month comparison (left half)
and the warm-season data set for the weekday/weekend comparison (right half).
March 2011
3-197
DRAFT - DO NOT CITE OR QUOTE
-------
Cold Months
Warm Months
Weekdays
Weekends
CJ
eft
a>
o "^
c a
01
o
150 -
a 100 -
S37 days, 47 year-round sites
mean
median
; ) 5"-95™
=> 1a-99m
459 days, 47 year-round sites
327 days; 50 warm-season sites
132 days, 50 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
hour hour hour h:u.r
Figure 3A-105. Diel patterns in 1-h avg ozone for the Los Angeles CSA between 2007 and 2009
using the year-round data set for the cold month/warm month comparison (left half)
and the warm-season data set for the weekday/weekend comparison (right half).
Cold Months
Warm Months
Weekdays
Weekends
« 150 -
O
=5 a 100-
ra (5
o>
C 50 -
C
425 days, 2 year-round sites
mean
median
306 days, 2 year-round sites
327 days; 8 warm-season sites
132 days, 8 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
hour hour hour hour
Figure 3A-106. Diel patterns in 1-h avg ozone for the Minneapolis CSA between 2007 and 2009
using the year-round data set for the cold month/warm month comparison (left half)
and the warm-season data set for the weekday/weekend comparison (right half).
March 2011
3-198
DRAFT - DO NOT CITE OR QUOTE
-------
Cold Months
Warm Months
Weekdays
Weekends
o
150 -
50 -
e37 days, 20 year-round sites
— mean
median
: > 5"-95™
=i 1a-99»
459 days, 20 year-round sites
327 days, 30 warm-season sites
132 days, 30 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
hci.r hour hour ho^r
Figure 3A-107. Diel patterns in 1-h avg ozone for the New York CSA between 2007 and 2009
using the year-round data set for the cold month/warm month comparison (left half)
and the warm-season data set for the weekday/weekend comparison (right half).
Cold Months
Warm Months
Weekdays
Weekends
150 -
§ 100 -
•S
50 -
637 days, 9 year-round sites
mean
median
\
\
\
\
459 days, 9 year-round sites
\
\
\
\
327 days, 17 warm-season sites
\
\
\
132 days, 17 warm-season sites
\
\
\
\
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
hct^r hour hour hc^r
Figure 3A-108. Diel patterns in 1-h avg ozone for the Philadelphia CSA between 2007 and 2009
using the year-round data set for the cold month/warm month comparison (left half)
and the warm-season data set for the weekday/weekend comparison (right half).
March 2011
3-199
DRAFT - DO NOT CITE OR QUOTE
-------
Cold Months
Warm Months
Weekdays
Weekends
m
o
X
'c
o
o
150 -
50 -
?37 days, 14 year-round sites
— mean
median
; ) 5"_g5"
=> 1a-99"
459 days, 14 year-round sites
327 days; 31 warm-season sites
132 days, 31 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
hour hour hour h:u.r
Figure 3A-109. Diel patterns in 1-h avg ozone for the Phoenix CBSA between 2007 and 2009
using the year-round data set for the cold month/warm month comparison (left half)
and the warm-season data set for the weekday/weekend comparison (right half).
Cold Months
Warm Months
Weekdays
Weekends
150 H
O
p ft 100 -
.0
en
i
CL
S37 days, 2 year-round sites
mean
median
^^ 5"-95™
; J 1a-99m
—T
459 days, 2 year-round sites
327 days, 14 warm-season sites
132 days, 14 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
hour hour hour h:u.r
Figure 3A-110. Diel patterns in 1-h avg ozone for the Pittsburgh CSA between 2007 and 2009
using the year-round data set for the cold month/warm month comparison (left half)
and the warm-season data set for the weekday/weekend comparison (right half).
March 2011
3-200
DRAFT - DO NOT CITE OR QUOTE
-------
Cold Months
Warm Months
Weekdays
Weekends
(J
O i. 100 -
Q.
(D -^
•if O
2 50 H
D -
424 days, 2 year-round sites
— mean
median
: > 5"-95™
306 days, 2 year-round sites
327 days, 12 warm-season sites
132 days, 12 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
hour "oir hour ho-r
Figure 3A-111. Diel patterns in 1-h avg ozone for the Salt Lake City CSA between 2007 and 2009
using the year-round data set for the cold month/warm month comparison (left half)
and the warm-season data set for the weekday/weekend comparison (right half).
Cold Months
Warm Months
Weekdays
Weekends
co
CO
(J
o
'E
o
c
<
c
(T>
CO
150 -
100-
50 -
637 days, 5 year-round sites
mean
median
=> 5"-95™
; > 1--99"
I
I
I
459 days, 5 year-round sites
I
n
i
i
327 days, 5 warm-season sites
i
i
132 days, 5 warm-season sites
i
i
i
i
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
hour "oir hour hour
Figure 3A-112. Diel patterns in 1-h avg ozone for the San Antonio CBSA between 2007 and 2009
using the year-round data set for the cold month/warm month comparison (left half)
and the warm-season data set for the weekday/weekend comparison (right half).
March 2011
3-201
DRAFT - DO NOT CITE OR QUOTE
-------
Cold Months
Warm Months
Weekdays
Weekends
(J
o
o
-o
stf
153 -
100-
637 days, 25 year-round sites
— mean
median
: > 5"-95™
,-—, !«_ 99111
459 days, 25 year-round sites
327 days, 31 warm-season sites
132 days, 31 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
hxr hour hour ho-r
Figure 3A-113. Diel patterns in 1-h avg ozone for the San Francisco CSA between 2007 and 2009
using the year-round data set for the cold month/warm month comparison (left half)
and the warm-season data set for the weekday/weekend comparison (right half).
Cold Months
Warm Months
Weekdays
Weekends
150 -
CO
O -2
CD a
!rf
CD
CO
50 -
3 -
637 days, 5 year-round sites
— mean
median
= 5"-95m
459 days, 5 year-round sites
327 days, 10 warm-season sites
132 days, 10 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
ho.r "ou' hour ha-r
Figure 3A-114. Diel patterns in 1-h avg ozone for the Seattle CSA between 2007 and 2009 using
the year-round data set for the cold month/warm month comparison (left half) and
the warm-season data set for the weekday/weekend comparison (right half).
March 2011
3-202
DRAFT - DO NOT CITE OR QUOTE
-------
Cold Months
Warm Months
Weekdays
Weekends
(J
150 -
100 -
D -
635 days, 3 year-round sites
mean
median
: ) 5"-95™
=i 1a-99"
459 days, 3 year-round sites
327 days; 16 warm-season sites
132 days, 16 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 QC:OC
hxr "cu- hour hour
Figure 3A-115. Diel patterns in 1-h avg ozone for the St. Louis CSA between 2007 and 2009
using the year-round data set for the cold month/warm month comparison (left half)
and the warm-season data set for the weekday/weekend comparison (right half).
March 2011
3-203
DRAFT - DO NOT CITE OR QUOTE
-------
Chapter 4. Exposure to Ambient Ozone
4.1. Introduction
1 The 2006 O3 AQCD evaluated O3 concentrations and exposures in multiple
2 microenvironments, discussed methods for estimating personal and population exposure via
3 monitoring and modeling, analyzed relationships between personal exposure and ambient
4 concentrations, and discussed the implications of using ambient O3 concentrations as an estimate of
5 exposure in epidemiologic studies. This chapter presents new information regarding exposure to
6 ambient O3 in the context of existing relevant information summarized in the 2006 O3 AQCD, which
7 in many areas remains definitive. A brief summary of findings from the 2006 O3 AQCD is presented
8 at the beginning of each section as appropriate.
9 Section 4.2 presents general exposure concepts describing the relationship between ambient
10 pollutant concentrations and personal exposure. Section 4.3 describes exposure measurement
11 techniques and studies that measured personal, ambient, indoor, and outdoor concentrations of O3
12 and related pollutants. Section 4.4 describes techniques for modeling local O3 concentrations,
13 microenvironmental concentrations, and personal and population exposure. Section 4.5 discusses the
14 implications of using ambient O3 concentrations to estimate exposure in epidemiologic studies,
15 including several factors that contribute to exposure error.
4.2. General Exposure Concepts
16 A theoretical model of personal exposure is presented to highlight measurable quantities and
17 the uncertainties that exist in this framework. An individual's time-integrated total exposure to O3
18 can be described based on a compartmentalization of the person's activities throughout a given time
19 period:
Equation 4-1
20 where ET = total (T) exposure over a time-period of interest, Cj = airborne O3 concentration at
21 microenvironmentj, and dt = portion of the time-period spent in microenvironment/ Equation 4-1
22 can be decomposed into a model that accounts for exposure to O3, of ambient (Ea) and nonambient
23 (Ena) origin of the form:
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments such as the Integrated Science Assessments (ISAs) and the Integrated Risk Information System (IRIS).
March 2011 4-1 DRAFT - DO NOT CITE OR QUOTE
-------
Equation 4-2
1 Ambient O3 is formed through photochemical reactions involving NOX, VOCs, and other
2 compounds, as described in Chapter 3. Although nonambient sources of O3 exist, such as O3
3 generators and laser printers, these sources are specific to individuals and may not represent
4 important sources of population exposure. Ozone concentrations generated by ambient and
5 nonambient sources are subject to spatial and temporal variability that can affect estimates of
6 exposure and influence epidemiologic effect estimates. Exposure parameters affecting interpretation
7 of epidemiologic studies are discussed in Section 4.5.
8 This assessment focuses on the ambient component of exposure because this is more relevant
9 to the NAAQS review. Ea can be expressed in terms of the fraction of time spent in various outdoor
10 and indoor microenvironments (Wallace et al, 2006, 089190: Wilson et al, 2000, 010288):
Equation 4-3
1 1 where /= fraction of the relevant time period (equivalent to dt in Equation 4-1), subscript o = index
12 of outdoor microenvironments, subscript /' = index of indoor microenvironments, subscript o,i =
13 index of outdoor microenvironments adjacent to a given indoor microenvironment /', and F^i =
14 infiltration factor for indoor microenvironment (i). Equation 4-3 is subject to the constraint Tf0 +
15 Z/i = 1 to reflect the total exposure over a specified time period, and each term on the right hand side
16 of the equation has a summation because it reflects various microenvironmental exposures. Here,
17 "indoors" refers to being inside any aspect of the built environment, e.g., home, office buildings,
18 enclosed vehicles (automobiles, trains, buses), and/or recreational facilities (movies, restaurants,
19 bars). "Outdoor" exposure can occur in parks or yards, on sidewalks, and on bicycles or motorcycles.
20 Fjnf is a function of the building air exchange characteristics. Assuming steady state ventilation
21 conditions, the infiltration factor is a function of the penetration (P) of O3, the air exchange rate (a)
22 of the microenvironment, and the rate of O3 loss (k) in the microenvironment; Fmf = Pa/(a+k).
23 In epidemiologic studies, Ca is often used in lieu of outdoor microenvironmental data to
24 represent these exposures based on the availability of data. Thus it is often assumed that C0 = Ca and
25 that the fraction of time spent outdoors can be expressed cumulatively as/,; the indoor terms still
26 retain a summation because infiltration differs among different microenvironments. If an
27 epidemiologic study employs only Ca, then the assumed model of an individual's exposure to
28 ambient O3, first given in Equation 4-3, is re-expressed solely as a function of Ca:
Equation 4-4
29 Meteorology, varying precursor emissions and O3 formation rates, spatial variability of O3
30 concentration, design of the epidemiologic study, and other factors determine whether or not
3 1 Equation 4-4 is a reasonable approximation for Equation 4-3. Errors and uncertainties inherent in use
March 201 1 4-2 DRAFT - DO NOT CITE OR QUOTE
-------
1 of Equation 4-4 in lieu of Equation 4-3 are described in Section 4.5 with respect to implications for
2 interpreting epidemiologic studies. Epidemiologic studies often use concentration measured at a
3 central site monitor to represent ambient concentration; thus a, the ratio between personal exposure
4 to ambient O3 and the ambient concentration of O3, is defined as:
Equation 4-5
Combination of Equation 4-4 and Equation 4-5 yields:
Equation 4-6
6 where a varies between 0 and 1. If a person's exposure occurs in a single microenvironment, the
7 ambient component of a microenvironmental O3 concentration can be represented as the product of
8 the ambient concentration and P. Wallace et al. (2006, 089190) note that time-activity data and
9 corresponding estimates of P for each microenvironmental exposure are needed to compute an
10 individual's a with accuracy. If local sources and sinks exist and are significant but not captured by
1 1 central site monitors, then the ambient component of the local outdoor concentration may be
12 estimated using dispersion models, land use regression models, receptor models, fine scale CTMs or
13 some combination of these techniques. These techniques are described in Section 4.4.
4.3. Exposure Measurement
4.3.1. Personal Monitoring Techniques
14 As described in the 2006 O3 AQCD (U.S. EPA, 2006, 088089). a passive sampler has been
15 developed and deployed to measure personal exposure to O3. This sampler uses a filter coated with
16 nitrite, which is converted to nitrate by O3 and then quantified by a technique such as ion
17 chromatography (Koutrakis et al., 1993, 202808). This method has been licensed and marketed by
18 Ogawa, Inc., Japan (Ogawa & Company, 2007, 090937). The cumulative sampling and the detection
19 limit of the passive badges makes them suitable for monitoring periods of 24 hours or greater, which
20 limits their ability to measure short-term daily fluctuations in personal O3 exposure. Longer
21 sampling periods give lower detection limits; use of the badges for a 6-day sampling period yields a
22 detection limit of 1 ppb, while a 24-hour sampling period gives a detection limit of approximately 5-
23 10 ppb. This can result in a substantial fraction of daily samples being below the detection limit
24 (Sarnat et al., 2005, 087531: Sarnat et al., 2006, 089784).
25 The nitrite-nitrate conversion reaction has also been used as the basis for an active sampler
26 consisting of a nitrite-coated glass tube through which air is drawn by a pump operating at
27 65 mL/min (Geyh et al., 1997, 086151: Geyh et al., 1999, 016908). The reported detection limit is
March 2011 4-3 DRAFT - DO NOT CITE OR QUOTE
-------
1 10 ppb-h, enabling the quantification of O3 concentrations measured over a few hours rather than a
2 full day (Geyh et al, 1999, 016908).
3 A portable active O3 monitor based on the UV photometric technique used for stationary
4 monitors (Chapter 3) has recently been approved as a FEM (75 FR 22126) (2010, 687659). This
5 monitor includes a Nafion tube in the inlet line to equalize humidity, reducing the effect of humidity
6 changes in different microenvironments (Wilson and Birks, 2006, 595155). Its size and weight
7 (approximately 10x20x30 cm; 2 kg) make it suitable for use in a backpack configuration. The
8 monitors are currently used by the U.S. National Park service as stationary monitors to measure O3
9 in several national parks (Chapter 3).
4.3.2. Indoor-Outdoor Concentration Relationships
10 The 2006 O3 AQCD (U.S. EPA, 2006, 088089) presented results from several studies on the
11 relationship between indoor O3 concentration and the O3 concentration immediately outside the
12 indoor microenvironment. These studies show that the indoor concentration is often substantially
13 lower than the outdoor concentration unless indoor sources are present. Low indoor O3
14 concentrations can be explained by reactions of O3 with surfaces and airborne constituents. Studies
15 have shown that O3 is deposited onto indoor surfaces where reactions produce secondary pollutants
16 such as formaldehyde (Reiss et al., 1995, 078727; Reiss et al., 1995, 078743). However, the indoor-
17 outdoor relationship is greatly affected by the air exchange rate; under conditions of high air
18 exchange rate, such as open windows, the indoor O3 concentration may approach the outdoor
19 concentration. Geyh et al. (2000, 001775) measured 6-day (approximately 144 hours) indoor and
20 outdoor concentrations at 116 homes in southern California, approximately equally divided between
21 the community of Upland and several mountain communities. The extended sampling period
22 resulted in a relatively low detection limit (1 ppb) for the passive samplers used. The Upland homes
23 were nearly all air-conditioned, while the mountain community homes were ventilated by opening
24 windows. During the O3 season, the indoor O3 concentration averaged over all homes was
25 approximately 24% of the overall mean outdoor concentration in Upland (11.8 versus 48.2 ppb),
26 while in the mountain communities, the indoor concentration was 36% of the outdoor concentration
27 (21.4 versus 60.1 ppb). This is consistent with the increased air exchange rate expected in homes
28 using window ventilation. In the non-ozone season, when homes are likely to be more tightly closed
29 to conserve heat, the ratios of indoor to outdoor concentration were 0.15 (3.2 versus 21.1 ppb) and
30 0.08 (2.8 versus 35.7 ppb) in Upland and the mountain communities, respectively. Avol et al. (1998,
31 018270) observed a mean (standard deviation) indoor-outdoor (I/O) ratio of 0.37 (0.25) for 239
32 matched 24-h samples collected between February and December at homes in the Los Angeles area.
33 The I/O ratio during summer was higher than the non-summer I/O ratio (0.43 versus 0.32). The
34 authors also reported a correlation of 0.58 between the indoor concentration and the outdoor
3 5 concentration, which was only slightly higher than the correlation between the indoor concentration
36 and the concentration at the neighborhood fixed-site monitor (0.49). Romieu et al. (1998, 049834)
37 reported a mean I/O ratio of 0.20 (SD = 0.18) in 145 homes in Mexico City for 14-day cumulative
March 2011 4-4 DRAFT - DO NOT CITE OR QUOTE
-------
1 samples, with the highest ratios observed in homes where windows were usually open during the day
2 and where there was no carpeting or air filters. Studies conducted in Nashville, TN and Toronto,
3 Canada both reported mean residential I/O ratios of approximately 0.1 (Lee et al, 2004, 055599; Liu
4 L-JS; Koutrakis et al., 1995, 039061V
5 Investigators have also measured I/O ratios for non-residential microenvironments, including
6 schools and vehicles. Romieu et al. (1998, 049834) reported that O3 concentrations measured during
7 school hours (10-day cumulative sample, 5 h/day) were 30-40% of concentrations immediately
8 outside the schools, while overall I/O ratios (14-day cumulative sample, 24 h/day) were
9 approximately 15%. The authors attribute this discrepancy to increased air exchange during the
10 school day due to opening doors and windows. Air exchange was also identified as an important
11 factor in the I/O ratios measured at eight French schools (Blondeau et al., 2005, 078044). In this
12 study, the I/O ratios based on simultaneous continuous measurements ranged from 0-0.45, increasing
13 with decreasing building tightness. Although no indoor measurements were made, Rundell et al.
14 (2006, 089785) report a mean daytime continuous outdoor O3 concentration of 106 ppb at one
15 university and four elementary school playing fields, indicating the potential for elevated O3
16 exposure in the school outdoor microenvironment.
17 Gradients in O3 concentrations observed near roadways provide evidence of the NO-O3
18 titration reaction that takes place in the ambient environment in which NO emitted from vehicles
19 reacts with O3 to produce NO2 (Finlayson-Pitts and Pitts, 1986, 035054). Ozone concentration has
20 been observed in several studies to increase with increasing distance from the roadway, both upwind
21 and downwind of the road (Section 3.6.2.1). Depending on wind direction, O3 concentrations near
22 the roadway have been found to be 20-80% of ambient concentrations at sites 400 m or more distant
23 from roads. This indicates that exposures in near-road, on-road and in-vehicle microenvironments,
24 while highly variable, may be lower than those in other microenvironments. A study on patrol cars
25 during trooper work shifts reported in-vehicle 9-h concentrations that were approximately 51% of
26 simultaneously measured roadside concentrations (mean of 11.7 versus 28.3 ppb) (Riediker et al.,
27 2003, 043761).
4.3.3. Personal-Ambient Concentration Relationships
4.3.3.1. Personal-Ambient Correlations
28 The relationship between personal exposure and ambient O3 concentrations has been evaluated
29 in several research studies, many of which were conducted prior to 2005 and are discussed in the
30 2006 O3 AQCD (U.S. EPA, 2006, 088089). Some studies evaluated subject-specific, or longitudinal
31 correlations, which describe multiple daily measurements for a single individual. These studies
32 indicate the inter-individual variability of personal-ambient correlations. Another type of correlation
33 is a pooled correlation, which combines data from multiple individuals over multiple monitoring
34 periods (e.g., days), providing an overall indicator of the personal-ambient relationship for all study
March 2011 4-5 DRAFT - DO NOT CITE OR QUOTE
-------
1 subjects. A third type of correlation is a community-average correlation, which correlates average
2 exposure across all study subjects with fixed-site monitor concentrations. Community-average
3 correlations are particularly informative for interpreting time-series epidemiologic studies, in which
4 ambient concentrations are used as a surrogate for community-average exposure.
5 The results of these studies indicate that personal exposures are moderately well correlated
6 with ambient concentrations, and that the ratio of personal exposure to ambient concentration is
7 higher in outdoor microenvironments and during the summer season. In situations where a lack of
8 correlation was observed, this may be due in part to a high proportion of personal measurements
9 below the detection limit. Chang et al. (2000, 001276) measured hourly personal exposures in
10 multiple microenvironments and found that the pooled correlation between personal exposure and
11 ambient concentration was highest for outdoor microenvironments (r = 0.68-0.91). Correlations in
12 residential indoor microenvironments were very low (r = 0.05-0.09), with moderate correlations
13 (0.34-0.46) in other indoor microenvironments such as restaurants and shopping malls. Liard et al.
14 (1999, 001426) evaluated community-average correlations based on 4-day mean personal O3
15 exposure measurements for adults and children and found a relatively high correlation (r = 0.83)
16 with ambient concentrations, even though 31-82% of the personal measurements were below the
17 detection limit. Sarnat et al. (2000, 001852) studied a population of older adults in Baltimore and
18 found that longitudinal correlations between 24-h personal exposure and ambient concentration
19 varied by subject and season, with somewhat higher correlations observed during summer (mean =
20 0.20) than in winter (mean = 0.06). Some evidence was presented that subjects living in well-
21 ventilated indoor environments have higher correlations than those living in poorly ventilated indoor
22 environments, although exceptions to this were also observed. A moderate pooled correlation of 0.61
23 was reported between 24-h avg personal and central-site measurements by Linn et al. (1996, 082508)
24 for a population of southern California schoolchildren who spent an average of 101-136 minutes
25 outdoors. The authors also report a correlation of 0.70 between central-site measurements and
26 concentrations outside the children's schools. Although the average school outdoor concentration
27 (34 ppb) was higher than the average central-site concentration (23 ppb) and the average personal
28 exposure concentration was lower (5 ppb) than the central-site value, the similarity between the
29 correlations indicate that central-site monitor concentrations can represent personal exposures in
30 addition to representing local outdoor concentrations. A similar result was observed in a study in
31 Vancouver, BC comparing three groups spending different amounts of time outdoors: health clinic
32 workers (0-25% of time outdoors), camp counselors (7.5-45% of time outdoors), and farm workers
33 (100% of time outdoors) (Brauer and Brook, 1997, 083339). Health clinic workers and camp
34 counselors were monitored 24 h/day, while farm workers were monitored during their work shift
35 (6-14 hours). In this study, the pooled correlations between personal exposure and fixed-site
36 concentration were similar among the groups, without a clear trend (r = 0.60, 0.42, and 0.64,
37 respectively), although the ratios of personal exposure to fixed-site monitor concentration increased
38 among the groups with increasing amount of time spent outdoors (0.28, 0.48, and 0.96, respectively).
March 2011 4-6 DRAFT - DO NOT CITE OR QUOTE
-------
1 This indicates that temporal variations in personal exposure to O3 are driven by variations in ambient
2 concentration, even for individuals that spend little time outdoors.
4.3.3.2. Personal-Ambient Ratios
3 Additional studies were summarized in the 2006 O3 AQCD (U.S. EPA, 2006, 088089)
4 evaluating the ratio of personal O3 exposure to ambient concentration or regressing personal
5 exposure on ambient concentration. O'Neill et al. (2003, 042752) studied a population of shoe
6 cleaners working outdoors in Mexico City and presented a regression model indicating a 0.56 ppb
7 increase in 6-h personal exposure for each 1 ppb increase in ambient concentration (95% CI: 0.43,
8 0.69). Regression analyses by Sarnat et al. for 24-h data from mixed populations of children and
9 older adults in Baltimore (2001, 019401) and Boston (2005, 087531) found differing results between
10 the two cities, with Baltimore subjects showing a near-zero slope (0.01) during the summertime
11 while Boston subjects showed a positive slope of 0.27 ppb personal exposure per 1 ppb ambient
12 concentration. In both cities, the winter slope was near zero. Differences between the study areas
13 may be due to differences in time spent outdoors, residential ventilation conditions, or other factors.
14 Xue et al. (2005, 087532) measured 6-day personal exposure of children in southern California and
15 found that the average ratio of personal exposure to ambient concentration was relatively stable
16 throughout the year at 0.3. These authors also regressed personal exposures on ambient
17 concentration after adjusting for time-activity patterns and housing characteristics and found a slope
18 of 0.54 ppb/ppb, with an R2 value of 0.58. Unadjusted regression slopes were not presented.
19 A few studies have been published since the 2006 O3 AQCD comparing personal exposures
20 with ambient concentrations, and these findings generally confirm the conclusions of the 2006 O3
21 AQCD. Sarnat et al. (2006, 089784) measured 24-h personal exposures for a panel of older adults in
22 Steubenville, OH during summer and fall 2000. Subjects were classified as high-ventilation or low-
23 ventilation based on whether they spent time in indoor environments with open windows. Regression
24 of personal exposures on ambient concentration found a higher slope for high-ventilation subjects
25 compared with low-ventilation subjects in both summer (0.18 versus 0.08) and fall (0.27 versus
26 0.20). Although no personal exposures were measured, McConnell et al. (2006, 089256) found that
27 average 24-h home outdoor O3 concentrations were within 6 ppb of O3 concentrations measured at
28 central-site monitors in each of three southern California communities, with a combined average
29 home outdoor concentration of 33 ppb compared to the central-site average of 36 ppb. Ramirez-
30 Aguilar et al. (2008, 098930) measured 48- to 72-h personal exposures of four groups of asthmatic
31 children aged 6-14 in Mexico City during 1998-2000. A moderate pooled correlation (r = 0.35) was
32 observed between these exposures and corresponding ambient concentrations. Regression of
33 personal exposures on ambient concentrations yielded a slope of 0.17 ppb/ppb (95% CI: 0.13, 0.21)
34 after adjustment for distance to the fixed-site monitor, time spent outdoors, an interaction term
35 combining these two variables, and an interaction term representing neighborhood and study group.
36 Taken together, results from previous and recently published studies indicate that while the
37 relationship between personal exposures and ambient concentrations varies due to a number of
March 2011 4-7 DRAFT - DO NOT CITE OR QUOTE
-------
1 factors, such as activity patterns, housing characteristics, and season, O3 concentrations measured at
2 central-site monitors are representative of day-to-day changes in average personal O3 exposure,
3 which is the important parameter for time-series epidemiologic studies. Another important finding is
4 that the magnitude of personal exposures is smaller than concentrations reported at fixed-site
5 monitors due to time spent indoors and the low indoor penetration of O3.
4.3.4. Co-Exposure to Other Pollutants and Environmental Stressors
4.3.4.1. Personal Exposure to Ozone and Co-pollutants
6 Personal exposure to O3 shows variable correlation with personal exposure to other pollutants,
7 with differences in correlation depending on factors such as season, city-specific characteristics, and
8 spatial variability of the co-pollutant. Suh and Zanobetti (2010, 677202) reported Spearman rank
9 correlation coefficients during spring and fall between 24-h avg O3 measurements and co-pollutants
10 of 0.14, 0.00, and -0.03 for PM2 5, EC, and NO2, respectively. Note the higher correlation observed
11 for PM2 5, a regional pollutant, in contrast with the extremely low correlations for the traffic-related
12 and more spatially variable pollutants EC and NO2. Chang et al. (2000, 001276) measured hourly
13 personal exposures to PM25 and O3 in summer and winter in Baltimore, Maryland. Correlations
14 between PM2 5 and O3 were 0.05 and -0.28 in summer and winter, respectively. Results indicate
15 personal O3 exposures were not significantly associated with personal PM25 exposures in either
16 summer or winter. These non-significant correlations may be attributed in part to the relatively low
17 personal O3 exposures observed in this study.
18 Studies conducted in Baltimore (Sarnat et al., 2001, 019401) and Boston (Sarnat et al., 2005,
19 087531) found differing results for the correlation between 24-h avg personal O3 and personal PM2 5
20 exposures, particularly during the winter season. Sarnat et al. (2001, 019401) found a positive slope
21 when regressing personal exposures of both total PM2 5 (0.21) and PM2 5 of ambient origin (0.22)
22 against personal O3 exposures during the summer season, but negative slopes (-0.05 and -0.18,
23 respectively) during the winter season. The summertime slope for personal PM2 5 exposure versus
24 personal O3 exposure was much higher for children (0.37) than for adults (0.07), which may be the
25 result of different activity patterns. This team of researchers also found a positive, although higher,
26 summer slope between 24-h avg personal O3 and personal PM2 5 in Boston (0.72) (Sarnat et al.,
27 2005, 087531). However, the winter slope was positive (1.25) rather than negative, as in Baltimore.
28 In both cities during both seasons, there was a wide range of subject-specific correlations between
29 personal O3 and personal PM2 5 PM2 5, with some subjects showing relatively strong positive
30 correlations (>0.75) and others showing strong negative correlations (<-0.50). The median
31 correlation in both cities was slightly positive in the summer and near zero (Boston) or slightly
32 negative (Baltimore) in the winter. These results indicate the potential effects of city-specific
33 characteristics, such as housing stock and building ventilation patterns, on relationships between O3
34 and co-pollutants.
March 2011 4-8 DRAFT - DO NOT CITE OR QUOTE
-------
4.3.4.2. Near-Road Exposure to Ozone and Co-pollutants
1 Beckerman et al. (2008, 096484) measured both 1-week and continuous concentrations of O3,
2 NO, NO2, NOX, PM2 5, PMi.o, and several VOCs (the BTEX compounds, MTBE, hexane, and THC)
3 in the vicinity of heavily traveled (annual average daily traffic [AADT] >340,000) roadways in
4 Toronto, Canada. Passive samplers were deployed for one week in August 2004. Ozone
5 concentrations were negatively correlated with all pollutants, with the exception of VOCs at one of
6 the monitoring sites which were suspected of being influenced by small area sources. Site specific
7 correlations are given in Figure 4-1. Correlations were -0.77 to -0.85 for NO2, -0.48 to -0.62 for NO,
8 and -0.55 to -0.63 for NOX. Pooled correlations using data from both sites were somewhat lower in
9 magnitude. PM2 5 and PMi.o correlations were -0.35 to -0.78 and -0.34 to -0.58, respectively. At the
10 monitoring site not influenced by small area sources, O3-VOC correlations ranged from -0.41 to
11 -0.66.
12 Beckerman et al. (2008, 096484) also made on-road measurements of multiple pollutants with
13 a instrumented vehicle. Concentrations were not reported, but correlations between O3 and other
14 pollutants were negative and somewhat greater in magnitude (i.e., more negative) than the near-road
15 correlations. SO2, CO, and BC were measured in the mobile laboratory, although not at the roadside,
16 and they all showed negative correlations with O3 when the data were controlled for site.
17 Correlations for continuous concentrations between O3 and co-pollutants were somewhat lower than
18 the 1-week correlations, except for O3-PM25 correlations. Correlations were -0.90, -0.66, -0.77, and -
19 0.89 for NO2, NO, NOX, and PMi 0 respectively. The continuous O3-PM2 5 correlation was -0.62,
20 which is in the range of the 1-week correlation.
-0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0
Pearson Correlation Coefficient
Source: Beckerman et al. (2008, 0964841
Figure 4-1. Correlations between 1-week ozone concentrations and co-pollutants.
March 2011
4-9
DRAFT - DO NOT CITE OR QUOTE
-------
4.3.4.3. Indoor Exposure to Ozone and Co-pollutants
1 Ambient O3 that infiltrates indoors reacts with organic compounds and other chemicals to
2 form oxidized products, as described in the 2006 O3 AQCD (U.S. EPA, 2006, 088089). It is
3 anticipated that individuals are exposed to these reaction products, although little evidence is
4 available on personal exposures. The reactions are similar to those occurring in the ambient air, as
5 summarized in Chapter 3. For example, O3 can react with terpenes and other compounds from
6 cleaning products, air fresheners, and wood products to form particulate and gaseous species, such as
7 formaldehyde (Aoki and Tanabe, 2007, 449034: Reiss et al, 1995, 078743V Ozone has also been
8 shown to react with material trapped on HVAC filters and generate airborne products (Beko et al.,
9 2007, 601190; Hyttinen et al., 2006, 134392). It is possible that these reaction products may have
10 health effects in addition to, or greater than, those from O3 itself (Anderson et al., 2007, 600123;
11 Weschler and Shields, 1997, 084530). Ozone may also react to form other oxidants, which then go
12 on to participate in additional reactions. White et al. (2010, 633874) found evidence that HONO or
13 other oxidants may have been present during experiments to estimate indoor OH concentrations,
14 indicating complex indoor oxidant chemistry. Rates of these reactions are dependent on indoor O3
15 concentration, temperature, and air exchange rate, making estimation of exposures to reaction
16 products difficult.
4.3.5. Population Proximity to Fixed-Site Ozone Monitors
17 The distribution of O3 monitors across urban areas varies between cities (Section 3.6.2.1), and
18 the population living near each monitor varies as well. It is not necessarily true that proximity to a
19 monitor determines the degree to which that monitor represents an individual's ambient exposure,
20 but proximity is one indicator. One way to calculate monitor representativeness is to calculate the
21 fraction of the urban population living within a certain radius of a monitor. Table 4-1 presents the
22 fraction of the population in selected cities living within 1, 5, 10, and 20 km of an O3 monitor. Values
23 are presented for both total population and for those under 18 years of age, a potentially susceptible
24 population to the effects of O3. The data indicate that relatively few people live within 1 km of an O3
25 monitor, while nearly all of the population in most cities lives within 20 km of a monitor. Many O3
26 monitors are sited at "neighborhood scale," intended to represent an area of the city with dimensions
27 in the 0.5-4 km range (Section 3.5.6.1). Looking at the results for a 5-km radius, generally 20-30%
28 of the population lives within this distance from an O3 monitor. Some cities have a greater
29 population in this buffer, such as Salt Lake City, while others have a lower percentage, such as
30 Minneapolis and Seattle. Percentages for children are generally similar to the total population, with
31 no clear trend.
March 2011 4-10 DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-1. Fraction of the 2009
selected U.S. cities
City
Atlanta CSA
Baltimore CSA
Birmingham CSA
Boston CSA
Chicago CSA
Dallas CSA
Denver CSA
Detroit CSA
Houston CSA
Los Angeles CSA
Minneapolis CSA
New York CSA
Philadelphia CSA
Phoenix CBSA
Pittsburgh CSA
Salt Lake City CSA
San Antonio CBSA
San Francisco CSA
Seattle CSA
St. Louis CSA
Population
Total
5,901,670
8,421,016
1,204,399
7,540,533
9,980,113
6,791,942
3,103,801
5,445,448
5,993,633
18,419,720
3,652,490
22,223,406
6,442,836
4,393,462
2,471,403
1,717,045
2,061,147
7,497,443
4,181,278
2,914,754
population living within
<18yr
1,210,932
1,916,106
281,983
1,748,918
2,502,454
1,530,877
675,380
1,411,875
1,387,851
4,668,441
872,497
5,284,875
1,568,878
873,084
563,309
460,747
484,473
1,675,711
918,309
720,746
Within
Total
0.3%
1.3%
1.4%
0.9%
1.5%
0.4%
1.7%
0.8%
1.5%
1.6%
0.3%
1.5%
0.9%
2.0%
1.5%
3.0%
0.5%
2.6%
0.3%
1.3%
1 km
<18yr
0.3%
1.1%
1.6%
0.9%
1.5%
0.4%
1.6%
0.9%
1.8%
1.7%
0.3%
1.7%
1.0%
2.4%
1.4%
3.0%
0.5%
2.9%
0.3%
1.5%
a specified distance of an ozone monitor in
Within 6 km
Total
8%
25%
22%
17%
28%
13%
35%
15%
26%
28%
5%
23%
22%
35%
22%
41%
12%
41%
5%
17%
<18yr
9%
24%
24%
16%
29%
13%
36%
17%
28%
29%
4%
23%
24%
41%
21%
38%
12%
40%
5%
18%
Within
Total
28%
57%
56%
49%
63%
45%
66%
42%
54%
77%
16%
51%
55%
74%
52%
79%
42%
81%
18%
52%
10km
<18yr
29%
55%
59%
47%
65%
44%
68%
44%
57%
79%
16%
50%
56%
79%
50%
79%
43%
81%
16%
53%
Within 20 km
Total
75%
89%
73%
85%
89%
87%
92%
77%
83%
98%
57%
91%
89%
96%
88%
95%
78%
98%
43%
80%
<18yr
77%
89%
74%
85%
91%
87%
93%
78%
84%
98%
56%
91%
89%
97%
88%
95%
80%
98%
39%
82%
1 Another approach is to divide the metropolitan area into sectors surrounding each monitor
2 such that every person in the sector lives closer to that monitor than any other. This facilitates
3 calculation of the fraction of the city's population represented (according to proximity) by each
4 monitor. In Atlanta, for example, the population fraction represented by each of the 11 monitors in
5 the city ranged from 2.9-22%. The two monitors closest to the city center (sites A and B on
6 Figure 3-24) accounted for 16% and 8% of the population, respectively. Site B has two listed
7 monitoring objectives, highest concentration and population exposure. The other monitor in Atlanta
8 with a listed objective of highest concentration is Site C, which represents the largest fraction of the
9 population (22%). The eight monitors with a primary monitoring objective of population exposure
10 account for 2.9-17% of the population per monitor.
11 Atlanta population fractions for children (<18 years of age) are similar to those for the general
12 population, but other populations show a different pattern of monitor representativeness. Older adults
13 (age 65 and up) were somewhat differently distributed with respect to the monitors, with most
14 monitors showing a difference of more than a percentage point compared to the general population.
15 Based on 2000 population data, the fraction of older adults closest to the two city center monitors (A
16 and B) was 4% higher and 2% lower, respectively, than the fraction for the population as a whole.
17 Site C showed the highest differential, with 21% of the total population but only 15% of the older
March 2011
4-11
DRAFT - DO NOT CITE OR QUOTE
-------
1 adult population. This indicates the potential for monitors to differentially represent potentially
2 susceptible populations.
4.4. Exposure Modeling
4.4.1. Concentration Surface Modeling
3 One approach to improve exposure estimates in urban areas involves construction of a
4 concentration surface over a geographic area, with concentration at locations between monitors
5 estimated using a model to compensate for missing data. The calculated O3 concentration surface can
6 then be used to estimate exposures outside residences, schools, workplaces, roadways, or other
7 locations of interest. This technique does not estimate exposure directly because it does not account
8 for activity patterns or concentrations in different microenvironments. There are three main types of
9 approaches: spatial interpolation of measured concentrations; statistical models using meteorological
10 variables, pollutant concentrations, and other predictors to estimate concentrations at receptors in the
11 domain; and rigorous first-principle models, such as chemistry-transport models or dispersion
12 models incorporating O3 chemistry. Some researchers have developed models that combine these
13 techniques. The models may be applied over urban, regional, or national spatial scales, and can be
14 used to estimate daily concentrations or longer-term averages. This discussion will focus on short-
15 term concentrations estimated across urban areas.
16 The 2006 O3 AQCD (U.S. EPA, 2006, 088089) discussed concentration surface models,
17 focusing on chemistry-transport models as well as geospatial and spatiotemporal interpolation
18 techniques (Christakos and Vyas, 1998, 029728: Christakos and Vyas, 1998, 011987: e.g.,
19 Georgopoulos et al, 1997, 083591). Recent research has continued to refine and extend the modeling
20 approaches. A few recent papers have compared different approaches for the same urban area.
21 Marshall et al. (2008, 193983) compared four spatial interpolation techniques for estimation of
22 O3 concentrations in Vancouver, BC. The investigators assigned a daily average O3 concentration to
23 each of the 51,560 postal-code centroids using one of the following techniques: (1) the concentration
24 from the nearest monitor within 10 km; (2) the average of all monitors within 10 km; (3) the inverse-
25 distance-weighted (IDW) average of all monitors in the area; and (4) the IDW average of the 3
26 closest monitors within 50 km. Method 1 (the nearest-monitor approach) and Method 4 (IDW-50
27 km) had similar mean and median estimated annual- and monthly-average concentrations, although
28 the 10th-90th percentile range was smaller for IDW-50. This is consistent with the averaging of
29 extreme values inherent in IDW methods. The Pearson correlation coefficient between the two
30 methods was 0.93 for monthly-average concentrations and 0.78 for annual-average concentrations.
31 Methods 2 and 3 were considered sub-optimal and were excluded from further analysis. In the case
32 of Method 2, a single downtown high-concentration monitor skewed the results in the vicinity,
33 partially as a result of the asymmetric layout of the coastal city of Vancouver. Method 3 was too
34 spatially homogenous because it assigned most locations a concentration near the regional average,
March 2011 4-12 DRAFT - DO NOT CITE OR QUOTE
-------
1 except for locations immediately adjacent to a monitoring site. CMAQ concentration estimates using
2 a 4 km><4 km grid were also compared to the interpolation techniques in this study. Mean and
3 median concentrations from CMAQ were approximately 50% higher than Method 1 and Method 4
4 estimates for both annual and monthly average concentrations. This may be due in part to the CMAQ
5 grid size, which was too coarse to reveal near-roadway decrements in O3 concentration due to
6 titration by NO. The IQR for the annual average was similar between CMAQ and the interpolation
7 techniques, but the monthly average CMAQ IQR was approximately twice as large, indicating a
8 seasonal effect. Bell (2006, 194358) compared CMAQ estimates for northern Georgia with
9 nearest-monitor and spatial interpolation techniques, including IDW and kriging. The area-weighted
10 concentration estimates from CMAQ indicated areas of spatial heterogeneity that were not captured
11 by approaches based on the monitoring network. The author concluded that some techniques, such as
12 spatial interpolation, were not suitable for estimation of exposure in certain situations, such as for
13 rural areas. Using the concentration from the nearest monitor resulted in an overestimation of
14 exposure relative to model estimates.
15 Land use regression (LUR) models have been developed to estimate levels of air pollutants,
16 predominantly NO2, as a function of several land use factors, such as land use designation, traffic
17 counts, home heating usage, point source strength, and population density (Briggs et al., 1997,
18 025950: Gilliland et al., 2005, 098820: Ryan and LeMasters, 2007, 156063V LUR, initially termed
19 regression mapping (Briggs et al., 1997, 025950). is a regression derived from monitored
20 concentrations as a function of data from a combination of the land use factors. The regression is
21 then used for predicting concentrations at multiple locations based on the independent variables at
22 those particular locations without monitors. Hoek et al. (2008, 195851) warn of several limitations of
23 LUR, including distinguishing real associations between pollutants and covariates from those of
24 correlated co-pollutants, limitations in spatial resolution from monitor data, applicability of the LUR
25 model under changing temporal conditions, and introduction of confounding factors when LUR is
26 used in epidemiologic studies. These limitations may partially explain the lack of LUR models that
27 have been developed for O3 at the urban scale. Brauer et al. (2008, 156292) evaluated the use of
28 LUR and IDW-based spatial-interpolation models in epidemiologic analyses for Vancouver, BC and
29 suggested that LUR is appropriate for directly-emitted pollutants with high spatial variability, such
30 as NO and BC, while IDW is appropriate for secondary pollutants such as NO2 and PM2 5 with less
31 spatial variability. Although this study did not evaluate an LUR model for O3, possibly due to the
32 lack of the required dedicated sampling campaign, the IDW approach would be expected to be
33 favored since O3 is a secondary pollutant. At a much larger spatial scale, EU-wide, Beelen et al.
34 (2009, 601157) compared a LUR model for O3 with ordinary kriging and universal kriging, which
35 incorporated meteorlogical, topographical, and land use variables to characterize the underlying
36 trend. The LUR model performed reasonably well at rural locations (5-km resolution), explaining a
37 higher percentage of the variability (R2 = 0.62) than for other pollutants. However, at the urban scale
38 (1-km resolution), only one variable was selected into the O3 LUR model (high-density residential
39 land use), and the R2 value was very low (0.06). Universal kriging was the best method for the large-
March 2011 4-13 DRAFT - DO NOT CITE OR QUOTE
-------
1 scale composite EU concentration map, for O3 as well as for NO2 and PMi0, with an R2 value for O3
2 of 0.70. The authors noted that these methods were not designed to capture spatial variation in
3 concentrations that are known to occur within tens of meters of roadways (Section 3.6.2.1), which
4 could partially explain poor model performance at the urban scale.
5 Titration of O3 with NO emitted by motor vehicles tends to reduce O3 concentrations near
6 roadways. McConnell et al. (2006, 089256) developed a regression model to predict residential O3
7 concentrations in southern California using estimates of residential NOX calculated from traffic data
8 with the CALINE4 line source dispersion model. The authors estimated that local traffic contributes
9 18% of NOX concentrations measured in the study communities, with the remainder coming from
10 regional background. Their regression model indicates that residential NOX reduces residential O3
11 concentrations by 0.51 ppb O3 per 1 ppb NOX, and that a 10th-90th percentile increase in local NOX
12 results in a 7.5 ppb decrease in local O3 concentrations. This intra-urban traffic-related variability in
13 O3 concentrations suggests that traffic patterns are an important factor in the relationship between
14 central site monitor and residential O3, and that differences in traffic density between the central site
15 monitor and individual homes could result in either an overestimate or underestimate of residential
16 O3.
17 A substantial number of researchers have used geostatistical methods and chemistry-transport
18 models to estimate O3 concentrations at urban, regional, national, and continental scales, both in the
19 U.S. and in other countries (Hooyberghs et al., 2006, 608180: e.g., Pakalapati et al., 2009, 615781V
20 In addition to short-term exposure assessment for epidemiologic studies, such models may also be
21 used for long-term exposure assessment, O3 forecasts (Sahu et al., 2009, 618189). or evaluating
22 emission control strategies (Gabusi and Volta, 2005, 606112). It is difficult to determine the utility of
23 these methods for exposure assessment; while improved local-scale estimates of outdoor
24 concentrations may contribute to better assignment of exposures, information on activity patterns is
25 needed to produce estimates of personal exposure.
4.4.2. Microenvironmental Models
26 Population-based methods, such as the Air Pollution Exposure (APEX) and Stochastic Human
27 Exposure and Dose Simulation (SHEDS) models, involve stochastic treatment of the model inputs
28 (Burke et al., 2001, 014050: U.S. EPA, 2009, 194009). These are described in detail in the 2008 NOX
29 ISA (U.S. EPA, 2008, 157073). in AX3.6.1. Stochastic models utilize distributions of pollutant-
30 related and individual-level variables, such as ambient and local O3 concentration contributions and
31 breathing rate respectively, to compute the distribution of individual exposures across the modeled
32 population. The models also have the capability to estimate received dose through a dosimetry
33 model. Using distributions of input parameters in the model framework rather than point estimates
34 allows the models to incorporate uncertainty and variability explicitly into exposure estimates (Zidek
35 et al., 2007, 190076). These models estimate time-weighted exposure for modeled individuals by
36 summing exposure in each microenvironment visited during the exposure period.
March 2011 4-14 DRAFT - DO NOT CITE OR QUOTE
-------
1 The initial set of input data for population exposure models is ambient air quality data, which
2 may come from a monitoring network or model estimates. Estimates of concentrations in a set of
3 microenvironments are generated either by mass balance methods or microenvironmental factors.
4 Microenvironments modeled include indoor residences; other indoor locations, such as schools,
5 offices, and public buildings; and vehicles. The sequence of microenvironments and exertion levels
6 during the exposure period is determined from characteristics of each modeled individual. The
7 APEX model does this by generating a profile for each simulated individual by sampling from
8 distributions of demographic variables such as age, gender, and employment; physiological variables
9 such as height and weight; and situational variables such as living in a house with a gas stove or air
10 conditioning. Activity patterns from a database such as Consolidated Human Activity Database
11 (CHAD) are assigned to the simulated individual using age, gender, and biometric characteristics
12 (U.S. EPA, 2009, 194010). Breathing rates are calculated for each activity based on exertion level,
13 and the corresponding received dose may then be computed. Summaries of individual- and
14 population-level metrics are produced, such as maximum exposure or dose, number of individuals
15 exceeding a specified exposure/dose threshold, and number of person-days at or above benchmark
16 exposure levels. The models also consider the nonambient contribution to total exposure.
17 Nonambient source terms are added to the infiltration of ambient pollutants to calculate the total
18 concentration in the microenvironment. Output from model runs with and without nonambient
19 sources can be compared to estimate the ambient contribution to total exposure and dose.
20 An analysis has been conducted for the APEX model to evaluate the contribution of
21 uncertainty in input parameters and databases to the uncertainty in model outputs (Langstaff, 2007,
22 090315). The Monte Carlo analysis indicates that the uncertainty in model exposure estimates for
23 asthmatic children during moderate exercise is small to moderate, with 95% confidence intervals of
24 at most ± 6 percentage points at exposures above 60, 70, and 80 ppb (8-h avg) However, APEX
25 appears to substantially underestimate the frequency of multiple high-exposure events for a single
26 individual. The two main sources of uncertainty identified were related to the activity pattern
27 database and the spatial interpolation of fixed-site monitor concentrations to other locations. One
28 area of potential improvement in the activity pattern database is additional information on children's
29 activities. Improved information on spatial variation of O3 concentrations, including in near-roadway
30 and indoor microenvironments, would also contribute to reduced uncertainty. Another area of need is
31 for improved personal exposure monitors with shorter averaging times to capture peak exposures and
32 lower detection limits to capture low indoor concentrations. A similar modeling approach has
33 recently been developed which is suitable for panel epidemiologic studies or for controlled human
34 exposure studies, in which activity pattern data specific to the individuals in the study can be
35 collected. Time-activity data is combined with questionnaire data on housing characteristics,
36 presence of indoor or personal sources, and other information to develop a personalized set of model
37 input parameters for each individual. This model, the Exposure Model for Individuals, is being
38 developed by EPA's National Exposure Research Laboratory.
March 2011 4-15 DRAFT - DO NOT CITE OR QUOTE
-------
1 Recent larger-scale human activity databases, such as those developed for the CHAD or the
2 National Human Activity Pattern Survey (NHAPS), have been designed to characterize exposure
3 patterns among much larger population subsets than can be examined during individual panel studies
4 (Klepeis et al, 2001, 002437: McCurdy et al., 2000, 000782). CHAD consists of a consolidation of
5 human activity data obtained during several panel studies in which diary or retrospective activity
6 data were obtained, while NHAPS acquired sample population time-activity data through surveys
7 about human activity (Klepeis et al., 2001, 002437). The complex human activity patterns across the
8 population (all ages) are illustrated in Figure 4—2 (Klepeis et al., 2001, 002437). which is presented
9 to illustrate the diversity of daily activities among the entire population as well as the proportion of
10 time spent in each microenvironment. For example, about 25% of the individuals reported being
11 outdoors or in a vehicle between 2:00 and 3:00 pm, when daily O3 levels are peaking, although about
12 half of this time was spent in or near a vehicle, where O3 concentrations are likely to be lower than
13 ambient concentrations. Different patterns would be anticipated when breaking down activity
14 patterns only for subgroups such as children or the elderly. Population exposures can be estimated
15 using O3 concentration data in each microenvironment.
I
fl
o
CD
Ofi
cd
CD
PH
j-j-HHRRSS^SSS
cflcscscticflcBcticScflc^cscs
ooooooooooooooooooooooooo
ooooooooooooooooooooooooo
f~v| i—I f1^ f/"] T^- Ij^-j V^Q [^--, QQ ^j^ ^^ ^H
r-
\o
Time of Day
Source: Reprinted with permission from Nature Publishing Group, Klepeis et al. (2001, 0024371.
Figure 4-2. Distribution of time that NHAPS respondents spent in ten microenvironments based
on smoothed 1-min diary data.
March 2011
4-16
DRAFT - DO NOT CITE OR QUOTE
-------
4.4.3. Hybrid Model Approaches
1 Georgopoulos et al. (2005, 080269) used the MENTOR/SHEDS modeling framework to
2 estimate O3 exposure in Philadelphia over a 2-week period in July 1999. They found that both the
3 50th and 95th percentile O3 concentrations were correlated with census-tract level outdoor
4 concentrations estimated by photochemical modeling combined with spatiotemporal interpolation,
5 and attributed this correlation to the lack of indoor sources of O3. Relationships between exposure
6 and concentrations at fixed-site monitors were not reported.
4.5. Implications for Epidemiologic Studies
4.5.1. Exposure Measurement Error
7 Exposure error can be an important contributor to variability in epidemiologic study results.
8 Community-level time-series studies may involve a population of thousands or millions of people
9 across an urban area whose exposure and health status is estimated over the course of a few years
10 using a short monitoring interval (hours to days). Community-averaged concentration is typically
11 used as a surrogate for ambient exposure in community time-series studies. Exposures and health
12 effects are spatially aggregated over the time intervals of interest because community time-series
13 studies are designed to examine health effects and their potential causes at the community level. A
14 longitudinal cohort epidemiology study, such as the ACS cohort study, typically involves hundreds
15 or thousands of subjects followed over several years or decades (e.g., Jerrett et al., 2009, 194160).
16 Concentrations are generally aggregated over time and by community to estimate exposures. In
17 addition, panel studies, which consist of a relatively small sample (typically tens) of study
18 participants followed over a period of days to months, have been used to examine the health effects
19 associated with exposure to ambient concentrations of air pollutants (e.g., Delfino et al., 1996,
20 080788). Panel studies may also apply a microenvironmental model to represent exposure to an air
21 pollutant.
22 Exposure error can mask epidemiologic associations between pollutants and health outcomes,
23 although this may be less of an issue for O3 because it is a regional pollutant with relatively low
24 spatial variability. For example, a study in Atlanta observed associations between HRV parameters
25 and two traffic-related pollutants (EC and NO2), but the associations were only statistically
26 significant when personal exposures to the pollutants were used in the regression model (Suh and
27 Zanobetti, 2010, 677202). No association was observed with ambient concentrations of these two
28 pollutants. In contrast, associations were similar for ambient concentrations and personal exposures
29 of O3 and PM2 5, another regional pollutant. Further discussion of the effect of spatial variability is
3 0 provided in S ection 4.5.3.
31 The importance of exposure misclassification varies with study design and is dependent on the
32 spatial and temporal aspects of the design. For example, the use of a community-averaged O3
March 2011 4-17 DRAFT - DO NOT CITE OR QUOTE
-------
1 concentration in a community time-series epidemiologic study may not allow for adequate
2 examination of the role of spatial variability. Other factors that could influence exposure estimates
3 include nonambient exposure; spatial and temporal variability, topography of the natural and built
4 environment, and meteorology; measurement errors; use of ambient O3 concentration as a surrogate
5 for ambient O3 exposure; and the presence of O3 in a mixture of pollutants. The following sections
6 will consider various sources of error and how they affect the interpretation of results from
7 epidemiologic studies of different designs.
4.5.2. Nonambient Ozone Exposure
8 For other criteria pollutants, nonambient sources can be an important contributor to total
9 personal exposure. There are relatively few indoor sources of O3, so personal O3 exposure is
10 expected to be dominated by ambient O3 in outdoor microenvironments and in indoor
11 microenvironments with high air exchange rates (e.g., with open windows). Even in
12 microenvironments where nonambient exposure is substantial, such as in a room with an O3
13 generator, this nonambient exposure is unlikely to be temporally correlated with ambient O3
14 exposure (Wilson and Suh, 1997, 077408). and therefore would not affect epidemiologic associations
15 between O3 and the health effect (Sheppard et al., 2005, 079176). In simulations of a nonreactive
16 pollutant, Sheppard et al. (2005, 079176) concluded that nonambient exposure does not influence the
17 health outcome effect estimate if ambient and nonambient concentrations are independent. It should
18 be noted that the effect estimate calculated from using personal exposure to ambient O3 rather than
19 ambient concentration will be increased in inverse proportion to the ratio of ambient exposure to
20 ambient concentration, and daily fluctuations in this ratio can widen the confidence intervals in the
21 ambient concentration effect estimate, but uncorrelated nonambient exposure will not bias the effect
22 estimate.
4.5.3. Spatiotemporal Variability
4.5.3.1. Spatial Variability
23 Compared with directly emitted pollutants such as CO and NOX, O3 exhibits relatively low
24 spatial variability across urban areas, as discussed in Chapter 3. Spatial variability contributes to
25 exposure error if the ambient O3 concentration measured at the central site monitor is used as an
26 ambient exposure surrogate and differs from the actual ambient O3 concentration outside a subject's
27 residence and/or worksite (in the absence of indoor O3 sources). Averaging data from a large number
28 of samplers will dampen intersampler variability, and use of multiple monitors over smaller land
29 areas may allow for more variability to be incorporated into an epidemiologic analysis.
30 Community exposure may not be well represented when monitors cover large areas with
31 several subcommunities having different sources and topographies, such as the Los Angeles CSA
32 (Chapter 3). Ozone monitors in Los Angeles had a much wider range of intermonitor correlations
March 2011 4-18 DRAFT - DO NOT CITE OR QUOTE
-------
1 (-0.06 to 0.97) than Atlanta (0.61 to 0.96) or Boston (0.56 to 0.97) using 2007-2009 data. Although
2 the negative and near-zero correlations in Los Angeles were observed for monitors located some
3 distance apart (>150 km), some closer monitor pairs had low positive correlations, likely due to
4 changes in topography and airflow patterns over short distances. The spatial variability in O3
5 concentration in 24 MS As across the U.S. was examined in the 2006 O3 AQCD (U.S. EPA, 2006,
6 088089). Spatial variability was examined by using Pearson correlation coefficients, values of the
7 90th percentile of the absolute difference in O3 concentrations, and CODs. No clear discernible
8 regional differences across the U.S. were found in the ranges of parameters analyzed.
9 However, spatial variation in atmospheric constituents that participate in O3 formation and
10 titration reactions contribute to spatial variability in some areas, particularly near roadways (Section
11 3.6.2.1). Liu et al. (1995, 039061) conducted a O3 exposure assessment study in Toronto, Canada
12 during the winter and summer of 1992. Results indicated that outdoor O3 concentrations exhibited
13 spatial variation over the study area. The spatial variation is believed to result from population and
14 traffic density. The results are consistent with a smaller-scale study by McKendry (1993, 677201) in
15 Montreal, Canada, in which the investigator examined spatial variation using measurements from
16 nine ambient monitoring sites and found the spatial pattern of O3 concentration to be more distinct in
17 the winter than in the summer.
18 Sarnat et al. (2010, 385852) studied the spatial variability of O3, along with PM2 5, NO2, and
19 CO, in the Atlanta, GA, metropolitan area and evaluated how spatial variability affects interpretation
20 of epidemiologic results, using time-series data for circulatory disease ED visits. The authors found
21 that associations with ambient O3 concentration were similar among all sites tested, including
22 multiple urban sites and a rural site some 38 miles from the city center. This result was also observed
23 for PM2 5, another regional pollutant. In contrast, the more spatially variable pollutants CO and NO2
24 showed different associations for the rural site than the urban sites, although the urban site
25 associations were similar to one another for CO. This suggests that choice of monitor may have little
26 impact on the results of O3 epidemiologic studies.
4.5.3.2. Seasonality
27 The relationship between personal exposure and ambient concentration has been found to vary
28 by season, with at least three factors potentially contributing to this variation: differences in building
29 ventilation (e.g., air conditioning or heater use versus open window ventilation), higher O3
30 concentrations during the O3 season contributing to increased exposure and improved detection by
31 personal monitors; and changes in activity pattern resulting in more time spent outside. Evidence has
32 been presented in studies conducted in several cities regarding the effect of ventilation on personal-
33 ambient and indoor-outdoor O3 relationships (see Sections 4.3.3 and 4.3.4). More limited evidence is
34 available regarding the specific effects of O3 detection limits and activity pattern changes on O3
35 relationships.
36 Several studies have found increased summertime correlations or ratios between personal
37 exposure and ambient concentration (Sarnat et al., 2000, 001852; Sarnat et al., 2005, 087531) or
March 2011 4-19 DRAFT - DO NOT CITE OR QUOTE
-------
1 between indoor and outdoor O3 concentrations (Avol et al., 1998, 018270; Geyh et al, 2000,
2 001775). However, others have found higher ratios in fall than in summer (Sarnat et al., 2006,
3 089784) or equivalent, near-zero ratios in winter and summer (Sarnat et al., 2001, 019401). possibly
4 because summertime use of air conditioners decreases building air exchange rates. It should be noted
5 that O3 concentrations during winter are generally much lower than summertime concentrations,
6 possibly obscuring wintertime relationships due to detection limit issues. Studies specifically
7 evaluating the effect of ventilation conditions on O3 relationships have found increased correlations
8 or ratios for individuals or buildings experiencing higher air exchange rates (Geyh et al., 2000,
9 001775: Romieu et al., 1998, 049834: Sarnat et al., 2000, 001852: Sarnat et al., 2006, 089784).
10 Increased correlations or ratios between personal exposure and ambient concentration, or
11 between indoor and outdoor concentration, are likely to reduce error in exposure estimates used in
12 epidemiologic studies. This suggests that studies conducted during the O3 season or in periods when
13 communities are likely to have high air exchange rates (e.g., during mild weather) may be less prone
14 to exposure error than studies conducted only during winter. Year-round studies that include both the
15 O3 and non-O3 seasons may have an intermediate level of exposure error.
4.5.4. Exposure to Co-pollutants and Ozone Reaction Products
16 Although indoor O3 concentrations are usually well below ambient concentrations, the same
17 reactions that consume O3 indoors form particulate and gaseous species, including other oxidants, as
18 summarized in Section 4.3.4.3. Exposures to these reaction products would therefore be expected to
19 be correlated with ambient O3 concentrations, and could potentially contribute to health effects
20 observed in epidemiologic studies. Current evidence regarding personal exposures to these reaction
21 products is extremely limited.
4.5.5. Exposure Estimation Methods in Epidemiologic Studies
22 The use of O3 measurements from central ambient monitoring sites is the most common
23 method for assigning exposure in epidemiologic studies. However, fixed-site measurements do not
24 account for the effects of spatial variation in O3 concentration, ambient and non-ambient
25 concentration differences, and varying activity patterns on personal exposures (Brown et al., 2009,
26 190895: Chang et al., 2000, 001276: Zeger et al., 2000, 001949). The use of fixed-site
27 concentrations results in minimal exposure error when: (1) O3 concentrations are uniform across the
28 region; (2) personal activity patterns are similar across the population; and (3) housing
29 characteristics, such as air exchange rate and indoor reaction rate, are constant over the study area.
30 Since these factors vary by location and population, there will be errors in the magnitude of total
31 exposure based solely on ambient monitoring data.
32 As described in Section 4.3.4, results from previous and recently published studies indicate
33 that while the relationship between personal exposures and ambient concentrations varies due to a
34 number of factors, such as activity patterns, housing characteristics, and season, O3 concentrations
March 2011 4-20 DRAFT - DO NOT CITE OR QUOTE
-------
1 measured at central-site monitors are representative of day-to-day changes in average personal O3
2 exposure, which is the important parameter for time-series epidemiologic studies. Another important
3 finding is that the magnitude of personal exposures is smaller than concentrations reported at fixed-
4 site monitors due to time spent indoors and the low indoor penetration of O3. his tends to support the
5 use of fixed-site concentrations as exposure estimates in epidemiologic studies.
6 Modeling approaches can also be used to estimate exposures for epidemiologic studies, as
7 discussed in Section 4.4. Geostatistical spatial interpolation techniques can provide finer-scale
8 estimates of local concentration over urban areas. A microenvironmental modeling approach
9 simulates exposure using empirical distributions of concentrations in specific microenvironments
10 together with human activity pattern data. The main advantage of the modeling approach is that it
11 can be used to estimate exposures over a wide range of population and scenarios. A main
12 disadvantage of the modeling approach is that the results of modeling exposure assessment must be
13 compared to an independent set of measured exposure levels (Klepeis, 1999, 001697). In addition,
14 resource-intensive development of validated and representative model inputs is required, such as
15 human activity patterns, distributions of air exchange rate, and deposition rate. Therefore, modeled
16 exposures are used much less frequently in epidemiologic studies.
4.6. Summary and Conclusions
4.6.1. Exposure Measurement
4.6.1.1. Measurement of Ozone Exposure
17 Passive badge samplers are the most widely used technique for measuring personal O3
18 exposure. The operate on the nitrite-nitrate conversion principle, and are convenient since they
19 require no pumps or wet chemistry in the field. They represent a cumulative (rather than continuous)
20 sample, and their detection limit makes them suitable for monitoring periods of 24 hours or greater.
21 his limits their applicability in measuring short-term daily fluctuations in personal exposure. Over a
22 24-h period, the detection limit of the badges is approximately 5-10 ppb, which may result in an
23 appreciable fraction of the samples being below the detection limit. An active sampler based on the
24 nitrite-nitrate conversion reaction is also available, with a reported detection limit of 10 ppb-h,
25 enabling measurement of sub-daily O3 concentrations. A portable continuous O3 monitor based on a
26 different principle, UV absorption, has recently become available. Its size and weight make it
27 suitable for use in a backpack configuration, although its use for personal exposure measurements
28 has been limited.
29 Several studies described in the 2006 O3 AQCD, along with a few new studies published since,
30 describe the relationship between indoor O3 concentration and the O3 concentration immediately
31 outside the indoor microenvironment. These studies show that the indoor concentration is often
32 substantially lower than the outdoor concentration unless indoor sources are present. Low indoor O3
March 2011 4-21 DRAFT-DO NOT CITE OR QUOTE
-------
1 concentrations can be explained by reaction of O3 with surfaces and airborne constituents. However,
2 the indoor-outdoor relationship is greatly affected by the air exchange rate; under conditions of high
3 air exchange rate, such as open windows, the indoor O3 concentration may approach the outdoor
4 concentration. In residential microenvironments, studies report indoor-outdoor ratios ranging from
5 approximately 0.1-0.4, with the highest ratios observed in the summer O3 season and for homes with
6 increased window ventilation. A correlation of 0.58 was reported between indoor and outdoor O3
7 concentrations, indicating that variations in outdoor concentration may be reflected indoors, though
8 the magnitude of the concentration is lower. Indoor-outdoor ratios at schools were similar, with
9 higher ratios observed during the school day when opening doors and windows may lead to
10 increased air exchange rates. In vehicles, high air exchange rates that would normally lead to high
11 interior-exterior concentration ratios are offset by O3 scavenging through vehicle-emitted NO,
12 resulting in reported in-vehicle concentrations that were approximately 50% of those measured at the
13 roadside.
14 The relationship between personal exposure and ambient O3 concentrations has been evaluated
15 in several research studies, many of which were conducted prior to 2005 and are discussed in the
16 2006 O3 AQCD. The results of these studies indicate that personal exposures are moderately well
17 correlated with ambient concentrations, and that the ratio of personal exposure to ambient
18 concentration is higher in outdoor microenvironments and during the summer season. In situations
19 where a lack of correlation was observed, this may be due in part to a high proportion of personal
20 measurements below the detection limit. Correlations reported for daily or multi-day measurements
21 range from approximately 0.3-0.8, with the upper end of the range reflecting longer-duration (4-day)
22 community average measurements that may limit the influence of inter-individual variability in
23 exposure. Hourly measurements in specific microenvironments show greater variability in
24 correlations between personal exposure and ambient concentration, with residential indoor
25 correlations <0.1 and outdoor correlations of 0.7-0.9. Slopes from regression analyses of personal
26 exposure on ambient concentration generally ranged from approximately 0.1-0.3. Higher slopes were
27 observed in studies that either adjusted for activity pattern and air exchange rate (0.54) or focused on
28 outdoor shoe cleaners (0.56), who may have increased exposure due to spending a substantial
29 fraction of the day outdoors. Ratios of personal exposure to ambient concentration showed similar
30 results, with a ratio of 0.3 reported for a year-round study in southern California, while ratios ranged
31 from 0.28-0.96 for outdoor workers, increasing with time spent outdoors.
32 Taken together, results from previous and recently published studies indicate that while the
33 relationship between personal exposures and ambient concentrations varies due to a number of
34 factors, such as activity patterns, housing characteristics, and season, O3 concentrations measured at
35 central-site monitors are representative of day-to-day changes in average personal O3 exposure,
36 which is the important parameter for time-series epidemiologic studies. Another important finding is
37 that the magnitude of personal exposures is smaller than concentrations reported at fixed-site
38 monitors due to time spent indoors and the low indoor penetration of O3.
March 2011 4-22 DRAFT - DO NOT CITE OR QUOTE
-------
4.6.1.2. Co-Exposure to Ozone and Other Pollutants
1 Individuals may be exposed to other pollutants in conjunction with exposure to O3. Personal
2 exposure to O3 shows variable association with personal exposure to other pollutants, with
3 differences in association depending on factors such as season, city-specific characteristics, and
4 spatial variability of the co-pollutant. For PM2 5, a rank correlation of 0.14 was reported between
5 daily O3 and PM2 5 exposures during spring and fall in Atlanta. Positive slopes were reported during
6 summer in both Baltimore and Boston, although the slopes were somewhat different (0.21 and 0.72,
7 respectively). The summertime slope in Baltimore was higher for children (0.37) than for adults
8 (0.07), which may be the result of different activity patterns and time spent outdoors. Additional
9 evidence of variation by season and city is provided by the differing signs of the wintertime slopes,
10 with Baltimore showing a negative slope and Boston showing a positive slope. Interindividual
11 variability likely played a role as well, since both cities showed a wide range (including both
12 negative and positive values) for individual-specific personal O3- PM2 5 slopes. For EC and NO2,
13 near-zero correlations were reported with O3 during spring and fall in Atlanta. These extremely low
14 correlations for the traffic-related and spatially variable pollutants EC and NO2 contrast with the
15 higher correlation observed for PM25, a regional pollutant.
16 In near-road and on-road microenvironments, correlations between O3 and traffic-related
17 pollutants are moderately to strongly negative, with the most strongly negative correlations observed
18 for NO2 (-0.8 to -0.9). This is consistent with the chemistry of NO oxidation, in which O3 is
19 consumed to form NO2. The more moderate negative correlations observed for PM2 5, PMLO, and
20 VOC may reflect reduced concentrations of O3 in more polluted environments due to other
21 scavenging reactions. A similar process occurs indoors, where infiltrated O3 reacts with airborne or
22 surface-associated materials to form secondary compounds, such as formaldehyde. Although such
23 reactions decrease indoor O3 exposure, they result in increasing exposure to other species which may
24 themselves have health effects.
4.6.2. Exposure Modeling
25 Exposures estimates in urban areas may be improved by constructing a concentration surface
26 over a geographic domain using a model to compensate for missing data. The calculated
27 concentration surface can then be used to estimate exposures outside residences, schools,
28 workplaces, roadways, or other locations of interest. This technique does not estimate exposure
29 directly because it does not account for activity patterns or concentrations in different
30 microenvironments. Most such modeling efforts have focused on the less-reactive pollutants PM or
31 NO2. In a study that extended CALINE4 NOX modeling results to evaluate the impact on residential
32 O3 concentrations (see Section 4.4.1), O3 concentrations were reduced by 0.51 ppb O3 per 1 ppb
33 NOX. This intra-urban traffic-related variability in O3 concentrations suggests that differences in
34 traffic density between the central site monitor and individual homes could result in either an
35 overestimate or underestimate of residential O3.
March 2011 4-23 DRAFT - DO NOT CITE OR QUOTE
-------
1 A separate class of models, known as microenvironmental models, estimate time-weighted
2 exposure for modeled individuals by summing exposure in each microenvironment visited during the
3 exposure period. Stochastic microenvironmental models, such as APEX and SHEDS, utilize
4 distributions of pollutant-related and individual-level variables, such as ambient and local O3
5 concentration contributions and breathing rate respectively, to compute the distribution of individual
6 exposures across the modeled population. The models also have the capability to estimate received
7 dose through a dosimetry model. Using distributions of input parameters in the model framework
8 rather than point estimates allows the models to incorporate uncertainty and variability explicitly into
9 exposure estimates. For the APEX model, an analysis has been conducted indicating that the
10 uncertainty in model exposure estimates for asthmatic children during moderate exercise is small to
11 moderate; however, APEX appears to substantially underestimate the frequency of multiple high-
12 exposure events for a single individual. Microenvironmental models, such as EMI, are also being
13 developed to use individual-specific information derived from measurements or questionnaires,
14 rather than population distributions, to estimate exposures. This approach is particularly suitable for
15 panel health studies where information is available for each participant, and may reduce uncertainty
16 in health effect estimates by improving exposure estimates.
4.6.3. Implications for Epidemiologic Studies
17 Exposure error can be an important contributor to variability in epidemiologic study results,
18 although this may be less of an issue for O3 because it is a secondary pollutant with relatively low
19 spatial variability across an urban area. For example, an epidemiologic study in Atlanta observed
20 similar associations between HRV parameters and either ambient concentrations or personal
21 exposures of O3 and PM2 5, another regional pollutant. The importance of exposure error varies with
22 study design and is dependent on the spatial and temporal aspects of the design. Several factors that
23 could influence exposure estimates include nonambient exposure, spatial and temporal variability,
24 and the presence of O3 in a mixture of pollutants. Nonambient exposure is unlikely to influence
25 health effect estimates because of the lack of indoor O3 sources and because indoor-generated O3
26 exposures are unlikely to be correlated with ambient O3 exposure. Compared with directly emitted
27 pollutants such as CO and NOX, O3 exhibits relatively low spatial variability across urban areas, as
28 discussed in Chapter 3. Averaging data from a large number of samplers will dampen intersampler
29 variability, and use of multiple monitors over smaller land areas may allow for more variability to be
30 incorporated into an epidemiologic analysis. Evidence from a study comparing the effect of spatial
31 variability on effect estimates for O3, PM2 5, NO2, and CO suggests that choice of monitor for more
32 spatially homogenous pollutants such as O3 may have little impact on the results of epidemiologic
33 studies. Season, however, may have a substantial effect due to much lower O3 concentrations during
34 the winter, along with the higher correlations between ambient concentrations and personal
35 exposures observed during the summer. Studies conducted during the O3 season or in periods when
36 communities are likely to have high air exchange rates (e.g., during mild weather) may be less prone
37 to exposure error than studies conducted only during winter. Year-round studies that include both the
March 2011 4-24 DRAFT - DO NOT CITE OR QUOTE
-------
1 O3 and non-O3 seasons may have an intermediate level of exposure error. Exposure to mixtures of
2 pollutants containing O3 also complicates interpretation of epidemiologic results. Moderate to strong
3 negative correlations between O3 and traffic-related pollutants, particularly NO2, make it difficult to
4 determine to what extent O3-based effect estimates quantitatively reflect the independent effect of O3
5 itself, or the effect of another pollutant or pollutants in the mixture. Interpretation of O3 effects in the
6 presence of PM is additionally complicated by the highly variable correlations observed, which
7 differ by city, season, and population characteristics (e.g., children versus adults). Although these
8 sources of exposure error should be considered in evaluating epidemiologic results, previous and
9 recently published exposure research indicate that O3 concentrations measured at central-site
10 monitors are indicative of day-to-day changes in average personal O3 exposure, making ambient
11 concentrations a useful parameter for epidemiologic studies.
March 2011 4-25 DRAFT - DO NOT CITE OR QUOTE
-------
References
A list of all references considered for inclusion in this chapter can be found at
http://hero.epa. gov/index.cfm?action=search.do&submit=Search&group_id=475
Ambient Air Monitoring Reference and Equivalent Methods: Designation of One New Equivalent Method (EPA Office of
Research and Development). 75 Fed. Reg. 22126-22127 (2010). 687659
Anderson, S. E.; Wells, J. R.; Fedorowicz, A.; Butterworth, L. F.; Meade, B. I; Munson, A. E. (2007). Evaluation of the
contact and respiratory sensitization potential of volatile organic compounds generated by simulated indoor air
chemistry. Toxicol Sci, 97: 355-363. http://dx.doi.org/10.1093/toxsci/kfm043 600123
Aoki, T.; Tanabe, S. (2007). Generation of sub-micron particles and secondary pollutants from building materials by ozone
reaction. Atmos Environ, 41: 3139-3150. http://dx.doi.Org/10.1016/j.atmosenv.2006.07.053 449034
Avol, E. L.; Navidi, W. C.; Colome, S. D. (1998). Modeling ozone levels in and around southern California homes. Environ
Sci Technol, 32: 463-468. 018270
Beckerman, B.; Jerrett, M.; Brook, J. R.; Verma, D. K.; Arain, M. A.; Finkelstein, M. M. (2008). Correlation of nitrogen
dioxide with other traffic pollutants near a major expressway. Atmos Environ, 42: 275-290. 096484
Beelen, R.; Hoek, G; Pebesma, E.; Vienneau, D.; de Hoogh, K.; Briggs, D. J. (2009). Mapping of background air pollution
at a fine spatial scale across the European Union. Sci Total Environ, 407: 1852-1867.
http://dx.doi.0rg/10.1016/i.scitotenv.2008.ll.048601157
Beko, G; Clausen, G; Weschler, C. J. (2007). Further studies of oxidation processes on filter surfaces: Evidence for
oxidation products and the influence of time in service. Atmos Environ, 41: 5202-5212.
http://dx.doi.0rg/10.1016/i.atmosenv.2006.07.063601190
Bell, M. L. (2006). The use of ambient air quality modeling to estimate individual and population exposure for human
health research: a case study of ozone in the Northern Georgia region of the United States. Environ Int, 32: 586-
593. 194358
Blondeau, P.; lordache, V; Poupard, O.; Genin, D.; Allard, F. (2005). Relationship between outdoor and indoor air quality
in eight French schools. Indoor Air, 15: 2-12. 078044
Brauer, M.; Brook, J. R. (1997). Ozone personal exposures and health effects for selected groups residing in the Fraser
Valley. Atmos Environ, 31: 2113-2121. 083339
Brauer, M.; Lencar, C.; Tamburic, L.; Koehoorn, M.; Demers, P.; Karr, C. (2008). A cohort study of traffic-related air
pollution impacts on birth outcomes. Environ Health Perspect, 116: 680-686. 156292
Briggs, D. J.; Collins, S.; Elliott, P.; Fischer, P.; Kingham, S.; Lebret, E.; Pryl, K.; Van Reeuwijk, H.; Smallbone, K.; Van
Der Veen, A. (1997). Mapping urban air pollution using GIS: a regression-based approach. Int J Geogr Inform Sci,
11:699-718.025950
Brown, K.; Sarnat, J.; Suh, H.; Coull, B.; Koutrakis, P. (2009). Factors influencing relationships between personal and
ambient concentrations of gaseous and particulate pollutants. Sci Total Environ, 407: 3754-3765. 190895
Burke, J. M.; Zufall, M. J.; Ozkaynak, H. (2001). A population exposure model for particulate matter: case study results for
PM2.5 in Philadelphia, PA. J Expo Sci Environ Epidemiol, 11: 470-489. 014050
Chang, L.-T; Koutrakis, P.; Catalano, P. J.; Suh, H. H. (2000). Hourly personal exposures to fine particles and gaseous
pollutants—results from Baltimore, Maryland. J Air Waste Manag Assoc, 50: 1223-1235. 001276
Christakos, G; Vyas, V M. (1998). A composite space/time approach to studying ozone distribution over eastern United
States. Atmos Environ, 32: 2845-2857. 029728
Christakos, G; Vyas, V M. (1998). A novel method for studying population health impacts of spatiotemporal ozone
distribution. Soc Sci Med, 47: 1051-1066. 011987
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments such as the Integrated Science Assessments (ISAs) and the Integrated Risk Information System (IRIS).
March 2011 4-26 DRAFT - DO NOT CITE OR QUOTE
-------
Delfino, R. I; Coate, B. D.; Zeiger, R. S.; Seltzer, J. M.; Street, D. H; Koutrakis, R (1996). Daily asthma severity in
relation to personal ozone exposure and outdoor fungal spores. Am J Respir Crit Care Med, 154: 633-641. 080788
Gabusi, V; Volta, M. (2005). Seasonal modelling assessment of ozone sensitivity to precursors in northern Italy. Atmos
Environ, 39: 2795-2804. http://dx.doi.Org/10.1016/j.atmosenv.2004.07.041 606112
Georgopoulos, P. G; Purushothaman, V.; Chiou, R. (1997). Comparative evaluation of methods for estimating potential
human exposure to ozone: photochemical modeling and ambient monitoring. J Expo Sci Environ Epidemiol, 7:
191-215. 083591
Georgopoulos, P. G; Wang, S.-W.; Vyas, V. M.; Sun, Q.; Burke, J.; Vedantham, R.; McCurdy, T.; Ozkaynak, H. (2005). A
source-to-dose assessment of population exposures to fine PM and ozone in Philadelphia, PA, during a summer
1999 episode. J Expo Sci Environ Epidemiol, 15: 439-457. 080269
Geyh, A. S.; Roberts, P. T.; Lurmann, F. W.; Schoell, B. M.; Avol, E. L. (1999). Initial field evaluation of the Harvard active
ozone sampler for personal ozone monitoring. J Expo Sci Environ Epidemiol, 9: 143-149. 016908
Geyh, A. S.; Wolfson, J. M.; Koutrakis, P.; Mulik, J. D.; Avol, E. L. (1997). Development and evaluation of a small active
ozone sampler. Environ Sci Technol, 31: 2326-2330. 086151
Geyh, A. S.; Xue, J.; Ozkaynak, H.; Spengler, J. D. (2000). The Harvard Southern California chronic ozone exposure study:
assessing ozone exposure of grade-school-age children in two southern California communities. Environ Health
Perspect, 108: 265-270. 001775
Gilliland, F.; Avol, E.; Kinney, P.; Jerrett, M.; Dvonch, T.; Lurmann, F.; Buckley, T.; Breysse, P.; Keeler, G; de Villiers, T.;
McConnell, R. (2005). Air pollution exposure assessment for epidemiologic studies of pregnant women and
children: lessons learned from the Centers for Children's Environmental Health and Disease Prevention Research.
Environ Health Perspect, 113: 1447-54. 098820
Hoek, G; Beelen, R.; de Hoogh, K.; Vienneau, D.; Gulliver, J.; Fischer, P.; Briggs, D. (2008). A review of land-use
regression models to assess spatial variation of outdoor air pollution. Atmos Environ, 42: 7561-7578. 195851
Hooyberghs, J.; Mensink, C.; Dumont, G; Fierens, F. (2006). Spatial interpolation of ambient ozone concentrations from
sparse monitoring points in Belgium. J Environ Monit, 8: 1129-1135. http://dx.doi.org/10.1039/b612607n 608180
Hyttinen, M.; Pasanen, P.; Kalliokoski, P. (2006). Removal of ozone on clean, dusty and sooty supply air filters. Atmos
Environ, 40: 315-325. 134392
Jerrett, M.; Burnett, R. T.; Pope, C. A.; Ito, K.; Thurston, G; Krewski, D.; Shi, Y; Calle, E.; Thun, M. (2009). Long-term
ozone exposure and mortality. N Engl J Med, 360: 1085-1095. http://dx.doi.org/10.1056/NEJMoa0803894 194160
Klepeis, N. E. (1999). An introduction to the indirect exposure assessment approach: modeling human exposure using
microenvironmental measurements and the recent National Human Activity Pattern Survey. Environ Health
Perspect, 107: 365-374. 001697
Klepeis, N. E.; Nelson, W. C.; Ott, W. R.; Robinson, J. P.; Tsang, A. M.; Switzer, P.; Behar, J. V; Hern, S. C.; Engelmann,
W. H. (2001). The National Human Activity Pattern Survey (NHAPS): a resource for assessing exposure to
environmental pollutants. J Expo Sci Environ Epidemiol, 11: 231-252. 002437
Koutrakis, P.; Wolfson, J. M.; Bunyaviroch, A.; Froehlich, S. E.; Hirano, K.; Mulik, J. D. (1993). Measurement of ambient
ozone using a nitrite-coated filter. Anal Chem, 65: 209-214. 202808
Langstaff, J. E. (2007). Analysis of uncertainty in ozone population exposure modeling [technical memorandum]. Research
Triangle Park, NC: U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards. 090315
Lee, K.; Parkhurst, W. J.; Xue, J.; Ozkaynak, H.; Neuberg, D.; Spengler, J. D. (2004). Outdoor/indoor/personal ozone
exposures of children in Nashville, Tennessee. J Air Waste Manag Assoc, 54: 352-359. 055599
Liard, R.; Zureik, M.; Le Moullec, Y; Soussan, D.; Glorian, M.; Grimfeld, A.; Neukirch, F. (1999). Use of personal passive
samplers for measurement of NO2, NO, and O3 levels in panel studies. Environ Res, 81: 339-348. 001426
Linn, W. S.; Shamoo, D. A.; Anderson, K. R.; Peng, R.-C.; Avol, E. L.; Hackney, J. D.; Gong, H. Jr (1996). Short-term air
pollution exposures and responses in Los Angeles area schoolchildren. J Expo Sci Environ Epidemiol, 6: 449-472.
082508
Liu L-JS; Koutrakis, P.; Leech, J.; Broder, I. (1995). Assessment of ozone exposures in the greater metropolitan Toronto
area. J Air Waste Manag Assoc, 45: 223-234. 039061
March 2011 4-27 DRAFT - DO NOT CITE OR QUOTE
-------
Marshall, J. D.; Nethery, E.; Brauer, M. (2008). Within-urban variability in ambient air pollution: comparison of estimation
methods. Atmos Environ, 42: 1359-1369. http://dx.doi.Org/10.1016/j.atmosenv.2007.08.012 193983
McConnell, R.; Berhane, K.; Yao, L.; Lurmann, F. W.; Avol, E.; Peters, J. M. (2006). Predicting residential ozone deficits
from nearby traffic. Sci Total Environ, 363: 166-174. 089256
McCurdy, T; Glen, G; Smith, L.; Lakkadi, Y. (2000). The National Exposure Research Laboratory's consolidated human
activity database. J Expo Sci Environ Epidemiol, 10: 566-578. 000782
McKendry, I. G. (1993). Ground-level ozone in Montreal, Canada. Atmos Environ, 27: 93-103.
http://dx.doi.org/10.1016/0957-1272(93)90049-0 677201
O'Neill, M. S.; Ramirez-Aguilar, M.; Meneses-Gonzalez, F.; Hernandez-Avila, M.; Geyh, A. S.; Sienra-Monge, J. J.;
Romieu, I. (2003). Ozone exposure among Mexico City outdoor workers. J Air Waste Manag Assoc, 53: 339-346.
042752
Ogawa & Company (2007). Ambient air passive sampler for NO-NO2, NOx, SO2, O3, NH3. Retrieved June 17, 2008 from
http://www.ogawausa.com/passive.html. 090937
Pakalapati, S.; Beaver, S.; Romagnoli, J. A.; Palazoglu, A. (2009). Sequencing diurnal air flow patterns for ozone exposure
assessment around Houston, Texas. Atmos Environ, 43: 715-723. http://dx.doi.Org/10.1016/j.atmosenv.2008.09.054
615781
Ramirez-Aguilar, M.; Barraza-Villarreal, A.; Moreno-Macias, H.; Winer, A. M.; Cicero-Fernandez, P.; Velez-Marquez, M.
G; Cortez-Lugo, M.; Sienra-Monge, J. J.; Romieu, I. (2008). Assessment of personal exposure to ozone in
asthmatic children residing in Mexico City. Salud Publica Mex, 50: 67-75. 098930
Reiss, R.; Ryan, P. B.; Koutrakis, P.; Tibbetts, S. J. (1995). Ozone reactive chemistry on interior latex paint. Environ Sci
Technol, 29: 1906-1912. 078727
Reiss, R.; Ryan, P. B.; Tibbetts, S. J.; Koutrakis, P. (1995). Measurement of organic acids, aldehydes, and ketones in
residential environments and their relation to ozone. J Air Waste Manag Assoc, 45: 811-822. 078743
Riediker, M.; Williams, R.; Devlin, R.; Griggs, T; Bromberg, P. (2003). Exposure to particulate matter, volatile organic
compounds, and other air pollutants inside patrol cars. Environ Sci Technol, 37: 2084-2093. 043761
Romieu, L; Lugo, M. C.; Colome, S.; Garcia, A. M.; Avila, M. H.; Geyh, A.; Velasco, S. R.; Rendon, E. P. (1998).
Evaluation of indoor ozone concentration and predictors of indoor-outdoor ratio in Mexico City. J Air Waste Manag
Assoc, 48: 327-335. 049834
Rundell, K. W.; Caviston, R.; Hollenbach, A. M.; Murphy, K. (2006). Vehicular air pollution, playgrounds, and youth
athletic fields. Inhal Toxicol, 18: 541-547. 089785
Ryan, P. H.; LeMasters, G. K. (2007). A review of land-use regression models for characterizing intraurban air pollution
exposure. Inhal Toxicol, 19: 127. 156063
Sahu, S. K.; Yip, S.; Holland, D. M. (2009). Improved space-time forecasting of next day ozone concentrations in the
eastern US. Atmos Environ, 43: 494-501. http://dx.doi.Org/10.1016/j.atmosenv.2008.10.028 618189
Sarnat, J. A.; Brown, K. W.; Schwartz, J.; Coull, B. A.; Koutrakis, P. (2005). Ambient gas concentrations and personal
particulate matter exposures: implications for studying the health effects of particles. Epidemiology, 16: 385-395.
087531
Sarnat, J. A.; Koutrakis, P.; Suh, H. H. (2000). Assessing the relationship between personal particulate and gaseous
exposures of senior citizens living in Baltimore, MD. J Air Waste Manag Assoc, 50: 1184-1198. 001852
Sarnat, J. A.; Schwartz, J.; Catalano, P. J.; Suh, H. H. (2001). Gaseous pollutants in particulate matter epidemiology:
Confounders or surrogates? Environ Health Perspect, 109: 1053-1061. 019401
Sarnat, S. E.; Coull, B. A.; Schwartz, J.; Gold, D. R.; Suh, H. H. (2006). Factors affecting the association between ambient
concentrations and personal exposures to particles and gases. Environ Health Perspect, 114: 649-654. 089784
Sarnat, S. E.; Klein, M.; Sarnat, J. A.; Flanders, W. D.; Waller, L. A.; Mulholland, J. A.; Russell, A. G; Tolbert, P. E. (2010).
An examination of exposure measurement error from air pollutant spatial variability in time-series studies. J Expo
Sci Environ Epidemiol, 20: 135-146. http://dx.doi.org/10.1038/jes.2009.10 385852
Sheppard, L.; Slaughter, J. C.; Schildcrout, J.; Liu L-JS; Lumley, T. (2005). Exposure and measurement contributions to
estimates of acute air pollution effects. J Expo Sci Environ Epidemiol, 15: 366-376. 079176
March 2011 4-28 DRAFT - DO NOT CITE OR QUOTE
-------
Suh, H. H.; Zanobetti, A. (2010). Exposure error masks the relationship between traffic-related air pollution and heart rate
variability. J Occup Environ Med, 52: 685-692. http://dx.doi.org/10.1097/JOM.Ob013e3181e8071f677202
U.S. EPA (2006). Air quality criteria for ozone and related photochemical oxidants (Report No. EPA/600/R-05/004AF).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid= 149923. 088089
U.S. EPA (2008). Integrated science assessment for oxides of nitrogen - Health criteria (Report No. EPA/600/R-08/071).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development.
http://cfpub.epa. go v/ncea/cfm/recordisplay. cfm?deid= 194645. 157073
U.S. EPA (2009). Consolidated Human Activity Database. Retrieved August 27, 2009 from http://www.epa.gov/chadnetl/.
194010
U.S. EPA (2009). Human exposure modeling - Air pollutants exposure model (APEX/TRIM.Expo Inhalation). Retrieved
August 27, 2009 from http://www.epa.gov/ttn/fera/human_apex.html. 194009
Wallace, L.; Williams, R.; Suggs, J.; Jones, P. (2006). Estimating contributions of outdoor fine particles to indoor
concentrations and personal exposures: effects of household characteristics and personal activities (Report No.
EPA/600/R-06/023). Research Triangle Park, NC: U.S. Environmental Protection Agency; National Exposure
Research Laboratory. 089190
Weschler, C. J.; Shields, H. C. (1997). Potential reactions among indoor pollutants. Atmos Environ, 31: 3487-3495. 084530
White, I. R.; Martin, D.; Munoz, M. P.; Petersson, F. K.; Henshaw, S. J.; Nickless, G; Lloyd-Jones, G. C.; Clemitshaw, K.
C.; Shallcross, D. E. (2010). Use of reactive tracers to determine ambient OH radical concentrations: Application
within the indoor environment. Environ Sci Technol, 44: 6269-6274. http://dx.doi.org/10.1021/es901699a 633874
Wilson, K. L.; Birks, J. W. (2006). Mechanism and elimination of a water vapor interference in the measurement of ozone
by UV absorbance. Environ Sci Technol, 40: 6361-6367. http://dx.doi.org/10.1021/es052590c 595155
Wilson, W. E.; Mage, D. T; Grant, L. D. (2000). Estimating separately personal exposure to ambient and nonambient
particulate matter for epidemiology and risk assessment: why and how. J Air Waste Manag Assoc, 50: 1167-1183.
010288
Wilson, W. E.; Suh, H. H. (1997). Fine particles and coarse particles: concentration relationships relevant to epidemiologic
studies. J Air Waste Manag Assoc, 47: 1238-1249. 077408
Xue, J.; Liu, S. V; Ozkaynak, H.; Spengler, J. D. (2005). Parameter evaluation and model validation of ozone exposure
assessment using Harvard Southern California Chronic Ozone Exposure Study data. J Air Waste Manag Assoc, 55:
1508-1515.087532
Zeger, S. L.; Thomas, D.; Dominici, F.; Samet, J. M.; Schwartz, J.; Dockery, D.; Cohen, A. (2000). Exposure measurement
error in time-series studies of air pollution: concepts and consequences. Environ Health Perspect, 108: 419-426.
001949
Zidek, J. V.; Shaddick, G; Meloche, J.; Chatfield, C.; White, R. (2007). A framework for predicting personal exposures to
environmental hazards. Environ Ecol Stat, 14: 411-431. 190076
March 2011 4-29 DRAFT - DO NOT CITE OR QUOTE
-------
Chapter 5. Dosimetry and Mode of
Action
5.1. Human and Animal Ozone Dosimetry
5.1.1. Introduction
1 Dosimetry refers to the measurement or estimation of the quantity of or rate at which a
2 chemical and/or its reaction products are absorbed and retained at target sites. The measurement of
3 the dose of reactive gases such as O3 can range in refinement from their concentration in the ambient
4 exposure atmosphere to the "effective" dose of the chemical or its reaction products that actively
5 participate in toxic reactions (Dahl, 1990, 094536). Thus, the units for the expression of the dose of
6 O3 might range from the concentration of gas in the air (units of ppm or mg/m3), to the quantity of
7 gas inhaled as the product of gas concentration x minute ventilation x time (units of ppm x L x h), to
8 the quantity of gas retained by the whole body, to the concentration of gas molecules that have been
9 absorbed or reacted with the tissue (moles/g tissue weight). In modeling studies, the dose rate is
10 often expressed as a flux per unit of surface area of a region of respiratory epithelium.
11 Ozone is a highly reactive though poorly water soluble gas. The latter feature is believed to be
12 the reason why it is able to penetrate into targets in the lower respiratory tract. The fact that it is so
13 chemically reactive has suggested to some that its effective dose at the target sites exists in the form
14 of secondary oxidation products such as aldehydes and peroxides. Reaction products are formed
15 when O3 interacts with components of the extracellular lining fluid (ELF) such as lipids and
16 antioxidants. Ozone toxicity is observed to some extent in the nasal cavity, however further toxicity
17 exists in the deep lung where the ELF thickness narrows allowing O3 to react directly with cells
18 protruding from the ELF and surface macrophages. Ozone uptake relates directly to these ELF
19 substrate reactions and is termed "reactive absorption." Thus the uptake of O3 is related to both the
20 concentration of O3 as well as the availability of substrates within the ELF.
21 Two types of measurement have been used to arrive at the O3 dose to target sites during
22 breathing: (1) measurement of removal of O3 from the air stream (termed "uptake"); and (2)
23 measurement of chemical reactions or the product of those reactions with tissues or with
24 biomolecules known to be present in tissues (termed "reactions"). The results of the above
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments such as the integrated Science Assessments (ISAs) and the Integrated Risk Information System (IRIS).
March 2011 5-1 DRAFT - DO NOT CITE OR QUOTE
-------
1 measurements have been incorporated into mathematical models for the purpose of explaining,
2 predicting, and extrapolating O3 dose in different exposure scenarios of interest.
3 This chapter is intended as an update of the past O3 AQCDs (U.S. EPA, 1996, 017831:
4 U.S. EPA, 2006, 088089) and restates the basic concepts derived from O3 dosimetry literature
5 presented in previous documents as well as introduces the recent relevant literature. Particular
6 attention is given to dosimetric factors influencing individual susceptibility to adverse effects from
7 O3 and factors that affect the ability to extrapolate between species (e.g., experimental animal to
8 human). As there have been few O3 dosimetry studies since the last AQCD, the reader is referred to
9 previous documents (U.S. EPA, 1996, 017831: U.S. EPA, 2006, 088089) for more detailed
10 discussion of the literature. Evaluation of the progress in the interpretation of past dosimetry studies,
11 as well as studies published since 2005, in the areas of reactions, uptake, and models for O3
12 dosimetry, is discussed in the following sections.
5.1.2. Ozone Reactions and Reaction Products
5.1.2.1. Summary of Findings from 2006 Ozone AQCD
13 Ozone dose can be examined by the chemical reactions or the products of these reactions that
14 result from O3 exposure. Since O3 is chemically reactive with a wide spectrum of biomolecules, it is
15 not feasible to delineate its many reaction products. Measurements of reaction formation have
16 included either the loss of a specific molecule and appearance of plausible products, or the addition
17 of O3-derived oxygen to biomolecules through the use of oxygen-18 labeling. In vitro exposure of
18 ELF shows that O3 disappearance from the gas phase depends on the characteristics of the ELF
19 substrates (Hu et al, 1994, 041323: Postlethwait et al, 1998, 086754).
20 To gain access to the underlying cellular compartments, O3 must dissolve at the air-liquid
21 interface of the airway surface and travel through the ELF layer. The ELF is comprised of the airway
22 surface lining that includes the periciliary layer and overlying mucus layer, and the alveolar surface
23 lining that includes the subphase of liquid and vesicular surfactant and the surfactant monolayer.
24 There is a progressive decrease in ELF thickness and increase in interfacial surface with progression
25 from the large airways to the alveolus, with the mucus coating becoming patchy in the distal
26 conducting airways (Figure 5-1). Some cells, such as macrophages, may protrude into the gas phase,
27 allowing for direct contact between O3 and cell membranes. The progressive thinning in the ELF
28 decreases the distance O3 must travel to reach the cellular tissue layer. A computational fluid
29 dynamics (CFD) model was able to predict experimentally measured O3 uptake, but only with nasal
30 mucus layer thickness considered (Cohen-Hubal et al., 1996, 043785). reaffirming the importance of
31 the resistance imparted by the ELF layer in dose and lesion patterns in the nasal passage.
32 Taking into account the high reactivity and low water solubility of O3, calculations suggest
33 that O3 will not penetrate ELF layers greater than 0.1 (im without being transformed to other more
34 long-lived reactive species, thus initiating a reaction cascade (Pryor, 1992, 042725). It follows that
March 2011 5-2 DRAFT - DO NOT CITE OR QUOTE
-------
1 the ELF should be considered an important target for 03 toxicity in the airways. Experimental
2 support for this concept comes from several studies which measure the total oxygen-addition product
3 of O3 reactions in the airways through the use of oxygen-18 labeled O3. High concentrations of O3
4 reaction products are found in the nasal lavage cells, bronchoalveolar lavage (BAL), mucus,
5 surfactant, and cells, and in the epithelial cells of the lower airways, providing evidence that O3
6 reacts at the air-liquid interface. Model calculations of the nasal cavity based on diffusion equations
7 and reaction rates of O3 with model substrates predict an O3 penetration distance (0.5 (im) less than
8 the thickness of the mucus layer (10 (im) (Santiago et al., 2001, 019841). Thus, O3 may cause injury
9 both by direct reaction with constituents of the lining layer and cells protruding from it, and by
10 initiating a reaction cascade that carries the oxidative burden deeper into the tissues.
a.
Bronchus
b.
Air
Liquid
in iiini
Tissue
Tissue
Air
Air
Source: Panel (a) reproduced with permission from McGraw-Hill (Weibel, 1980, 0798481.
Figure 5-1. Structure of lower airways with progression from the large airways to the alveolus.D
Panel (a) illustrates basic airway anatomy. Structures are epithelial cells, EP;
basement membrane, BM; smooth muscle cells, SM; and fibrocartilaginous coat, FC.
Panel (b) illustrates the relative amounts of liquid, tissue, and blood with distal
progression. In the bronchi there is a thick surface lining over a relatively thick layer
of tissues. With distal progress, the lining diminishes allowing increased access of
compounds crossing the air-liquid interface to the tissues and the blood.
March 2011
5-3
DRAFT - DO NOT CITE OR QUOTE
-------
1 Ozone may interact with many of the components that make up the ELF including
2 phospholipids, neutral lipids, free fatty acids, proteins, and low molecular weight antioxidants
3 (Perez-Gil, 2008, 198890: Uppu et al., 1995, 076212V The composition of ELF will vary as a
4 function of anatomic location, species, strain, and likely exposure due to substrate depletion.
5 Reduced substrates enter the ELF from the cellular layer or are transported across the cell layer from
6 blood plasma, or can be regenerated from previously oxidized moieties that are reduced by other
7 reduction reactions.
8 Ozone reacts with the double bond of lipids such as unsaturated fatty acids, a large component
9 of ELF, to form stable and less reactive ozonide, aldehyde, and hydroperoxide reaction products via
10 chemical reactions such as the Criegee ozonolysis mechanism (Figure 5-2) (Pryor et al., 1991,
11 042482). Lipid ozonation products, such as the aldehydes hexanal, heptanal, and nonanal, have been
12 recovered after O3 exposure in human bronchial alveolar lavage fluid (BALF), rat BALF, isolated rat
13 lung, and in vitro systems (Frampton et al., 1999, 040757; Postlethwait et al., 1998, 086754; Pryor et
14 al., 1996, 082697). It was estimated that 88% of the O3 that does not come in contact with
15 antioxidants will react with unsaturated fatty acids including phospholipids or neutral lipids in the
16 ELF (Uppu et al., 1995, 076212). Nonanal has been suggested as a relatively specific biomarker for
17 O3 exposure since the monounsaturated fatty acid parent compound, oleic acid, does not undergo
18 autoxidation (Pryor et al., 1996, 082697). Adducts of the aldehyde 4-hydroxynonenal were found in
19 human alveolar macrophages after O3 exposure (Hamilton et al., 1998, 086157). Polyunsaturated
20 fatty acid (PUFA) reactions are limited by the availability of O3 since lipids are so abundant in the
21 ELF. Yields of O3-induced aldehydes were increased by the decrease in other substrates such as
22 ascorbic acid (AH2) (Postlethwait et al., 1998, 086754). PUFA reactions may not generate sufficient
23 bioactive materials to account for acute cell injury, however only modest amounts of products may
24 be necessary to induce cytotoxicity (Postlethwait and Ultman, 2001, 196805; Postlethwait et al.,
25 1998, 086754).
March 2011 5-4 DRAFT - DO NOT CITE OR QUOTE
-------
RHC = CH
PUFA
03-
ozone
o o
! I
RHC — CH-
trioxolane
RHC = O —O + RHC = O
carbonyl oxide aldehyde
either in
the —
absence
ofH2O
/O~O\ or in the
RHC CH— presence
\o/ ofH2O
Criegee ozonide
XOH
-* RHC - >
OOH
hydroxyhydropemxy cpd.
aldehyde
H2°2
hydrogen
peroxide
Source: (U.S. EPA, 2006,
Figure 5-2. Schematic overview of ozone interaction with PUFA in ELF and lung cells. It should
be noted that not all secondary reaction products are shown.
1 Cholesterol is the most abundant neutral lipid in human ELF. Reaction of cholesterol with O3
2 results in biologically active cholesterol products such as the oxysterols, (3-epoxide and 6-oxo-3,5-
3 diol (Murphy and Johnson, 2008, 197792: Pulfer and Murphy, 2004, 076673: Pulfer et al, 2005,
4 076663). Product yields will depend on ozonolysis conditions, however cholesterol ozonolysis
5 products were formed in similar abundance to phospholipid-derived ozonolysis products in rat ELF
6 (Pulfer and Murphy, 2004, 076673).
7 Antioxidant substances appear to be an important cellular defense against O3. The level and
8 type of antioxidant present in ELF varies between species, regions of the respiratory tract, and can be
9 altered by O3 exposure. Endogenous antioxidants such as urate (UA), ascorbate (AH2), and reduced
10 glutathione (GSH) display high intrinsic reactivities toward O3, but do not possess equal O3
11 reactivity. In individual and in limited composite mixtures, UA was the most reactive antioxidant
12 tested, followed by AH2 (Mudway and Kelly, 1998, 000273). In human BALF samples, the mean
13 consumption of AH2 was greater than UA (Mudway et al., 1996, 080730). GSH was consistently less
14 reactive than UA or AH2 (Kanofsky and Sima, 1995, 075973: Mudway and Kelly, 1998, 000273:
15 Mudway et al., 1996, 080730). In a red cell based system, AH2 augmented the in vitro uptake of O3
16 by sixfold as computed by the mass balance across the exposure chamber (Ballinger et al., 2005,
17 076649). In addition, O3-induced cell membrane oxidation required interactions with AH2 and GSH,
18 but not UA or the vitamin E analog Trolox, however UA could block reactions of O3 with AH2
19 (Ballinger et al., 2005, 076649). The consumption of antioxidants by O3 was linear with time and
20 positively correlated with initial substrate concentration and chamber O3 concentration (Mudway and
21 Kelly, 1998, 000273: Mudway et al., 1996, 080730). However, estimated in vitro O3 uptake was not
22 proportional to the production of O3-derived aldehydes from red cell membrane exposure (Ballinger
23 et al., 2005, 076649). Studies with rats exposed to O3 show consumption of ascorbate that correlates
March 2011
5-5
DRAFT - DO NOT CITE OR QUOTE
-------
1 with O3 exposure (Gunnison and Hatch, 1999, 087204: Gunnison et al, 1996, 080803: Vincent et al,
2 1996, 080778).
3 ELF also contains proteins present in blood plasma as well as proteins secreted by surface
4 epithelial cells. Ozone reactions with proteins have been studied by their in vitro reactions as well as
5 reactions of their constituent amino acids (the most reactive of which are cysteine, histidine,
6 methionine, tyrosine, and tryptophan). Ozone reaction with S-containing biomolecules has been
7 shown to follow the following order: thiosulfate > ascorbate > cysteinmethionine > glutathione
8 (Kanofsky and Sima, 1995, 075973). Rate constants for the reaction of amino acids with O3 vary
9 between investigations due to differing reaction conditions and assumptions; however aliphatic
10 amino acids consistently are very slow to react with O3 (e.g., alanine: 25-100 moles/L/sec) (Hoigne
11 and Bader, 1983, 625266: Ignatenko and Cherenkevich, 1985, 625265: Kanofsky and Sima, 1995,
12 075973: Pryor et al., 1984, 595130). Uppu et al. (1995, 076212) predicted that 12% of inhaled O3
13 that does not react with antioxidants will react with proteins in the ELF, whereas 88% will react with
14 PUFAs.
15 ELF exists as a complex mixture, thus it is important to look at O3 reactivity in substrate
16 mixtures. Individual antioxidant consumption rates decreased as the substrate mixture complexity
17 increased (e.g., antioxidant mixtures and albumin addition) (Mudway and Kelly, 1998, 000273).
18 However, O3 reactions with AH2 predominated over the reaction with lipids, when exposed to
19 substrate solution mixtures (Postlethwait et al., 1998, 086754). It was suggested that O3 may react
20 with other substrates once AH2 concentrations within the reaction plane fall sufficiently.
21 Additionally, once AH2 was consumed, the absorption efficiency diminished, allowing inhaled O3 to
22 be distributed to more distal airways (Postlethwait et al., 1998, 086754). Multiple studies have
23 concluded O3 is more reactive with AH2 and UA than with the weakly reacting GSH (or cysteine or
24 methionine) or with amino acid residues and protein thiols (Cross et al., 1992, 625299: Kanofsky
25 and Sima, 1995, 075973).
5.1.2.2. Recent Publications
26 Further experiments on the reaction kinetics between O3 and the antioxidants present in the
27 ELF have been conducted since the last review. To quantify these reactions, Kermani, et al. (2006,
28 195643) evaluated the interfacial exposure of aqueous solutions of UA, AH2, and GSH (50-200 (iM)
29 with O3 (1-5 ppm). Similar to the results of Mudway and Kelly (1998, 000273). this study found the
30 hierarchy in reactivity between O3 and these antioxidants to be UA>AH2»GSH. UA and AH2
31 shared a 1:1 stoichiometry with O3, whereas 2.5 moles of GSH were consumed per mole of O3.
32 Using these stoichiometries, reaction rate constants were derived (5.8x 104moles/L/sec,
33 5.5 xlO4 moles/L/sec, and 57.5/M075/sec for the reaction of O3 with UA, AH2, and GSH,
34 respectively). These values are similar to those derived from data presented in Mudway and Kelly
35 (1998, 000273). Other studies reported reactive rate constants that are two to three orders of
36 magnitude larger, however these studies used higher concentrations of O3 and antioxidants under less
March 2011 5-6 DRAFT - DO NOT CITE OR QUOTE
-------
1 physiologically relevant experimental conditions (Giamalva et al, 1985, 595129; Kanofsky and
2 Sima, 1995, 075973: Pryor et al., 1984, 595130V
3 A series of studies used new techniques to investigate the reaction products resulting from
4 initial air-liquid interface interactions of O3 with ELF components (e.g., antioxidants and proteins) in
5 ~1 millisecond (Enami et al., 2008, 195834: Enami et al., 2008, 195833: Enami et al., 2009, 197791:
6 Enami et al., 2009, 195835: Enami et al., 2009, 195621). Solutions of aqueous UA, AH2, GSH,
7 a-tocopherol (a-TOH), and protein cysteines (CyS) were sprayed as microdroplets in O3(g)/N2
8 mixtures at atmospheric pressure and analyzed by electrospray mass spectrometry. These recent
9 studies demonstrated different reactivity toward AH2, UA, and GSH by O3 in the gas phase
10 compared to the liquid phase thus supporting the relevance of reactions between gas phase O3 and
11 ELF.
12 As was seen in previous studies (Kanofsky and Sima, 1995, 075973: Kermani et al., 2006,
13 195643). the hierarchy of reactivity of these ELF components with O3(g) was determined to be AH2
14 ~ UA > CyS > GSH. There was some variance between the reaction rates and product formation of
15 UA, AH2, and GSH with O3(g) as investigated by Enami et al. versus O3(aq) as described previously.
16 UA was more reactive than AH2 toward O3(aq), but in reactions with O3(g), these antioxidants have
17 equivalent reactivity (Enami et al., 2008, 195834). As O3 is a kinetically slow one-electron acceptor
18 but very reactive O-atom donor, products of the interaction of O3 with UA, AH2, GSH, CyS, and a-
19 TOH result from addition of n O-atoms (n = 1-4). These products included epoxides (e.g., U-O"),
20 peroxides (e.g. U-O2"), and ozonides (e.g., U-O3"). For instance, GSH was oxidized to sulfonates
21 (GSO37GSO32~), not glutathione disulfide (GSSG) by O3(g) (Enami et al., 2009, 197791). However,
22 it is possible that other oxidative species are oxidizing GSH in vivo, since sulfonates are not detected
23 in O3 exposed ELF whereas GSSG is. This is also supported by the fact that O3 is much less reactive
24 with GSH than other antioxidants, such that < 3% of O3 will be scavenged by GSH when in
25 equimolar amounts with AH2 (Enami et al., 2009, 197791).
26 Ozonolysis product yields and formation were affected by pH. Acidified conditions (pH ~
27 3-4), such as those that may result from acidic particulate exposure or pathological conditions like
28 asthma (pH ~ 6), decreased the scavenging ability of UA and GSH for O3; such that at low pH, the
29 scavenging of O3 must be taken over by other antioxidants, such as AH2 (Enami et al., 2008, 195834:
30 Enami et al., 2009, 197791). Also, under acidic conditions (pH ~ 5), the ozonolysis products of AH2
31 shifted from the innocuous dehydroascorbic acid to the more persistent products, ascorbate ozonide
32 and threonic acid (Enami et al., 2008, 195833). It is possible that the acidification of the ELF by
33 acidic co-pollutant exposure will increase the toxicity of O3 by preventing some antioxidant
34 reactions and shifting the reaction products to more persistent compounds.
35 The ELF is a complex mixture of lipids, proteins, and antioxidants that serve as the first
36 barrier and target for inhaled O3. The thickness of the lining fluid and mucus layer is an important
37 determinant of the dose of O3 to the tissues. The antioxidant substances present in the ELF appear in
38 most cases to limit interaction of O3 with underlying tissues and to prevent penetration of O3 deeper
March 2011 5-7 DRAFT - DO NOT CITE OR QUOTE
-------
1 into the lung. However, new findings indicate that in some cases, the antioxidants might themselves
2 be participating in harmful reactions. The formation of toxic reaction products is likely related to the
3 concentration of antioxidants present and the quenching ability of the lining fluid. New findings also
4 emphasize the importance of gaseous O3 acting at the air-liquid interface in initiating a reaction
5 cascade, and point out important distinctions in the reaction rates and product formation between
6 gaseous and aqueous O3 reactivity toward airway antioxidants. In addition to the highly reactive O3,
7 secondary oxidation products formed in the aqueous phase might penetrate into the cells and cause
8 injury (Section 5.2).
5.1.3. Ozone Uptake
5.1.3.1. Summary of Findings from the 2006 Ozone AQCD
9 Past AQCDs (U.S. EPA, 1996, 017831: U.S. EPA, 2006, 088089) provide information on the
10 majority of literature relevant to understanding the state of the science in O3 dosimetry. One method
1 1 of addressing the question of O3 dosimetry is to measure the amount of O3 removed from the air
12 stream during breathing (termed "uptake"). The percentage of the O3 in the air stream that is
13 removed is termed "uptake efficiency." Uptake studies have utilized both bolus and continuous O3
14 exposure techniques as well as dosimetry modeling to investigate uptake efficiency and distribution
15 of O3 uptake between upper and lower respiratory tract regions. A number of studies that have
16 measured the fractional O3 uptake efficiency of the human respiratory tract (FRT), upper airways
17 (FUA), and lower respiratory tract (FLRT) are presented in Table 5-1 as a review.
Table 5-1. Human respiratory tract uptake efficiency data
Reference Mouth/Nosea InspinitayFlow VT(mL,
fB(bpm)b FRT FUA FLRT
CONTINUOUS EXPOSURE
Gerritvetal. (1988. 0408991 M 509 832
N 456 754
M/N 350 832
M/N 634 778
Gerrityetal. (1994, 041314)° M 1,360 1,650
M 1,360 1,239
Gerrityetal. (1995, 042785) Mouthpiece 330 825
Wiesteretal. (1996, 041280) M 539 631
N 514 642
Santiago et al. (2001,019841) N 50
N 250
Rigasetal. (2000, OJ0454) Face mask 480 1,100
18 0.40 0.91
18 0.36 0.91
12 0.41 0.93
24 0.38 0.89
25 0.81 0.37 0.43
35 0.78 0.41 0.36
12 0.91 0.27 0.95
16 0.76
16 0.73
0.80d
0.33
27.6 0.86
BOLUS EXPOSURE
Huetal. (1992, 042794) Mouthpiece 250
0.96 0.46
March 2011
5-8
DRAFT - DO NOT CITE OR QUOTE
-------
Ultmanetal. (1994. 041746)
Ultmanetal. (2004. 057197)
Nodelman and Ultman (1999, 015112)
Mouthpiece
Mouthpiece
M
M
Nasal Cannula
Nasal Cannula
Mouthpiece
Mouthpiece
250
250
490
517
150
1,000
150
1,000
500e
500
450e
574
500
500
500
500
15
15
32.7
27
18
120
18
120
0.30
0.47
0.87
0.91
0.90
0.45
0.80
0.25
0.95
0.90
aM = mouth exposure by natural breathing; N = nasal exposure by natural breathing; M/N = pooled data from mouth and nasal exposure; mouthpiece =
exposure by mouthpiece; FRT= total RT uptake; FUA= upper airway uptake; FLRT= lower RT uptake.
bfB is either measured or is computed from flows and VT.
"Total RT uptake reported by Gerrity et al. (1988, 0408991 and Gerrity et al. (1994, 0413141 did not include the contribution from UA uptake efficiency during
expiration. The data include an expiratory UA contribution, assuming it equals inspiratory UA uptake efficiency.
dFuAfrom Santiago et al. (2001, 0198411 represents nasal absorption (Fnose).
eVT is computed from flow and fB.
Target Sites for Ozone Dose
1 A primary uptake site of O3 delivery to the lungs is believed to be the centriacinar region
2 (CAR). The CAR refers to the zone at the junction of the conducting airways and the gas exchange
3 region. This area is also considered the proximal alveolar region (PAR) and is defined as the first
4 generation distal to the terminal bronchioles. Contained within the CAR, the respiratory bronchioles
5 were confirmed as the site receiving the greatest O3 dose (18O mass/lung weight) in resting O3
6 exposed rhesus monkeys, when not considering the nose (Plopper et al., 1998, 087203).
7 Furthermore, the greatest cellular injury occurred in the vicinity of the respiratory bronchioles and
8 was dependent on the delivered O3 dose to these tissues. However, 18O label was detected to a lesser
9 extent in other regions of the tracheobronchial airway tree, showing that O3 is delivered to these
10 compartments as well, resulting in a smaller dose. Models predict that the net O3 dose (O3 flux to air-
11 liquid interface) gradually decreases distally from the trachea toward the end of the tracheobronchial
12 region (TB) and then rapidly decreases in the pulmonary region (Miller et al., 1985, 040307).
13 However, the tissue dose (O3 flux to liquid-tissue interface) is low in the trachea, increases to a
14 maximum in the terminal bronchioles and the CAR, and then rapidly decreases distally into the
15 pulmonary region. These models are limited by the exclusion of the upper respiratory tract as well as
16 reactions occurring between ELF constituents and O3 after the 16th generation, representing the
17 CAR region (Miller et al., 1985, 040307V
Nasopharyngeal Removal and Dose of Ozone
18 In both animals and humans, about 50% of the absorbed O3 is removed in the head (nose,
19 mouth, and pharynx), about 7% in the larynx/trachea, and about 43% in the lungs (Hatch et al., 1989,
20 041799; Hu et al., 1992, 042794). The nasopharyngeal region provides a defense against O3 entering
21 the lungs by removing half of the inhaled O3 from the airstream. The limiting factors in nasal O3
22 uptake are simultaneous diffusion and chemical reaction of O3 in the nasal ELF layer (Santiago et al.,
23 2001, 019841). The lining layer in the nose is thicker than in the lungs, but, like for the lungs,
March 2011
5-9
DRAFT - DO NOT CITE OR QUOTE
-------
1 mathematical estimates predict that O3 penetrates less than the thickness of the ELF layer leaving the
2 reaction products as likely the agents damaging the nasal tissue. The percentage of O3 taken up is
3 inversely related to flow rate and weakly related to inlet O3 concentration. It was hypothesized that
4 the nonlinear reaction kinetics of the nose could result from the depleting substrates in the nasal ELF
5 becoming the limiting factor of the reaction (Santiago et al., 2001, 019841).
6 Uptake efficiencies have been calculated for various segments of the upper airways
7 (Table 5-1). Gerrity et al. (1995, 042785) reported unidirectional uptake efficiencies of O3 inhaled
8 from a mouthpiece of 17.6% from the mouth to vocal cords, 12.8% from the vocal cords to the upper
9 trachea (totaling 27.0%), 11.5% from the upper trachea to the main bifurcation carina (totaling
10 35.5%), and essentially zero between the carina and the bronchus intermedius (totaling 32.5%).
11 These values are lower than those calculated by Hu et al. (1992, 042794) that reported uptake
12 efficiencies of 21, 36, 44, and 46% between the mouth and the vocal cords, the upper trachea, the
13 main bifurcation carina, and the bronchus intermedius, respectively. The lower efficiencies seen in
14 Gerrity et al. (1995, 042785) may have resulted from the mouthpiece scrubbing O3 from the breath
15 during inhalation. Overall, the nasopharyngeal region removes half of the inhaled O3 by reactions in
16 the nasal ELF. The exact uptake efficiency will change due to variations in flow rate and inhaled
17 concentration.
Pulmonary Ozone Uptake and Dose
18 O3 uptake in rats is approximately 54% efficient (Hatch et al., 1989, 041799). while in humans
19 at rest it ranges from 80-95% efficient (Hu et al., 1992, 042794). Approximately 43% of inhaled O3
20 is absorbed in the lungs of both humans and animals. Uptake efficiency is affected by changes in a
21 number of variables, including tidal volume (VT), breathing frequency (fB), exposure time, minute
22 volume, and O3 concentration. Simulations from the Overton et al. (1996, 080733) single-path
23 anatomical respiratory tract model, where the upper and lower respiratory tracts were modeled but
24 uptake by the upper airways was not considered, predicted that fractional uptake and PAR O3 dose
25 increased with VT. Likewise, experimental studies found that O3 uptake is positively correlated with
26 changes in VT (Gerrity et al., 1988, 040899: Ultman et al., 2004, 057197). Also, O3 exposure leads to
27 a reflex mediated increase in fB and reduction in VT, hypothesized to be protective by decreasing the
28 dose delivered in the lung (Gerrity et al., 1994, 041314). While maintaining a constant minute
29 volume, a decrease in VT will result in an increase in fB. Nasal flow rate (Santiago et al., 2001,
30 019841) and fB are inversely related to O3 uptake, such that an increase in fB will decrease uptake
31 efficiency (Figure 5-3) (Gerrity et al., 1988, 040899: Ultman et al., 2004, 057197: Wiester et al.,
32 1996, 041280). Modeling also predicted a decrease in fractional uptake with increased fB, but an
33 increase in PAR dose with increased fB (Overton et al., 1996, 080733). Similarly, increased fB (80 -
34 160 bpm) and rapid shallow breathing in rats resulted in a decrease in midlevel tracheal 18O content
35 and an increase in 18O content in the mainstem bronchi (Alfaro et al., 2004, 053551). This
36 dependence may be a result of frequency-induced alterations in contact time that affects the first-
March 2011 5-10 DRAFT - DO NOT CITE OR QUOTE
-------
1 order absorption rate for O3 (Postlethwait et al, 1994, 044219). Also, an association with O3 uptake
2 efficiency was found with minute volume and exposure time.
i.u -
£ 0.9 -
.Si
"o
E
UJ
fl)
™ 0.8 -
O.
3
0.7 -
+ "+ + + o
+:?+o e (g
+ +
+ + +o" + °
o
0 *© °°
+
o*- o
0 0
o
o
O 28 Women
+ 32 Men 0
20 30 40
Breathing Frequency (bpm)
50
Source: Used with permission from Health Effects Institute, Ultman et al. (2004, 0571971
Figure 5-3. Total ozone uptake efficiency as a function of breathing frequency at a minute
ventilation of 30 L/min. Subjects breathed 0.25 ppm ozone oronasally through a
breathing mask.
The uptake efficiency was well correlated with breathing frequency (r = -0.723, p < 0.001) and
tidal volume (not illustrated; r = 0.490, p < 0.001).
3 Increasing flow leads to deeper penetration of O3 into the lung, such that a smaller fraction of
4 O3 is absorbed in the upper airways and shifts uptake to the central and lower airways and
5 respiratory airspaces (Hu et al., 1994, 041323; Nodelman and Ultman, 1999, 015112; Ultman et al.,
6 1994, 041746). Hu et al. (1994, 041323; Ultman et al., 1994, 041746) found that O3 absorption
7 increases relative to the respiratory tract penetration volume (Vp) of a bolus O3 dose (Figure 5-4).
8 Ozone uptake efficiency and Vp are not affected by bolus O3 concentration (Gerrity et al., 1988,
9 040899; Hu et al., 1992, 042794; Kabel et al., 1994, 095597). indicating that O3 uptake is a linear
10 absorption process, where the diffusion and chemical reaction rates of O3 are proportional to the O3
11 concentration. This relationship was not true for nasal cavity uptake, which was proportional to O3
12 concentration in the inlet air (Santiago et al., 2001, 019841). or in exercising individuals, where
13 uptake efficiency had a weak but significant negative dependence on O3 concentration (Rigas et al.,
14 2000, 010454). Overall O3 inhalation uptake is over 80% efficient, but the exact efficiency that
15 determines how much O3 is available at longitudinally distributed compartments in the lung is
16 sensitive to changes in VT, fe, and exposure time. Increased fB will shift the O3 uptake from the upper
17 airways to the central airways and respiratory airspaces.
March 2011
5-11
DRAFT - DO NOT CITE OR QUOTE
-------
I.U -
o
tj
2
"• 0.5-
0)
1
o.o -
v
JA VP50% V
'/
*.'
/•w*
t
D , ^ •
§ .»•*'
20 40 60 80 100 120 140 160 180 200
Penetration Volume (ml_)
Source: Adapted with permission from Health Effects Institute, Ultman et al. (2004, 0571971
Figure 5-4. Ozone uptake fraction as a function of volumetric penetration (Vp) in a representative
subject. [Each point represents the ozone uptake of a bolus inspired through a
mouthpiece by the subject. The volumes, VUA and VD, are the volume of the upper
airways and anatomical dead space, respectively, and VP50% is the Vp at which 50%
of the inspired bolus was absorbed. In 47 healthy subjects, Ultman et al. (2004,
057197) found that VP50% was well correlated with VD and better correlated with the
volume of the conducting airways, i.e., VD minus VUA.
Mode of Breathing
1 Ozone uptake and distribution is highly sensitive to the mode of breathing. Variability in
2 conducting airways volume had a weaker influence on O3 absorption during nasal breathing
3 compared to oral breathing. This could be a result of O3 scrubbing in the nasal passageways that are
4 bypassed by oral breathing. Studies by Ultman and colleagues using bolus inhalation demonstrate
5 that O3 uptake fraction is greater during nasal breathing than during oral breathing at each Vp (Kabel
6 et al., 1994, 095597: Nodelman and Ultman, 1999, 015112: Ultman et al., 1994, 041746). However,
7 oral breathing results in deeper penetration of O3 into the lung with a higher absorbed fraction in the
8 pharyngeal, lower airways, and respiratory airways (Nodelman and Ultman, 1999, 015112). Similar
9 results were obtained from O3 uptake studies in dogs (Yokoyama and Frank, 1972, 039756):
10 however earlier human studies suggest that oral or oronasal breathing results in a higher O3 uptake
11 efficiency than nasal breathing (Gerrity et al., 1988, 040899: Wiester et al., 1996, 041280). These
12 human studies measured total respiratory tract absorption after continuous O3 exposure using a
13 pharyngeal sampling tube, which may decrease sensitivity and lead to measurement errors.
Interindividual Variability
14 Similarly exposed individuals vary in the amount of actual dose received with intersubject
15 differences in fractional O3 uptake by the respiratory tract accounting for the majority of total
March 2011
5-12
DRAFT - DO NOT CITE OR QUOTE
-------
1 variation in O3 uptake efficiency (Bush et al., 1996, 080763: Rigas et al., 2000, 010454: Santiago et
2 al., 2001, 019841). Interindividual variability accounted for between 10-50% of the absolute
3 variability in O3 uptake measurements (Rigas et al., 2000, 010454: Santiago et al., 2001, 019841).
4 When concentration, time, and minute ventilation are held constant, fractional absorption ranges
5 from 0.80 to 0.91 (Rigas et al., 2000, 010454). It has been hypothesized that interindividual variation
6 in O3 uptake is the result of substituting a dose surrogate, such as exposure concentration or inhaled
7 dose, for the actual O3 dose delivered to the tissues.
8 Variability in local dose may be attributed to differences in the pulmonary physiology. Since
9 the conducting airways (CA) remove the majority of inhaled O3 before it reaches the gas exchange
10 region, the volume and surface area of the upper airways will influence O3 uptake. Models predict
11 that fractional O3 uptake and PAR dose (flux of O3 to the PAR surfaces divided by exposure
12 concentration) increase with decreasing TB volume and decreasing TB region expansion. On the
13 contrary, alveolar expansion had minimal effect on uptake efficiency as little O3 reaches the
14 peripheral lung (Bush et al., 2001, 016665: Overton et al., 1996, 080733). Ozone uptake is virtually
15 complete by the time O3 reaches the alveolar spaces of the lung (Postlethwait et al., 1994, 044219).
16 Experimental studies have found that differences in CA volumes may account for 75% of the
17 variation in absorption between subjects (Ultman et al., 2004, 057197). In support of this concept,
18 regression analysis showed that O3 absorption is positively correlated with anatomical dead space
19 (VD) and CA volume (i.e., VD minus VUA), but not total lung capacity (TLC), forced vital capacity
20 (FVC), or functional residual capacity (FRC) (Bush et al., 1996, 080763: Hu et al., 1994, 041323:
21 Postlethwait et al., 1994, 044219: Ultman et al., 2004, 057197). Variability in VD is correlated more
22 with the variability in the CA than the upper airways. Similarly, uptake was correlated with changes
23 in individual bronchial cross-sectional area, indicating that changes in cross-sectional area available
24 for gas diffusion are related to overall O3 retention (Reeser et al., 2005, 195718: Ultman et al., 2004,
25 057197). These studies provide support to the pulmonary physiology, especially the CA volume and
26 surface area, playing a key role in variability of O3 uptake between individuals.
27 When absorption data between genders is normalized to Vp/VD, then the variability attributed
28 to gender differences is no longer distinguishable (Bush et al., 1996, 080763). A physiologically
29 based pharmacokinetic (PBPK) model simulating O3 uptake indicates that regional extraction of O3
30 is relatively insensitive to age, but extraction per unit surface area is two- to eightfold higher in
31 infants compared to adults, due to the fact that children under age 5 have much a much smaller
32 airway surface area in the extrathoracic (nasal) and pulmonary regions (Sarangapani et al., 2003,
33 054581).
Correlation of Dose and Response
34 Two studies have investigated the correlation of O3 uptake with the pulmonary function
35 responses to O3 exposure (Gerrity et al., 1994, 041314: Reeser et al., 2005, 195718). These studies
36 found that the large subject-to-subject variability in %AFEVi response to O3 does not appear to have
March 2011 5-13 DRAFT - DO NOT CITE OR QUOTE
-------
1 a dosimetric explanation. Reeser et al. (2005, 195718) found no significant relationship between
2 %AFEVi and fractional absorption of O3 using the bolus method. Contrary to previous findings, the
3 percent change in dead space volume of the respiratory tract (%AVD) did not correlate with O3
4 uptake, possibly due to the contraction of dead space caused by airway closure. Gerrity et al. (1994,
5 041314) found that intersubject variability in FEVi and airway resistance was not related to
6 differences in the O3 dose delivered to the lower airways, whereas minute ventilation was predictive
7 of FEVi decrement. No study has yet demonstrated that subjects show a consistent pattern of O3
8 retention when re-exposed over weeks of time, as has been shown to be the case for the FEVi
9 response, or that within-subject variation in FEVi response is related to fluctuations in O3 uptake.
10 On the contrary, cellular injury and inflammation have been found to correlate with the site-
11 specific O3 dose. Contained within the CAR, the respiratory bronchioles were confirmed as the site
12 receiving the greatest O3 dose (18O mass/lung weight) and sustained the greatest cellular injury in O3
13 (0.4 and 1.0 ppm) exposed resting rhesus monkeys (Plopper et al., 1998, 087203). The respiratory
14 bronchioles, having the highest concentration of local O3 dose, were also the site of significant GSH
15 reduction.
Co-Pollutant and Sequential Ozone Exposure
16 Previous continuous O3 exposure (0.12 or 0.36 ppm) decreased bolus O3 uptake, possibly due
17 to depletion of compounds able to react with O3 (Asplund et al., 1996, 082505; Rigas et al., 1997,
18 083602). Conversely, O3 (0.36 ppm) bolus uptake was increased with prior NO2 (0.36 or 0.72 ppm)
19 or SO2 (0.36 ppm) exposure (Rigas et al., 1997, 083602). It was hypothesized that this increased
20 fractional absorption could be due to increased production of reactive substrates in the ELF due to
21 oxidant-induced airway inflammation.
Physical Activity
22 Exercise increases the overall exposure of the lung due in most part to the increased volume of
23 air passing through the lung. Exercise increases breathing frequency and flow rate. According to
24 present thinking, doubling minute ventilation is assumed to lead to a doubling of dose, however, the
25 linearity of the dose relative to ventilation relationship has not been carefully studied. A recent study
26 by Sawyer et al. (2007, 195142) showed that doubling minute ventilation led to only a 1.6-fold
27 higher dose of O3 in the lower airway. In addition to increasing the quantity of O3 in the lung,
28 exercise also has been shown to lead to a switch to oronasal breathing. By increasing flow to what is
29 common in moderate exercise, the upper airways absorbed a smaller fraction of the O3 (-0.50 at
30 quiet breathing to 0.10 at exercise); however, the trachea and more distal conducting airways
31 received higher doses than during quiet breathing (0.65 absorbed in the lower conducting airways,
32 and 0.25 absorbed in the respiratory zone) (Hu et al., 1994, 041323). The same shift in the O3 dose
33 distribution to deeper into the lung occurred in other studies mimicking the effects of exercise
34 (Nodelman and Ultman, 1999, 015112).
March 2011 5-14 DRAFT - DO NOT CITE OR QUOTE
-------
5.1.3.2. Recent Publications
1 Few new studies have investigated the uptake of O3 in the respiratory tract since the end of the
2 last O3 assessment (U.S. EPA, 2006, 088089). The studies that have been conducted agree with the
3 results presented above and do not change the dosimetry conclusions of the last document.
4 Past studies have shown that O3-induced epithelial damage to the lung occurs with a
5 reproducible pattern of severity between daughter branches of individual bifurcations that is
6 dependent on the O3 concentration-time profile of the inhaled gas. A 3-dimensional computational
7 fluid dynamics model was created to investigate the dose-response relationship leading to the
8 distribution of damage in a single airway bifurcation (Taylor et al, 2007, 195717). The model
9 consisted of one parent branch and two symmetrical daughter branches with a branching angle of 90°
10 and a sharp carinal ridge. Various flow scenarios were simulated using Reynolds numbers (Re)
11 ranging from 100 to 500. The Re that corresponds to a certain airway generation is dependent upon
12 both lung size and minute ventilation, such that the range in Re from 100-500 would encompass
13 generations 1-5, 3-7, and 6-10 for an adult during quiet breathing, light exertion, and heavy exercise,
14 respectively, whereas the same Re range corresponds to generations 0-4, 1-6, and 4-8 for a 4-year-
15 old child. Consistent with early physical models of Schroter and Sudlow (1969, 071359). the model
16 predicted that during inspiration, the velocity and O3 concentration distribution were axisymmetric
17 throughout the parent branch, but skewed towards the inner wall within the daughter branches.
18 During expiration, the model predicted that the velocity and O3 concentration distribution was
19 slightly skewed towards the outer walls of the daughter branches. Hot spots of wall flux existed at
20 the carina during inspiration and expiration with Re >100. Additional hot spots were found during
21 expiration on the parent branch wall downstream of the branching region.
22 Past studies investigating nasal uptake of O3 have shown that the nose partially protects the
23 rest of the respiratory tract from damage from inspired O3 (Gerrity et al., 1988, 040899; Santiago et
24 al., 2001, 019841). Sawyer et al. (2007, 195142) further investigated nasal uptake of O3 in healthy
25 adults during exercise. Fractional O3 uptake, acoustic rhinometry (AR), and nasal NO measurements
26 were taken on ten adults (8 W, 2 M) exposed to 0.2 ppm O3 before and after moderate exercise at
27 two flow rates (10 and 20 L/min). The percent nasal uptake of O3 was -50% greater at 10 L/min
28 compared to 20 L/min both pre- and postexercise. However, the inhaled O3 delivery rate to the lung
29 (i.e., flow rate X [O3 ppm] X nasal O3 penetration) was 1.6-fold greater at the higher flow than at the
30 lower flow (2.5 compared to 0.9 ppm-L/min). Exercise did not affect O3 uptake at either flow rate,
31 but did significantly increase nasal volume (Vn) and AR measurements of nasal cross-sectional area
32 (MCA, CSA2, and CSA3) (p< 0.05). Conversely, exercise decreased nasal resistance (Rn) (p < 0.01)
33 and NO production (p > 0.05). The change in Vn and CSA2:MCA ratio was correlated with the
34 percent change in nasal uptake, however the overall effect was small and sensitive to elimination of
35 outliers and gender segregation.
36 Smoking history, with its known increase in mucus production, was not found to significantly
37 affect the fractional uptake of a bolus dose of O3 in apparently healthy smokers with limited smoking
March 2011 5-15 DRAFT - DO NOT CITE OR QUOTE
-------
1 history (Bates et al, 2009, 195727). Despite similar internal O3 dose distribution, the smokers
2 exhibited greater pulmonary responses to O3 bolus exposures, measured as FEVi decrements and
3 increases in the normalized slope of the alveolar plateau (SN). This is contrary to previous studies
4 conducted in smokers with a greater smoking history that found decreased O3 induced decrements in
5 FEVj in smokers (Emmons and Foster, 1991, 042430: Frampton et al., 1997, 082692).
6 Recent studies have reiterated the importance of intersubject variation in O3 uptake. The
7 intersubject variability in nasal O3 uptake determined by Sawyer et al. (2007, 195142) ranged from
8 26.8 to 65.4% (pre- and postexercise). A second study investigating the use of the CO2 expirogram to
9 quantify pulmonary responses to O3 found that intersubject variability accounted for 50% of the
10 overall variance in the study (Taylor et al., 2006, 195731).
11 In summary, O3 uptake efficiency is sensitive to a number of factors. As discussed before, the
12 characteristics of the ELF layer is a key determinant in the dose of O3 that reaches the tissue layer.
13 Fractional absorption will decrease with increased flow and increase proportional to VT. Decreased
14 uptake efficiency due to increased fB and oronasal breathing, as occurs during exercise, will shift the
15 O3 dose distribution deeper and lead to a greater dose to the lower respiratory tract. Individual total
16 airway O3 uptake efficiency is also sensitive to large changes in O3 concentration, exposure time, and
17 VE. Major sources of variability in absorption of O3 include O3 concentration, exposure time,
18 breathing frequency, minute volume, and tidal volume, but the interindividual variation is the
19 greatest source of variability uptake efficiency. However, to this date, studies have failed to show
20 that the large differences in biological response between subjects (FEVi, BAL cell inflammatory
21 response, etc.) are explainable by the differences in O3 uptake. Recent studies have provided
22 evidence for hot spots of O3 flux around bifurcations in the airways.
5.1.4. Species Homology, Sensitivity, and Animal-to-Human Dose
Extrapolation
5.1.4.1. Summary of Findings from 2006 Ozone AQCD
23 The 2006 O3 AQCD (U.S. EPA, 2006, 088089) discussed the suitability of animal models for
24 comparison with human O3 exposure and concluded that the acute and chronic functional responses
25 of laboratory animals to O3 appear qualitatively homologous to human responses. Thus, animal
26 studies can provide important data in determining cause-effect relationships between exposure and
27 health outcome that would be impossible to collect in human studies. Still, care must be taken when
28 comparing quantitative dose-response relationships in animal models to humans due to obvious
29 interspecies differences.
30 Physiological and anatomical differences exist between experimental species. Primates are
31 oronasal breathers with a dichotomous branching lung structure, while rodents are obligate nasal
32 breathers with a monopodial branching lung structure. In addition, rodents have fewer terminal
33 bronchioles, the major site of O3 uptake, compared to humans or canines (McBride, 1992, 078532).
March 2011 5-16 DRAFT - DO NOT CITE OR QUOTE
-------
1 Past studies of the effect of body size on resting oxygen consumption also suggest that rodents inhale
2 more volume of air per lung mass than primates. These distinctions as well as differences in nasal
3 structure between primates and rodents could affect the site and amount of O3 uptake. Also, because
4 of their higher body surface to volume ratio, rodents can rapidly lower body temperature during
5 exposure leading to lowered O3 dose and toxicity (Iwasaki et al., 1998, 086165; Slade et al., 1997,
6 082708; Watkinson et al., 2003, 050547). In addition to lowering the O3 dose to the lungs, this
7 hypothermic response may cause: (1) lower metabolic rate, (2) altered enzyme kinetics, and (3)
8 altered membrane function. The thermoregulatory mechanisms also may affect disruption of heart
9 rate which may lead to: (1) decreased cardiac output (CO), (2) lowered blood pressure (BP), and (3)
10 decreased tissue perfusion (Watkinson et al., 2003, 050547). These responses have not been
11 observed in humans except at very high exposures, thus further complicating extrapolation of effects
12 from animals to humans.
13 Sensitivity to heath effects from O3 varies between and within species, as well as between
14 endpoints. Rodents appear to have a slightly higher tachypneic response to O3 and are less sensitive
15 to changes in pulmonary function test than humans (U.S. EPA, 1996, 017831). However, rats do
16 experience attenuation of pulmonary function and tachypneic ventilatory responses, similar to
17 humans (Wiester et al., 1996, 080829). Hatch et al. (1986, 040472) reported that guinea pigs were
18 the most responsive to O3-induced inflammatory cell and protein influx. Rabbits were the least
19 responsive and rats, hamsters, and mice were intermediate responders. Further analysis of this study
20 by Miller et al. (1988, 041545) found that the protein levels in guinea pigs increased more rapidly
21 with predicted pulmonary tissue dose than in rats and rabbits. Alveolar macrophages isolated from
22 guinea pigs and humans mounted similar qualitative and quantitative cytokine responses to in vitro
23 O3 (0.1-1.0 ppm for 60 minutes) exposure (Arsalane et al., 1995, 077430).
24 Humans and animal models are similar in the pattern of regional O3 dose, but absolute values
25 differ. Hatch et al. (1994, 038953) reported that exercising humans exposed to oxygen-18 labeled O3
26 (0.4 ppm) accumulated 4-5 times higher concentrations of O3 reaction product in BAL cells,
27 surfactant and protein fractions compared to resting rats similarly exposed (0.4 ppm). It was
28 necessary to expose resting rats to 2 ppm O3 to achieve the same BAL accumulation of 18O reaction
29 product that was observed in humans exposed to 0.4 ppm with intermittent heavy exercise. The
30 concentration of 18O reaction product in BAL paralleled the accumulation of BAL protein and
31 cellular effects of the O3 exposure observed such that these responses to 2.0 ppm O3 were similar to
32 those of the 0.4 ppm O3 in exercising humans.
33 As O3 absorption and activity relies on ELF antioxidant substances as described in Section 5.2,
34 variability in antioxidant concentrations and metabolism between species may affect dose and O3-
35 induced health outcomes. Guinea pigs and mice have a lower basal activity of GSH transferase and
36 GSH peroxidase, and lower vitamin E levels in the lung compared to rats (Ichinose et al., 1988,
37 041805; Sagai et al., 1987, 041704). Nasal lavage fluid analysis shows that humans have a higher
38 proportion of their nasal antioxidants as UA and low levels of AH2 whereas mice, rats, or guinea pigs
March 2011 5-17 DRAFT - DO NOT CITE OR QUOTE
-------
1 have high levels of AH2 and undetectable levels of UA (Figure 5-5a). GSH is not detected in the
2 nasal lavage of most of these species, but is present in monkey nasal lavage. Guinea pigs and rats
3 have a higher antioxidant to protein ratio in nasal lavage and BAL fluid than humans (Hatch, 1992,
4 043901). The BALF profile differs from the nasal lavage (Figure 5-5b). Humans have a higher
5 proportion of GSH and less AH2 making up their BALF content compared to the guinea pigs and rats
6 (Hatch, 1992, 043901: Slade et al, 1993, 042865V Similar to the nose, rats have the highest
7 antioxidant to protein mass ratio found in BALF (Slade et al., 1993, 042865). Antioxidant defenses
8 also vary with age (Servais et al., 2005, 195667) and exposure history (Duan et al., 1996, 080791).
9 Duan et al. (1993, 086326; 1996, 080791) reported that differences in antioxidant levels between
10 species and lung regions did not appear to be the primary factor in O3 induced tissue injury.
11 However, a close association between site-specific O3 dose, the degree of epithelial injury, and
12 reduced glutathione depletion was later revealed in monkeys (Plopper et al., 1998, 087203).
March 2011
5-18
DRAFT - DO NOT CITE OR QUOTE
-------
a.
Rat
Guinea pig
Human
200 400 600 800
Antioxidant/ Protein, nanomoles / gram
b.
Rat
Guinea pig
Human
1000
D Ascorbic acid
• Uric acid
DGIutathione
D Ascorbic acid
• Uric acid
D Glutathione
0 50 100 150 200 250
Antioxidant/ Protein, nanomoles / gram
Source: Adapted with permission from CRC Press, Inc., Hatch (1992, 0439011 and with permission from Slade et al. (1993, 0428651
Figure 5-5. Species comparison of antioxidant / protein ratios of: (a) nasal lavage fluid and,
(b) bronchoalveolar lavage fluid.
5.1.4.2. Recent Publications
1 There have been few new publications examining interspecies differences in dosimetry and
2 response to O3 since the last AQCD. These studies do not overtly change the conclusions discussed
3 above from the previous document (U.S. EPA, 2006, 088089).
March 2011
5-19
DRAFT - DO NOT CITE OR QUOTE
-------
1 A quantitative comparison of O3 transport in the airways of rats, dogs, and humans was
2 conducted using a three-compartment airways model, based on upper and lower airway casts and
3 mathematical calculation for alveolar parameters (Tsujino et al, 2005, 195842). The model was
4 designed as cylindrical tubes with constant volume and one-dimensional gas movement and no
5 airway branching patterns. It used data for solubility of O3 as well as measured nasopharyngeal
6 removal rates of O3 published previously. This model examined how interspecies anatomical and
7 physiological differences affect intra-airway O3 concentrations and the amount of gas absorbed (10%
8 O3 exposure). Peak, real-time, and mean O3 concentrations were higher in the upper and lower
9 airways of humans compared to rats and dogs, but lowest in the alveoli of humans. The amount of O3
10 absorbed was lowest in humans when normalized by body weight (8.47 x 10"8 g/kg compared to
11 1.1 x 10"7 in rats and 1.46 x 1Q"7 in dogs). The intra-airway concentration decreased distally in all
12 species. Sensitivity analysis demonstrated that VT, fe, and upper and lower airways surface area had
13 a significant impact on model results. The model is limited in that it did not account for chemical
14 reactions in the ELF or consider gas diffusion as a driving force for O3 transport. Also, the model
15 was run at a respiratory rate of 16/min simulating a resting individual, however exercise may cause a
16 further deviation from animal models as was seen in Hatch et al. (1994, 038953).
17 To further understand the genetic basis for age-dependent differential response to O3, adult
18 (15 weeks old) and neonatal (15-16 days old) mice from 8 genetically diverse strains were examined
19 for O3-induced (0.8 ppm for 5 hours) pulmonary injury and lung inflammation (Vancza et al., 2009,
20 596419). Ozone exposure increased polymorphonuclear leukocytes (PMN) influx in all strains of
21 mice tested, but significantly in only some sensitive strains, suggesting a genetic background effect.
22 This strain difference was not due to differences in delivered dose of O3 to the lung, evidenced by
23 18O lung enrichment. The sensitivity of strains for O3-induced increases in BAL protein and PMNs
24 was different for different strains of rats suggesting that genetic factors contribute to heightened
25 responses. Interestingly, adult mice accumulated more than twice the levels of 18O reaction product
26 of O3 than corresponding strain neonates. Thus, it appeared that the infant mice showed a two- to
27 threefold higher response than the adults when expressed relative to the accumulated O3 reaction
28 product in their lungs. The apparent decrease in delivered O3 dose in neonates could be a result of a
29 more rapid loss of body temperature in infant rats incident to maternal separation and chamber air
30 flow.
31 The three-dimensional detail of the nasal passages of immature Rhesus macaque monkeys
32 were analyzed for developing predictive dosimetry models and exposure-dose-response relationships
33 (Carey et al., 2007, 195752). In doing so the authors report that the relative amounts of the five
34 epithelial cell types in the nasal airways of monkeys remains consistent between infancy and
35 adulthood (comparing to (Gross et al., 1982, 040121; Gross et al., 1987, 625447)). Ozone exposures
36 (0.5 ppm, 8 h/day under acute [5 days] and episodic conditions [5 replicates of the acute paradigm
37 spaced a week apart]) confirmed that the ciliated respiratory and transitional epithelium were the
38 most sensitive cell types in the nasal cavity, showing 50-80% decreases in epithelial thickness and
March 2011 5-20 DRAFT - DO NOT CITE OR QUOTE
-------
1 epithelial cell volume. The character and location of nasal lesions resulting from O3 exposure are
2 similar between adult and infant monkeys similarly exposed. However, infant monkeys did not
3 undergo nasal airway epithelial remodeling or adaptation that occurs in adult animals and they may
4 develop persistent necrotizing rhinitis following episodic longer-term exposures.
5 In summary, for all species there are limitations that must be considered when attempting to
6 extrapolate to human O3 exposures. Rats required 4-5 times higher exposure to O3 to achieve
7 comparable increases in BAL protein and PMNs to exercising humans. New studies have shown that
8 varied O3 response in different mouse strains was not due to differences in delivered dose of O3 to
9 the lung but more likely genetic sensitivity, and that infant mice show greater toxicity relative to the
10 their smaller lung dose than adults. Even though interspecies differences limit quantitative
11 comparison between species, the acute and chronic functional responses of laboratory animals to O3
12 appear qualitatively homologous to those of the human making them a useful tool in determining
13 mechanistic and cause-effect relationships with O3 exposure.
5.2. Possible Pathways/Modes of Action
5.2.1. Introduction
14 As described in the previous section, O3 is a highly reactive oxidant gas with low water
15 solubility. Its diffusion into the fluid/tissue compartment of the respiratory tract occurs by reactive
16 absorption. This process depends on the availability of substrates such as antioxidants, lipids,
17 proteins, and carbohydrates and results in their oxidative modification. Because of its chemical
18 reactivity, inhaled O3 directly targets components residing on the airways and alveolar surfaces,
19 including ELF and surface macrophages. Although the O3 molecule is consumed and may not reach
20 the apical plasma membrane of airways and alveolar epithelium, secondary oxidation products
21 transmit signals to the epithelium, nociceptive sensory nerve fibers and, if present, dendritic cells,
22 mast cells and eosinophils. Thus, O3 effects are mediated by components of ELF and by the multiple
23 cell types found in the respiratory tract.
24 Three distinct short-term responses have been well-characterized in humans challenged with
25 O3: decreased pulmonary function, airways inflammation, and increased bronchial reactivity. In
26 addition, evidence has been accumulating that O3 exposure exacerbates, and possibly causes, asthma
27 and allergic airways disease in humans. Effects on the nasal airways and distal lung of humans,
28 including inflammation and injury, have also been described. Animal studies have demonstrated a
29 wide range of respiratory system effects. While the respiratory tract is the primary target tissue,
30 cardiovascular and other organ effects occur following short- and long-term exposures of animals to
31 O3. Mechanisms responsible for these effects are incompletely understood.
32 This section of the ISA highlights findings of studies published since the last O3 AQCD
33 (U.S. EPA, 2006, 088089) which provide insight into the biological pathways underlying the effects
March 2011 5-21 DRAFT-DO NOT CITE OR QUOTE
-------
1 of O3. Older studies which represent the current state of the science are also discussed. Studies
2 conducted at more environmentally-relevant concentrations of O3 are of greater interest, since
3 mechanisms responsible for effects at low O3 concentrations may not be identical to those occurring
4 at high O3 concentrations. In fact, some evidence suggests a concentration-dependent hierarchy of
5 effects. The following subsections describe the current understanding of potential pathways and
6 modes of action responsible for the pulmonary and extrapulmonary effects of O3 exposure.
5.2.2. Formation of Secondary Oxidation Products in the Respiratory Tract
7 Since O3 does not diffuse far into the aqueous layer of the ELF without reacting, it is not likely
8 to directly impact the underlying cells of the respiratory tract (Pryor, 1992, 042725). This does not
9 preclude direct reactions with the plasma membranes of cells extending beyond the ELF such as
10 surface macrophages. The secondary oxidation products formed in the ELF following O3 exposure
11 are primarily responsible for ozone's effects at the molecular, cellular and tissue level. The amount
12 and type of secondary oxidation product formed are important determinants of the anatomic sites of
13 reaction and injury due to O3 exposure, as will be discussed below.
14 Although not itself a free radical, ozone's effects are primarily mediated through free radical
15 reactions. Free radicals are generated during O3-mediated oxidation reactions (Pryor, 1994, 075987).
16 Subsequent reactions of these radical species produce cytotoxic nonradicals such as ozonides and
17 aldehydes (Cueto et al., 1992, 042770: Pryor, 1976, 038940). These effects are reduced by the
18 presence of the lipid-soluble free radical scavenger alpha-tocopherol (Fujita et al., 1987, 004280;
19 Pryor, 1976, 038940: Pryor, 1994, 075987). Although O3 can react with all hydrocarbons, its
20 reactivity towards specific groups varies greatly (Pryor, 1992, 042725). Polyunsaturated fatty acids
21 are one preferred target of O3. Following reaction of O3 with unsaturated fatty acids in the ELF,
22 measurable amounts of aldehydes were found in human bronchoalveolar lavage (BAL) fluid
23 (Frampton et al., 1999, 040757: Mudway and Kelly, 2000, 010452). Peroxidation of membrane
24 lipids is an important mechanism underlying O3-induced injury (Mudway and Kelly, 2000, 010452:
25 Pryor, 1976, 038940). This could occur by free-radical reactions initiated by O3 in the ELF or by
26 direct effects of O3 on membranes of cells, like surface macrophages, which extend beyond the ELF.
27 Markers of lipid peroxidation have been demonstrated in lung tissue and BAL fluid following O3
28 exposure and are enhanced in alpha-tocopherol deficient animals (Mudway and Kelly, 2000,
29 010452). Ozone-mediated lipid peroxidation leads to the rapid formation of eicosanoids, another
30 class of secondary oxidation products (discussed below). Ozonized cholesterol species have been
31 measured in BAL fluid (Pulfer et al., 2005, 076663) and in isolated surfactant (Pulfer and Murphy,
32 2004, 076673). indicating that O3 reacts with the cholesterol found in surfactant. In addition, O3
33 attacks ELF proteins through reactions with cysteine, methionine, tryptophan and tyrosine residues
34 (Mudway and Kelly, 2000, 010452). This results in protein oxidation and carbonylation (Mudway
35 and Kelly. 2000. 010452).
March 2011 5-22 DRAFT - DO NOT CITE OR QUOTE
-------
1 The ELF contains numerous antioxidants including alpha-tocopherol, albumin, ascorbate,
2 ceruloplasmin, glutathione, lactoferrin, mucins, urate and transferrin (Freed et al., 1999, 011829;
3 Mudway et al., 2006, 196536). Ascorbate, glutathione and urate are present in relatively high
4 concentrations in the surface liquid of human conducting airways and are known to be preferred
5 targets of O3. These antioxidants are thought to be the first line of defense against inhaled O3,
6 preventing free radical reactions with cellular proteins and lipids (Mudway and Kelly, 2000,
7 010452). In vitro studies have demonstrated consumption of water-soluble antioxidants and the
8 formation of oxidation products by O3 as well as a reactive hierarchy with O3 (Cross et al., 1992,
9 625299; Mudway and Kelly, 1998, 000273). When examined as a single antioxidant in solution,
10 urate exhibited the greatest reactivity, followed by ascorbate and glutathione (Mudway and Kelly,
11 1998, 000273). Results using mixtures have demonstrated greater complexity (Mudway and Kelly,
12 2000, 010452).
13 Although ELF constituents such as antioxidants may protect against the deleterious effects of
14 O3, there is some evidence that antioxidants may paradoxically facilitate O3-mediated damage. This
15 apparent contradiction should be viewed in terms of the concentration-dependent role of the ELF
16 antioxidants. Studies in vitro using red cell ghosts as a target showed that aqueous phase reactions
17 between O3 and the low molecular weight antioxidants ascorbate and glutathione generated
18 secondary oxidation products capable of perturbing membrane proteins and lipids (Ballinger et al.,
19 2005, 076649). Reactions between O3 and these antioxidant species exhibited a biphasic
20 concentration response, with oxidation of protein and lipid occurring at lower, but not higher,
21 concentrations of antioxidant. In this way, endogenous reactants led to the formation of secondary
22 oxidation products which were injurious and also led to quenching reactions which were protective.
23 Aqueous phase reactions between O3 and urate or bovine serum albumin did not result in membrane
24 oxidation (Ballinger et al., 2005, 076649). Further, the presence of urate or bovine serum albumin
25 protected against lipid and protein oxidation resulting from the reaction of O3 and ascorbate
26 (Ballinger et al., 2005, 076649). Thus, the formation of secondary oxidation products mediated by
27 some antioxidants was opposed by quenching reactions involving other antioxidants.
28 Local scavenging of inhaled O3 by antioxidants in specific respiratory regions has been
29 demonstrated in vivo (Gunnison and Hatch, 1999, 087204; Mudway et al., 1999, 001270). Urate, but
30 not ascorbate or glutathione, was depleted in nasal lavage fluid during exposure of human subjects to
31 0.2 ppm O3 for 2 hours indicating that urate is the predominant antioxidant with respect to O3
32 reactivity in the nasal cavity (Mudway et al., 1999, 001270). In addition, depletion of urate during O3
33 exposure was associated with a small but significant increase in plasma urate levels (Mudway et al.,
34 1999, 001270). Efforts to identify the predominant antioxidant(s) in other respiratory tract regions
35 and in other species have failed to yield definitive results. In one study, glutathione was increased,
36 rather than decreased, in BAL fluid and bronchial wash fluid 1.5 h following a 2 h exposure of
37 human subjects to 0.2 ppm O3 (Blomberg et al., 1999, 001267).
March 2011 5-23 DRAFT - DO NOT CITE OR QUOTE
-------
1 Since exposure to O3 often leads to airway inflammation characterized by neutrophilia and
2 since neutrophil-derived oxidants often scavenge ELF antioxidants, concentrations of ELF
3 antioxidants were examined during airways neutrophilia which generally occurs 4-6 hours
4 postexposure to O3 (Gunnison and Hatch, 1999, 087204: Long et al., 2001, 057301: Mudway et al,
5 1999, 011833). In humans exposed to 0.2 ppm O3 for 2 hours, urate, glutathione and alpha-
6 tocopherol levels remained unchanged in BAL fluid 6 hours postexposure while ascorbate was
7 decreased significantly in both BAL fluid and plasma (Mudway et al., 1999, 011833). A second
8 study involving the same protocol reported a loss of ascorbate from bronchial wash fluid and BAL
9 fluid, representing proximal and distal airway ELF respectively, as well as an increase in oxidized
10 glutathione in both compartments (Mudway et al., 2001, 025327). No change was observed in ELF
11 urate levels in response to O3 (Mudway et al., 2001, 025327). Further, O3 exposure (0.8 ppm,
12 4 hours) in female rats resulted in a 50% decrease in BAL fluid ascorbate immediately postexposure
13 (Gunnison and Hatch, 1999, 087204). These studies suggested a role for ascorbate and glutathione in
14 protecting against oxidative stress associated with inflammation. On the other hand, a study in
15 hamsters exposed to 3 ppm O3 for 6 hours found no depletion of ascorbate, glutathione or alpha-
16 tocopherol in BAL fluid (Long et al., 2001, 057301). Instead an increase in BAL fluid urate and a
17 decrease in plasma ascorbate were observed (Long et al., 2001, 057301).
18 Although it is known that ELF antioxidants are variably distributed among regions of the
19 respiratory tract, mechanisms underlying this variability are not well-understood. It is thought that
20 both plasma ultrafiltrate and locally secreted substances contribute to the antioxidant content of the
21 ELF (Freed et al., 1999, 011829: Mudway et al., 2006, 196536). In the case of urate, the major
22 source appears to be the plasma (Peden et al., 1995, 076189). Repletion of urate in nasal lavage fluid
23 was demonstrated during sequential nasal lavage in human subjects (Mudway et al., 1999, 001270).
24 When these subjects were exposed to O3, nasal lavage urate was significantly decreased while
25 plasma urate levels was significantly increased (Mudway et al., 1999, 001270). In addition,
26 concentrations of urate were increased by cholinergic stimulation of the airways which suggests that
27 increased mucosal gland secretions can be an important source (Peden et al., 1995, 076189).
28 Regulation of ascorbate, glutathione and alpha-tocopherol concentrations within the ELF is less clear
29 than that of urate (Mudway et al., 2006, 196536). In a sequential nasal lavage study in humans,
30 wash-out of ascorbate and glutathione occurred, indicating the absence of rapidly acting repletion
31 mechanisms (Mudway et al., 1999, 001270). Other studies discussed above demonstrated increases
32 in BALF glutathione and decreases in BALF and plasma ascorbate levels several hours following O3
33 exposure (Blomberg et al., 1999, 001267: Mudway et al., 1999, 011833: Mudway et al., 2001,
34 025327). Furthermore, high levels of dehydroascorbate, the oxidized form of ascorbate, have been
35 reported in human ELF (Mudway et al., 2006, 196536). Other investigators have demonstrated
36 cellular uptake of oxidized ascorbate by several cell types leading to intracellular reduction and
37 export of reduced ascorbate (Welch et al., 1995, 644675).
March 2011 5-24 DRAFT - DO NOT CITE OR QUOTE
-------
1 A further consideration is the compromised status of ELF antioxidants in disease states such as
2 asthma (Mudway and Kelly, 2000, 010452). This could possibly be due to ongoing inflammation
3 which causes antioxidant depletion or to abnormal antioxidant transport or synthesis (Mudway and
4 Kelly, 2000, 010452). For example, basal ascorbate levels were significantly lower and basal levels
5 of oxidized glutathione and urate were significantly higher in bronchial wash fluid and BAL fluid of
6 mild asthmatics compared with healthy control subjects (Mudway et al, 2001, 025327). Differences
7 in ELF antioxidant content have also been noted between species. These observations have led to the
8 suggestion that the amount and composition of ELF antioxidants, the capacity to replenish
9 antioxidants in the ELF or the balance between beneficial and injurious interactions between
10 antioxidants and O3 may contribute to O3 sensitivity which varies between individuals and species
11 (Mudway and Kelly, 2000, 010452: Mudway et al., 1999, 001270: Mudway et al., 2006, 196536).
12 The complexity of these interactions was demonstrated by a study in which O3 exposure resulted in
13 similar increases in airway neutrophils and decreases in pulmonary function in both mild asthmatics
14 and healthy controls, despite differences in ELF antioxidant concentrations prior to O3 exposure
15 (Mudway et al., 2001, 025327). Further, the O3-induced increase in oxidized glutathione and
16 decrease in ascorbate observed in ELF of healthy controls was not observed in mild asthmatics
17 (Mudway et al., 2001, 025327). While the authors concluded that basal ascorbate and oxidized
18 glutathione concentrations were not predictive of responsiveness to O3, they also suggested that the
19 increased basal urate concentrations in the mild asthmatics may have played a protective role
20 (Mudway et al., 2001, 025327). Thus compensatory mechanisms resulting in enhanced total
21 antioxidant capacity may play a role in modulating responses to O3.
22 Several studies in animals evaluated the relationships between 18O-labeled O3 dose markers,
23 injury markers and ascorbate concentrations following O3 exposure. In female rats exposed to
24 0.8 ppm O3 for 4 hours, BAL indicators of injury and inflammation (protein and neutrophil number)
25 and 18O reaction product were increased inversely with the reduction in ascorbate (Gunnison and
26 Hatch, 1999, 087204). In another study, aging rats (9 and 24 months old) were shown to have 49%
27 and 64% lower ascorbate in lung tissue, respectively, than 2-month-old rats (Vincent et al., 1996,
28 080778). However, aging-induced ascorbate loss did not increase the accumulation of 18O reaction
29 products following O3 exposure (0.4-0.8 ppm, 2-6 hours). Pregnancy and lactation also caused lower
30 ascorbate content in BAL and nasal lavage fluid and was associated with an increase in accumulation
31 of 18O reaction products following O3 exposure (Gunnison and Hatch, 1999, 087204). Kari et al.
32 (1997, 086171) observed that a 3-week caloric restriction (75%) in rats abrogated the toxicity of O3
33 (2 ppm, 2 hours), measured as BAL fluid increases in protein, fibronectin and neutrophils, which
34 was seen in normally fed rats. Accompanying this resistance to O3 toxicity, was a 30% higher basal
35 BAL fluid ascorbate concentration, a rapid accumulation of ascorbate into the lungs to levels 60%
36 above normal and reduction (30%) in the accumulation of 18O reaction product in the lungs. These
37 investigations demonstrated an inverse relationship between ascorbate levels and O3 dose and
38 provide evidence for ascorbate playing a protective role following O3 exposure in these studies.
March 2011 5-25 DRAFT - DO NOT CITE OR QUOTE
-------
1 Many investigations have focused on antioxidant deficiency and supplementation as
2 modulators of O3-mediated effects. Ascorbate deficiency has been shown to increase the effects of
3 acute, but not chronic, O3 exposure in guinea pigs and humans (Kodavanti et al., 1995, 077440;
4 Slade et al., 1989, 059465). Supplementation with ascorbate and alpha-tocopherol was protective in
5 healthy adults who were on an ascorbate-deficient diet and exposed to 0.4 ppm O3 for 2 hours while
6 exercising (Samet et al., 2001, 019034). In this study, the protective effect consisted of a smaller
7 reduction in forced expiratory volume in one second (FEVi) following O3 exposure (Samet et al.,
8 2001, 019034). However the inflammatory response (influx of neutrophils and levels of IL-6)
9 measured in BAL fluid 1 hour after O3 exposure was not different between supplemented and non-
10 supplemented subjects (Samet et al., 2001, 019034). Supplementation with ascorbate and alpha-
11 tocopherol also protected against pulmonary function decrements and nasal inflammatory responses
12 which were associated with high levels of ambient O3 in asthmatic children living in Mexico City
13 (Romieu et al., 2002, 034711; Sienra-Monge et al., 2004, 196422). Similarly, supplementation with
14 ascorbate, alpha-tocopherol and beta-carotene improved pulmonary function in Mexico City
15 streetworkers (Romieu et al., 1998, 086756). However, ascorbate and alpha-tocopherol
16 supplementation failed to ameliorate the pulmonary function decrements or airways neutrophilia
17 observed in humans exposed to 0.2 ppm O3 for 2 hours (Mudway et al., 2006, 196536). It was
18 suggested that supplementation may be ineffective in the absence of antioxidant deficiency (Mudway
19 et al., 2006, 196536). Furthermore, protective effects of supplementation with alpha-tocopherol
20 alone have not been observed in humans (Mudway and Kelly, 2000, 010452).
21 Recent studies in animals demonstrated protection against O3-induced effects using gamma-
22 tocopherol supplementation in models of allergic rhinosinusitis (Wagner et al., 2009, 201574) and
23 lower airway allergic inflammation (Wagner et al., 2007, 596420). Previous studies demonstrated
24 that supplementation with alpha-tocopherol was ineffective in these models (Wagner et al., 2007,
25 596420). Other investigators found that alpha-tocopherol deficiency led to an increase in liver lipid
26 peroxidation (Sato et al., 1980, 039738) and a drop in liver alpha-tocopherol levels following O3
27 exposure (Vasu et al., 2010, 201561). A recent study used alpha-tocopherol transfer protein null mice
28 as a model of alpha-tocopherol deficiency and demonstrated an altered adaptive response of the lung
29 genome to O3 exposure (Vasu et al., 2010, 201561). Taken together, these studies provide evidence
30 that the tocopherol system modulates O3-induced responses.
31 Other antioxidants have been shown to confer resistance to O3-induced injury. In a recent
32 study, lung hyperpermeability in response to O3 was unexpectedly reduced in mice deficient in the
33 glutamate-cysteine ligase modifier subunit gene compared with sufficient mice (Johansson et al.,
34 2010, 644476). Since the lungs of these mice exhibited 70% glutathione depletion, protection against
35 O3-induced injury was unexpected (Johansson et al., 2010, 644476). However it was found that
36 several other antioxidant defenses, including metallothionein, were upregulated in response to O3 to
37 a greater degree in the glutathione-deficient mice compared with sufficient mice (Johansson et al.,
38 2010, 644476). The authors suggested that resistance to O3-induced lung injury was due to
March 2011 5-26 DRAFT - DO NOT CITE OR QUOTE
-------
1 compensatory augmentation of antioxidant defenses (Johansson et al, 2010, 644476). Antioxidant
2 effects have also been attributed to Clara cell secretory protein (CCSP) and surfactant protein A
3 (SP-A). CCSP was found to modulate the susceptibility of airways epithelium to injury in mice
4 exposed to O3 (0.2 or 1 ppm for 8 hours) by an unknown mechanism (Plopper et al., 2006, 596410).
5 SP-A protected against O3-induced airways inflammation and injury, possibly by acting as a
6 sacrificial substrate (Haque et al., 2007, 597606).
7 A role for plasma antioxidants in modulating O3-induced respiratory effects has also been
8 suggested (Aibo et al., 2010, 378559). In this study, pretreatment of rats with a high dose of
9 acetaminophen resulted in increased levels of plasma cytokines and the influx of inflammatory cells
10 into the lung following 6 h exposure to 0.25 and 0.5 ppm O3 (Aibo et al., 2010, 378559). These
11 effects were not observed in response to O3 alone. Although not measured in this study, glutathione
12 depletion in the liver is known to occur in acetaminophen toxicity. Since liver glutathione is the
13 source of plasma glutathione, acetaminophen treatment may have lowered plasma glutathione levels
14 and altered the redox balance in the vascular compartment. These findings indicate an
15 interdependence between respiratory tract, plasma and liver responses to O3, possibly related to
16 glutathione status.
17 Another important consideration is the non-uniformity of the injury response to O3 throughout
18 the respiratory tract. Several mechanisms have been proposed to explain this phenomenon. First,
19 dosimetry may be a key determinant since the sites receiving the largest dose might be expected to
20 exhibit the greatest injury or inflammation (Plopper et al., 1998, 087203; Postlethwait et al., 2000,
21 003000). An important corollary is that the uneven distribution of mucus in the respiratory tract
22 airways may influence the capacity of O3 to reach the aqueous layer (Mudway and Kelly, 2000,
23 010452). Secondly, the non-homogeneous formation of cytotoxic products in the ELF may account
24 for the variable response (Postlethwait et al., 2000, 003000). The thickness of the ELF varies along
25 the respiratory tract, being greater in the upper airways and less more distally. Further the
26 composition of the ELF varies along the respiratory tract. Thus, the amount of protective
27 antioxidants and other scavengers in various respiratory tract regions is likely to limit the formation
28 of cytotoxic products. Similarly, the availability of reactants which are precursors of potent cytotoxic
29 products may facilitate their formation in a particular region. For example, the formation of highly
30 electrophilic aldehydes from unsaturated fatty acids may occur in all respiratory tract regions while
31 the formation of oxidized surfactant lipids is likely restricted to the alveolar region and respiratory
32 bronchioles where surfactant is found. Thus, region-specific formation of particular oxidation
33 products may dictate patterns of epithelial injury in the respiratory tract.
34 The relationship between site-specific O3 dose, epithelial injury and glutathione concentration
35 was investigated by Plopper (1998, 087203). Adult rhesus monkeys were exposed for 2 hours to 0.4
36 and 1.0 ppm O3, which was labeled with 18O, and tissues were analyzed immediately postexposure
37 for dose, epithelial injury and glutathione levels. Results indicated that exposure to 1 ppm O3
38 resulted in the greatest epithelial injury in the respiratory bronchioles although injury was observed
March 2011 5-27 DRAFT - DO NOT CITE OR QUOTE
-------
1 at all of the airway sites but not in the lung parenchyma. Exposure to 0.4 ppm O3 resulted in
2 epithelial injury only in the respiratory bronchioles. Local O3 dose was found to be highly variable
3 among the different sites with the greatest levels found in the respiratory bronchioles and lowest
4 levels found in the parenchyma following 1 ppm O3. Glutathione levels varied in the different
5 airways sites in monkeys exposed to filtered air. Exposure to 1 ppm O3 decreased glutathione levels
6 only in the respiratory bronchioles. This study demonstrated a close relationship between the
7 exposure dose of O3 (uptake of 18O) and the degree of initial epithelial injury at a particular site in
8 the respiratory tract. Glutathione depletion observed only at that site suggests that glutathione played
9 a protective role during O3 exposure.
10 While the formation of secondary oxidation products is the key event leading to O3-mediated
11 effects, scavenging and/or metabolism of those products is likely to be an important determinant of
12 outcomes. One such mechanism may be scavenging of oxidized lipids via the macrophage receptor
13 with collagenous structure (MARCO) expressed on the cell surface of alveolar macrophages. A
14 recent study demonstrated increased gene expression of MARCO in the lungs of an O3-resistant C3H
15 mouse strain (HeJ) but not in an O3-sensitive, genetically nearly identical strain (OuJ) (Dahl et al.,
16 2007, 196986). Upregulation of MARCO occurred in mice exposed to 0.3 ppm O3 for 24-48 hours;
17 inhalation exposure for 6 hours at this concentration was insufficient for this response. Animals
18 lacking the MARCO receptor exhibited greater inflammation and injury, as measured by BAL
19 neutrophils, protein and isoprostanes, following exposure to 0.3 ppm O3 (Dahl et al., 2007, 196986).
20 MARCO also protected against the inflammatory effects of oxidized surfactant lipids (Dahl et al.,
21 2007, 196986). Scavenging of oxidized lipids may limit O3-induced injury since ozonized
22 cholesterol species formed in the ELF (Pulfer and Murphy, 2004, 076673; Pulfer et al., 2005,
23 076663) stimulate apoptosis and cytotoxicity (Gao et al., 2009, 200764; Sathishkumar et al., 2007,
24 097758; Sathishkumar et al., 2007, 197785; Sathishkumar et al., 2009, 201549) in vitro. While these
25 studies have focused on the alveolar compartment (alveolar macrophages, surfactant lipids),
26 comparable pathways have yet to be elucidated in the conducting airways. A second mechanism
27 likely to impact O3-mediated effects is the metabolism of secondary oxidation products catalyzed by
28 antioxidant enzymes such as glutathione peroxidase, glutathione S-transferases (GST) and
29 NADPH:quinone oxidoreductase 1 (NQO1). Evidence for the importance of GST and NQO1 in
30 modulating the effects of O3 is presented below (see Gene-Environment Interactions).
31 Secondary oxidation products formed as a result of O3 exposure initiate numerous responses at
32 the cellular, tissue and whole organ level of the respiratory system. These responses include the
33 activation of neural reflexes, injury and inflammation, hyperpermeability, increased bronchial
34 reactivity and altered host defenses, as will be discussed below. In addition, the enhancement of
35 asthma and allergic responses demonstrated as a result of O3 exposure is likely due to secondary
36 oxidation products. Exposure to O3 also results in effects on other organ systems such as the
37 cardiovascular, hepatic and central nervous systems. Mechanisms underlying these extrapulmonary
38 responses are not well understood. It is unlikely that lipid ozonides and other secondary oxidation
March 2011 5-28 DRAFT - DO NOT CITE OR QUOTE
-------
1 products, which are bioactive and cytotoxic in the respiratory system, gain access to the vascular
2 space (Chuang et al., 2009, 197202). However O3 exposure may result in systemic oxidative stress,
3 as suggested by studies in humans reporting an association between O3 exposure and both levels of
4 plasma 8-isoprostanes and the presence of peripheral blood lymphocyte micronuclei (Chen et al.,
5 2006, 196504: Chen et al., 2007, 145956).
5.2.2.1. Summary
6 The initial key event in ozone's toxicity pathway is the formation of secondary oxidation
7 products in the respiratory tract. Pathways for the removal of those products are also of great
8 importance. Due to the highly reactive nature of O3, direct reactions most likely involve components
9 of the ELF and/or plasma membranes of surface macrophages which extend beyond the ELF.
10 Reaction products likely mediate ozone's effects on respiratory tract epithelium.
5.2.3. Activation of Neural Reflexes
11 Acute O3 exposure results in reversible effects on lung function parameters through activation
12 of neural reflexes. The involvement of bronchial C-fibers, a type of nociceptive sensory nerve, has
13 been demonstrated in dogs (Coleridge et al., 1993, 038695: Schelegle et al., 1993, 039203) and the
14 involvement of nociceptive sensory nerves has been demonstrated in humans (Passannante et al.,
15 1998, 030114). Furthermore there is evidence that substance P (SP), a tachykinin which is known to
16 be released from C-fibers, plays a role in O3-mediated effects (Hazbun et al., 1993, 043914: Krishna
17 etal. 1997.084262).
18 The response to O3 in humans is characterized by substernal discomfort, especially on deep
19 inspiration, accompanied by involuntary truncation of inspiration (Hazucha et al., 1989, 041909).
20 This leads to decreased inspiratory capacity and to decreased forced vital capacity (FVC) and forced
21 expiratory volume in one second (FEVi), as measured by spirometry, and is accompanied by a
22 decreased tidal volume and increased respiratory frequency in human subjects during exercise
23 (Hazucha et al., 1989, 041909). For example, these pulmonary function responses have been noted
24 immediately after a 1-h exposure to 0.3 ppm O3, resolving by 6 hours after exposure (Schelegle et
25 al., 1991, 042491). and during and immediately after a 4-h exposure to 0.2 ppm O3 in exercising
26 humans (Aris et al., 1993, 038275: Balmes et al., 1996, 080830). Although spirometric changes
27 began to return to baseline shortly after exposure, small residual spirometric decrements were
28 reported at 24 hours postexposure (Hazucha et al., 1996, 043923). Spirometric changes in FEVi and
29 FVC were not due to changes in respiratory muscle strength (Hazucha et al., 1989, 041909).
30 Changes in FVC or symptoms were not modified by treatment with bronchodilators such as atropine
31 (Beckett et al., 1985, 039758). Thus, parasympathetic involvement in the O3-mediated decreases in
32 lung volume was not significant (Mudway and Kelly, 2000, 010452). However, the loss of vital
33 capacity was reversible with intravenous administration of the rapid-acting opioid agonist,
34 sufentanyl, indicating an involvement of opioid receptor-containing nerve fibers and/or more central
March 2011 5-29 DRAFT - DO NOT CITE OR QUOTE
-------
1 neurons (Passannante et al, 1998, 030114). The effects of sufentanyl may be attributed to blocking
2 C-fiber stimulation by O3 since activation of opioid receptors downregulates C-fiber function
3 (Belvisi et al., 1992, 644681). There is some evidence that eicosanoids play a role in the neural
4 reflex since cyclooxygenase inhibition with indomethacin (Alexis et al., 2000, 013072; Schelegle et
5 al., 1987, 041706) or ibuprofen, which also blocks some lipoxygenase activity (Hazucha et al., 1996,
6 043923). before exposure to O3 significantly blunted the spirometric responses. In the latter study,
7 ibuprofen treatment resulted in measurable decreases in BAL levels of PGE2 and TXB2 at 1 hour
8 postexposure (Hazucha et al., 1996, 043923). Although an earlier study demonstrated that PGE2
9 stimulated bronchial C-fibers (Coleridge et al., 1976, 038612: Coleridge et al., 1993, 038695) and
10 suggested that PGE2 mediated O3-induced decreases in pulmonary function, no correlation was
11 observed between the degree of ibuprofen-induced inhibition of BAL PGE2 levels and blunting of
12 the spirometric response to O3 (Hazucha et al., 1996, 043923). Nonetheless, recent studies continue
13 to provide evidence that arachidonic acid metabolites, as well as oxidative stress, contribute to
14 human responsiveness to O3 (Alfaro et al., 2007, 196567).
15 A delay in onset of O3-induced pulmonary function responses has been noted in numerous
16 studies. Recently the delay was characterized in terms of changes in breathing frequency (Schelegle
17 et al., 2007, 195841). In humans exposed to O3, no change in breathing frequency was observed until
18 a certain cumulative inhaled dose of O3 had been reached. Subsequently, the magnitude of the
19 change in breathing frequency was correlated with the inhaled dose rate (Schelegle et al., 2007,
20 195841). These investigators proposed that initial reactions of O3 with ELF resulted in a time-
21 dependent depletion of ELF antioxidants, and that activation of neural reflexes occurred only after
22 the antioxidant defenses were overwhelmed (Schelegle et al., 2007, 195841).
23 There is a large range of pulmonary function responses to O3 among healthy young adults
24 (Balmes et al., 1996, 080830: Hazucha et al., 2003, 048168). Since individual responses are
25 relatively consistent across time, it is thought that responsiveness reflects an intrinsic characteristic
26 of the subject (Mudway and Kelly, 2000, 010452). Older adults are generally not responsive to O3
27 (Hazucha et al., 2003, 048168). while obese young women may be more responsive than lean young
28 women (Bennett et al., 2007, 418827). The lack of spirometric responsiveness is not attributable to
29 the presence of endogenous endorphins, which could potentially block C-fiber stimulation by O3, as
30 demonstrated in a study involving intravenous administration of naloxone immediately following the
31 O3 exposure to weak responders (Passannante et al., 1998, 030114). Currently, the mechanisms
32 underlying the inter-individual variation in responsiveness to O3 are not known. It has been proposed
33 that some of the variation in response may be genetically determined (Yang et al., 2005, 077211).
34 More discussion on this topic is found below (Section 5.2.9.1).
35 In addition to the spirometric changes, mild airways obstruction occurs as a result of O3
36 exposure. This pulmonary function decrement is generally measured as specific airway resistance
37 (sRaw) which is the product of airway resistance and thoracic gas volume. Increased sRaw occurred
38 fairly rapidly in exercising humans, peaking at 1-6 hours and resolving by 24 hours following O3
March 2011 5-30 DRAFT - DO NOT CITE OR QUOTE
-------
1 exposure (Aris et al, 1993, 038275: Balmes et al, 1996, 080830: Hazucha et al., 1996, 043923:
2 Schelegle et al., 1991, 042491). Small but statistically significant increases in sRaw during O3
3 exposure (0.2 ppm for 4 hours with intermittent exercise) and immediately following O3 exposure
4 (0.3 ppm for 1 hour, moderate exercise or 0.4 ppm for 2 hours, intermittent exercise) were observed
5 in several studies (Aris et al., 1993, 038275: Balmes et al., 1996, 080830: Hazucha et al., 1996,
6 043923: Schelegle et al., 1991, 042491). These changes in sRaw correlated with changes in
7 inflammatory and injury endpoints measured 18 hours postexposure, as will be discussed below, but
8 not with the time course or degree of spirometric change measured during exposure (Aris et al.,
9 1993, 038275: Balmes et al., 1996, 080830: Schelegle et al., 1991, 042491). In addition, a small but
10 persistent increase in airways resistance associated with narrowing of small peripheral airways
11 (measured as changes in isoV FEF25-75) was demonstrated in O3-exposed humans (0.35 ppm for
12 130 minutes with intermittent exercise) (Weinmann et al., 1995, 077206: Weinmann et al., 1995,
13 038645). A similar study (0.4 ppm O3 for 2 hours with intermittent exercise) found decreases in
14 FEF25-75 concomitant with increases in residual volume, which is suggestive of small airways
15 dysfunction (Kreit et al., 1989, 041817). In separate studies, a statistically significant increase in
16 residual volume (Hazucha et al., 1989, 041909) and a statistically significant decrease in FEF25_75
17 (Horstman et al., 1995, 075834) were observed following O3 exposure.
18 Mechanisms underlying the rapid increase in airways resistance following O3 exposure are
19 incompletely understood. However pretreatment with atropine was found to decrease baseline sRaw
20 and prevent O3-induced increases in sRaw (Beckett et al., 1985, 039758). indicating the involvement
21 of muscarinic cholinergic receptors of the parasympathetic nervous system. Interestingly, atropine
22 pretreatment partially blocked the decrease in FEVi, but had no effect on the decrease in FVC,
23 breathing rate, tidal volume or respiratory symptoms (Beckett et al., 1985, 039758). Thus pulmonary
24 function decrements measured as FEVi may reflect both restrictive and obstructive type changes in
25 airways responses. Using a beta-adrenergic agonist, it was shown that smooth muscle contraction,
26 not increased airways mucus section, was responsible for O3-induced increases in airways resistance
27 (Beckett et al., 1985, 039758). Furthermore, tachykinins may contribute to O3-mediated increases in
28 airways resistance. Bronchopulmonary C fibers mediate local axon responses by releasing
29 tachykinins such as SP. Tachykinins bind to neurokinin (NK) receptors resulting in responses such as
30 bronchoconstriction. In one study in which bronchial biopsies were performed and studied by
31 immunohistochemistry, SP was substantially diminished in submucosal sensory nerves 6 hours
32 following O3 exposure (0.2 ppm O3 for 2 hours with exercise) (Krishna et al., 1997, 084262). A
33 statistically significant correlation was observed between loss of SP immunoreactivity from neurons
34 in the bronchial mucosa and changes in FEVi measured 1 hour postexposure (Krishna et al., 1997,
35 084262). Another study found that SP was increased in lavage fluid of human subjects immediately
36 after O3 challenge (0.25 ppm O3 for 1 hour with heavy exercise) (Hazbun et al., 1993, 043914) These
37 results provide evidence that the increased airways resistance observed following O3 exposure is due
March 2011 5-31 DRAFT - DO NOT CITE OR QUOTE
-------
1 to vagally-mediated responses and possibly by local axon reflex responses through
2 bronchopulmonary C fiber-mediated release of SP.
3 In responsive individuals, a striking degree of response attenuation occurs following repeated
4 daily exposures to O3. This phenomena has been reported for both lung function and symptoms such
5 as upper airway irritation, nonproductive cough and substernal discomfort and pain upon deep
6 inspiration (Folinsbee et al, 1980, 038880: Hackney et al, 1977, 038282: Horvath et al., 1981,
7 039221). Repeated daily exposures also led to an attenuation of the sRaw response in exercising
8 humans (Christian et al., 1998, 029925). It is well-established that a young O3 responder will no
9 longer be responsive on the fourth or fifth day of consecutive daily O3 exposure (0.4 ppm O3 for
10 4 hours or 0.2 ppm O3 for 4 hours) and that after developing this tolerance it takes up to 7-10 days of
11 non-exposure in order for the subject to regain O3 responsiveness (Christian et al., 1998, 029925:
12 Devlin et al., 1997, 083577: Folinsbee et al., 1980, 038880: Hackney et al., 1977, 038282: Horvath
13 et al., 1981, 039221: Linn et al., 1982, 039646). One group reported persistent small airway
14 dysfunction despite attenuation of the FEVi response on the third day of consecutive O3 exposure
15 (Frank et al., 2001, 093491). Studies in animals also indicate an attenuation of the physiologic
16 response as measured by breathing patterns and tidal volume following five consecutive days of O3
17 exposure (Tepper et al., 1989, 041991). The mechanisms underlying this attenuation in humans and
18 animal models are not well understood (Devlin et al., 1997, 083577). although some studies have
19 implicated an alteration in lung antioxidant capacity, enhanced mucus production or factors related
20 to epithelial hyperplasia following O3 exposure (Devlin et al., 1997, 083577). These potential
21 mechanisms and others are discussed below. Adaptation of O3-induced bradycardic responses, which
22 also result from activation of neural reflexes, have also been reported in animal studies (Hamade and
23 Tankersley, 2009, 596386: Watkinson et al., 2001, 016245).
24 Finally, the degree of acute decrease of vital capacity in young, healthy adults does not
25 correlate (positively) with the degree of neutrophilic inflammation observed in their airways or the
26 degree of airways obstruction elicited by O3 exposure (Aris et al., 1993, 038275: Aris et al., 1995,
27 075945: Balmes et al., 1996, 080830: Schelegle et al., 1991, 042491). implying that different
28 mechanisms are at work. Further evidence is provided by the finding that pretreatment with
29 ibuprofen attenuated O3-induced lung function changes and increases in BAL fluid PGE2 levels, but
30 had no effect on neutrophilia (Hazucha et al., 1996, 043923).
5.2.3.1. New Cellular and Molecular Insights
31 Recent studies in animals provide new information regarding the effects of O3 on reflex
32 responses mediated by bronchopulmonary C-fibers, a type of nociceptive sensory nerve. Ozone
33 exposure in mice was found to selectively activate a subset of receptors on bronchopulmonary C-
34 fibers which are TRPA1 ion channels (Taylor-Clark and Undem, 2010, 377143). TRPA1 ion
35 channels are members of the TRP family of ion channels, which are known to mediate the responses
36 of sensory neurons to inflammatory mediators (Caceres et al., 2009, 628549). In addition to TRPA1
March 2011 5-32 DRAFT - DO NOT CITE OR QUOTE
-------
1 ion channels possibly playing a key role in O3-induced decrements in pulmonary function, they may
2 mediate allergic asthma (Caceres et al., 2009, 628549). Activation of TRPA1 ion channels following
3 O3 exposure is likely due to the formation of secondary products such as aldehydes and
4 prostaglandins (Taylor-Clark and Undem, 2010, 377143). Ozonation of unsaturated fatty acids in the
5 ELF results in the generation of aldehydes (Frampton et al., 1999, 040757). For example
6 4-hydroxynonenal and 4-oxononenal are derived from the peroxidation of omega-6 unsaturated fatty
7 acids (Taylor-Clark et al., 2008, 628565; Trevisani et al., 2007, 628590). 4-oxononenal is a stronger
8 electrophile than 4-hydroxynonenal and exhibits greater potency towards the TRPA1 channels
9 (Taylor-Clark et al., 2008, 628565). TRPA1 channels can be activated by aldehydes and other
10 electrophiles through covalent modification of cysteine and lysine residues (Trevisani et al., 2007,
11 628590). In addition, PGE2 is known to sensitize TRPA1 channels (Bang et al., 2007, 628545) and,
12 as noted previously, PGE2 (or other products of cyclooxygenase-catalyzed reactions) has been
13 proposed to mediate the pulmonary function changes observed following O3 exposure in humans.
14 In addition to stimulating central nervous system reflexes, bronchopulmonary C fibers mediate
15 local axon responses by releasing neuropeptides such as SP, neurokinin (NK) A and calcitonin gene-
16 related peptide (CGRP). Oslund et al. (2008, 195654) demonstrated that NK-1 receptor blockade had
17 no effect on O3-stimulated physiologic responses such as tidal volume and breathing frequency in
18 rats over the 8-h exposure period. However, SP and NK receptors contributed to vagally-mediated
19 bronchoconstriction in guinea pigs 3 days after a single exposure to O3 (2 ppm for 4 hours) (Verhein
20 etal. 2011. 670295).
5.2.3.2. Summary
21 A key event in ozone's toxicity pathway is the activation of neural reflexes which leads to
22 decrements in pulmonary function. Evidence is accumulating that secondary oxidation products are
23 responsible for this effect. Eicosanoids have been implicated in humans while eicosanoids and
24 aldehydes are effective in animal models. Different receptors on bronchial C-fibers have been shown
25 to mediate separate effects of O3 on pulmonary function. Nociceptor sensory nerves are involved in
26 the involuntary truncation of respiration which results in decreases in FVC, FEVi, tidal volume and
27 an increase in respiratory frequency and pain upon deep inspiration. Opioids block these responses
28 while atropine does not. New evidence in an animal model suggests that TRPA1 receptors on
29 bronchial C-fibers mediate this pathway. Ozone exposure also results in activation of vagal sensory
30 nerves and a mild increase in airways obstruction measured as increased sRaw. Atropine and beta-
31 adrenergic agonists blocked this response in one study indicating that the airway obstruction was due
32 to bronchoconstriction. Other studies in humans implicated SP release from bronchial C-fibers
33 resulting in airway narrowing due to either neurogenic edema or bronchoconstriction. New evidence
34 in an animal model suggests that the SP-NK receptor pathway caused bronchoconstriction following
35 O3 exposure. Considerable inter-individual variability exists in O3 responsiveness measured by
36 decrements in pulmonary function. Further, attenuation of these pulmonary function decrements
March 2011 5-33 DRAFT - DO NOT CITE OR QUOTE
-------
1 occurs following O3 exposure for several consecutive days. Mechanisms responsible for these effects
2 are not known but may be related to inherent differences in neural sensitivity.
5.2.4. Respiratory Tract Injury and Inflammation
3 As described above, O3 reacts with components of the ELF resulting in the generation of
4 secondary oxidation products. Higher concentrations of these products may directly injure
5 respiratory tract epithelium. Lower concentrations may initiate cellular responses including cytokine
6 generation, adhesion molecule expression and modification of tight junctions leading to
7 inflammation and increased permeability across airways epithelium (Dahl et al., 2007, 196986;
8 Mudway and Kelly, 2000, 010452). Subsequent epithelial remodeling may also occur (Mudway and
9 Kelly, 2000, 010452V
10 Injury and inflammation have been observed in many different regions of the respiratory tract
11 following O3 exposure (Plopper et al., 1998, 087203: Postlethwait et al., 2000, 003000V The nasal
12 airways, conducting airways and distal airways (i.e. respiratory bronchioles or centriacinar region
13 depending on the species) have all been identified as sites of O3-mediated injury and inflammation
14 (Mudway and Kelly, 2000, 010452). One study found greater injury in conducting airways
15 downstream of bifurcations where local doses of O3 were higher (Postlethwait et al., 2000, 003000).
16 Although the extent of O3-induced injury is variable along the respiratory tract, common features of
17 the injury response have been noted (Mudway and Kelly, 2000, 010452). In the conducting airways,
18 necrosis of ciliated cells and degranulation of secretory cells has been observed and in the alveolar
19 region, necrosis of Type 1 pneumocytes occurs.
20 Further, O3-induced injury and inflammation responses are variable between species. For
21 example, Dormans et al. (1999, 040766) found that rats, mice, and guinea pigs all exhibited
22 O3-induced (0.2 - 0.4 ppm for 3-56 days) inflammation; however, guinea pigs were the most
23 sensitive with respect to alveolar macrophage elicitation and pulmonary cell density in the
24 centriacinar region. Mice were the most sensitive to bronchiolar epithelial hypertrophy and
25 biochemical changes (e.g. lactate dehydrogenase, glutathione reductase, glucose-6-phosphate
26 dehydrogenase activity), and had the slowest recovery from O3 exposure. All species displayed
27 increased collagen in the ductal septa and large lamellar bodies in Type II pneumocytes at the longest
28 exposure and highest concentration, whereas this response occurred in the rat and guinea pig at
29 lower O3 levels (0.2 ppm) as well. Since no dose metric was measured, it is possible that some of
30 these differences may be attributable to disparate total inhaled dose or local organ dose. Overall, the
31 authors rated mice as most susceptible, followed by guinea pigs, then rats (Dormans et al., 1999,
32 040766). Rats were also less sensitive to epithelial necrosis and inflammatory responses from O3
33 (1.0 ppm for 8 h) than monkeys and ferrets, which manifested a similar response (Sterner-Kock et
34 al., 2000, 013033). These data suggest that ferrets may be a good animal model for O3-induced
35 airway effects due to the similarities in pulmonary structure between primates and ferrets.
March 2011 5-34 DRAFT - DO NOT CITE OR QUOTE
-------
1 While injury and inflammation often accompany each other, in many cases epithelial injury
2 precedes observable inflammatory effects and in other cases inflammation leads to injury of the
3 surrounding cells and tissues. In addition to being species-dependent, the acute injury response is
4 focal, site-specific and dependent on exposure parameters and the time that elapsed since exposure
5 (Postlethwait et al., 2000, 003000). The presence of shed epithelial cells in the BAL, increases in
6 levels of BAL lactate dehydrogenase and protein and increased epithelial permeability have been
7 observed and are indicative of epithelial injury. In addition, histologic analysis has demonstrated
8 damage to tight junctions between epithelial cells, suggesting an increase in epithelial permeability.
9 Several studies have measured epithelial permeability as the flux of the small solute 99mTc-
10 DTPA which was introduced into the air spaces in different regions of the respiratory tract. An early
11 study demonstrated increased pulmonary epithelial permeability, measured as the clearance of
12 99mTc-DTPA, in humans exposed for 2 h to 0.4 ppm O3 while exercising moderately (Kehrl et al.,
13 1987, 040824). Another study found that increased epithelial permeability occurred at 1-3 hours and
14 18-20 hours postexposure and did not resolve for several days (Foster and Stetkiewicz, 1996,
15 079920). Increased bronchial permeability was also observed in dogs immediately after and 18 hours
16 postexposure to 0.2 ppm O3 for 6 hours (Freed et al., 1996, 080798). Increased epithelial
17 permeability has been proposed to play a role in allergic sensitization (Matsumura, 1970, 050626). in
18 activation of neural reflexes and in stimulation of smooth muscle receptors (Dimeo et al., 1981,
19 039662). Studies in animals have also demonstrated increased vascular permeability, as measured by
20 BAL protein and albumin (Costa et al., 1985, 040273: Hu et al., 1982, 039418).
21 An important hallmark of acute O3 exposure in humans and animals is neutrophilic airways
22 inflammation. Although neutrophil influx into nasal airways has been demonstrated in human
23 subjects (Graham and Koren, 1990, 042299). most studies of neutrophil influx have focused on the
24 lower airways (Aris et al., 1993, 038275: Hazucha et al., 1996, 043923). The time course of this
25 response and its resolution is slower than that of the decrements in pulmonary function (Hazucha et
26 al., 1996, 043923). In general, airways neutrophilia is observable within 1-2 hours, peaks at
27 4-6 hours and is returning to baseline levels at 24 h following exposure to O3 in exercising humans
28 involving 0.4 ppm for 2 hours (Devlin et al., 1991, 040359) or 0.3 ppm for 1 hour (Schelegle et al.,
29 1991, 042491). Since the influx and persistence of neutrophils in airways following O3 exposure
30 correlates with the temporal profile of epithelial injury (Hu et al., 1982, 039418). neutrophils are
31 likely to be injurious. However, neutrophils can contribute to the repair of O3-injured epithelium by
32 removing necrotic epithelial cells (Mudway and Kelly, 2000, 010452: Vesely et al., 1999, 051045).
33 The degree of airways inflammation due to O3 is thought to have more important long-term
34 consequences than the more quickly resolving changes in pulmonary function since airways
35 inflammation is often accompanied by tissue injury (Balmes et al., 1996, 080830).
36 The influx of inflammatory cells in the airways of human subjects has been assessed by
37 bronchoscopy and by morphometric measurements in bronchial mucosal biopsies. Bronchoscopy can
38 be used to sample fluid from all of the airways and from the lung parenchyma distal to the wedged
March 2011 5-35 DRAFT - DO NOT CITE OR QUOTE
-------
1 bronchoscope tip or just the more proximal portion. While many studies have documented a distal
2 lung inflammatory response using conventional (pooled) BAL, a lesser number of studies have
3 documented inflammation in the proximal airways using "proximal" BAL sampling (i.e., liquid
4 aspirated from a 20-30 mL initial lavage aliquot after wedging the bronchoscope or from the left
5 main bronchus transiently isolated by inflation of proximal and distal balloons) or by using bronchial
6 mucosal biopsy (Aris et al., 1993, 038275; Schelegle et al., 1991, 042491). Airways neutrophilia was
7 observed at 1 and 6 hours postexposure in proximal airways BAL but only at 6 hours postexposure
8 in BAL from all of the airways combined (Schelegle et al., 1991, 042491). This result demonstrated
9 that measurements made in proximal airways BAL better reflected the earliest phase of airways
10 inflammation than measurements made in conventional multi-aliquot (pooled) BAL.
11 Inter-individual variability in the neutrophilic response has been noted (Devlin et al., 1991,
12 040359: Holz et al., 1999, 058731: Schelegle et al., 1991, 042491). One study demonstrated a
13 threefold difference in airways neutrophilia, measured as percent of total cells in proximal BAL,
14 among human subjects exposed to 0.3 ppm O3 for 1 hour while exercising (Schelegle et al., 1991,
15 042491). while a 20-fold difference was demonstrated in BAL neutrophils following exposure to
16 0.08-0.10 ppm O3 for 6.6 hours while exercising (Devlin et al., 1991, 040359). Reproducibility of
17 intra-individual responses to 0.25 ppm O3, measured as sputum neutrophilia, was demonstrated by
18 Holz (1999, 058731). Few studies have examined the dose-responsiveness of airways neutrophilia in
19 Os-exposed humans (Devlin et al., 1991, 040359: Holz et al., 1999, 058731V No dose-
20 responsiveness was observed in healthy human subjects exposed for 1 hour to 0.125 and 0.25 ppm
21 O3 and a statistically significant increase in sputum neutrophilia was observed only at the higher
22 dose (Holz et al., 1999, 058731). However, dose-dependent and statistically significant increases in
23 BAL neutrophils and the inflammatory mediator IL-6 were reported following exposure to 0.08 and
24 0.1 ppm O3 for 6.6 hours in exercising humans (Devlin et al., 1991, 040359). Additional evidence is
25 provided by a meta-analysis of the O3 dose-inflammatory response in controlled human exposure
26 studies involving exposure to 0.08-0.6 ppm O3 for 60-396 minutes (Mudway and Kelly, 2004,
27 399328). Results demonstrated a linear relationship between inhaled O3 dose (determined as the
28 product of concentration, ventilation and time) and BAL neutrophils at 0-6 hours and 18-24 hours
29 following O3 exposure (Mudway and Kelly, 2004, 399328).
30 Ozone exposure results in alterations in other airways inflammatory cells besides neutrophils.
31 Numbers of lymphocytes and total cells in BAL fluid were decreased early after O3 exposure,
32 preceding the neutrophil influx (Blomberg et al., 1999, 001267: Krishna et al., 1997, 084262:
33 Mudway and Kelly, 2000, 010452). The decrease in total cells was thought to reflect decreases in
34 airway macrophages, although it was not clear whether the cells were necrotic or whether membrane
35 adhesive properties were altered making them more difficult to obtain by lavage (Blomberg et al.,
36 1999, 001267: Frampton et al., 1997, 086111: Mudway and Kelly, 2000, 010452: Mudway et al.,
37 1999, 011833: Pearson and Bhalla, 1997, 082686). Recent studies have demonstrated increases in
38 numbers of sputum monocytes and dendritic-like cells (Alexis et al., 2010, 628538) (discussed
March 2011 5-36 DRAFT - DO NOT CITE OR QUOTE
-------
1 further in section 5.2.7). Increases in submucosal mast cells were observed 1.5 hours after a 2-h
2 exposure of healthy human subjects to 0.2 ppm O3 (Blomberg et al., 1999, 001267) and increases in
3 BAL mast cell number were observed 18 hours after O3 exposure (Frampton et al., 1997, 086111).
4 Mast cells may play an important role in mediating neutrophil influx since they are an important
5 source of several pro-inflammatory cytokines and since their influx precedes that of the neutrophils
6 (Blomberg et al., 1999, 001267; Stenfors et al., 2002, 030473). Further, a study using mast cell-
7 deficient mice demonstrated decreased neutrophilic inflammation in response to O3 compared with
8 wild type mice (Kleeberger et al., 1993, 044203). The mechanisms involved in clearing O3-provoked
9 inflammation remain to be clarified.
10 The cellular and molecular signals involved in injury and inflammatory responses following
11 O3 exposure have been extensively evaluated (U.S. EPA, 2006, 088089). Eicosanoids are one class
12 of secondary oxidation products which may be formed rapidly following O3 exposure and which
13 may mediate injury and inflammation. Eicosanoids are metabolites of arachidonic acid, a 20-carbon
14 polyunsaturated fatty acid, which is released from membrane phospholipids by phospholipase
15 A2-mediated catalysis. Activation of phospholipase A2 occurs by several cell signaling pathways
16 and may be triggered by O3-mediated lipid peroxidation of cellular membranes (Rashba-Step et al.,
17 1997, 628562). Additionally, cellular phospholipases A2, C and D may be activated by lipid
18 ozonation products (Kafoury et al., 1998, 016913). While the conversion of arachidonic acid to
19 prostaglandins, leukotrienes and other eicosanoid products is generally catalyzed by
20 cyclooxygenases and lipoxygenases, non-enzymatic reactions also occur during oxidative stress
21 leading to the generation of a wide variety of eicosanoids and reactive oxygen species. Further, the
22 release of arachidonic acid from phospholipids is accompanied by the formation of
23 lysophospholipids which are precursors for platelet activating factors. Thus, formation of
24 eicosanoids, reactive oxygen species and platelet activating factors accompanies O3-mediated lipid
25 peroxidation.
26 Additional cell signaling mediators are generated subsequent to O3 exposure. Secondary
27 reaction products may stimulate airway macrophages to produce cytokines such as IL-1, IL-6 and
28 TNF-a which in turn activate IL-8 production by epithelial cells. Ozone exposure is also known to
29 upregulate the vascular endothelial adhesion molecules P-selectin and ICAM-1 (Blomberg et al.,
30 1999, 001267; Krishna et al., 1997, 084262) and to increase the expression of pro-inflammatory
31 mediators GM-CSF, Gro-a and IL-8 (Mudway and Kelly, 2000, 010452). In addition, lung epithelial
32 cells may release ATP in response to O3 exposure (Ahmad et al., 2005, 196429). ATP and its
33 metabolites (catalyzed by ecto-enzymes) can bind to cellular purinergic receptors resulting in
34 activation of cell signaling pathways (Picher et al., 2004, 644780). One such metabolite, adenine, is
35 capable of undergoing oxidation leading to the formation of urate which, if present in high
36 concentrations, could activate inflammasomes and result in caspase 1 activation and the maturation
37 and secretion of IL-lbeta and IL-18 (Dostert et al., 2008, 155753).
March 2011 5-37 DRAFT - DO NOT CITE OR QUOTE
-------
1 Many studies have focused on cell signaling pathways leading to airways neutrophilia in
2 humans or animal models which generally peaks 4-6 hours after exposure. Although IL-8 has been
3 proposed to play a role in neutrophil chemotaxis, measurements of IL-8 in lavage fluid from humans
4 exposed to O3 show increases that are too late to account for this effect (Mudway and Kelly, 2000,
5 010452). However, the profiles of PGE2 and IL-6 responses suggest that they may play a role in
6 neutrophil chemotaxis. A study in mice demonstrated that PAF may be important in this response
7 (Longphre et al., 1999, 001199). while macrophage inflammatory protein-2 (MIP-2) and ICAM-1
8 have also been implicated in a rat model (Bhalla and Gupta, 2000, 015036). Other studies have
9 investigated mechanisms involved in earlier or later phases of neutrophilic airways inflammation and
10 inflammation occurring in the nasal airways and more distal lung (U.S. EPA, 2006, 088089).
11 One set of studies in humans focused on the earliest phase of airways inflammation (1-2 hours
12 following exposure). Exercising subjects were exposed to 0.2 ppm O3 for 2 hours and bronchial
13 biopsy tissues were obtained 1.5 and 6 hours after exposure (Blomberg et al., 1999, 001267; Bosson
14 et al., 2003, 051687: Bosson et al., 2009, 399331: Stenfors et al., 2002, 030473). Results
15 demonstrated upregulation of vascular endothelial adhesion molecules P-selectin and ICAM-1 at
16 both 1.5 and 6 hours (Blomberg et al., 1999, 001267: Stenfors et al., 2002, 030473). Submucosal
17 mast cell numbers were increased at 1.5 hours in the biopsy samples without an accompanying
18 increase in neutrophil number (Blomberg et al., 1999, 001267). Pronounced neutrophil infiltration
19 was observed at 6 h in the bronchial mucosa (Stenfors et al., 2002, 030473). Surprisingly,
20 suppression of the NFkB and AP-1 pathways at 1.5 hours and a lack of increased IL-8 at 1.5 or
21 6 hours in bronchial epithelium was observed (Bosson et al., 2009, 399331). The authors suggested
22 that vascular endothelial adhesion molecules, rather than redox sensitive transcription factors, are
23 key to early neutrophil recruitment in response to O3.
24 Multi-day exposure to O3 has been found to dampen the inflammatory response, but not the
25 injury response, compared with a single day exposure (Christian et al., 1998, 029925: Devlin et al.,
26 1997, 083577). In human subjects exposed for 4 hours to 0.2 ppm O3 during moderate exercise,
27 decreased numbers of BAL neutrophils were observed after 4 days of consecutive exposure
28 compared with responses after 1 day (Christian et al., 1998, 029925). Results indicated an
29 attenuation of the inflammatory response in both proximal airways and distal lung. However
30 repeated exposure did not result in attenuation of the injury markers lactate dehydrogenase and
31 protein in the BAL. Similar results were found in a study of humans undergoing heavy intermittent
32 exercise who were exposed for 2 hours to 0.4 ppm O3 for 5 consecutive days (Devlin et al., 1997,
33 083577). In this latter study, partial recovery of the inflammatory response was noted 10 days
34 following the exposure (Devlin et al., 1997, 083577). In an animal study conducted in parallel (Van
35 Bree et al., 2002, 035452). full susceptibility to O3 challenge following exposure to O3 for 5
36 consecutive days required 15-20 days recovery. Further, no attenuation of cellular proliferation in
37 terminal bronchioles was observed during the 5 consecutive days of O3 exposure. In a separate study
38 in rats involving repeated O3 exposures, a lack of attenuation of the injury marker lavagable protein,
March 2011 5-38 DRAFT - DO NOT CITE OR QUOTE
-------
1 the persistence of macrophages in the centriacinar region, and histological evidence of progressive
2 tissue injury was demonstrated (Tepper et al., 1989, 041991). Thus, the inflammatory response
3 resembled that of the pulmonary function response which was attenuated with repeated short-term
4 O3 exposure in both human subjects and animals (Christian et al., 1998, 029925; Hackney et al.,
5 1977, 038282: Horvath et al., 1981, 039221). Findings that injury, measured by BAL markers or by
6 histopathology, persisted in the absence of inflammation or pulmonary function decrements suggests
7 that, despite adaptation, repeated exposure to O3 may have serious long-term consequences such as
8 airway remodeling. The mechanisms involved in clearing O3-provoked inflammation remain to be
9 clarified.
10 Increases in markers of inflammation and of injury occurred to a comparable degree in
11 subjects with mild (least sensitive) and more remarkable (more sensitive) spirometric responses to
12 O3 (Balmes et al., 1996, 080830). Two other studies using similar exposure protocols found that
13 acute spirometric changes were not positively correlated with cellular and biochemical indicators of
14 inflammation (Aris et al., 1993, 038275; Schelegle et al., 1991, 042491). However inflammation was
15 correlated with changes in sRaw (Balmes et al., 1996, 080830). In another study, pretreatment with
16 ibuprofen had no effect on neutrophilia although it blunted the spirometric response (Hazucha et al.,
17 1996, 043923). Taken together, results from these studies indicate different mechanisms underlying
18 the spirometric and inflammatory responses to O3.
19 In contrast, a common mechanism underlying both inflammation and impaired pulmonary
20 function was suggested by (Krishna et al., 1997, 084262). This study, conducted in exercising
21 humans exposed to 0.2 ppm O3 for 2 hours, demonstrated a correlation between loss of SP
22 immunoreactivity from neurons in the bronchial mucosa and numbers of neutrophils and epithelial
23 cells (shed epithelial cells are an index of injury) in the BAL 6 h postexposure. Furthermore, the loss
24 of SP immunoreactivity was correlated with the observed changes in FEVi. SP is a neuropeptide
25 released by sensory nerves which mediates neurogenic edema and bronchoconstriction (Krishna et
26 al., 1997, 084262). Further, another study found that SP was increased in lavage fluid of human
27 subjects immediately after O3 challenge (Hazbun et al., 1993, 043914). Taken together, these
28 findings suggest O3-mediated stimulation of sensory nerves leading to activation of central and local
29 axon reflexes as a common effector pathway leading to impaired pulmonary function and
30 inflammation.
31 Chronic exposure to O3 has been studied in animal models. In the nasal airways, exposure to
32 O3 for days or weeks results in mucous cell metaplasia of nasal transitional epithelium (Harkema et
33 al., 1999, 001209: Hotchkiss et al., 1991, 042441). This remodeling effect was characterized by
34 neutrophilic infiltration, a loss of sensitive nasal epithelial cells, the proliferation of resistant
35 epithelial cells and mucin gene overexpression (Cho et al., 1999, 011985). Bronchiolitis, bronchiolar
36 metaplasia of alveolar ducts, proliferation of Type 2 pneumocytes and fibrosis were reported
37 following chronic O3 exposure (Mudway and Kelly, 2000, 010452). It was suggested that remodeling
38 of respiratory epithelium may lead to enhanced resistance or tolerance to O3 (Mudway and Kelly,
March 2011 5-39 DRAFT - DO NOT CITE OR QUOTE
-------
1 2000, 010452). Deposition of collagen in the airways and sustained lung function decrements
2 especially in small airways have also been demonstrated as a response to chronic O3 exposure
3 (Chang et al, 1992, 042387: Mudway and Kelly, 2000, 010452).
5.2.4.1. New Cellular and Molecular Insights
4 Recent investigations in animal models have elucidated additional mechanisms involved in
5 O3-induced inflammation and injury. In one study, tachykinins working through NK-1 and CGRP
6 receptors were found to contribute to airways epithelial injury, but not to neutrophil influx, in O3-
7 exposed rats (Oslund et al., 2008, 195654: Oslund et al., 2009, 201539). Key roles for CXCR2, a
8 receptor for the cytokines KC and MIP-2, and for IL-6 in O3-mediated neutrophil influx were
9 demonstrated in mice (Johnston et al., 2005, 596393: Johnston et al., 2005, 596394). Activation of
10 JNK and p38 pathways and cathepsin-S were also found to be important in this response (Williams
11 et al., 2007, 628609: Williams et al., 2008, 628607: Williams et al., 2009, 628605). Furthermore,
12 matrix metalloproteinase-9 (MMP-9) protected against O3-induced airways inflammation and injury
13 in mice (Yoon et al.. 2007. 596422).
14 Williams et al. (2007, 597545) found that the toll-like receptor (TLR) adaptor protein MyD88
15 was important in mediating O3-induced neutrophilia in mice exposed to 3 ppm O3 for 3 hours, with
16 TLR4 and TLR2 contributing to the speed of the response. Moreover, MyD88, TLR2 and TLR4
17 contributed to inflammatory gene expression in this model and O3 upregulated MyD88, TLR4 and
18 TLR4 gene expression. These results complement those of Hollingsworth et al. (2004, 097816) who
19 demonstrated airways neutrophilia following acute (2 ppm O3 for 3 hours) and subchronic O3
20 exposure (0.3 ppm for 3 days) in a mouse model (Hollingsworth et al., 2004, 097816). In this study,
21 airways neutrophilia was not dependent on TLR4 (Hollingsworth et al., 2004, 097816). Ozone
22 effects on lung hyperpermeability, which is often correlated with neutrophil influx, were previously
23 found to require a functioning TLR4 (Kleeberger et al., 2000, 014895).
24 Other studies focused on the role of hyaluronan in mediating a later phase (24 hours) of
25 O3-induced inflammation in mice (Garantziotis et al., 2009, 597603: Garantziotis et al., 2010,
26 624947). Hyaluronan is an extracellular matrix component which is normally found in the ELF as a
27 large polymer. Exposure to 2 ppm O3 for 3 hours resulted in elevated levels of soluble low molecular
28 weight hyaluronan in the BAL fluid 24 hours postexposure (Garantziotis et al., 2009, 597603:
29 Garantziotis et al., 2010, 624947). Ozone may have caused the depolymerization of hyaluronan to
30 soluble fragments which are known to be endogenous ligands of the CD44 receptor and TLR4 in the
31 macrophage (Jiang et al., 2005, 628556). Binding of hyaluronan fragments to the CD44 receptor
32 activates hyaluronan clearance, while binding to TLR4 results in signaling through MyD88 to
33 produce chemokines that stimulate the influx of inflammatory cells (Jiang et al., 2005, 628556).
34 Activation of NFkB occurred in both airway epithelia and alveolar macrophages 24 hours
35 postexposure to O3. Increases in BAL pro-inflammatory factors KC, IL-lbeta, MCP-1, tumor
36 necrosis factor-a (TNF-a) and IL-6 observed 24 hours following O3 exposure were found to be
March 2011 5-40 DRAFT - DO NOT CITE OR QUOTE
-------
1 partially dependent on TLR4 while increases in BAL inflammatory cells, which consisted mainly of
2 macrophages, were dependent on CD44. BAL inflammatory cells number and injury markers
3 following O3 exposure were similar in wild-type and TLR4-deficient animals.
5.2.4.2. Summary
4 Injury and inflammation are key events in ozone's toxicity pathway. Secondary oxidation
5 products have been implicated in a number of these processes. Although there may be inter-species
6 differences with respect to specific mediators, mechanisms involved in the acute responses to O3
7 include epithelial injury and airways neutrophilia. Longer-term exposures may result in mucus cell
8 metaplasia of nasal epithelium or airways remodeling and fibrosis. Work from several laboratories in
9 humans and animal models suggest that O3 triggers the release of tachykinins such as SP from
10 airway sensory nerves which could contribute to downstream effects including injury and
11 inflammation. New investigations show that O3 exposure leads to the generation of hyaluronan
12 fragments which activate TLR4 and CD44-dependent signaling pathways in macrophages and result
13 in a greater turnover of macrophage populations in the lung. Activation of these pathways occurs
14 later than the acute neutrophilic response suggesting that they may contribute to longer-term effects
15 of O3. The mechanisms involved in clearing O3-provoked inflammation remain to be clarified.
16 Similar to the pulmonary function responses discussed in the previous section, considerable
17 inter-individual variability exists in O3 responsiveness as measured by airways neutrophilia. Further,
18 attenuation of the inflammatory response occurs following O3 exposure for several consecutive days.
19 However evidence suggests that injury may continue despite the dampening of the inflammatory
20 response during repeated exposures. Mechanisms responsible for inter-individual variability and
21 response attenuation, or the lack thereof, are not known. It should be noted that inflammation, as
22 measured by airways neutrophilia, is not correlated with decrements in pulmonary function as
23 measured by spirometry. Consequently, spirometric measures are not a good surrogate for the degree
24 of inflammation in any given individual following O3 exposure. Furthermore, airways neutrophilia
25 may not be a good indicator of O3-mediated lung injury.
5.2.5. Increased Bronchial Reactivity
26 In addition to causing mild airways obstruction as discussed above, acute O3 exposure results
27 in reversible increases in bronchial reactivity by mechanisms which are not well understood. These
28 effects may be more significant in human subjects with already compromised airways
29 (Section 5.2.6). Bronchial reactivity is generally determined in terms of a response to a challenge
30 agent. Non-specific bronchial reactivity in humans is assessed by measuring the effect of inhaling
31 increasing concentrations of a bronchoconstrictive drug on lung mechanics (sRaw or FEVi).
32 Methacholine is most commonly employed but histamine and other agents are also used. Specific
33 bronchial reactivity is assessed by measuring effects in response to an inhaled allergen in individuals
March 2011 5-41 DRAFT - DO NOT CITE OR QUOTE
-------
1 (or animals) already sensitized to that allergen. An increase in sRaw in response to non-specific or
2 specific challenge agents indicates airways hyperresponsiveness (AHR).
3 Ozone may sensitize bronchial smooth muscle to stimulation through a direct effect on smooth
4 muscle or through effects on the sensory nerves in the epithelium or on the motor nerves innervating
5 the smooth muscle (Holtzman et al, 1979, 039220: O'Byrne et al, 1983, 041366: O'Byrne et al,
6 1984, 040066). One possibility may be related to O3-mediated increases in epithelial permeability,
7 which would improve access of agonist to smooth muscle receptors (Holtzman et al., 1979, 039220).
8 Neurally-mediated effects have been demonstrated in several studies. In one, pretreatment with
9 atropine was found to block O3-induced AHR, suggesting the involvement of cholinergic
10 postganglionic pathways (Holtzman et al., 1979, 039220). Other studies in animals demonstrated
11 that Os-induced AHR involved vagally-mediated responses (Freed et al., 1996, 080798) and local
12 axon reflex responses through bronchopulmonary C fiber-mediated release of SP (Joad et al., 1996,
13 082711). Further, pretreatment with capsaicin to deplete nerve fibers of SP blocked O3-mediated
14 bronchial reactivity measured as AHR (Tepper et al., 1993, 628570).
15 Some evidence suggested the involvement of arachidonic acid metabolites (Fabbri et al., 1985,
16 040276: Seltzer et al., 1986, 040383). An early study found AHR in exercising humans immediately
17 postexposure to 0.6 ppm O3 for 2 hours, which was associated with increases in BAL neutrophils and
18 cyclooxygenase products (Seltzer et al., 1986, 040383). Other investigators found that ibuprofen
19 pretreatment had no effect on AHR following exposure to 0.4 ppm O3 for 2 hours, although
20 spirometric responses were blunted (Hazucha et al., 1996, 043923). This study indicated that the
21 arachidonic acid metabolites whose generation was blocked by ibuprofen, an inhibitor of
22 cyclooxygenase and some lipoxygenase activity, (i.e. prostaglandins, thromboxanes and some
23 leukotrienes) did not play a role in AHR. Experiments in dogs demonstrated a close correlation
24 between O3-induced AHR and airway neutrophilic inflammation measured in tissue biopsies
25 (Holtzman et al., 1983, 039745). Furthermore, the increased AHR observed in dogs following O3
26 exposure was inhibited by neutrophil depletion (O'Byrne et al., 1983, 041366) and by pre-treatment
27 with inhibitors of arachidonic acid metabolism. In one of these studies, indomethacin pre-treatment
28 did not prevent airway neutrophilia in response to O3 providing evidence that the subset of
29 arachidonic acid metabolites whose generation was inhibitable by the cyclooxygenase inhibitor
30 indomethacin (i.e., prostaglandins and thromboxanes) was not responsible for neutrophil influx
31 (O'Byrne et al., 1984, 040066). Taken together, these findings suggest that arachidonic acid
32 metabolites, but probably not prostaglandins or thromboxanes, may be involved in the AHR response
33 following O3 exposure in dogs.
34 Later it was recognized that increased bronchial reactivity can be both a rapidly occurring and
35 persistent response to O3 (Foster and Freed, 1999, 001202). Secondary oxidation products of O3 and
36 tachykinins have been proposed as early mediators of the response and inflammation-derived
37 products have been proposed as mediators of the later response (Foster and Freed, 1999, 001202).
38 Some studies have suggested an involvement of IL-1 (Park et al., 2004, 644864) and TNF-a (Cho et
March 2011 5-42 DRAFT - DO NOT CITE OR QUOTE
-------
1 al, 2001, 016160; Shore et al., 2001, 018993). Furthermore, multiday exposure to O3 has been found
2 to dampen the AHR response compared with a single day exposure (Dimeo et al., 1981, 039662).
5.2.5.1. New Cellular and Molecular Insights
3 Recent studies in animal models provide new evidence for mechanisms underlying increased
4 bronchial reactivity. In guinea pigs, AHR was found to be mediated by different pathways at 1 and
5 3 days postexposure to a single dose of O3 (2 ppm for 4 hours) (Verhein et al., 2011, 670295; Yost et
6 al., 2005, 597549). At 1 day, airway hyperreactivity was due to activation of airway parasympathetic
7 nerves rather than to a direct effect on smooth muscle (Yost et al., 2005, 597549). This effect
8 occurred as a result of O3-stimulated release of major basic protein from eosinophils (Yost et al.,
9 2005, 597549). Major basic protein is known to block inhibitory M2 muscarinic receptors which
10 normally dampen acetylcholine release from parasympathetic nerves (Yost et al., 2005, 597549). The
11 resulting increase in acetylcholine release caused an increase in smooth muscle contraction
12 following O3 exposure (Yost et al., 2005, 597549). Eosinophils played a different role 3 days
13 postexposure to O3 in guinea pigs (Yost et al., 2005, 597549). Ozone-mediated influx of eosinophils
14 into lung airways resulted in a different population present 3 days postexposure compared to those
15 present at one day (Yost et al., 2005, 597549). At this point, eosinophil-derived major basic protein
16 increased smooth muscle responsiveness to acetylcholine which also contributed to AHR (Yost et al.,
17 2005, 597549). However, the major effect of eosinophils was to protect against vagal hyperreactivity
18 (Yost et al., 2005, 597549). The authors suggested that these beneficial effects were due to the
19 production of nerve growth factor (Yost et al., 2005, 597549). Further work by these investigators
20 demonstrated a key role for IL-lbeta in mediating AHR three days postexposure to O3 (Verhein et
21 al., 2011, 670295). In this study, IL-lbeta increased nerve growth factor and SP which acted through
22 the NK1 receptor to cause vagally-mediated bronchoconstriction (Verhein et al., 2011, 670295). The
23 mechanism by which SP caused acetylcholine release from parasympathetic nerves following O3
24 exposure was not determined (Verhein et al., 2011, 670295). Two studies by other investigators
25 demonstrated that SP released from airways nociceptive neurons contributed to O3-induced AHR
26 measured 3 hours postexposure to 2 ppm O3 in ferrets (Wu et al., 2003, 628616; Wu et al., 2008,
27 597548). These authors further suggested that SP expression in airway neurons was upregulated by
28 IL-1 which was released in response to O3 (Wu et al., 2008, 597548). Taken together, the above
29 study results indicate that mechanisms involved in O3-mediated AHR can vary over time
30 postexposure and that eosinophils and SP can play a role.
31 Evidence for cytokine and chemokine involvement in the AHR response to O3 has recently
32 been described. Williams et al. (2008, 597546) demonstrated that the Th2 cytokine IL-13 contributed
33 to AHR, as well as to airways neutrophilia, following exposure of mice to 3 ppm O3 for 3 hours.
34 Other studies in mice have demonstrated a key role for CXCR2, the chemokine receptor for the
35 neutrophil chemokines KC and MIP-2, but not for IL-6 in O3-mediated AHR (Johnston et al., 2005,
36 596394; Johnston et al., 2005, 596393). In contrast, CXCR2 and IL-6 were both required for
March 2011 5-43 DRAFT - DO NOT CITE OR QUOTE
-------
1 neutrophil influx in this model (Johnston et al., 2005, 596393; Johnston et al, 2005, 596394). as
2 discussed above.
3 Other studies have focused on the role of TLR4. Rollings-worth measured AHR, as well as
4 airways neutrophilia, in mice 6 and 24 hours following acute (2 ppm O3 for 3 hours) and subchronic
5 (0.3 ppm for 3 days) exposure to O3 (Hollingsworth et al., 2004, 097816). TLR4 is a key component
6 of the innate immune system and is responsible for the immediate inflammatory response seen
7 following challenge with endotoxin and other pathogen-associated substances. In the 2004 study, a
8 functioning TLR4 was required for the full AHR response following O3 exposure but not for airways
9 neutrophilia (Hollingsworth et al., 2004, 097816). These findings are complemented by an older
10 study demonstrating that O3 effects on lung hyperpermeability required a functioning TLR4
11 (Kleeberger et al., 2000, 014895). Williams et al. (2007, 597545) found that TLR2, TLR4 and the
12 TLR adaptor protein MyD88 contributed to AHR in mice exposed to 3 ppm O3 for 3 hours. Ozone
13 was also found to upregulate MyD88, TLR4 and TLR4 gene expression in this model (Williams et
14 al., 2007, 597545).
15 A newly recognized mechanistic basis for O3-induced AHR is provided by studies focusing on
16 the role of hyaluronan following O3 exposure in mice (Garantziotis et al., 2010, 624947)
17 (Garantziotis et al., 2009, 597603). Briefly, TLR4 and CD44 were found to mediate AHR in response
18 to O3 and hyaluronan. Hyaluronan is an extracellular matrix component which is normally found in
19 the ELF as a large polymer. Exposure to 2 ppm O3 for 3 hours resulted in enhanced AHR and
20 elevated levels of soluble low molecular weight hyaluronan in the BAL fluid 24 hours postexposure
21 (Garantziotis et al., 2009, 597603; Garantziotis et al., 2010, 624947). Ozone may have caused the
22 depolymerization of hyaluronan to soluble fragments which are known to be endogenous ligands of
23 the CD44 receptor and TLR4 in the macrophage (Jiang et al., 2005, 628556). In the two recent
24 studies, O3-induced AHR was attenuated in CD44 and TLR4-deficient mice (Garantziotis et al.,
25 2009, 597603; Garantziotis et al., 2010, 624947). Hyaluronan fragment-mediated stimulation of
26 AHR was found to require functioning CD44 receptor and TLR4 (Garantziotis et al., 2009, 597603;
27 Garantziotis et al., 2010, 624947). In contrast, high-molecular-weight hyaluronan blocked AHR in
28 response to O3 (Garantziotis et al., 2009, 597603). In another study high-molecular-weight
29 hyaluronan enhanced repair of epithelial injury (Jiang et al., 2005, 628556). These studies provide a
30 link between innate immunity and the development of AHR following O3 exposure, and indicate a
31 role for TLR4 in increasing bronchial smooth muscle reactivity.
5.2.5.2. Summary
32 Increased bronchial reactivity is a key event in the toxicity pathway of O3. It can be both a
33 rapidly occurring and persistent response, although adaptation can also occur during multi-day
34 exposures. Both direct effects on smooth muscle and neurally-mediated effects on smooth muscle
35 have been proposed to contribute to AHR following O3 exposure. Currently, more evidence has
36 accumulated for the latter mechanism. In humans exposed to O3, atropine was found to block the
March 2011 5-44 DRAFT - DO NOT CITE OR QUOTE
-------
1 early AHR response indicating the involvement of cholinergic postganglionic pathways. Inhibition
2 of arachidonic acid metabolism was ineffective in blocking this response in humans while mixed
3 results were found in animal models. Studies in O3-exposed animals have demonstrated a role for SP
4 release from bronchial C fibers in mediating neurally-mediated effects on smooth muscle. Later
5 phases of increased bronchial reactivity may involve the induction of IL-lbeta which in turn
6 upregulates SP production. In guinea pigs, eosinophil-derived major basic protein contributed to the
7 stimulation of cholinergic postganglionic pathways. A novel role for hyaluronan in mediating the
8 later phase effects of O3 has recently been demonstrated. High molecular weight polymers of
9 hyaluronan normally found in the ELF were degraded following O3 exposure in mice. The resulting
10 hyaluronan fragments stimulated AHR in a TLR4- and CD44 receptor-dependent manner. Previous
11 work has shown that O3-mediated increases in lung permeability required a functioning TLR4
12 suggesting a possible relationship between increased epithelial permeability and AHR in this model.
13 Other cytokines and chemokines have been implicated in the AHR response to O3 in animals models.
5.2.6. Exacerbation and Induction of Asthma and Allergic Responses
14 In individuals with asthma, there is increased responsiveness to bronchoconstrictor challenge.
15 This results from a combination of structural and physiological factors including increased inner-
16 wall thickness, smooth muscle responsiveness and mucus secretion. Although inflammation is likely
17 to contribute, its relationship to AHR is not clear (U.S. EPA, 2006, 088089). However, some
18 asthmatics have higher baseline levels of neutrophils, lymphocytes, eosinophils and mast cells in
19 bronchial washes and bronchial biopsy tissue (Stenfors et al., 2002, 030473). Evidence is
20 accumulating that O3 exposure exacerbates asthmatic and allergic responses in sensitive individuals.
21 Further, some studies suggest that O3 exposure leads to the development of asthmatic and allergic
22 responses.
23 In order to determine whether asthmatics exhibit greater sensitivity to O3, several older studies
24 compared pulmonary function responses in asthmatic and non-asthmatic subjects following O3
25 exposure. While the majority focused on measurements of FEVi and FVC and found no differences
26 between the two groups (Holz et al., 1999, 058731: Koenig et al., 1987, 041521: Linn et al., 1978,
27 038874: Mudway et al., 2001, 025327: Scannell et al., 1996, 080755: Stenfors et al., 2002, 030473).
28 there were notable exceptions. In one study, greater airways obstruction in asthmatics compared with
29 non-asthmatic subjects was observed immediately following a 2-h exposure to 0.4 ppm O3 with
30 intermittent exercise (Kreit et al., 1989, 041817). These changes were measured as statistically
31 significant greater decreases in FEVi and FEF25_75 in the absence of a bronchoconstrictor challenge
32 (Kreit et al., 1989, 041817). These results suggest that this group of asthmatics responded to
33 O3-exposure with a greater degree of vagally-mediated bronchoconstriction compared with the non-
34 asthmatics. A second study demonstrated a statistically significant greater decrease in FEVi and
35 FEVi/FVC in asthmatics compared with non-asthmatics exposed to 0.12 ppm O3 for 7.6 hours with
36 light exercise (Horstman et al., 1995, 075834). These responses were accompanied by wheezing and
March 2011 5-45 DRAFT - DO NOT CITE OR QUOTE
-------
1 inhaler use in the asthmatics (Horstman et al., 1995, 075834). Aerosol bolus dispersion
2 measurements demonstrated a statistically significant greater change in asthmatics compared with
3 non-asthmatics which was suggestive of O3-induced small airway dysfunction (Horstman et al.,
4 1995, 075834). Furthermore, a statistically significant correlation was observed between the degree
5 of baseline airway status and the FEVi response to O3 in the asthmatic subjects (Horstman et al.,
6 1995, 075834). A third study found similar decreases in FVC and FEVi in both asthmatics and non-
7 asthmatics exposed to 0.4 ppm O3 for 2 hours with mild exercise (Alexis et al., 2000, 013072).
8 However, a statistically significant decrease in FEF75, a measure of small airway function, was
9 observed in asthmatics but not in non-asthmatics (Alexis et al., 2000, 013072). Taken together these
10 latter studies indicate that while the magnitude of restrictive type spirometric decline was similar in
11 asthmatics and non-asthmatics, that obstructive type changes (i.e. bronchoconstriction) were greater
12 in asthmatics. Further, asthmatics exhibited greater sensitivity to O3 in terms of small airways
13 function.
14 Since asthma exacerbations occur in response to allergens and/or other triggers, some studies
15 have focused on O3-induced changes in AHR following a bronchoconstrictor challenge. No
16 difference in sensitivity to methacholine bronchoprovocation was observed between asthmatics and
17 non-asthmatics exposed to 0.4 ppm O3 for 2 hours with moderate exercise (Kreit et al., 1989,
18 041817). However, increased bronchial reactivity to inhaled allergens was demonstrated in mild
19 allergic asthmatics (Torres et al., 1996, 078122: Kehrl et al., 1999, 022101: Molfino et al., 1991,
20 042379) and in allergen-sensitized guinea pigs following O3 exposure (Sun et al., 1997, 082724).
21 Similar, but modest, responses were reported for individuals with allergic rhinitis (Torres et al., 1996,
22 078122). Further, the contractile response of isolated airways was increased by pre-exposure to O3 in
23 human subjects sensitized and challenged with allergen (Roux et al., 1999, 001264).
24 In terms of airways neutrophilia, larger responses were observed in asthmatics compared to
25 non-asthmatics subjects exposed to O3 in some (Balmes et al., 1997, 086092: Basha et al., 1994,
26 075950: Scannell et al., 1996, 080755) but not all (Mudway et al., 2001, 025327) of the older
27 studies. Further, statistically significantly increases in myeloperoxidase levels (an indicator of
28 neutrophil activation) in bronchial washes was observed in mild asthmatics compared with non-
29 asthmatics, despite no difference in O3-stimulated neutrophil influx between the 2 groups following
30 exposure to 0.2 ppm O3 for 2 hours with mild exercise (Stenfors et al., 2002, 030473).
31 Eosinophils and associated proteins are thought to affect muscarinic cholinergic receptors
32 which are involved in vagally-mediated bronchoconstriction (Mudway and Kelly, 2000, 010452).
33 Studies described in Section 5.2.5.1 which demonstrated a key role of eosinophils in O3-mediated
34 AHR may not be entirely relevant to humans given the large numbers of eosinophils normally
35 present in guinea pig airways compared with humans (Yost et al., 2005, 597549). However airways
36 eosinophilia often occurs in human allergic airways disease, suggesting greater plausibility of this
37 mechanism in allergic asthmatics. Furthermore, O3 exposure sometimes often results in airways
38 eosinophilia in allergic subjects or animal models. For example, eosinophilia of the nasal and other
March 2011 5-46 DRAFT - DO NOT CITE OR QUOTE
-------
1 airways was observed in individuals with preexisting allergic disease following O3 inhalation (Peden
2 et al, 1995, 076189: Vagaggini et al, 2002, 035191). Further, O3 exposure increased allergic
3 responses, such as eosinophilia and augmented intraepithelial mucosubstances, in the nasal airways
4 of ovalbumin (OVA)-sensitized rats (Wagner et al., 2002, 026079). In contrast, Stenfors (2002,
5 030473) found no stimulation of eosinophil influx measured in bronchial washes and BAL fluid of
6 mild asthmatics following exposure to a lower concentration of O3.
7 The role of mast cells in O3-mediated asthma exacerbations has been investigated. Mast cells
8 are thought to play a key role in O3-induced airways inflammation, since airways neutrophilia was
9 decreased in mast cell-deficient mice exposed to O3 (Kleeberger et al., 1993, 044203). However,
10 another study found that mast cells were not involved in the development of increased bronchial
11 reactivity in O3-exposed mice (Noviski et al., 1999, 001198). Nonetheless, mast cells release a wide
12 variety of important inflammatory mediators which may lead to asthma exacerbations (Stenfors et
13 al., 2002, 030473). A large increase in mast cell number in bronchial submucosa was observed in
14 non-asthmatics and a significant decrease in mast cell number in bronchial epithelium was observed
15 in mild asthmatics 6 hours following exposure to 0.2 ppm O3 for 2 hours during mild exercise
16 (Stenfors et al., 2002, 030473). While these results point to an O3-mediated flux in bronchial mast
17 cell populations which differed between the non-asthmatics and mild asthmatics, interpretation of
18 these findings is difficult. Furthermore, mast cell number did not change in airway lavages in either
19 group in response to O3 (Stenfors et al., 2002, 030473)
20 Cytokine profiles in the airways have been investigated as an indicator of O3 sensitivity.
21 Differences in epithelial cytokine expression were observed in bronchial biopsy samples in non-
22 asthmatic and asthmatic subjects both at baseline and 6 h postexposure to 0.2 ppm O3 for 2 hours
23 (Bosson et al., 2003, 051687). The asthmatic subjects had a higher baseline expression of IL-4 and
24 IL-5 compared to non-asthmatics. In addition, expression of IL-5, IL-8, GM-CSF, and ENA-78 in
25 asthmatics was increased significantly following O3 exposure compared to non-asthmatics (Bosson
26 et al., 2003, 051687). Some of these (IL-4, IL-5 and GM-CSF) are T helper type 2 (Th2)-related
27 cytokines or neutrophil chemoattractants, and play a role in IgE production, airway eosinophilia and
28 suppression of Thl-cytokine production (Bosson et al., 2003, 051687). These findings suggest a link
29 between adaptive immunity and enhanced sensitivity of asthmatics to O3.
5.2.6.1. New Cellular and Molecular Insights
30 Since asthmatics may have enhanced sensitivity to O3, one recent study investigated whether
31 O3 exposure exacerbated asthmatic responses in persistent asthmatics. These subjects had been
32 therapeutically treated with inhaled corticosteroids for several months prior to the study. Exposure of
33 these subjects to 0.2 ppm O3 for 2 hours with mild exercise resulted in decrements in FEVi and FVC
34 and an increase in sRaw immediately postexposure (Stenfors et al., 2010, 386512). In addition, large
35 increases in neutrophil number and myeloperoxidase levels (an indicator of neutrophil activation) in
36 airway lavages and in mast cell number in bronchial submucosa tissue obtained by biopsy was
March 2011 5-47 DRAFT - DO NOT CITE OR QUOTE
-------
1 observed 18 hours postexposure (Stenfors et al., 2010, 386512). No change in bronchial wash or
2 BAL fluid mast cell number or eosinophil number in any compartment was observed. (Stenfors et
3 al., 2010, 386512). These results suggest that some standard treatments for asthma may not protect
4 against the effects of environmental O3 and that this group may be particularly prone to asthma
5 exacerbations given that airways neutrophilia occurred despite the inhaled corticosteroid treatment
6 (Stenfors et al., 2010, 386512).
7 Other recent studies in humans and animals provide evidence that O3 causes the development
8 of allergic responses and/or asthma. Several of these suggested that O3-mediated activation of innate
9 immunity contributes to O3-induced, as well as to O3-amplified, allergic responses. In one study
10 involving human subjects, the authors hypothesized that O3 exposure would result in recruitment of
11 activated innate immune cells to the airways. Healthy individuals were exposed to 0.08 ppm O3 for
12 6.6 hours with intermittent exercise and airways inflammation was characterized in induced sputum
13 18 h postexposure (Alexis et al., 2010, 628538). Previous studies demonstrated that induced sputum
14 contains liquid and cellular constituents of the lining fluid from central conducting airways (Alexis
15 et al., 2001, 190013) and also identified these airways as a site of preferential O3 absorption during
16 exercise (Hu et al., 1994, 041323). Ozone exposure resulted in sputum neutrophilia and increased
17 numbers of airways monocytes and dendritic-like cells (Alexis et al., 2010, 628538). In addition,
18 increased expression of cell surface markers characteristic of innate immunity and antigen
19 presentation (i.e. CD-14 and HLA-DR) was demonstrated on airways monocytes (Alexis et al., 2010,
20 628538). Enhanced antigen presentation contributes to exaggerated T cell responses and promotes
21 Th2 inflammation and an allergic phenotype (Lay et al., 2007, 196610). Upregulation of pro-
22 inflammatory cytokines in sputum was also demonstrated in O3-exposed subjects (Alexis et al.,
23 2010, 628538). One of these cytokines, IL-12p70, correlated with numbers of dendritic-like cells in
24 the sputum, and is an indicator of dendritic cell activation (Alexis et al., 2010, 628538). These
25 authors have previously reported that O3 activates monocytes and macrophages (Lay et al., 2007,
26 196610) which could play a role in exacerbating existing asthma by activating allergen-specific
27 memory T-cells. The current study confirms these findings and extends them by suggesting a
28 potential mechanism whereby O3-activated dendritic cells could stimulate naive T-cells to promote
29 the development of asthma (Alexis et al., 2010, 628538). A companion study by these same
30 investigators (described in detail in Section 5.2.9.1) provides evidence of dendritic cell activation,
31 measured as increased expression of HLA-DR, in a subset of the human subjects (GSTM1 null)
32 exposed to 0.4 ppm O3 for 2 hours with intermittent exercise (Alexis et al., 2009, 628542)
33 Another recent study demonstrated O3-mediated activation of the innate immune system and
34 linked it to the development of non-specific AHR in a mouse model (Pichavant et al., 2008, 596409).
35 Repeated exposure to 1 ppm O3 for 3 hours induced non-specific AHR measured 24 hours following
36 the last exposure (Pichavant et al., 2008, 596409). This response was found to require NKT cells,
37 which are effector lymphocytes of innate immunity, as well as IL-17 and airways neutrophilia
38 (Pichavant et al., 2008, 596409). Since glycolipids such as galactosyl ceramide are ligands for the
March 2011 5-48 DRAFT - DO NOT CITE OR QUOTE
-------
1 invariant CD1 receptor on NKT cells and serve as endogenous activators of NKT cells, a role for
2 O3-oxidized lipids in activating NKT cells was proposed (Pichavant et al., 2008, 596409). The
3 authors contrasted this innate immunity pathway with that of allergen-provoked specific AHR which
4 involves adaptive immunity, the cytokines IL-4, IL -13, IL-17, and airways eosinophilia (Pichavant
5 et al., 2008, 596409). Interestingly, NKT cells were required for both the specific AHR provoked by
6 allergen and the non-specific AHR provoked by O3 (Pichavant et al., 2008, 596409). Different
7 cytokine profiles of the NKT cells from allergen and O3-exposed mice in mediating was proposed to
8 account for the different pathways (Pichavant et al., 2008, 596409). More recently, NKT cells have
9 been found to function in both innate and adaptive immunity (Vivier et al., 2011, 676697).
10 Priming of the innate immune system by O3 was reported by Hollingsworth et al., (2007,
11 597609). In this study, exposure of mice to 2 ppm O3 for 3 hours led to nonspecific AHR at 24 and
12 48 hours postexposure, an effect which subsided by 72 hours (Hollingsworth et al., 2007, 597609).
13 However in mice treated with aerosolized endotoxin immediately following O3 exposure, AHR was
14 greatly enhanced at 48 and 72 hours postexposure (Hollingsworth et al., 2007, 597609). Ozone pre-
15 exposure was found to reduce the number of inflammatory cells, to increase cytokine production and
16 total protein in the BAL fluid and to increase systemic IL-6 following exposure to endotoxin
17 (Hollingsworth et al., 2007, 597609). Furthermore, O3 stimulated the apoptosis of alveolar
18 macrophages 24 hours postexposure, an effect which was greatly enhanced by endotoxin treatment.
19 Apoptosis of blood monocytes was also observed in response to the combined exposures
20 (Hollingsworth et al., 2007, 597609). Ozone pre-exposure enhanced the response of lung
21 macrophages to endotoxin (Hollingsworth et al., 2007, 597609). Taken together, these findings
22 demonstrated that O3 exposure increased innate immune responsiveness to endotoxin. The authors
23 proposed that this effect was mediated by TLR4-dependent pathways since O3 increased surface
24 expression of TLR4 on macrophages (Hollingsworth et al., 2007, 597609). More recently, these
25 authors demonstrated that hyaluronan contributed to the O3-primed response to endotoxin (Li et al.,
26 2010, 670282). In this study, exposure of mice to 1 ppm O3 for 3 h resulted in enhanced responses to
27 endotoxin, which was mimicked by intratracheal instillation of hyaluronan fragments (Li et al., 2010,
28 670282). Hyaluronan, like O3, was also found to induce TLR4 receptor peripheralization in the
29 membrane (Hollingsworth et al., 2007, 597609: Li et al., 2010, 670282). an effect which is
30 associated with enhanced responses to endotoxin. This study and previous ones by the same
31 investigators showed elevation of BAL hyaluronan in response to O3 exposure (Garantziotis et al.,
32 2009, 597603: Garantziotis et al., 2010, 624947: Li et al., 2010, 670282). providing evidence that
33 ozone's effects on innate immunity are at least in part mediated by hyaluronan fragments. The
34 authors note that excessive TLR4 signaling can lead to lung injury and suggest that O3 may be
35 responsible for an exaggerated innate immune response which may underlie lung injury and
36 decreased host defense (Li et al., 2010, 670282) (Section 5.2.7.1).
37 An interaction between allergen and O3 in the induction of nonspecific AHR was shown in
38 another animal study (Larsen et al., 2010, 628560). Mice were sensitized with the aerosolized
March 2011 5-49 DRAFT - DO NOT CITE OR QUOTE
-------
1 allergen OVA on 10 consecutive days followed by exposure to O3 (0.1-0.5 ppm for 3 hours) (Larsen
2 et al., 2010, 628560). While allergen sensitization alone did not alter airway responsiveness to a
3 nonspecific challenge, O3 exposure of sensitized mice resulted in nonspecific AHR at 6 and 24 hours
4 postexposure (Larsen et al., 2010, 628560). The effects of O3 on AHR were independent of airways
5 eosinophilia and neutrophilia (Larsen et al., 2010, 628560). However, OVA pretreatment led to
6 goblet cell metaplasia which was enhanced by O3 exposure (Larsen et al., 2010, 628560). It should
7 be noted that OVA sensitization using only aerosolized antigen in this study is less common than the
8 usual procedure for OVA sensitization achieved by one or more initial systemic injections of OVA
9 and adjuvant followed by repeated inhalation exposure to OVA.
10 Furthermore, O3 was found to act as an adjuvant for allergic sensitization (Hollingsworth et
11 al., 2010, 635786). In this study in mice, oropharyngeal aspiration of OVA on day 0 and day 6 failed
12 to lead to allergic sensitization unless mice were first exposed to 1 ppm O3 for 2 hours
13 (Hollingsworth et al., 2010, 635786). The O3-mediated response involved Th2 (IL-4, IL-5 and IL-9)
14 and Thl7 cytokines (IL-17) and was dependent on a functioning TLR4 (Hollingsworth et al., 2010,
15 635786). Ozone exposure also activated OVA-bearing dendritic cells in the thoracic lymph nodes, as
16 measured by the presence of the CD86 surface marker, which suggests naive T-cell stimulation and
17 the involvement of Th2 pathways (Hollingsworth et al., 2010, 635786). Thus ozone's adjuvant
18 effects may be due to activation of both innate and adaptive immunity.
19 Results of recent studies demonstrate participation of innate immune pathways and also
20 suggest involvement of adaptive immune pathways in both the induction and the exacerbation of
21 allergic responses and AHR by O3.
5.2.6.2. Summary
22 Collectively these older and more recent studies provide insight into ozone's ability to provoke
23 asthma exacerbations in humans. Greater airways inflammation and/or greater bronchial reactivity
24 have been demonstrated in asthmatics compared to non-asthmatics. This pre-existing inflammation
25 and altered baseline bronchial reactivity may contribute to the enhanced bronchoconstriction seen in
26 asthmatics exposed to O3. Furthermore, inflammation may contribute to O3-mediated AHR. Animal
27 studies have demonstrated a role for eosinophil-derived proteins in mediating these effects. Since
28 airways eosinophilia occurs in both allergic humans and allergic animal models, this pathway may
29 underlie the exacerbation of allergic asthma by O3. In addition, differences have been noted in
30 epithelial cytokine expression in bronchial biopsy samples of healthy and asthmatic subjects. A Th2
31 phenotype, indicative of adaptive immune system activation and enhanced allergic responses, was
32 observed before O3 exposure and was increased by O3 exposure in asthmatics. Since eosinophilia is a
33 hallmark of a Th2 phenotype, these findings support links between allergic asthma, sensitivity to O3
34 and adaptive immunity. Studies in humans and animal models also provide evidence for activation of
35 innate immunity by O3. In humans, O3 exposure resulted in increased numbers of airways monocytes
36 and dendritic-like cells. Altered expression of cell surface markers characteristic of innate immunity
March 2011 5-50 DRAFT - DO NOT CITE OR QUOTE
-------
1 and antigen presentation was observed on monocytes and macrophages. Recruitment of these
2 activated immune cells could lead to activation of allergen-specific memory T-cells in allergic
3 individuals and result in the exacerbation of existing asthma in response to an allergen trigger. In
4 animal studies, O3 exposure primed the innate immune system and led to increased endotoxin-
5 induced AHR by a mechanism involving hyaluronan and TLR4. The exaggerated immune response
6 to O3 + endotoxin could lead to a more pronounced lung injury response to a bacterial trigger.
7 Enhanced bronchial reactivity, airways eosinophilia, Th2 phenotype, recruitment of activated innate
8 immune cells, and enhanced responsiveness to endotoxin all provide biological plausibility for
9 epidemiologic evidence of asthma exacerbations associated with exposure to O3. Thus, the influx of
10 immunomodulatory cells and the activation of innate and adaptive immunity lead to the exacerbation
11 of asthma and allergic responses which is emerging as a key event in the toxicity pathway of O3.
12 Recent studies in humans and animal models also provide evidence that O3 exposure causes
13 induction of AHR and allergic responses. Both activation of innate immunity and promotion of
14 adaptive immunity have been implicated. In humans, O3 exposure resulted in increased numbers of
15 dendritic-like cells and levels of a cytokine associated with dendritic cell activation in the sputum,
16 suggesting the presence of a population of activated dendritic cells which could stimulate naive
17 T-cells to promote the development of asthma. Evidence for activated dendritic cells was also found
18 in GSTM1 null human subjects (Section 5.2.9.1) and in allergen-sensitized animals exposed to O3. In
19 the latter study, O3 acted as an adjuvant for allergic sensitization and the development of AHR by a
20 mechanism involving TLR4. In a different animal model, O3-induced AHR required the presence of
21 NKT cells and IL-17, both of which indicate innate immune system activation. Ozone-induced
22 goblet cell metaplasia has also been demonstrated. These findings suggest that O3 may be capable of
23 causing new onset asthma and allergic responses in humans. Thus, promotion of adaptive immunity
24 and activation of innate immunity leads to the induction of AHR and allergic responses which is
25 emerging as a key event in the toxicity pathway of O3.
5.2.7. Impaired Host Defense
26 O3 impacts host defense by a variety of different mechanisms. Animal models have
27 demonstrated decreased mucociliary particle clearance and effects on alveolar macrophages,
28 including inhibited phagocytosis and production of reactive oxygen intermediates, and altered
29 chemotaxis and adhesion in response to O3 exposure (U.S. EPA, 2006, 088089). Ozone has been
30 shown to target SP-A, resulting in a decrease in its function (U.S. EPA, 2006, 088089). In addition,
31 reduced clearance of bacterial pathogens and enhanced susceptibility to bacterial lung infections
32 were observed in rodents exposed acutely to O3 (Gilmour et al., 1993, 039620; U.S. EPA, 2006,
33 088089). Further, O3-induced alterations in immune function have been demonstrated (Jakab et al.,
34 1995, 039548: U.S. EPA, 2006, 088089).
March 2011 5-51 DRAFT - DO NOT CITE OR QUOTE
-------
5.2.7.1. New Cellular and Molecular Insights
1 As described above, priming of the innate immune system by O3 was reported by
2 Hollingsworth et al. (2007, 597609). Besides effects on AHR, exposure of mice to 2 ppm O3 for
3 3 hours reduced inflammatory cell influx in the airways in response to endotoxin (Hollingsworth et
4 al., 2007, 597609). In addition, O3 exposure stimulated the apoptosis of alveolar macrophages
5 24 hours postexposure, an effect which was enhanced by a subsequent treatment with endotoxin
6 (Hollingsworth et al., 2007, 597609). Apoptosis of circulating blood monocytes was also observed in
7 response to O3 and endotoxin (Hollingsworth et al., 2007, 597609). The authors attributed these
8 effects to the increased surface expression of TLR4 and increased signaling in macrophages
9 observed in the study (Hollingsworth et al., 2007, 597609). It was proposed that the resulting
10 decrease in airways inflammatory cells could account for O3-mediated decreased clearance of
11 bacterial pathogens observed in numerous animal models (Hollingsworth et al., 2007, 597609). A
12 more recent study by these investigators provided evidence that hyaluronan mediates O3-priming of
13 innate immunity and suggested that exaggerated innate immune responses may underlie lung injury
14 and decreased host defense (Li et al., 2010, 670282).
15 Recent studies also demonstrated SP-A oxidation by O3. SP-A is an important innate immune
16 protein which plays a number of roles in host defense including acting as opsonin for the recognition
17 of some pathogens (Haque et al., 2009, 200767). These investigations demonstrated that O3-
18 mediated carbonylation of SP-A was associated with impaired macrophage phagocytosis in vitro
19 (Mikerov et al., 2008, 596405). Furthermore, O3 exposure in mice was found to increase
20 susceptibility to pneumonia infection in mice through an impairment of SP-A dependent
21 phagocytosis (Mikerov et al., 2008, 201537: Mikerov et al., 2008, 597493).
22 Another recent study demonstrated impaired antigen-specific immunity following subacute O3
23 exposure (0.6 ppm, 10 h/day for 15 days) in mice (Feng et al., 2006, 596381). Specifically, O3
24 exposure altered the lymphocyte subset and cytokine profile and impacted thymocyte early
25 development leading to immune dysfunction.
5.2.7.2. Summary
26 Collectively these older and more recent studies in animal models provide several mechanisms
27 by which O3 exposure could enhance susceptibility to lung infections. Both decreased mucociliary
28 particle clearance and decreased numbers and function of alveolar macrophage have been
29 implicated. Recent studies suggest that O3-mediated oxidation of SP-A oxidation and priming of the
30 innate immune system may contribute to decreased pathogen clearance. Immune dysfunction outside
31 of the lung has also been demonstrated. Thus, immune system modulation resulting in impaired host
32 defense is emerging as a key event in ozone's toxicity pathway.
March 2011 5-52 DRAFT - DO NOT CITE OR QUOTE
-------
5.2.8. Extrapulmonary Effects
1 Extrapulmonary effects of O3 have been noted for decades (U.S. EPA, 2006, 088089). One
2 such effect is hypothermia, which in rodents occurs subsequent to the activation of neural reflexes
3 and involves the parasympathetic nervous system (Watkinson et al., 2001, 016245). Other
4 mechanisms are likely to be involved in extrapulmonary effects. It has been proposed that lipid
5 oxidation products resulting from reaction of O3 with lipids in the ELF are responsible for systemic
6 effects, however it is not known whether they gain access to the vascular space (Chuang et al., 2009,
7 197202). Alternatively, extrapulmonary release of diffusible mediators may initiate or propagate
8 inflammatory responses in the vascular or in systemic compartments (Cole and Freeman, 2009,
9 597507). For example, one such mediator, the cytokine IL-6, is known to have pleitropic effects such
10 as inducing the acute phase response, activating the hypothalamus-pituitary-adrenal axis and altering
11 serum cholesterol levels (Tarrant, 2010, 644810).
5.2.8.1. Cardiovascular Effects
12 Effects of O3 on the cardiovascular system have been demonstrated in both humans and animal
13 models (U.S. EPA, 2006, 088089). Several mechanisms have been proposed to account for these
14 responses (Perepu et al., 2010, 385020). First, O3 may impair alveolar-arterial oxygen transfer and
15 reduce the supply of arterial oxygen to the myocardium. This may have a greater impact in
16 individuals with compromised cardiopulmonary systems. Gong et al. (1998, 029938) provided
17 evidence of a small decrease in arterial oxygen saturation in human subjects exposed to O3. In
18 addition, Delaunois et al. (1998, 015779) demonstrated pulmonary vasoconstriction in O3-exposed
19 rabbits. Although of interest, the contribution of this pathway to O3-induced cardiovascular effects
20 remains uncertain. Secondly, O3 may trigger neural reflexes which stimulate the autonomic nervous
21 system and alter electrophysiologic responses of the heart. For example, bradycardia, altered HRV
22 and arrhythmia have been demonstrated in animals exposed to O3 (Arito et al., 1990, 042285;
23 Hamade and Tankersley, 2009, 596386: Watkinson et al., 2001, 016245). Third, O3-induced
24 pulmonary inflammation may lead to inflammatory or injury responses in the cardiovascular system
25 (Cole and Freeman, 2009, 597507).
26 Some recent studies have suggested that O3 may alter the systemic vasculature. In resting
27 humans, exposure to fine particulate matter (PM) + O3 resulted in arterial vasoconstriction and
28 increased diastolic blood pressure during the 2-h exposure (Brook et al., 2002, 024987). However, a
29 recent study in humans observed no changes in HRV or blood pressure in healthy nonsmokers
30 exposed at rest to 0.12 ppm O3 for 2 hours (Brook et al., 2009, 195611). Thus, the previously
31 observed effects of the combined O3 and PM exposure may have been attributable to PM alone. A
32 role for O3 in modulating endothelin, a potent vasoconstrictor, has also been proposed. Studies in
33 animals found that O3-induced endothelin system genes in the lung and increased circulating levels
34 of endothelin (Thomson et al., 2005, 087554: Thomson et al., 2006, 097483).
March 2011 5-53 DRAFT - DO NOT CITE OR QUOTE
-------
Recent Cellular and Molecular Insights
1 Ozone-induced perturbations of the cardiovascular system were recently investigated in young
2 mice and monkeys (Chuang et al., 2009, 197202). Young mice exposed to 0.5 ppm O3 for 5 days
3 demonstrated increased heart rate and blood pressure. Decreases in endothelial-dependent
4 vasorelaxation and NO homeostasis were observed in arterial tissue. Oxidative and nitrosative stress
5 were demonstrated in lung and aortic tissue following O3 exposure. Mitochondrial DNA was
6 damaged in lung and aortas from young mice and infant rhesus monkeys similarly exposed to O3.
7 Since altered NO homeostasis, mitochondrial DNA damage and oxidative stress are known to
8 contribute to the development of atherosclerosis, young Apo E null mice were exposed
9 subchronically to O3 (0.5 ppm O3 for 5 days/week over 8 weeks). Ozone exposure augmented the
10 aortic lesion areas compared with controls exposed to filtered air, suggesting an acceleration of
11 atherogenesis. This is the first study to definitively identify the systemic vasculature as a target of
12 O3-induced effects.
13 A second recent study in animals demonstrated effects on the heart due to chronic O3 exposure
14 (Perepu et al., 2010, 385020). Rats were exposed to 0.8 ppm O3 for 28 and 56 days and isolated
15 hearts were subjected to ischemia-reperfusion injury. Ozone exposure enhanced the sensitivity to
16 injury in this model, as demonstrated by decreased cardiac function compared with control rats that
17 were exposed to filtered air. Further, markers of lipid peroxidation and inflammation were greater in
18 the hearts of O3-exposed rats. This study is the first to definitively identify the heart as a target for
19 O3-induced effects.
20 Further evidence for O3-induced effects in the systemic vasculature and heart is provided by a
21 recent chronic study in rats (Kodavanti et al., In Press, 666323). Episodic exposure to 0.4 ppm O3 for
22 16 weeks (5 h/day for 1 day/week) resulted in increased aortic levels of mRNA for biomarkers of
23 oxidative stress, thrombosis, vasoconstriction and proteolysis. Ozone exposure also increased lectin-
24 like oxidized-low density lipoprotein receptor-l(LOX-l) mRNA and protein levels in the aorta.
25 Depletion of cardiac mitochondrial phospholipid fatty acids was also observed. Taken together, these
26 results suggest a role for circulating oxidized lipids in mediating the effects of O3.
5.2.8.2. Hepatic Effects
27 Changes in hepatic gene expression have also been noted following O3 exposure. Specifically,
28 downregulation of gene families related to lipid, fatty acid and carbohydrate metabolism was
29 demonstrated in the liver of O3-exposed mice (Last et al., 2005, 596400). In addition, transcription of
30 enzymes involved in xenobiotic metabolism was decreased. Impairment of hepatic drug metabolism
31 was suggested in an older study which found that mice exposed to O3 had prolonged pentobarbital
32 sleeping time (Graham et al., 1985, 040289). Further evidence of hepatic effects is provided by a
33 recent study in which exposure to 0.25 and 0.5 ppm O3 for 6 hours resulted in exacerbation of drug-
34 induced liver injury in mice pre-exposed to acetaminophen. This included a greater increase in
March 2011 5-54 DRAFT - DO NOT CITE OR QUOTE
-------
1 hepatic neutrophil accumulation and greater alteration in gene expression profiles in mice exposed to
2 O3 and acetaminophen compared with either exposure alone (Aibo et al., 2010, 378559).
5.2.8.3. Summary
3 Collectively, these older and more recent studies in animal models provide evidence for
4 extrapulmonary effects of O3. Although it was suggested that these effects are directly mediated by
5 secondary oxidation products formed in the lung as a result of O3 exposure, there is no evidence that
6 these species enter the circulation. Alternatively, extrapulmonary effects may be due to activation of
7 neural reflexes or to release of diffusible mediators which may initiate or propagate inflammatory
8 responses in the vascular or systemic compartments. Recent studies suggest that oxidative/nitrosative
9 stress contributes to O3-induced cardiovascular effects. Thus, systemic inflammation and vascular
10 oxidative/nitrosative stress are emerging as key events in the toxicity pathway of O3.
5.2.9. Factors Affecting Responses to Ozone
11 Responses to O3 are variable within the population and the basis for this variability is not clear
12 (Mudway and Kelly, 2000, 010452). Research has focused on the role of gene-environment
13 interactions, preexisting conditions, adaptive mechanisms and lifestage in influencing the responses
14 to O3. Co-exposure to other pollutants has also been considered.
5.2.9.1. Gene-Environment Interactions
15 The significant inter-individual variation in responses to O3 infers that genetic background is
16 an important determinant of susceptibility to O3 (Cho and Kleeberger, 2007, 195616; Kleeberger et
17 al., 1997, 095736). Strains of mice which are prone or resistant to O3-induced effects have been used
18 to systematically identify candidate susceptibility genes. Genome wide linkage analyses (also known
19 as positional cloning) demonstrated quantitative trait loci for O3-induced lung inflammation and
20 hyperpermeability on chromosome 17 (Kleeberger et al., 1997, 095736) and chromosome 4
21 (Kleeberger et al., 2000, 014895). respectively, using these recombinant inbred strains of mice. More
22 specifically these studies found that Tnf, whose protein product is the inflammatory cytokine TNF-a,
23 and Tlr4, whose protein product is TLR4, were candidate susceptibility genes (Kleeberger et al.,
24 1997, 095736; Kleeberger et al., 2000, 014895). Other investigations in inbred mouse strains found
25 that differences in expression of certain proteins, such as CCSP (Broeckaert et al., 2003, 055490) and
26 MARCO (Dahl et al., 2007, 196986). are responsible for phenotypic characteristics, such as
27 epithelial permeability and scavenging of oxidized lipids, respectively, which confer sensitivity to
28 03.
29 Genetic polymorphisms have received increasing attention as modulators of O3-mediated
30 effects. Functionally relevant polymorphisms in candidate susceptibility genes have been studied at
31 the individual and population level in humans and also in animal models. Genes whose protein
March 2011 5-55 DRAFT - DO NOT CITE OR QUOTE
-------
1 products are involved in antioxidant defense/oxidative stress and xenobiotic metabolism, such as
2 glutathione-S-transferase Ml (GSTM1) andNADPH:quinone oxidoreductase 1 (NQO1), have also
3 been major focuses of these efforts. This is because oxidative stress resulting from O3 exposure is
4 thought to contribute to the pathogenesis of asthma and because xenobiotic metabolism detoxifies
5 secondary oxidation products formed by O3 which contribute to oxidative stress (Islam et al., 2008,
6 097348). TNF-a is of interest since it is linked to a candidate O3 susceptibility gene and since it plays
7 a key role in initiating airways inflammation (Li et al., 2006, 090972). Polymorphisms of genes
8 coding for GST Ml, NQO1 and TNF-a have been associated with altered susceptibility to O3-
9 mediated effects (Bergamaschi et al., 2001, 052670: Corradi et al., 2002, 035448: Li et al., 2006,
10 090972: Romieu et al., 2004, 056796: Yang et al., 2005, 077211). Additional studies have focused on
11 functional variants in other genes involved in antioxidant defense such as catalase (CAT),
12 myeloperoxidase, heme oxygenase (HMOX-1) and manganese superoxide dismutase (MnSOD)
13 (Islam et al., 2008, 097348: Wenten et al., 2009, 597084). These studies are discussed below.
14 GSTM1 is a phase II antioxidant enzyme which is transcriptionally regulated by NF-E2-
15 related factor 2-antioxidant response element (Nrf2-ARE) pathway. A large proportion (40-50%) of
16 the general public (across ethnic populations) has the GSTM1 null genotype, which has been linked
17 to an increased risk of adverse health effects due to exposure to air pollutants (London, 2007,
18 093279). A role for GSTs in metabolizing electrophiles such as 4-hydroxynonenal, which is a
19 secondary oxidation product formed following O3 exposure, has been demonstrated (Awasthi et al.,
20 2004, 644649). A recent study found that the GSTM1 genotype modulated the time course of the
21 neutrophilic inflammatory response following acute O3 exposure (0.4 ppm for 2 hours with
22 intermittent exercise) in healthy adults (Alexis et al., 2009, 628542). In GSTMl-null and sufficient
23 subjects, O3-induced sputum neutrophilia was similar at 4 hours. However, neutrophilia resolved by
24 24 hours in sufficient subjects but not in GSTMl-null subjects. It is not known whether this effect
25 was due to the persistence of pro-inflammatory stimuli, impaired production of downregulators or
26 impaired neutrophil apoptosis and clearance. In addition, O3 exposure increased the expression of the
27 surface marker CD 14 in airway neutrophils of GSTM-1 null subjects compared with sufficient
28 subjects. Furthermore, numbers of airway macrophages were decreased at 4 and 24 hours following
29 O3 exposure in GSTM1-sufficient subjects (Alexis et al., 2009, 628542). Airways macrophages in
30 GSTM1 null subjects were greater in number and found to have greater oxidative burst and
31 phagocytic capability than those of sufficient subjects. Airways macrophages and dendritic cells
32 from GSTM1 null subjects exposed to O3 expressed higher levels of the surface marker HLA-DR,
33 suggesting activation of the innate immune system (Alexis et al., 2009, 628542). These differences
34 in inflammatory responses between the GSTM1 null and sufficient subjects may provide biological
35 plausibility for the differences in O3-mediated effects reported in controlled human exposure studies
36 (Bergamaschi et al., 2001, 052670: Corradi et al., 2002, 035448). It should also be noted that
37 GSTM1 genotype did not affect the acute pulmonary function (spirometric) response to O3 which
38 provides additional evidence for separate mechanisms underlying ozone's effects on pulmonary
March 2011 5-56 DRAFT - DO NOT CITE OR QUOTE
-------
1 function and inflammation in adults (Alexis et al., 2009, 628542). However, GSTM1 null asthmatic
2 children were previously found to be more sensitive to the effects of O3 on pulmonary function than
3 GSTM1 sufficient asthmatic children (Romieu et al., 2004, 056796).
4 NQO1 catalyzes the 2-electron reduction by NADPH of quinones to hydroquinones.
5 Depending on the substrate, it is capable of both protective detoxification reactions and redox
6 cycling reactions resulting in the generation of reactive oxygen species. A recent study using NQO1-
7 null mice demonstrated that NQO1 contributes to O3-induced oxidative stress, AHR and
8 inflammation in mice (Voynow et al., 2009, 194311). These experimental results may provide
9 biological plausibility for the increased biomarkers of oxidative stress and increased pulmonary
10 function decrements observed in O3-exposed individuals bearing both the wild-type NQO1 gene and
11 the null GSTM1 gene (Bergamaschi et al., 2001, 052670: Corradi et al., 2002, 035448).
12 Two studies reported relationships between TNF promoter variants and O3-induced effects in
13 humans. In one study, O3-induced change in lung function was significantly lower in adult subjects
14 with TNF promoter variants -308A/A and -308G/A compared with adult subjects with the variant -
15 308G/G (Yang et al., 2005, 077211). This response was modulated by a specific polymorphism of
16 LTA (Yang et al., 2005, 077211). a previously identified candidate susceptibility gene whose protein
17 product is lymphotoxin-a (Kleeberger et al., 1997, 095736). In the second study, an association
18 between the TNF promoter variant -308G/G and decreased risk of asthma and lifetime wheezing in
19 children was found (Li et al., 2006, 090972). The protective effect on wheezing was modulated by
20 ambient O3 levels and by GSTM1 and GSTP1 polymorphisms. The authors suggested that the
21 TNF-308 G/G genotype may have a protective role in the development of childhood asthma (Li et
22 al., 2006, 090972).
23 Similarly, a promoter variant of the gene HMOX-1, consisting of a smaller number of (GT)n
24 repeats, was associated with a reduced risk for new-onset asthma in non-Hispanic white children
25 (Islam et al., 2008, 097348). The number of (GT)n repeats in this promoter has been shown to be
26 inversely related to the inducibility of HMOX-1. A modulatory effect of O3 was demonstrated since
27 the beneficial effects of this polymorphism were seen only in children living in low O3 communities
28 (Islam et al., 2008, 097348). This study also identified an association between a polymorphism of the
29 CAT gene and increased risk of new-onset asthma in Hispanic children; however no modulation by
30 O3 was seen (Islam et al., 2008, 097348). No association was observed in this study between a
31 MnSOD polymorphism and asthma (Islam et al., 2008, 097348).
32 Studies to date indicate that some variability in individual responsiveness to O3 may be
33 accounted for by functional genetic polymorphisms. Further, the effects of gene-environment
34 interactions may be different in children and adults.
5.2.9.2. Preexisting Diseases and Conditions
35 Several preexisting diseases and conditions have been described which modulate the response
36 to O3 exposure. Atopy and asthma are important factors in humans. For example, asthmatics were
March 2011 5-57 DRAFT - DO NOT CITE OR QUOTE
-------
1 more susceptible to O3-mediated inflammation (Balmes et al, 1997, 086092; Basha et al, 1994,
2 075950; Scannell et al., 1996, 080755). while not exhibiting any increase in responsiveness as
3 measured by spirometry. Ozone exposure resulted in eosinophilia of the nasal and lower airways in
4 individuals with preexisting allergic disease (Peden et al., 1995, 076189; Vagaggini et al., 2002,
5 035191) and increased bronchial reactivity to challenge with inhaled specific antigen in mild allergic
6 asthmatics (Torres et al., 1996, 078122; Kehrl et al., 1999, 022101; Molfino et al., 1991, 042379).
7 Increased bronchial reactivity was also reported in individuals with allergic rhinitis (Torres et al.,
8 1996, 078122). It has been proposed that sensitivity is conferred by the presence of greater numbers
9 of resident airway inflammatory cells in disease states such as asthma (Mudway and Kelly, 2000,
10 010452).
11 In addition, smoking and COPD status are important determinants since responsiveness to O3,
12 as measured by spirometry, is decreased in individuals with these conditions (U.S. EPA, 2006,
13 088089). Furthermore, obesity may alter susceptibility. In a reanalysis of the data of Hazucha (2003,
14 048168). increasing body mass index in young women was associated with increased O3
15 responsiveness (Bennett et al., 2007, 418827). In animal models, diet-induced obesity augmented
16 inflammation and injury, as measured by BAL markers, as well as innate AHR, in mice exposed
17 acutely to O3 (Johnston et al., 2008, 597625). In contrast, the inflammatory response following sub-
18 acute exposure to O3 was dampened by obesity in a different mouse model (Shore et al., 2009,
19 201551). Finally, nutritional status may impact the response to O3. Many investigations have focused
20 on antioxidant deficiency as a modulator of O3-mediated effects (see above). Although results of
21 these studies are mixed, studies in humans demonstrate that supplementation with ascorbate and
22 alpha-tocopherol was protective against O3-induced pulmonary function deficits in healthy adults
23 who were ascorbate-deficient (Samet et al., 2001, 019034) and in asthmatic children living in
24 Mexico City (Romieu et al., 2002, 034711; Sienra-Monge et al., 2004, 196422). Furthermore,
25 supplementation with ascorbate, alpha-tocopherol and beta-carotene was found to be protective in
26 Mexico City streetworkers (Romieu et al., 1998, 086756).
5.2.9.3. Lifestage: Postnatal development
27 An interesting set of studies conducted over the last 10 years in the infant rhesus monkey has
28 identified numerous O3-mediated perturbations in the developing lung and immune system (Plopper
29 et al., 2007, 596412). These investigations were prompted by the dramatic rise in the incidence of
30 childhood asthma and focused on the possible role of O3 and allergens in promoting remodeling of
31 the epithelial-mesenchymal trophic unit during postnatal development of the tracheobronchial
32 airway wall. Rhesus monkeys were used in these studies because the branching pattern and
33 distribution of airways in this model are more similar to humans than are those of rodents to humans.
34 In addition, a model of allergic airways disease, which exhibits the main features of human asthma,
35 had already been established in the adult rhesus monkey. Studies in infant monkeys were designed to
36 determine whether repeated exposure to O3 altered postnatal growth and development, and if so,
March 2011 5-58 DRAFT - DO NOT CITE OR QUOTE
-------
1 whether such effects were reversible. In addition, exposure to O3 was evaluated for its potential to
2 increase the development of allergic airways disease. Infant rhesus monkeys (30 days old) were
3 exposed to cyclic episodic O3 over a 5-month period. This involved biweekly cycles of alternating
4 O3 (5 consecutive days of 0.5 ppm O3, 8 h/day) and filtered air (9 consecutive days). Some animals
5 were sensitized to house dust mite allergen (HDMA) and then exposed to HDMA aerosol for 2 h/day
6 on days 3-5 of either filtered air or O3 exposure.
7 Key findings are numerous. First, baseline airway resistance and AHR in the infant monkeys
8 were dramatically increased by combined exposure to both HDMA and O3 (Joad et al., 2006,
9 596390; Schelegle et al., 2003, 053778). Secondly, O3 exposure alone led to a large increase in BAL
10 eosinophils (Schelegle et al., 2003, 053778) while HDMA exposure alone led to a large increase of
11 eosinophils in airways tissue (Joad et al., 2006, 596390; Schelegle et al., 2003, 053778). Thirdly, the
12 growth pattern of distal airways was significantly changed by exposure to O3 alone and in
13 combination with HDMA. More specifically, longer and narrower airways resulted and the number
14 of conducting airway generations between the trachea and the gas exchange area was decreased
15 (Fanucchi et al., 2006, 096491). This latter effect was not ameliorated by a recovery period of
16 6 months. Fourthly, exposure to both HDMA and O3 altered the abundance and distribution of
17 CD25+ lymphocytes in the airways (Miller et al., 2009, 596406). Lastly, several effects were seen at
18 the level of the epithelial mesenchymal trophic unit in response to O3. These include altered
19 organization of the airways epithelium (Schelegle et al., 2003, 053778). increased abundance of
20 mucous goblet cells (Schelegle et al., 2003, 053778). disruption of the basement membrane zone
21 (Evans et al., 2004, 596379). reduced innervation (Larson et al., 2004, 057062). increased
22 neuroendocrine-like cells (Joad et al., 2006, 596390). and altered orientation and abundance of
23 smooth muscle bundles (Plopper et al., 2007, 596412; Tran et al., 2004, 628626). Six months of
24 recovery in filtered air led to reversal of some but not all of these effects (Evans et al., 2004, 596379;
25 Kajekar et al., 2007, 567661; Plopper et al., 2007, 596412). The authors concluded that cyclic
26 challenge of infant rhesus monkeys to allergen and O3 during the postnatal period compromised
27 airway growth and development and resulted in changes which favor allergic airways responses
28 (Plopper et al., 2007, 596412V
29 Nasal mucous membranes are also a target of O3-mediated effects. The infant rhesus monkey
30 was used as a model since its nasal airways are similar to those of children (Carey et al., 2007,
31 195752). Lesions in airways epithelium in the developing nasal passages of immature monkeys were
32 determined following both acute (5 consecutive days of 0.5 ppm 8 h/day) and cyclic episodic (as
33 described above) O3 exposure. Similar effects were observed in response to acute and episodic O3.
34 Histological analysis demonstrated necrotizing rhinitis in the nasal mucosa lining of the main nasal
35 chamber and focal regions of epithelial exfoliation, especially in the anterior maxilloturbinate. An
36 anterior to posterior decrease was observed in the severity of these lesions. Morphometric analysis
37 demonstrated a 65% reduction in the mean thickness of the nasal epithelium in the anterior
38 maxilloturbinate and loss of volume density of airway cilia, epithelial cytoplasm and nuclei. The
March 2011 5-59 DRAFT - DO NOT CITE OR QUOTE
-------
1 authors reported that the O3-induced nasal lesions observed in this study are similar to those reported
2 for adult monkeys. However, unlike the adult monkeys, no epithelial hyperplasia or metaplasia was
3 observed in the young monkeys suggesting that persistent necrotizing rhinitis may be the long term
4 sequelae in the absence of protective adaptations.
5 Effects of O3 on early postnatal airways development has also been studied in rats. A recent
6 study demonstrated that O3 exposure during critical postnatal periods resulted in increased SP nerve
7 fiber density in lung smooth muscle (Hunter et al., 2010, 382064). The authors proposed that O3 may
8 lead to enhanced responsiveness of airway sensory nerves. Another study found increases in
9 immediate-early gene responses in airways epithelium of rats exposed postnatally to O3 (Johnston et
10 al., 2006, 097439). Further, neonatal mice exhibit strain-specific differential susceptibility to O3
11 (Vancza et al., 2009, 596419).
5.2.9.4. Lifestage: Aging
12 On the other side of the lifestage spectrum is aging. The spirometric response to O3 is lost in
13 humans as they age (Drechsler-Parks, 1995, 076085: Hazucha et al., 2003, 048168). In mice,
14 physiological responses to O3 were also attenuated with age (Hamade et al., 2010, 666324).
15 Mechanisms accounting for this effect have not been well-studied but could include altered number
16 and sensitivity of receptors or altered signaling pathways involved in neural reflexes.
5.2.9.5. Adaptation
17 The decrease in pulmonary function and increase in bronchoconstriction, airways
18 inflammation and bronchial reactivity observed on the first and second days of consecutive daily
19 exposure in response to O3 were not seen after 4 or 5 days (see above). Several mechanisms have
20 been postulated. First, the upregulation of antioxidant defenses (or conversely, a decrease in critical
21 O3-reactive substrates) may protect against O3-mediated adverse effects. Increases in antioxidant
22 content of the BAL have been demonstrated by Devlin (1997, 083577). Tepper (1989, 041991). and
23 others. Second, IL-6 was demonstrated to be an important mediator of adaptation (McKinney et al.,
24 1998, 086751). Third, a protective role for increases in mucus producing cells and mucus
25 concentrations in the airways has also been proposed (Devlin et al., 1997, 083577). Fourth, epithelial
26 hyperplasia or metaplasia may decrease susceptibility to subsequent O3 challenge (Carey et al., 2007,
27 195752: Harkema et al., 1987, 040816: Harkema et al., 1987, 041496). These morphologic changes
28 have been observed in nasal and lower airways. Although there is some evidence to support these
29 possibilities, there is no consensus on mechanisms underlying adaptation. Recent studies
30 demonstrating that O3 activates TRP receptors suggest that modulation of TRP receptor number or
31 sensitivity by repeated O3 exposures may also contribute to adaptation.
March 2011 5-60 DRAFT - DO NOT CITE OR QUOTE
-------
5.2.9.6. Co-Exposures with Particulate Matter
1 Numerous studies have investigated the effects of co-exposure to O3 and PM because of the
2 prevalence of these pollutants in ambient air. Results are highly variable and depend on whether
3 exposures are simultaneous or sequential, the type of PM employed and the endpoint examined. In
4 humans, simultaneous exposure to O3 (0.12 ppm for 2 h at rest) and CAPs resulted in a diminished
5 systemic IL-6 response compared with exposure to CAPs alone (Urch et al., 2010, 387113).
6 Exposure to O3 alone did not alter blood IL-6 levels (Urch et al., 2010, 387113). The authors
7 provided evidence that O3 mediated a switch to shallow breathing which may have accounted for this
8 effect (Urch et al., 2010, 387113). Further, simultaneous exposure to CAPs and O3, but not exposure
9 to either alone, resulted in increased diastolic blood pressure in human subjects (Fakhri et al., 2009,
10 191914). In some strains of mice, pre-exposure to O3 (0.5 ppm for 2 hours) modulated the effects of
11 carbon black PM on heart rate, HRV and breathing patterns (Hamade and Tankersley, 2009, 596386).
12 Another recent study in mice demonstrated that treatment with carbon nanotubes followed 12 hours
13 later by O3 exposure (0.5 ppm for 3 hours) resulted in a dampening of some of the pulmonary effects
14 of carbon nanotubes measured as markers of inflammation and injury in the BAL (Han et al., 2008,
15 596387). The authors suggest that this may represent "cross-tolerance." Harkema et al. (2005,
16 078340) found that epithelial and inflammatory responses in the airways of rats were enhanced by
17 co-exposure to O3 and LPS (used as a model of biogenic PM) or to O3 and OVA (used as a model of
18 an aeroallergen). Furthermore, one recent study demonstrated maternal-fetal effects of PM exposure
19 on O3 responses. In this study, maternal exposure to PM resulted in augmented lung mediators of
20 inflammation, airway epithelial mucous metaplasia and enhanced O3-mediated AHR in young mice
21 (Auten et al., 2009, 200760). Overall, these findings are hard to interpret but demonstrate the
22 complexity of interactions between PM and O3 exposures.
5.2.9.7. Summary
23 Collectively, these older and more recent studies provide evidence for mechanisms which may
24 underlie the variability in responsiveness seen among individuals. Certain functional genetic
25 polymorphisms, pre-existing conditions and diseases, lifestages and co-exposures contribute to
26 enhanced susceptibility to O3. Adaptation may also be important, but it is incompletely understood,
27 both in terms of the pathways involved and the resulting consequences.
5.2.10. Overall Summary
28 Key events in the toxicity pathway of O3 have been identified in humans and animal models.
29 They include the formation of secondary oxidation products in the lung, activation of neural reflexes,
30 pulmonary injury and inflammation and increased bronchial reactivity. In addition, evidence is
31 accumulating that immune system modulation may lead to impaired host defense and the
March 2011 5-61 DRAFT - DO NOT CITE OR QUOTE
-------
1 exacerbation and/or induction of asthma and allergic responses (Figure 5-6). Systemic inflammation
2 and vascular oxidative/nitrosative stress may be critical to the extrapulmonary effects of O3.
Ozone + Respiratory Tract
w
Formation of secondary oxidation products
/ I \
Activation of
neural reflexes
s|/lnspiratory Mild
Capacity bronchoconstriction
Increased
bronchial „.
. . ^-
reactivity
(AHR)
\
Decrements in pulmonary function
\
Exacerbation/induction of
asthma and allergic responses
permeability
Airways •
neutrophilia
Allergic priming
& sensitization
Inflammation
and injury
Repair or
remodeling
Immune system
modulation
.-"' \
Decreased
pathogen clearance
Impaired host defense/
respiratorytract infections
Figure 5-6. Schematic depicting key events in ozone's toxicity pathway. Solid arrows denote
pathways for which there is greater certainty. Broken arrows represent pathways of
emerging interest.
5.2.11. Gaps in Knowledge
3 Despite a vast body of knowledge regarding the effects of O3 exposure, the current
4 understanding of mechanisms underlying important health effects in humans is incomplete.
5 Additional research will be useful to elucidate the biologic pathways by which exposure to O3:
6 • Primes the immune system, including promotion of adaptive immunity and activation of
7 innate immunity
8
9
Alters early postnatal development of the lung and immune system
Affects the cardiovascular system
March 2011
5-62
DRAFT - DO NOT CITE OR QUOTE
-------
References
A list of all references considered for inclusion in the dosimetry section of this chapter can be found at
http://hero.epa. gov/index.cfm?action=search.do&submit=Search&group_id=409
A list of all references considered for inclusion in the mode of action section of this chapter can be found at
http://hero.epa.gov/index.cfm?action=search.do&submit=Search&group_id=392
Ahmad, S.; Ahmad, A.; McConville, G; Schneider, B. K.; Allen, C. B.; Manzer, R.; Mason, R. I; White, C. W. (2005).
Lung epithelial cells release ATP during ozone exposure: signaling for cell survival. Free Radic Biol Med, 39: 213-
226. http://dx.doi.0rg/10.1016/j.freeradbiomed.2005.03.009 196429
Aibo, D. I.; Birmingham, N. R; Lewandowski, R.; Maddox, J. R; Roth, R. A.; Ganey, R E.; Wagner, J. G; Harkema, J. R.
(2010). Acute Exposure to Ozone Exacerbates Acetaminophen-Induced Liver Injury in Mice. Toxicol Sci, 115: 267-
285. http://dx.doi.org/10.1093/toxsci/kfq034 378559
Alexis, N. E.; Lay, J. C.; Hazucha, M.; Harris, B.; Hernandez, M. L.; Bromberg, R A.; Kehrl, H.; Diaz-Sanchez, D.; Kim,
C.; Devlin, R. B.; Peden, D. B. (2010). Low-level ozone exposure induces airways inflammation and modifies cell
surface phenotypes in healthy humans. Inhal Toxicol, 22: 593-600. http://dx.doi.org/10.3109/08958371003596587
628538
Alexis, N. E.; Zhou, H.; Lay, J. C.; Harris, B.; Hernandez, M. L.; Lu, T S.; Bromberg, P. A.; Diaz-Sanchez, D.; Devlin, R.
B.; Kleeberger, S. R.; Peden, D. B. (2009). The glutathione-S-transferase Mu 1 null genotype modulates ozone-
induced airway inflammation in human subjects. J Allergy Clin Immunol, 124: 1222-1228.
http://dx.doi.0rg/10.1016/i.iaci.2009.07.036628542
Alexis, N.; Soukup, J.; Nierkens, S.; Becker, S. (2001). Association between airway hyperreactivity and bronchial
macrophage dysfunction in individuals with mild asthma. Am J Physiol Lung Cell Mol Physiol, 280: L369-L375.
190013
Alexis, N.; Urch, B.; Tarlo, S.; Corey, P.; Pengelly, D.; O'Byrne, P.; Silverman, F. (2000). Cyclooxygenase metabolites play
a different role in ozone-induced pulmonary function decline in asthmatics compared to normals. Inhal Toxicol, 12:
1205-1224. 013072
Alfaro, M. R; Putney, L.; Tarkington, B. K.; Hatch, G. E.; Hyde, D. M.; Schelegle, E. S. (2004). Effect of rapid shallow
breathing on the distribution of 1 SO-labeled ozone reaction product in the respiratory tract of the rat. Inhal Toxicol,
16: 77-85. 053551
Alfaro, M. F.; Walby, W. F.; Adams, W. C.; Schelegle, E. S. (2007). Breath condensate levels of 8-isoprostane and
leukotriene B4 after ozone inhalation are greater in sensitive versus nonsensitive subjects. Exp Lung Res, 33: 115-
33. http://dx.doi.org/10.1080/01902140701364367 196567
Aris, R. M.; Christian, D.; Hearne, P. Q.; Kerr, K.; Finkbeiner, W. E.; Balmes, J. R. (1993). Ozone-induced airway
inflammation in human subjects as determined by airway lavage and biopsy. Am J Respir Crit Care Med, 148:
1363-1372.038275
Aris, R. M.; Tager, I.; Christian, D.; Kelly, T; Balmes, J. R. (1995). Methacholine responsiveness is not associated with
03-induced decreases in FEV1. Chest, 107: 621-628. 075945
Arito, H.; Uchiyama, L; Arakawa, H.; Yokoyama, E. (1990). Ozone-induced bradycardia and arrhythmia and their relation
to sleep-wakefulness in rats. Toxicol Lett, 52: 169-178. http://dx.doi.org/10.1016/0378-4274(90)90151-6 042285
Arsalane, K.; Gosset, P.; Vanhee, D.; Voisin, C.; Hamid, Q.; Tonnel, A.-B.; Wallaert, B. (1995). Ozone stimulates synthesis
of inflammatory cytokines by alveolar macrophages in vitro. Am J Respir Cell Mol Biol, 13: 60-68. 077430
Asplund, P. T; Ben-Jebria, A.; Rigas, M. L.;Ultman, J. S. (1996). Longitudinal distribution of ozone absorption in the
lung: effect of continuous inhalation exposure. Arch Environ Occup Health, 51: 431-438. 082505
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments such as the Integrated Science Assessments (ISAs) and the Integrated Risk Information System (IRIS).
March 2011 5-62 DRAFT - DO NOT CITE OR QUOTE
-------
Auten, R. L.; Potts, E. N.; Mason, S. N.; Fischer, B.; Huang, Y; Foster, W. M. (2009). Maternal exposure to particulate
matter increases postnatal ozone-induced airway hyperreactivity in juvenile mice. Am J Respir Crit Care Med, 180:
1218-1226. http://dx.doi.org/10.1164/rccm.200901-0116OC 200760
Awasthi, Y. C.; Yang, Y; Tiwari, N. K.; Patrick, B.; Sharma, A.; Li, J.; Awasthi, S. (2004). Regulation of 4-
hydroxynonenal-mediated signaling by gluathione S-transferases. Free Radic Biol Med, 37: 607-619.
http://dx.doi.0rg/10.1016/i.freeradbiomed.2004.05.033644649
Ballinger, C. A.; Cueto, R.; Squadrito, G; Coffin, J. R; Velsor, L. W.; Pryor, W. A.; Postlethwait, E. M. (2005). Antioxidant-
mediated augmentation of ozone-induced membrane oxidation. Free Radic Biol Med, 38: 515-526.
http://dx.doi.0rg/10.1016/i.freeradbiomed.2004.ll.009076649
Balmes, J. R.; Aris, R. M.; Chen, L. L.; Scannell, C.; Tager, I. B.; Finkbeiner, W.; Christian, D.; Kelly, T.; Hearne, P. Q.;
Ferrando, R.; Welch, B. (1997). Effects of ozone on normal and potentially sensitive human subjects part I: Airway
inflammation and responsiveness to ozone in normal and asthmatic subjects. Boston, MA: Health Effects Institute.
086092
Balmes, J. R.; Chen, L. L.; Scannell, C.; Tager, L; Christian, D.; Hearne, P. Q.; Kelly, T.; Aris, R. M. (1996). Ozone-
induced decrements in FEV1 and FVC do not correlate with measures of inflammation. Am J Respir Crit Care
Med, 153: 904-909. 080830
Bang, S.; Kim, K. Y; Yoo, S.; Kim, Y. G; Hwang, S. W. (2007). Transient receptor potential Al mediates acetaldehyde-
evoked pain sensation. Eur J Neurosci. 26: 2516-2523. http://dx.doi.Org/10.llll/i.1460-9568.2007.05882.x 628545
Basha, M. A.; Gross, K. B.; Gwizdala, C. J.; Haidar, A. H.; Popovich, J. Jr (1994). Bronchoalveolar lavage neutrophilia in
asthmatic and healthy volunteers after controlled exposure to ozone and filtered purified air. Chest, 106: 1757-1765.
075950
Bates, M. L.; Brenza, T. M.; Ben-Jebria, A.; Bascom, R.; Ultman, J. S. (2009). Longitudinal distribution of ozone
absorption in the lung: Comparison of cigarette smokers and nonsmokers. Toxicol Appl Pharmacol, 236: 270-275.
195727
Beckett, W. S.; McDonnell, W. F.; Horstman, D. H.; House, D. E. (1985). Role of the parasympathetic nervous system in
acute lung response to ozone. J Appl Physiol, 59: 1879-1885. 039758
Belvisi, M. G; Stretton, C. D.; Verleden, G. M.; Ledingham, S. J.; Yacoub, M. H.; Barnes, P. J. (1992). Inhibition of
cholinergic neurotransmission in human airways by opioids. J Appl Physiol, 72: 1096-1100. 644681
Bennett, W. D.; Hazucha, M. J.; Folinsbee, L. J.; Bromberg, P. A.; Kissling, G. E.; London, S. J. (2007). Acute pulmonary
function response to ozone in young adults as a function of body mass index. Inhal Toxicol, 19:1147-1154.
http://dx.doi.org/10.1080/08958370701665475418827
Bergamaschi, E.; De Palma, G; Mozzoni, P.; Vanni, S.; Vettori, M. V; Broeckaert, F.; Bernard, A.; Mutti, A. (2001).
Polymorphism of quinone-metabolizing enzymes and susceptibility to ozone-induced acute effects. Am J Respir
Crit Care Med, 163: 1426-1431. 052670
Bhalla, D. K.; Gupta, S. K. (2000). Lung injury, inflammation, and inflammatory stimuli in rats exposed to ozone. J
Toxicol Environ Health, 59: 211-228. 015036
Blomberg, A.; Mudway, I. S.; Nordenhall, C.; Hedenstrom, H.; Kelly, F. J.; Frew, A. J.; Holgate, S. T; Sandstrom, T.
(1999). Ozone-induced lung function decrements do not correlate with early airway inflammatory or antioxidant
responses. Eur Respir J, 13: 1418-1428. 001267
Bosson, J.; Blomberg, A.; Pourazar, J.; Mudway, I. S.; Frew, A. J.; Kelly, F. J.; Sandstrom, T. (2009). Early suppression of
NFkappaB and IL-8 in bronchial epithelium after ozone exposure in healthy human subjects. Inhal Toxicol, 21:
913-919. http://dx.doi.org/10.1080/08958370802657389 399331
Bosson, J.; Stenfors, N.; Bucht, A.; Helleday, R.; Pourazar, J.; Holgate, S. T; Kelly, F. J.; Sandstrom, T; Wilson, S.; Frew,
A. J.; Blomberg, A. (2003). Ozone-induced bronchial epithelial cytokine expression differs between healthy and
asthmatic subjects. Clin Exp Allergy, 33: 777-782. 051687
Broeckaert, F.; Clippe, A.; Wattiez, R.; Falmagne, P.; Bernard, A. (2003). Lung hyperpermeability, Clara-cell secretory
potein (CC16), and susceptibility to ozone of five inbred strains of mice. Inhal Toxicol, 15: 1209-1230. 055490
Brook, R. D.; Brook, J. R.; Urch, B.; Vincent, R.; Rajagopalan, S.; Silverman, F. (2002). Inhalation of fine particulate air
pollution and ozone causes acute arterial vasoconstriction in healthy adults. Circulation, 105: 1534-1536. 024987
March 2011 5-63 DRAFT - DO NOT CITE OR QUOTE
-------
Brook, R. D.; Urch, B.; Dvonch, J. T.; Bard, R. L.; Speck, M.; Keeler, G; Morishita, M.; Marsik, F. J.; Kamal, A. S.;
Kaciroti, N.; Harkema, J.; Corey, P.; Silverman, F.; Gold, D. R.; Wellenius, G; Mittleman, M. A.; Rajagopalan, S.;
Brook, J. R. (2009). Insights into the mechanisms and mediators of the effects of air pollution exposure on blood
pressure and vascular function in healthy humans. Hypertension, 54: 659-U420.
http://dx.doi.org/10.1161/hvpertensionaha.109.130237 195611
Bush, M. L.; Asplund, P. T.; Miles, K. A.; Ben-Jebria, A.; Ultman, J. S. (1996). Longitudinal distribution of O3 absorption
in the lung: gender differences and intersubject variability. J Appl Physiol, 81: 1651-1657. 080763
Bush, M. L.; Zhang, W.; Ben-Jebria, A.; Ultman, J. S. (2001). Longitudinal distribution of ozone and chlorine in the human
respiratory tract: simulation of nasal and oral breathing with the single-path diffusion model. Toxicol Appl
Pharmacol, 173: 137-145. 016665
Caceres, A. L; Brackmann, M.; Elia, M. D.; Bessac, B. F.; del Camino, D.; DAmours, M.; Witek, J. S.; Fanger, C. M.;
Chong, J. A.; Hayward, N. J.; Homer, R. J.; Cohn, L.; Huang, X.; Moran, M. M.; Jordt, S. E. (2009). A sensory
neuronal ion channel essential for airway inflammation and hyperreactivity in asthma. PNAS, 106: 9099-9104.
http://dx.doi.org/10.1073/pnas.0900591106628549
Carey, S. A.; Minard, K. R.; Trease, L. L.; Wagner, J. G; Garcia, G. J.; Ballinger, C. A.; Kimbell, J. S.; Plopper, C. G;
Corley, R. A.; Postlethwait, E. M.; Harkema, J. R.; Einstein, D. R. (2007). Three-dimensional mapping of ozone-
induced injury in the nasal airways of monkeys using magnetic resonance imaging and morphometric techniques.
Toxicol Pathol, 35: 27-40. http://dx.doi.org/10.1080/01926230601072343 195752
Chang, L.-Y; Huang, Y; Stockstill, B. L.; Graham, J. A.; Grose, E. C.; Menache, M. G; Miller, F. J.; Costa, D. L.; Crapo, J.
D. (1992). Epithelial injury and interstitial fibrosis in the proximal alveolar regions of rats chronically exposed to a
simulated pattern of urban ambient ozone. Toxicol Appl Pharmacol, 115: 241-252. http://dx.doi.org/10.1016/0041-
008X(92)90329-Q 042387
Chen, C.; Arjomandi, M.; Balmes, J.; Tager, I.; Holland N (2007). Effects of Chronic and Acute Ozone Exposure on Lipid
Peroxidation and Antioxidant Capacity in Healthy Young Adults. Environ Health Perspect, 115: 1732-1737.
http://dx.doi.org/10.1289/ehp.10294 145956
Chen, C.; Arjomandi, M.; Qin, H.; Balmes, J.; Tager, I.; Holland, N. (2006). Cytogenetic damage in buccal epithelia and
peripheral lymphocytes of young healthy individuals exposed to ozone. Mutagenesis, 21: 131-137.
http://dx.doi.org/10.1093/mutage/gel007 196504
Cho, H. Y; Hotchkiss, J. A.; Harkema, J. R. (1999). Inflammatory and epithelial responses during the development of
ozone-induced mucous cell metaplasia in the nasal epithelium of rats. Toxicol Sci, 51: 135-145. 011985
Cho, H. Y; Kleeberger, S. R. (2007). Genetic mechanisms of susceptibility to oxidative lung injury in mice. Free Radic
BiolMed,42: 433-445. http://dx.doi.Org/10.1016/j.freeradbiomed.2006.ll.021 195616
Cho, H.-Y; Zhang, L.-Y; Kleeberger, S. R. (2001). Ozone-induced lung inflammation and hyperreactivity are mediated via
tumor necrosis factor-"alpha" receptors. Am J Physiol, 280: L537-L546. 016160
Christian, D. L.; Chen, L. L.; Scannell, C. H.; Ferrando, R. E.; Welch, B. S.; Balmes, J. R. (1998). Ozone-induced
inflammation is attenuated with multiday exposure. Am J Respir Crit Care Med, 158: 532-537. 029925
Chuang, G. C.; Yang, Z.; Westbrook, D. G; Pompilius, M.; Ballinger, C. A.; White, R. C.; Krzywanski, D. M.; Postlethwait,
E. M.; Ballinger, S. W. (2009). Pulmonary ozone exposure induces vascular dysfunction, mitochondrial damage,
and atherogenesis. Am J Physiol Lung Cell Mol Physiol, 297: L209-L216.
http://dx.doi.org/10.1152/aiplung.00102.2009 197202
Cohen-Hubal, E. A.; Kimbell, J. S.; Fedkiw, P. S. (1996). Incorporation of nasal-lining mass-transfer resistance into a CFD
model for prediction of ozone dosimetry in the upper respiratory tract. Inhal Toxicol, 8: 831-857. 043785
Cole, M. P.; Freeman, B. A. (2009). Promotion of cardiovascular disease by exposure to the air pollutant ozone. Am J
Physiol Lung Cell Mol Physiol, 297: L209-L216. 597507
Coleridge, H. M.; Coleridge, J. C. G; Ginzel, K. H.; Baker, D. G; Banzett, R. B.; Morrison, M. A. (1976). Stimulation of
'irritant' receptors and afferent C-fibers in the lungs by prostaglandins. Nature, 264: 451-453. 038612
Coleridge, J. C. G; Coleridge, H. M.; Schelegle, E. S.; Green, J. F. (1993). Acute inhalation of ozone stimulates bronchial
C-fibers and rapidly adapting receptors in dogs. J Appl Physiol, 74: 2345-2352. 038695
March 2011 5-64 DRAFT - DO NOT CITE OR QUOTE
-------
Corradi, M.; Alinovi, R.; Goldoni, M.; Vettori, M.; Folesani, G; Mozzoni, P.; Cavazzini, S.; Bergamaschi, E.; Rossi, L.;
Mutti, A. (2002). Biomarkers of oxidative stress after controlled human exposure to ozone. Toxicol Lett, 134: 219-
225. 035448
Costa, D. L.; Schafrank, S. N.; Wehner, R. W.; Jellett, E. (1985). Alveolar permeability to protein in rats differentially
susceptible to ozone. J Appl Toxicol, 5: 182-186. http://dx.doi.org/10.1002/jat.2550050309 040273
Cross, C. E.; Motchnik, P. A.; Bruener, B. A.; Jones, D. A.; Kaur, R; Ames, B. N.; Halliwell, B. (1992). Oxidative damage
to plasma constituents by ozone. FEES Lett, 298: 269-272. http://dx.doi.org/10.1016/0014-5793(92)80074-Q
625299
Cueto, R.; Squadrito, G. L.; Bermudez, E.; Pryor, W. A. (1992). Identification of heptanal and nonanal in bronchoalveolar
lavage from rats exposed to low levels of ozone. Biochem Biophys Res Commun, 188: 129-134.
http://dx.doi.org/10.1016/0006-291X(92)92359-6 042770
Dahl, A. R. (1990). Dose concepts for inhaled vapors and gases. Toxicol Appl Pharmacol, 103: 185-197. 094536
Dahl, M.; Bauer, A. K.; Arredouani, M.; Soininen, R.; Tryggvason, K.; Kleeberger, S. R.; Kobzik, L. (2007). Protection
against inhaled oxidants through scavenging of oxidized lipids by macrophage receptors MARCO and SR-AI/II. J
Clin Invest, 117: 757-764. http://dx.doi.org/10.1172/JCI29968 196986
Delaunois, A.; Segura, P.; Montano, L. M.; Vargas, M. H.; Ansay, M.; Gustin, P. (1998). Comparison of ozone-induced
effects on lung mechanics and hemodynamics in the rabbit. Toxicol Appl Pharmacol, 150: 58-67. 015779
Devlin, R. B.; Folinsbee, L. J.; Biscardi, E; Hatch, G; Becker, S.; Madden, M. C.; Robbins, M.; Koren, H. S. (1997).
Inflammation and cell damage induced by repeated exposure of humans to ozone. Inhal Toxicol, 9: 211-235.
083577
Devlin, R. B.; McDonnell, W. E; Mann, R.; Becker, S.; House, D. E.; Schreinemachers, D.; Koren, H. S. (1991). Exposure
of humans to ambient levels of ozone for 6.6 hours causes cellular and biochemical changes in the lung. Am J
Respir Cell Mol Biol, 4: 72-81. 040359
Dimeo, M. J.; Glenn, M. G; Holtzman, M. J.; Sheller, J. R.; Nadel, J. A.; Boushey, H. A. (1981). Threshold concentration
of ozone causing an increase in bronchial reactivity in humans and adaptation with repeated exposures. Am Rev
Respir Dis, 124: 245-248. 039662
Dormans, J. A. M. A.; Van Bree, L.; Boere, A. J. F.; Marra, M.; Rombout, P. J. A. (1999). Interspecies differences in time
course of pulmonary toxicity following repeated exposure to ozone. Inhal Toxicol, 11: 309-329. 040766
Dostert, C.; Petrilli, V; Van Bruggen, R.; Steele, C.; Mossman, B. T; Tschopp, J. (2008). Innate Immune Activation
Through Nalp3 Inflammasome Sensing of Asbestos and Silica. Science, 320: 674-677. 155753
Drechsler-Parks, D. M. (1995). The dose-response relationship in older men exposed to ozone. Exp Gerontol, 30: 65-75.
076085
Duan, X.; Buckpitt, A. R.; Pinkerton, K. E.; Ji, C.; Plopper, C. G. (1996). Ozone-induced alterations in glutathione in lung
subcompartments of rats and monkeys. Am J Respir Cell Mol Biol, 14: 70-75. 080791
Duan, X.; Buckpitt, A. R.; Plopper, C. G. (1993). Variation in antioxidant enzyme activities in anatomic subcompartments
within rat and rhesus monkey lung. Toxicol Appl Pharmacol, 123: 73-82. 086326
Emmons, K.; Foster, W. M. (1991). Smoking cessation and acute airway response to ozone. Arch Environ Occup Health,
46: 288-295. 042430
Enami, S.; Hoffmann, M. R.; Colussi, A. J. (2008). Acidity enhances the formation of a persistent ozonide at aqueous
ascorbate/ozone gas interfaces. PNAS, 105: 7365-7369. 195833
Enami, S.; Hoffmann, M. R.; Colussi, A. J. (2008). Ozonolysis of uric acid at the air/water interface. J Phys Chem B, 112:
4153-4156. http://dx.doi.org/10.1021/ip712010k 195834
Enami, S.; Hoffmann, M. R.; Colussi, A. J. (2009). How phenol and alpha-tocopherol react with ambient ozone at
gas/liquid interfaces. J Phys Chem A, 113: 7002-7010. http://dx.doi.org/10.1021/ip901712k 195835
Enami, S.; Hoffmann, M. R.; Colussi, A. J. (2009). Ozone oxidizes glutathione to a sulfonic acid. Chem Res Toxicol, 22:
35-40. http://dx.doi.org/10.1021/tx800298jlO.1021/tx800298j 197791
March 2011 5-65 DRAFT - DO NOT CITE OR QUOTE
-------
Enami, S.; Hoffmann, M. R.; Colussi, A. J. (2009). Simultaneous detection of cysteine sulfenate, sulfinate, and sulfonate
during cysteine interfacial ozonolysis. J Phys Chem B, 113: 9356-9358. http://dx.doi.org/10.1021/jp904316n
195621
Evans, M. J.; Fanucchi, M. V; Baker, G. L.; Van Winkle, L. S.; Pantle, L. M.; Nishio, S. J.; Schelegle, E. S.; Gershwin, L.
J.; Miller, L. A.; Hyde, D. M.; Plopper, C. G. (2004). The remodelled tracheal basement membrane zone of infant
rhesus monkeys after 6 months of recovery. Clin Exp Allergy, 34: 1131-1136. http://dx.doi.Org/10.llll/j.1365-
2222.2004.02004.x CEA2004 596379
Fabbri, L. M.; Aizawa, H.; O'Byrne, P. M.; Bethel, R. A.; Walters, E. H.; Holtzman, M. J.; Nadel, J. A. (1985). An anti-
inflammatory drug (BW755C) inhibits airway hyperresponsiveness induced by ozone in dogs. J Allergy Clin
Immunol, 76: 162-166. http://dx.doi.org/10.1016/0091-6749(85)90695-5 040276
Fakhri, A. A.; Ilic, L. M.; Wellenius, G. A.; Urch, B.; Silverman, R; Gold, D. R.; Mittleman, M. A. (2009). Autonomic
effects of controlled fine particulate exposure in young healthy adults: Effect modification by ozone. Environ
Health Perspect, 117: 1287-1292. http://dx.doi.org/10.1289/ehp.0900541 191914
Fanucchi, M. V; Plopper, C. G; Evans, M. J.; Hyde, D. M.; Van Winkle, L. S.; Gershwin, L. J.; Schelegle, E. S. (2006).
Cyclic exposure to ozone alters distal airway development in infant rhesus monkeys. Am J Physiol Lung Cell Mol
Physiol, 291: L644-L650. http://dx.doi.org/10.1152/ajplung.00027.2006 096491
Feng, R.; He, W.; Ochi, H.; Castranova, V (2006). Ozone exposure impairs antigen-specific immunity but activates IL-7-
induced proliferation of CD4-CD8-thymocytes in BALB/c mice. J Toxicol Environ Health A, 69: 1511-1526.
http://dx.doi.org/10.1080/15287390500468696596381
Folinsbee, L. J.; Bedi, J. R; Horvath, S. M. (1980). Respiratory responses in humans repeatedly exposed to low
concentrations of ozone. Am Rev Respir Dis, 121: 431-439. 038880
Foster, W. M.; Freed, A. N. (1999). Regional clearance of solute from peripheral airway epithelia: recovery after sublobar
exposure to ozone. J Appl Physiol, 86: 641-646. 001202
Foster, W. M.; Stetkiewicz, P. T. (1996). Regional clearance of solute from the respiratory epithelia: 18—20 hpostexposure
to ozone. J Appl Physiol, 81:1143-1149. 079920
Frampton, M. W.; Morrow, P. E.; Torres, A.; Cox, C.; Voter, K. Z.; Utell, M. J.; Gibb, F. R.; Speers, D. M. (1997). Ozone
responsiveness in smokers and nonsmokers. Am J Respir Crit Care Med, 155: 116-121. 082692
Frampton, M. W.; Morrow, P. E.; Torres, A.; Voter, K. Z.; Whitin, J. C.; Cox, C.; Speers, D. M.; Tsai, Y; Utell, M. J.
(1997). Effects of ozone on normal and potentially sensitive human subjects Part II: airway inflammation and
responsiveness to ozone in nonsmokers and smokers. Boston, MA: Health Effects Institute. 086111
Frampton, M. W.; Pryor, W. A.; Cueto, R.; Cox, C.; Morrow, P. E.; Utell, M. J. (1999). Ozone exposure increases aldehydes
in epithelial lining fluid in human lung. Am J Respir Crit Care Med, 159: 1134-1137. 040757
Frank, R.; Liu, M. C.; Spannhake, E. W.; Mlynarek, S.; Macri, K.; Weinmann, G. G. (2001). Repetitive ozone exposure of
young adults: evidence of persistent small airway dysfunction. Am J Respir Crit Care Med, 164: 1253-1260.
093491
Freed, A. N.; Chou, C. L.; Fuller, S. D.; Croxton, T. L. (1996). Ozone-induced vagal reflex modulates airways reactivity in
rabbits. Respir Physiol Neurobiol, 105: 95-102. 080798
Freed, A. N.; Cueto, R.; Pryor, W. A. (1999). Antioxidant transport modulates peripheral airway reactivity and
inflammation during ozone exposure. J Appl Physiol, 87: 1595-1603. 011829
Fujita, M.; Sasayama, S.; Ohno, A.; Nakajima, H.; Asanoi, H. (1987). Importance of angina for development of collateral
circulation. Heart, 57: 139-143. 004280
Gao, X.; Raghavamenon, A. C.; DAuvergne, O.; Uppu, R. M. (2009). Cholesterol secoaldehyde induces apoptosis in J774
macrophages via mitochondrial pathway but not involving reactive oxygen species as mediators. Biochem Biophys
Res Commun, 389: 382-387. http://dx.doi.Org/10.1016/j.bbrc.2009.09.005 200764
Garantziotis, S.; Li, Z.; Potts, E. N.; Kimata, K.; Zhuo, L.; Morgan, D. L.; Savani, R. C.; Noble, P. W.; Foster, W. M.;
Schwartz, D. A.; Hollingsworth, J. W. (2009). Hyaluronan mediates ozone-induced airway hyperresponsiveness in
mice. JBiol Chem, 284: 11309-11317. http://dx.doi.org/10.1074/ibc.M802400200 597603
March 2011 5-66 DRAFT - DO NOT CITE OR QUOTE
-------
Garantziotis, S.; Li, Z.; Potts, E. N.; Lindsey, J. Y; Stober, V. P.; Polosukhin, V. V; Blackwell, T. S.; Schwartz, D. A.;
Foster, W. M.; Hollingsworth, J. W. (2010). TLR4 is necessary for hyaluronan-mediated airway
hyperresponsiveness after ozone inhalation. Am J Respir Crit Care Med, 181: 666-675.
http://dx.doi.org/10.1164/rccm.200903-0381OC624947
Gerrity, T. R.; Biscardi, F.; Strong, A.; Garlington, A. R.; Brown, J. S.; Bromberg, P. A. (1995). Bronchoscopic
determination of ozone uptake in humans. J Appl Physiol, 79: 852-860. 042785
Gerrity, T. R.; McDonnell, W. R; House, D. E. (1994). The relationship between delivered ozone dose and functional
responses in humans. Toxicol Appl Pharmacol, 124: 275-283. 041314
Gerrity, T. R.; Weaver, R. A.; Berntsen, J.; House, D. E.; O'Neil, J. J. (1988). Extrathoracic and intrathoracic removal of O3
in tidal-breathing humans. J Appl Physiol, 65: 393-400. 040899
Giamalva, D.; Church, D. R; Pryor, W. A. (1985). A comparison of the rates of ozonation of biological antioxidants and
oleate and linoleate esters. Biochem Biophys Res Commun, 133: 773-779. 595129
Gilmour, M. I.; Park, P.; Doerfler, D.; Selgrade, M. K. (1993). Factors that influence the suppression of pulmonary
antibacterial defenses in mice exposed to ozone. Exp Lung Res, 19: 299-314.
http://dx.doi.org/10.3109/01902149309064348039620
Gong, H. Jr; Wong, R.; Sarma, R. J.; Linn, W. S.; Sullivan, E. D.; Shamoo, D. A.; Anderson, K. R.; Prasad, S. B. (1998).
Cardiovascular effects of ozone exposure in human volunteers. Am J Respir Crit Care Med, 158: 538-546. 029938
Graham, D. E.; Koren, H. S. (1990). Biomarkers of inflammation in ozone-exposed humans: comparison of the nasal and
bronchoalveolar lavage. Am J Respir Crit Care Med, 142: 152-156. 042299
Graham, J. A.; Menzel, D. B.; Mole, M. L.; Miller, F. J.; Gardner, D. E. (1985). Influence of ozone on pentobarbital
pharmacokinetics in mice. Toxicol Lett, 24: 163-170. http://dx.doi.org/10.1016/0378-4274(85)90053-0 040289
Gross, E. A.; Starr, T. B.; Randall, H. W.; Morgan, K. T. (1987). Morphometric analysis of the primate nasal cavity
[Abstract]. Toxicologist, 7: 193. 625447
Gross, E. A.; Swenberg, J. A.; Fields, S.; Popp, J. A. (1982). Comparative morphometry of the nasal cavity in rats and
mice. JAnat, 135: 83-88. 040121
Gunnison, A. F.; Hatch, G E. (1999). OS-induced inflammation in prepregnant, pregnant, and lactating rats correlates with
O3 dose estimated by ISO. Am J Physiol, 276: L332-L340. 087204
Gunnison, A. F.; Hatch, G. E.; Crissman, K.; Bowers, A. (1996). Comparative sensitivity of lactating and virgin female rats
to ozone-induced pulmonary inflammation. Inhal Toxicol, 8: 607-623. 080803
Hackney, J. D.; Linn, W. S.; Mohler, J. G; Collier, C. R. (1977). Adaptation to short-term respiratory effects of ozone in
men exposed repeatedly. J Appl Physiol, 43: 82-85. 038282
Hamade, A. K.; Misra, V.; Rabold, R.; Tankersley, C. G. (2010). Age-related changes in cardiac and respiratory adaptation
to acute ozone and carbon black exposures: Interstrain variation in mice. Inhal Toxicol, 22: 84-94.
http://dx.doi.org/10.3109/08958378.2010.503974666324
Hamade, A. K.; Tankersley, C. G. (2009). Interstrain variation in cardiac and respiratory adaptation to repeated ozone and
particulate matter exposures. Am J Physiol Regul Integr Comp Physiol, 296: R1202-R1215.
http://dx.doi.org/10.1152/aipregu.90808.2008596386
Hamilton, R. F.; Li, L.; Eschenbacher, W. L.; Szweda, L.; Holian, A. (1998). Potential involvement of 4-hydroxynonenal in
the response of human lung cells to ozone. Am J Physiol, 274: L8-L16. 086157
Han, S. G; Andrews, R.; Gairola, C. G; Bhalla, D. K. (2008). Acute pulmonary effects of combined exposure to carbon
nanotubes and ozone in mice. Inhal Toxicol, 20: 391-398. http://dx.doi.org/10.1080/08958370801904014 596387
Haque, R.; Umstead, T. M.; Freeman, W. M.; Floros, J.; Phelps, D. S. (2009). The impact of surfactant protein-A on ozone-
induced changes in the mouse bronchoalveolar lavage proteome. Proteome Science, 7: 12.
http://dx.doi.org/10.1186/1477-5956-7-12200767
Haque, R.; Umstead, T. M.; Ponnuru, P.; Guo, X.; Hawgood, S.; Phelps, D. S.; Floros, J. (2007). Role of surfactant protein-
A (SP-A) in lung injury in response toacute ozone exposure of SP-A deficient mice. Toxicol Appl Pharmacol, 220:
72-82. http://dx.doi.0rg/10.1016/j.taap.2006.12.017 597606
March 2011 5-67 DRAFT - DO NOT CITE OR QUOTE
-------
Harkema, J. R.; Hotchkiss, J. A.; Ban, E. B.; Bennett, C. B.; Gallup, M.; Lee, J. K.; Basbaum, C. (1999). Long-lasting
effects of chronic ozone exposure on rat nasal epithelium. Am J Respir Cell Mol Biol, 20: 517-529. 001209
Harkema, J. R.; Plopper, C. G; Hyde, D. M.; StGeorge, J. A.; Dungworth, D. L. (1987). Effects of an ambient level of
ozone on primate nasal epithelial mucosubstances: quantitative histochemistry. Am J Pathol, 127: 90-96.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMCl899606/. 040816
Harkema, J. R.; Plopper, C. G; Hyde, D. M.; StGeorge, J. A.; Wilson, D. W.; Dungworth, D. L. (1987). Response of the
macaque nasal epithelium to ambient levels of ozone: a morphologic and morphometric study of the transitional
and respiratory epithelium. Am J Pathol, 128: 29-44. 041496
Harkema, J. R.; Wagner, J. G. (2005). Epithelial and inflammatory responses in the airways of laboratory rats coexposed to
ozone and biogenic substances: Enhancement of toxicant-induced airway injury. Exp Toxicol Pathol, 57: 129-141.
http://dx.doi.0rg/10.1016/i.etp.2005.05.013078340
Hatch, G. E. (1992). Comparative biochemistry of airway lining fluid. In RA Parent (Ed.), Comparative biology of the
normal lung: v. I, treatise on pulmonary toxicology (pp. 617-632). Boca Raton, FL: CRC Press, Inc. 043901
Hatch, G. E.; Slade, R.; Harris, L. P.; McDonnell, W. F.; Devlin, R. B.; Koren, H. S.; Costa, D. L.; McKee, J. (1994). Ozone
dose and effect in humans and rats: a comparison using oxygen-18 labeling and bronchoalveolar lavage. Am J
Respir Crit Care Med, 150: 676-683. 038953
Hatch, G. E.; Slade, R.; Stead, A. G; Graham, J. A. (1986). Species comparison of acute inhalation toxicity of ozone and
phosgene. J Toxicol Environ Health, 19: 43-53. http://dx.doi.org/10.1080/15287398609530905 040472
Hatch, G. E.; Wiester, M. J.; Overton, J. H. Jr; Aissa, M. (1989). Respiratory tract dosimetry of [18]O-labeled ozone in rats:
implications for a rat-human extrapolation of ozone dose. In Atmospheric ozone research and its policy
implications (pp. 553-560). Nijmegend, the Netherlands: Elsevier Science Publishers B. V. 041799
Hazbun, M. E.; Hamilton, R.; Holian, A.; Eschenbacher, W. L. (1993). Ozone-induced increases in substance P and 8-epi-
prostaglandin F2"alpha" in the airways of human subjects. Am J Respir Cell Mol Biol, 9: 568-572. 043914
Hazucha, M. J.; Bates, D. V.; Bromberg, PA. (1989). Mechanism of action of ozone on the human lung. J Appl Physiol,
67: 1535-1541.041909
Hazucha, M. J.; Folinsbee, L. J.; Bromberg, P. A. (2003). Distribution and reproducibility of spirometric response to ozone
by gender and age. J Appl Physiol, 95: 1917-1925. 048168
Hazucha, M. J.; Madden, M.; Pape, G; Becker, S.; Devlin, R.; Koren, H. S.; Kehrl, H.; Bromberg, P. A. (1996). Effects of
cyclo-oxygenase inhibition on ozone-induced respiratory inflammation and lung function changes. Eur J Appl
Physiol, 73: 17-27. 043923
Hoigne, J.; Bader, H. (1983). Rate constants of reactions of ozone with organic and inorganic compounds in water - II:
Dissociating organic compounds. Water Res, 17: 185-194. http://dx.doi.org/10.1016/0043-1354(83)90099-4 625266
Hollingsworth, J. W.; Cook, D. N.; Brass, D. M.; Walker, J. K. L.; Morgan, D. L.; Foster, W. M.; Schwartz, D. A. (2004).
The role of Toll-like receptor 4 in environmental airway injury in mice. Am J Respir Crit Care Med, 170: 126-132.
097816
Hollingsworth, J. W.; Free, M. E.; Li, Z.; Andrews, L. N.; Nakano, H.; Cook, D. N. (2010). Ozone activates pulmonary
dendritic cells and promotes allergic sensitization through a Toll-like receptor 4-dependent mechanism. J Allergy
Clinlmmunol, 125: 1167-1170. 635786
Hollingsworth, J. W.; Maruoka, S.; Li, Z.; Potts, E. N.; Brass, D. M.; Garantziotis, S.; Fong, A.; Foster, W. M.; Schwartz,
D. A. (2007). Ambient ozone primes pulmonary innate immunity in mice. J Immunol, 179: 4367-4375.
http://www.jimmunol.Org/cgi/content/abstract/179/7/4367. 597609
Holtzman, M. J.; Cunningham, J. H.; Sheller, J. R.; Irsigler, G. B.; Nadel, J. A.; Boushey, H. A. (1979). Effect of ozone on
bronchial reactivity in atopic and nonatopic subjects. Am Rev Respir Dis, 120: 1059-1067. 039220
Holtzman, M. J.; Fabbri, L. M.; O'Byrne, P. M.; Gold, B. D.; Aizawa, H.; Walters, E. H.; Alpert, S. E.; Nadel, J. A. (1983).
Importance of airway inflammation for hyperresponsiveness induced by ozone. Am Rev Respir Dis, 127: 686-690.
039745
Holz, O.; Jorres, R. A.; Timm, P.; Mucke, M.; Richter, K.; Koschyk, S.; Magnussen, H. (1999). Ozone-induced airway
inflammatory changes differ between individuals and are reproducible. Am J Respir Crit Care Med, 159: 776-784.
058731
March 2011 5-68 DRAFT - DO NOT CITE OR QUOTE
-------
Horstman, D. H.; Ball, B. A.; Brown, I; Gerrity, T.; Folinsbee, L. J. (1995). Comparison of pulmonary responses of
asthmatic and nonasthmatic subjects performing light exercise while exposed to a low level of ozone. Toxicol Ind
Health, 11: 369-385. 075834
Horvath, S. M.; Gliner, J. A.; Folinsbee, L. J. (1981). Adaptation to ozone: duration of effect. Am Rev Respir Dis, 123:
496-499. 039221
Hotchkiss, J. A.; Harkema, J. R.; Henderson, R. F. (1991). Effect of cumulative ozone exposure on ozone-induced nasal
epithelial hyperplasia and secretory metaplasia in rats. Exp Lung Res, 15: 589-600. 042441
Hu, P. C.; Miller, F. J.; Daniels, M. J.; Hatch, G. (1982). Protein accumulation in lung lavage fluid following ozone
exposure. Environ Res, 29: 377-388. http://dx.doi.org/10.1016/0013-9351(82)90039-1 039418
Hu, S. C.; Ben-Jebria, A.; Ultman, J. S. (1992). Longitudinal distribution of ozone absorption in the lung: quiet respiration
in healthy subjects. J Appl Physiol, 73: 1655-1667. 042794
Hu, S.-C.; Ben-Jebria, A.; Ultman, J. S. (1994). Longitudinal distribution of ozone absorption in the lung: effects of
respiratory flow. J Appl Physiol, 77: 574-583. 041323
Hunter, D. D.; Wu, Z.; Dey, R. D. (2010). Sensory neural responses to ozone exposure during early postnatal development
in rat airways. Am J Respir Cell Mol Biol, xx: xx-xx. http://dx.doi.org/10.1165/rcmb.2009-0191OC 382064
Ichinose, T.; Arakawa, K.; Shimojo, N.; Sagai, M. (1988). Biochemical effects of combined gases of nitrogen dioxide and
ozone: II species differences in lipid peroxides and antioxidative protective enzymes in the lungs. Toxicol Lett, 42:
167-176. 041805
Ignatenko, A. V.; Cherenkevich, S. N. (1985). [Reactivity of amino-acids and proteins in reactions with ozone]. Kinet
Catal, 26: 1332-1335. 625265
Islam, T.; McConnell, R.; Gauderman, W. J.; Avol, E.; Peters, J. M.; Gilliland, F. D. (2008). Ozone, oxidant defense genes
and risk of asthma during adolescence. Am J Respir Crit Care Med, 177: 388-395.
http://dx.doi.org/10.1164/rccm.200706-863OC097348
Iwasaki, T.; Takahashi, M.; Saito, H.; Arito, H. (1998). Adaptation of extrapulmonary responses to ozone exposure in
conscious rats. Ind Health, 36: 57-60. 086165
Jakab, G. J.; Spannhake, E. W.; Canning, B. J.; Kleeberger, S. R.; Gilmour, M. I. (1995). The effects of ozone on immune
function. Environ Health Perspect, 2: 77-89. 039548
Jiang, D.; Liang, J.; Fan, J.; Yu, S.; Chen, S.; Luo, Y; Prestwich, G. D.; Mascarenhas, M. M.; Garg, H. G; Quinn, D. A.;
Homer, R. J.; Goldstein, D. R.; Bucala, R.; Lee, P. J.; Medzhitov, R.; Noble, P. W. (2005). Regulation of lung injury
and repair by Toll-like receptors and hyaluronan. Nat Med, 11: 1173-1179. http://dx.doi.org/10.1038/nml315
628556
Joad, J. P.; Kott, K. S.; Brie, J. M. (1996). The local C-fiber contribution to ozone-induced effects on the isolated guinea
pig lung. Toxicol Appl Pharmacol, 141: 561-567. 082711
Joad, J. P.; Kott, K. S.; Brie, J. M.; Peake, J. L.; Plopper, C. G; Schelegle, E. S.; Gershwin, L. J.; Pinkerton, K. E. (2006).
Structural and functional localization of airway effects from episodic exposure of infant monkeys to allergen and/or
ozone. Toxicol Appl Pharmacol, 214: 237-243. http://dx.doi.Org/10.1016/j.taap.2005.12.012 596390
Johansson, E.; Wesselkamper, S. C.; Shertzer, H. G; Leikauf, G. D.; Dalton, T. P.; Chen, Y. (2010). Glutathione deficient
C57BL/6J mice are not sensitized to ozone-induced lung injury. Biochem Biophys Res Commun, 396: 407-412.
http://dx.doi.0rg/10.1016/i.bbrc.2010.04.105644476
Johnston, C.; Holm, B.; Gelein, R.; Finkelstein, J. (2006). Postnatal lung development: Immediate-early gene responses
post ozone and LPS exposure. Inhal Toxicol, 18: 875-883. http://dx.doi.org/10.1080/08958370600822466 097439
Johnston, R. A.; Mizgerd, J. P.; Shore, S. A. (2005). CXCR2 is essential for maximal neutrophil recruitment and
methacholine responsiveness after ozone exposure. Am J Physiol Lung Cell Mol Physiol, 288: L61-L67.
http://dx.doi.org/10.1152/ajplung.00101.2004 00101.2004 596393
Johnston, R. A.; Schwartzman, I. N.; Flynt, L.; Shore, S. A. (2005). Role of interleukin-6 in murine airway responses to
ozone. Am J Physiol Lung Cell Mol Physiol, 288: L390-L397. http://dx.doi.org/10.1152/ajplung.00007.2004
596394
March 2011 5-69 DRAFT - DO NOT CITE OR QUOTE
-------
Johnston, R. A.; Theman, T. A.; Lu, F. L.; Terry, R. D.; Williams, E. S.; Shore, S. A. (2008). Diet-induced obesity causes
innate airway hyperresponsiveness to methacholine and enhances ozone-induced pulmonary inflammation. J Appl
Psychol, 104: 1727-1735. http://dx.doi.org/10.1152/japplphysiol.00075.2008 597625
Jorres, R.; Nowak, D.; Magnussen, H.; Speckin, R; Koschyk, S. (1996). The effect of ozone exposure on allergen
responsiveness in subjects with asthma or rhinitis. Am J Respir Crit Care Med, 153: 56-64. 078122
Kabel, J. R.; Ben-Jebria, A.; Ultman, J. S. (1994). Longitudinal distribution of ozone absorption in the lung: comparison of
nasal and oral quiet breathing. J Appl Physiol, 77: 2584-2592. 095597
Kafoury, R. M.; Pryor, W. A.; Squadrito, G. L.; Salgo, M. G; Zou, X.; Friedman, M. (1998). Lipid ozonation products
activate phospholipases A2, C, and D. Toxicol Appl Pharmacol, 150: 338-349. 016913
Kajekar, R.; Pieczarka, E. M.; Smiley-Jewell, S. M.; Schelegle, E. S.; Fanucchi, M. V; Plopper, C. G. (2007). Early
postnatal exposure to allergen and ozone leads to hyperinnervation of the pulmonary epithelium. Respir Physiol
Neurobiol, 155: 55-63. http://dx.doi.Org/10.1016/j.resp.2006.03.002 567661
Kanofsky, J. R.; Sima, P. D. (1995). Reactive absorption of ozone by aqueous biomolecule solutions: implications for the
role of sulfhydryl compounds as targets for ozone. Arch Biochem Biophys, 316: 52-62. 075973
Kari, F.; Hatch, G; Slade, R.; Crissman, K.; Simeonova, P. P.; Luster, M. (1997). Dietary restriction mitigates ozone-
induced lung inflammation in rats: a role for endogenous antioxidants. Am J Respir Cell Mol Biol, 17: 740-747.
086171
Kehrl, H. R.; Peden, D. B.; Ball, B. A.; Folinsbee, L. J.; Horstman, D. H. (1999). Increased specific airway reactivity of
persons with mild allergic asthma after 7.6 hours of exposure to 0.16 ppm ozone. J Allergy Clin Immunol, 104:
1198-1204.022101
Kehrl, H. R.; Vincent, L. M.; Kowalsky, R. J.; Horstman, D. H.; O'Neil, J. J.; McCartney, W. H.; Bromberg, P. A. (1987).
Ozone exposure increases respiratory epithelial permeability in humans. Am Rev Respir Dis, 135: 1124-1128.
040824
Kermani, S.; Ben-Jebria, A.; Ultman, J. S. (2006). Kinetics of ozone reaction with uric acid, ascorbic acid, and glutathione
at physiologically relevant conditions. Arch Biochem Biophys, 451: 8-16.
http://dx.doi.0rg/10.1016/i.abb.2006.04.015 195643
Kleeberger, S. R.; Levitt, R. C.; Zhang, L.-Y.; Longphre, M.; Harkema, J.; Jedlicka, A.; Eleff, S. M.; DiSilvestre, D.;
Holroyd, K. J. (1997). Linkage analysis of susceptibility to ozone-induced lung inflammation in inbred mice. Nat
Genet, 17: 475-478. 095736
Kleeberger, S. R.; Reddy, S.; Zhang, L.-Y.; Jedlicka, A. E. (2000). Genetic susceptibility to ozone-induced lung
hyperpermeability: role of toll-like receptor 4. Am J Respir Cell Mol Biol, 22: 620-627. 014895
Kleeberger, S. R.; Seiden, J. E.; Levitt, R. C.; Zhang, L.-Y. (1993). Mast cells modulate acute ozone-induced inflammation
of the murine lung. Am J Respir Crit Care Med, 148: 1284-1291.044203
Kodavanti, U. P.; Costa, D. L.; Dreher, K. L.; Crissman, K.; Hatch, G. E. (1995). Ozone-induced tissue injury and changes
in antioxidant homeostasis in normal and ascorbate-deficient guinea pigs. Biochem Pharmacol, 50: 243-251.
http://dx.doi.org/10.1016/0006-2952(95)00122-0 077440
Kodavanti, U. P.; Thomas, R.; Ledbetter, A. D.; Schladweiler, M. C.; Shannahan, J. H.; Wallenborn, J. G; Lund, A. K.;
Campen, M. J.; Butler, E. O.; Gottipolu, R. R.; Nyska, A.; Richards, J. E.; Andrews, D.; Jaskot, R. H.; McKee, J.;
Kotha, S. R.; Patel, R. B.; Parianandi, N. L. (In Press). Vascular and cardiac impairments in rats Inhaling ozone and
diesel exhaust particles. Environ Health Perspect. http://dx.doi.org/10.1289/ehp.1002386 666323
Koenig, J. Q.; Covert, D. S.; Marshall, S. G; Van Belle, G; Pierson, W. E. (1987). The effects of ozone and nitrogen
dioxide on pulmonary function in healthy and in asthmatic adolescents. Am J Respir Crit Care Med, 136: 1152-
1157. 041521
Kreit, J. W.; Gross, K. B.; Moore, T. B.; Lorenzen, T. J.; DArcy, J.; Eschenbacher, W. L. (1989). Ozone-induced changes in
pulmonary function and bronchial responsiveness in asthmatics. J Appl Physiol, 66: 217-222. 041817
Krishna, M. T; Springall, D.; Meng, Q.-H.; Withers, N.; Macleod, D.; Biscione, G; Frew, A.; Polak, J.; Holgate, S. (1997).
Effects of ozone on epithelium and sensory nerves in the bronchial mucosa of healthy humans. Am J Respir Crit
Care Med, 156: 943-950. http://ajrccm.atsjournals.Org/cgi/content/full/156/3/943. 084262
March 2011 5-70 DRAFT - DO NOT CITE OR QUOTE
-------
Larsen, S. T.; Matsubara, S.; McConville, G; Poulsen, S. S.; Gelfand, E. W. (2010). Ozone increases airway hyperreactivity
and mucus hyperproduction in mice previously exposed to allergen. J Toxicol Environ Health A, 73: 738-747.
http://dx.doi.org/10.1080/15287391003614034628560
Larson, S. D.; Schelegle, E. S.; Walby, W. R; Gershwin, L. J.; Fanuccihi, M. V; Evans, M. J.; Joad, J. P.; Tarkington, B. K.;
Hyde, D. M.; Plopper, C. G. (2004). Postnatal remodeling of the neural components of the epithelial-me senchymal
trophic unit in the proximal airways of infant rhesus monkeys exposed to ozone and allergen. Toxicol Appl
Pharmacol, 194: 211-220. 057062
Last, J. A.; Gohil, K.; Mathrani, V. C.; Kenyon, N. J. (2005). Systemic responses to inhaled ozone in mice: cachexia and
down-regulation of liver xenobiotic metabolizing genes. Toxicol Appl Pharmacol, 208: 117-126.
http://dx.doi.0rg/10.1016/i.taap.2005.02.001 596400
Lay, J. C.; Alexis, N. E.; Kleeberger, S. R.; Roubey, R. A.; Harris, B. D.; Bromberg, P. A.; Hazucha, M. J.; Devlin, R. B.;
Peden, D. B. (2007). Ozone enhances markers of innate immunity and antigen presentation on airway monocytes in
healthy individuals. J Allergy Clin Immunol, 120: 719-22. http://dx.doi.Org/10.1016/j.jaci.2007.05.005 196610
Li, Y.-R; Gauderman, W. J.; Avol, E.; Dubeau, L.; Gilliland, R D. (2006). Associations of tumor necrosis factor G-308A
with childhood asthma and wheezing. Am J Respir Crit Care Med, 173: 970-976.
http://dx.doi.org/10.1164/rccm.200508-1256OC090972
Li, Z.; Potts, E. N.; Piantadosi, C. A.; Foster, W. M.; Hollingsworth, J. W. (2010). Hyaluronan fragments contribute to the
ozone-primed immune response to lipopolysaccharide. J Immunol, 185: 6891-6898.
http://dx.doi.org/10.4049/jimmunol. 1000283 670282
Linn, W. S.; Buckley, R. D.; Spier, C. E.; Blessey, R. L.; Jones, M. P.; Fischer, D. A.; Hackney, J. D. (1978). Health effects
of ozone exposure in asthmatics. Am Rev Respir Dis, 117: 835-843. 038874
Linn, W. S.; Medway, D. A.; Anzar, U. T; Valencia, L. M.; Spier, C. E.; Tsao FS-D; Fischer, D. A.; Hackney, J. D. (1982).
Persistence of adaptation to ozone in volunteers exposed repeatedly for six weeks. Am Rev Respir Dis, 125: 491-
495. 039646
London, S. J. (2007). Gene-air pollution interactions in asthma. Proc Am Thorac Soc, 4: 217-220. 093279
Long, N. C.; Suh, J.; Morrow, J. D.; Schiestl, R. H.; Krishna Murthy, G. G; Brain, J. D.; Frei, B. (2001). Ozone causes lipid
peroxidation but little antioxidant depletion in exercising and nonexercising hamsters. J Appl Physiol, 91: 1694-
1700. 057301
Longphre, M.; Zhang, L.-Y.; Harkema, J. R.; Kleeberger, S. R. (1999). Ozone-induced pulmonary inflammation and
epithelial proliferation are partially mediated by PAF. J Appl Physiol, 86: 341-349. 001199
Matsumura, Y (1970). The effects of ozone, nitrogen dioxide, and sulfur dioxide on the experimentally induced allergic
respiratory disorder in guinea pigs. II. The effects of ozone on the absorption and the retention of antigen in the
lung. Am Rev Respir Dis, 102: 438-443. 050626
McBride, J. T. (1992). Architecture of the tracheobronchial tree. In RA Parent (Ed.), Comparative biology of the normal
lung (pp. 49-61). Boca Raton, FL: CRC Press. 078532
McKinney, W. J.; Jaskot, R. H.; Richards, J. H.; Costa, D. L.; Dreher, K. L. (1998). Cytokine mediation of ozone-induced
pulmonary adaptation. Am J Respir Cell Mol Biol, 18: 696-705. 086751
Mikerov, A. N.; Gan, X.; Umstead, T. M.; Miller, L.; Chinchilli, V. M.; Phelps, D. S.; Floras, J. (2008). Sex differences in
the impact of ozone on survival and alveolar macrophage function of mice after Klebsiella pneumoniae infection.
Respir Res, 9: 24. http://dx.doi.org/10.1186/1465-9921-9-24 597493
Mikerov, A. N.; Haque, R.; Gan, X.; Guo, X.; Phelps, D. S.; Floras, J. (2008). Ablation of SP-A has a negative impact on
the susceptibility of mice to Klebsiella pneumoniae infection after ozone exposure: Sex differences. Respir Res, 9:
77. http://dx.doi.org/10.1186/1465-9921-9-77 201537
Mikerov, A. N.; Umstead, T. M.; Gan, X.; Huang, W.; Guo, X.; Wang, G; Phelps, D. S.; Floras, J. (2008). Impact of ozone
exposure on the phagocytic activity of human surfactant protein A (SP-A) and SP-A variants. Am J Physiol Lung
Cell Mol Physiol, 294: L121-L130. http://dx.doi.org/10.1152/ajplung.00288.2007 596401
Miller, F. J.; Overton, J. H. Jr; Jaskot, R. H.; Menzel, D. B. (1985). Amodel of the regional uptake of gaseous pollutants in
the lung: I. The sensitivity of the uptake of ozone in the human lung to lower respiratory tract secretions and
exercise. Toxicol Appl Pharmacol, 79: 11-27. http://dx.doi.org/10.1016/0041-008X(85)90364-3 040307
March 2011 5-71 DRAFT - DO NOT CITE OR QUOTE
-------
Miller, F. I; Overton, J. H.; Gerrity, T. R.; Graham, R. C. (1988). Interspecies dosimetry of reactive gases. In U Mohr; D
Dungworth; R McClellan; G Kimmerle; W Stober; J Lewkowski (Eds.), Inhalation Toxicology: The Design and
Interpretation of Inhalation Studies and their use in Risk Assessment (pp. 139-155). New York, NY: Springer-Verlag
Inc. 041545
Miller, L. A.; Gerriets, J. E.; Tyler, N. K.; Abel, K.; Schelegle, E. S.; Plopper, C. G; Hyde, D. M. (2009). Ozone and
allergen exposure during postnatal development alters the frequency and airway distribution of CD25+ cells in
infant rhesus monkeys. Toxicol Appl Pharmacol, 236: 39-48. http://dx.doi.org/10.10167j.taap.2008.12.031 596406
Molfmo, N. A.; Wright, S. C.; Katz, L; Tarlo, S.; Silverman, F.; McClean, P. A.; Szalai, J. P.; Raizenne, M.; Slutsky, A. S.;
Zamel, N. (1991). Effect of low concentrations of ozone on inhaled allergen responses in asthmatic subjects.
Lancet, 338: 199-203. 042379
Mudway, I. S.; Behndig, A. F.; Helleday, R.; Pourazar, J.; Frew, A. J.; Kelly, F. J.; Blomberg, A. (2006). Vitamin
supplementation does not protect against symptoms in ozone-responsive subjects. Free Radic Biol Med, 40: 1702-
1712.196536
Mudway, I. S.; Blomberg, A.; Frew, A. J.; Holgate, S. T; Sandstrom, T; Kelly, F. J. (1999). Antioxidant consumption and
repletion kinetics in nasal lavage fluid following exposure of healthy human volunteers to ozone. Eur Respir J, 13:
1429-1438. 001270
Mudway, I. S.; Housley, D.; Eccles, R.; Richards, R. J.; Datta, A. K.; Tetley, T. D.; Kelly, F. J. (1996). Differential depletion
of human respiratory tract antioxidants in response to ozone challenge. Free Radic Res, 25: 499-513. 080730
Mudway, I. S.; Kelly, F. J. (1998). Modeling the interactions of ozone with pulmonary epithelial lining fluid antioxidants.
Toxicol Appl Pharmacol, 148: 91-100. 000273
Mudway, I. S.; Kelly, F. J. (2000). Ozone and the lung: a sensitive issue. Mol Aspects Med, 21: 1-48. 010452
Mudway, I. S.; Kelly, F. J. (2004). An investigation of inhaled ozone dose and the magnitude of airway inflammation in
healthy adults: Online data supplement. Am J Respir Crit Care Med, 169: 1089-1095.
http://dx.doi.org/10.1164/rccm.200309-1325PP399328
Mudway, I. S.; Krishna, M. T; Frew, A. J.; MacLeod, D.; Sandstrom, T; Holgate, S. T; Kelly, F. J. (1999). Compromised
concentrations of ascorbate in fluid lining the respiratory tract in human subjects after exposure to ozone. Occup
Environ Med, 56: 473-481.011833
Mudway, I. S.; Stenfors, N.; Blomberg, A.; Helleday, R.; Dunster, C.; Marklund, S. L.; Frew, A. J.; Sandstrom, T; Kelly, F.
J. (2001). Differences in basal airway antioxidant concentrations are not predictive of individual responsiveness to
ozone: A comparison of healthy and mild asthmatic subjects. Free Radic Biol Med, 31: 962-974. 025327
Murphy, R. C.; Johnson, K. M. (2008). Cholesterol, reactive oxygen species, and the formation of biologically active
mediators. J Biol Chem, 283: 15521-15525. http://dx.doi.org/10.1074/jbc.R700049200 197792
Nodelman, V; Ultman, J. S. (1999). Longitudinal distribution of chlorine absorption in human airways: A comparison to
ozone absorption. J Appl Physiol, 87: 2073-2080. http://jap.physiology.Org/cgi/content/full/87/6/2073. 015112
Noviski, N.; Brewer, J. P.; Skornik, W. A.; Galli, S. J.; Drazen, J. M.; Martin, T. R. (1999). Mast cell activation is not
required for induction of airway hyperresponsiveness by ozone in mice. J Appl Physiol, 86: 202-210. 001198
O'Byrne, P. M.; Walters, E. H.; Aizawa, H.; Fabbri, L. M.; Holtzman, M. J.; Nadel, J. A. (1984). Indomethacin inhibits the
airway hyperresponsiveness but not the neutrophil influx induced by ozone in dogs. Am Rev Respir Dis, 130: 220-
224. 040066
O'Byrne, P.; Walters, E.; Gold, B.; Aizawa, H.; Fabbri, L.; Alpert, S.; Nadel, J.; Holtzman, M. (1983). Neutrophil depletion
inhibits airway hyperresponsiveness induced by ozone exposure. Am Rev Respir Dis, 130: 214-219. 041366
Oslund, K. L.; Hyde, D. M.; Putney, L. F.; Alfaro, M. F.; Walby, W. F.; Tyler, N. K.; Schelegle, E. S. (2008). Activation of
neurokinin-1 receptors during ozone inhalation contributes to epithelial injury and repair. Am J Respir Cell Mol
Biol, 39: 279-288. http://dx.doi.org/10.1165/rcmb.2008-0009OC 195654
Oslund, K. L.; Hyde, D. M.; Putney, L. F.; Alfaro, M. F.; Walby, W. F.; Tyler, N. K.; Schelegle, E. S. (2009). Activation of
calcitonin gene-related peptide receptor during ozone inhalation contributes to airway epithelial injury and repair.
Toxicol Pathol, 37: 805-813. http://dx.doi.org/10.1177/0192623309345691 201539
March 2011 5-72 DRAFT - DO NOT CITE OR QUOTE
-------
Overton, J. H.; Graham, R. C.; Menache, M. G; Mercer, R. R.; Miller, F. J. (1996). Influence of tracheobronchial region
expansion and volume on reactive gas uptake and interspecies dose extrapolations. Inhal Toxicol, 8: 723-745.
080733
Park, J. W.; Taube, C.; Swasey, C.; Kodama, T.; Joetham, A.; Balhorn, A.; Takeda, K.; Miyahara, N.; Allen, C. B.;
Dakhama, A.; Kim, S. H.; Dinarello, C. A.; Gelfand, E. W. (2004). Interleukin-1 receptor antagonist attenuates
airway hyperresponsiveness following exposure to ozone. Am J Respir Cell Mol Biol, 30: 830-836.
http://dx.doi.org/10.1165/rcmb.2003-0373OC644864
Passannante, A. N.; Hazucha, M. J.; Bromberg, P. A.; Seal, E.; Folinsbee, L.; Koch, G. (1998). Nociceptive mechanisms
modulate ozone-induced human lung function decrements. J Appl Physiol, 85: 1863-1870. 030114
Pearson, A. C.; Bhalla, D. K. (1997). Effects of ozone on macrophage adhesion in vitro and epithelial and inflammatory
responses in vivo: The role of cytokines. J Toxicol Environ Health, 50: 143-157. 082686
Peden, D. B.; Setzer, R. W. Jr; Devlin, R. B. (1995). Ozone exposure has both a priming effect on allergen-induced
responses and an intrinsic inflammatory action in the nasal airways of perennially allergic asthmatics. Am J Respir
CritCareMed, 151: 1336-1345. 076189
Perepu, R. S.; Garcia, C.; Dostal, D.; Sethi, R. (2010). Enhanced death signaling in ozone-exposed ischemic-reperfused
hearts. Mol Cell Biochem, 336: 55-64. http://dx.doi.org/10.1007/sll010-009-0265-4 385020
Perez-Gil, J. (2008). Structure of pulmonary surfactant membranes and films: The role of proteins and lipid-protein
interactions. Biochim Biophys Acta, 1778: 1676-1695. http://dx.doi.Org/10.1016/j.bbamem.2008.05.003 198890
Pichavant, M.; Goya, S.; Meyer, E. H.; Johnston, R. A.; Kim, H. Y; Matangkasombut, P.; Zhu, M.; Iwakura, Y; Savage, P.
B.; DeKruyff, R. H.; Shore, S. A.; Umetsu, D. T. (2008). Ozone exposure in a mouse model induces airway
hyperreactivity that requires the presence of natural killer T cells and IL-17. J Exp Med, 205: 385-393.
http://dx.doi.org/10.1084/iem.20071507596409
Picher, M.; Burch, L. H.; Boucher, R. C. (2004). Metabolism of P2 receptor agonists in human airways: Implications for
mucociliary clearance and cystic fibrosis. J Biol Chem, 279: 20234-20241.
http://dx.doi.org/10.1074/ibc.M400305200644780
Plopper, C. G; Hatch, G. E.; Wong, V; Duan, X.; Weir, A. J.; Tarkington, B. K.; Devlin, R. B.; Becker, S.; Buckpitt, A. R.
(1998). Relationship of inhaled ozone concentration to acute tracheobronchial epithelial injury, site-specific ozone
dose and glutathione depletion in rhesus monkeys. Am J Respir Cell Mol Biol, 19: 387-399. 087203
Plopper, C. G; Mango, G. W.; Hatch, G. E.; Wong, V. J.; Toskala, E.; Reynolds, S. D.; Tarkington, B. K.; Stripp, B. R.
(2006). Elevation of susceptibility to ozone-induced acute tracheobronchial injury in transgenic mice deficient in
Clara cell secretory protein. Toxicol Appl Pharmacol, 213: 74-85. http://dx.doi.Org/10.1016/j.taap.2005.09.003
596410
Plopper, C. G; Smiley-Jewell, S. M.; Miller, L. A.; Fanucchi, M. V; Evans, M. J.; Buckpitt, A. R.; Avdalovic, M.;
Gershwin, L. J.; Joad, J. P.; Kajekar, R.; Larson, S.; Pinkerton, K. E.; Van Winkle, L. S.; Schelegle, E. S.; Pieczarka,
E. M.; Wu, R.; Hyde, D. M. (2007). Asthma/allergic airways disease: Does postnatal exposure to environmental
toxicants promote airway pathobiology? Toxicol Pathol, 35: 97-110. http://dx.doi.org/10.1080/01926230601132030
596412
Postlethwait, E. M.; Cueto, R.; Velsor, L. W.; Pryor, W. A. (1998). 03-induced formation of bioactive lipids: Estimated
surface concentrations and lining layer effects. Am J Physiol, 274: L1006-L1016. 086754
Postlethwait, E. M.; Joad, J. P.; Hyde, D. M.; Schelegle, E. S.; Brie, J. M.; Weir, A. J.; Putney, L. F.; Wong, V. J.; Velsor, L.
W.; Plopper, C. G. (2000). Three-dimensional mapping of ozone-induced acute cytotoxicity in tracheobronchial
airways of isolated perfused rat lung. Am J Respir Cell Mol Biol, 22: 191-199. 003000
Postlethwait, E. M.; Langford, S. D.; Bidani, A. (1994). Determinants of inhaled ozone absorption in isolated rat lungs.
Toxicol Appl Pharmacol, 125: 77-89. 044219
Postlethwait, E. M.; Ultman, J. S. (2001). Airspace surface chemistry mediates OS-induced lung injury. Hum Ecol Risk
Assess, 7: 1145-1159. http://dx.doi.org/10.1080/20018091094907 196805
Pryor, W. A. (1976). Free radical reactions in biology: Initiation of lipid autoxidation by ozone and nitrogen dioxide.
Environ Health Perspect, 16: 180-181. 038940
March 2011 5-73 DRAFT - DO NOT CITE OR QUOTE
-------
Pryor, W. A. (1992). How far does ozone penetrate into the pulmonary air/tissue boundary before it reacts? Free Radic Biol
Med, 12: 83-88. http://dx.doi.org/10.1016/0891-5849(92)90060-T 042725
Pryor, W. A. (1994). Mechanisms of radical formation from reactions of ozone with target molecules in the lung. Free
Radic Biol Med, 17: 451-465. 075987
Pryor, W. A.; Bermudez, E.; Cueto, R.; Squadrito, G. L. (1996). Detection of aldehydes in bronchoalveolar lavage of rats
exposed to ozone. Toxicol Sci, 34: 148-156. 082697
Pryor, W. A.; Das, B.; Church, D. F. (1991). The ozonation of unsaturated fatty acids: Aldehydes and hydrogen peroxide as
products and possible mediators of ozone toxicity. Chem Res Toxicol, 4: 341-348. 042482
Pryor, W. A.; Giamalva, D. H.; Church, D. F. (1984). Kinetics of ozonation. 2. Amino acids and model compounds in water
and comparisons to rates in nonpolar solvents. J Am Chem Soc, 106: 7094-7100 . 595130
Pulfer, M. K.; Murphy, R. C. (2004). Formation of biologically active oxysterols during ozonolysis of cholesterol present in
lung surfactant. J Biol Chem, 279: 26331-26338. 076673
Pulfer, M. K.; Taube, C.; Gelfand, E.; Murphy, R. C. (2005). Ozone exposure in vivo and formation of biologically active
oxysterols in the lung. J Pharmacol Exp Ther, 312: 256-264. 076663
Rashba-Step, J.; Tatoyan, A.; Duncan, R.; Ann, D.; Pushpa-Rehka, T R.; Sevanian, A. (1997). Phospholipid peroxidation
induces cytosolic phospholipase A2 activity: Membrane effects versus enzyme phosphorylation. Arch Biochem
Biophys, 343: 44-54. http://dx.doi.org/10.1006/abbi.1997.0134 628562
Reeser, W. H.; Lee, G. M.; Taylor, A.; Wang, L.; Arnold, S. F.; Ultman, J. S.; Ben-Jebria, A. (2005). Uptake of ozone in
human lungs and its relationship to local physiological response. Inhal Toxicol, 17: 699-707.
http://dx.doi.org/10.1080/08958370500224433 195718
Rigas, M. L.; Ben-Jebria, A.; Ultman, J. S. (1997). Longitudinal distribution of ozone absorption in the lung: Effects of
nitrogen dioxide, sulfur dioxide, and ozone exposures. Arch Environ Occup Health, 52: 173-178. 083602
Rigas, M. L.; Catlin, S. N.; Ben-Jebria, A.; Ultman, J. S. (2000). Ozone uptake in the intact human respiratory tract:
Relationship between inhaled dose and actual dose. J Appl Physiol, 88: 2015-2022. 010454
Romieu, I.; Meneses, F.; Ramirez, M.; Ruiz, S.; Padilla, R. P.; Sienra, J. J.; Gerber, M.; Grievink, L.; Dekker, R.; Walda, L;
Brunekreef, B. (1998). Antioxidant supplementation and respiratory functions among workers exposed to high
levels of ozone. Am J Respir Crit Care Med, 158: 226-232. 086756
Romieu, I.; Sienra-Monge, J. J.; Ramirez-Aguilar, M.; Moreno-Macias, H.; Reyes-Ruiz, N. L; Estela del Rio-Navarro, B.;
Hernandez-Avila, M.; London, S. J. (2004). Genetic polymorphism of GSTM1 and antioxidant supplementation
influence lung function in relation to ozone exposure in asthmatic children in Mexico City. Thorax, 59: 8-10.
056796
Romieu, I.; Sienra-Monge, J. J.; Ramirez-Aguilar, M.; Tellez-Rojo, M. M.; Moreno-Macias, H.; Reyes-Ruiz, N. L; Del
Rio-Navarro, B. E.; Ruiz-Navarro, M. X.; Hatch, G; Slade, R.; Hernandez-Avila, M. (2002). Antioxidant
supplementation and lung functions among children with asthma exposed to high levels of air pollutants. Am J
Respir Crit Care Med, 166: 703-709. 034711
Roux, E.; Hyvelin, J.-M.; Savineau, J.-P; Marthan, R. (1999). Human isolated airway contraction Interaction between air
pollutants and passive sensitization. Am J Respir Crit Care Med, 160: 439-445. 001264
Sagai, M.; Arakawa, K.; Ichinose, T; Shimojo, N. (1987). Biochemical effects on combined gases of nitrogen dioxide and
ozone I Species differences of lipid peroxides and phospholipids in lungs. Toxicology, 46: 251-265.
http://dx.doi.org/10.1016/0300-483X(87)90207-1 041704
Samet, J. M.; Hatch, G. E.; Horstman, D.; Steck-Scott, S.; Arab, L.; Bromberg, P. A.; Levine, M.; McDonnell, W. F.;
Devlin, R. B. (2001). Effect of antioxidant supplementation on ozone-induced lung injury in human subjects. Am J
Respir Crit Care Med, 164: 819-825. 019034
Santiago, L. Y; Hann, M. C.; Ben-Jebria, A.; Ultman, J. S. (2001). Ozone absorption in the human nose during
unidirectional airflow. J Appl Physiol, 91: 725-732. http://jap.physiology.Org/cgi/content/abstract/91/2/725. 019841
Sarangapani, R.; Gentry, P. R.; Covington, T. R.; Teeguarden, J. G; Clewell HJ 3rd (2003). Evaluation of the potential
impact of age- and gender-specific lung morphology and ventilation rate on the dosimetry of vapors. Inhal Toxicol,
15: 987-1016. http://dx.doi.org/10.1080/713857276 054581
March 2011 5-74 DRAFT - DO NOT CITE OR QUOTE
-------
Sathishkumar, K.; Gao, X.; Raghavamenon, A. C.; Parinandi, N.; Pryor, W. A.; Uppu, R. M. (2009). Cholesterol
secoaldehyde induces apoptosis in H9c2 cardiomyoblasts through reactive oxygen species involving mitochondrial
and death receptor pathways. Free Radic Biol Med, 47: 548-558.
http://dx.doi.0rg/10.1016/i.freeradbiomed.2009.05.020201549
Sathishkumar, K.; Murthy, S. N.; Uppu, R. M. (2007). Cytotoxic effects of oxysterols produced during ozonolysis of
cholesterol in murine GT1-7 hypothalamic neurons. Free Radic Res, 41: 82-88.
http://dx.doi.org/10.1080/10715760600950566 197785
Sathishkumar, K.; Xi, X.; Martin, R.; Uppu, R. M. (2007). Cholesterol secoaldehyde, an ozonation product of cholesterol,
induces amyloid aggregation and apoptosis in murine GT1-7 hypothalamic neurons. J Alzheimers Dis, 11: 261-274.
097758
Sato, S.; Shimura, S.; Hirose, T; Maeda, S.; Kawakami, M.; Takishima, T; Kimura, S. (1980). Effects of long-term ozone
exposure and dietary vitamin E in rats. Tohoku J Exp Med, 130: 117-128.
http://www.journalarchive.jst.go.jp/english/jnlabstract_en.php?cdjournal=tjeml920&cdvol=130&noissue=2&startp
age=117. 039738
Sawyer, K.; Brown, J.; HazuchaM; Bennett, W. D. (2007). The effect of exercise on nasal uptake of ozone in healthy
human adults. J Appl Physiol, 102: 1380-1386. http://dx.doi.org/10.1152/iapplphvsiol.00269.2006 195142
Scannell, C.; Chen, L.; Aris, R. M.; Tager, I.; Christian, D.; Ferrando, R.; Welch, B.; Kelly, T.; Balmes, J. R. (1996).
Greater ozone-induced inflammatory responses in subjects with asthma. Am J Respir Crit Care Med, 154: 24-29.
080755
Schelegle, E. S.; Adams, W. C.; Siefkin, A. D. (1987). Indomethacin pretreatment reduces ozone-induced pulmonary
function decrements in human subjects. Am Rev Respir Dis, 136: 1350-1354. 041706
Schelegle, E. S.; Carl, M. L.; Coleridge, H. M.; Coleridge, J. C. G; Green, J. F. (1993). Contribution of vagal afferents to
respiratory reflexes evoked by acute inhalation of ozone in dogs. J Appl Physiol, 74: 2338-2344. 039203
Schelegle, E. S.; Miller, L. A.; Gershwin, L. J.; Fanucchi, M. V; Van Winkle, L. S.; Gerriets, J. E.; Walby, W. F.; Mitchell,
V; Tarkington, B. K.; Wong, V. J.; Baker, G. L.; Pantle, L. M.; Joad, J. P.; Pinkerton, K. E.; Wu, R.; Evans, M. J.;
Hyde, D. M.; Plopper, C. G. (2003). Repeated episodes of ozone inhalation amplifies the effects of allergen
sensitization and inhalation on airway immune and structural development in Rhesus monkeys. Toxicol Appl
Pharmacol, 191: 74-85. 053778
Schelegle, E. S.; Siefkin, A. D.; McDonald, R. J. (1991). Time course of ozone-induced neutrophilia in normal humans. Am
J Respir Crit Care Med, 143: 1353-1358. 042491
Schelegle, E. S.; Walby, W. F.; Adams, W. C. (2007). Time course of ozone-induced changes in breathing pattern in healthy
exercising humans. J Appl Physiol, 102: 688-697. http://dx.doi.org/10.1152/iapplphvsiol.00141.2006 195841
Schroter, R. C.; Sudlow, M. F. (1969). Flow patterns in models of the human bronchial airways. Respir Physiol Neurobiol,
7: 341-355.071359
Seltzer, J.; Bigby, B. G; Stulbarg, M.; Holtzman, M. J.; Nadel, J. A.; Ueki, I. F.; Leikauf, G. D.; Goetzl, E. J.; Boushey, H.
A. (1986). OS-induced change in bronchial reactivity to methacholine and airway inflammation in humans. J Appl
Physiol, 60: 1321-1326. 040383
Servais, S.; Boussouar, A.; Molnar, A.; Douki, T; Pequignot, J. M.; Favier, R. (2005). Age-related sensitivity to lung
oxidative stress during ozone exposure. Free Radic Res, 39: 305-316.
http://dx.doi.org/10.1080/10715760400011098 195667
Shore, S. A.; Lang, J. E.; Kasahara, D. L; Lu, F. L.; Verbout, N. G; Si, H.; Williams, E. S.; Terry, R. D.; Lee, A.; Johnston,
R. A. (2009). Pulmonary responses to subacute ozone exposure in obese vs. lean mice. J Appl Physiol, 107: 1445-
1452. http://dx.doi.org/10.1152/japplphysiol.00456.2009 201551
Shore, S. A.; Schwartzman, I. N.; Le Blanc, B.; Krishna Murthy, G. G; Doerschuk, C. M. (2001). Tumor necrosis factor
receptor 2 contributes to ozone-induced airway hyperresponsiveness in mice. Am J Respir Crit Care Med, 164: 602-
607.018993
Sienra-Monge, J. J.; Ramirez-Aguilar, M.; Moreno-Macias, H.; Reyes-Ruiz, N. I.; Del Rio-Navarro, B. E.; Ruiz-Navarro,
M. X.; Hatch, G; Crissman, K.; Slade, R.; Devlin, R. B.; Romieu, I. (2004). Antioxidant supplementation and nasal
inflammatory responses among young asthmatics exposed to high levels of ozone. Clin Exp Immunol, 138: 317-
322. http://dx.doi.0rg/10.llll/i.1365-2249.2004.02606.x 196422
March 2011 5-75 DRAFT - DO NOT CITE OR QUOTE
-------
Slade, R.; Crissman, K.; Norwood, I; Hatch, G. (1993). Comparison of antioxidant substances in bronchoalveolar lavage
cells and fluid from humans, guinea pigs, and rats. Exp Lung Res, 19: 469-484. 042865
Slade, R.; Highfill, J. W.; Hatch, G. E. (1989). Effects of depletion of ascorbic acid or nonprotein sulfhydryls on the acute
inhalation toxicity of nitrogen dioxide, ozone, and phosgene. Inhal Toxicol, 1: 261-271. 059465
Slade, R.; Watkinson, W. R; Hatch, G. E. (1997). Mouse strain differences in ozone dosimetry and body temperature
changes. Am J Physiol, 272: L73-L77. 082708
Stenfors, N.; Bosson, J.; Helleday, R.; Behndig, A. F.; Pourazar, J.; Tornqvist, H.; Kelly, F. J.; Frew, A. J.; Sandstrom, T.;
Mudway, I. S.; Blomber, A. (2010). Ozone exposure enhances mast-cell inflammation in asthmatic airways despite
inhaled corticosteroid therapy. Inhal Toxicol, 22: 133-139. http://dx.doi.org/10.3109/08958370903005736 386512
Stenfors, N.; Pourazar, J.; Blomberg, A.; Krishna, M. T.; Mudway, I.; Helleday, R.; Kelly, F. J.; Frew, A. J.; Sandstrom, T.
(2002). Effect of ozone on bronchial mucosal inflammation in asthmatic and healthy subjects. Respir Med, 96: 352-
358. 030473
Sterner-Kock, A.; Kock, M.; Braun, R.; Hyde, D. M. (2000). Ozone-induced epithelial injury in the ferret is similar to
nonhuman primates. Am J Respir Crit Care Med, 162: 1152-1156. 013033
Sun, J.; Koto, H.; Chung, K. F. (1997). Interaction of ozone and allergen challenges on bronchial responsiveness and
inflammation in sensitised guinea pigs. Int Arch Allergy Immunol, 112: 191-195. 082724
Tarrant, J. M. (2010). Blood cytokines as biomarkers of in vivo toxicity in preclinical safety assessment: Considerations for
their use. Toxicol Sci, 117: 4-16. 644810
Taylor, A. B.; Borhan, A.; Ultman, J. S. (2007). Three-dimensional simulations of reactive gas uptake in single airway
bifurcations. Ann Biomed Eng, 35: 235-249. http://dx.doi.org/10.1007/sl0439-006-9195-4 195717
Taylor, A. B.; Lee, G. M.; Nellore, K.; Ben-Jebria, A.; Ultman, J. S. (2006). Changes in the carbon dioxide expirogram in
response to ozone exposure. Toxicol Appl Pharmacol, 213: 1-9. http://dx.doi.Org/10.1016/j.taap.2005.09.009
195731
Taylor-Clark, T. E.; McAlexander, M. A.; Nassenstein, C.; Sheardown, S. A.; Wilson, S.; Thornton, J.; Carr, M. J.; Undem,
B. J. (2008). Relative contributions of TRPA1 and TRPV1 channels in the activation of vagal bronchopulmonary C-
fibres by the endogenous autacoid 4-oxononenal. J Physiol, 586: 3447-3459.
http://dx.doi.org/10.1113/iphvsiol.2008.153585628565
Taylor-Clark, T. E.; Undem, B. J. (2010). Ozone activates airway nerves via the selective stimulation of TRPA1 ion
channels. J Physiol, 588: 423-433. http://dx.doi.org/10.1113/jphysiol.2009.183301 377143
Tepper, J. S.; Costa, D. L.; Fitzgerald, S.; Doerfler, D. L.; Bromberg, P. A. (1993). Role of tachykinins in ozone-induced
acute lung injury in guinea pigs. JAppl Physiol, 75: 1404-1411.
http://jap.physiology.org/cgi/content/abstract/75/3/1404. 628570
Tepper, J. S.; Costa, D. L.; Lehmann, J. R.; Weber, M. F.; Hatch, G. E. (1989). Unattenuated structural and biochemical
alterations in the rat lung during functional adaptation to ozone. Am J Respir Crit Care Med, 140: 493-501. 041991
Thomson, E.; Kumarathasan, P.; Goegan, P.; Aubin, R. A.; Vincent, R. (2005). Differential regulation of the lung endothelin
system by urban particulate matter and ozone. Toxicol Sci, 88: 103-113. 087554
Thomson, E.; Kumarathasan, P.; Vincent, R. (2006). Pulmonary expression of preproET-1 and preproET-3 mRNAs is
altered reciprocally in rats after inhalation of air pollutants. Exp Biol Med, 231: 979-984.
http://ebm.rsmjournals.com/cgi/content/abstract/231/6/979. 097483
Tran, M. U.; Weir, A. J.; Fanucchi, M. V; Rodriguez, A. E.; Pantle, L. M.; Smiley-Jewell, S. M.; Van Winkle, L. S.; Evans,
M. J.; Miller, L. A.; Schelegle, E. S.; Gershwin, L. J.; Hyde, D. M.; Plopper, C. G. (2004). Smooth muscle
hypertrophy in distal airways of sensitized infant rhesus monkeys exposed to house dust mite allergen. Clin Exp
Allergy, 34: 1627-1633. http://dx.doi.Org/10.llll/i.1365-2222.2004.02057.x628626
Trevisani, M.; Siemens, J.; Materazzi, S.; Bautista, D. M.; Nassini, R.; Campi, B.; Imamachi, N.; Andre, E.; Patacchini, R.;
Cottrell, G. S.; Gatti, R.; Basbaum, A. L; Bunnett, N. W.; Julius, D.; Geppetti, P. (2007). 4-Hydroxynonenal, an
endogenous aldehyde, causes pain and neurogenic inflammation through activation of the irritant receptor TRPA1.
PNAS, 104: 13519-13524. http://dx.doi.org/10.1073/pnas.0705923104628590
Tsujino, L; Kawakami, Y; Kaneko, A. (2005). Comparative simulation of gas transport in airway models of rat, dog, and
human. Inhal Toxicol, 17: 475-485. http://dx.doi.org/10.1080/08958370590964476 195842
March 2011 5-76 DRAFT - DO NOT CITE OR QUOTE
-------
U.S. EPA (1996). Air quality criteria for ozone and related photochemical oxidants (Report No. EPA/600/P-93/004AF).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development. (NTIS
No. PB94173127) 017831
U.S. EPA (2006). Air quality criteria for ozone and related photochemical oxidants (Report No. EPA/600/R-05/004AF).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=149923. 088089
Ultman, J. S.; Ben-Jebria, A.; Arnold, S. F. (2004). Uptake distribution of ozone in human lungs: intersubject variability in
physiologic response. University Park, PA: Health Effects Institute. http://pubs.healtheffects.org/view.php?id=70.
057197
Ultman, J. S.; Ben-Jebria, A.; Hu, S.-C. (1994). Noninvasive determination of respiratory ozone absorption: The bolus-
response method (Report No. research report no. 69). Cambridge, MA: Health Effects Institute. 041746
Uppu, R. M.; Cueto, R.; Squadrito, G. L.; Pryor, W. A. (1995). What does ozone react with at the air/lung interface? Model
studies using human red blood cell membranes. Arch Biochem Biophys, 319: 257-266. 076212
Urch, B.; Speck, M.; Corey, P.; Wasserstein, D.; Manno, M.; Lukic, K. Z.; Brook, J. R.; Liu, L.; Coull, B.; Schwartz, J.;
Gold, D. R.; Silverman, F. (2010). Concentrated ambient fine particles and not ozone induce a systemic interleukin-
6 response in humans. Inhal Toxicol, 22: 210-218. http://dx.doi.org/10.3109/08958370903173666 387113
Vagaggini, B.; Taccola, M.; Clanchetti, S.; Carnevali, S.; Bartoli, M. L.; Bacci, E.; Dente, F. L.; Di Franco, A.; Giannini,
D.; Paggiaro, P. L. (2002). Ozone exposure increases eosinophilic airway response induced by previous allergen
challenge. Am J Respir Crit Care Med, 166: 1073-1077. 035191
Van Bree, L.; Dormans, J. A. M. A.; Koren, H. S.; Devlin, R. B.; Rombout, P. J. A. (2002). Attenuation and recovery of
pulmonary injury in rats following short-term, repeated daily exposure to ozone. Inhal Toxicol, 14: 883-900.
035452
Vancza, E. M.; Galdanes, K.; Gunnison, A.; Hatch, G; Gordon, T (2009). Age, strain, and gender as factors for increased
sensitivity of the mouse lung to inhaled ozone. Toxicol Sci, 107: 535-543. http://dx.doi.org/10.1093/toxsci/kfn253
596419
Vasu, V. T; Oommen, S.; Lim, Y; Valacchi, G; Hobson, B.; Eiserich, J. P.; Leonard, S. W.; Traber, M. G; Cross, C. E.;
Gohil, K. (2010). Modulation of ozone-sensitive genes in alpha-tocopherol transfer protein null mice. Inhal Toxicol,
22: 1-16.201561
Verhein, K. C.; Hazari, M. S.; Moulton, B. C.; Jacoby, I. W.; Jacoby, D. B.; Fryer, A. D. (2011). Three days after a single
exposure to ozone the mechanism of airway hyperreactivity is dependent upon substance P and nerve growth factor.
Am J Physiol Lung Cell Mol Physiol, 300: L176-L184. http://dx.doi.org/10.1152/ajplung.00060.2010 670295
Vesely, K. R.; Schelegle, E. S.; Stovall, M. Y; Harkema, J. R.; Green, J. F.; Hyde, D. M. (1999). Breathing pattern response
and epithelial labeling in ozone-induced airway injury in neutrophil-depleted rats. Am J Respir Cell Mol Biol, 20:
699-709. 051045
Vincent, R.; Vu, D.; Hatch, G; Poon, R.; Dreher, K.; Guenette, J.; Bjarnason, S.; Potvin, M.; Norwood, J.; McMullen, E.
(1996). Sensitivity of lungs of aging Fischer 344 rats to ozone: assessment by bronchoalveolar lavage. Am J
Physiol, 271: L555-L565. 080778
Vivier, E.; Raulet, D. H.; Moretta, A.; Caligiuri, M. A.; Zitvogel, L.; Lanier, L. L.; Yokoyama, W. M.; Ugolini, S. (2011).
Innate or adaptive immunity? The example of natural killer cells. Science, 331: 44-49.
http://dx.doi.org/10.1126/science.1198687676697
Voynow, J. A.; Fischer, B. M.; Zheng, S.; Potts, E. N.; Grover, A. R.; Jaiswal, A. K.; Ohio, A. J.; Foster, W. M. (2009).
NAD(P)H quinone oxidoreductase 1 is essential for ozone-induced oxidative stress in mice and humans. Am J
Respir Cell Mol Biol, 41: 107-113. 194311
Wagner, J. G; Harkema, J. R.; Jiang, Q.; Illek, B.; Ames, B. N.; Peden, D. B. (2009). Gamma-tocopherol attenuates ozone-
induced exacerbation of allergic rhinosinusitis in rats. Toxicol Pathol, 37: 481-491.
http://dx.doi.org/10.1177/0192623309335630201574
Wagner, J. G; Hotchkiss, J. A.; Harkema, J. R. (2002). Enhancement of nasal inflammatory and epithelial responses after
ozone and allergen coexposure in brown Norway rats. Toxicol Sci, 67: 284-294.
http://toxsci.oxfordjournals.Org/cgi/reprint/67/2/284.pdf 026079
March 2011 5-77 DRAFT - DO NOT CITE OR QUOTE
-------
Wagner, J. G; Jiang, Q.; Harkema, J. R.; Illek, B.; Patel, D. D.; Ames, B. N.; Peden, D. B. (2007). Ozone enhancement of
lower airway allergic inflammation is prevented by gamma-tocopherol. Free Radic Biol Med, 43:1176-1188.
http://dx.doi.0rg/10.1016/i.freeradbiomed.2007.07.013596420
Watkinson, W. P.; Campen, M. J.; Nolan, J. P.; Costa, D. L. (2001). Cardiovascular and systemic responses to inhaled
pollutants in rodents: effects of ozone and particulate matter. Environ Health Perspect, 109: 539-546.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1240578/. 016245
Watkinson, W. P.; Campen, M. J.; Wichers, L. B.; Nolan, J. P.; Costa, D. L. (2003). Cardiac and thermoregulatory responses
to inhaled pollutants in healthy and compromised rodents: modulation via interaction with environmental factors.
Environ Res, 92: 35-47. 050547
Weibel, E. R. (1980). Design and structure of the human lung. In Fishman, A. P (Ed.), Assessment of pulmonary function
New York, NY: McGraw-Hill. 079848
Weinmann, G G; Liu, M. C.; Proud, D.; Weidenbach-Gerbase, M.; Hubbard, W.; Frank, R. (1995). Ozone exposure in
humans: Inflammatory, small and peripheral airway responses. Am J Respir Crit Care Med, 152: 1175-1182.
http://ajrccm.atsjournals.Org/cgi/content/abstract/152/4/l 175. 077206
Weinmann, G. G; Weidenbach-Gerbase, M.; Foster, W. M.; Zacur, H.; Frank, R. (1995). Evidence for ozone-induced small-
airway dysfunction: lack of menstrual-cycle and gender effects. Am J Respir Crit Care Med, 152: 988-996. 038645
Welch, R. W.; Wang, Y; Grossman A Jr; Park, J. B.; Kirk, K. L.; Levine, M. (1995). Accumulation of vitamin C (ascorbate)
and its oxidized metabolite dehydroascorbic acid occurs by separate mechanisms. J Biol Chem, 270: 12584-12592.
http://dx.doi.org/10.1074/ibc.270.21.12584644675
Wenten, M.; Gauderman, W. J.; Berhane, K.; Lin, P. C.; Peters, J.; Gilliland, F. D. (2009). Functional variants in the
catalase and myeloperoxidase genes, ambient air pollution, and respiratory-related school absences: An example of
epistasis in gene-environment interactions. Am J Epidemiol, 170: 1494-1501. http://dx.doi.org/10.1093/aje/kwp310
597084
Wiester, M. J.; Stevens, M. A.; Menache, M. G; McKee, J. L. Jr; Gerrity, T. R. (1996). Ozone uptake in healthy adult males
during quiet breathing. Toxicol Sci, 29: 102-109. 041280
Wiester, M. J.; Tepper, J. S.; Winsett, D. W.; Crissman, K. M.; Richards, J. H.; Costa, D. L. (1996). Adaptation to ozone in
rats and its association with ascorbic acid in the lung. Toxicol Sci, 31: 56-64. 080829
Williams, A. S.; Nath, P.; Leung, S. Y; Khorasani, N.; McKenzie, A. N. J.; Adcock, I. M.; Chung, K. F. (2008).
Modulation of ozone-induced airway hyperresponsiveness and inflammation by interleukin-13. Eur Respir J, 32:
571-578. http://dx.doi.org/10.1183/09031936.00121607 597546
Williams, A. S.; Eynott, P. R.; Leung, S. Y; Nath, P.; Jupp, R.; De Sanctis, G. T.; Resnick, R.; Adcock, I. M.; Chung, K. F.
(2009). Role of cathepsin S in ozone-induced airway hyperresponsiveness and inflammation. Pulm Pharmacol Ther,
22: 27-32. http://dx.doi.Org/10.1016/i.pupt.2008.ll.002628605
Williams, A. S.; Issa, R.; Durham, A.; Leung, S. Y; Kapoun, A.; Medicherla, S.; Higgins, L. S.; Adcock, I. M.; Chung, K.
F. (2008). Role of p38 mitogen-activated protein kinase in ozone-induced airway hyperresponsiveness and
inflammation. Eur J Pharmacol, 600: 117-122. http://dx.doi.Org/10.1016/j.ejphar.2008.09.031 628607
Williams, A. S.; Issa, R.; Leung, S. Y; Nath, P.; Ferguson, G. D.; Bennett, B. L.; Adcock, I. M.; Chung, K. F. (2007).
Attenuation of ozone-induced airway inflammation and hyper-responsiveness by c-Jun NH2 terminal kinase
inhibitor SP600125. J Pharmacol Exp Ther, 322: 351-359. http://dx.doi.org/10.1124/jpet.107.121624 628609
Williams, A. S.; Leung, S. Y; Nath, P.; Khorasani, N. M.; Bhavsar, P.; Issa, R.; Mitchell, J. A.; Adcock, I. M.; Chung, K. F.
(2007). Role of TLR2, TLR4, and MyD88 in murine ozone-induced airway hyperresponsiveness and neutrophilia. J
ApplPhysiol, 103: 1189-1195. http://dx.doi.org/10.1152/iapplphvsiol.00172.2007597545
Wu, Z. X.; Barker, J. S.; Batchelor, T. P.; Dey, R. D. (2008). Interleukin (IL)-l regulates ozone-enhanced tracheal smooth
muscle responsiveness by increasing substance P (SP) production in intrinsic airway neurons of ferret. Respir
Physiol Neurobiol, 164: 300-311. http://dx.doi.Org/10.1016/j.resp.2008.07.019 597548
Wu, Z. X.; Satterfield, B. E.; Dey, R. D. (2003). Substance P released from intrinsic airway neurons contributes to ozone-
enhanced airway hyperresponsiveness in ferret trachea. J Appl Physiol, 95: 742-750.
http://jap.physiology.org.jproxy.lib.ecu.edU/cgi/content/abstract/95/2/742. 628616
March 2011 5-78 DRAFT - DO NOT CITE OR QUOTE
-------
Yang, I. A.; Holz, O.; Jorres, R. A.; Magnussen, H.; Barton, S. I; Rodriguez, S.; Cakebread, J. A.; Holloway, J. W.;
Holgate, S. T. (2005). Association of tumor necrosis factor-"alpha" polymorphisms and ozone-induced change in
lung function. Am J Respir Crit Care Med, 171: 171-176. 077211
Yokoyama, E.; Frank, R. (1972). Respiratory uptake of ozone in dogs. Arch Environ Occup Health, 25: 132-138. 039756
Yoon, H. K.; Cho, H. Y; Kleeberger, S. R. (2007). Protective role of matrix metalloproteinase-9 in ozone-induced airway
inflammation. Environ Health Perspect, 115: 1557-1563. http://dx.doi.org/10.1289/ehp.10289 596422
Yost, B. L.; Gleich, G. J.; Jacoby, D. B.; Fryer, A. D. (2005). The changing role of eosinophils in long-term hyperreactivity
following a single ozone exposure. Am J Physiol Lung Cell Mol Physiol, 289: L627-L635.
http://dx.doi.org/10.1152/aiplung.00377.2004597549
March 2011 5-79 DRAFT - DO NOT CITE OR QUOTE
-------
Chapter 6. Integrated Health Effects of
Short-Term Ozone Exposure
6.1. Introduction
1 This chapter reviews, summarizes, and integrates the evidence for various health outcomes
2 associated with short-term exposures to O3 (hours to days). Numerous controlled human exposure,
3 epidemiologic and toxicological studies have permitted evaluation of the relationships of short-term
4 O3 exposure with a range of endpoints related to respiratory effects (Section 6.2), cardiovascular
5 effects (Section 6.3), and mortality (Sections 6.2, 6.3, and 6.6). A smaller number of studies are
6 available to assess the effects of O3 on other physiological systems such as the central nervous
7 system (Section 6.4), liver and metabolism (Section 6.5.1), and cutaneous and ocular tissues (Section
8 6.5.2).
9 Evidence for major health effect categories (e.g., respiratory, cardiovascular, mortality) is
10 described in individual sections that include a brief summary of conclusions from the 2006 O3
11 AQCD and an evaluation of recent evidence that is intended to build upon evidence from previous
12 reviews. Within each section, results are organized by health endpoint (e.g., lung function,
13 pulmonary inflammation) then by specific scientific discipline (e.g., controlled human exposure,
14 epidemiology, and toxicology). Each major section (e.g., respiratory, cardiovascular, mortality)
15 concludes with an integrated summary of the findings and a conclusion regarding causality. Based
16 upon the framework described in Chapter 1, a determination of causality is made for a broad health
17 effect category, such as respiratory effects, with coherence and plausibility being based on evidence
18 available across disciplines and also across the suite of related health endpoints, including cause-
19 specific mortality.
6.2. Respiratory Effects
20 Based on evidence integrated across human controlled exposure, epidemiologic, and
21 toxicological studies, the 2006 O3 AQCD concluded that there was clear, consistent evidence of a
22 causal relationship between short-term O3 exposure and respiratory effects (U.S. EPA, 2006,
23 088089). Contributing to this conclusion were consistent and coherent observations across scientific
24 disciplines of associations of short-term O3 exposures with pulmonary function decrements and
25 increases in lung inflammation, lung permeability, airway hyperresponsiveness, respiratory
26 symptoms, and respiratory-related hospitalizations and emergency department (ED) visits.
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments such as the Integrated Science Assessments (ISAs) and the Integrated Risk Information System (IRIS).
March 2011 6-1 DRAFT-DO NOT CITE OR QUOTE
-------
1 Controlled human exposure studies have provided strong and quantifiable exposure-response
2 data on the human health effects of O3. The most salient observations from studies reviewed in the
3 1996 and 2006 O3 AQCDs (U.S. EPA, 1996, 017831: U.S. EPA, 2006, 088089) were that: (1) young
4 healthy adults exposed to O3 concentrations^ 80 ppb develop significant reversible, transient
5 decrements in pulmonary function if minute ventilation (VE) or duration of exposure is increased
6 sufficiently; (2) children experience similar spirometric responses but lesser symptoms from O3
7 exposure relative to young adults; (3) O3-induced spirometric responses are decreased in older
8 individuals relative to young adults; (4) there is a large degree of intersubject variability in
9 physiologic and symptomatic responses to O3, but responses tend to be reproducible within a given
10 individual over a period of several months; and (5) subjects exposed repeatedly to O3 for
11 several days develop a tolerance to successive exposures, as demonstrated by an attenuation of
12 spirometric and symptomatic responses, that is lost after about a week without exposure.
13 Substantial evidence for biologically plausible O3-induced respiratory morbidity has been
14 derived from coherence between toxicological and controlled human exposure studies examining
15 parallel endpoints. For example, O3-induced decrements in lung function have also been observed in
16 animals, and as in humans, tolerance or adaptation has been demonstrated in animal models. Both
17 humans and rodents exhibit increased airway hyperresponsiveness. This is an important consequence
18 of exposure to ambient O3, because the airways are then predisposed to narrowing upon inhalation of
19 a variety of ambient stimuli. Additionally, airway hyperresponsiveness tends to resolve more slowly
20 and appears less subject to attenuation. Increased permeability and inflammation have been observed
21 in the airways of humans and animals alike after O3 exposure, and although these aspects are not
22 necessarily associated with immediate changes in lung function or hyperresponsiveness, the potential
23 relationship between repetitive bouts of acute inflammation and the development of chronic
24 respiratory disease is unknown. Another feature of O3 exposure-related respiratory morbidity is
25 impaired host defense and reduced resistance to lung infection, which has been strongly supported
26 by toxicological evidence and to a limited extent by human data. Respiratory infection in early life is
27 associated with increased incidence of asthma in humans.
28 In epidemiologic studies, acute O3-related respiratory morbidity has been assessed most
29 frequently using lung function. Several studies of healthy children attending camps as well as studies
30 of outdoor workers, groups exercising outdoors, and children with asthma support O3 effects on lung
31 function at ambient levels (U.S. EPA, 1996, 017831: U.S. EPA, 2006, 088089V In addition to lung
32 function, ambient O3 has been positively associated with respiratory symptoms (e.g., cough, wheeze,
33 shortness of breath), especially in large U.S. panel studies of asthmatic children (Gent et al., 2003,
34 052885: Mortimer et al., 2000, 013255). The respiratory health effects of acute O3 exposure are most
35 clearly indicated in asthmatic children and subjects with increased outdoor exposures. In contrast
36 with other respiratory health endpoints, the association between short-term O3 exposure and
37 respiratory mortality is less clearly indicated. Although O3 has been consistently associated with
38 nonaccidental and cardiopulmonary mortality, the contribution of respiratory causes to these findings
March 2011 6-2 DRAFT - DO NOT CITE OR QUOTE
-------
1 has been uncertain as the few studies that have examined mortality specifically from respiratory
2 causes have reported inconsistent associations with ambient O3 exposures.
3 As discussed throughout this section, consistent with the strong body of evidence presented in
4 the 2006 O3 AQCD, recent studies continue to support associations between short-term O3 exposure
5 and respiratory health effects, in particular, lung function decrements in controlled human exposure
6 studies, airway inflammatory responses in toxicological studies, and respiratory-related
7 hospitalizations and ED visits. Recent epidemiologic studies contribute new evidence on susceptible
8 populations and of associations of ambient O3 exposures with biological markers of airway
9 inflammation and oxidative stress, which is consistent with the extensive evidence from human
10 controlled exposure and toxicological studies. Furthermore, extending the potential continuum of
11 well-established O3-associated respiratory effects, new multicity studies and a multicontinent study
12 demonstrate associations between ambient O3 and respiratory-related mortality.
6.2.1. Lung Function
6.2.1.1. Controlled Human Exposure
13 This section focuses on studies in which volunteers were exposed, for periods of up to 8 hours
14 to O3 concentrations ranging from 40 to 500 ppb, while at rest or during exercise of varying
15 intensity. Responses to acute O3 exposures in the range of ambient concentrations include decreased
16 inspiratory capacity; mild bronchoconstriction; rapid, shallow breathing patterns during exercise; and
17 symptoms of cough and pain on deep inspiration (PDI). Reflex inhibition of inspiration results in a
18 decrease in forced vital capacity (FVC) and total lung capacity (TLC) and, in combination with mild
19 bronchoconstriction, contributes to a decrease in the forced expiratory volume in 1 second (FEVi).
20 As individuals may experience small changes in various health endpoints from exercise, diurnal
21 variation, or other effects in addition to those of O3 during the course of an exposure, the term "O3-
22 induced" is used herein to designate effects that have been corrected for such extraneous responses
23 as measured during filtered air (FA) exposures.
Pulmonary Function Effects of Ozone Exposure in Healthy Subjects
Acute Exposure of Healthy Subjects
24 The majority of controlled human exposure studies have investigated the effects of exposure to
25 O3 in young healthy nonsmoking adults (18-35 years of age). These studies typically use fixed
26 concentrations of O3 under carefully regulated environmental conditions and subject activity levels.
27 The magnitude of respiratory effects (decrements in spirometry and symptomatic response) in these
28 individuals is a function of O3 concentration (C), minute ventilation (VE), and exposure duration.
29 Any physical activity will increase minute ventilation and therefore the dose of inhaled O3. Dose of
30 inhaled O3 to the lower airways is also increased due to a shift from nasal to oronasal breathing with
March 2011 6-3 DRAFT - DO NOT CITE OR QUOTE
-------
1 a consequential decrease in O3 scrubbing by the upper airways. Thus, the intensity of physiological
2 response following an acute exposure will be strongly associated with minute ventilation.
3 There is a rapid recovery of O3-induced spirometric responses and symptoms; 40 to 65%
4 recovery appears to occur within about 2 hours following exposure (Folinsbee and Hazucha, 1989,
5 041732). For example, following a 2-h exposure to 400 ppb O3 with intermittent exercise,
6 Nightingale et al. (2000, 000796) observed a 13.5% mean decrement in FEVi. By 3 hours
7 postexposure, however, only a 2.7% FEVi decrement persisted. Partial recovery also occurs
8 following cessation of exercise despite continued exposure to O3 (Folinsbee et al., 1977, 038283)
9 and at low O3 concentrations during exposure (Hazucha et al., 1992, 042789). A slower recovery
10 phase, especially after exposure to higher O3 concentrations, may take at least 24 hours to complete
11 (Folinsbee and Hazucha, 2000, 001701: Folinsbee et al., 1993, 043781). Repeated daily exposure
12 studies at higher concentrations typically show that FEVi response to O3 is enhanced on the
13 second day of exposure. This enhanced response suggests a residual effect of the previous exposure,
14 about 22 hours earlier, even though the pre-exposure spirometry may be the same as on the
15 previous day. The absence of the enhanced response with repeated exposure at lower O3
16 concentrations may be the result of a more complete recovery or less damage to pulmonary tissues
17 (Folinsbee et al., 1994, 044189).
18 For healthy young adults exposed at rest for 2 hours, 500 ppb is the lowest O3 concentration
19 reported to produce a statistically significant O3-induced group mean FEVi decrement (6.4%, n=10,
20 (Folinsbee et al., 1978, 039060) and 6.7%, n=13, (Horvath et al., 1979, 039222)). Airway resistance
21 was not clearly affected during at-rest exposure to these O3 concentrations. For exposures of 1-2
22 hours to > 120 ppb O3, statistically significant symptomatic responses and effects on FEVi are
23 observed when VE is sufficiently increased by exercise. For instance, with very heavy continuous
24 exercise (VE= 89 L/min), an O3-induced group mean decrement of 9.7% in FEVi has been reported
25 for healthy young adults exposed for 1 hour to 120 ppb O3 (Gong et al., 1986, 040465). Symptoms
26 are present and decrements in forced expiratory volumes and flows occur at 160-240 ppb
27 O3 following 1 hour of continuous heavy exercise (VE « 55 to 90 L/min (Adams and Schelegle,
28 1983, 039822: Avol et al., 1984, 040221: Folinsbee et al., 1984, 040065: Gong et al., 1986, 040465)
29 and following 2 hours of intermittent heavy exercise (VE « 65-68 L/min) (Kulle et al., 1985, 040311:
30 Linn et al., 1986, 040481: McDonnell et al., 1983, 040680). With heavy intermittent exercise
31 (15-min intervals of rest and exercise [VE = 68 L/min]), symptoms of breathing discomfort and a
32 group mean O3-induced decrement of 3.4% in FEVi occurred in young healthy adults exposed for 2
33 hours to 120 ppb O3 (McDonnell et al., 1983, 040680).'
34 For prolonged (6.6 hours) exposures relative to shorter exposures, significant pulmonary
35 function responses and symptoms have been observed at lower O3 concentrations and at a moderate
36 level of exercise (VE = 40 L/min). The results from studies using 6.6 hours of constant or square-
37 wave (S-W) exposures are illustrated in Figure 6-1. Exposure to 40 ppb for 6.6 hours produces small,
1 In total, subjects were exposed to O3 for 2.5 hours. Intermittent exercise periods, however, were only conducted for the first 2 hours of
exposure and FEVi was determined 5 minutes after the exercise was completed.
March 2011 6-4 DRAFT - DO NOT CITE OR QUOTE
-------
1 statistically insignificant changes in FEVi that are relatively similar to responses from FA exposure
2 (Adams, 2002, 093690). Volunteers exposed to 60 ppb O3 experience group mean O3-induced FEVi
3 decrements of about 3% (Adams, 2006, 087681; Brown et al, 2008, 195140)1; those exposed to
4 80 ppb have group mean decrements which range from 6 to 8% (Adams, 2003, 042245; Adams,
5 2006, 087681; Horstman et al., 1990, 042187; McDonnell et al., 1991, 042384); at 100 ppb, group
6 mean decrements range from 8 to 14% (Horstman et al., 1990, 042187; McDonnell et al., 1991,
7 042384); and at 120 ppb, group mean decrements of 13 to 16% are observed (Adams, 2002, 093690;
8 Folinsbee et al., 1988, 040898; Horstman et al., 1990, 042187). As illustrated in Figure 6-1, there is a
9 smooth dose-response curve without evidence of a threshold for exposures between 40 and 120 ppb
10 O3. Taken together, these data indicate that mean FEVi is clearly decreased by 6.6-h exposures to
11 60 ppb O3 and higher concentrations in subjects performing moderate exercise.
20 -i
-O C. 15 -
o c
3 QJ
10 -
» Adams (2006)
A Adams (2003)
* Adams (2002)
0 Folinsbee etal. (1988)
n Horstman etal. (1990)
McDonnell etal. (1991)
McDonnell etal. (2007)
O
O
0.02 0.04
0.06 0.08 0.1
Ozone (ppm)
0.12 0.14
Source: Brown et al. (2008, 1951401
Figure 6-1. Cross-study comparison of mean ozone-induced FEVi decrements following 6.6
hours of constant, square-wave exposure to ozone.rDuring each hour of the
exposures, subjects were engaged in moderate exercise for 50 minutes and rest for
10 minutes. Following the third hour, subjects had an additional 35 minute rest
period for lunch. The McDonnell et al. (2007, 093104) curve illustrates the predicted
FEVi decrement at 6.6 hours as a function of ozone concentration for a 23-year old
(the average age of subjects that participated in the illustrated studies). Note that
this curve was not "fitted" to the plotted data. Error bars (where available) are the
standard error of responses. The data at 0.08 and 0.12 ppm have been offset for
illustrative purposes.
1 Adams (2006, 087681) did not find effects on FEVi at 60 ppb to be statistically significant. In an analysis of the Adams (2006, 087681)
data, even after removal of potential outliers, Brown et al. (2008, 195140) found the average effect on FEVi at 60 ppb to be small, but
highly statistically significant (p < 0.002) using several common statistical tests.
March 2011
6-5
DRAFT - DO NOT CITE OR QUOTE
-------
1 As opposed to constant or S-W concentration patterns used in the studies described above,
2 many more recent studies conducted at the levels of 40-80 ppb have used variable O3 concentration
3 patterns. It has been suggested that a triangular (variable concentration) exposure profile can
4 potentially lead to higher FEVi responses than S-W profiles at overall equivalent inhaled O3 doses.
5 Hazucha et al. (1992, 042789) were the first to investigate the effects of variable versus constant
6 concentration exposures on responsiveness to O3. In their study, volunteers were randomly exposed
7 to a triangular concentration profile that increased linearly from 0-240 ppb for the first 4 hours of the
8 8-h exposure, then decreased linearly from 240 to 0 ppb over the next 4 hours of the 8-h exposure,
9 and to an S-W exposure of 120 ppb O3 for 8 hours. While the total inhaled O3 doses at 4 hours and
10 8 hours for the S-W and the triangular concentration profile were almost identical, the FEVi
11 response was dissimilar. For the S-W exposure, FEVi declined -5% by the fifth hour and then
12 remained at that level. With the triangular O3 profile, there was minimal FEVi response over the first
13 3 hours followed by a rapid decrease in FEVi (-10.3%) over the next 3 hours. During the seventh
14 and eighth hours, mean FEVi decrements improved to -6.3% as the O3 concentration decreased from
15 120 to 0 ppb (mean = 60 ppb). These findings illustrate that the severity of symptoms and the
16 magnitude of spirometric responses are time-dependent functions of inhaled dose rate with periods
17 of both effect development and recovery during the course of an exposure.
18 Subsequently, others have also demonstrated that variable concentration exposures can elicit
19 greater FEVi and symptomatic responses than S-W exposures (Adams, 2003, 042245; Adams, 2006,
20 196494: Adams, 2006, 087681). Adams (2006, 196494) reproduced the findings of Hazucha et al.
21 (1992, 042789) at 120 ppb. However, Adams (2003, 042245: 2006, 087681) found that responses
22 from an 80 ppb O3 (average) triangular exposure did not differ significantly from those observed in
23 the 80 ppb O3 S-W exposure at 6.6 hours. Nevertheless, FEVi and symptoms were significantly
24 different from pre-exposure at 4.6 hours (when the O3 concentration was 150 ppb) in the triangular
25 exposure, but not until 6.6 hours in the S-W exposure. At the lower O3 concentration of 60 ppb, no
26 temporal pattern differences in FEVi responses between S-W and triangular exposure profiles could
27 be discerned (Adams, 2006, 087681). However, total symptom scores were significantly increased
28 for the 60 ppb triangular (but not the S-W) exposure following 5.6 and 6.6 hours of exposure. At
29 40 ppb, triangular and S-W patterns produced responses similar to FA exposure (Adams, 2002,
30 093690: Adams, 2006, 087681).
31 For exposures of 60 ppb and greater, these studies (Adams, 2003, 042245: Adams, 2006,
32 087681: Adams, 2006, 196494: Hazucha et al., 1992, 042789) demonstrate that during triangular
33 exposure protocols, volunteers may develop greater spirometric and/or symptomatic responses
34 during and following peak O3 concentrations as compared to responses over the same time interval
35 of S-W exposures. This observation is not unexpected since the inhaled dose rate during peaks of the
36 triangular protocols approached twice that of the S-W protocols, e.g., 150 ppb versus 80 ppb. At time
37 intervals toward the end of an exposure, inhaled dose rates for the triangular protocols were less than
38 those of S-W. At these later time intervals, there is some recovery of responses during triangular
39 exposure protocols, whereas there is a continued development of or a plateau of responses in the
March 2011 6-6 DRAFT - DO NOT CITE OR QUOTE
-------
1 S-W exposure protocols. Thus, responses during triangular protocols relative to S-W protocols may
2 be expected to diverge and be greater following peak exposures and then converge toward the end of
3 an exposure. The ensuing discussion on exposures between 40 and 80 ppb will focus on
4 postexposure effects where the influence of triangular and S-W concentration patterns are minimal,
5 i.e., FEVi pre-to-post effects are similar (although not identical) between triangular and S-W
6 protocols having equivalent average exposure concentrations.
7 Schelegle et al. (2009, 618629) recently investigated the effects of 6.6 hours variable O3
8 exposure protocols at mean concentrations of 60, 70, 80, and 87 ppb on respiratory symptoms and
9 pulmonary function in young healthy adults (16 F, 15 M; 21.4 ± 0.6 years). The mean FEVi
10 (±standard error) decrements at 6.6 hours (end of exposure relative to pre-exposure) were -0.80 ±
11 0.90%, 2.72 ± 1.48%, 5.34 ± 1.42%, 7.02 ± 1.60%, and 11.42 ± 2.20% for exposure to FA, 60, 70,
12 80, and 87 ppb O3, respectively. Statistically significant decrements in FEVi and increases in total
13 subjective symptom scores (p < 0.05) were found following exposure to mean concentrations of 70,
14 80, and 87 ppb O3 relative to FA. Statistically significant effects were not found at 60 ppb. One of
15 the expressed purposes of the Schelegle et al. (2009, 618629) study was to determine the minimal
16 mean O3 concentration that produces a statistically significant decrement in FEVi and symptoms in
17 healthy individuals completing 6.6-h exposure protocols. At 70 ppb, Schelegle et al. (2009, 618629)
18 observed significant effects. At 60 ppb, a 3.5% FEVi decrement was not found to be statistically
19 significant. However, the slightly smaller 2.9% FEVi decrement at 60 ppb observed by Adams
20 (2006, 087681) was found to be statistically significant by Brown et al. (2008, 195140).
21 More recently, Kim et al. (In Press, 674869) investigated the effects of a 6.6-h exposure to
22 60 ppb O3 on pulmonary function and respiratory symptoms in young healthy adults (32 F, 27 M;
23 25.0 ± 0.5 year) that were roughly half GSTM1-null and half GSTM1-positive. Sputum neutrophil
24 levels were also measured in a subset of the subjects (13 F, 11 M). The mean FEVi (±standard error)
25 decrements at 6.6 hours (end of exposure relative to pre-exposure) were significantly different (p =
26 0.008) between the FA (0.002 ± 0.46%) and O3 (1.76 ± 0.50%) exposures. The inflammatory
27 response following O3 exposure was also significantly (p<0.001) increased relative to the FA
28 exposure. Respiratory symptoms were not affected by O3 exposure. There was also no significant
29 effect of GSTM1 genotype on FEVi or inflammatory responses.
30 Consideration of the minimal O3 concentration producing statistically significant effects
31 following 6.6-h exposures warrants additional discussion. As discussed above, numerous studies
32 have demonstrated statistically significant O3-induced group mean FEVi decrements of 6-8% at
33 80 ppb. Schelegle et al. (2009, 618629) have now reported statistically significant O3-induced group
34 mean FEVi decrement of 6%, as well as respiratory symptoms, at 70 ppb. At 60 ppb, there is
35 information available from 4 separate studies (Adams, 1998, 670457)'(Adams. 2006, 087681; Kim
36 et al., In Press, 674869: Schelegle et al., 2009, 618629V The group mean O3-induced FEVi
1 The American Petroleum Institute has declined to provide a copy of this report to EPA.
March 2011
6-7
DRAFT - DO NOT CITE OR QUOTE
-------
1 decrements observed in these studies were 3.6% by Adams (1998, 670457)1. 2.8% (triangular
2 exposure) and 2.9% (S-W exposure) by Adams (2006, 087681). 3.5% by Schelegle et al. (2009,
3 618629). and 1.8% by Kim et al. (In Press, 674869). Based on data from these three studies, at
4 60 ppb, the weighted-average group mean O3-induced FEVi decrement (i.e., adjusted for FA
5 responses) is 2.7% (n=150) (Adams, 1998, 670457: Adams, 2006, 087681: Kim et al., In Press,
6 674869: Schelegle et al., 2009, 618629). Although not found to be statistically significant in the
7 original studies, these group mean changes in FEVi at 60 ppb are consistent between studies, i.e.,
8 none observed an average improvement in lung function with following a 6.6-h exposure to 60 ppb
9 O3. Indeed, as was illustrated in Figure 6-1, the FEVi responses at 60 ppb fall on a smooth dose-
10 response curve for exposures between 40 and 120 ppb O3. Furthermore, in a re-analysis of the
11 60 ppb S-W data from Adams (2006, 087681). Brown et al. (2008, 195140) found the mean effects
12 on FEVi to be highly statistically significant (p<0.002) using several common statistical tests even
13 after removal of 3 potential outliers. The time-course and magnitude of FEVi responses at 40 ppb
14 resemble those occurring during FA exposures (Adams, 2002, 093690: Adams, 2006, 087681).
15 Taken together, the available evidence shows that detectable effects of O3 on group mean FEVi
16 persist down to 60 ppb, but not 40 ppb in young healthy adults exposed for 6.6 hours during
17 moderate exercise.
Intersubject Variability in Response of Healthy Subjects
18 Consideration of group mean changes is important in discerning if observed effects are due to
19 O3 exposure rather than chance alone. Inter-individual variability in responses is, however,
20 considerable and pertinent to assessing the fraction of the population that might actually be affected
21 during an O3 exposure. Hackney et al. (1975, 039208) first recognized a wide range in the sensitivity
22 of subjects to O3. The range in the subjects' ages (29 to 49 years) and smoking status (0 to 50 pack
23 years) in the Hackney et al. (1975, 039208) study are now understood to affect the spirometric and
24 symptomatic responses to O3. Subsequently, DeLucia and Adams (1977, 038281) examined
25 responses to O3 in six healthy non-smokers and found that two exhibited notably greater sensitivity
26 to O3. Since that time, numerous studies have documented considerable variability in responsiveness
27 to O3 even in subjects recruited to assure homogeneity in factors recognized or presumed to affect
28 responses.
29 An individual's FEVi response to a 2-h O3 exposure is generally reproducible over
30 several months and presumably reflects the intrinsic responsiveness of the individual to O3 (Hazucha
31 et al., 2003, 048168: McDonnell WF 3rd; Horstman et al., 1985, 040283). The frequency distribution
32 of individual FEVi responses following these relatively short exposures becomes skewed as the
33 group mean response increases, with some individuals experiencing large reductions in FEVi (Kulle
34 et al., 1985, 040311: Weinmann et al., 1995, 076022). For 2-h exposures with intermittent exercise
35 causing a predicted average FEVi decrement of 10%, individual decrements ranged from
36 approximately 0 to 40% in white males aged 18-36 years (McDonnell et al., 1997, 084266). For an
1 This information is from page 133 of Adams (2006, 087681).
March 2011
6-8
DRAFT - DO NOT CITE OR QUOTE
-------
1 average FEVi decrement of 13%, Ultman et al. (2004, 057197) reported FEVi responses ranging
2 from a 4% improvement to a 56% decrement in young healthy adults (32 M, 28 F) exposed for 1
3 hour to 250 ppb O3. One-third of the subjects had FEVi decrements of >15%, and 7% of the subjects
4 had decrements of >40%.
5 Consistent with the 1- to 2-h studies, the distribution of individual responses following 6.6-h
6 exposure studies becomes skewed with increasing exposure concentration and magnitude of the
7 group mean FEVi response (McDonnell, 1996, 082679). Figure 6-2 illustrates frequency
8 distributions of individual FEVi responses observed in 31 young healthy adults following 6.6-h
9 exposures between 0 and 80 ppb. Schelegle et al. (2009, 618629) found >10% FEVi decrements in
10 16, 19, 29, and 42% of individuals exposed for 6.6 hours to 60, 70, 80, and 87 ppb, respectively. Just
11 as there are differences in mean decrements between studies having similar exposure scenarios
12 (Figure 6-1 at 80 and 120 ppb), there are also differences in the proportion of individuals affected
13 with >10% FEVi decrements. At 80 ppb, the proportion affected with >10% FEVi decrements was
14 17% (n=30) by Adams (2006, 087681)1. 26% (n=60) by McDonnell (1996, 082679). and 29%(n=31)
15 by Schelegle et al. (2009, 618629). At 60 ppb, the proportion with >10% FEVi decrements was 20%
16 (n=30) by Adams (1998, 670457)2. 3% (n=30) by Adams (2006, 087681)5. 16% (n=31) by Schelegle
17 et al. (2009, 618629). and 5% (n=59) by Kim et al. (In Press, 674869). Based on these studies, the
18 weighted average proportion of individuals with >10% FEVi decrements is 10% following exposure
19 to 60 ppb. Due to insufficient data, these proportions were not corrected for responses to FA
20 exposure where lung function typically improves in healthy adults. For example, uncorrected versus
21 O3-induced (i.e., adjusted for response during FA exposure) proportions of individuals having >10%
22 FEVi decrements in the Adams (2006, 087681)3 study were, respectively, 3% versus 7% at 60 ppb
23 and 17% versus 23% at 80 ppb. Thus, uncorrected proportions underestimate the actual fraction of
24 healthy individuals affected.
1 Not assessed by Adams (2006, 087681). the proportion was provided in Figure 8-1B of U.S. EPA(2006, 088089).
2 This information is from page 761 of Adams (2002, 093690).
3 Not assessed by Adams (2006, 087681). uncorrected and O3-induced proportions are from Figures 8-1B and 8-2, respectively, of the 2006
03 AQCD (2006, 088089).
March 2011 6-9 DRAFT - DO NOT CITE OR QUOTE
-------
J2
"§ 30'
2*25-
W 20'
M—
Percent c
3 Ui O Ui
n
-
0 ppb
0%
nn
—
60 ppb
16%
in
n
70 ppb
—
19%
Hn n
nn
-
80 ppb
—
29%
Inl In
-10 0 10 20 30 -10 0 10 20 30 -10 0
FEV., Decrement (%)
10
20
30
10
20
30
Source: Adapted with permission from American Thoracic Society, Schelegle et al. (2009, 6186291
Figure 6-2. Frequency distributions of FEVi decrements observed by Schelegle et al. (2009) in
young healthy adults (16 F, 15 M) following 6.6-h exposures to ozone or filtered air. D
During each hour of the exposures, subjects were engaged in moderate exercise for
50 minutes and rest for 10 minutes. Following the third hour, subjects had an
additional 35-minute rest period for lunch. Subjects were exposed to a triangular
ozone concentration profile having the average ozone concentration provided in
each panel. As average ozone concentration increased, the distribution of
responses became asymmetric with a few individuals exhibiting large FEVi
decrements. The percentage indicated in each panel is the portion of subjects
having a FEVi decrement in excess of 10%.
1 Given considerable inter-individual variability in responses, the interpretation of biologically
2 small group mean decrements requires careful consideration. Following prolonged 6.6-h exposures
3 to an average level of 60 ppb O3, data available from three studies yield a weighted-average group
4 mean O3-induced FEVi decrement (i.e., adjusted for FA responses) of 3.3% (n=91) (Adams, 1998,
5 670457: Adams, 2006, 087681: Schelegle et al., 2009, 618629). The data from these studies also
6 yield a weighted-average proportion (uncorrected for FA responses) of subjects with >10% FEVi
7 decrements of 13% (n=91) (Adams, 1998, 670457: Adams, 2006, 087681: Schelegle et al., 2009,
8 618629). In an individual with relatively "normal" lung function, recognizing technical and
9 biological variability in measurements, confidence can be given that within-day changes in FEVi of
10 > 5% are clinically meaningful (American Thoracic Society, 1991, 044889: Pellegrino et al., 2005,
11 626521). Here focus is given to individuals with >10% decrements in FEVi since some individuals
12 in the Schelegle et al. (2009, 618629) study experienced 5-10% FEVi decrements following
13 exposure to FA. The data are not available to the EPA to determine the O3-induced proportion for
14 either the Adams (1998, 670457) or Schelegle et al. (2009, 618629) studies. As already stated,
15 however, this uncorrected proportion likely underestimates that actual proportion of healthy
16 individuals experiencing O3-induced FEVi decrements in excess of 10%. Therefore, by considering
17 uncorrected responses and those individuals having >10% decrements, 13% is an underestimate of
18 the proportion of healthy individuals that are likely to experience clinically meaningful changes in
19 lung function following exposure for 6.6 hours to 60 ppb O3 during moderate exercise. Although
20 none of these studies (Adams, 1998, 670457: Adams, 2006, 087681: Schelegle et al., 2009, 618629)
21 reported FEVi decrements at 60 ppb to be statistically significant, Brown et al. (2008, 195140) found
22 those from Adams (2006, 087681) to be highly statistically significant. The forgoing discussion
23 shows that even where group mean decrements are biologically small and of debatable statistical
March 2011
6-10
DRAFT - DO NOT CITE OR QUOTE
-------
1 significance, a considerable fraction of exposed individuals experience clinically meaningful
2 decrements in lung function.
Responses in Individuals with Pre-Existing Disease
3 Individuals with respiratory disease are of primary concern in evaluating the health effects of
4 O3 because a given change in function is likely to have more impact on a person with preexisting
5 function impairment and reduced reserve.
6 Possibly due to the age of subjects studied, patients with COPD performing light to moderate
7 exercise do not generally experience statistically significant pulmonary function decrements
8 following 1- and 2-h exposures to < 300 ppb O3 (Kehrl et al, 1985, 040294: Linn et al, 1982,
9 039645: Linn et al., 1983, 040672: Solic et al., 1982, 039610V Following a 4 h exposure to 240 ppb
10 O3 during exercise, Gong et al. (1997, 083593) found an O3-induced FEVi decrement of 8% in
11 COPD patients which was not statistically different from the decrement of 3% in healthy subjects.
12 Demonstrating the need for control exposures and presumably due to exercise, four of the patients in
13 the Gong et al. (1997, 083593) study had FEVi decrements of >14% following both the FA and O3
14 exposures. Although the clinical significance is uncertain, small transient decreases in arterial blood
15 oxygen saturation have also been observed in some of these studies.
16 Based on studies reviewed in the 1996 and 2006 O3 AQCD (U.S. EPA, 1996, 017831:
17 U.S. EPA, 2006, 088089). asthmatic subjects appear to be at least as sensitive to acute effects of O3
18 as healthy nonasthmatic subjects. Horstman et al. (1995, 075834) found the O3-induced FEVi
19 decrement in mild-to-moderate asthmatics to be significantly larger than in healthy subjects (19%
20 versus 10%, respectively) exposed to 160 ppb O3 during exercise for 7.6-h exposure. In asthmatics, a
21 significant positive correlation between O3-induced spirometric responses and baseline lung function
22 was observed, i.e., responses increased with severity of disease. Such differences in pulmonary
23 function between asthmatics and healthy individuals were not found in shorter duration studies.
24 Alexis et al. (2000, 013072) and Torres et al. (1996, 078122) reported a tendency for slightly greater
25 FEVi decrements in asthmatics than healthy subjects. Several studies reported similar responses
26 between asthmatics and healthy individuals (Basha et al., 1994, 075950: Hiltermann et al., 1995,
27 078494: Scannell et al., 1996, 080755). The lack of differences in the Hiltermann et al. (1995,
28 078494) and Basha et al. (1994, 075950) studies was not surprising, however, given extremely small
29 sample sizes and corresponding lack of statistical power. One study reported a tendency for
30 asthmatics to have smaller O3-induced FEVi decrements than healthy subjects (3% versus 8%,
31 respectively) when exposed to 200 ppb O3 for 2 hours during exercise (Mudway et al., 2001,
32 025327). However, the asthmatics in that study also tended to be older than the healthy subjects,
33 which could partially explain their lesser response since FEVi responses to O3 diminish with age.
34 Some, but not all, studies have also reported that asthmatics have a somewhat exaggerated
35 airway inflammatory response to acute O3 exposure relative to healthy control subjects (e.g., (Basha
36 et al., 1994, 075950: Hiltermann et al., 1997, 084979: Hiltermann et al., 1999, 013196: Holz et al.,
37 2002, 041632: McBride et al., 1994, 043912: Michelson et al., 1999, 001147: Newson et al., 2000,
March 2011 6-11 DRAFT-DO NOT CITE OR QUOTE
-------
1 000853: Peden, 2001, 025355: Peden et al, 1995, 076189: Peden et al., 1997, 085842: Scannell et
2 al., 1996, 080755: Vagaggini et al., 1999, 001210)). For example, at 18 hours post-O3 exposure
3 (200 ppb, 4 hours with exercise) and corrected for FA responses, Scannell et al. (1996, 080755)
4 found significantly increased neutrophils in 18 asthmatics (12%) compared to 20 healthy subjects
5 (4.5%). This difference in inflammatory response was observed despite no group differences in
6 spirometric responses to O3.
7 Vagaggini et al. (2010, 387127) exposed mild-to-moderate asthmatics (n=23; 33 ± 11 years) to
8 300 ppb O3 for 2 hours with moderate exercise. Although the group mean O3-induced FEVi
9 decrement was only 4%, eight subjects were categorized as "responders" with >10 FEVi decrements.
10 There were no baseline differences between responders and nonresponders. At 6 hours post O3
11 exposure, sputum neutrophils were significantly increased by 15% relative to FA in responders. The
12 neutrophil increase in responders was also significantly greater than the 0.2% increase in
13 nonresponders. Across all subjects, there was a significant (r=0.61, p = 0.015) correlation between
14 changes in FEVi and changes in sputum neutrophils. Prior studies have reported that inflammatory
15 responses do not appear to be correlated with lung function responses in either asthmatic or healthy
16 subjects (Balmes et al., 1996, 080830: Balmes et al., 1997, 086092: Devlin et al., 1991, 040359:
17 Holz et al., 1999, 058731). Interestingly, the nonresponders in the Vagaggini et al. (2010, 387127)
18 study experienced a significant O3-induced 11.3% increase in sputum eosinophils, while responders
19 had an insignificant 2.6% decrease. Six of the subjects were NQOlwt and GSTM1 null, but this
20 phenotype was not found to be associated with the changes in lung function or inflammatory
21 responses to O3.
22 A few recent studies have evaluated the effects of corticosteroid usage on the response of
23 asthmatics to O3. Vagaggini et al. (2007, 196638) evaluated whether corticosteroid usage would
24 prevent O3-induced lung function decrements and inflammatory responses in a group of subjects
25 with mild persistent asthma (n=9; 25 ± 7 years). In this well designed study, asthmatics were
26 randomly exposed on four occasions to 270 ppb O3 or FA for 2 hours with moderate exercise.
27 Exposures were preceded by four days of treatment with prednisone or placebo. Pretreatment with
28 corticosteroids prevented an inflammatory response in induced sputum at 6 hours postexposure.
29 FEVi responses were, however, not prevented by corticosteroid treatment and were roughly
30 equivalent to those observed following placebo. Vagaggini et al. (2001, 025343) also found
31 budesonide to decrease airway neutrophil influx in asthmatics following O3 exposure. In contrast,
32 inhalation of corticosteroid budesonide failed to prevent or attenuate O3-induced responses in healthy
33 subjects as assessed by measurements of lung function, bronchial reactivity and airway inflammation
34 (Nightingale et al., 2000, 000796). High doses of inhaled fluticasone and oral prednisolone have
35 each been reported to reduce inflammatory responses to O3 in healthy individuals (Holz et al., 2005,
36 077170).
37 More recently, Stenfors et al. (2010, 386512) exposed persistent asthmatics (n=13; aged
38 33 years) receiving chronic inhaled corticosteroid therapy to 200 ppb O3 for 2 hours with moderate
39 exercise. An average O3-induced FEVi decrement of 8.4% was observed, whereas, only a 3.0%
March 2011 6-12 DRAFT - DO NOT CITE OR QUOTE
-------
1 FEVi decrement is predicted for similarly exposed age-matched healthy controls (McDonnell et al.,
2 2007, 093104). At 18 hours postexposure, there was a significant O3-induced increase in BAL
3 neutrophils, but not eosinophils. Bronchial biopsy also showed a significant O3-induced increase in
4 mast cells. This study suggests that the protective effect of acute corticosteroid therapy against
5 inflammatory responses to O3 in asthmatics demonstrated by Vagaggini et al. (2007, 196638) may be
6 lost with continued treatment regimes.
Factors Modifying Responsiveness to Ozone
7 Physical activity increases VE and therefore the dose of inhaled O3. Consequently, the intensity
8 of physiological response during and following an acute O3 exposure will be strongly associated
9 with minute ventilation. Apart from inhaled O3 dose and related environmental factors (e.g., repeated
10 daily exposures), individual-level factors, such as health status, age, gender, ethnicity, race, smoking
11 habit, diet, and SES have been considered as potential modulators of a physiologic response to such
12 exposures.
13 Children, adolescents, and young adults (<18 years of age) appear, on average, to have nearly
14 equivalent spirometric responses to O3, but have greater responses than middle-aged and older adults
15 when exposed to comparable O3 doses (U.S. EPA, 1996, 017831). Symptomatic responses to O3
16 exposure, however, appear to increase with age until early adulthood and then gradually decrease
17 with increasing age (U.S. EPA, 1996, 017831). For subjects aged 18-36 years, McDonnell et al.
18 (1999, 010939) reported that symptom responses from O3 exposure also decrease with increasing
19 age. Diminished symptomatic responses in children and the elderly might put these groups at
20 increased risk for continued O3 exposure. Once lung growth and development reaches the peak
21 (18-20 years of age in females and early twenties in males), pulmonary function, which is at its
22 maximum as well, begins to decline progressively with age as does O3 sensitivity.
23 In healthy individuals, the fastest rate of decline in O3 responsiveness appears between the
24 ages of 18 and 35 years (Passannante et al., 1998, 030114: Seal et al., 1996, 044251). more so for
25 females then males (Hazucha et al., 2003, 048168). A model based on laboratory data estimates
26 approximately a 1.1% reduction in FEVi per year over the above age range (Seal et al., 1996,
27 044251). During the middle age period (35-55 years), O3 sensitivity continues to decline but at a
28 much lower rate. Beyond this age (>55 years), acute O3 exposure elicits minimal spirometric
29 changes. Whether the same age-dependent pattern of O3 sensitivity decline also holds for
30 nonspirometric pulmonary function, airway reactivity or inflammatory endpoints has not been
31 determined. Although there is considerable evidence that spirometric and symptomatic responses to
32 O3 exposure decrease with age beyond young adulthood, this evidence comes from cross-sectional
33 analysis and has not been confirmed by longitudinal studies of the same individuals.
34 Several studies have suggested that physiological differences between sexes may predispose
35 females to a greater susceptibility to O3. Lower plasma and nasal lav age fluid (NLF) levels of uric
36 acid (the most prevalent antioxidant) in females, the initial defense mechanism of O3 neutralization
37 in airway surface liquid, may be a contributing factor (Housley et al., 1996, 080811). Consequently,
March 2011 6-13 DRAFT - DO NOT CITE OR QUOTE
-------
1 reduced absorption of O3 in the upper airways may promote its deeper penetration. Dosimetric
2 measurements have shown that the absorption distribution of O3 is independent of gender when
3 absorption is normalized to anatomical dead space (Bush et al, 1996, 080763). Thus, a gender-
4 related differential removal of O3 by uric acid seems to be minimal. In general, the physiologic
5 response of young healthy females to O3 exposure appears comparable to the response of young
6 males (Hazucha et al., 2003, 048168). During the follicular phase of the menstrual cycle, lung
7 function response to O3 may be enhanced (Fox et al., 1993, 043906). but Seal et al. (1996, 044251)
8 later reported no effect of menstrual cycle phase in their analysis of responses of 150 women. Seal et
9 al. (1996, 044251) conceded that the methods used by Fox et al. (1993, 043906) more precisely
10 defined menstrual cycle phase.
11 Only one controlled human exposure study (Seal et al., 1993, 039357) has compared lung
12 function responses of whites (93 M, 94 F) and blacks (92 M, 93 F) exposed to a range of O3
13 concentrations (0-400 ppb). The main effects of gender-race group and O3 concentration were
14 statistically significant (both at p < 0.001), although the interaction between gender-race group and
15 O3 concentration was not significant (p = 0.13). These findings indicate some overall difference
16 between the gender-race groups that is independent of O3 concentration, i.e., the concentration-
17 response curves for the four gender-race groups are parallel. In a multiple comparison procedure on
18 data collapsed across all O3 concentrations for each gender-race group, both black men and black
19 women had significantly larger decrements in FEVi than did white men. The authors noted that the
20 O3 dose per unit of lung tissue would be greater in blacks and females than whites and males,
21 respectively. That this difference in tissue dose might have affected responses to O3 cannot be ruled
22 out. The college students recruited for the Seal et al. (1993, 039357) study are probably from better
23 educated and more SES advantaged families, thus reducing potential for these variables as
24 confounding factors. In a follow-up analysis, Seal et al. (1996, 044251) reported that, of three SES
25 categories, individuals in the middle SES category showed greater concentration-dependent decline
26 in percent-predicted FEVi (4-5% at 400 ppb O3) than low and high SES groups. The authors did not
27 have an "immediately clear" explanation for this finding.
28 Smokers are less responsive to O3 than nonsmokers. Spirometric and plethysmographic
29 pulmonary function decline, nonspecific airway hyperreactivity, and inflammatory response of
30 smokers to O3 were all weaker than data reported for nonsmokers. Although all of these responses
31 are intrinsically related, the functional association between them, as in nonsmokers, has been weak.
32 Similarly, the time course of development and recovery of these effects as well their reproducibility
33 was not different from nonsmokers. Chronic airway inflammation with desensitization of bronchial
34 nerve endings and an increased production of mucus may plausibly explain the pseudo-protective
35 effect of smoking (Frampton et al., 1997, 082692: Torres et al., 1997, 084265).
36 The first line of defense against oxidative stress is antioxidants-rich ELF which scavenge free
37 radicals and limit lipid peroxidation. Exposure to O3 depletes the antioxidant level in nasal ELF
38 probably due to scrubbing of O3 (Mudway et al., 1999, 001270). however, the concentration and the
39 activity of antioxidant enzymes either in ELF or plasma do not appear to be related to O3
March 2011 6-14 DRAFT - DO NOT CITE OR QUOTE
-------
1 responsiveness (Avissar et al., 2000, 012528: Blomberg et al, 1999, 001267: Samet et al, 2001,
2 019034). Carefully controlled studies of dietary antioxidant supplementation have demonstrated
3 some protective effects of a-tocopherol and ascorbate on spirometric lung function from O3 but not
4 on the intensity of subjective symptoms and inflammatory response including cell recruitment,
5 activation and a release of mediators (Samet et al., 2001, 019034: Trenga et al., 2001, 019845).
6 Dietary antioxidants have also been reported to attenuate O3-induced bronchial hyperresponsiveness
7 in asthmatics (Trenga et al., 2001, 019845).
8 Several studies (Bergamaschi et al., 2001, 052670) have reported that genetic polymorphisms
9 of antioxidant enzymes may modulate pulmonary function and inflammatory response to O3
10 challenge. It appears that healthy carriers of NQO1 wild type in combination with GSTM1 null
11 genotype are more responsive to O3. Adults with GSTM1 null only genotype did not show O3
12 hyperresponsiveness. In contrast, asthmatic children with GSTM1 null genotype (Romieu et al.,
13 2004, 056796) were reported to be more responsive to O3. However, in a controlled exposure of
14 mild-to-moderate asthmatics (n=23; 33 ± 11 years) to 300 ppb O3 for 2 hours with moderate
15 exercise, Vagaggini et al. (2010, 387127) found that six of the subjects had a NQOlw^ and GSTM1
16 null, but this genotype was not associated with the changes in lung function or inflammatory
17 responses to O3.
18 Kim et al. (In Press, 674869) also recently reported that GSTM1 genotype was not predictive
19 of FEVi responses in young healthy adults (32 F, 27 M; 25.0 ± 0.5 year) that were roughly half
20 GSTMl-null and half GSTM1-sufficient. Sputum neutrophil levels, measured in a subset of the
21 subjects (13 F, 11 M), were also not significantly associated with GSTM1 genotype.
22 In a study of healthy volunteers with GSTM1 sufficient (n=19; 24 ± 3) and GSTM1 null
23 (n=16; 25 ± 5) genotypes exposed to 400 ppb O3 for 2 hours with exercise, Alexis et al. (2009,
24 628542) found that inflammatory responses but not lung function responses to O3 were dependent on
25 genotype. At 4 hours post O3 exposure, both GSTM1 genotypes had significant increases in sputum
26 neutrophils with a tendency for a greater increase in GSTM1 sufficient than nulls. At 24 hours
27 postexposure, sputum neutrophils had returned to baseline levels in the GSTM1 sufficient
28 individuals. In the GSTM1 null subjects, however, sputum neutrophil levels increased from 4 hours
29 to 24 hours and were significantly greater than both baseline levels and levels at 24 hours in the
30 GSTM1 sufficient individuals. Since there was no FA control in the Alexis et al. (Alexis et al., 2009,
31 628542) study, effects of the exposure other than O3 itself cannot be ruled out. In general, the
32 findings between studies are inconsistent and additional, better controlled studies are needed to
33 clarify the influence of genetic polymorphism on O3 responsiveness.
34 In a retrospective analysis of data from 541 healthy, nonsmoking, white males between the
35 ages of 18-35 years from 15 studies conducted at the U.S. EPA Human Studies Facility in Chapel
36 Hill, North Carolina, McDonnell et al. (2010, 383972) found that increased body mass index (BMI)
37 was found to be associated with enhanced FEVi responses. The BMI effect was of the same order of
38 magnitude but in the opposite direction of the age effect where by FEVi responses diminish with
39 increasing age. In a similar retrospective analysis, Bennett et al. (2007, 418827) found enhanced
March 2011 6-15 DRAFT - DO NOT CITE OR QUOTE
-------
1 FEVi decrements following O3 exposure with increasing BMI in a group of 75 healthy, nonsmoking,
2 women (age 24 ± 4 years; BMI range 15.7 to 33.4), but not 122 healthy, nonsmoking, men (age 25 ±
3 4 years; BMI range 19.1 to 32.9). In the women, greater O3-induced FEVi decrements were seen in
4 overweight (BMI >25) than in normal weight (BMI from 18.5 to 25), and in normal weight than in
5 underweight (BMI <18.5) (P trend< 0.022). Higher BMI may be a risk factor for adverse pulmonary
6 effects associated with O3 exposure.
Repeated Ozone Exposure Effects
7 Based on studies reviewed in previous O3 AQCDs (U.S. EPA, 1986, 017607: U.S. EPA, 1996,
8 017831: U.S. EPA, 2006, 088089). several conclusions can be drawn about repeated 1- to 2-h O3
9 exposures. Repeated exposures to O3 causes an enhanced (i.e., greater) pulmonary function response
10 on the second day of exposure. The enhanced response appears to depend to some extent on the
11 magnitude of the initial response (Horvath et al., 1981, 039221). Small responses to the first O3
12 exposure are less likely to result in an enhanced response on the second day of O3 exposure
13 (Folinsbee et al., 1994, 044189). With continued daily exposures (i.e., beyond the second day) there
14 is a substantial (or even total) attenuation of pulmonary function responses, typically on the third to
15 fifth days of repeated O3 exposure. This attenuation of responses is lost in 1 week (Kulle et al., 1982,
16 040668: Linn et al., 1982, 039646) or perhaps 2 weeks (Horvath et al., 1981, 039221) without O3
17 exposure. In temporal conjunction with pulmonary function changes, symptoms induced by O3 (e.g.,
18 cough, pain on deep inspiration, and chest discomfort), are increased on the second exposure day and
19 attenuated with repeated O3 exposure thereafter (Folinsbee et al., 1980, 038880: Folinsbee et al.,
20 1998, 038663: Foxcroft and Adams, 1986, 040463: Linn et al., 1982, 039646). In longer-duration
21 (4-6.6 hours), lower-concentration studies that do not cause an enhanced second-day response, the
22 attenuation of response to O3 appears to proceed more rapidly (Folinsbee et al., 1994, 044189).
23 Inflammatory markers from BALF on the day following both 2 hours (Devlin et al., 1997, 083577)
24 and 4 hours (Christian et al., 1998, 029925: Torres et al., 2000, 005654) repeated O3 exposure for
25 4 days indicate that there is ongoing cellular damage irrespective of the attenuation of some cellular
26 inflammatory responses of the airways, lung function and symptoms response.
6.2.1.2. Epidemiology
27 Among epidemiologic studies reviewed in the 1996 and 2006 O3 AQCDs, increases in ambient
28 O3 exposure were consistently associated with lung function decrements in groups with higher
29 expected personal O3 exposures and higher exertion levels, including children attending summer
30 camps and adults exercising or working outdoors (U.S. EPA, 1996, 017831: U.S. EPA, 2006,
31 088089). An equally strong body of epidemiologic evidence demonstrated O3-associated lung
32 function decrements in children, especially those with pre-existing respiratory disease such as
33 asthma. These epidemiologic findings, in particular, ambient O3-associated decreases in lung
34 function in healthy populations with increased outdoor exposures, are well-supported by findings
35 from human controlled exposure studies. Recent epidemiologic studies contributed mixed evidence
March 2011 6-16 DRAFT - DO NOT CITE OR QUOTE
-------
1 of association between ambient O3 exposure and lung function; however, a majority of studies
2 conducted in asthmatic children indicated decreases in lung function in association with increases in
3 ambient O3 exposure. Newer data on children attending camps, outdoor workers, and other healthy
4 populations were limited, and across these studies, ambient O3 exposure was associated with both
5 decreases and increases in lung function. Recent studies build upon the extant body of evidence by
6 providing additional data to assess important lags of O3 exposure associated with decrements in lung
7 function; confounding by co-pollutants; and potential susceptibility due to corticosteroid (CS) use,
8 genetic polymorphisms, obesity, and diet.
Populations with Increased Outdoor Exposures
9 Few epidemiologic studies characterizing acute O3-related respiratory morbidity have
10 accounted for time spent outdoors, which may be an important determinant of interindividual
11 variability in personal O3 exposure. Relative to other epidemiologic studies, studies of subjects
12 engaged in outdoor recreation, exercise, or work may be more comparable to controlled exposure
13 studies because of better-estimated personal O3 exposures and examination of O3 effects during
14 exertion when the dose of O3 reaching the lungs may be higher because of higher ventilation and
15 inhalation of larger volumes of air. Characteristics and ambient O3 concentration data from these
16 epidemiologic studies are presented in Table 6-1. The collective body of evidence clearly
17 demonstrates decrements in lung function in association with O3 exposures during outdoor exertion
18 or exercise (Figures 6-3 to 6-5 and Tables 6-2 to 6-4). A large number of older studies comprise a
19 majority of the supporting evidence, whereas recent studies, which were far fewer in number,
20 provide less compelling evidence. In addition to the consistency of associations among
21 epidemiologic studies, the parallel findings from human controlled exposure studies indicating that
22 lower O3 exposures induce decrements in lung function when combined with exercise as compared
23 with exposures during rest (Section 6.2.1.2), strengthen the evidence for increases in ambient O3
24 exposure producing decrements in lung function.
March 2011 6-17 DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-1. Mean and upper percentile concentrations of ozone in epidemiologic studies examining
lung function in populations with increased outdoor exposures
Study
Korricketal. (1998, 026841)
Thurstonetal. (1997,0776451
Spektoretal. (1988, 041710)
Spektoretal. (1988, 040904)
Spektoretal. (1991,042612)
Neasetal. (1999, 0034661
Girardotetal. (2006,0882711
Selwynetal. (1985, 041356)
Thaller et al. (2008, 1958691
Higginsetal. (1990, 0421951
Avoletal. (1990. 0423661
Burnett etal. (1990, 670386)
Raizenneetal. (1989,
0417001
Braueretal. (1996, 0807541
Castillejos etal. (1995,
0784851
Romieuetal. (1998, 0867561
Nickmilder etal. (2007,
0907101
Brunekreef etal. (1994,
0451611
Hoeketal. (1993, 0430091
Braun-Fahrlanderetal. (1994,
0386651
Hoppe et al. (2003, 0556181
Location
Mt. Washington, NH
Connecticut River Valley,
CT
Tuxedo, NY
Fairview Lake, NJ
Fairview Lake, NJ
Philadelphia, PA
Great Smoky Mountain
National Park, TN
Houston, TX
Galveston, TX
San Bernardino, CA
Idyllwild, CA
LakeCouchining,
Ontario, CA
Lake Erie, Ontario, CA
British Columbia,
Canada
Mexico City, Mexico
Mexico City, Mexico
southern Belgium
Netherlands
Wageningen,
Netherlands
southern Switzerland
Munich, Germany
Years/Season
1991, 1992
Warm season
1991-1993
Warm season
1985
Warm season
1984
Warm season
1988
Warm season
1993
Warm season
2002-2004
Warm season
1981
Warm season
2002-2004
Warm season
1987
Warm season
1988
Warm season
1983
Warm season
1986
Warm season
1993
Warm season
1990
Warm season
1996
Warm season
2002
Warm season
1981
Warm season
1989
Warm season
1989
Warm season
1992
Warm season
M^tri,. r™^ntr*n™ Middle/Upper Percentile
(Ppbj Concentrations (ppb)
Hike duration ,n
(2-1 2 h) w
1-hmax 83.6
1-havg NR
1-havga 53
1-havga 69
12-havg 57.5
(9:00-21:00) 55.9
Hike duration ,B,b
(2-9 h) 4M
15-mmax 47
1-hmax NR
1-havga 59
1-havga 94
1-havga 59
1-havga 71
1-hmax 40
1-hmax 149
1-hmax 123
1-hmax MP
8-h max NR
1-hmax 40
24-h avg NR
1-hmax NR
21-74
Range: 20-1 60
Range: 21-124
Range (1-hmax): 40->100
Range (1-h max): 40-150
Campl) IQR: 19.8 (Camp 1)
Camp 2) IQR:21.9(Camp2)
Range: 25.0-74.2
Range: 4-135
Median: 35
Range: 19-118
25-245
Approximate range (1-h max): 60-1 60C
Maximum: 95
Range (1-h max): < 10-143
Range: 13-84
49-365
NR
24.5-112.7°
18.9-81. 1d
Range: 10-100
Range: 25-1 20
Range: 20-80
, n h High days: 65.9 Max (high days): 86
i^-nmax Control days: 27.2 Max (control days): 39
IQR = interquartile range, NR = not reported, Max = maximum
a1-h avg, at the time of afternoon lung function measurement.
Personal exposure estimates were derived based on time-activity diary data.
"Quantitative results not presented. Concentrations estimated from data presented in a figure.
Concentrations converted from pg/m to ppb using the conversion factor of 0.51 assuming standard temperature (25°C) and pressure (1 atm).
1 In the 1996 O3 AQCD (U.S. EPA, 1996, 017831). studies of children attending summer camps
2 were noted for their on-site measurement of ambient O3 and repeated assessment of lung function
3 over 1- to 2-week periods (Avol et al., 1990, 042366: Berry et al., 1991, 042377: Burnett et al., 1990,
4 670386: Higgins et al., 1990, 042195: Raizenne et al., 1987, 040903: Raizenne et al., 1989, 041700:
5 Spektor and Lippmann, 1991, 042612: Spektor et al., 1988, 040904: Thurston et al., 1997, 077645).
6 In groups mostly comprising healthy children across heterogeneous geographic locations,
March 2011
6-18
DRAFT - DO NOT CITE OR QUOTE
-------
1 decrements in FEVi were consistently observed in association with ambient O3 exposures averaged
2 over the 1-8 hours preceding lung function measurement (Figure 6-3 and Table 6-2). Kinney et al.
3 (1996, 079203) corroborated this consistency in a reanalysis combining data from nonasthmatic
4 subjects from six studies (Avol et al., 1990, 042366; Burnett et al., 1990, 670386; Higgins et al.,
5 1990, 042195; Raizenne et al., 1987, 040903; Spektor and Lippmann, 1991, 042612; Spektor et al.,
6 1988, 040904). Based on uniform statistical methods, a 40-ppb increase in concurrent-hour O3
7 concentration was associated with a -20 mL (95% CI: -25, -14) change in afternoon FEVi (Kinney et
8 al., 1996, 079203). Study-specific effect estimates ranged between a 0.76 and 48 mL decrease (per
9 40 ppb O3) and were observed in locations with mean afternoon 1-h avg O3 concentrations between
10 53 and 123 ppb. In contrast with these previous findings, in a recent cross-sectional analysis of 72
11 children attending 6 different camps in Belgium, children at camps with higher daily 1-h max O3
12 concentrations did not consistently have greater decreases in intraday FEVi or FEVi/FVC
13 (Nickmilder et al., 2007, 090710). In camp studies, associations between O3 exposure and peak
14 expiratory flow (PEF) were more variable as characterized by the wider range in effect estimates and
15 wider 95% CIs (Figure 6-4 and Table 6-3). Nonetheless, most effect estimates indicated decreases in
16 PEF in association with ambient O3 exposure, with the largest effect estimated in a group of
17 asthmatic campers (Thurston et al., 1997, 077645).
March 2011 6-19 DRAFT - DO NOT CITE OR QUOTE
-------
Spektoretal. (1988)
Spektorand Lippmann
(1991)
Raizenne etal. (1987)
Burnettetal.(1990)
Higgins etal. (1990)
Avol etal. (1990)
Kinney etal. (1996)
PEF (ml/sec)
Spektor etal. (1988)
Raizenne etal. (1987)
Burnettetal.(1990)
Higgins etal. (1990)
Avol etal. (1990)
Kinney etal. (1996)
Neasetal. (1999)a
Thurstonetal. (1997)
Population
Nonasthmatic campers
Nonasthmatic campers
Nonasthmatic campers
Nonasthmatic campers
Nonasthmatic campers
Nonasthmatic campers
Pooled estimate
Nonasthmatic campers
Nonasthmatic campers
Nonasthmatic campers
Nonasthmatic campers
Nonasthmatic campers
Pooled estimate
Nonasthmatic and
asthmatic campers
Asthmatic campers
-160 -140 -120 -100 -80 -60 -40 -20 0 20 40 60 80
Effect Estimate (95% Cl)
"Neas et al. (1999, 003466) examined 12-h avg O3 exposure (9:00 a.m. - 9:00 p.m.). All of the other studies examined O3 exposure in the 1
hour preceding afternoon lung function measurement.
Figure 6-3. Changes in FEVi (ml) or PEF (mL/sec) in association with ambient ozone exposure
in studies of children attending summer camp. [Effect estimates are standardized to
a 40- or 30-ppb increase in 1-h or 12-h avg ozone, respectively. All effect estimates
are from single pollutant models.
March 2011
6-20
DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-2. Additional characteristics and quantitative data for studies represented in Figure 6-3
Study
Location
Population
Effect Estimate (95% Cl)a
FEV, (ml)
Spektoretal. (1988, 040904)
Spektorand Lippmann (1991, 042612)
Raizenneetal. (1989, 041700)
Burnett etal. (1990, 670386)
Higginsetal. (1990, 042195)
Avol etal. (1991,042613)
Kinnevetal. (1996, 079203)
Lake Fairview, NJ
Lake Fairview, NJ
Lake Erie, Ontario, Canada
LakeCouchiching, Ontario,
San Bernadino, CA
Pine Springs, CA
Pooled analysis
Nonasthmatic campers
Nonasthmatic campers
Nonasthmatic campers
Canada Nonasthmatic campers
Nonasthmatic campers
Nonasthmatic campers
Nonasthmatic campers
-20.0 (-32.5, -7.5)
-51. 6 (-72.8, -30.4)
-7.6 (-42. 1,26.9)
-11.6 (-19. 4, -3.8)
-33.6 (-49.3, -17.9)
-12.8 (-23.0, -2.6)
-20.0 (-25.5, -14.5)
PEF (ml/sec)
Spektoretal. (1988, 040904)
Raizenneetal. (1989, 041700)
Burnett etal. (1990, 670386)
Higginsetal. (1990, 042195)
Avol etal. (1991,042613)
Kinnevetal. (1996, 079203)
Neasetal. (1999, 003466)
Thurstonetal. (1997,077645)
Lake Fairview, NJ
Lake Erie, Ontario, Canada
LakeCouchiching, Ontario,
San Bernardino, CA
Pine Springs, CA
Pooled analysis
Philadelphia, PA
CT River Valley, CT
Nonasthmatic campers
Nonasthmatic campers
Canada Nonasthmatic campers
Nonasthmatic campers
Nonasthmatic campers
Nonasthmatic campers
Nonasthmatic and asthmatic campers
Asthmatic campers
-80.0 (-147.3, -17.0)
-106.4 (-209.9, -2.9)
-4.0 (-30.7, 22.7)
-44.0 (-105.0, 17.2)
86.8(31.9,141.7)
6.8 (-19.1, 32.7)
-27.5 (-70.8, 15.75)
-146.7 (-261.7, -31.7)
aAII effect estimates are standardized to a 40-ppb increase in 1-h avg 03, except that from Neasetal. (1999, 003466), which is standardized to a 30-ppb
increase in 12-h avg 03'
1 Similar to the camp studies, studies of subjects exercising outdoors collectively show that low-
2 level exposures O3 (range of mean concentrations: 40-149 ppb) during short periods (10-60 minutes)
3 of moderate to heavy exercise are associated with decreases in lung function, with stronger evidence
4 of association observed among adults than among children (Figure 6-4 and Table 6-3). These studies
5 were noted for the repeated examination of subjects over days with a wide range in ambient O3
6 concentrations and assessment of O3 exposures during discrete outdoor exercise periods. Further,
7 results from these studies were consistent with those from human controlled exposure studies
8 indicating that lower concentrations of O3 exposures induced lung function decrements when
9 combined with exercise as compared with exposures during rest. In the more limited set of studies of
10 adult day-hikers that examined variable multihour exposures during one period of exercise, results
11 were mixed (Girardot et al, 2006, 088271: Korrick et al, 1998, 026841). Both Girardot et al. (2006,
12 088271) (n = 354) and Korrick et al. (1998, 026841) (n = 530) were large studies of predominantly
13 white, healthy adults hiking in Great Smoky Mountains National Park, TN and Mt. Washington, NH,
14 respectively. Korrick et al. (1998, 026841) reported a posthike decline of 1.4% (95% CI: -2.4, -0.30)
15 in FEVi per 30-ppb increase in 8-h avg O3. In contrast, Girardot et al. (2006, 088271) found that O3
16 exposure was associated with a posthike increase in FEVi (0.72% [95% CI: -0.46, 1.90]) per 30 ppb
17 increase in 8-h avg O3). In Korrick et al. (1998, 026841). effect estimates for O3 with FVC,
18 FEVi/FVC, FEF25-75%, and PEF were negative but associated with wide 95% CIs; however, similar
19 associations in Girardot et al. (2006, 088271) were in mixed directions. To explain discrepancies in
20 findings between studies, Girardot et al. (2006, 088271) pointed to their exclusion of 367 subjects
21 (61%) for failure to provide at least 2 acceptable spirometry tests, compared with 31% excluded in
22 Korrick et al. (1998, 026841). Excluded subjects in Girardot et al. (2006, 088271) had significantly
March 2011
6-21
DRAFT - DO NOT CITE OR QUOTE
-------
1 higher mean O3 exposure, tended to be regular exercisers, and had small posthike increases in lung
2 function, which also may have contributed to weak findings for O3. Additional explanations provided
3 by Girardot et al. (2006, 088271) included their use of a larger number of untrained technicians and
4 shorter mean duration of hike (5 hours versus 8 hours).
Study
Korricketal. (1998)
Girardot etal. (2006)
Selwyn et al. (1 985)a
Spektoretal. (1988)
Brunekreef etal. (1994)
Hoppe etal. (2003)
Castillejosetal. (1995)
Hoek etal. (1993)
Braun-Fahrlanderetal. (1994)
Population
Adults hiking
Adults hiking
Adults exercising
Adults exercising
Adults exercising
Adults exercising
Children exercising
Children exercising
Children exercising
Parameter
FEV1 (% change x 1 0) -•-
PEF (nWsec) — •-
-40 -20 0
Effect Estimate (95% Cl)
"The 95% CI was constructed using a standard error that was estimated from the p-value.
Figure 6-4. Changes in FEVi (ml or percent change) or PEF (mL/sec) in association with
ambient ozone exposures of adults and children during outdoor exercise. [Effect
estimates are standardized to a 40-ppb increase in ozone exposures in the range of
15 minutes to 1 hour, and a 30-ppb increase for mean ozone exposures in the range
of 3 to 8 hours. All effect estimates are from single pollutant models.
March 2011
6-22
DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-3. Additional characteristics and quantitative data for studies represented in Figure 6-4
Study
Location
Population
Parameter
O3 Averaging
Time
Effect Estimate
(95% Cl)a
Korricketal. (1996, 026481)
It. Washington, NH
Adult day hikers FEVi (percent change) 8-h avg
-1.4 (-2.4,-0.30)
Girardotetal. (2006,0882711
Great Smoky Mt, TN
Adult day hikers FEVi (percent change) 8-h avg
0.72 (-0.46, 1.90
Selwynetal. (1985. 041356)b
Houston, TX
Adults exercising FEV, (ml)
15-mmax
-16 (-31.1,-0.87)°
Spektoretal. (1962, 0147101
Tuxedo, NY
Adults exercising FEV, (ml)
30-m avg
-54 (-84.1,-26.6)
Brunekreefetal. (1994, 045161)
Netherlands
Adults exercising FEVi (ml)
10-mto1-h
-20.8 (-41.2,-0.42)
Hoppe et al. (2003, 0556181
Munich, Germany
Adults exercising FEV, (percent change) 3-h avg (8:00-11:00) -0.01 (-0.10, 0.09
Castillejos et al. (1995, 078485.)
Mexico City, Mexico
Children exercising FEV, (percent change) 1-havg
-0.48 (-0.72, -0.24)
Hoeketal. (1993, 0430091
Wageningen, Netherlands Children exercising PEF (mL/sec)
1 -h avg
-2.8 (-7.4, 1.9)
Braun-Fahrlanderetal. (1994, 0386661 Switzerland
Children exercising
1 -h avg
27.2 (-7.2, 61.6)
-88 (-166.4,-9.6)
'Effect estimates are standardized to a 40-ppb increase in 03 exposures in the range of 15 minutes to 1 hour and a 30-ppb increase for mean 03
exposures in the range of 3 to 8 hours.
The 95% Cl was constructed using a standard error that was estimated from the p-value.
1 Ambient O3 exposure has been associated consistently with decrements in lung function
2 among outdoor workers (Figure 6-5 and Table 6-4). In particular, Brauer et al. (1996, 080754) was
3 noted for the low ambient O3 concentrations (workshift mean [SD]: 26.0 ppb [11.8]), long outdoor
4 workshifts of the 58 berry pickers (11 hours) and a larger O3-associated decrease in afternoon FEVi
5 (-152 mL [95% CI: -183, -121] per 40 ppb increase in 1-h max O3) compared with those observed in
6 studies of exercising adults with higher exertion levels. Brauer et al. (1996, 080754) also found that
7 workday O3 exposure was associated with a greater decrement in FEVi on the next morning
8 (-180 mL [95% CI: -227, -133] per 40 ppb increase in 1-h max O3), indicating a delayed or persistent
9 effect. Consistent with previous findings, a recent study of lifeguards in Galveston, TX found that O3
10 exposure during 6-8 hour workshifts was associated with decrements in FEVi/FVC (Thaller et al.,
11 2008, 195869). In this study, 142 mostly white males, 16-27 years of age were followed for at least
12 one summer from 2002 to 2004. Among all subjects, a 40 ppb increase in 1-h max O3 was associated
13 with a 0.4% decrease (95% CI: -0.8, 0) in afternoon FEVi/FVC. A similar magnitude of effect was
14 estimated in a co-pollutant model that included daily max NO2 and daily avg PM2 5. Ozone was not
15 associated with either FEVi or FVC individually.
March 2011
6-23
DRAFT - DO NOT CITE OR QUOTE
-------
Study
Population Parameter Lag Subgroup
Thaller etal. (2008) Lifeguards FVC (% change) 0
FEV1/FVC(% change) 0
Braueretal. (1996) Berrypickers FEV1 (L) 0
1
Romieuetal.(1998) Streetworkers FEV1 (L)
Hoppe etal. (2003) Forestry workers FEV1 (L)
PEF (L/sec)
0 Placebo
Antioxidant supplement
0
0 i
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4
Effect Estimate (95% Cl)
0.6 0.8
Figure 6-5. Changes in lung function parameters in association with ambient ozone exposures
among outdoor workers. [Effect estimates are standardized to a 40-ppb increase for
1/2-h or 1-h max ozone and a 30-ppb increase for 8-h max ozone. All effect estimates
are from single pollutant models.
Table 64. Additional characteristics and quantitative data for studies represented in
Study
Thaller etal. (2008,
1958691
Braueretal. (1996,
080754)
Romieuetal. (1998,
0867561
Hoppe et al. (2003,
055618)
Location
Galveston, TX
British Columbia,
Canada
Mexico City, Mexico
Munich, Germany
Population
Lifeguards
Berry pickers
Streetworkers
Forestry
workers
Parameter
FVC (percent change)
FEWFVC (percent
change)
FEV, (ml)
FEV, (ml)
FEV, (ml)
PEF (mL/sec)
Averaging Lgg SubgrQup
8-h max 0
1-h max .
Placebo
1-h max 0 Antioxidant
supplement
1/2-h max n
(13:00-16:00)
Figure 6-5
Effect Estimate
(95% Cl)a
0.24 (-0.28, 0.72)
-0.40 (-0.80, 0)
-152 (-183, -121)
-180 (-227, -133)
-71. 6 (-113.9, -29.3)
-17.6 (-68.6, 33.4)
-56 (-118.4, 6.4)
-460 (-81 6, -107)
'Effect estimates are standardized to a 40-ppb increase for 1/2-h or 1-h max 03anda 30-ppb increase for 8-h max 03.
Asthmatic Children
1 Studies of asthmatic children generally demonstrate that increases in ambient O3 exposure are
2 associated with decrements in PEF and FEVi (Figures 6-6 and 6-7 and Tables 6-6 and 6-7).
3 Characteristics and ambient O3 concentration data from these epidemiologic studies are presented in
4 Table 6-5. The most geographically representative data were provided by the 7-U.S. city Inner-City
5 Asthma Study (ICAS) of 861 children with persistent asthma and atopy (O'Connor et al., 2008,
March 2011
6-24
DRAFT - DO NOT CITE OR QUOTE
-------
1 156818). Using year-round data, investigators observed that a 20-ppb increase in the lag 1-5 avg of
2 24-h avg O3 was associated with decreases of 0.41 (95% CI: -1.0, 0.21) in percent predicted FEVi
3 and of 0.22 (95% CI: -0.86, 0.43) in percent predicted PEF. Ozone was associated with larger
4 decreases in lung function in co-pollutant models with PM2 5 and NO2. Lag 1-5 avg O3 also was
5 negatively associated with morning PEF in the 1993 National Cooperative Inner City Asthma Study
6 (NCICAS) of different children (n = 846) from the same cities plus Boston, MA (Mortimer et al.,
7 2002, 030281). Consistent with human controlled exposure studies (Section 6.2.1.2), Mortimer et al.
8 (2002, 030281) found that increasing O3 exposure was associated with an increased incidence^>f
9 10% declines in PEF (OR: 1.30 [95% CI: 1.04, 1.61] per 30 ppb increase in lag 1-5 of 8-h avg O3),
10 demonstrating that O3 exposure is related to clinically important changes in lung function in
11 asthmatic children.
12 In addition to these multicity studies and the study of asthmatic children attending summer
13 camps described earlier (Thurston et al., 1997, 077645). several smaller studies conducted in the
14 U.S., Mexico City, and Europe also found associations between ambient O3 exposure and
15 decrements in PEF among asthmatic children (Figures 6-6 and 6-7 and Tables 6-6 and 6-7).
March 2011 6-25 DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-5. Mean and Upper Percentile Concentrations of Ozone in Epidemiologic Studies Examining
Lung Function in Asthmatic Children
Study
Mortimer etal. (2002,0302811
O'Connor etal. (2008, 1568181
Lewis et al. (2005, 0810791
Rabinovitch et al. (2004, 0967531
Dales et al. (2009, 594285)
Liu et al. (2009, 1920031
Romieuetal. (1996, 0807481
Romieuetal. (1997, 0858071
Romieu et al. (2002, 0347111
Romieu et al. (2004, 056796)
Romieu et al. (2006, 0909691
Barraza-Villarreal et al. (2008,
156254)
Romieu et al. (2009, 5487881
Hernandez-Cadena et al. (2009,
594283
Gielenetal. (1997,0835921
Hoppe et al. (2003, 0556181
Wiwatanadate and Trakultivakorn
(2010.3877061
Jalaudin et al. (2000, 0119291
Location
8 U.S. communities
(NCICAS)
7 U.S. communities
(ICAS)
Detroit, Ml
Denver, CO
Windsor, ON, Canada
Northern Mexico City,
Mexico
Southern Mexico City,
Mexico
Mexico City, Mexico
Mexico City, Mexico
Mexico City, Mexico
Amsterdam,
Netherlands
Munich, Germany
Chiang Mai, Thailand
Sydney, Australia
Years/Season
1993
Warm season
1998-2001
All-year
2001-2002
All-year
2000-2003
Cold season
2005
Cold season
1991-1992
Warm and cold
season
1991-1992
Warm and cold
season
1998-2000
All-year
2003-2005
All-year
2005
Warm season
1995
Warm season
1992-1995
Warm season
August 2005-June
2006
February-December
1994
Metric
8-h avg
(10:00a.m.-6:00p.m.)
24-h avg
8-h max
1-h max
24-h avg
1-h max
1-h max
1-h max
8-h max
1-h max
8-h max
1-h max
24-h avg
1-h max
8-h max
1/2-hmax
24-h avg
24-h avg
Mean
Concentration
(PPb)
48
NR
Eastside: 40.4°
Westside:41.4°
28.2°
14.1
27.2
190
196
66.2
102
31.6
86.5
26.3
74.5
33.5
High 03 days: 65.9
Control days: 27.2
17.5
12
Middle/Upper Percentile
Concentrations (ppb)
Approximate IQR = 15a
Approximate median: 20a
Approximate range: 2-50a
Overall IQR: 16.0
Overall range: 14.8-92.0
Median: 30.0
Range: 0-70.0
Median: 13.0; IQR: 8.8-17.8
Median: 27.0; IQR: 21. 8-32.8
Range: 40-370
Range: 40-390
Range: 11.1-142.5
Range: 12-309
IQR: 22.0 (8-h); Range: 4.9-
86.3
IQR:48.0
IQR: 17.9-35.3; Range: 9.0-
62.8
IQR: 46.5-92.5; Range: 26.0-
165.0
Range: 13.8-55.4
High 03days: 65.9-70.4 (range)
90th percentile: 26.82
Range: 5.55-34.65
IQR: 8.3
Maximum: 43
NCICAS = National Cooperative Inner-City Asthma Study, IQR = interquartile range, ICAS = Inner City Asthma Study, NR = Not Reported
'Quantitative results not presented. Concentrations estimated from data presented in a figure.
"Measured at sites established by investigators.
March 2011
6-26
DRAFT - DO NOT CITE OR QUOTE
-------
Study
Lag
Subgroup
O'Connor etal. (2008) % predicted FEV1 1-5 avg
% predcited PEF
Mortimer etal. (2002) PEF (% change) 1-5 avg
Lewis etal. (2005) FEV1 (% change) 2
Gielen etal. (1997) PEF (% change) 2
Dales etal. (2009) % predicted FEV1 0
Liu etal. (2009) FEV1 (% change) 0
CS user
With URI
-6 -4 -2 0
Effect Estimate (95% Cl)
Figure 6-6. Changes in lung function parameters (percent-predicted or %change) in association
with ambient ozone exposures among asthmatic children. (€S = corticosteroid, URI
= Upper respiratory infection. Effect estimates are standardized to a 40-, 30-, and 20-
ppb increase for 1-h max, 8-h max, and 24-h avg ozone, respectively. All effect
estimates are from single pollutant models.
Table 6-6. Additional characteristics and
Study
O'Connor etal. (2008,
1568181
Mortimer etal. (2002,
0302811
Lewis et al. (2005, 0810791
Gielen etal. (1997,0835921
Dales et al. (2009, 5942851
Liu et al. (2009, 1920031
Location/
Population
7 U.S. communities
Asthmatic children
8 U.S. communities
Asthmatic children
Detroit, Ml
Asthmatic children
Amsterdam,
Netherlands
Asthmatic children
Windsor, ON, Canada
Asthmatic children
Windsor, ON, Canada
Asthmatic children
03
Lag
1 -5 avg
1 -5 avg
2
2
0
0
quantitative data for studies represented in Figure 6-6
O3 Averaging
Time
24-h avg
8-h avg
(10:00-18:00)
8-h max
8-h max
1-h max
24-h avg
Parameter
% predicted FEV,
% predicted PEF
% predicted PEF
percent change, lowest
daily FEV,
percent change, PEF
% predicted FEV,
percent change, FEV,
Effect Estimate
Subgroup (gs% c|)a
-0.41 (-1.03,0.21)
-0.22 (-0.86, 0.43)
-1.2 (-2.1, -0.26)
CSuser -8.0 (-13.5, -2.1)
With URI -5.4 (-11. 3, 1.0)
-1.34 (-2.58, -0.10)
-0.47 (-11. 67, 2.25)
-0.89 (-3.5, 1.8)
'Effect estimates are standardized to a 40-, 30-, and 20-ppb increase for 1 -h max, 8-h max, and 24-h avg 03, respectively.
March 2011
6-27
DRAFT - DO NOT CITE OR QUOTE
-------
Study
Thurstonetal. (1997)
Romieuetal. (1996)
Romieuetal. (1997)
Romieuetal. (2002)
Romieu et al. (2004)
Romieuetal. (2006)
Hoppe et al. (2003)
Jalaludinetal. (2000)
Wiwatanadate and
Trakultivakorn(2010)
Rabinovitchetal. (2004)
Parameter
PEF (L/min)
Evening PEF (L/min)
Evening PEF (L/min)
FEV1 (ml/10)
FEF25%-75% (L/min)
FEV1 (ml/10)
FEV1 (ml/10)
FEV1 (ml/10)
Daily change PEF (L/min)
Daily avg PEF (L/min)
Morning FEV1 (ml/10)
Lag
0
2
0
2
1
1
1
0-4 avg
0
0
0
0-2 avg
Subgroup
Placebo — •-
Antioxidant —
Placebo, moderate/severe
asthma ~
Antioxidant, moderate/severe
asthma
Placebo, GSTM1 positive •-
GSTP1 He/lie Ile/Val — •—
GSTP1 Val/Val
»
— •
-10
-6 -2 2
Effect Estimate (95% Cl)
Figure 6-7. Changes in lung function parameters (L/min or mL/10) in association with ambient
ozone exposures among asthmatic children. IAHR = airway hyperresponsiveness.
Effect estimates are standardized to a 40-, 30-, and 20-ppb increase for 1-h (or 1/2-h)
max, 8-h max, and 24-h avg ozone, respectively. All effect estimates are from single
pollutant models.
March 2011
6-28
DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-7. Additional characteristics and quantitative data for studies represented in Figure 6-7
Study
Thurstonetal. (1997,0776451
Romieuetal. (1996. 080748)
Romieuetal. (1997. 085807)
Romieu et al. (2002, 0347111
Romieu et al. (2004, 0567961
Romieu et al. (2006, 0909691
Barraza-Villarreal et al. (2008,
1562541
Hoppe et al. (2003, 0556181
Jalaudin et al. (2000, 0119291
Wiwatanadate and Trakultivakorn
(2010,3877061
Rabinovitch et al. (2004, 0967531
Location/
Population
CT River Valley, CT
Asthmatic campers
Northern
Mexico City, Mexico
Asthmatic children
southern
Mexico City, Mexico
Asthmatic children
Mexico City, Mexico
Asthmatic children
Moderate/severe
asthmatics
Mexico City, Mexico
Asthmatic children
Mexico City, Mexico
Asthmatic children
Mexico City, Mexico
Children
Munich, Germany
Asthmatic children
Sydney, Australia
Asthmatic children
Chiang Mai,
Thailand
Asthmatic children
Denver, CO
Asthmatic children
03
Lag
0
0
2
0
2
1
1
1
1-5
avg
0
0
0
0-3
avg
03
Averaging
Time
1 -h avg
1-h max
1-h max
1-h max
1-h max
1-h max
8-h max
1/2-hmax
24-h avg
24-h avg
1-h max
Parameter
PEF (L/min)
Evening PEF
(L/min)
Evening PEF
(L/min)
FEV,(mL/10)
FEF25%-75%
(L/min)
FEV,(mL/10)
FEV,(mL/10)
FEV,(mL/10)
Daily change
PEF (L/min)
Daily avg PEF
(L/min)
Morning FEV,
(mL/10)
Subgroup
Placebo
Antioxidant
Placebo, moderate/severe asthma
Antioxidant, moderate/severe ashtma
Placebo, GSTM1 null
Placebo, GSTM1 sufficient
Supplement, GSTM1 null
Supplement, GSTM1 sufficient
GSTP1 lie/lie or I le/Val
GSTP1 Val/Val
Nonasthmatics
asthmatics
NoAHR
AHR
Effect Estimate
95% Cl)a
-8.8 (-15.7, -1.90)
-0.45 (-2. 16, 1.26)
-1.50 (-3.60, 0.53)
-1.45 (-2.88, -0.02)
-0.1 7 (-1.95, 1.62)
-0.36 (-1.35, 0.63)
0.08 (-0.98, 1.13)
-1.88 (-3.42, -0.34)
-0.07 (-1.5, 1.37)
-2.4 (-4.3, -0.52)
-0.50 (-1.86, 0.85)
-0.1 4 (-1.95, 1.67)
0.24 (-1.43, 1.92)
-0.86 (-1.1 8, 0.08)
0.84 (-0.42, 2.1)
-2.9 (-9. 1,3.3)
-0.22 (-3.8, 3.4)
-8.4 (-19.64, 2.84)
-0.71 (-2.6, 1.17)
-5.2 (-8.27, -2.18)
1.0 (-1.6, 3.6)
5.3 (-0.24, 10.8)
AHR = airway hyperresponsiveness.
'Effect estimates are standardized to a 40, 30, and 20 ppb increase for 1-h (or 1/2-h) max, 8-h max, and 24-h avg 03, respectively.
1 Among the studies that examined FEVi, evidence of association with ambient O3 exposure
2 was stronger in particular subgroups of asthma severity, comorbid conditions, or antioxidant capacity
3 (Jalaludin et al., 2000, 011929: Lewis et al., 2005, 081079: Romieu et al., 2004, 056796: Romieu et
4 al., 2006, 090969) than among asthmatics overall (Barraza-Villarreal et al., 2008, 156254: Lewis et
5 al., 2005, 081079: Romieu et al., 2002, 034711). Demonstrating varying susceptibilities within a
6 group of asthmatic children, Jalaudin et al. (2000, 011929) estimated a greater effect in asthmatics
7 with airway hyperresponsiveness (AHR), and Hoppe et al. (2003, 055618) found that 20% of their
8 asthmatic subjects experienced a greater than 10% decline in FEVi in association with O3 exposure,
9 Additionally, in a group of 86 asthmatic children in Detroit, MI, ambient O3 exposure was associated
10 with decreases in lung function primarily among CS users and subjects reporting concurrent
11 presence of symptoms related to an upper respiratory infection (URI) but not among asthmatics
12 overall (Lewis et al., 2005, 081079). In the group with a URI, 30 ppb increases in lags 1 and 2 of 8-h
13 max O3 were associated with a 6.1% decrease (95% CI: -10.4, -1.6) and a 5.4% decrease
March 2011
6-29
DRAFT - DO NOT CITE OR QUOTE
-------
1 (95% CI: -11.3, 0.10) in lowest daily FEVi, respectively. Ozone was associated with larger decreases
2 in FEVi in two-pollutant models with PMi0 or PM2 5.
3 In human controlled exposure studies, CS treatment of asthmatics generally has not prevented
4 Os-induced FEVi decrements (Section 6.2.1.3). In contrast, among epidemiologic studies, use of
5 inhaled CS has shown both protective (Delfino et al., 2002, 093740: Mortimer et al, 2000, 013255)
6 and exacerbating (Gent et al., 2003, 052885) effects on respiratory symptoms. Among recent studies,
7 effect modification on lung function responses also is mixed. In Lewis et al. (2005, 081079).
8 analyses of interactions between O3 and CS use indicated stronger associations among CS users than
9 among CS nonusers (quantitative results not reported for CS nonusers). Among the 11 (12.8%) CS
10 users, a 30 ppb increase in lag 2 of 8-h max O3 was associated with an 8.0% decrease (95% CI: -
11 13.5, -2.1) in lowest daily FEVi and a 6.7% increase (95% CI: 0.60, 13.2) in diurnal FEVi
12 variability. Lags 1 and 3-5 avg of 8-h max O3 were estimated to have less impact as were similar lags
13 of 24-h avg O3. Ozone exposures were estimated to produce larger changes in lung function in two-
14 pollutant models with PMi0 or PM2 5. The authors purported CS use to be a proxy for greater asthma
15 severity, based on observations that CS users had higher mean FEVi variability and lower mean
16 daily FEVi. In contrast to Lewis et al. (2005, 081079). Hernandez-Cadena et al. (2009, 594283)
17 observed greater O3-related decrements in post-albuterol FEVi among the 60 CS nonusers than
18 among the 25 CS users. In two winter-only studies, consideration of CS use did not largely influence
19 associations between ambient O3 and lung function parameters (Liu et al., 2009, 192003:
20 Rabinovitch et al., 2004, 096753).
21 Although studies have varied in populations and season examined, recent evidence suggests
22 that the inconsistency in effect modification by CS use may, at least in part, be explained by
23 differences in severity of asthmatics included and definition of CS use. In Hernandez-Cadena et al.
24 (2009, 594283). the group of CS nonusers included both intermittent and persistent asthmatics. In
25 Lewis et al. (2005, 081079). most moderate to severe asthmatics (91%) were included in the group
26 of CS users (use for at least 50% of study days). Liu et al. (2009, 192003) did not provide
27 information on asthma severity; however, they defined CS use more stringently as daily use.
28 Differences in asthma severity and definition of CS use may explain why both CS use and nonuse
29 could serve as indicators of severe or uncontrolled asthma. Additionally, investigators did not assess
30 adherence to reported CS regimen, and misclassification of CS use may bias findings.
31 O3 is a powerful oxidant, and antioxidant capacity may influence susceptibility to ambient O3
32 exposure (Sections 5.1.2 and 6.2.1.4). Human controlled exposure studies have demonstrated
33 protective effects of a-tocopherol (vitamin E) and ascorbate (vitamin C) on O3-induced lung function
34 decrements (Section 6.2.1.4), and epidemiologic studies of asthmatic children conducted in
35 Mexico City have had similar findings. In an antioxidant supplementation trial, among moderate to
36 severe asthmatic children, ambient O3 exposure was associated with a greater decrease in FEVi in
37 the placebo group than in the supplementation group (Romieu et al., 2002, 034711) (Figure 6-7 and
38 Table 6-7). Romieu et al. (2009, 548788) observed positive interactions between O3 and diets higher
39 in fruits and vegetables index (FVI) and Mediterranean pattern index (MDI). The FVI and MDI were
March 2011 6-30 DRAFT - DO NOT CITE OR QUOTE
-------
1 each constructed as a 4-level variable to represent increasing consumption of vitamins C and E, and
2 the MDI additionally represented the intake of omega-3 fatty acids, which also have anti-
3 inflammatory effects. At lag 0-4 avg O3 concentrations > 38 ppb, FVI was associated with an
4 increase in FEVi (137 mL per unit increase in FVI [95% CI: 8, 266]). This protective effect of FVI
5 was diminished at O3 concentrations < 25 ppb (65 mL increase in FEVi per unit increase in FVI
6 [95% CI: -70, 200]). Similar results were obtained for MDI.
7 Antioxidant capacity also can be characterized by the activity of xenobiotic metabolizing
8 enzymes. Ambient O3 exposure has been associated with greater decreases in lung function among
9 asthmatic GSTM1 null children, especially among those not supplemented with antioxidant vitamins
10 (Romieu et al., 2004, 056796). Human controlled exposure studies have also indicated greater
11 susceptibility of GSTM1 null subjects, but primarily in conjunction with the NQO1 wild type
12 genotype (Section 6.2.1.4). Effect modification by the GSTP1 variant is unclear. Romieu et al.
13 (2006, 090969) observed that asthmatic children with GSTP1 lie/lie or Ile/Val (associated with
14 greater oxidative metabolism activity) had larger O3-associated decreases in FEVi (Figure 6-7 and
15 Table 6-7). Also unexpectedly, O3 exposure was associated with an increase in FEVi among
16 asthmatics the GSTP1 Val/Val variant, which is associated with reduced antioxidant capacity.
17 Studies of asthmatic children restricted to winter months provided little evidence of an
18 association between ambient O3 exposure and changes in lung function as studies reported both
19 positive and negative associations among various lags of O3 exposure and lung function parameters
20 (Dales et al., 2009, 594285: Liu et al., 2009, 192003: Rabinovitch et al., 2004, 096753V In colder
21 months when children remain primarily indoors, O3, which has low penetration indoors and lack of
22 indoor sources, may have weaker effects. As noted in previous AQCDs (U.S. EPA, 1996, 017831:
23 U.S. EPA, 2006, 088089) and for endpoints such as respiratory hospital admissions, ED visits, and
24 mortality, associations with O3 are generally greater in the warm season.
Asthmatic Adults
25 Relative to studies in asthmatic children, studies of asthmatic adults were limited in number
26 and did not provide strong evidence of acute changes in lung function in association with ambient O3
27 exposure. Characteristics and ambient O3 concentration data from these studies are presented in
28 Table 6-8. One exception was the recent study of 16- to 27-year-old lifeguards in Galveston, TX, that
29 found larger O3-associated decrements in FEVi/FVC among the 16 asthmatic lifeguards (-1.6%
30 [95% CI: -2.8, -0.4] per 40 ppb increase in 1-h max O3) than among the 126 nonasthmatic lifeguards
31 (-0.40% [95% CI: -0.80, 0] per 40 ppb increase in 1-h max O3) (Brooks, personal communication,
32 2010, 644155). In one of the few studies that conducted personal monitoring, neither personal O3
33 exposure nor stationary site O3 concentrations was associated with PEF in a group of asthmatic
34 children and adults (Delfino et al., 1997, 084531). Khatri et al. (2009, 594282) aimed to estimate
35 personal O3 exposures of 38 asthmatic and 13 healthy nonsmoking adults in Atlanta, GA using
36 central site measurements plus time-activity data. They found atopy to be a stronger susceptibility
37 factor than asthma (Khatri et al., 2009, 594282). Investigators reported a larger decrease in percent
March 2011 6-31 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
predicted FEVi/FVC per 30 ppb increase in lag 2 of 8-h max O3 among the 38 atopic (asthmatic or
healthy) subjects (-12 mL [95% CI: -3, -21]) than among asthmatic subjects (-4.7 mL [95% CI: -11,
2.3]). Additionally, among asthmatics, O3 was associated with an increase in FEVi. Based on
correlations observed between decreases in lung function and decreases in quality of life scores,
investigators inferred the O3-associated decreases in lung function to be clinically significant. They
further suggested that atopy may influence responses to ambient O3 exposure because during the
summer, high ambient O3 concentrations may increase allergenicity of pollens.
O3 was not found to have a strong effect on the lung function of asthmatic adults in panel
studies conducted in Europe and Asia during low ambient O3 periods. In a group of 11 subjects in
Rome, Italy followed for 1 month each in spring and winter, lag 0 O3 was associated with a decrease
in percent predicted FEVi and FVC; however, associations with lags 0-1 and 0-2 avg O3 were mostly
positive (Lagorio et al, 2006, 089800). The authors attributed the lack of negative association for O3
to the stable clinical condition of asthmatics. However, this argument was weakened by observations
that NO2 was consistently associated with larger decreases in FEVi and FVC. Park et al. (2005,
088673) followed asthmatics 16-75 years in age in Incheon, Korea, during a period of dust storms
when PMio concentrations fluctuated widely but O3 concentrations remained relatively steady.
Whereas PMi0 was associated with decreases in PEF, O3 was associated with increases in daily
average PEF (2.2 L/min [95% CI: -1.0, 5.5] per 30 ppb increase in 8-h max O3), suggesting that
during dust storms, PMi0 effects may dominate, especially because personal O3 exposures are
expected to be low as a result of limited outdoor activity.
Table 6-8. Mean and upper percentile concentrations of ozone in epidemiologic studies examining
lung function in asthmatic adults
Study
Khatri et al. (2009,
5942821
Thaller etal. (2008,
1958691
Delfino etal. (1997,
0845311
Lagorio et al. (2006,
0898001
Park et al. (2005,
0886731
Location
Atlanta, GA
Galveston, TX
Alpine, CA
Rome, Italy
Incheon,
Korea
Years/Season
2003, 2005, 2006
Warm season
2002-2004
Warm season
1994
Warm season
1999
Spring and winter
March-June 2002
Metric
8-h max
1-h max
1 2-h avg
personal
(8:00-20:00)
24-h avg
24-h avg
Mean
Concentration (ppb)
59a
NR
18
Spring: 36.2°
Winter: 8.0°
Dust event days: 23.6
Control days: 25.1
Middle/Upper Percentile
Concentrations (ppb)
Range: 44-73
Median: 35
Range: 19-118
90th percentile: 52
Range: 0-80
IQR: 8.6 (Spring), 5.1 (Winter)"
Overall range: 3.4-48.6°
NR
21
22
23
NR = Not reported
'Personal exposure estimates were derived based on time spent in the vicinity of various 03 monitors.
"Concentrations converted from pg/m3to ppb using the conversion factor of 0.51 assuming standard temperature (25°C) and pressure (1 atm).
Populations Not Restricted To Asthmatics
Studies have examined associations between ambient O3 exposure and lung function
decrements in the general population and in other potentially populations such as children and older
adults. Limited data are available in populations restricted to healthy populations. Characteristics and
March 2011
6-32
DRAFT - DO NOT CITE OR QUOTE
-------
1 ambient O3 concentration data from these studies are presented in Table 6-9. The 2006 O3 AQCD
2 identified children as a potentially susceptible group based on consistent evidence of association
3 between ambient O3 exposure and decrements in FEVi and PEF (U.S. EPA, 2006, 088089)
4 (Figure 6-8 and Table 6-10). Whereas most of these studies did not distinguish between effects in
5 healthy and asthmatic children, Hoppe et al. (2003, 055618) found larger effects in asthmatic
6 children. In contrast, Avol et al. (1998, 086365) found that healthy children, children with asthma,
7 and children with wheeze had similar FEVi responses to ambient O3 exposure. A recent study of 56
8 healthy children in Vienna, Austria did not find an association between O3 and decrements in total
9 lung capacity; however, this study was restricted to the cold season (Neuberger et al., 2004, 093249)
Table 6-9. Mean and upper percentile concentrations of ozone in epidemiologic studies examining
lung function in populations not restricted to asthmatic subjects
Study
Alexeef et al.
(2007, 1958621
Alexeef et al.
(2008, 1958641
Naeheretal.
(1999.0335681
Avol etal. (1998,
0863651
Linn etal. (1996,
0825081
Gold etal. (1999,
0869191
Scarlett et al.
(1996,0811581
Ward et al. (2002,
0258391
Ulmeretal. (1997,
0836251
Hoppe etal.
(2003, 0556181
Steinvil et al.
(2009, 5487801
Chen etal. (1999,
0111491
Son etal. (2010,
6466551
Location
Greater Boston, MA
Vinton, VA
6 southern CA
communities
Rubidoux, Upland,
Torrence, CA
Mexico City, Mexico
Surrey, England
Birmingham and
Sandwell, England
Freudenstadt and
Villingen, Germany
Munich, Germany
Tel Aviv, Israel
3 Taiwan communities
Ulsan, Korea
Years/Season
1995-2005
All-year
1995-1996
Warm season
Spring and summer
1992-1993, 1993-
1994
Fall and spring
1991
Winter, spring, fall
1994
Warm season
1997
Winter and summer
1994
March -October
1992-1995
Warm season
2002-2007
All-year
1995-1996
May-January
2003-2007
All-year
Metric
24-h avg
8-h max
24-h avg
personal
24-h avg
24-h avg
8-h max
24-h avg
1/2-hmax
1/2-hmax
8-h avg
(10:00 a.m. -
6:00 p.m.)
1-h max
8-h max
Mean
Concentration (ppb)
24.4a
53.7
NR
34a
52.0
50.7
Winter median: 13.0
Summer median: 22.0
Freudenstadt median: 50.6
Villingen medina: 32.1
High days: 65.9
Control days: 27.2
41.1
NR
35.86
Middle/Upper Percentile
Concentrations (ppb)
NR
Range: 17.0-87.6
Approximate range: 5-1 60b
Range: 7-86a
IQR: 25
Range: 7.9-103
Range: 6.8-128
Winter range: 2-33
Summer range: 10-41
Freudenstadt 5th-95th: 22.5-89.7
Villingen 5th-95th: 0.5-70.1
Max (high days): 86
Max (control days): 39
IQR: 34.7-48.7
Range: 6.5-72.8
Range: 19.7-110.3
Median: 36.30
Range: 9.80-59.53
NR = Not Reported, IQR = interquartile range.
'Measured at sites established by investigators.
'Quantitative results not presented. Concentrations estimated from data presented in a figure.
March 2011
6-33
DRAFT - DO NOT CITE OR QUOTE
-------
Study
Linnetal. (1996)
Avoletal.(1998)a
Gold etal. (1999)
Scarlett etal. (1996)
Ward et al. (2002)
Hoppe etal. (2003)
Ulmeretal.(1997)a
Chen etal. (1999)
Alexeetf etal. (2008)
Alexeef etal. (2007)
Naeher etal. (1999)
Steinvil etal. (2009)
Son etal. (2010)
Population
Children
Children
Children
Children
Healthy children
Children
Children
Children
Adults
Adults
Healthy women
Adult worn en
Children and adults
Parameter
Intraday change FEV1 (ml/10)
Intraday change FEV1 (ml/10)
Morning APEF (L/min)b
FEV1 (ml/10)
PEF (L/min)
FEV1 (% change)
FVC (% change)
FEV1 (ml/10)
FEV1 (ml/10)
% predicted FEV1
% predicted FEV1
PEF (L/min)
FEV1/FVC (ml/10)
FEV1 (% predicted)
Lag Subgroup
1 -1
0-1 avg GSTP1 He/He • —
GSTP1 HeA/al ValA/al •
Nonobese •
Obese •—
NoAHR •
0-2 avg •
-14 -11 -8 -5 -2 1
Effect Estimate (95% Cl)
"The 95% CI was constructed using a standard error that was estimated from the p-value.
bAPEF refers to the daily deviation from the mean PEF across study days.
Figure 6-8. Changes in lung function parameters in association with ambient ozone exposures
in studies not restricted to asthmatic populations. IAHR = airway
hyperresponsiveness. Effect estimates are standardized to a 40-, 30-, and 20-ppb
increase fora 1-h (or 1/2-h) max, 8-h max, and 24-h avg ozone exposures,
respectively. All effect estimates are from single pollutant models.
March 2011
6-34
DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-10. Additional characteristics and quantitative data for studies represented in Figure 6-8
Study
Linn etal. (1996,
0825081
Avol etal. (1998,
0863651
Gold etal. (1999,
086919)
Scarlett etal. (1996,
081158)
Ward et al. (2002,
025839)
Hoppe et al. (2003,
055618)
Ulmeretal. (1997,
083625)
Chen etal. (1999,
011149)
Alexeeff etal. (2008,
195864)
Alexeef etal. (2007,
195862)
Naeher etal. (1999,
033568)
Steinvil et al. (2009,
548780)
Son etal. (2010,
646655)
Location/
Population
3 southern CA
communities
Children
3 southern CA
communities
Children
Mexico City, Mexico
Children
Surrey, England
Children
Birmingham and
Sandwell, England
Healthy children
Munich, Germany
Children
Freudenstadt and
Villingen, Germany
Children
3 Taiwan communities
Children
Greater Boston, MA
Adults
Greater Boston, MA
Adults
Vinton, VA
Healthy women
Tel Aviv, Israel
Healthy adults
Ulsan, Korea
Children and adults
O3Lag
0
0 (personal)
1
MOavq
1
0
0-6 avg
0
1
1
0-1 avg
0-1 avg
0
0-4 avg
7
0-2 avg
03
Averaging
Time
1 -h avg
24-h avg
24-h avg
8-h max
24-h avg
1/2-hmax
1/2-hmax
1-h max
24-h avg
24-h avg
8-h max
8-h avg
(10:00-18:00)
8-h max
Parameter Subgroup
I ntraday change FEVi
(mL/10)
I ntraday change FEVi
(mL/10)
Morning APEF (L/min)c
FEV, (mL/10)
PEF (L/min)
FEV, (percent change)
FVC (percent change)
FEV, (mL/10)
FEV, (mL/10)
FEV, (% oredicted) GSTP1 lie/lie
i-nv, (/o predicted) GSTR1 ||e/Vg| Vg|/Vg|
Nonobese
FEV, (% predicted) °QbeAsHeR
AHR
PEF (L/min)
FEV,/FVC (mL/10)
FEV, (% predicted)
Effect
Estimate
(95% Cl)a
-1.1 6 (-2.06, -0.26)
-4.08 (-10.7, 2.6)b
-1.80 (-3.76, 0.16)
-13.0 (-22.8, -3.2)
-0.08 (-0.57, 0.41)
-3. 16 (-8.31, 2.0)
-11.1 (-22.0, -0.18)
-1.4 (-4.3, 1.4)
-2.5 (-4.9, -0.1)
-5.9 (-10.4, 1.3)b
-2.56 (-4.91, -0.21)
-1.0 (-2.2, 0.19)
-2.3 (-3.5, -1.0)
-1.5 -2.5, -0.52)
-2.6 -5.1, -1.9)
-1.7 (-2.7, -0.73)
-4.1 (-6.3, -1.8)
-2.5 (-5.2, 0.04)
-5.1 (-8.7, -1.5)
-9.6 (-19.6, -0.64)
-1.36 (-2.65, -0.08)
AHR = airway hyperresponsiveness.
'Effect estimates are standardized to a 40, 30, and 20 ppb increase for 1-h (or 1/2-h) max, 8-h max, and 24-h avg 03, respectively.
The 95% Cl was constructed using a standard error that was estimated from the p-value.
CAPEF refers to the daily deviation from the mean PEF across study days.
1 A cross-sectional study was conducted of 2,102 children and adults living near a
2 petrochemical plant in Ulsan, Korea (Son et al., 2010, 646655). The mean percent predicted FEVi
3 was 82.85%, indicating a large proportion of subjects with diminished lung function. Multiple O3
4 exposure metrics, including concentrations averaged across 13 city monitors, concentrations from
5 the nearest monitor, inverse distance-weighted concentrations, and estimates from kriging, were
6 associated with decrements in lung function. Among single-day lags (0-2) and 2- (lag 0-1 or 1-2) and
7 3-day (lag 0-2) avg of 8-h max O3 exposure (kriged), lag 0-2 avg was associated with the largest
8 decrements in percent predicted FEVi (-1.36 [95% CI: -2.65, -0.08] per 30 ppb increase in 8-h max
9 O3) and FVC (-16.8 [95% CI: -20.0, -13.6]). Ozone effect estimates showed small changes in
10 magnitude in two-pollutant models with PMi0, NO2, SO2, or CO. An important limitation of this
11 study was the lack of adjustment for meteorological factors.
12 Robust findings from human controlled exposure studies demonstrate O3-induced spirometric
13 responses in children and young adults but diminished responses in older adults (Section 6.2.1.4).
14 While epidemiologic investigation of adults has been limited, studies find associations in healthy
15 adults and older adults. Naeher et al. (1999, 033568) observed associations between ambient O3
16 exposure and decreases in PEF among healthy women, ages 19-43 years. In a large cross-sectional
March 2011
6-35
DRAFT - DO NOT CITE OR QUOTE
-------
1 study of 2,380 healthy adults in Tel Aviv, Israel, lag 7 of 8-h avg (10:00 a.m. to 6:00 p.m.) O3 was
2 associated with a decrease in FEVi/FVC (-96 mL [95% CI: -196, -6.4] per 30 ppb increase in O3);
3 however, overall, the study provided weak evidence for an effect of O3, as associations of other lags
4 of exposure (single day lags 0-7 and 0-6 avg) with FEVi, FVC, and FEVi/FVC were positive
5 (Steinvil et al, 2009, 548780). Whereas Hoppe et al. (2003, 055618) did not find ambient O3
6 exposure-associated decreases in lung function among elderly subjects, findings from the Normative
7 Aging Study demonstrated that ambient O3 exposure was associated with decrements in FEVi and
8 FVC among 900 mostly white, healthy men (mean [SD] age = 68.9 [7.2] years) from the Greater
9 Boston, MA area (Alexeeff et al., 2008, 195864). This study in the Greater Boston area conducted
10 spirometry once every 3 years for 10 years in, a large proportion of whom were middle-aged or
11 elderly. Among all subjects, a 20 ppb increase in lag 0-1 avg of 24-h avg O3 was associated with a
12 1.7% decrease (95% CI: -2.6, -0.72) in FEVi, which was the largest decrement observed among all
13 lags of O3 exposure (1- to 7-day avg) examined (Alexeeff et al., 2008, 195864). Additionally,
14 consistent with findings from human controlled exposure studies (Section 6.2.1.4), larger effects
15 were estimated in specific groups, namely, obese adults, adults with AHR, and adults with the
16 GSTP1 Ile/Val or Val/Val variant (Alexeeff et al., 2007, 195862: Alexeeff et al., 2008, 195864)
17 (Figure 6-8 and Table 6-10). Larger O3-related decrements in FEVi and FVC were also observed in
18 subjects with long GT dinucleotide repeats in the promoter region of the antioxidant enzyme heme
19 oxygenase-1 (Alexeeff et al., 2008, 195864). which has been associated with reduced inducibility
20 (Hiltermann et al., 1998, 086158). The largest O3-related percent decreases in lung function were
21 observed in the group of obese subjects with AHR (-5.3% FEVi [95% CI: -8.3, -2.4] per 20 ppb
22 increase in lag 0-1 avg of 24-h avg O3).
Lag Structure in Ambient Ozone Exposure-associated Lung Function Decrements
23 Controlled human exposure studies demonstrate decreases in lung function within 2 hours to 2
24 days of O3 exposure, depending on the exposure regimen, with an attenuation of effect after 3-5 days
25 after a daily exposure regimen. Consistent with these findings, studies of subjects engaged in
26 outdoor recreation, exercise, or work indicate decreases in lung function in association with O3
27 exposures over the duration of activity. Among the few studies of subjects with increased outdoor
28 exposures that examined other lags of O3 exposure, some found no persistence of effects (Hoppe et
29 al., 2003, 055618: Spektor et al., 1991, 042383). whereas others found that the effects of O3
30 exposure carried over to the next day (Brauer et al., 1996, 080754: Spektor et al., 1988, 041710).
31 Collectively, epidemiologic studies in other populations have examined associations with
32 single-day O3 concentrations lagged from 0 to 7 days as well concentrations averaged over
33 2-10 days. Some studies have found decreases in lung function associated with same-day or
34 previous-day O3 exposures (Alexeeff et al., 2008, 195864: Chen et al., 1999, 011149: Jalaludin et al.,
35 2000, 011929: Lewis et al., 2005, 081079: Romieu et al., 1996, 080748: Romieu et al., 1997,
36 085807: Ross et al., 2002, 042749: Son et al., 2010, 646655). Relatively fewer epidemiologic studies
March 2011 6-36 DRAFT - DO NOT CITE OR QUOTE
-------
1 have found associations with longer lags of ambient O3 exposures (5-7 days) (Hernandez-Cadena et
2 al, 2009, 594283: Steinvil et al, 2009, 548780: Wiwatanadate and Trakultivakorn, 2010, 387706).
3 Additionally, in many studies, multiday averages of O3 exposure (2-10 days) were associated
4 with decreases in lung function (Alexeeff et al., 2007, 195862: Barraza-Villarreal et al., 2008,
5 156254: Gold et al., 1999, 086919: Liu et al., 2009, 192003: Mortimer et al., 2002, 030281: Naeher
6 et al., 1999, 033568: O'Connor et al., 2008, 156818: Son et al., 2010, 646655: Ward et al., 2002,
7 025839). indicating that exposures accumulated over several days may be important or may be
8 subject to less measurement error. Collectively, among studies that examined a range of single-day
9 lags and multiday averages, evidence did not overwhelmingly point to stronger immediate, delayed,
10 or cumulative effects of O3 exposure on lung function. Some studies indicated stronger effects of
11 multiday O3 exposures (Gold et al., 1999, 086919: Mortimer et al., 2002, 030281: Naeher et al.,
12 1999, 033568: Ward et al., 2002, 025839). whereas many others did not find a consistent trend
13 (Alexeeff et al., 2008, 195864: Lagorio et al., 2006, 089800: Lewis et al., 2005, 081079: Liu et al.,
14 2009, 192003: Son et al., 2010, 646655: Steinvil et al., 2009, 548780: Wiwatanadate and
15 Trakultivakorn, 2010, 387706).
Summary of Epidemiologic Studies of Lung Function
16 The cumulative body of epidemiologic evidence strongly supports associations between
17 ambient O3 exposure and decrements in lung function among children, in particular, those with
18 asthma and those with increased outdoor exposures. Consistent with findings from human controlled
19 exposure studies, epidemiologic evidence demonstrates ambient O3-associated decrements in lung
20 function in adults exercising or working outdoors. Although recent epidemiologic studies contributed
21 mixed results, most studies of asthmatic children indicated negative associations between O3
22 exposure and decrements in lung function (Figures 6-6 and 6-7 and Tables 6-6 and 6-7). Whereas
23 previous evidence was weak, new evidence indicates that O3 exposure may be associated with
24 decrements in lung function in older adults. Effect modification by any individual susceptibility
25 factor was examined only in one to two studies; however, O3-associated lung function decrements
26 were increased with obesity, reduced activity of antioxidant enzymes, AHR, or concurrent URL High
27 dietary antioxidant intake was found to decrease susceptibility to O3-associated decreases in lung
28 function. A small proportion of studies of lung function have evaluated confounding by
29 co-pollutants; however, in studies that provided quantitative (Figure 6-9 and Table 6-11) or graphical
30 results (Dales et al., 2009, 594285: Liu et al., 2009, 192003) of co-pollutant modeling, most O3
31 effect estimates did not change considerably in magnitude when adjusted for PM2 5, PMi0, NO2, or
32 SO2. Ambient O3 exposures accumulated over several days are associated with decreases in lung
33 function; however, in epidemiologic studies, there is uncertainty around the relative effects of
34 immediate, delayed, or cumulative O3 exposures.
March 2011 6-37 DRAFT - DO NOT CITE OR QUOTE
-------
Study Parameter
O'Connor et al. (2008) % predicted FEV1
Subgroup
Lewis et al. (2005)
% predicted PEF
% change, lowest daily
FEV1
Thaller et al. (2008) % change, FEV1/FVC
Romieu et al. (1996) Evening PEF (L/min)
Romieu et al. (1997) Evening PEF (L/min)
Goldetal. (1999)
Morning APEF (%)a
Romieu et al. (2002) FEV1 (ml/10)
Jalaludin et al. (2000) APEF (L/min)a
Chenetal. (1999)
Sonetal. (2010)
Wiwatanadate and
Trakultivakorn (2010)
FEV1 (ml/10)
% predicted FEV1
Daily avg PEF (L/min)
CS users
CS nonusers
With URI
Without URI
Nonasthmatics
Asthmatics
Placebo
Vitamin C/E
-10 -9 -8 -7 -6 -5 -4-3-2-101 23
Effect Estimate (95% Cl)
aAPEF refers to the daily deviation from the mean PEF across study days.
Figure 6-9. Comparison of ozone-lung function effect estimates in single- and co-pollutant
models. [IDS = corticosteroid, URI = Upper respiratory infection. Effect estimates are
standardized to a 40-, 30-, and 20-ppb increase for 1-h max, 8-h max, and 24-h avg
ozone, respectively. Effect estimates depicted as black circles are from single
pollutant models, and effect estimates depicted as open circles are from
co-pollutant models.
March 2011
6-38
DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-11. Additional characteristics and quantitative data for studies presented in Figure 6-9
Study
O'Connor etal.
(2008, 1568181
Lewis et al. (2005,
0810791
Thaller etal. (2008,
1958691
Romieuetal. (1996,
0807481
Romieuetal. (1997,
0858071
Gold etal. (1999,
0869191
Romieu et al. (2002,
0347111
Jalaudin etal.
(2000, 0119291
Chen etal. (1999,
0111491
Son etal. (2010,
6466551
Wiwatanadate and
Trakultivakorn
(2010,3877061
Location/ O3
Population Lag
7 U.S.
communities 1-5
Asthmatic avg
children
Detroit, Ml
Asthmatic 2
children
Galveston, TX
Outdoor 0
workers
Mexico City,
Mexico g
Asthmatic
children
Mexico City,
Mexico n
Asthmatic
children
Mexico City, , ,„
Mexico "
Children 9
Mexico City,
Mexico /
Asthmatic
children
Sydney,
Australia n
Children with u
wheeze
3 Taiwan ,
communities
Incheon, Korea n -
Children and " ,n
adults av9
Chiang Mai,
Thailand 5
Asthmatic
children
03
Averaging
Time
«„
8-h max
1-h max
1-h max
1-h max
24-h avg
1-h max
24-h avg
1-h max
8-h max
24-h avg
Parameter
% predicted
FEV,
percent change,
lowest daily
FEV,
percent change,
FEV,/FVC
Evening PEF
(L/min)
Evening PEF
(L/min)
Morning APEF
FEV,(mL/10)
APEF"
FEV,(mL/10)
% predicted
FEV,
Daily avg PEF
(L/min)
O3 Effect
_...„,,„..„ Estimate in
Subgroup sjng|e.po||utant
Model (95% Cl)a
-0.41 (-1.03, 0.21)
CSuser -8.0 (-13.5, -2.1)
WithURI -5.4 (-11. 3, 1.0)
Nonasthmatic -0.4 (-0.8, 0)
Asthmatic -1.6 (-2.8, -0.4)
-1.50 (-3.60, 0.53)
-1.45 (-2.88, -0.02)
-3.1 (-4.7, -1.4)
Placebo 1.88 (-3.43, -0.34)
Vitamin C/E -0.07 (-1.52, 1.37)
-1.84 (-3.48, -0.19)
-2.56 (-4.91, -0.21)
-1.36 (-2.65, -0.08)
-2.6 (-5.2, 0)
O3 Effect Estimate
in Co-pollutant
Model (95% Cl)a
-0.54 (-1.27, 0.19) with
PM2.5, N02
-0.3 (-16.0, 18.0) with
PM2.5
-8.1 (-15.0, -0.60)
-0.6 (-1.2,0) with PM25,
N02
-1.7 (-3.1, -0.1)
-0.66 (-3. 16, 1.85) with
PM2.5
-2.20 (-3.96, -0.44) with
PM,0
-5.7 (-8.9, -2.0) with
PM2.5C
-1.84 (-3.55, -0.13) with
PM,0, N02
-0.02 (-1.73, 1.69)
-1.76 (-3.42, -0.11) with
PM,0, N02
-3.40 (-6.07, -0.73) with
N02
-1.83 (-3.4, -0.25) with
PM,0
-3.2 (-6.2, -0.2) with S02
'Effect estimates were standardized to a 40-, 30-, and 20-ppb increase for 1-h max, 8-h max, and 24-h avg 03, respectively.
bAPEF refers to the daily deviation from the mean PEF across sampling days.
"Effect estimate is based on a multi-pollutant model of the joint effects of an increase in 03and PM2.5.
6.2.1.3. Toxicology
1 The 2006 O3 AQCD found that pulmonary function decrements occur in a number of species
2 with acute exposures^ 1 week), ranging from 0.25 to 0.4 ppm O3 (U.S. EPA, 2006, 088089).
3 Information published more recently adds to the evidence of ventilation defects induced by acute or
4 subchronic exposure. Rats exposed to 0.5 ppm O3 for 2 or 6 days, either continuously or
5 alternatingly, were analyzed by magnetic resonance imaging (MRI). Although the lung capacity of
6 the animals was unaffected by O3 exposure, ventilation defects were evident, based on delayed and
7 incomplete or heterogeneous lung filling. This effect increased with the duration of exposure, and
8 was more prevalent and severe in animals exposed alternatingly (12 h/day) as opposed to
9 continuously (22 h/day). Among rats exposed over six days for 12 h/day, 85% exhibited ventilation
March 2011
6-39
DRAFT - DO NOT CITE OR QUOTE
-------
1 defects. The authors suggest that the delayed filling of lung lobes or segments is likely a result of an
2 increase in airway resistance brought about by narrowing of the peripheral small airways
3 (Cremillieux et al, 2008, 180454). Lung resistance and elastance were unaffected in allergen
4 sensitized mice exposed solely to 0.5 ppm O3 once a week for 4 weeks (Farraj et al., 2010, 380846).
5 However, co-exposure to O3 and diesel exhaust particles increased lung resistance. Long-term
6 exposure to O3 during development may ultimately affect pulmonary function by altering lung
7 morphology (see Chapter 7).
6.2.2. Airway Hyperresponsiveness
8 Airway hyperresponsiveness refers to a condition in which the conducting airways undergo
9 enhanced bronchoconstriction in response to a variety of stimuli. Airway responsiveness is typically
10 quantified by measuring changes in pulmonary function (e.g., FEVi or specific airway resistance)
11 following the inhalation of an aerosolized specific (allergen) or nonspecific (e.g., methacholine)
12 bronchoconstricting agent or another stimulus such as exercise or cold air. Asthmatics are generally
13 more sensitive to bronchoconstricting agents than nonasthmatics, and the use of an airway challenge
14 to inhaled bronchoconstricting agents is a diagnostic test in asthma. Standards for airway
15 responsiveness testing have been developed for the clinical laboratory (American Thoracic Society.,
16 2000, 090799). although variation in methodology for administering the bronchoconstricting agent
17 may affect the results (Cockcroft et al., 2005, 090805). There is a wide range of airway
18 responsiveness in nonasthmatic people, and responsiveness is influenced by wide range of factors,
19 including cigarette smoke, pollutants, respiratory infections, occupational exposures, and respiratory
20 irritants. Since the 2006 O3 AQCD, no epidemiology studies have examined airway responsiveness
21 as a biological endpoint.
6.2.2.1. Controlled Human Exposures
22 Beyond its direct effect on lung function, O3 exposure causes an increase in airway
23 responsiveness in human subjects as indicated by a reduction in the concentration of methacholine
24 required to produce a given reduction in FEVi or increase in sRaw. Increased airway responsiveness
25 is an important consequence of exposure to ambient O3, because the airways are then predisposed to
26 narrowing upon inhalation of a variety of ambient stimuli including specific allergens, SO2, and cold
27 air.
28 O3 exposure of asthmatic subjects, who characteristically have increased airway
29 responsiveness at baseline, can cause further increases in responsiveness (Kreit et al., 1989, 041817).
30 Similar relative changes in airway responsiveness are seen in asthmatics and health controls exposed
31 to O3 despite their markedly different baseline airway responsiveness. Several studies (Torres et al.,
32 1996, 078122: Kehrl et al., 1999, 022101: Molfino et al., 1991, 042379) have been published
33 suggesting an increase in specific (i.e., allergen-induced) airway reactivity. An important aspect of
March 2011 6-40 DRAFT - DO NOT CITE OR QUOTE
-------
1 increased airway responsiveness after O3 exposure is that this may represent a plausible link between
2 ambient O3 exposure and increased hospital admissions for asthma.
3 Changes in airway responsiveness after O3 exposure appear to resolve more slowly than
4 changes in FEVi or respiratory symptoms (Folinsbee and Hazucha, 2000, 001701). Furthermore, in
5 studies of repeated exposure to O3, changes in airway responsiveness tend to be somewhat less
6 susceptible to attenuation with consecutive exposures than changes in FEVi (Dimeo et al., 1981,
7 039662: Folinsbee et al., 1994, 044189: Gong et al., 1997, 082696: Kulle et al., 1982, 040668V
8 Increases in airway responsiveness do not appear to be strongly associated with decrements in lung
9 function or increases in symptoms (Aris et al., 1995, 075945).
10 Since the 2006 O3 AQCD, no controlled human exposure studies have been conducted to
11 examine the mechanistic aspects of O3-induced airway hyperresponsiveness. The mechanism of
12 O3-induced increases in airway responsiveness is poorly understood, but it appears to be associated
13 with a number of cellular and biochemical changes in airway tissue. Although inflammation could
14 play a role in the increase in airway responsiveness, cyclooxygenase inhibitors have not been
15 effective at blocking the O3-induced influx of PMNs into BALF (Hazucha et al., 1996, 043923: Ying
16 et al., 1990, 042334). Therefore, O3-induced airway responsiveness may not be due to the presence
17 of PMNs in the airway or to the release of arachidonic acid metabolites. Rather, it seems likely that
18 the mechanism for this response is multifactorial as discussed in detail below.
6.2.2.2. Toxicology
19 In addition to studies with human subjects, a number of species, including nonhuman
20 primates, dogs, cats, rabbits, and rodents, have been used to examine the effect of O3 exposure on
21 airway hyperresponsiveness. With a few exceptions, commonly used animal models have been
22 guinea pigs, rats, or mice acutely exposed to high O3 concentrations (1-3 ppm) to induce airway
23 hyperresponsiveness. These high dose models are helpful for determining underlying mechanisms of
24 general airway hyperresponsiveness, but have questionable relevance for extrapolation to potential
25 airway responses in humans exposed to ambient levels of O3.
26 A limited number of studies have observed airway hyperresponsiveness in rodents and guinea
27 pigs after exposure to less than 0.3 ppm O3. As previously reported in the 2006 O3 AQCD, one study
28 demonstrated that a very low concentration of O3 (0.05 ppm) induced airway hyperresponsiveness in
29 certain strains of rats suggesting a genetic component (Depuydt et al., 1999, 011995). More recently,
30 Chhabra and colleagues (2010, 677665) demonstrated that exposure of OVA-sensitized guinea pigs
31 to 0.12 ppm for 2 h/d for 4 weeks produced specific airway hyperresponsiveness to an inhaled OVA
32 challenge. Interestingly, dietary supplementation of the guinea pigs with vitamins C and E
33 ameliorated a portion of the airway hyperresponsiveness as well as indices of inflammation and
34 oxidative stress in this study. Larsen and colleagues did an O3 concentration-response study in mice
35 sensitized by 10 daily inhalation treatments with an OVA aerosol (Larsen et al., 2010, 628560).
36 Although airway responsiveness to methacholine was increased in non-sensitized animals exposed to
37 a single 3-h exposure to 0.5, but not 0.1 or 0.25, ppm O3, airway hyperresponsiveness was observed
March 2011 6-41 DRAFT - DO NOT CITE OR QUOTE
-------
1 after exposure to 0.1 and 0.25 ppm O3 in OVA-sensitized mice. Shore and colleagues (Johnston et
2 al., 2005, 596394) have also demonstrated O3-induced airway hyperresponsiveness in mice after
3 exposure to 0.3 ppm O3 for 3 hours. Adaptation to this effect was observed in mice that were
4 exposed to the same concentration of O3 for 72 hours and showed no evidence of airway
5 hyperresponsiveness. Thus, recent toxicological studies have demonstrated that O3-induced airway
6 hyperresponsiveness occurs in monkeys, guinea pigs, and mice after either acute or repeated
7 exposure to relevant concentrations of O3.
8 The mechanisms by which O3 enhances the airway responsiveness to either specific (e.g.,
9 OVA) or non-specific (e.g., methacholine) bronchoprovocation are not clear and appear to be
10 associated with complex cellular and biochemical changes in the conducting airways. Considerable
11 research effort has been directed towards exploring the causes of O3-induced airway
12 hyperresponsiveness, but the majority of such studies have been conducted at high concentrations of
13 O3. It is clear that inflammation plays a key role in O3-induced airway hyperresponsiveness, although
14 the precise mediators and cells that are involved have not been identified at relevant concentrations
15 of O3. Because inflammation is likely to play a role in O3-induced airway hyperresponsiveness, the
16 mechanism for this response may be multifactorial, involving the presence of cytokines, prostanoids,
17 or neuropeptides; activation of macrophages, eosinophils, or mast cells; and epithelial damage that
18 increases direct access of mediators to the smooth muscle or receptors in the airways that are
19 responsible for reflex bronchoconstriction. Johnston et al. (2005, 596394) demonstrated that airway
20 hyperresponsiveness occurred in both wild type and IL-6 knockout mice exposed to 0.3 ppm O3
21 despite reduction in markers of lung injury and inflammation in O3-exposed IL-6 knockout mice.
22 This same group of investigators has demonstrated the involvement of natural killer T cells, obesity,
23 CXCR2, leptin, and IL-17 in O3-induced airway hyperresponsiveness albeit at exposure
24 concentrations of 1-2 ppm O3 (Garantziotis et al., 2010, 624947; Johnston et al., 2005, 596393; Lu et
25 al., 2006, 597955; Pichavant et al., 2008, 596409; Shore et al., 2003, 057302; Voynow et al., 2009,
26 194311; Williams et al., 2007, 597545) have been proposed for airway hyperresponsiveness induced
27 by single exposures to O3 at 1-3 ppm. Thus, a number of potential mediators and cells may play a
28 role in O3-induced airway hyperresponsiveness, but mechanistic studies are needed at more relevant
29 concentrations of O3.
30 In order to evaluate the ability of O3 to enhance specific and non-specific airway
31 responsiveness, it is important to understand the role of adaptation in ozone's effects. Several studies
32 have clearly demonstrated that some adverse effects caused by acute exposure are absent after
33 repeated exposures to O3. The ability of the pulmonary system to adapt to repeated insults to O3 is
34 complex, however, and experimental findings for adaptation to O3-induced airway
35 hyperresponsiveness are inconsistent. As described above, airway hyperresponsiveness was observed
36 in mice after a 3-h exposure but not in mice exposed continuously for 72 hours to 0.3 ppm (Johnston
37 et al., 2005, 596394). However, the Chhabra study demonstrated O3-induced airway
38 hyperresponsiveness in guinea pigs exposed for 2 h/day for 10 days (Chhabra et al., 2010, 677665).
39 Besides the obvious species disparity, these studies differ in that the mice were exposed continuously
March 2011 6-42 DRAFT - DO NOT CITE OR QUOTE
-------
1 for 72 hours, whereas the guinea pigs were exposed intermittently over 10 days, suggesting that
2 adaptation might be lost with periods of rest in between O3 exposures. This type of reasoning is the
3 basis for the episodic exposure protocol used in the infant rhesus monkey studies (Plopper et al.,
4 2007, 596412).
6.2.3. Pulmonary Inflammation, Injury and Oxidative Stress
6.2.3.1. Controlled Human Exposures
5 In addition to physiological pulmonary responses, respiratory symptoms, and airway
6 hyperresponsiveness, O3 exposure has been shown to result in epithelial permeability and respiratory
7 tract inflammation. As reported in studies reviewed in the 1996 and 2006 O3 AQCDs (U.S. EPA,
8 1996, 017831; U.S. EPA, 2006, 088089). acute O3 exposure initiates an acute inflammatory response
9 throughout the respiratory tract which may persist for at least 18-24 hours postexposure. A meta-
10 analysis of 21 studies (Mudway and Kelly, 2004, 057299) showed that PMN influx in healthy
11 subjects is significantly associated (statistically) with total O3 dose (i.e., the product of O3
12 concentration, exposure duration, and VE).
13 The presence of neutrophils (PMNs) in the lung has long been accepted as a hallmark of
14 inflammation and is an important indicator that O3 causes inflammation in the lungs. Neutrophilic
15 inflammation of tissues indicates activation of the innate immune system and requires a complex
16 series of events which are normally followed by processes that clear the evidence of acute
17 inflammation. Inflammatory effects have been assessed in vivo by lavage (proximal airway and
18 bronchoalveolar), bronchial biopsy, and more recently, induced sputum. A single acute exposure
19 (1-4 hours) of humans to moderate concentrations of O3 (0.2-0.6 ppm) while exercising at moderate
20 to heavy levels results in a number of cellular and biochemical changes in the lung, including an
21 inflammatory response characterized by increased numbers of PMNs, increased permeability of the
22 epithelial lining of the respiratory tract, cell damage, and production of proinflammatory cytokines
23 and prostaglandins (U.S. EPA, 2006, 088089). These changes also occur in humans exposed to 80
24 and 100 ppb O3 for 6-8 hours (Alexis et al., 2010, 628538: Devlin et al., 1991, 040359: Peden et al.,
25 1997, 085842). Soluble mediators of inflammation such as the cytokines (e.g., IL-6, IL-8) and
26 arachidonic acid metabolites (e.g., prostaglandin [PG]E2, PGF2a, thromboxane, and leukotrienes
27 [LTs] such as LTB4) have been measured in the BALF of humans exposed to O3. In addition to their
28 role in inflammation, many of these compounds have bronchoconstrictive properties and may be
29 involved in increased airway responsiveness following O3 exposure. The possible relationship
30 between repetitive bouts of acute inflammation in humans caused by O3 and the development of
31 chronic respiratory disease is unknown.
32 Studies reviewed in the 2006 O3 AQCD (U.S. EPA, 2006, 088089) reported that inflammatory
33 responses do not appear to be correlated with lung function responses in either asthmatic or healthy
34 subjects (Balmes et al., 1996, 080830: Balmes et al., 1997, 086092: Devlin et al., 1991, 040359:
March 2011 6-43 DRAFT - DO NOT CITE OR QUOTE
-------
1 Holz et al, 1999, 058731). However, Vagaggini et al. (2010, 387127) recently reported a significant
2 (r=0.61, p=0.015) correlation between changes in FEVi and changes in sputum neutrophils in mild-
3 to-moderate asthmatics (n=23; 33 ± 11 years) exposed to 300 ppb O3 for 2 hours with moderate
4 exercise. Significant inflammatory responses to O3 exposures that did not elicit significant
5 spirometric responses have also been observed (Holz et al., 2005, 077170; McBride et al., 1994,
6 043912V
7 The time course of the inflammatory response to O3 in humans has not been fully
8 characterized. Different markers exhibit peak responses at different times. Studies in which lavages
9 were performed 1 hour after O3 exposure (1 hour at 0.4 ppm or 4 hours at 0.2 ppm) have
10 demonstrated that the inflammatory responses are quickly initiated (Devlin et al., 1996, 042840;
11 Schelegle et al., 1991, 042491; Torres et al., 1997, 084265). Inflammatory mediators and cytokines
12 such as IL-8, IL-6, and PGE2 are greater at 1 hours than at 18 hours post-O3 exposure (Devlin et al.,
13 1996, 042840; Torres et al., 1997, 084265). However, IL-8 still remain elevated at 18 hours post-O3
14 (4 hours at 0.2 ppm O3 versus FA) in healthy subjects (Balmes et al., 1996, 080830). Schelegle et al.
15 (1991, 042491) found increased PMNs in the "proximal airway" lavage at 1, 6, and 24 hours after O3
16 exposure (4 hours at 0.2 ppm O3), with a peak response at 6 hours. Although, at 18-24 hours after O3
17 exposure, PMNs remain elevated relative to 1 hour postexposure (Schelegle et al., 1991, 042491;
18 Torres et al., 1997, 084265).
19 Alexis et al. (2010, 628538) recently reported that a 6.6-h exposure with moderate exercise to
20 80 ppb O3 caused an increased sputum neutrophil levels at 18 hours postexposure in young healthy
21 adults (n=15; 24 ± 1 years). In a prior study, Alexis et al. (2009, 628542) found genotype effects on
22 inflammatory responses but not lung function responses to a 2 h-exposure to 400 ppb O3. At 4 hours
23 post O3 exposure, both GSTM1 genotypes had significant increases in sputum neutrophils with a
24 tendency for a greater increase in GSTM1-sufficient than null individuals. At 24 hours postexposure,
25 neutrophils had returned to baseline levels in the GSTM1-sufficient individuals. In the GSTMl-null
26 subjects, however, neutrophil levels increased further from 4 hours to 24 hours and were
27 significantly greater than both baseline levels and 24-h levels in GSTM1-sufficient individuals.
28 Alexis et al. (2009, 628542) found that GSTM1-sufficient individuals (n=19; 24 ± 3 years) had a
29 decrease in macrophage levels at 4-24 hours postexposure to 400 ppb O3 for 2 hours with exercise.
30 Effects of the exposure apart from O3 can not be ruled out in the Alexis et al. (2009, 628542; 2010,
31 628538) studies, however, since no FA exposure was conducted.
32 Kim et al. (In Press, 674869) has more recently shown a significant (p < 0.001) increase in
33 sputum neutrophil levels following a 6.6-h exposure to 60 ppb O3 relative to FA in young healthy
34 adults (13 F, 11 M; 25.0 ± 0.5 years). There was no significant effect of GSTM1 genotype (half
35 GSTMl-null) on the inflammatory responses observed in these individuals. Previously,
36 inflammatory responses had only been evaluated down to a level of 80 ppb.
37 Inflammatory responses to O3 exposure have also been studied in asthmatic subjects (Basha et
38 al., 1994, 075950; Peden et al., 1997, 085842; Scannell et al., 1996, 080755). In these studies,
39 asthmatics showed significantly more neutrophils in the BALF (18 hours postexposure) than did
March 2011 6-44 DRAFT - DO NOT CITE OR QUOTE
-------
1 similarly exposed healthy individuals. In one of these studies (Peden et al, 1997, 085842). which
2 included only allergic asthmatics who tested positive for Dematophagoides farinae antigen, there
3 was an eosinophilic inflammation (twofold increase), as well as neutrophilic inflammation (threefold
4 increase). In a study of subjects with intermittent asthma exposed to 0.4 ppm O3 for 2 hours,
5 increases in eosinophil cationic protein, neutrophil elastase and IL-8 were found to be significantly
6 increased 16 hours postexposure and comparable in induced sputum and BALF (Hiltermann et al.,
7 1999, 013196). Scannell et al. (1996, 080755) also reported that IL-8 tends to be higher in the BALF
8 of asthmatics compared to nonasthmatics following O3 exposure, suggesting a possible mediator for
9 the significantly increased neutrophilic inflammation in those subjects. Bosson et al. (2003, 051687)
10 found significantly greater epithelial expression of IL-5, IL-8, granulocyte-macrophage colony-
11 stimulating factor (GM-CSF) and epithelial cell-derived neutrophil-activating peptide 78 (ENA-78)
12 in asthmatics compared to healthy subjects following exposure to 0.2 ppm O3 for 2 hours. In
13 contrast, Stenfors et al. (2002, 030473) did not detect a difference in the O3-induced increases in
14 neutrophil numbers between 15 mild asthmatic and 15 healthy subjects by bronchial wash at the 6
15 hours postexposure time point. However, the asthmatics were on average 5 years older than the
16 healthy subjects in this study, and it is not yet known how age affects inflammatory responses. It is
17 also possible that the time course of neutrophil influx differs between healthy and asthmatic
18 individuals.
19 Vagaggini et al. (2002, 035191) investigated the effect of prior allergen challenge on responses
20 in mild asthmatics exposed for 2 hours to 0.27 ppm O3 or filtered air. At 6 hours postexposure,
21 eosinophil numbers in induced sputum were found to be significantly greater after O3 than after air
22 exposures. Studies such as this suggest that the time course of eosinophil and neutrophil influx
23 following O3 exposure can occur at levels detectable within the airway lumen by as early as 6 hours.
24 They also suggest that the previous or concurrent activation of proinflammatory pathways within the
25 airway epithelium may enhance the inflammatory effects of O3. For example, in an in vitro study of
26 primary human nasal epithelial cells and BEAS-2B cell line, cytokine production induced by
27 rhinovirus infection was enhanced synergistically by concurrent exposure to O3 at 0.2 ppm for 3
28 hours (Spannhake et al., 2002, 030637).
29 Markers from BALF following both 2 hours (Devlin et al., 1997, 083577) and 4 hours
30 (Christian et al., 1998, 029925: Torres et al., 2000, 005654) repeated O3 exposures (up to 5 days)
31 indicate that there is ongoing cellular damage irrespective of the attenuation of some cellular
32 inflammatory responses of the airways, pulmonary function, and symptom responses. Devlin et al.
33 (1997, 083577) found that several indicators of inflammation (e.g., PMN, IL-6, PGE2, fibronectin)
34 were attenuated after 5 days of exposure (i.e., values were not different from FA). However, other
35 markers (LDH, IL-8, total protein, epithelial cells) did not show attenuation, suggesting that tissue
36 damage probably continues to occur during repeated exposure. Christian et al. (1998, 029925)
37 showed decreased numbers of neutrophils and a decrease in IL-6 levels in healthy adults after 4 days
38 of exposure versus the single exposure to 0.2 ppm O3 for 4 hours. Torres et al. (2000, 005654) also
39 found both functional and BALF cellular responses to O3 were abolished at 24 hours postexposure
March 2011 6-45 DRAFT - DO NOT CITE OR QUOTE
-------
1 following the fourth exposure day. However, levels of total protein, IL-6, IL-8, reduced glutathione
2 and ortho-tyrosine were still increased significantly. In addition, visual scores (bronchoscopy) for
3 bronchitis and erythema and the numbers of neutrophils in bronchial mucosal biopsies were
4 increased. Results indicate that, despite an attention of some markers of inflammation in BALF and
5 pulmonary function decrements, inflammation within the airways persists following repeated
6 exposure to O3. The continued presence of cellular injury markers indicates a persistent effect that
7 may not necessarily be recognized due to the attenuation of spirometric and symptom responses.
8 A number of studies show that O3 exposures increases epithelial cell permeability through
9 direct (technetium-99m labeled diethylene triamine pentaacetic acid, 99mTc-DTPA, clearance) and
10 indirect (e.g., increased BALF albumin, protein) techniques. Kehrl et al. (1987, 040824) showed
11 increased 99mTc-DTPA clearance in healthy young adults at 75 minutes postexposure to 0.4 ppm O3
12 for 2 hours. Foster and Stetkiewicz (1996, 079920) have shown that increased 99mTc-DTPA clearance
13 persists for at least 18-20 hours post-O3 exposure (130 minutes to average O3 concentration of
14 0.24 ppm), and the effect is greater at the lung apices than at the base. Increased BALF protein,
15 suggesting O3-induced changes in epithelial permeability, have also been reported at 1 hour and
16 18 hours postexposure (Balmes et al., 1996, 080830: Devlin et al., 1997, 083577). Meta-analysis of
17 results from 21 publications (Mudway and Kelly, 2004, 057299). showed that increased BALF
18 protein is associated with total inhaled O3 dose (i.e., the product of O3 concentration, exposure
19 duration, and VE). Changes in permeability associated with acute inflammation may provide
20 increased access of inhaled antigens, particles, and other inhaled substances deposited on lung
21 surfaces to the smooth muscle, interstitial cells, and the blood.
6.2.3.2. Epidemiology
22 In the 2006 O3 AQCD, epidemiologic evidence of O3-associated changes in biological markers
23 of airway inflammation was limited to observations of increases in upper airway nasal lavage levels
24 of inflammatory cell counts, eosinophilic cationic protein, and myeloperoxidases (U.S. EPA, 2006,
25 088089). As a consequence of advances in less invasive methods to collect biological samples
26 repeatedly from subjects in the field, the number of recent studies assessing ambient O3-related
27 changes in lower airway inflammation and oxidative stress has increased dramatically. Although
28 most biomarkers were not specific to the lung, most studies collected exhaled breath, exhaled breath
29 condensate (EEC), nasal lavage fluid, or induced sputum with the aim of monitoring inflammatory
30 responses in airways, as opposed to monitoring systemic responses in blood. These recent studies
31 form a larger base to establish coherence with findings from human experimental and animal
32 toxicological studies that have measured similar endpoints and provide further biological plausibility
33 for associations of ambient O3 with respiratory symptoms and lung function. These endpoints also
34 allow assessment of potential O3-related acute respiratory morbidity in populations that are less
35 likely to experience increases in respiratory symptoms, including healthy populations and groups
36 with increased outdoor exposures.
March 2011 6-46 DRAFT - DO NOT CITE OR QUOTE
-------
1 Despite the strengths of biomarker studies, several limitations are recognized that may limit
2 the interpretations of associations between ambient O3 exposure and changes in biomarker levels.
3 For example, the clinical relevance of the observed magnitudes of changes has not been well
4 characterized (Duramad et al., 2007, 625792: Murugan et al, 2009, 625839). The inadequate
5 understanding of the changes in biomarker levels in relation to other endpoints of respiratory
6 morbidity may provide an explanation for the mixed results observed in studies that evaluate
7 multiple biomarkers in addition to lung function or respiratory symptoms. The lack of standardized
8 methodology for collection, low sensitivity and specificity of many assay methods, and poor
9 characterization of subject factors that contribute to inter-individual variability, including asthma
10 severity and recent medication use, are sources of uncertainty that may contribute to the
11 inconsistency of findings among studies.
12 In recent studies, the biomarker most frequently measured was exhaled nitric oxide (eNO),
13 likely related to its ease of collection in the field and automated measurement. NO acts as a signaling
14 molecule in numerous biological processes; however, studies pointed to observations of inducible
15 nitric oxide synthase activation and NO production by proinflammatory cytokines, macrophages,
16 neutrophils, and epithelial cells in the lung (Barnes and Liew, 1995, 083814) to support analysis of
17 eNO as an indicator of airway inflammation. Further support is provided by observations of higher
18 eNO in asthmatics, especially in those with poorly controlled asthma (Jones et al., 2001, 625816;
19 Kharitonov and Barnes, 2000, 625817). Other biological media analyzed included EEC, induced
20 sputum, and nasal lavage fluid, all of which are hypothesized to contain aerosolized particles and/or
21 cells from fluid lining the lower and upper airways (Balbi et al., 2007, 625784; Howarth et al., 2005,
22 625805; Hunt, 2002, 625808). These fluids contain cytokines, cells, and markers of oxidative stress
23 that mediate inflammatory responses underlying asthma pathogenesis and exacerbation. Ozone has
24 been demonstrated to increase formation of reactive oxygen species (ROS) and oxidation products in
25 airways (Section 5.1.2) ((Frampton et al., 1999, 040757; Mudway and Kelly, 2000, 010452). and
26 oxidative stress has been linked to asthma by regulating expression of cytokines and activity of
27 inflammatory cells in airways (Heidenfelder et al., 2009, 190026). Recent studies examined
28 8-isoprostane, which is a prostaglandin F2a-like compound produced by ROS via the nonenzymatic
29 peroxidation of arachidonic acid in membrane phospholipids (Morrow et al., 1990, 625835). EEC
30 8-isoprostane levels are consistently higher in asthmatics than in nonasthmatics and increase upon
31 asthma exacerbation (Baraldi et al., 2003, 625802). Studies also measured thiobarbituric acid
32 reactive substances (TEARS) to represent oxidative stress. TEARS are derived from oxidative
33 degradation of lipids and sugars (Janero, 1990, 625809).
34 Table 6-12 presents the characteristics and ambient O3 concentration data from recent studies
35 assessing associations between O3 exposure and biological markers of airway inflammation and
36 oxidative stress. Many recent studies reported positive associations between short-term ambient O3
37 exposure and increases in airway inflammation and oxidative stress, in particular, studies of
38 asthmatic children in Mexico City (Figures 6-10 and 6-11 and Tables 6-13 and 6-14). Further, diet
39 and antioxidant intake were identified as potential susceptibility factors.
March 2011 6-47 DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-12. Mean and upper percentile ozone concentrations in studies examining biological markers
of airway inflammation and oxidative stress
Study
Qian et al. (2009, 5487931
Khatri et al. (2009, 5942821
Ferdinands etal. (2008,
1564331
Adamkiewicz et al. (2007,
187925)
Delfinoetal. (2010, 647222)
Liu et al. (2009, 1920031
Sienra-Mongeetal. (2004,
1964221
Barraza-Villarreal et al.
(2008, 1562541
Romieu et al. (2008,
1799081
Chimentietal. (2009,
4188281
Rodriguez et al. (2007,
0928421
Location
6 U.S. communities
(SOCS)
Atlanta, GA
Atlanta, GA
Steubenville, OH
Los Angeles, CA
Windsor, ON, Canada
Mexico City, Mexico
Mexico City, Mexico
Mexico City, Mexico
Sicily, Italy
Perth, Australia
Years
1997-1999
All-year
2003, 2005,
2006
Warm
season
2004
Warm
season
2000
Cold season
2005-2007
All-year
2005
Cold season
1999-2000
All-year
2003-2005
All-year
2004
All-year
NR
All-year
1996-2001
All-year
Metrics
8-h max
8-h max
1-h max
24-h avg
1-h max
24-h avg
24-h avg
1-h max
8-h max
8-h max
1-h max
8-h max
8-h avg
(07:00-
15:00)
24-h avg
1-h max
Mean
Concentration (ppb)
33.6
59a
71
15.3
19.8
Warm season: 33.3
Cool season: 20.6
14.1
27.2
66.2
31.6
86.5
31.1
Fall: 32.7 (week), 35.1 (race)0
Winter: 37.0 (week), 30.8
(race)"
Summer: 51.2 (week), 46.1
(race)"
28
33
Middle/Upper Percentile
Concentrations (ppb)
1.6-91.5
Range: 44-73
Median: 61
IQR: 54-67
Maximum: 32.2
Maximum: 61.6
Range: 8.04-76.4 (warm season), 6.17-44.9
(cool season)
Median: 13.0; IQR: 8.8-17.8
Median: 27.0; IQR: 21. 8-32.8
Range: 11.1-142.5
IQR: 22.0 (8-h); Range: 4.9-86.3
IQR:48.0; Range: NR
Median: 31. 4
Range: 9.8-60.7
NR
Range: 9-74
Range: 12-95
IQR = interquartile range , NR = Not Reported.
'Personal exposure estimates were derived based on time spent in the vicinity of various 03 monitors.
Concentrations converted from pg/m to ppb using the conversion factor of 0.51 assuming standard temperature (25°C) and pressure (1 atm).
March 2011
6-48
DRAFT - DO NOT CITE OR QUOTE
-------
Study
Qianetal. (2009)
Liu etal. (2009)
Barraza-Villarreal et al.
(2008)
Population eNO units Lag Subgroup
Asthmatic children ppb 0
Asthmatic children % change 0
Children
ppb
Khatri et al. (2009) Adults ppb
Delfino etal. (2010) Olderadults ppb
Nonasthmatics
Asthmatics
0-4 avg Cool season
Warm season*
Adamkiewicz et al.
(2004)
Olderadults
ppb
»
•
:s
-2024
Change in concentration (95% Cl)
Figure 6-10. Associations of ambient ozone exposure with changes in concentrations of exhaled
nitric oxide (eNO).
All results are from single-pollutant models. Effect estimates were standardized to a 30- or 20-
ppb increase for 8-h max or 24-h avg ozone, respectively.
Table 6-13. Additional characteristics and quantitative data for studies presented in Figure 6-10
Study
Qian et al. (2009, 5487931
Liu et al. (2009, 1920031
Barraza-Villarreal et al. (2008, 1562541
Khatri et al. (2009, 5942821
Delfino etal. (2010, 647222)
Adamkiewicz et al. (2007, 1879251
Location/
Population
6 U.S. communities
Asthmatic children
Windsor, ON, Canada
Asthmatic children
Mexico City, Mexico
Asthmatic children
Atlanta, GA
Los Angeles, CA
Steubenville, Ohio
O3Lag
0
0
0
2
0-4 avg
0
O3 Averaging Time
8-h max
24-h avg
8-h max
8-h max
24-h avg
24-h avg
eNO Units
ppb
percent change
ppb
ppb
ppb
ppb
Subgroup
Nonasthmatics
Asthmatics
Cool season
Warm season
Effect Estimate
(95% Cl)a
-0.27 (-0.39, -0.15)
-0.1 7 (-0.30, -0.01)
4.5 3.5, 5.9)
4.2 4.0, 4.5)
4.17(0.14,8.2)
4.06(1.25,6.87)
-0.10 (-2.31, 2.11)
-1.74 (-3.64, 0.17)
"Effect estimates were standardized to a 30- or 20-ppb increase for 8-h max or 24-h avg 03, respectively.
March 2011
6-49
DRAFT - DO NOT CITE OR QUOTE
-------
Study
Liu et al. (2009)
Romieu et al. (2008)
Population
Outcome
Lag Subgroup
A .u .• u-i.j EBC 8-isoprostane (%
Asthmatic children u \ °
change)
EBC TEARS (% change)
Asthmatic children EBC MDA (In pg/ml)
Sienra-Monge etal. (2004) Asthmatic children Nasal lavage GSx (In pg/ml) 3
Nasal lavage Uric acid (In
pg/ml)
Sienra-Monge etal. (2004) Asthmatic children Nasal lavage IL-8 (In pg/ml) 3
Nasal lavage IL-6 (In pg/ml)
Barraza-Villarreal et al.
(2008)
Khatri et al. (2009)
Children
Adults
Nasal lavage IL-8 (pg/ml) 0
Blood eosinophils (%
change)
Placebo
Vit C/E
Placebo
Vit C/E
Placebo
Vit C/E
Placebo
Vit C/E
Nonasthmatics
Asthmatics
-0.5 0.5 1.5
Change in concentration (95% Cl)
Figure 6-11. Associations of ambient ozone exposure with biological markers of airway oxidative
stress and airway inflammation. [EBC = exhaled breath condensate, TBARS =
thiobarbituric acid reactive substances, MDA = malondialdehyde, GSx = glutathione,
IL-8 = interleukin 8, IL-6 = interleukin 6, Vit C/E = group supplemented with vitamins
C and E. Effect estimates were standardized to a 30- or 20-ppb increase for 8-h max
or 24-h avg ozone, respectively.
March 2011
6-50
DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-14. Additional characteristics and quantitative data for studies presented in Figure 6-11
Study
Liu et al. (2009, 1920031
Romieu et al. (2008, 1799081
Sienra-Monge et al. (2004, 1964221
Barraza-Villarreal et al. (2008, 1562541
Khatri et al. (2009, 5942821
Location/ _ . O3 Averaging
Population 3 g Time
Windsor, ON, Canada n 0, h
Asthmatic children u M'n avg
Mexico City, Mexico n Q ,
Asthmatic children ° 8'h max
Mexico City, Mexico , „ ,
Asthmatic children J H'n max
Mex,o City, Mexico „ H „,„
K'GA 2 8-h™
Outcome
EEC 8-isoprostane
(percent change)
EEC TEARS (percent
change)
EBCMDA(lnpg/mL)
Nasal lavage GSx (In
pg/mL)
Nasal lavage Uric
acid (In pg/mL)
Nasal lavage IL-8 (In
pg/mL)
Nasal lavage IL-6 (In
pg/mL)
Nasal lavage IL-8
(pg/mL)
Blood eosinophils
(percent change)
Subgroup
Placebo
VitC/E
Placebo
\/it r/F
Placebo
VitC/E
VitC/E
Nonasthmatics
Asthmatics
Effect estimate
(95% Cl)a
0.10 (-0.09, 0.34
0.07 (-0.1 8, 0.41
0.28 (0.03, 0.54)
-0.08 -0.14, -0.02)
-0.02 -0.07, 0.03)
-0.10 (-0.26, 0.06)
0.1 2 (-0.04, 0.29)
0.23 (0.02, 0.45)
0.09 (-0.14, 0.31)
0.32(0.14,0.50)
0.01 (-0.1 7, 0.19)
1.62(1.36, 1.88)
1.61 (1.42, 1.80)
1.83(0.62,4.28)
EBC = exhaled breath condensate, TEARS = thiobarbituric acid reactive substances, MDA= malondialdehyde, GSx = glutathione, IL-8 = interleukin 8, IL-6
= interleukin 6, Vit C/E = group supplemented with vitamins C and E.
'Effect estimates were standardized to a 30- or 20-ppb increase for 8-h max or 24-h avg 03, respectively.
Asthmatic Subjects
1 Among asthmatics, evidence of association between O3 and eNO was inconsistent, with
2 studies reporting positive and negative associations. In the multicity (Boston, MA; New York, NY;
3 Denver, CO; Philadelphia, PA; San Francisco, CA; and Madison, WI) salmeterol ((3-2 adrenergic
4 agonist) trial of 119 persistent asthmatics, 12-65 years of age, increases in O3 exposure were
5 associated with statistically significant decreases in eNO (Qian et al., 2009, 548793) eNO was
6 measured every 2-4 weeks over a 16-week period between February 1997 and January 1999 and
7 related to 8-h max O3 exposures (single-day lags 0 to 4 days and 0-4 day avg). Among all subjects,
8 increases in lag 0 and 0-4 avg O3 were associated with the largest decreases in eNO (-0.27 ppb [95%
9 CI: -0.39, -0.15] per 30 ppb increase in lag 0 of 8-h max O3). Subgroup analyses did not reveal
10 strong heterogeneity in response among salmeterol, CS, or placebo groups. Associations of NO2 and
11 PMio with eNO were positive and statistically significant in all three treatment groups, suggesting
12 that the counterintuitive findings for O3 were not simply due to the reduction of inflammatory
13 responses by medication use. The authors suggested that at higher O3 exposures, O3 may rapidly
14 react with NO in airways to form reactive nitrogen species such as peroxynitrite. In the cross-
15 sectional study of adults in Atlanta, GA, Khatri et al. (2009, 594282) observed that a 30-ppb increase
16 in lag 1 of 8-h max O3 was associated with a 4.17-ppb increase in eNO (95% CI: 0.14, 8.2) among
17 asthmatics. Consistent with eNO results, O3 was also positively associated with blood eosinophils,
18 which are believed to be the main effector cells that initiate and sustain inflammation in asthma and
March 2011
6-51
DRAFT - DO NOT CITE OR QUOTE
-------
1 allergy (Schmekel et al, 2001, 625849). In one of the few studies to compare effects between
2 asthmatics and nonasthmatics, Barraza-Villarreal et al. (2008, 156254) found that O3 may increase
3 eNO in both asthmatic and healthy children. Among asthmatics, a 30-ppb increase in lag 0 of 8-h
4 max O3 was associated with a 1.45-ppb increase (95% CI: 1.39, 1.50) in eNO, and the association
5 remained statistically significant in a co-pollutant model with PM2 5. A slightly larger effect was
6 estimated for nonasthmatics.
7 Similar to other studies restricted to winter months, Liu et al. (2009, 192003) (described in
8 Section 6.2.1.2) reported a negative association between O3 and eNO. Results for EEC levels of
9 TEARS and 8-isoprostane also did not provide strong evidence of O3 effects on airway oxidative
10 stress. SO2, NO2, and PM25 had larger, positive estimated effects on all three biomarkers, suggesting
11 that in the winter when O3 concentrations are low, other more dominant pollutants may have stronger
12 effects on respiratory health endpoints.
13 Several studies of asthmatic children in Mexico City demonstrated associations between acute
14 changes in ambient O3 and changes in an array of proinflammatory and oxidative stress mediators
15 (Barraza-Villarreal et al., 2008, 156254: Romieu et al., 2008, 179908: Romieu et al., 2009, 548788:
16 Sienra-Monge et al., 2004, 196422). By also examining differences in responses by antioxidant
17 intake, these studies, as a whole, provided strong evidence that inhaled O3 may be an important
18 source of ROS in airways and/or may increase airway inflammation via oxidative stress-mediated
19 mechanisms. All of these studies measured nasal lavage levels of the cytokine IL-8. Although IL-8 is
20 not believed to be a key mediator of the asthmatic response, higher IL-8 levels have been described
21 in asthmatics. Further, observations of increased IL-8 expression as a consequence of increased
22 expression of other proinflammatory cytokines by secondary reaction products suggest that IL-8 may
23 be a nonspecific downstream indicator of systemic oxidative stress and inflammation (Section 5.1.2).
24 Other markers analyzed in these studies included eNO, TEARS, EEC pH, and IL-6, which have
25 shown stronger relationships with asthma and airway inflammation. For example, EEC pH, which
26 reflects the proton-buffering capacity of ammonium in airways, is consistently lower in asthmatics,
27 decreases upon acute asthma exacerbation, and is negatively correlated with airway levels of
28 proinflammatory cytokines (Carpagnano et al., 2005, 625789: Hunt et al., 2000, 002173: Kostikas et
29 al., 2002, 625821).
30 Romieu et al. (2008, 179908) analyzed malondialdehyde (MDA, a TEARS) in EEC samples
31 collected biweekly between January and October 2004. A 30-ppb increase in lag 0 of 8-h max O3
32 was associated with a 0.29 nmol (95% CI: 0.03, 0.54) increase in log-transformed MDA. Similar
33 results were reported for lag 1 and 2-day cumulative exposure, and associations were robust to the
34 addition of PM25 into models. Approximately 25% of EEC samples had nondetectable levels of
35 MDA, and the random assignment of concentrations between 0 and 4.1 nmol may have contributed
36 random measurement error to the estimated O3 effects. Because MDA represents less than 1% of
37 lipid peroxides and is present at low concentrations, its reliability as a marker of oxidative stress in
38 vivo has been questioned. However, the authors pointed to their observations of statistically
39 significant associations of EEC MDA levels with FEVi, FVC, and nasal lavage IL-8 levels to
March 2011 6-52 DRAFT - DO NOT CITE OR QUOTE
-------
1 support its analysis as a biologically-relevant indicator of airway inflammation. They also used
2 recent observations of increases in exhaled MDA during acute asthma exacerbation in children
3 (Corradi et al, 2003, 625790) to assert that their findings were evidence for O3-associated increased
4 respiratory morbidity.
5 As with eNO, Barraza-Villarreal et al. (2008, 156254) observed that O3 was associated with
6 similar changes in nasal lavage IL-8 and EEC pH among asthmatic and nonasthmatic children.
7 Among asthmatics, a 30-ppb increase in lag 0 of 8-h max O3 was associated with a 1.61 pg/mL
8 increase (95% CI: 1.42, 1.80) in IL-8, and a 0.10-unit decrease (95% CI: -0.18, -0.01) in EEC pH,
9 and results remained statistically significant in co-pollutant models with PM2 5. In the same cohort of
10 asthmatic and nonasthmatic children, a diet high in FVI was found to protect against O3-related
11 increases in nasal lavage IL-8 (Romieu et al., 2009, 548788). At high ambient O3 levels (> 38 ppb, 8-
12 h max), a 1-unit increase in FVI was associated with a 0.219 decrease (95% CI: -0.38, -0.05) in the
13 natural log of IL-8, suggesting that a diet rich in antioxidants may protect against O3-stimulated
14 nasal inflammation by scavenging ROS. The protective effect was diminished by about 49% at O3
15 levels of 25 ppb or lower.
16 That high levels of antioxidants may protect against O3-associated airway inflammation was
17 also observed by Sienra-Monge et al. (2004, 196422). For 12 weeks, 59 asthmatics received a daily
18 vitamin C and E supplement, and 58 received a placebo. At baseline, 6 weeks, and 12 weeks,
19 investigators measured nasal lavage levels of IL-6, IL-8, uric acid, and total glutathione. While the
20 roles of these markers in the inflammatory cascade of asthma are not well characterized, they have
21 been shown to be induced by experimental O3 exposure (Torres et al., 2000, 005654; Mudway et al.,
22 1999, 001270). IL-6 is involved in recruitment of inflammatory cells to sites of tissue damage. Uric
23 acid and glutathione are ROS scavengers that are present in the RTLF and are observed to be
24 consumed in the initial phase of antioxidant defense against inhaled O3 (Section 5.1.2). Consistent
25 with findings from human clinical and animal studies (Section 5.1.2), the results in the placebo
26 group from Sienra-Monge et al. (2004, 196422) indicate that ambient O3 exposure may initiate an
27 antioxidant response, as indicated by decreases in nasal lavage levels of uric acid and glutathione
28 (Figure 6-11 and Table 6-14). However, despite decreases in the levels of endogenous antioxidants,
29 O3 was positively associated with increases in the inflammatory cytokines, IL-6 and IL-8 (Figure 6-
30 11 and Table 6-14). Results in the vitamin C/E supplementation group indicated that augmenting the
31 circulating levels of antioxidants may confer some protection against O3-associated inflammation in
32 nasal passages. Per a 30-ppb increase in 3-day 8-h max O3, a smaller increase in IL-6 was observed
33 in the vitamin C/E group (0.03 In pg/mL [95% CI: -0.28, 0.33]) than in the placebo group (0.43 In
34 pg/mL [95% CI: 0.16, 0.71]). These observations were supported by other findings in the same
35 cohort that O3-associated increases in respiratory symptoms were higher in subjects with reduced
36 activity in oxidative metabolism genes (Romieu et al., 2006, 090969) (Section 6.2.3.1). In Sienra-
37 Monge et al. (2004, 196422). O3 was associated with increases in uric acid in the placebo group
38 across O3 lags but decreases in glutathione in both the placebo and supplementation group.
39 Therefore, the results do not clearly delineate the interactions among inhaled O3, endogenous
March 2011 6-53 DRAFT - DO NOT CITE OR QUOTE
-------
1 antioxidants, and dietary supplementations of antioxidants. It is also unclear what may be the
2 optimal levels of vitamins C and E to confer protection against O3-associated respiratory effects.
3 Whereas this study found that supplementation of vitamin-deficient subjects to increase vitamin C
4 and E levels to five times above the recommended daily allowance may confer protection, the results
5 from Romieu et al. (2009, 548788) suggested that antioxidant intake from usual diets may be
6 sufficient.
Populations not Restricted to Asthmatic Subjects
7 Although limited in number and sample sizes, consistent with the collective body of evidence,
8 recent studies of subjects engaged in outdoor activities mostly supported a positive association
9 between O3 exposure and airway inflammation. The exception was a well-designed panel study in
10 which 16 adolescent long-distance runners in Atlanta, GA, were followed before and after exercise
11 for 10 days in August 2004 (Ferdinands et al., 2008, 156433). Effect estimates for lags 0, 1, and 2 of
12 1-h max O3 with EEC pH were positive, indicating O3-associated decreases in airway inflammation.
13 In a cross-sectional study of children at camps in south Belgium, although O3 was not associated
14 with lung function, an association was found for eNO. Children at camps with 1-h max O3
15 concentrations above 85.2 ppb had greater increases in intraday eNO compared with children at
16 camps with O3 concentrations below 51 ppb. A benchmark dose analysis indicated that the threshold
17 for an O3-induced increase of 4.3 ppb eNO (indicating increased airway inflammation) was 68.6 ppb
18 forthe 1-h max and 56.3 ppb forthe 8-hr max O3. Among 9 adult male runners in Sicily, Italy
19 examined 3 days before and 20 hours after 3 races in fall, winter, and summer, weekly average O3
20 concentrations (8-h avg, 7:00-15:00) were positively correlated with apoptosis of airway cells
21 (Spearman's r = 0.76, p < 0.0005) and bronchial epithelial cell differential counts (Spearman's r =
22 0.467, p < 0.05) but not with neutrophil or macrophage cell counts or levels of the proinflammatory
23 cytokines TNF-a and IL-8 (Chimenti et al., 2009, 418828). These limited data from Chimenti et al.
24 (2009, 418828) suggested that O3 exposure during exercise may increase airway epithelial injury and
25 activate anti-inflammatory mechanisms such as apoptosis; however, studies with a larger number of
26 subjects and repeated measures are needed to strengthen the evidence.
27 Panel studies examining O3-associated changes in eNO in elderly subjects produced
28 contrasting findings. Both studies were similar in that outdoor O3 was monitored by investigators in
29 the vicinity of subjects' residences and cool season-specific results were presented. However, several
30 differences were noteworthy, including geographic location, inclusion of healthy subjects, and
31 examination of multiday average exposures. Delfino et al. (2010, 647222) followed 60 elderly
32 subjects with coronary artery disease in the Los Angeles, CA area for two 6-week periods, one in the
33 warm season and one in the cool season, although the exact months were not specified. Multiday
34 averages of O3 (3- to 9-day) were associated with statistically significant increases in eNO, with
35 effect estimates increasing with increasing number of averaging days. Additionally, in contrast to
36 most other studies, a strong positive effect was estimated for the cooler season (4.06 ppb [95% CI:
37 1.25, 6.87]) per 20-ppb increase in 5-day avg O3), whereas no association was observed for the warm
March 2011 6-54 DRAFT - DO NOT CITE OR QUOTE
-------
1 season (-0.01 ppb [95% CI: -2.31, 2.11]). Despite the unusual findings, they were similar to findings
2 from another study of Los Angeles area adult asthmatics conducted between October and November,
3 in which O3 was associated with a decrease in indoor activity (Eiswerth et al., 2005, 196443).
4 Adamkiewicz et al. (2004, 087925) did not find a positive association and eNO in a mixed group of
5 older adults (ages 54-91 years) comprising healthy subjects and those with asthma or COPD. The
6 study was conducted in Steubenville, OH between September and December, and as was observed in
7 most other studies conducted during colder months, O3 (concurrent 1 hour and 24 hours preceding
8 eNO collection) was associated with decreases (statistically nonsignificant) in eNO. Associations
9 with other O3 exposure lags were not examined. The authors attributed the negative associations to
10 weak, but negative correlations of O3 with NO and PM2 5, which were associated with small,
11 statistically significant increases in eNO.
Summary of Epidemiologic Studies of Pulmonary Inflammation and Oxidative
Stress
12 Many recent studies reported positive associations between short-term ambient O3 exposure
13 and increases in airway inflammation and oxidative stress, in particular, studies of asthmatic children
14 in Mexico City. Further, diets high in antioxidant vitamin content and antioxidant vitamin
15 supplementation were identified as factors that may protect against O3-associated increases in airway
16 inflammation. Limited evidence suggested that ambient O3 exposure may increase airway
17 inflammation in subjects with increased outdoor exposures and older adults. In the few studies that
18 evaluated co-pollutant models, O3 effect estimates showed small changes in magnitude but little
19 change in statistical significance (Barraza-Villarreal et al., 2008, 156254; Liu et al., 2009, 192003;
20 Romieu et al., 2008, 179908).
21 Several recent studies simultaneously assessed associations of O3 with lung function and
22 biological markers of airway inflammation. In most cases, the results differed among endpoints, and
23 whether evaluated at the same or different lags of O3 exposure, associations were generally stronger
24 for biological markers of airway inflammation than for lung function (Barraza-Villarreal et al., 2008,
25 156254; Escamilla-Nunez et al., 2008, 594284; Nickmilder et al., 2007, 090710). These findings are
26 consistent with those from human controlled exposure studies that indicate a lack of correlation
27 between inflammatory and spirometric responses induced by O3 exposure. Studies have suggested
28 that O3-related respiratory morbidity may occur via multiple mechanisms with varying time courses
29 of action, and the examination of a limited number of O3 exposure lags in these aforementioned
30 studies may explain some of the inconsistencies in associations of O3 with different respiratory
31 health endpoints.
32 Collectively, studies examined associations with single-day O3 concentrations lagged from 0 to
33 5 days, as well concentrations averaged over 2 to 9 days. Lag 0 O3 exposure was most frequently
34 examined and consistently associated with increased airway inflammation and oxidative stress.
35 However, among studies that examined single- and multi-day lags, multiday average O3 exposures
36 were associated with greater increases in airway inflammation and oxidative stress (Delfino et al.,
March 2011 6-55 DRAFT - DO NOT CITE OR QUOTE
-------
1 2010, 647222: Liu et al, 2009, 192003: Sienra-Monge et al, 2004, 196422). Human controlled
2 exposure studies have similarly found that several indicators of airway inflammation remain elevated
3 following exposures to O3 repeated over multiple days (Section 6.2.3.1).
6.2.3.3. Toxicology
4 The 2006 O3 AQCD states that the "extensive human clinical and animal toxicological
5 evidence, together with the limited available epidemiologic evidence, is clearly indicative of a causal
6 role for O3 in inflammatory responses in the airways." Numerous recent in vitro and in vivo studies
7 add to these observations of O3-induced inflammation and injury, and provide new information
8 regarding the underlying mechanisms (Aibo et al., 2010, 378559: Carey et al., 2007, 195752:
9 Castagna et al., 2009, 596372: Cho et al., 2007, 596373: Dahl et al., 2007, 196986: Damera et al.,
10 2009, 596375: Fakhrzadeh et al., 2008, 596380: Farraj et al., 2010, 380846: Garantziotis et al., 2010,
11 624947: Han et al., 2008, 596387: Hicks et al., 2010, 624932: Huffman et al., 2006, 596388: Inoue
12 et al., 2008, 197803: Jang et al., 2005, 195638: Janic et al., 2005, 483658: Johnston et al., 2005,
13 596393: Johnston et al., 2005, 596394: Johnston et al., 2006, 097439: Johnston et al., 2007, 596392:
14 Kenyon et al., 2006, 596396: Kooter et al., 2007, 596397: Manzer et al., 2006, 596404: Oslund et al.,
15 2008, 195654: Oslund et al., 2009, 201539: Oyarzun et al., 2005, 596407: Plopper et al., 2006,
16 596410: Servais et al., 2005, 195667: Vancza et al., 2009, 596419: Voynow et al., 2009, 194311:
17 Wagner et al., 2007, 596420: Wang et al., 2007, 596421: Yoon et al., 2007, 596422V
18 The similarities of non-human primates to humans make them attractive models in which to
19 study the pulmonary response to O3. A single 6 h exposure of adult male Cynomolgus monkeys to
20 1 ppm O3 induced significant increases in inflammatory and injury markers, including BAL
21 neutrophils, total protein, alkaline phosphatase, IL-6, IL-8, and G-CSF (Hicks et al., 2010, 624932).
22 Gene expression analysis confirmed the increases in IL-8, which has been previously observed in
23 O3-induced pulmonary injury in rhesus monkeys (Chang MM-J; Wu et al., 1998, 011983). Anti-
24 inflammatory IL-10 was also elevated, but fold changes in IL-10 and G-CSF were relatively low and
25 highly variable. The single exposure also caused necrosis and sloughing of the epithelial lining of the
26 most distal portions of the terminal bronchioles and the respiratory bronchioles. Bronchiolitis,
27 alveolitis, parenchymal and centriacinar inflammation were also observed. A second exposure
28 protocol (two exposures with a 2-week inter-exposure interval) resulted in similar inflammatory
29 responses, with the exception of total protein and alkaline phosphatase levels which were attenuated,
30 indicating some adaptation with respect to injury.
31 Exposure of adult BALB/c mice to 0.01 ppm O3 for 4 hours increased BAL levels of
32 keratinocyte chemoattractant (KC; IL-8 homologue) (~ sixfold), IL-6 (~12-fold), and TNF-a (~
33 twofold) (Damera et al., 2010, 380255). Additionally, O3 increased BAL neutrophils by 21% without
34 changes in other cell types. A trend of increased neutrophils with increased O3 concentration
35 (0.12-2 ppm) was observed in BALB/c mice exposed for 3 hours (Jang et al., 2005, 195638).
36 Although alterations in the epithelium of the airways were not evident in 129J mice after 4 hours of
37 exposure to 0.2 ppm O3 (Plopper et al., 2006, 596410). detachment of the bronchiolar epithelium
March 2011 6-56 DRAFT - DO NOT CITE OR QUOTE
-------
1 was observed in SD rats after 5 days or 60 days of exposure to 0.25 ppm O3 (Oyarzun et al., 2005,
2 596407). Subacute (65 hours) exposure to 0.3 ppm O3 induced pulmonary inflammation and
3 enhanced vascular permeability in mixed strain mice (129/Ola and C57BL/6) (Inoue et al., 2008,
4 197803). Three hours or 72 hours of exposure to 0.3 ppm O3 resulted in similar levels of IL-6
5 expression in the lungs of C57BL/6 mice (Johnston et al., 2005, 596394). along with increases in
6 BAL protein, sTNFRl, and sTNFR2. Increased neutrophils were observed only after the 72-h
7 exposure, and neither exposure resulted in detectable levels of IL-6 or KC protein. Levels of BAL
8 protein, sTNFRl, and sTNFR2 were higher in the 72-h exposure group than in the 3-h exposure
9 group. In another study, the same subacute (72 hours) exposure protocol elicited increases in BALF
10 protein, IP-10, sTNFRl, macrophages, neutrophils, and IL-6, IL-la, and IL-1(3 expression (Johnston
11 et al., 2007, 596392). Yoon et al. (2007, 596422) exposed C57BL/6J mice continuously to 0.3 ppm
12 O3 for 6, 24, 48, or 72 hours, and observed elevated levels of KC, MIP-2, metalloproteinases, and
13 inflammatory cells in the lungs at various time points.
14 After exposing adult C57BL mice to 0.5 ppm O3 for 3 hours, Han et al. (2008, 596387)
15 observed early (5 hours postexposure) increases in BAL TNF-a and IL-1(3, which diminished by
16 24 hours postexposure. Total BAL protein was elevated at 24 hours, but there were only minimal or
17 negligible changes in LDH, total cells, or PMNs. Ozone increased BAL mucin levels (with statistical
18 significance by 24 hours postexposure), and significantly elevated surfactant protein D at both time
19 points. Prior intratracheal (IT) exposure to multiwall carbon nanotubes enhanced most of these
20 effects, but the majority of responses to the combined exposure were not greater than those to
21 nanotubes alone. Ozone exposure did not induce markers of oxidative stress in lung tissue, BAL, or
22 serum. Consistent with this study, Aibo et al. (2010, 378559) did not detect changes in BAL
23 inflammatory cell numbers in the same mouse strain after a 6-h exposure to 0.25 or 0.5 ppm. The
24 majority of inflammatory cytokines (pulmonary or circulating) were not significantly changed (as
25 assessed 9 hours post O3 exposure).
26 In a study examining age, strain, and gender as factors for susceptibility to O3 in mice,
27 increased BAL neutrophils were observed in four strains of neonates 24 hours after exposure to
28 0.8 ppm O3 for 5 hours (Vancza et al., 2009, 596419). Three of these strains also exhibited increased
29 BAL protein, although the two endpoints were not necessarily consistently correlated in a given
30 strain. Inflammation and injury were observed in adult mice as well. A study assessing NQO1 as a
31 susceptibility factor was conducted by Voynow et al. (2009, 194311). Specific effects of this gene on
32 O3 responses are discussed in Chapter 8; only ozone's effects in wild type C57BL/6 mice are
33 described here. Exposure to 1 ppm for 3 hours increased BAL total cells, neutrophils, and KC; these
34 responses were greatest at 24 hours postexposure. F2-isoprostane (8-isoprostane), a marker of
35 oxidative stress, was also elevated by O3, peaking at 48 hours postexposure.
36 Atopic asthma appears to be a risk factor for more severe O3 induced airway inflammation in
37 humans (Balmes et al., 1997, 086092: Scannell et al., 1996, 080755). and allergic animal models are
38 often used to investigate the effects of O3 on this susceptible population. Farraj et al. (2010, 380846)
39 exposed allergen-sensitized adult male BALB/c mice to 0.5 ppm O3 for 5 hours once per week for
March 2011 6-57 DRAFT - DO NOT CITE OR QUOTE
-------
1 4 weeks. Ovalbumin-sensitized mice exposed to O3 had significantly increased BAL eosinophils by
2 85% and neutrophils by 103% relative to OVA sensitized mice exposed to air, but these changes
3 were not evident upon histopathologic evaluation of the lung, and no O3 induced lesions were
4 evident in the nasal passages. Ozone increased BAL levels of N-acetyl-glucosaminidase (NAG; a
5 marker of injury) and protein. DEP co-exposure (2.0 mg/m3, nose only) inhibited these responses.
6 Wagner et al. (2007, 596420) exposed the relatively O3-resistant Brown Norway rat strain to 1 ppm
7 O3 after sensitizing and challenging with OVA. Rats were exposed for 2 days, and airway
8 inflammation was assessed one day later. Filtered air for controls contained less than 0.02 ppm O3.
9 Histopathology indicated O3 induced site-specific lung lesions in the centriacinar regions,
10 characterized by wall thickening partly due to inflammatory cells influx. BAL neutrophils were
11 elevated by O3 in allergic rats, and modestly increased in non-allergic animals (not significant). A
12 slight (but not significant) increase in macrophages was observed, but eosinophil numbers were not
13 affected by O3. Soluble mediators of inflammation (Cys-LT, MCP-1, and IL-6) were elevated by O3
14 in allergic animals but not non-allergic rats. Treatment with yT, which neutralizes oxidized lipid
15 radicals and protects lipids and proteins from nitrosative damage, did not alter the morphologic
16 character or severity of the centriacinar lesions caused by O3, nor did it reduce neutrophil influx. It
17 did, however, significantly reduce O3-induced soluble inflammatory mediators in allergic rats.
Mechanisms of Injury
18 Since O3 has been well established as a causative agent of airway inflammation and injury, the
19 majority of recent research has focused on the underlying mechanisms. A brief description of some
20 of the recent contributions to this area of research is provided here; more detailed descriptions of the
21 mechanisms behind O3-mediated injury and inflammation can be found in the mode of action
22 chapter (Chapter 5). There are several signaling pathways responsive to changes in oxidation status,
23 which tend to be influenced at different levels in different host backgrounds. The molecular
24 mechanisms of TNF receptor-mediated lung injury induced by O3 and associated signaling pathways
25 (NF-KB, MAPK/AP-1) have been examined (Cho et al., 2007, 596373: Fakhrzadeh et al., 2008,
26 596380). along with the changes in gene expression which characterize O3-induced stress and
27 inflammation (Wang et al., 2007, 596421). Other contributors to injury and inflammation include the
28 IL-1 and neurokinin receptors (Johnston et al., 2007, 596392; Oslund et al., 2008, 195654).
29 calcitonin gene-related peptide receptor activation (Oslund et al., 2009, 201539). CXCR2, a receptor
30 for neutrophil chemokines (Johnston et al., 2005, 596393). and NQO1 (Voynow et al., 2009,
31 194311). an enzyme involved in oxidative stress. Studies indicate a role for oxidative stress in
32 mediating inflammation (Jang et al., 2005, 195638; Wagner et al., 2007, 596420). Protective roles
33 have been identified for nitric oxide synthase (Kenyon et al., 2006, 596396). metallothionein (Inoue
34 et al., 2008, 197803). matrix metalloproteinases (Yoon et al., 2007, 596422). Clara cell secretory
35 protein (Plopper et al., 2006, 596410). and the recognition of oxidized lipids by alveolar
36 macrophages (Dahl et al., 2007, 196986).
March 2011 6-58 DRAFT - DO NOT CITE OR QUOTE
-------
6.2.4. Respiratory Symptoms and Medication Use
6.2.4.1. Epidemiology
1 In epidemiologic studies, respiratory symptom data are typically collected by having subjects
2 or parents record symptoms such as wheeze, cough, and shortness of breath and medication use in a
3 diary without direct supervision by study staff. Several limitations of symptom reports are well-
4 recognized: recall error if not recorded daily, differences among subjects in the interpretation of
5 symptoms, biased reporting between asthmatic and healthy participants and between known high and
6 low pollution days, and occurrence in a smaller percent of the population compared with changes in
7 lung function and mediators of airway inflammation. Nonetheless, symptom diaries remain a
8 convenient and useful tool to collect individual-level data from a large number of subjects and allow
9 the modeling associations of daily changes in O3 exposure with daily changes in respiratory
10 morbidity. Furthermore, they represent an overt clinical effect and may have greater impact on
11 subjects' activities. Importantly, most of the limitations described above are sources of random
12 measurement error that can bias effect estimates to the null or increase the uncertainty around effect
13 estimates.
14 Table 6-15 presents the characteristics and ambient O3 concentration data from studies
15 assessing associations between O3 exposure and respiratory symptoms and medication use. Most
16 studies have been conducted in asthmatics, and the collective body of epidemiologic evidence
17 strongly supports associations between acute increases in ambient O3 exposure and increases in
18 respiratory symptoms in children with asthma (U.S. EPA, 1996, 017831: U.S. EPA, 2006, 088089)
19 (Figures 6-12 and Tables 6-16). Evidence also indicated that O3 exposure is likely to be associated
20 with increased use of asthma medication (Figure 6-13 and Table 6-17). The effect of O3 exposure on
21 respiratory symptoms in subjects with other pre-existing diseases and in healthy children is not
22 clearly indicated (Figure 6-14 and Table 6-18).
March 2011 6-59 DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-15. Mean and upper percentile ozone concentrations in studies examining respiratory
symptoms, asthma medication use, and activity levels
Study
Mortimer etal. (2002,
0302811
O'Connor etal. (2008,
1568181
Schildc rout etal.
(2006, 0898121
Apte et al. (2008,
1958651
Gent et al. (2003,
0528851
Thurstonetal. (1997,
0776451
Triche et al. (2006,
0932741
Khatri et al. (2009,
5942821
Ross et al. (2002,
0427491
Rabinovitch et al.
(2004, 0967531
Mann etal. (2010,
6358271
Linn etal. (1996,
0825081
Ostroetal. (2001,
0167021
Delfino et al. (2003,
0504601
Eiswerth et al. (2005,
1964431
Romieuetal. (1996,
0807481
Romieuetal. (1997,
0858071
Gold etal. (1999,
0869191
Romieu et al. (2006,
0909691
Escamilla-Nunez et
al. (2008, 5942841
Gielenetal. (1997,
0835921
HoekandBrunekreef
(1995,0461841
Just etal. (2002,
0354291
Feo Brito et al. (2007,
0932591
Park et al. (2005,
0886731
Moon et al. (2009,
1902971
Location
8 U.S. communities:
(NCICAS)
7 U.S. communities:
(ICAS)
8 U.S. communities
(CAMP)
Multiple U.S. cities
CT, southern MA
Connecticut River
Valley, CT
southwestern VA
Atlanta, GA
Moline, Illinois
Denver, CO
Fresno/Clovia,
California
Rubidoux, Upland,
Torrence, CA
Los Angeles, CA
Los Angeles, CA
Los Angeles, CA
northern Mexico City,
Mexico
southern Mexico City,
Mexico
Mexico City, Mexico
Mexico City, Mexico
Mexico City, Mexico
Amsterdam,
Netherlands
Deurne and
Enkhuizen,
Netherlands
Paris, France
Ciudad Real and
Puertollano, Spain
Incheon, Korea
4 cities, South Korea
Years/Season
1993
Warm season
1998-2001
All-year
1993-1995
Warm season
1994-1998
All-year
2001
Warm season
1991-1993
Warm season
1995-1996
Warm season
2003, 2005, 2006
Warm season
1994
April-October
2000-2003
Cold season
2000-2005
All-year
1992-1993, 1993-
1994
Fall and spring
1993
Warm season
1999-2000
Cold season
1983
Cold season
1991-1992
Warm and cold
season
1991-1992
Warm and cold
season
1991
Winter, spring, fall
1998-2000
All-year
2003-2005
All-year
1995
Warm season
1989
March-July
1996
April-June
2000-2001
Warm season
March-June 2002
April-May, 2003
Warm season
Metrics
8-h avg
(10:00 a.m. to
6:00 p.m.)
24-h avg
1-h max
24-h avg
Workday avg
(8:00 a.m. -
5:00 p.m.)
1 -h avg
8-h avg
1-h max
8-h max
8-h max
8-h avg
1-h max
8-h max
24-h avg
1-h max
1-h max
8-h max
1-h max
1-h max
1-h max
24-h avg
8-h max
1-h max
8-h max
1-h max
8-h max
1-h max
8-h avg
1-h max
24-h avg
24-h avg
Mean
Concentration (ppb)
48
NR
NR
25.5
36.6
58.6
51.3
83.6
54.5
59b
41.5
28.2C
NR
34b
Los Angeles: 59.5
Pasadena: 95.8
25.4
17.1
NR
190
196
52.0
66.2
102
31.6
86.5
33.5
Deurne: 57
Enkhuizen: 59
29.5
65.9 (Ciudad Real!"
56.8 (Puertollano)d
Dust event days: 23.6
Control days: 25.1
NR
Middle/Upper Percentile
Concentrations (ppb)
Approximate IQR = 15a
Approximate median: 20a
Approximate range: 2-50a
Range in medians: 43.5-65.8
Range in 10th-90th: 23.3-53.3
Range: 2.5-67.3
Range: 2.5-107.1
Range: 27.1-125.5
Range: 21. 4-99.6
Range: 20-1 60
Range: 23.5-87.6
Range: 44-73b
Range: 8.9-78.3
Median: 30.0C
Range: 0-70.0C
Median: 49.4
Range: 3.7-120.0
Range: 7-86b
Range: 10-130
Range: 10-220
Range: 4-52
Range: 3-37
NR
Range: 40-370
Range: 40-390
IQR: 25
Range: 7.9-103
Range: 11.1-142.5
Range: 12-309
IQR: 22.0 (8-h); Range: 4.9-86.3
IQR:48.0; Range: NR
Range: 13.8-55.4
Range: 22-1 07
Range: 4-114
5-61
Range: 45.4-101. 5° (Ciudad Real),
11.2-70.5d (Puertollano)
NR
NR
March 2011
6-60
DRAFT - DO NOT CITE OR QUOTE
-------
1993^
Rodriguez et al. D,h A.lletr-ii- 1996-2001 24-h avg 28 Range: 9-74
(2007, 092842) KerTn' AUSlralla All-year _ 1-hmax _ 33 _ Range: 12-95 _
NCICAS= National Cooperative Inner-City Asthma Study, IQR = interquartile range, ICAS = Inner City Asthma Study, NR = Not Reported, CAMP =
Childhood Asthma Management Program
'Quantitative results not presented. Concentrations estimated from data presented in a figure.
Personal exposure estimates were derived based on time spent in the vicinity of various 03 monitors.
"Measured at sites established by investigators.
Concentrations converted from jjg/m to ppb using the conversion factor of 0.51 assuming standard temperature (25°C) and pressure (1 atm).
Asthmatic Subjects
1 The strong body of evidence demonstrating associations between ambient O3 exposure and
2 respiratory symptoms among asthmatic children and adults mostly comprises several single-region
3 or single-city studies (Figure 6-12 and Table 6-16). In contrast, U.S. multicity studies provided
4 mixed evidence of O3-associated increases in respiratory symptoms among asthmatic children
5 (Mortimer et al., 2002, 030281: O'Connor et al., 2008, 156818: Schildcrout et al., 2006, 089812). In
6 particular, NCICAS (Mortimer et al., 2000, 013255: Mortimer et al., 2002, 030281) and ICAS
7 (O'Connor et al., 2008, 156818) of different children from mostly the same cities produced
8 contrasting findings. In the NCICAS cohort, lag 1-4 avg O3 was positively associated with morning
9 asthma symptoms (OR: 1.35 [95% CI: 1.04, 1.69] per 30-ppb increase in 8-h avg O3) (Mortimer et
10 al., 2002, 030281). Ozone effect estimates decreased slightly in magnitude in two-pollutant models
11 withSO2(OR: 1.23 [95% CI: 0.94, 1.61]) or NO2 (OR: 1.14 [95% CI: 0.85, 1.59]). In the ICAS
12 cohort (described in section 6.2. 1.2), associations of 19-day avg O3 with wheeze and nighttime
13 asthma were positive and negative, respectively (O'Connor et al., 2008, 156818). NCICAS was
14 conducted during the warm season, and symptom data were collected daily (Mortimer et al., 2000,
15 013255: Mortimer et al., 2002, 030281). whereas in ICAS, every 2 months, parents reported the
16 number of days with respiratory symptoms over the previous 2 weeks (O'Connor et al., 2008,
17 156818). Because of the two-week symptom reporting period, ICAS investigators were precluded
18 from examining associations with single-day and shorter-duration O3 exposure periods.
19 Evidence of O3-associated respiratory symptoms was also weak in another recent U.S.
20 multicity study (Albuquerque, NM; Baltimore, MD; Boston, MA; Denver, CO; San Diego, CA;
21 Seattle, WA; St. Louis, MO; and Toronto, Canada) of 990 asthmatic children (Schildcrout et al.,
22 2006, 089812). In this study, symptom data were collected daily and analyses were restricted to high
23 O3 periods between May and September. In meta-analyses that combined city-specific estimates, a
24 40-ppb increase in daily lag 0 of 1-h max O3 was associated with any asthma symptom with an OR
25 (95% CI) of 1.08 (0.89, 1.31). Odds ratios for lags 1 and 2 and the 3-day sum of O3 were near 1.0.
26 Because O3 analyses were restricted to summer months, the median follow-up of subjects was
27 2 months, and data were available from an average of 12 subjects per day per city, the study may
28 have lacked sufficient power to perform city-specific analyses stratified by season, which the authors
29 suggested may have been necessary to discern O3-related effects.
March 2011 6-61 DRAFT - DO NOT CITE OR QUOTE
-------
Study
Mortimer et al. (2002)
O'Connor et al. (2008)
Schildcroutetal. (2006)
Thurston etal. (1997)
Rabinovitchetal. (2004)
Mann etal. (2010)
Ostroetal. (2001)
Delfinoetal. (2003)
Romieuetal. (1996)
Romieuetal. (1997)
Escamilla-Nunezetal. (2008)
Romieu et al. (2006)
Gielenetal. (1997)
Just etal. (2002)
Jalaludinetal. (2004)
Ross et al. (2002)
Park et al. (2005)
Outcome
Morning symptoms
Wheeze/cough
Asthma symptoms
Chest symptoms
Daytime symptoms
Wheeze
Wheeze
Bothersome symptoms
LRS
LRS
Cough
Difficulty breathing
LRS
URS
Asthma symptoms
Nose irritation
Wheeze
Morning symptom score
Night symptoms
Lag Subgroup
1 -5 avg
7 cities with NO2 data -
1-19 avg —
0
0
0-2 avg
0 All —i
Fungi allergic —
3 -•
g
0
0
0
0-5 avg GSTM1 positive
GSTM1 null
GSTP1 lie/lie I le/Val
GSTP1 Val/Val
0
2
1-3 avg
0
Q
•
-o
• —
-•
— —
I
*
O
•
»
*
•
^
•
-*-
-•-
012345
Odds ratio (95% Cl)
Figure 6-12. Associations of ambient ozone exposure with respiratory symptoms in asthmatic
subjects. All studies are of asthmatic children except for Ross et al. (2002, 042749)
which includes asthmatic children and adults and Park et al. (2005, 088673) which
includes asthmatic adults. LRS = lower respiratory symptoms, URS = upper
respiratory symptoms. Effect estimates are standardized to a 40-, 30-, and 20-ppb
increase for 1-h max, 8-h max (or 8-h avg), and 24-h avg ozone, respectively. Effect
estimates depicted as black circles are from single-pollutant models, and effect
estimates depicted as open circles are from co-pollutant models.
March 2011
6-62
DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-16. Additional characteristics and quantitative data for studies presented in Figure 6-12
Study
Mortimer etal. (2002,0302811
O'Connor etal. (2008, 1568181
Schildcroutetal. (2006, 089812)
Thurstonetal. (1997,0776451
Rabinovitch et al. (2004, 0967531
Mann etal. (2010, 635827)
Ostroetal. (2001,0167021
Delfino et al. (2003, 0504601
Romieuetal. (1996, 0807481
Romieuetal. (1997, 0858071
Escamilla-Nunezetal. (2008,
5942841
Romieu et al. (2006, 0909691
Gielenetal. (1997,0835921
Just etal. (2002, 035429)
Jalaludinetal. (2004, 0565951
Ross et al. (2002, 0427491
Park et al. (2005, 0886731
Location/
Population
8 U.S. communities
Asthmatic children
7 U.S. communities
Asthmatic children
8 U.S. communities
Asthmatic children
CT River Valley, CT
Asthmatic campers
Denver, CO
Asthmatic children
Fresno/Clovia, California
Asthmatic children
Los Angeles, CA
Asthmatic children
Los Angeles, CA
Asthmatic children
northern Mexico City,
Mexico
Asthmatic children
southern Mexico City,
Mexico
Asthmatic children
Mexico City, Mexico
Asthmatic children
Mexico City, Mexico
Asthmatic children
Amsterdam, Netherlands
Asthmatic children
Paris, France
Asthmatic children
Sydney, Australia
Asthmatic children
Moline, Illinois
Asthmatic children and
adults
Incheon, Korea
Asthmatic adults
03
Lag
1 -4 avg
1-19
avg
0
0
0-2 avg
0
3
0
0
0
0
0-5 avg
0
0-2 avg
2
1 -3 avg
0
O3 Averaging
Time
8-h avg
(10:00-18:00)
24-h avg
1-h max
1-h max
1-h max
8-h max
1-h max
1-h max
1-h max
1-h max
1-h max
1-h max
8-h max
8-h avg
1-h max
8-h avg
24-h avg
Outcome
Morning symptoms
Wheeze/cough
Asthma symptoms
Chest symptoms
Daytime symptoms
Wheeze
Wheeze
Symptoms
bothersome/
interfering with
activity
LRS
LRS
Wheeze
Difficulty breathing
LRS
URS
Asthma symptoms
Nose irritation
Wheeze
Morning symptoms
Night symptoms
Subgroup
All 8 cities
7 cities with N02
data
All
Fungi allergic
GSTM1 sufficient
GSTM1 null
GSTP1 lie/lie
Ile/Val
GSTP1 Val/Val
Odds Ratio
(95% Cl)a
1.35(1.04, 1.69)
1.23(0.94, 1.61 withS02
1.25(0.94,1.67
1.14(0.85, 1.55) with N02
1.02(0.86,1.21)
1.03(0.86, 1.21) with PM25,
N02
1.08(0.89, 1.31)
1.28(1.1, 1.5)
1.34(1.01, 1.77)
1.00(0.84,1.19)
1.06(0.84, 1.34)
0.94(0.88,1.00)
1.09(1.04, 1.14)
1.07(1.02, 1.12)
1.06(1.02, 1. 10) with PM 2.5
1.09(1.04, 1.14)
1.09(1.01, 1. 19) with PM 2.5
1.08(1.03, 1.14)
1.10(0.98, 1.24)
1.17(1.02, 1.33
1.06(0.94,1.20
1.30(1.10, 1.53)
1.04(0.75, 1.45)
1.16(1.02, 1.32)
2.7 (0.28, 25.8)
4.2(1.42,12.4)
1.21(0.92,1.59)
1.13(1.05, 1.22)
1.07(0.97, 1.19)
LRS = Lower respiratory symptoms, URS = Upper respiratory symptoms.
'Effect estimates were standardized to a 40-, 30-, and 20-ppb increase for 1-h max, 8-h max (or 8-h avg), and 24-h avg 03, respectively.
1 Previous O3 AQCDs acknowledged uncertainty regarding confounding by airborne allergens
2 or increased susceptibility of atopic asthmatics as few studies considered allergen exposures or
3 allergic sensitization of subjects (U.S. EPA, 1996, 017831: U.S. EPA, 2006, 088089). A growing
4 body of studies aimed to address this gap in knowledge, although results were mixed. Mortimer et al.
5 (2000, 013255) found that although O3 was associated with greater decrements in PEF and incidence
6 of asthma symptoms among nonatopic asthmatics, associations were stronger among asthmatics with
7 higher residential exposures to cat or cockroach allergen. In a recent study of asthmatic children in
March 2011
6-63
DRAFT - DO NOT CITE OR QUOTE
-------
1 Fresno, CA, most associations of single- and multiday lags of O3 exposure (0-14 days) with wheeze
2 were near or below 1.0 (Mann et al., 2010, 635827). The estimated effects did not differ between
3 fungi allergic and fungi nonallergic subjects (ascertained by skin prick test), although the odds ratio
4 was larger among cat nonallergic subjects than among cat allergic. In this study, many subjects were
5 allergic to multiple allergens; however, associations were not compared between subjects with any
6 versus no allergic sensitization. Feo Brito et al. (2007, 093259) followed a group of 137 pollen-
7 allergic asthmatics residing in two cities in central Spain during 1 pollen season (May-June 2000 or
8 2001). In the industrial Puertollano, a 40-ppb increase in lag 3 of 1-h max O3 was associated with a
9 14.3% increase (95% CI: 3.6, 26.0) in the number of subjects reporting symptoms, adjusting only for
10 time trend. A smaller, statistically nonsignificant effect estimate was obtained for pollen. Conversely,
11 in the less industrialized and less polluted Ciudad Real, pollen significantly contributed to increased
12 incidence of respiratory symptoms, whereas O3 did not. While co-pollutant modeling was not
13 conducted, in both locations, O3 and pollen concentrations were weakly correlated, indicating that
14 the findings for O3 were not likely confounded by pollen. Rather, the results suggested that O3 and
15 pollen may have independent effects that vary between locations, depending on the mix of airborne
16 pollutants.
17 Several studies conducted in multiple cohorts of asthmatic children in Mexico City, Mexico
18 demonstrated O3-associated increases in respiratory symptoms (Escamilla-Nuiiez et al., 2008,
19 594284: Romieu et al., 1996, 080748: Romieu et al., 1997, 085807: Romieu et al., 2006, 090969).
20 Recent studies expanded on earlier evidence by providing new information to assess important lags
21 of O3 exposure and factors that may contribute to heterogeneity in symptom responses to ambient O3
22 exposure. For example, Romieu et al. (2006, 090969) and Escamilla-Nunez et al. (2008, 594284)
23 found that the magnitudes of association of ambient O3 exposure with respiratory symptoms and
24 medication use increased with increasing averaging days of O3 exposure. Combined evidence from
25 Romieu et al. (1996, 080748) and Romieu et al. (1997, 085807) indicated that among single-day lags
26 of exposure, lag 0 O3 had the greatest estimated effect on respiratory symptoms. Additionally,
27 although the interaction between O3 and season was not statistically significant, Escamilla-Nunez et
28 al. (2008, 594284) estimated larger effects during the warm season (May-September).
29 Although Romieu et al. (2006, 090969) did not observe differences in associations between O3
30 and lung function by GST polymorphisms (Section 6.2.1.2), they did observe effect modification for
31 respiratory symptoms. Compared with GSTM1 sufficient subjects and GSTP1 lie/lie or Ile/Val
32 subjects, respectively, larger effects were estimated for GSTM1 null subjects and for GSTP1 Val/Val
33 subjects (Figure 6-12 and Table 6-16). Ozone had the greatest estimated effect on difficulty breathing
34 in asthmatics who were both GSTM1 null and GSTP1 Val/Val (OR: 1.49 [95% CI: 1.14, 1.93]).
35 These results add to the body of epidemiologic evidence that deficiencies in oxidant metabolism may
36 increase susceptibility to O3-related respiratory morbidity and also are consistent with findings from
37 human controlled exposure studies indicating increased responsiveness to O3 among GSTM1 null
38 subjects (Section 6.2.1.4). As was discussed in Section 6.2.1.2, compared with the GSTM1 genotype,
39 evidence for susceptibility related to GSTP1 polymorphisms is less certain. Romieu et al. (2006,
March 2011 6-64 DRAFT - DO NOT CITE OR QUOTE
-------
1 090969) found that the GSTP1 Val/Val variant was associated with a lesser O3-associated decrement
2 in lung function but greater risk of respiratory symptoms. Whereas some studies have reported
3 greater risk of asthma among GSTP1 lie/lie or Ile/Val subjects (Hemmingsen et al., 2001, 670827;
4 Mapp et al., 2002, 670826). others have reported greater risk among GSTP1 Val/Val subjects (Tamer
5 et al., 2004, 199914). In Romieu et al. (2006, 090969). GSTP1 lie/lie was associated with greater
6 severity of asthma, and Lee et al. (2004, 090971) also reported greater risk of air pollution-
7 associated asthma among GSTP1 lie/lie subjects.
8 The 2006 O3 AQCD concluded that ambient O3 was likely associated with increased asthma
9 medication use (U.S. EPA, 2006, 088089). Although evidence in recent studies was mixed (Park et
10 al., 2005, 088673: Romieu et al., 2006, 090969: Schildcrout et al., 2006, 089812). the overall body
11 of evidence supports the previous conclusion (Figure 6-13 and Table 6-17). The effects are estimated
12 with greater uncertainty as indicated by the wide 95% CIs. The wide 95% CIs have been attributed to
13 a smaller number of study subjects reporting medication use and the low frequency of use over the
14 study period. Within most studies, findings were similar for respiratory symptoms and asthma
15 medication use. For example, Romieu et al. (2006, 090969) and Escamilla-Nunez et al. (2008,
16 594284) observed positive, statistically significant associations of O3 with both respiratory
17 symptoms and bronchodilator use. Schildcrout et al. (2006, 089812) and Park et al. (2005, 088673).
18 did not observe a positive association for either respiratory symptoms or rescue inhaler use. In
19 contrast, Romieu et al. (1996, 080748) and Rabinovitch et al. (2004, 096753) observed that O3 was
20 positively associated with daytime respiratory symptoms but not with bronchodilator use.
March 2011
6-65
DRAFT - DO NOT CITE OR QUOTE
-------
Study
Schildcrout et al. (2006)
Thurston et al. (1997)
Ostroetal. (2001)
Romieuetal. (1996)
Romieuetal. (1997)
Romieuetal. (2006)
Gielenetal. (1997)
Jalaludin et al. (2004)
Ross et al. (2002)
Parketal. (2005)
Outcome
Rescue inhaler use
Beta-agonist use
Extra medication use
Bronchodilator use
Bronchodilator use
Bronchodilator use
Bronchodilator use
Beta-agonist use/no steroid
ICS use
Asthma medication use
Inhaler use
Lag Subgroup
1
0 — •-
0
0-5 avg GSTP1 lie/lie Ile/Val •—
GSTP1 Val/Val
1 -3 avg
0 — •
•r
0.5
1.5
Odds ratio (95% Cl)
Figure 6-13. Associations of ambient ozone exposure with asthma medication use. All studies
are of asthmatic children except for Ross et al. (2002, 042749) which included
asthmatic children and adults and Park et al. (2005, 088673) which included
asthmatic adults. ICS = corticosteroid use. Effect estimates are standardized to a
40-, 30-, and 20-ppb increase for 1-h max, 8-h max (or 8-h avg), and 24-h avg ozone,
respectively. All effect estimates are from single pollutant models.
March 2011
6-66
DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-17. Additional characteristics and quantitative data for studies presented in Figure 6-13.
Study
Schildcroutetal. (2006, 089812)
Thurstonetal. (1997,0776451
Ostroetal. (2001,0167021
Romieuetal. (1996, 0807481
Romieuetal. (1997, 0858071
Romieu et al. (2006, 0909691
Gielenetal. (1997,0835921
Jalaludinetal. (2004, 0565951
Ross et al. (2002, 0427491
Park et al. (2005, 0886731
Location/
Population
8 U.S. communities
Asthmatic children
CT River Valley, CT
Asthmatic campers
Los Angeles, CA
Asthmatic children
northern Mexico City, Mexico
Asthmatic children
southern Mexico City, Mexico
Asthmatic children
Mexico City, Mexico
Asthmatic children
Amsterdam, Netherlands
Asthmatic children
Sydney, Australia
Asthmatic children
Moline, Illinois
Asthmatic children and adults
Incheon, Korea
Asthmatic adults
O3Lag
0
0
1
0
0
0-5 avg
0
1
1-3 avg
0
OsAj/eragi
1-h max
1-h max
1-h max
1-h max
1-h max
1-h max
8-h max
1-h max
8-h avg
24-h avg
9 Outcome
Rescue inhaler use
Beta-agonist use
Extra medication use
Bronchodilatoruse
Bronchodilatoruse
Bronchodilatoruse
Bronchodilatoruse
Beta-agonist use/no
steroid
ICS use
Asthma medication use
Inhaler use
_ . Odds Ratio
Subgroup (g5% C|)a
1.01(0.89,
1.17(0.96,
1.10(1.03,
0.97 (0.93,
1.02(1.00,
GSTP1 lie/lie I le/Val 0.96(0.90,
GSTP1 Val/Val 1.10(1.02,
1.10(0.78,
1.08(0.89,
1.08(0.96,
1.08(1.00,
0.99 (0.94,
1.15)
1.44)
1.19)
1.01)
1.05)
1.02)
1.19)
1.55)
1.32)
1.21)
1.17)
1.05)
ICS= Inhaled corticosteroid use.
'Effect estimates were standardized to a 40-, 30-, and 20-ppb increase for 1-h max, 8-h max (or 8-h avg), and 24-h avg 03, respectively.
1 While investigation is limited, studies have indicated that O3 exposure may be associated with
2 diminished activity level in asthmatics (Delfino et al., 2003, 050460; Eiswerth et al., 2005, 196443;
3 Khatri et al., 2009, 594282; O'Connor et al., 2008, 156818). In a panel study of asthmatics in
4 Los Angeles CA, Delfino et al. (2003, 050460) found that a 40-ppb increase in 1-h max O3 was
5 associated with symptoms that interfered with daily activity with an OR (95% CI) of 7.41 (1.18,
6 43.2). In a cross-sectional study of asthmatic adults in Atlanta, GA (described in Section 6.2.1.2),
7 Khatri et al. (2009, 594282) observed that a 30-ppb increase in lag 2 of 8-h max O3 was associated
8 with a 0.69-point decrease (95% CI: -1.28, -0.11) in the Juniper quality of life score, which
9 incorporates indices for symptoms, mood, and activity limitations (7-point scale). Eiswerth et al.
10 (2005, 196443) examined the activities of 64 asthmatics (age 16 years and older) in Los Angeles, CA
11 in fall of 1983. A 40-ppb increase in 1-h max O3 was associated with a 0.24% (95% CI: 0.08, 0.40%)
12 lower probability of participation in indoor activities. The associations with outdoor activities were
13 positive but statistically nonsignificant. Although the authors acknowledged that their findings were
14 unexpected and may have been influenced by lack of control for potential confounders, they
15 interpreted the decrease in indoor activities as rest replacing chores. Collectively, these findings
16 highlight the potentially broader impact of O3 exposure on quality of life among asthmatics.
17 Several studies have reported positive associations between long lags of O3 exposure (14-day
18 and 30-day distributed lags or 19-day avg) and school absenteeism among asthmatic children (Chen
19 et al., 2000, 011931; Gilliland et al., 2001, 013232; O'Connor et al., 2008, 156818). Whereas Chen et
20 al. (2000, 011931) and O'Connor et al. (2008, 156818) examined absences for any reason, Gilliland
21 et al. (2001, 013232) found associations with absences for respiratory causes. Despite this evidence,
March 2011
6-67
DRAFT - DO NOT CITE OR QUOTE
-------
1 several limitations have been noted, including the uncertain biological relevance of long lag periods
2 of O3 exposure and the potential for residual seasonal confounding when examining long lag periods
3 of exposure.
Populations not Restricted to Asthmatic Subjects
4 Aside from subjects with increased outdoor exposures, the collective body of epidemiologic
5 evidence has been mixed regarding associations between acute O3 exposure on respiratory symptoms
6 in healthy subjects (Figure 6-14 and Table 6-18). The strongest effects were estimated in a subgroup
7 of infants with asthmatic mothers (Triche et al, 2006, 093274). Triche et al. (2006, 093274)
8 followed 691 infants in southwestern VA followed for 83 days between June and August of 1995
9 and/or 1996 and found that a 20-ppb increase in lag 0 of 24-h avg O3 was associated with odds ratios
10 (95% CI) of 2.34 (1.02, 5.37) for wheeze and of 3.63 (1.81, 7.28) for difficulty breathing among the
11 61 infants with asthmatic mothers. In analyses that included all subjects, the estimated effects were
12 smaller and statistically nonsignificant. Odds ratios increased in magnitude in co-pollutant models
13 that included PM2 5 or PMi0-2.5. While these results suggested that children with asthmatic mothers
14 may be at greater risk of O3-related respiratory morbidity, the authors acknowledged that mothers
15 with asthma may be more likely to report symptoms in their children and that transient wheeze in
16 infants and may not predict respiratory morbidity later in life. Gold et al. (1999, 086919) reported an
17 association between ambient O3 exposure and phlegm in children in Mexico, City; however, they
18 acknowledged being unable to distinguish between the effects of O3 and PM2 5 exposure. Several
19 other studies of school-aged children reported null or negative associations between ambient O3
20 exposure and respiratory symptoms (Hoek and Brunekreef, 1995, 046184; Moon et al., 2009,
21 190297: Rodriguez et al., 2007, 092842V In a large study of 696 children (ages <13 years) in 4
22 regions in South Korea, Moon et al. (2009, 190297) observed a positive association with LRS
23 (cough, phlegm, or wheeze) in the region of Jeju island (OR: 1.08 [95% CI: 0.96, 1.21] per 30 ppb
24 increase in lag 0 8-h avg O3); however, the odds ratios were lower close to 1.0 in other cities and
25 analyses with other symptoms. In a crosssectional study of 4,200 adult workers from 100 office
26 buildings across the U.S., O3 was positively associated with building-related URS (nasal congestion
27 or sore throat) and LRS (wheeze, shortness of breath, or chest tightness) (Apte et al., 2008, 195865).
28 Investigators suggested that the findings may have been attributable to formaldehyde and organic
29 acids produced from O3-initiated reactions within buildings; however, additional data on indoor
30 levels of volatile organic compounds, indoor O3, and infiltration rates would likely be required to
31 characterize the relationship between ambient O3 concentrations and building-related symptoms.
March 2011 6-68 DRAFT - DO NOT CITE OR QUOTE
-------
Study
Triche et al. (2006)
Linnetal. (1997)
Goldetal. (1999)
Hoek and Brunekreef
(1995)
Moon et al. (2009)
Apte et al. (2008)
Outcome
Wheeze
Lag
0
Subgroup
All
With asthmatic mothers
Evening symptom score 0
Phlegm 0
Cough 0
LRS 0
LRS 0
1 2
Odds ratio (95% Cl)
Figure 6-14. Associations of ambient ozone exposure with respiratory symptoms in studies not
restricted to asthmatic populations. U.RS = lower respiratory symptoms. Effect
estimates are standardized to a 40-, 30-, and 20-ppb increase for 1-h max, 8-h max,
and 24-h avg ozone, respectively. Effect estimates depicted as black circles are from
single-pollutant models, and effect estimates depicted as open circles are from
co-pollutant models.
1
2
Table 6-18. Additional characteristics and quantitative data for studies presented in Figure 6-14.
Study
Triche et al. (2006, 0932741
Linnetal. (1996, 082508)
Goldetal. (1999,0869191
Hoek and Brunekreef (1995,
0461841
Moon et al. (2009, 1902971
Apte et al. (2008, 1958651
Location/ Population
southwestern VA
Infants
3 southern CA communities
Children
Mexico City, Mexico
Children
Deurne and Enkhuizen,
Netherlands
Children
4 cities, South Korea
Children
Multiple U.S. cities
Office building workers
03
Lag
0
0
1
0
0
0
O3 Averaging
Time
8-h max
24-h avg
24-h avg
1-h max
24-h avg
24-h avg
Outcome Subgroup
All subjects
Wheeze With asthmatic
mothers
Evening symptom
score
Phlegm
Cough
LRS
LRS
Odds Ratio
(95% Cl)a
1.60(0.85,3.0)
1.79(1.05,3.05)
with PM25
2.34(1.02,5.37)
2.99(1.44,6.23)
with PM2.5
0.38(0.11,1.29)
1.04(1.00, 1.07)
0.95(0.85, 1.06)
0.96(0.90,1.03)
1.30(1.00, 1.70)
LRS = Lower respiratory symptoms
'Effect estimates were standardized to a 40-, 30-, and 20-ppb increase for 1-h max, 8-h max, and 24-h avg 03,
respectively.
6.2.4.2. Summary of Epidemiologic Studies of Respiratory Symptoms and Asthma
Medication Use
With a majority of investigation focused on asthmatic subjects, most studies find that short-
term ambient O3 exposure is positively associated with respiratory symptoms and asthma medication
March 2011
6-69
DRAFT - DO NOT CITE OR QUOTE
-------
1 use. While evidence for effects on school absenteeism is less compelling, there is some evidence that
2 short-term O3 exposure is associated with reduced activity levels among asthmatics. Although
3 studies were limited in number, O3 was positively associated with respiratory symptoms in infants
4 with family history of asthma and office building workers (Apte et al., 2008, 195865; Triche et al.,
5 2006, 093274).
6 Collectively, studies examined associations with single-day O3 concentrations lagged from 0 to
7 5 days as well concentrations averaged over 2 to 19 days. While lag 0 O3 exposures were
8 consistently associated with respiratory symptoms, several studies that examined a range of exposure
9 lags found larger effect estimates for multiday averages (3-day to 6-day) of O3 exposure (Escamilla-
10 Nunez et al., 2008, 594284: Just et al., 2002, 035429: Mortimer et al., 2002, 030281: Rabinovitch et
11 al., 2004, 096753: Romieu et al., 2006, 090969: Ross et al., 2002, 042749). These findings for
12 multiday average of O3 exposure indicate that exposures accumulated over several days may be
13 important or may be subject to less measurement error. In the limited analysis of confounding by
14 co-pollutants, O3 effect estimates showed small changes in magnitude but little change in statistical
15 significance (Escamilla-Nunez et al., 2008, 594284: O'Connor et al., 2008, 156818: Triche et al.,
16 2006. 093274).
17 Several recent studies concurrently examined associations of ambient O3 exposure with
18 respiratory symptoms and lung function and reported conflicting associations whether evaluated at
19 the same or different lag of exposure. Associations were generally stronger for respiratory symptoms
20 than for lung function (Khatri et al., 2009, 594282: Rabinovitch et al., 2004, 096753: Romieu et al.,
21 2006, 090969). These findings suggest that O3-associated respiratory morbidity may occur via
22 multiple mechanisms with varying time courses of action, and the examination of a limited number
23 of exposure lags in these aforementioned studies may explain some of the inconsistencies in
24 associations of O3 exposure with different respiratory health endpoints.
6.2.5. Lung Host Defenses
25 The mammalian respiratory tract has a number of closely integrated defense mechanisms that,
26 when functioning normally, provide protection from the adverse effects of a wide variety of inhaled
27 particles and microbes. For simplicity, these interrelated defenses can be divided into two major
28 parts: (1) nonspecific (transport and phagocytosis) and (2) specific (immunologic) defense
29 mechanisms. A variety of sensitive and reliable methods have been used to assess the effects of O3
30 on these components of the lung's defense system to provide a better understanding of the health
31 effects associated with the inhalation of this pollutant. The previous O3 AQCD (U.S. EPA, 2006,
32 088089) states that animal toxicological studies provide extensive evidence that acute O3 exposures
33 as low as 0.08 to 0.5 ppm can cause increases in susceptibility to infectious diseases due to
34 modulation of lung host defenses. This section discusses the various components of host defenses,
35 such as the mucociliary escalator, the phagocytic and regulatory role of the alveolar macrophages
36 (AMs), the adaptive immune system, and integrated mechanisms that are studied by investigating the
37 host's response to experimental pulmonary infections.
March 2011 6-70 DRAFT - DO NOT CITE OR QUOTE
-------
6.2.5.1. Mucociliary Clearance
1 The mucociliary system is one of the lung's primary defense mechanisms. It protects the
2 conducting airways by trapping and quickly removing material that has been deposited or is being
3 cleared from the alveolar region by migrating alveolar macrophages. Ciliary movement directs
4 particles trapped on the overlying mucous layer toward the pharynx, where the mucus is swallowed
5 or expectorated.
6 The effectiveness of mucociliary clearance can be determined by measuring such biological
7 activities as the rate of transport of deposited particles; the frequency of ciliary beating; structural
8 integrity of the ciliated cells; and the size, number, and distribution of mucus-secreting cells. Once
9 this defense mechanism has been altered, a buildup of both viable and nonviable inhaled substances
10 can occur on the epithelium and may jeopardize the health of the host, depending on the nature of the
11 uncleared substance. Impaired mucociliary clearance can result in an unwanted accumulation of
12 cellular secretions, increased infections, chronic bronchitis, and complications associated with
13 chronic obstructive pulmonary disease. A number of previous studies with various animal species
14 have examined the effect of O3 exposure on mucociliary clearance and reported morphological
15 damage to the cells of the tracheobronchial tree from acute and sub-chronic exposure to 0.20 up to
16 1.0 ppm of O3. The cilia were either completely absent or had become noticeably shorter or blunt.
17 After placing these animals in a clean-air environment, the structurally damaged cilia regenerated
18 and appeared normal (U.S. EPA, 1986, 017607). Based on such morphological observations, related
19 effects such as ciliostasis, increased mucus secretions, and a slowing of mucociliary transport rates
20 might be expected. However, no measurable changes in ciliary beating activity have been reported
21 due to O3 exposure alone. Essentially no data are available on the effects of prolonged exposure to
22 O3 on ciliary functional activity or on mucociliary transport rates measured in the intact animal. In
23 general, functional studies of mucociliary transport have observed a delay in particle clearance soon
24 after acute exposure. Decreased clearance is more evident at higher doses (1 ppm), and there is some
25 evidence of tolerance/adaptation for these effects (U.S. EPA, 1986, 017607). However, no recent
26 studies have evaluated the effects of O3 on mucociliary clearance.
6.2.5.2. Alveolobronchiolar Transport Mechanism
27 In addition to the transport of particles deposited on the mucous surface layer of the
28 conducting airways, particles deposited in the deep lung may be removed either up the respiratory
29 tract or through interstitial pathways to the lymphatic system. The pivotal mechanism of
30 alveolobronchiolar transport involves the movement of alveolar macrophages (AMs) with
31 phagocytized particles to the bottom of the mucociliary escalator. Failure of the AMs to phagocytize
32 and sequester the deposited particles from the vulnerable respiratory membrane can lead to particle
33 entry into the interstitial spaces. Once lodged in the interstitium, particle removal is more difficult
34 and, depending on the toxic or infectious nature of the particle, its interstitial location may allow the
35 particle to set up a focus for pathologic processes. Although some studies show reduced early
March 2011 6-71 DRAFT - DO NOT CITE OR QUOTE
-------
1 (tracheobronchial) clearance after O3 exposure, late (alveolar) clearance of deposited material is
2 accelerated, presumably due to macrophage influx (which in itself can be damaging due to proteases
3 and oxidative reactions in these cells). In an important older study investigating the effects of longer
4 term O3 exposure on alveolobronchiolar clearance, rats were exposed to an urban pattern of O3
5 (continuous 0.06 ppm, 7 days/week with a slow rise to a peak of 0.25 ppm and subsequent decrease
6 to 0.06 ppm over a 9 h period for 5 days/week) for 6 weeks and were exposed 3 days later to
7 chrysotile asbestos, which can cause pulmonary fibrosis and neoplasia (Pinkerton et al., 1989,
8 042102). After 30 days, the lungs of the O3-exposed animals had twice the number and mass of
9 asbestos fibers as the air-exposed rats. New evaluations of O3 effects on alveolar clearance have not
10 been performed.
6.2.5.3. Alveolar Macrophages
11 Within the gaseous exchange region of the lung, the first line of defense against
12 microorganisms and nonviable particles that reach the alveolar surface is the AM. This resident
13 phagocyte is responsible for a variety of activities, including the detoxification and removal of
14 inhaled particles, maintenance of pulmonary sterility, and interaction with lymphocytes for
15 immunologic protection. Under normal conditions, AMs seek out particles deposited on the alveolar
16 surface and ingest them, thereby sequestering the particles from the vulnerable respiratory
17 membrane. To adequately fulfill their defense function, the AMs must maintain active mobility, a
18 high degree of phagocytic activity, and an optimally functioning biochemical and enzyme system. As
19 discussed in previous AQCDs, short periods of O3 exposure can cause a reduction in the number of
20 free AMs available for pulmonary defense, and these AMs are more fragile, less phagocytic, and
21 have decreased lysosomal enzyme activities. In results from earlier work in rabbits, a 2 h exposure to
22 0.1 ppm O3 inhibited phagocytosis and a 3 h exposure to 0.25 ppm decreased lysosomal enzyme
23 activities (Driscoll et al., 1987, 040803: Hurst et al., 1970, 015591). Effects on in vitro viability of
24 AM were observed at 0.06 ppm (Weissbecker et al., 1969, 015806). A few recent studies have
25 evaluated ozone's effects on macrophage function. At 1 and 24 h after a 4 h exposure of marine toads
26 (Bufo marinus) to 0.8 ppm O3, macrophages exhibited reduced in vitro capacity to phagocytize
27 fluorescent polystyrene microspheres. However, this effect did not persist at 48 h postexposure and
28 exposure did not affect yields of pulmonary macrophages (Dohm et al., 2005, 180452). In another
29 study, in vitro exposure to 0.03 ppm O3 for five minutes significantly decreased macrophage-like cell
30 mobility in response to pathogen-related chemotactic stimulation (Klestadt et al., 2005, 130425).
31 Additionally, O3 mediated oxidation of surfactant proteins reduced their ability to enhance
32 phagocytosis of both gram-positive and gram-negative bacteria by macrophages (Mikerov et al.,
33 2008, 596405). A single controlled human exposure study reviewed in the 1996 O3 AQCD found
34 decrements in the ability of alveolar macrophages to phagocytize microorganisms upon exposure to
35 0.08 to 0.1 ppm O3 for 6.6 h during moderate exercise (Devlin et al., 1991, 040359).
36 Collectively, these studies demonstrate that O3 can affect multiple steps or aspects required for
37 proper macrophage function.
March 2011 6-72 DRAFT - DO NOT CITE OR QUOTE
-------
6.2.5.4. Infection and Adaptive Immunity
General Effects on the Immune System
1 The effects of O3 on the immune system are complex and dependent on the exposure regimen
2 and the observation period. It appears that the T-cell-dependent functions of the immune system are
3 more affected than B-cell-dependent functions (U.S. EPA, 2006, 088089). Generally, there is an
4 early immunosuppressive effect that subsides with continued O3 exposure, resulting in either a return
5 to normal responses or an enhancement of immune responses. However, this is not always the case
6 as Aranyi (1983, 040512) showed decreased T-cell mitogen reactions in mice after chronic (90-day)
7 exposure to 0.1 ppm O3. Earlier studies report changes in cell populations in lymphatic tissues
8 (U.S. EPA, 2006, 088089). A more recent study in mice demonstrated that numbers of certain T cell
9 subsets in the spleen were reduced after exposure to 0.6 ppm O3 (lOh/day x 15d) (Feng et al, 2006,
10 596381).
11 O3 has also been found to alter responses to antigenic stimulation. For example, antibody
12 responses to a T-cell-dependent antigen were suppressed after a 56-day exposure of mice to 0.8 ppm
13 O3, and a 14-day exposure to 0.5 ppm O3 decreased the antiviral antibody response following
14 influenza virus infection (Jakab and Hmieleski, 1988, 041806); the latter impairment may pave the
15 way for lowered resistance to reinfection. The immune response is highly influenced by the temporal
16 relationship between O3 exposure and antigenic stimulation. When O3 exposure preceded Listeria
17 infection, there were no effects on delayed-type hypersensitivity or splenic lymphoproliferative
18 responses; however, when O3 exposure occurred during or after Listeria infection was initiated, these
19 immune responses were suppressed (van Loveren et al., 1988, 041847). In another study, a reduction
20 in mitogen activated T-cell proliferation was observed after exposure to 0.6 ppm for 15 d, and could
21 be ameliorated by antioxidant supplementation. Antigen-specific proliferation decreased by 60%,
22 indicating attenuation of the acquired immunity needed for subsequent memory responses (Feng et
23 al., 2006, 596381). Generally, continuous exposure to O3 impairs immune responses for the first
24 several days of exposure, followed by an adaptation to O3 that allows a return of normal immune
25 responses. Most species show little effect of O3 exposures prior to immunization, but show a
26 suppression of responses to antigen in O3 exposures post-immunization. In a recent study, exposure
27 of mice to 0.6 ppm O3 skewed the ex-vivo cytokine responses elicited by non-specific stimulation
28 toward inflammation, decreasing IL-2 and increasing IFN-y (Feng et al., 2006, 596381).
Models of Microbial Infection
Bacterial infection
29 A relatively large body of evidence shows that O3 increases susceptibility to bacterial
30 infections. Known contributing factors are impaired mucociliary streaming, altered
31 chemotaxis/motility, defective phagocytosis of bacteria, decreased production of lysosomal enzymes
32 or superoxide radicals by alveolar macrophages, and decreased IFN-y levels. In animal models of
March 2011 6-73 DRAFT - DO NOT CITE OR QUOTE
-------
1 bacterial infection, exposure to 0.08 ppm O3 increases mortality, regardless of whether O3 exposure
2 precedes or follows infection. Exercise and co-pollutants can enhance ozone's effects in infectivity
3 models. Recent studies in mice continue to demonstrate increased susceptibility to experimental
4 infectious pneumonia with exposure to O3, albeit at high levels (2 ppm) (Mikerov et al., 2008,
5 597493: Mikerov et al., 2008, 201537).
6
Viral infection
7 Only a few studies, described in previous AQCDs, have examined the effects of O3 exposure
8 on the outcome of viral respiratory infection. Some studies show increased mortality, while others
9 show diminished severity and increased survival time. In vitro cell culture studies of human
10 bronchial epithelial cells indicate O3-induced exacerbation of human rhinovirus infection
11 (Spannhake et al., 2002, 030637). New studies on the interactions of O3 and viral infections have not
12 been published. Natural killer (NK) cells, which destroy virally infected cells and tumors in the lung,
13 appear to be inhibited by higher doses of O3 and either unaffected or stimulated at lower doses.
14 Several studies show decreases in NK cell activity following acute exposures ranging from 0.8 to
15 1 ppm (Burleson et al., 1989, 042214: Gilmour and Jakab, 1991, 042391: Van Loveren et al., 1990,
16 042264). However, Van Loveren et al. (1990, 042264) showed that a 1-week exposure to 0.2 or
17 0.4 ppm O3 increased NK cell activity, and an urban pattern of exposure (base of 0.06 ppm with
18 peaks of 0.25 ppm) had no effect on NK cell activity after 1, 3, 13, 52, or 78 weeks of exposure
19 (Selgrade et al., 1990, 042363). A more recent study demonstrated a 35% reduction in NK cell
20 activity after exposure of mice to 0.6 ppm O3 (lOh/day x 15d) (Feng et al., 2006, 596381). The
21 defective IL-2 production demonstrated in this study may impair NK cell activation. Alternatively,
22 NK cell surface charge may be altered by ROS, decreasing their adherence to target cells (Nakamura
23 and Matsunaga. 1998. 625133).
6.2.6. Allergic and Asthma-Related Responses
24 Effects resulting from combined exposures to O3 and allergens have been studied in a variety
25 of animal species, generally as models of experimental asthma. Pulmonary function and airways
26 hyperresponsiveness in animal models of asthma are discussed in Sections 6.2.1.7 and 6.2.2.2.
27 Previous evidence indicates that O3 exposure skews immune responses toward an allergic phenotype.
28 For example, Gershwin et al. (1981, 039729) reported that O3 (0.8 and 0.5 ppm for 4 days) exposure
29 caused a 34-fold increase in the number of IgE (allergic antibody)-containing cells in the lungs of
30 mice. In general, the number of IgE-containing cells correlated positively with levels of anaphylactic
31 sensitivity. In humans, allergic rhinoconjunctivitis symptoms are associated with increases in
32 ambient O3 concentrations (Riediker et al., 2001, 051776). Five weeks of continuous exposure to
33 0.4 ppm O3 (but not 0.1 or 0.2 ppm O3) augmented sneezing and nasal secretions in a guinea pig
34 model of nasal allergy. Nasal eosinophils and allergic antibody levels in serum were also elevated by
35 exposure to concentrations as low as 0.2 ppm (lijima and Kobayashi, 2004, 596389). Short-term
36 exposure (2 days) to 1 ppm O3 exacerbated allergic rhinitis and lower airway allergic inflammation
March 2011 6-74 DRAFT - DO NOT CITE OR QUOTE
-------
1 in Brown Norway rats, a rat strain that is comparatively less sensitive to O3 than other rats or humans
2 (Wagner et al, 2007, 596420; Wagner et al, 2009, 201574). OVA-sensitized rats were intranasally
3 challenged with OVA on days 1 and 2, and exposed to 0 or 1 ppm O3 (8 h/day) on days 4 and 5.
4 Analysis at day 6 indicated that O3 exposure enhanced intraepithelial mucosubstances in the nose
5 and airways, induced cys-LTs, MCP-1, and IL-6 production in BALF, and upregulated expression of
6 the proallergic cytokines IL-5 and IL-13. These changes were not evident in non-allergic controls.
7 All of these responses were blunted by gamma-tocopherol (yT; vitamin E) therapy. yT neutralizes
8 oxidized lipid radicals, and protects lipids and proteins from nitrosative damage from NO-derived
9 metabolites. Farraj et al. (2010, 380846) exposed allergen-sensitized adult male BALB/c mice to
10 0.5 ppm O3 for 5 hours once per week for 4 weeks. Ozone exposure and O3/DEP (2.0 mg/m3) co-
11 exposure of OVA-sensitized mice elicited significantly greater serum IgE levels than in DEP-
12 exposed OVA-sensitized mice (98% and 89% increases, respectively). Ozone slightly enhanced
13 levels of BAL IL-5, but despite increases in IgE, caused a significant decrease in BAL IL-4 levels.
14 IL-10, IL-13, and IFN-y levels were unaffected. In addition to ozone's pro-allergic effects, it could
15 also make airborne allergens more allergenic. When combined with NO2, O3 has been shown to
16 enhance nitration of common protein allergens, which may increase their allergenicity (Franze et al.,
17 2005. 066088).
6.2.7. Hospital Admissions, Emergency Department Visits, and Physicians
Vis its
6.2.7.1. Summary of Findings from 2006 Ozone AQCD
18 The 2006 O3 AQCD (U.S. EPA, 2006, 088089) evaluated numerous respiratory ED visits and
19 hospital admissions studies, which consisted primarily of time-series studies conducted in the U.S.,
20 Canada, Europe, South America, Australia and Asia. Upon collectively evaluating the scientific
21 evidence, the 2006 O3 AQCD (U.S. EPA, 2006, 088089) concluded that "the overall evidence
22 supports a causal relationship between acute ambient O3 exposures and increased respiratory
23 morbidity resulting in increased ED visits and [hospital admissions] during the warm season"
24 (U.S. EPA, 2006, 088089). This conclusion is "strongly supported by the human clinical, animal
25 toxicologicfal], and epidemiologic evidence for [O3-induced] lung function decrements, increased
26 respiratory symptoms, airway inflammation, and airway hyperreactivity" (U.S. EPA, 2006, 088089).
27 Since the completion of the 2006 O3 AQCD (U.S. EPA, 2006, 088089). relatively fewer
28 studies conducted in the U.S., Canada, and Europe have examined the association between short-
29 term exposure to ambient O3 and respiratory hospital admissions and ED visits, but a growing
30 number of studies have been conducted in Asia. This section focuses primarily on multicity studies
31 because they examine the effect of O3 on respiratory-related hospital admissions and ED visits over a
32 large geographic area using a consistent statistical methodology. Single-city studies that encompass a
33 large number of hospital admissions or ED visits, or included a long study-duration were also
March 2011 6-75 DRAFT - DO NOT CITE OR QUOTE
-------
1 evaluated because these studies have more power to detect whether an association exists between
2 short-term exposure to O3 and respiratory hospital admissions and ED visits compared to smaller
3 single-city studies. Additional single-city studies were also evaluated within this section, if they were
4 conducted in locations not represented by the larger single-city and multicity studies, or examined
5 population-specific characteristics not included in the larger studies that may modify the association
6 between short-term exposure to O3 and respiratory-related hospital admissions or ED visits. The
7 remaining single-city studies identified were not evaluated in this section due to factors such as
8 inadequate study design or insufficient sample size.
9 It should be mentioned that when examining the association between short-term O3 exposure
10 and respiratory health effects that require medical attention, it is important to distinguish between
11 hospital admissions and ED visits. This is because it is likely that a small percentage of respiratory
12 ED visits will be admitted to the hospital; therefore, respiratory ED visits may represent potentially
13 less serious, but more common outcomes. As a result, in the following sections respiratory hospital
14 admission and ED visit studies are evaluated individually. Additionally, within each section, results
15 are presented as either a collection of respiratory diagnoses or as individual diseases (e.g., asthma,
16 COPD, pneumonia and other respiratory infections) in order to evaluate the potential effect of short-
17 term O3 exposure on each respiratory-related outcome. Table 6-19 presents the studies discussed
18 within this section along with the air quality characteristics of the city, or across all cities, included in
19 each study.
March 2011 6-76 DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-19. Mean and upper percentile concentrations of respiratory-related hospital admission and
emergency department visit studies evaluated
Study
Katsouyannietal.
(2009, 1998991"'°
Cakmak et al. (2006,
0932721
Biggeri et al. (2005,
0873951°
Dales et al. (2006,
0907441
Lin etal. (2008,
1958561
Wong et al. (2009,
1967221°
Medina-Ramon etal.
(2006, 087721V
Yang et al. (2005,
0901 841k
Zanobetti and
Schwartz (2006,
0901951
Silverman and Ito
(2010,3862521
Stieb et al. (2009,
1958581
Tolbert et al. (2007,
0903 16V
Darrowetal. (2011,
2028001
Ito et al. (2007,
1565941
Strickland etal. (2010,
6248781
Arbex et al. (2009,
1843341
Location
90 U.S. cities
(NMMAPS)"
32 European cities
(APHEA)d
12 Canadian cities
10 Canadian cities
4 Italian cities'
11 Canadian cities
11 New York
regions
Hong Kong
36 U.S. cities
Vancouver,
Canada
Boston, MA
New York, NY
7 Canadian cities
Atlanta, GA
Atlanta, GA
New York, NY
Atlanta, GA
Sao Paulo, Brazil
Type of Visit (ICD9/10)
Hospital Admissions:
NM MAPS: All respiratory (460-51 9)
APHEA: All respiratory (460-51 9)
1 2 Canadian cities: All respiratory (460-51 9)e
Hospital Admissions:
All respiratory (466, 480-486, 490, 491, 492,
493, 494, 496)
Hospital Admissions:
All respiratory (460-51 9)
Hospital Admissions:
Respiratory disorders (486, 768.9, 769, 770.8,
786,799.0,799.1)
Hospital Admissions:
Respiratory diseases (466, 490-493, 496)
Hospital Admissions:
All respiratory (460-51 9)
COPD (490-496)
Acute respiratory diseases (460-466, 480-487)
Hospital Admissions:
COPD (490-496, excluding 493)
Pneumonia (480-487)
Hospital Admissions:
COPD (490-492, 494, 496)
Hospital Admissions:
Pneumonia (480-487)
Hospital Admissions:
Asthma (493)
Emergency Department Visits:
Asthma (493)
COPD (490-492, 494-496)
Respiratory infection (464, 466, 480-487)
Emergency Department Visits:
All respiratory (460-465, 460.0, 466.1, 466.11,
466.19, 477, 480-486, 491, 492, 493, 496,
786.07, 786.09)
Emergency Department Visits:
All respiratory (460-466, 477, 480-486, 491,
492, 493, 496, 786.09)
Emergency Department Visits:
Asthma (493)
Emergency Department Visits:
Asthma (493)
Wheeze (786.07, 786.09)
Emergency Department Visits:
COPD (J40-44)
Metric
1-h max
24-h avg
8-h max
24-h avg
8-h max1
8-h max1
8-h max
24-h avg
24-h avg
8-h max
24-h avg
8-h max
8-h max
1-h max
24-h avg
Commute
Day-time
Night-time
8-h max
8-h max
1-h max
Mean
Concentration (ppbf
NMMAPS:
50th: 34.9-60.0
APHEA:
50th: 11.0-38.1
12 Canadian cities:
50th: 6.7-8.3
17.4
Warm season9: 5.7-60.0
17.0
44.1
18.8
Warm season9: 45.8
Cool season: 27.6
All year: 14.1
Winter: 13.2
Spring: 19.4
Summer: 13.8
Fall: 10.0
22.4°
Warm':41.0b
18.4
53.0
8-h max: 53
1-h max: 62
24-h avg: 30
Commute: 35"
Day-time: 45"
Night-time: 14"
All year: 30.4
Warm months": 42.7
Cold months: 18.0
All year: 45.4°
Warm": 55.2°
Cold": 34.5°
48.8
Range of
Concentrations (ppbf
NMMAPS:
75th: 46.8-68.8
APHEA:
75th: 15.3-49.4
12 Canadian cities:
75th: 8.9-1 2.4
Max: 38.0-79.0
95th: 86.1-90.0"
Max: 107.5-115.1
95th: 24.9-46.0
75th: 54.0
Max: 217.0
75th: 25.9
Max: 100.3
NR
Max: 38.6
75th: 31.0
95th: 47.6
75th: 53
90th: 68
75th: 19.3-28.6
75th: 67.0
90th: 82.1
Max: 147.5
8-h
max:
75th:
67 24-h avg:
Max: 75th: 37
148 Max: 81
1-h Commute:
max: 75th: 45
75th: Max: 106
76
Max:
180
All year:
95th: 68.0
Warm months:
95th: 77.0
Cold months:
95th: 33.0
NR
75th: 61.0
Max: 143.8
Day-
time:
75th: 58
Max:
123
Night-
time:
75th: 22
Max: 64
March 2011
6-77
DRAFT - DO NOT CITE OR QUOTE
-------
Study Location
202801? aL(2°°9' 6 Italian cities
195868) 3L (2°°9' Toronto, Canada
VHIen.euveetal.(2006, Toront0] Canada
Sinclair etal. (2010, Atl, . „.
•3oci7<\s Atlanta, GA
joo/n )
Type of Visit (ICD9/10)
Emergency Department Visits:
Wheezing
Physician Visits:
Asthma (493)
Physician Visits:
Allergic rhinitis (177)
Physician Visits:
Asthma
Upper respiratory infection
Lower respiratory infection
.... Mean
lvletnc Concentration (ppb)a
Bh r Summer": 21.1-44.3
H-nmax winter: 11.5-27.9
1-hmax 33.3
8-h max 30.0
Total Study Period:
All-year: 44.0
25 mo Period:
All-year: 47.9
„, ,v Warm: 61. 2
o-h max ^..M. 0-7 Q
Cold. 27.8
28 mo Period:
All-year: 40.7
Warm: 51. 8
Cold: 26.0
Range of
Concentrations (ppb)8
NR
95th: 66
Max: 121
Max: 98.7
NR
aSome studies did not present an overall value for the mean, middle and/or upper percentiles of the 03 distribution; as a result, the range of the mean, middle, and/or
upper percentiles across all of the cities included in the study are presented.
bStudy only presented median concentrations.
"Study presented concentrations as pg/m3Concentration was converted to ppb using the conversion factor of 0.51 assuming standard temperature (25°C) and
pressure (1 atm).
A subset of the European and U.S. cities included in the mortality analyses were used in the hospital admissions analyses: 8 of the 32 European cities and 14 of 90
U.S. cities.
"Hospital admission data was coded using three classifications (ICD-10-CA, ICD-9, and ICD-9-CM). Attempts were made by the original investigators to convert
diagnosis from ICD-10-CA back to ICD-9.
'Only 4 of the 8 cities included in the study collected 03 data.
'Warm season defined as May-September.
h95th percentile only presented for 3 of the 5 cities with 03data.
'03 measured from 10:00 a.m. to 6:00 p.m.
'Only 35 of the 36 cities included in the analysis had 03 data.
kStudy defined seasons as Winter (January-March), Spring (April-June); Summer (July-September), and Fall (October-December).
'Study only examined warm months (April-August).
"Study only examined warm months (April-October).
"Commute (7:00 a.m. to 10:00 a.m., 4:00 p.m.
p.m. to 7:00 p.m.); Day-time (8:00 a.m. to 7:00 p.m.); Night-time (12:00 a.m. to 6:00 a.m.).
"Warm/Summer months defined as April-September.
pMeans represent population-weighted 03 concentrations.
"VVarm months defined as May-October and Cold months defined as November-April.
r03 measured from 8:00 a.m. to 4:00 p.m.
sThis study did not report the ICD codes used for the conditions examined. The 25-month period represents August 1998-August2000, and the 28-month period
represents September 2000-December 2002. This study defined the warm months as April - October and the cold months as November-March.
6.2.7.2. Hospital Admission Studies
Respiratory Diseases
1 The association between exposure to an air pollutant, such as O3, and daily respiratory-related
2 hospital admissions has primarily been examined using all respiratory-related hospital admissions
3 within the range of ICD-9 codes 460-519. Newly identified studies attempt to further examine the
4 effect of O3 exposure on respiratory-related hospital admissions through a multicity design that
5 examines O3 effects across countries using a standardized methodology; multicity studies that
6 examine effects within one country; and multi- and single-city studies that attempt to examine
7 potential modifiers of the O3-respiratory-related hospital admission relationship.
8 The Air Pollution and Health: A European and North American Approach (APHENA) study
9 combined data from existing multicity study databases from Canada, Europe (APHEA2)
10 (Katsouyanni et al., 2001, 019008). and the U.S. (NMMAPS) (Samet et al., 2000, 010269) in order
11 to "develop more reliable estimates of the potential acute effects of air pollution on human health
12 [and] provide a common basis for [the] comparison of risks across geographic areas" (Katsouyanni
March 2011
6-78
DRAFT - DO NOT CITE OR QUOTE
-------
1 et al, 2009, 199899). In an attempt to address both of these issues, the investigators conducted
2 extensive sensitivity analyses to evaluate the robustness of the results to different model
3 specifications (e.g., penalized splines [PS] versus natural splines [NS]) and the extent of smoothing
4 to control for seasonal and temporal trends. The trend analyses consisted of subjecting the models to
5 varying extent of smoothing selected either a priori (e.g., 3 df/year, 8 df/year, and 12 df/year) or by
6 using the absolute sum of the residuals of the partial autocorrelation function (PACF). However, the
7 investigators did not identify the model they deemed to be the most appropriate for comparing the
8 results across study locations. As a result, when discussing the results across the three study locations
9 below, the 8 df/year results are presented for both the PS and NS models because: (1) 8 df/year is
10 most consistent with the extent of temporal adjustment used in previous and recent large multicity
11 studies in the U.S. (e.g., NMMAPS); (2) the risk estimates for 8 df/year and 12 df/year are
12 comparable for all three locations; (3) the models that used the PACF method did not report the
13 actual degrees of freedom chosen; and (4) the 3 df/year and the PACF method resulted in negative O3
14 risk estimates, which is inconsistent with the results obtained using more aggressive seasonal
15 adjustments. Additionally, when comparing results across studies in figures, only the results from
16 one of the spline models (e.g., NS) is presented because it has been previously demonstrated that
17 alternative spline models result in relatively similar effect estimates (Health Effects Institute, 2003,
18 042829). However, it should be noted that the underlying data and model specifications could result
19 in varying degrees of bias and precision in effect estimates with different spline models (Ostro et al.,
20 2006, 087991).
21 Katsouyanni et al. (2009, 199899) examined respiratory hospital admissions (ICD-9: 460-519)
22 for people aged 65 years and older using 1-h max O3 data. The extent of hospital admission and O3
23 data varied across the 3 datasets: Canadian dataset included 12 cities with data for 3 years
24 (1993-1996) per city; European dataset included 8 cities with each city having data for between 2
25 and 8 years from 1988-1997; and U.S. dataset included 14 cities with each city having data for
26 between 4 and 10 years from 1985-1994 and 7 cities having only summer O3 data. The investigators
27 used a three-stage hierarchical model to account for within-city, within region, and between region
28 variability. Results were presented individually for each region (Figure 6-15; Table 6-20). Ozone and
29 PMio concentrations were weakly correlated in all locations in the summer (r=0.27-0.40), but not in
30 the winter. In the Canadian cities, using all-year data, a 40-ppb increase in 1-h max O3
31 concentrations at lag 0-1 was associated with an increase in respiratory hospital admissions of 8.9%
32 (95% CI: 0.79, 16.8%) in a PS model and 8.1% (95% CI: 0.24, 16.8%) in aNS model. The results
33 were somewhat sensitive to the lag day selected, reduced when using a single-day lag (e.g., lag 1)
34 (PS: 6.0%; NS: 5.5%) and increased when using a distributed lag model (PS: 18.6%; NS: 20.4%).
35 When adjusting for PMio, the magnitude of the effect estimate was slightly larger in the NS model
36 (5.1% [95% CI: -6.6, 18.6%]) compared to the PS model (3.1% [95% CI: -8.3, 15.9%]); however,
37 the co-pollutant analysis was only conducted using a 1-day lag. The large confidence intervals for
38 both models could be attributed to the reduction in days included in the co-pollutant analyses as a
39 result of the every-6th-day PM sampling schedule. When restricting the analysis to the summer
March 2011 6-79 DRAFT - DO NOT CITE OR QUOTE
-------
1 months, stronger associations were observed between O3 and respiratory hospital admissions across
2 the lags examined, ranging from -22 to 37% (the study does not specify whether these effect
3 estimates are from a NS or PS model). Because O3 concentrations across the cities included in the
4 Canadian dataset (Katsouyanni et al. (2009, 199899) are low (median concentrations ranging from
5 6.7-8.3 ppb [Table 6-19]), the standardized increment of 40 ppb for a 1-h max increase in O3
6 concentrations does not accurately reflect the observed risk of O3-related respiratory hospital
7 admissions. Although this increment accurately characterizes the distribution of 1-h max O3
8 concentrations across the U.S. and European datasets, it misrepresents the observed O3
9 concentrations in the Canadian dataset. As a result in summary figures, for comparability, effect
10 estimates from the Canadian dataset are presented for both a 5.1-ppb increase in 1-h max O3
11 concentrations (i.e., an approximate interquartile range [IQR] increase in O3 concentrations across
12 the Canadian cities) as well as the standardized increment used throughout the ISA.
13 In Europe, weaker but positive associations were also observed in year round analyses; 2.9%
14 (95% CI: 0.63, 5.0%) in the PS model and 1.6% (95% CI: -1.7, 4.2%) in the NS model at lag 0-1 for
15 a 40-ppb increase in 1-h max O3 concentrations. Additionally, at lag 1, associations between O3 and
16 respiratory hospital admissions were also reduced, but in contrast to the lag 0-1 analysis, greater
17 effects were observed in the NS model (2.9% [95% CI: 1.0, 4.9%]) compared to the PS model (1.5%
18 [95% CI: -2.2, 5.4]). Unlike the Canadian analysis, a distributed lag model provided limited evidence
19 of an association between O3 and respiratory hospital admissions. To compare with the Canadian
20 results, when adjusting for PMi0 at lag 1, effect estimates were increased in the PS model (2.5%
21 [95% CI: 0.39-4.8%]) and remained robust in the NS model (2.4% [95% CI: 0.08, 4.6%]). However,
22 the European analysis also examined the effect of adjusting for PMi0 at lag 0-1 and found results
23 were attenuated in both models (PS: 0.8% [95% CI: -2.3, 4.0%]; NS: 0.8% [95% CI: -1.8, 3.6%]).
24 Unlike the Canadian and U.S. datasets, the European dataset consisted of daily PM data. The
25 investigators did not observe stronger associations in the summer-only analyses for the European
26 cities at lag 0-1 (PS: 0.4% [95% CI: -3.2, 4.0%]; NS: 0.2% [95% CI: -3.3, 3.9%]), but did observe
27 some evidence for larger effects during the summer, an -2.5% increase, at lag 1 in both models (the
28 study does not present the extent of temporal smoothing used for these models).
29 For the U.S. in year round analyses, the investigators reported a 1.4% (95% CI: -0.9, 3.9%)
30 increase in the PS model and 2.4% (95% CI: 0.0, 4.9%) increase in the NS model in respiratory
31 hospital admissions at lag 0-1 for a 40-ppb increase in 1-h max O3 concentrations with similar results
32 for both models at lag 1. The distributed lag model provided results similar to those observed in the
33 European dataset with the PS model (1.1% [95% CI: -3.0, 5.3%]), but larger effects in the NS model
34 (3.3% [95% CI: 0.02, 6.8%]), which is consistent with the Canadian results. When adjusting for
35 PM10 using the U.S. data (i.e., every-6th-day PM data), results were attenuated at lag 0-1 (PS: 0.6%
36 [95% CI: -2.0, 3.3%]; NS: 1.4% [95% CI: -1.3, 4.2%]) which is consistent with the results presented
37 for the European dataset. However, at lag 1, U.S. risk estimates remained robust to the inclusion of
38 PMio in co-pollutant models as was observed in the Canadian and European datasets. Compared to
39 the all-year analyses, the investigators did not observe stronger associations in the summer-only
March 2011 6-80 DRAFT - DO NOT CITE OR QUOTE
-------
1 analysis at either lag 0-1 (-2.2%) or lag 1 (-2.8%) in both the PS and NS models (the study does not
2 present the extent of temporal smoothing used for these models).
Location
U.S.
Canada
Europe
•_«(,
1
1
0-1
0-1 — '
DL(0-2)
0-1
1
1
j.
la
1
la
0-1
0-la
DL(0-2)
DL(0-2)a
1
_L
la
01
-J.
0-la
DL(0-2)
DL(0-2)a
1
1
0-1 —
0-1
1
J. H
n 1 L
• All-Year
— O —
— • —
-O
0
— • Summer
•
• All Yi — ir
+-
O—
-•-
• ^
— • —
A_
• ^
• ^
^
— • —
— •— All-Year
— O —
-• —
•O
0
»
1
-10
10 15 20 25 30 35 40
% Increase
Figure 6-15. Percent increase in respiratory hospital admissions from natural spline models for
a 40-ppb increase in 1-h max ozone concentrations for each location of the APHENA
study. D
Black circles = all-year results; open circles = all-year results in co-pollutant model with PMi0;
and red circles = summer only results. For Canada, lag days with an "a" next to them represent
the risk estimates standardized to an approximate IQR of 5.1 ppb for a 1-h max increase in
ozone concentrations.
March 2011
6-81
DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-20. Corresponding effect estimates for Figure 6-15
Location Season Lag3 Co-pollutant
% Increase (95% Cl)b
U.S.
All-year 1
1 PM10
2.62 (0.63, 4.64)
2. 14 (-0.08, 4.40)
0-1 2.38 (0.00, 4.89)
0-1 PM10
DL(0-2)
Summer 0-1
1.42 (-1.33, 4.23)
3.34 (0.02-6.78)
2. 14 (-0.63, 4.97)
1 2.78 (-0.02, 5.71)
Canada All-year 1
1a
1 PM10
1a PM10
0-1
0-1 a
DL(0-2)
DL(0-2)a
Summer 1
1a
5.54 (-0.94, 12.4)
0.69 (-0.1 2, 1.50)a
5.13 (-6.62, 18.6)
0.64 (-0.87, 2.20)a
8.12(0.24, 16.8)
1.00(0.03,2.00)3
20.4 (4.07, 40.2)
2.4(0.51,4.40)3
21.4(15.0,29.0)
2.50(1.80,3.30)3
0-1 32.0(18.6,47.7)
0-1 a
DL(0-2)
DL(0-2)a
Europe All-year 1
1 PM10
0-1
0-1 PM10
DL(0-2)
Summer 1
0-1
3.60(2.20,5.10)3
37.1 (11.5,67.5)
4.1 (1.40,6.80)3
2.94(1.02,4.89)
2.38 (0.08, 4.64)
1.58 (-1.71, 4.15)
0.87 (-1.79, 3.58)
0.79 (-4.46, 6.37)
2.46 (-0.63, 5.54)
0.24 (-3.32, 3.91)
aFor Canada, lag days with an "a" next to them represent the risk estimates standardized to an approximate IQR of 5.1 ppbfor a 1-h max increase in 03
concentrations.
bUnless noted, risk estimates standardized to 40 ppbfor a 1-h max increase in 03 concentrations.
1 Several additional multicity studies examined respiratory disease hospital admissions in
2 Canada and Europe. Cakmak et al. (2006, 093272) evaluated the association between ambient O3
3 concentrations and respiratory hospital admissions for all ages in 10 Canadian cities from April 1993
4 to March 2000. The primary objective of this study was to examine the potential modification of the
5 effect of ambient air pollution on daily respiratory hospital admissions (defined as acute bronchitis
6 and bronchiolitis [ICD-9: 466], pneumonia [480-486], bronchitis [490, 491], emphysema [492],
7 asthma [493], bronchiectasis [494], and COPD [496]) by education and income using a time-series
8 analysis conducted at the city-level. The authors calculated a pooled estimate across cities for each
9 pollutant using a random effects model by first selecting the lag day with the strongest association
10 from the city-specific models. For O3, the mean lag day across cities that provided the strongest
11 association and for which the pooled effect estimate was calculated was 1.2 days. In this study,
12 all-year O3 concentrations were used in the analysis, and additional seasonal analyses were not
March 2011
6-82
DRAFT - DO NOT CITE OR QUOTE
-------
1 conducted. Cakmak et al. (2006, 093272) reported a 4.4% increase (95% CI: 2.2, 6.5%) in
2 respiratory hospital admissions for a 20 ppb increase in 24 h average O3 concentrations. The
3 investigators only examined the potential effect of confounding by other pollutants through the use
4 of a multi-pollutant model, which is difficult to interpret. Cakmak et al. (2006, 093272) also
5 conducted an extensive analysis of potential modifiers, specifically gender, educational attainment,
6 and family income, on the association between air pollution and respiratory hospital admissions.
7 When stratifying by gender, the increase in respiratory hospital admissions due to short-term O3
8 exposure were similar in males (5.2% [95% CI: 3.0, 7.3%]) and females (4.2% [95% CI: 1.8, 6.6%]).
9 In addition, the examination of effect modification by income found no consistent trend across the
10 quartiles of family income. However, there was evidence that individuals with an education level
11 less than the 9th grade were disproportionately affected by O3 exposure (4.6% [95% CI: 1.8, 7.5%])
12 compared to individuals that completed grades 9-13 (1.7% [95% CI: -1.9, 5.3%]), some university or
13 trade school (1.4% [95% CI: -2.0, 5.1%]), or have a university diploma (0.66% [95% CI: -3.3,
14 4.7%]). The association between O3 and individuals with an education level less than the 9th grade
15 was the strongest association across all of the pollutants examined.
16 A multicity study conducted in Europe by Biggeri et al. (2005, 087395) examined the
17 association between short-term O3 exposure and respiratory hospital admissions (ICD-9: 460-519)
18 for all ages in four Italian cities from 1990 to 1999. In this study, O3 was only measured during the
19 warm season (May-September). The authors examined associations between daily respiratory
20 hospital admissions and short-term O3 exposure at the city-level using a time-series analysis. Pooled
21 estimates were calculated by combining city-specific estimates using fixed and random effects
22 models. The investigators found no evidence of an association between O3 exposure and respiratory
23 hospital admissions in the warm season in both the random (0.1% [95% CI: -5.2, 5.7%]; distributed
24 lag 0-3) and fixed effects (0.1% [95% CI: -5.2, 5.7%]; distributed lag 0-3) models for a 30-ppb
25 increase in 8-h max O3 concentrations.
26 In another multicity study conducted in Canada, Dales et al. (2006, 090744) examined the
27 association between all-year ambient O3 concentrations and neonatal (ages 0-27 days) respiratory
28 hospital admissions (defined as respiratory asphyxia [ICD-9: 799], respiratory failure [769],
29 unspecified birth asphyxia [768.9], other respiratory problems after birth [770.8], and pneumonia
30 [486]) in 11 Canadian cities from 1986 to 2000. The investigators used a statistical analysis approach
31 similar to Cakmak et al. (2006, 093272) (i.e., time-series analysis to examine city-specific
32 associations, and then a random effects model to pool estimates across cities). The authors reported
33 that for O3, the mean lag day across cities that provided the strongest association was 2 days. The
34 authors reported a 5.4% (95% CI: 2.9, 8.0%) increase in neonatal respiratory hospital admissions for
35 a 20-ppb increase in 24-h avg O3 concentrations at lag-2 days. The results from Dales et al. (2006,
36 090744) provide support for the associations observed in a smaller scale study that examined O3
37 exposure and pediatric respiratory hospital admissions in New York state (Lin et al., 2008, 195856).
38 Lin et al. (2008, 195856) observed a positive association between O3 and pediatric (i.e., <18 years)
39 respiratory admissions (ICD-9: 466, 490-493, 496) at lag 2 (results not presented quantitatively) in a
March 2011 6-83 DRAFT - DO NOT CITE OR QUOTE
-------
1 two-stage Bayesian hierarchical model analysis of 11 geographic regions of New York from 1991 to
2 2001.
3 Wong et al. (2009, 196722) examined the potential modification of the relationship between
4 ambient O3 (along with NO2, SO2, and PMi0) and respiratory hospital admissions (ICD-9: 460-519;
5 ICD-10: J40-J47) by influenza intensity in Hong Kong for the period 1996 - 2002. Influenza
6 intensity was defined using the proportion of weekly specimens positive for influenza A or B. In
7 models that examined the baseline effect (i.e., without taking into consideration influenza intensity)
8 of short-term O3 exposure, the authors found a 3.6% (95% CI: 1.9, 5.3%) and 3.2% (95% CI: 1.0,
9 5.4%) increase in respiratory hospital admissions at lag 0-1 for a 30-ppb increase in 8-h max O3
10 concentrations for the all age ariS 65 age groups, respectively. When examining influenza intensity,
11 Wong et al. (2009, 196722) reported that the association between short-term exposure to O3 and
12 respiratory hospital admissions was stronger with higher levels of influenza intensity: additional
13 increase in respiratory hospital admissions above baseline of 1.4% (95% CI: 0.24, 2.6%) for all age
14 groups and 2.4% (95% CI: 0.94, 3.8%) for those 65 and older when influenza activity increased from
15 0% to 10%. No difference in effects was observed when stratifying by sex.
Cause-Specific Respiratory Outcomes
16 In the 2006 O3 AQCD a limited number of studies were identified that examined the effect of
17 short-term O3 exposure on cause-specific respiratory hospital admissions. The limited evidence
18 "reported positive O3 associations with... asthma and COPD, especially... during the summer or
19 warm season" (U.S. EPA, 2006, 088089). Of the studies evaluated since the completion of the 2006
20 AQCD, more have focused on identifying whether O3 exposure is associated with specific
21 respiratory-related hospital admissions, including COPD, pneumonia, and asthma, but the overall
22 body of evidence remains small.
Chronic Obstructive Pulmonary Disease
23 Medina-Ramon et al. (2006, 087721) examined the association between short-term exposure
24 to ambient O3 and PMi0 concentrations and Medicare hospital admissions among individuals^
25 65 years of age for COPD (ICD-9: 490-496 excluding 493) in 35 cities in the U.S. for the years
26 1986-1999. The cities included in this analysis were selected because they monitored PMi0 on a
27 daily basis. In this study, city-specific results were obtained using a monthly time-stratified case-
28 crossover analysis. A meta-analysis was then conducted using random effects models to combine the
29 city-specific results. All cities measured O3 from May through September, while only 16 of the cities
30 had year-round measurements. The authors reported a 1.6% increase (95% CI: 0.48, 2.9%) in COPD
31 admissions for lag 0-1 in the warm season for a 30-ppb increase in 8-h max O3 concentrations. When
32 examining single-day lags, stronger associations were observed for lag 1 (2.9% [95% CI: 1.8, 4.0%])
33 compared to lag 0 (-1.5% [95% CI: -2.7, -0.24%]). The authors found no evidence of associations in
34 the cool season (-1.9% [95% CI: -3.6, -0.06%]; lag 0-1) or year round (0.24% [95% CI: -0.78,
35 1.2%]; lag 0-1). In a co-pollutant model using warm season data, the association between O3 and
March 2011 6-84 DRAFT - DO NOT CITE OR QUOTE
-------
1 COPD hospital admissions was robust to the inclusion of PMi0 in the model (results not presented
2 quantitatively). The authors conducted additional analyses to examine potential modification of the
3 warm season estimates for O3 and COPD admissions by several city-level characteristics: percentage
4 living in poverty, emphysema mortality rate (as an indication of smoking), daily summer apparent
5 temperature, and percentage of households using central air conditioning. Of the city-level
6 characteristics examined, stronger associations were only reported for cities with a larger variability
7 in daily apparent summer temperature.
8 In a single-city study conducted in Vancouver from 1994-1998, a location with low ambient
9 O3 concentrations (Table 6-19), Yang et al. (2005, 090184) examined the association between O3 and
10 COPD (ICD-9: 490-492, 494, 496). Ozone was moderately inversely correlated with CO (r=-0.56),
11 NO2 (r=-0.32), and SO2 (r=-0.34), and weakly inversely correlated with PM10 (r=-0.09), suggesting
12 that the observed O3 effect is likely not only due to a positive correlation with other pollutants. Yang
13 et al. (2005, 090184) examined 1- to 7-day (e.g., (0-6 days) lagged moving averages and observed an
14 8.8% (95% CI: -12.5, 32.6%) increase in COPD admissions for lag 0-3 per 20 ppb increase in 24-h
15 avg O3 concentrations. In two-pollutant models at lag 0-3, O3 effect estimates were robust to the
16 inclusion of NO2, SO2, and PMi0 in the model, but were increased slightly when adding CO (Figure
17 6-20; Table 6-22).
Pneumonia
18 In addition to COPD, Medina-Ramon et al. (2006, 087721) examined the association between
19 short-term exposure to ambient O3 and PMi0 concentrations and Medicare hospital admissions
20 among individuals > 65 years of age for pneumonia (ICD-9: 480-487). The authors reported an
21 increase in pneumonia hospital admissions in the warm season (2.5% [95% CI: 1.6, 3.5%] for a 30-
22 ppb increase in 8-h max O3 concentrations; lag 0-1). Similar to the results observed for COPD
23 hospital admissions, pneumonia hospital admissions associations were stronger at lag 1 (2.6% [95%
24 CI: 1.8, 3.4%]) compared to lag 0 (0.06% [95% CI: -0.72, 0.78%]), and no evidence of an
25 association was observed in the cool season or year round. In two-pollutant models, the association
26 between O3 exposure and pneumonia hospital admissions was robust to the inclusion of PMi0
27 (results not presented quantitatively). The authors also examined potential effect modification of the
28 warm season estimates for O3-related pneumonia hospital admissions, as was done for COPD, by
29 several city-level characteristics. Stronger associations were reported in cities with a lower
30 percentage of central air conditioning use. In the cities examined, the percentage of households
31 having central air conditioning ranged from 6 to 93%. Additionally, the authors found no evidence of
32 effect modification of the O3-pneumonia hospital admission relationship when examining the other
33 city-level characteristics.
34 Results from a single-city study conducted in Boston did not support the results presented by
35 Medina-Ramon et al. (2006, 087721). Zanobetti and Schwartz (2006, 090195) examined the
36 association of O3 and pneumonia Medicare hospital admissions for the period 1995-1999. Ozone
37 was weakly positively correlated with PM2 5 (r=0.20) and weakly inversely correlated with black
38 carbon, NO2, and CO (-0.25, -0.14, and -0.30, respectively). In an all-year analysis, the investigators
March 2011 6-85 DRAFT - DO NOT CITE OR QUOTE
-------
1 reported a 3.8% (95% CI: -7.9, -0.1%) decrease in pneumonia admissions for a 20-ppb increase in
2 24-h average O3 concentrations at lag 0 and a 6.0% (95% CI: -11.1, -1.4%) decrease for the average
3 of lags 0 and 1. It should be noted that the mean daily counts of pneumonia admissions was low for
4 this study, -14 admissions per day compared to -271 admissions per day for Medina-Ramon et al.
5 (2006, 087721). but in additional analyses in Boston positive associations with other pollutants and
6 pneumonia hospital admissions was observed (Zanobetti and Schwartz, 2006, 090195).
Asthma
1 There are relatively fewer studies that examined the association between short-term exposure
8 to O3 and asthma hospital admissions, presumably due to the limited power given the relative rarity
9 of asthma hospital admissions compared to ED or physician visits. A study from New York City
10 examined the association of 8-h max O3 concentrations with severe acute asthma admissions (i.e.,
11 those admitted to the Intensive Care Unit [ICU]) during the warm season in the years 1999 through
12 2006 (Silverman and Ito, 2010, 386252). In this study, O3 was moderately correlated with PMi0
13 (r=0.59). When stratifying by age, the investigators reported positive associations with ICU asthma
14 admissions for the 6- to 18-year age group (26.8% [95% CI: 1.4, 58.2%] for a 30-ppb increase in
15 maximum 8-h avg O3 concentrations at lag 0-1), but little evidence of associations forthe other age
16 groups examined (<6 years, 19-49, 50+, and all ages). However, positive associations were observed
17 for each age-stratified group and all ages for non-ICU asthma admissions, but again the strongest
18 association was reported forthe 6- to 18-years age group (28.2% [95% CI: 15.3, 41.5%]; lag 0-1). In
19 two-pollutant models, O3 effect estimates for both non-ICU and ICU hospital admissions remained
20 robust to adjustment for PM2 5. In an additional analysis, using a smooth function, the authors
21 examined whether the shape of the C-R curve for O3 and asthma hospital admissions (i.e., both
22 general and ICU for all ages) is linear. To account for the potential confounding effects of PM2 5,
23 Silverman and Ito (2010, 386252) also included a smooth function of PM25 lag 0-1. When
24 comparing the curve to a linear fit line the authors found that the linear fit is a reasonable
25 approximation of the concentration-response relationship between O3 and asthma hospital
26 admissions around and below the level of the current NAAQS (Figure 6-16).
March 2011 6-86 DRAFT - DO NOT CITE OR QUOTE
-------
co
Ozone: All Ages
cc
cc
o
o>
o
NAAQS
iiniini i ii
20
40
80
T
100
60
Ozone
Source: Used with permission from American Academy of Allergy, Asthma & Immunology, Silverman and Ito (2010, 3862521
Figure 6-16. Estimated relative risks (RRs) of ozone-related asthma hospital admissions
allowing for possible nonlinear relationships using natural splines. [The average of
0 day and 1 day lagged ozone was used in a two-pollutant model with PM2.s lag 0-1,
adjusting for temporal trends, day of the week, and immediate and delayed weather
effects. The solid lines are smoothed fit data, with long broken lines indicating 95%
confidence bands. The density of lines at the bottom of the figure indicates sample
size.
6.2.7.3. Emergency Department Visit Studies
1 Overall, relatively fewer studies have examined the association between short-term exposure
2 to O3 and respiratory-related ED visits, compared to hospital admissions. In the 2006 O3 AQCD
3 (U.S. EPA, 2006, 088089). positive, but inconsistent, associations were observed between O3 and
4 respiratory-related ED visits with effects generally occurring during the warm season. Since the
5 completion of the previous AQCD, larger studies have been conducted, in terms of sample size,
6 study duration, and in some cases multiple cities, to examine the association between O3 and ED
7 visits for all respiratory diseases, COPD, and asthma.
March 2011
6-87
DRAFT - DO NOT CITE OR QUOTE
-------
Respiratory Disease
1 A large single-city study conducted in Atlanta, by Tolbert et al. (2007, 090316). and
2 subsequently reanalyzed by Darrow et al. (2011, 202800). provides evidence for an association
3 between short-term exposures to ambient O3 concentrations and respiratory ED visits. Tolbert et al.
4 (2007, 090316) examined the association between air pollution, both gaseous pollutants and PM and
5 its components, and respiratory disease ED visits, defined as: asthma (ICD-9: 493, 786.07, and
6 786.09), COPD (491, 492, and 496), upper respiratory infection (URI) (460-465, 460.0, and 477),
7 pneumonia (480-486), and bronchiolitis (466.1, 466.11, and 466.19), in all ages from 1993 to 2004.
8 The correlations between O3 and the other pollutants examined ranged from 0.2 for CO and SO2 to
9 0.5-0.6 for the PM measures. Using an a priori average of lags 0-2 for each air pollutant examined,
10 the authors reported a 3.9% (95% CI: 2.7, 5.2%) increase in respiratory ED visits for a 30-ppb
11 increase in 8-h max O3 concentrations during the warm season [defined as May-October in other
12 studies using the same data (Peel et al. (2005, 056305): Strickland et al. (2010, 624878))1. In
13 co-pollutant models, the O3 associations with respiratory ED visits remained robust with CO, NO2,
14 and PMi0 (results not presented quantitatively).
15 Darrow et al. (2011, 202800) examined the same data as Tolbert et al. (2007, 090316) to
16 explore differences in the association between O3 exposure and respiratory-related ED visits due to
17 the use of various exposure metrics. The O3 exposure metrics examined in this study included: 8-h
18 max, 1-h max, 24-h average, commuting period (7:00 a.m. to 10:00 a.m.; 4:00 p.m. to 7:00 p.m.),
19 day-time (8:00 a.m. to 7:00 p.m.) and night-time (12:00 a.m. to 6:00 a.m.). To examine the
20 association between the various O3 exposure metrics and respiratory ED visits, the authors used a
21 time-stratified case-crossover approach, selecting control days as those days within the same
22 calendar month and maximum temperature as the case day. Darrow et al. (2011, 202800) found at
23 lag 1, the results were somewhat variable across exposure metrics. The strongest associations with
24 respiratory ED visits were found when using the 8-h max, 1-h max, and day-time exposure metrics
25 with weaker associations using the 24-h avg and commuting period exposure metrics; a negative
26 association was observed when using the night-time exposure metric (Figure 6-17).
March 2011 6-88 DRAFT - DO NOT CITE OR QUOTE
-------
1 . UO •
§
0 1.02 -
2l
*•" Q)
£ g- 1.01 -
«l
E
s
-= 0 99 -
Partial
Spearman r .
4|
-
I
n
1 0.95 0.93 0.83 0.78 0.04
x >> x
5 « 10
E E
OO
E -T
O -«T
O «
Source: Used with permission from Nature Publishing Group, Darrowetal. (2011, 2028001.
Figure 6-17. Risk ratio for respiratory ED visits and different ozone exposure metrics in Atlanta
from 1993-2004.
1 Orazzo et al. (2009, 202801) examined respiratory ED visits for ages 0-2 years in 6 Italian
2 cities from 1996 to 2000. However, instead of identifying respiratory ED visits using the traditional
3 approach of selecting ICD codes as was done by Tolbert et al. (2007, 090316) and Darrow et al.
4 (2011, 202800). Orazzo et al. (2009, 202801) used data on wheeze extracted from medical records as
5 an indicator of lower respiratory disease. This study examined daily counts of wheeze in relation to
6 air pollution using a time-stratified case-crossover approach in which control days were matched on
7 day of week in the same month and year as the case day. The authors found no evidence of an
8 association between 8-h max O3 concentrations and respiratory ED visits in children aged 0-2 years
9 in models that examined both single-day lags and moving averages of lags from 0-6 days in year-
10 round and seasonal analyses (i.e., warm and cool seasons). In all-year analyses, the percent increase
11 in total wheeze ranged from -1.4% to -3.3% for a 0-1 to 0-6 day lag, respectively.
COPD
12 Stieb et al. (2009, 195858) also examined the association between short-term O3 exposure and
13 COPD (ICD9: 490-492, 494-496) ED visits in 7 Canadian cities. Across cities, in an all-year
14 analysis, O3 was found to be positively associated with COPD ED visits (4.0% [95% CI: -0.54,
15 8.6%] at lag 2 for a 20-ppb increase in 24-h avg O3 concentrations). In seasonal analyses, larger
16 effects were observed between O3 and COPD ED visits during the warm season (i.e., April-
17 September) 6.8% [95% CI: 0.11, 13.9%] (lag day not specified); with no associations observed in the
March 2011
6-89
DRAFT - DO NOT CITE OR QUOTE
-------
1 winter season. As stated previously, in analyses of sub-daily time scales, the authors observed no
2 evidence of consistent associations between any pollutant and respiratory outcome.
3 In a single-city study, Arbex et al. (2009, 184334) examined the association between COPD
4 and several ambient air pollutants, including O3, in Sao Paulo, Brazil for the years 2001-2003 for
5 individuals over the age of 40. Associations between O3 exposure and COPD ED visits were
6 examined in both single-day lag (0-6 days) and polynomial distributed lag models (0-6 days). In
7 all-year analyses, O3 was not found to be associated with an increase in COPD ED visits (results not
8 presented quantitatively). The authors also conducted stratified analyses to examine the potential
9 modification of the air pollutant-COPD ED visits relationship by age (e.g., 40-64, >64) and sex. In
10 these analyses O3 was found to have an increase in COPD ED visits for women, but not for men or
11 either of the age groups examined.
Asthma
12 In a study of 7 Canadian cities, Stieb et al. (2009, 195858) also examined the association
13 between exposure to air pollution (i.e., CO, NO2, O3, SO2, PMio, PM2 5, and O3) and asthma ED
14 visits. Associations between short-term O3 exposure and asthma (ICD9: 493) ED visits were
15 examined at the city-level and then pooled using either fixed or random effects models depending on
16 whether heterogeneity among effect estimates was found to be statistically significant. Across cities,
17 in an all-year analysis, the authors found that short-term O3 exposure was associated with a positive
18 increase (3.5% [95% CI: 0.33, 6.8%] at lag 2 for a 20-ppb increase in 24-h avg O3 concentrations) in
19 asthma ED visits. The authors did not present the results from seasonal analyses for asthma, but do
20 state that no associations were observed between any pollutant and respiratory ED visits in the
21 winter season. Stieb et al. (2009, 195858) also examined associations between respiratory ED visits
22 and sub-daily time scales (i.e., 3-h avg of ED visits versus 3-h avg pollutant concentrations) and
23 found no evidence of consistent associations between any pollutant and respiratory outcome.
24 Several large single-city studies have also provided evidence of an association between asthma
25 ED visits and ambient O3 concentrations. Ito et al. (2007, 156594) examined the association between
26 short-term exposure to air pollution and asthma ED visits for all ages in New York City from 1999 to
27 2002. Ito et al. (2007, 156594) used three different weather models with varying extent of smoothing
28 to account for temporal relationships and multicollinearity among pollutants and meteorological
29 variables (i.e., temperature and dew point) to examine the effect of model selection on the air
30 pollutant-asthma ED visit relationship. When examining O3, the authors reported a positive
31 association with asthma ED visits, during the warm season across the models (ranging from 8.6 to
32 16.9%) and an inverse association in the cool season (ranging from -23.4 to -25.1%), at lag 0-1 for a
33 30-ppb increase in 8-h max O3 concentrations. Using a simplified version of the weather model used
34 in NMMAPS analyses (i.e., terms for same-day temperature and 1-3 day average temperature), Ito et
35 al. (2007, 156594) found that O3 effects were not substantially changed in co-pollutant models with
36 PM2 5, NO2, SO2, and CO during the warm season (Figure 6-20; Table 6-22).
37 Strickland et al. (2010, 624878) examined the association between O3 exposure and pediatric
38 asthma ED visits (ages 5-17 years; ICD-9 codes 493.0; 786.09 before October 1, 1998 or 786.07
March 2011 6-90 DRAFT - DO NOT CITE OR QUOTE
-------
1 after October 1, 1998) in Atlanta between 1993 and 2004 using the same air quality data as Darrow
2 et al. (2011, 202800) and Tolbert et al. (2007, 090316). In this study, the authors developed a
3 statistical model using hospital-specific time-series data that is essentially equivalent to a time-
4 stratified case-crossover analysis (i.e., using interaction terms between year, month, and day-of-week
5 to mimic the approach of selecting referent days within the same month and year as the case day).
6 The authors observed a 6.4% (95% CI: 3.2, 9.6%) increase in ED visits for a 30-ppb increase in 8-h
7 max O3 concentrations at lag 0-2 in an all-year analysis. In seasonal analyses, stronger associations
8 were observed during the warm season (i.e., May-October) (8.4% [95% CI: 4.4, 12.7%]; lag 0-2)
9 than the cold season (4.5% [95% CI: -0.82, 10.0%]; lag 0-2). In co-pollutant analyses O3 effect
10 estimates were not substantially changed when controlling for other pollutants (CO, NO2, PM2 5
11 elemental carbon, PM2 5 sulfate) (results not presented quantitatively). The authors also examined the
12 C-R relationship between O3 exposure and pediatric asthma ED visits and found that both quintile
13 and loess dose-response analyses (Figure 6-18) suggest that there are elevated associations with O3
14 at relatively low concentrations, between 30 and 40 ppb, with stronger evidence at concentrations of
15 40 ppb and above. These dose-response analyses do not provide evidence of a threshold level.
Ozone Warm Season
CM
o
"ro
ce
o
ro ir>
CO -
CD
30 40 50 60 70 80
Concentration (ppb)
Source: Used with permission from American Thoracic Society, Strickland et al. (2010, 6248781
Figure 6-18. Loess dose-response estimates and twice-standard error estimates from
generalized additive models for associations between 3-day avg ozone
concentrations and ED visits for pediatric asthma. [The reference for the rate ratio is
the estimated rate at the 5th percentile of the pollutant concentration. Estimates are
presented for the 5th percentile through the 95th percentile of pollutant
concentrations due to instability in the dose-response estimates at the distribution
tails.
March 2011
6-91
DRAFT - DO NOT CITE OR QUOTE
-------
6.2.7.4. Outpatient and Physician Visit Studies
1 Several studies have examined the association between ambient O3 concentrations and
2 physician or outpatient (non-hospital, non-ED) visits for acute conditions in various geographic
3 locations. Burra et al. (2009, 195868) examined asthma physician visits among patients aged 1-17
4 and 18-64 years in Toronto, Canada from 1992 to 2001. The authors found little or no evidence of an
5 association between asthma physician visits and O3; however, seasonal analyses were not conducted.
6 It should be noted that in this study, most of the relative risks for O3 were less than one and
7 statistically significant, perhaps indicating an inverse correlation with another pollutant or an artifact
8 of the strong seasonality of asthma visits. Villeneuve et al. (2006, 091179) also focused on physician
9 visits to examine the effect of short-term O3 exposure on allergic rhinitis among individuals aged 65
10 or older in Toronto from 1995 to 2000. The authors did not observe any evidence of an association
11 between allergic rhinitis physician visits and ambient O3 concentrations in single-day lag models in
12 an all-year analysis (results not presented quantitatively).
13 In a study conducted in Atlanta, Sinclair et al. (2010, 386271) examined the association of
14 acute asthma and respiratory infection (e.g., upper respiratory infections and lower respiratory
15 infections) outpatient visits from a managed care organization with ambient O3 concentrations as
16 well as multiple PM size fractions and species from August 1998 through December 2002. The
17 authors separated the analysis into two time periods (the first 25 months of the study period and the
18 second 28 months of the study period), in order to compare the air pollutant concentrations and
19 relationships between air pollutants and acute respiratory visits for the 25-month time-period
20 examined in Sinclair et al. (2004, 088696) to an additional 28-month time-period of available ARIES
21 data. The authors found little evidence of an association between O3 and asthma, for both children
22 and adults, or respiratory infection visits in all-year analyses and seasonal analyses. For example, a
23 slightly elevated relative risk (RR) for childhood asthma was observed during the 25-month period
24 in the cold season (RR: 1.12 [95% CI: 0.86, 1.41]; lag 0-2 for a 30-ppb increase in 8-h max O3), but
25 not in the warm season (RR: 0.97 [95% CI: 0.86, 1.10]; lag 0-2). During the 28-month period at lag
26 0-2, a slightly larger positive effect was observed during the warm season (RR: 1.06 [95% CI: 0.97,
27 1.17]), compared to the cold season (RR: 1.03 [95% CI: 0.87, 1.21]). Overall, these results contradict
28 those from Strickland et al. (2010, 624878) discussed above. Although the mean number of asthma
29 visits and O3 concentrations in Sinclair et al. (2010, 386271) and Strickland et al. (2010, 624878) are
30 similar the difference in results between the two studies could be attributed to the severity of O3-
31 induced asthma exacerbations (i.e., more severe symptoms requiring a visit to a hospital) and
32 behavior, such as delaying a visit to the doctor for less severe symptoms.
6.2.7.5. Summary
33 The results of the recent studies evaluated largely support the conclusion of the 2006 O3
34 AQCD. While fewer studies were published overall since the previous review, several multicity
35 studies (e.g., (Cakmak et al., 2006, 093272: Dales et al., 2006, 090744) and a multi-continent study
March 2011 6-92 DRAFT - DO NOT CITE OR QUOTE
-------
1 (Katsouyanni et al., 2009, 199899) provide supporting evidence for an association between short-
2 term O3 exposure and an increase in respiratory-related hospital admissions and ED visits.
3 Collectively, in the studies evaluated, both single-city and multicity, there is continued evidence for
4 increases in both hospital admissions and ED visits when examining all respiratory outcomes
5 combined. Additionally, new studies support an association between short-term O3 exposure and
6 asthma (e.g., (Stieb et al., 2009, 195858: Strickland et al., 2010, 624878) and COPD (e.g., (Medina-
7 Ramon et al., 2006, 087721; Stieb et al., 2009, 195858) hospital admissions and ED visits, with more
8 limited evidence for pneumonia (e.g., (Medina-Ramon et al., 2006, 087721; Zanobetti and Schwartz,
9 2006, 090195). In seasonal analyses, stronger associations were observed in the warm season or
10 summer months compared to the cold season, particularly for asthma (e.g., (Ito et al., 2007, 156594;
11 Strickland et al., 2010, 624878) and COPD (e.g., (Medina-Ramon et al., 2006, 087721) (Figure 6-19;
12 Table 6-21), which is consistent with the conclusions of the 2006 O3 AQCD. There is also continued
13 evidence that children are particularly susceptible to O3-induced respiratory effects (Dales et al.,
14 2006, 090744; Silverman and Ito, 2010, 386252; Strickland et al., 2010, 624878). Studies that
15 focused on respiratory-related outpatient or physician visits found no evidence of an association with
16 short-term O3 exposure, but this could be attributed to the severity of O3-induced respiratory effects
17 requiring more immediate treatment or behavioral factors that result in delayed visits to a physician.
Study
Wongetal. (2009; 196722)
Cakmak etal. (2006)93272)
Dales etal. (2006; 90744)
Orazzoetal.(2009, 202800)a
Katsouyanni et al. (2009; 199899)
Katsouyanni et al. (2009; 199899)
Katsouyanni et al. (2009; 199899)
Katsouyanni et al. (2009; 199899)b
Ho
g Kong
Dar
.
al. (2009; 202800)
(2007,090316)
.
Tolbertetal. (20,
Bigger! etal. (2005; 87395)c
Katsouyanni et al. (2009; 199899)
Katsouyanni et al. (2009; 199899)
Katsouyanni et al. (2009; 199899)
Katsouyanni et al. (2009; 199899)b
Sti
Stiebetal. (2009; 195858)
Medina-Ramon etal. (2006; 87721)
Vang etal. (2006;90184)
Stieb et al. (2009; 195858)e
Medina-Ramon etal. (2006; 87721)
Medina-Ramon etal. (2006; 87721)
Zanobetti and Schwartz (2006; 90195)
Medina-Ramon etal. (2006; 87721)
Medina-Ramon etal. (2006; 87721)
Medina-Ramon etal. (2006; 87721)
10 Ca
11 Canadian cities
6 Italian cities
APHENA-Europe
APHENA-U.S.
APHENA-Canada
APHENA-Canada
Atlanta
Atlanta
8 Italian cities
APHENA-Europe
APHENA-U.S.
APHENA-Canada
APHENA-Canada
New York
New York
New York
Atlanta
36 U.S. cities
Vancouver
36 U.S. cities
36 U.S. cities
36 U.S. cities
36 U.S. cities
36 U.S. cities
Age
0-27 days
0-2
65+
65+
Childrer
6-18
2
0-6
DL( 2)
DL( 2)
DL( 2)
65+
65+
65+
65+
"Wheeze used as indicator of lower respiratory disease.
b APHENA-Canada results standardized to approximate IQR of 5.1 ppb for 1 h max O3 concentrations.
c Study included 8 cities; but of those 8, only 4 had O3 data.
Von-ICU effect estimates.
eThe study did not specify the lag day of the summer season estimate.
Figure 6-19.Percent increase in respiratory-related hospital admission and ED visits in studies
that presented all-year and seasonal analyses. [Effect estimates are for a 20-ppb
increase in 24 hours; 30-ppb increase in 8-h max; and 40-ppb increase in 1-h max
ozone concentrations. Black=AII-year analysis; Red=Summer only analysis;
Blue=Winter only analysis.
March 2011
6-93
DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-21. Corresponding Effect Estimates for Figure 6-19
Study
ED Visit or
Hospital
Admission
Location
Age
Lag
Avg Time
% Increase
(95% Cl)
Respiratory
All-year
Wong et al. (2009, 1967221
Cakmak et al. (2006, 0932721
Dales et al. (2006, 0907441
Orazzoetal. (2011, 20280Q)a
Katsouyanni et al. (2009, 1998991
Hospital Admission
Hospital Admission
Hospital Admission
ED Visit
Hospital Admission
Hong Kong
10 Canadian cities
11 Canadian cities
6 Italian cities
APHENA-Europe
APHENA-U.S.
APHENA-Canada
APHENA-Canada
All
All
0-27 days
0-2
65+
65+
65+
65+
0-1
1.2
2
0-6
0-1
0-1
DL(0-2)
DL(0-2)°
8-h max
24-h avg
24-h avg
8-h max
1-h max
1-h max
1-h max
1-h max
3.58(1.90,5.29)
4.38(2.19,6.46)
5.41 (2.88, 7.96)
-3.34 (-11. 2, 5.28)
1.58 (-1.71, 4.15)
2.38 (0.00, 4.89)
20.4 (4.07, 40.2)
2.4(0.51,4.40)
Warm
Darrowetal. (2011,202800)
Tolbert et al. (2007, 0903161
Bigger! et al. (2005, 0873951°
Katsouyanni et al. (2009, 1998991
ED Visit
ED Visit
Hospital Admission
Hospital Admission
Atlanta
Atlanta
8 Italian cities
APHENA-Europe
APHENA-U.S.
APHENA-Canada
APHENA-Canada
All
All
All
65+
65+
65+
65+
1
0-2
0-3
0-1
0-1
DL(0-2)
DL(0-2)°
8-h max
8-h max
8-h max
1-h max
1-h max
1-h max
1-h max
2.08(1.25,2.91)
3.90 (2.70, 5.20)
0.06 (-5.24, 5.66)
0.24 (-3.32, 3.91)
2. 14 (-0.63, 4.97)
37.1 (11.5,67.5)
4.1(1.40,6.80)
Asthma
All-year
Stieb et al. (2009, 1958581
Strickland et al. (2010, 624878)
ED Visit
ED Visit
7 Canadian cities
Atlanta
All
Children
2
0-2
24-h avg
8-h max
3.48 (0.33, 6.76)
6.38(3.19,9.57)
Warm
Silverman and Ito (2010, 3862521"
Ito et al. (2007, 1565941
Strickland et al. (2010, 624878)
Silverman and Ito (2010, 386252)"
Hospital Admission
ED Visit
ED Visit
Hospital Admission
New York
New York
Atlanta
New York
All
All
Children
6-18
0-1
0-1
0-2
0-1
8-h max
8-h max
8-h max
8-h max
12.5(8.27, 16.7)
16.9(10.9,23.4)
8.43(4.42, 12.7)
28.2(15.3,41.5)
Cold
Ito et al. (2007, 156594)
Strickland et al. (2010, 624878)
ED Visit
ED Visit
New York
Atlanta
All
Children
0-1
0-2
8-h max
8-h max
-23.4 (-27.3, -19.3)
4.52 (-0.82, 10.1)
COPD
All-year
Stieb et al. (2009, 195858)
Medina-Ramon etal. (2006,
087721)
Yang et al. (2005, 090184)
ED Visit
Hospital Admission
Hospital Admission
7 Canadian cities
36 U.S. cities
Vancouver
All
65+
65+
2
0-1
0-3
24-h avg
8-h max
24-h avg
4.03 (-0.54, 8.62)
0.24 (-0.78, 1.21)
8.80 (-12.5, 32.6)
Warm
Stieb et al. (2009, 195858)6
Medina-Ramon etal. (2006,
087721)
ED Visit
Hospital Admission
7 Canadian cities
36 U.S. cities
All
65+
NR
0-1
24-h avg
8-h max
6.76(0.11,13.9)
1.63(0.48,2.85)
Cold
Medina-Ramon etal. (2006,
087721)
Hospital Admission
36 U.S. cities
65+
0-1
8-h max
-1.85 (-3.60, -0.06)
Pneumonia
All-year
Zanobetti and Schwartz (2006,
090195)
Medina-Ramon etal. (2006,
087721)
Hospital Admission
Hospital Admission
Boston
36 U.S. cities
65+
65+
0-1
0-1
24-h avg
8-h max
-5.96 (-11.1, -1.36)
1.81 (-0.72, 4.52)
March 2011
6-94
DRAFT - DO NOT CITE OR QUOTE
-------
Warm
Medina-Ramon etal. (2006,
0877211
Hospital Admission 36 U.S. cities
65+
8-hmax 2.49(1.57,3.47)
Cold
Medina-Ramon etal. (2006,
0877211
Hospital Admission 36 U.S. cities
65+
8-hmax -4.88 (-6.59,-3.14)
'Wheeze used as indicator of lower respiratory disease.
bAPHENA-Canada results standardized to approximate IQR of 5.1 ppb for 1-h max 03 concentrations.
"Study included 8 cities, but of those 8 only 4 had 03 data.
dNon-ICU effect estimates.
eThe study did not specify the lag day of the summer season estimate.
1 Although limited in number, the studies that examined the potential confounding effects of
2 co-pollutants found that O3 effect estimates remained relatively robust upon the inclusion of PM and
3 gaseous pollutants in two-pollutant models (Medina-Ramon et al., 2006, 087721; Strickland et al,
4 2010, 624878: Tolbert et al., 2007, 090316) (Figure 6-20; Table 6-22).
Study
Location Age Lag Copollutant
Respiratory
Katsouyanni et al. (2009; 199899) APHENA-U.S. 65+ 1
APHENA-Europe
APHENA-Canada
Yang etal. (2006; 90184)
Itoetal. (2007;156594)
PM10
a PM10
COPD
Vancouver 65+ 0-3
CO
NO2
SO2
PM10
Ail-Year
New York All 0-1
CO
NO2
SO2
PM2.5
"
Asthma
^
Summer
-10 -5 0 5 10 15 20 25 30
% Increase
Figure 6-20. Percent increase in respiratory-related hospital admissions and ED visits for
studies that presented single and co-pollutant model results. [Effect estimates are
fora 20-ppb increase in 24 hours; 30-ppb increase in 8-h max; and 40-ppb increase
in 1-h max ozone concentrations. An "a" represent risk estimates from APHENA-
Canada standardized to an approximate IQR of 5.1 ppb for a 1-h max increase in
ozone concentrations. Black = results from single-pollutant models; Red = results
from co-polluetant models with PMi0or PM2.s; Yellow = results from co-pollutant
models with CO; Blue = results from co-pollutant models with N02; Green = results
from co-pollutant models with S02.
March 2011
6-95
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
9
10
11
12
13
14
15
Table 6-22. Corresponding
Study3
effect estimates for Figure 6-20
Location Age Lag
Co-Pollutant
% Increase (95% Cl)
All-year
Respiratory
Katsouyanni et al. (2009, 1998991
APHENA-U.S. 65+ 1
APHENA-Europe
APHENA-Canada
PM10
PM10
PM10
PM10
2.62 (0.63, 4.64)
2. 14 (-0.08, 4.40)
2.94(1.02,4.89)
2.38 (0.08, 4.64)
5.54 (-0.94, 12.4)
0.69 (-0.1 2, 1.50)°
5.13 (-6.62, 18.6)
0.64 (-0.87, 2.20)°
COPD
Yang et al. (2005, 0901841
Vancouver 65+ 0-3
CO
N02
S02
PM10
8.80 (-12.5, 32.6)
22.8 (-2. 14, 50.7)
11.1 (-10.4,37.6)
13.4 (-8.40, 40.2)
11.1 (-8.40, 37.6)
Summer
Asthma
Ito et al. (2007, 1565941
New York All 0-1
CO
N02
S02
PM2.5
16.9(10.9,23.4)
18.1(12.1,24.5)
10.2(4.29, 16.4)
13.1 (7.16, 19.5)
12.7(6.37,19.3)
'Averaging times: Katsouyanni et al. (2009,199899) = 1-h max; Yang et al. (2005, 090184) = 24-h avg; and Ito et al. (2007,156594) = 8-h max.
"Risk estimates standardized to an approximate IQR of 5.1 ppbfora 1-h max increase in 03concentrations.
Additionally, a preliminary examination of the C-R relationship found no evidence of a
threshold between short-term O3 exposure and pediatric asthma ED visits (Silverman and Ito, 2010,
386252; Strickland et al., 2010, 624878). Overall, the new body of evidence supports an association
between short-term O3 exposure and respiratory-related hospital admissions and ED visits, with
additional evidence for stronger associations during the warm season for specific respiratory
outcomes such as asthma and COPD.
6.2.8. Respiratory Mortality
The 2006 O3 AQCD found inconsistent evidence for an association between short-term O3
exposure and respiratory mortality (U.S. EPA, 2006, 088089). Although some studies reported a
strong positive association between O3 exposure and respiratory mortality, additional studies
reported a small association or no association. Recent multicity studies found consistent positive
associations between short-term O3 exposure and respiratory mortality, specifically during the
summer months.
The APHENA study, described earlier in Section 6.2.7.2, (Katsouyanni et al., 2009, 199899)
found consistent positive associations for respiratory mortality in all-year analyses with stronger
associations in analyses restricted to the summer season. Additional multicity studies from the U.S.
March 2011
6-96
DRAFT - DO NOT CITE OR QUOTE
-------
1 (Zanobetti and Schwartz, 2008, 101596). Europe (Samoli et al., 2009, 195855). and Italy (Stafoggia
2 et al., 2010, 625034) that conducted summer season analyses provide additional support for an
3 association between short-term O3 exposure and respiratory mortality.
4 Of the studies evaluated, only the APHENA study (Katsouyanni et al., 2009, 199899) and the
5 Italian multicity study (Stafoggia et al., 2010, 625034) conducted an analysis of the potential for
6 co-pollutant confounding of the O3-respiratory mortality relationship. In the APHENA study, in the
7 European dataset, when focusing on the natural spline model with 8 df/year (as discussed in Section
8 6.2.7.2) and lag 1 results (as discussed in Section 6.6.2.1), respiratory mortality risk estimates were
9 robust to the inclusion of PMi0 in co-pollutant models in all-year analyses with O3 respiratory
10 mortality risk estimates increasing in the Canadian and U.S. datasets. In summer season analyses,
11 respiratory O3 mortality risk estimates were robust in the U.S. dataset and attenuated in the European
12 dataset. Similarly, in the Italian multicity study (Stafoggia et al., 2010, 625034). which was limited
13 to the summer season, respiratory mortality risk estimates were attenuated in co-pollutant models
14 with PMio. Based on the APHENA and Italian multicity results, O3 respiratory mortality risk
15 estimates appear to be moderately to substantially sensitive (e.g., increased or attenuated) to
16 inclusion of PMi0. However, in the APHENA study, the mostly every-6th-day sampling schedule for
17 PMio in the Canadian and U.S. datasets greatly reduced their sample size and limits the interpretation
18 of these results.
6.2.9. Summary and Causal Determination
19 The 2006 O3 AQCD concluded that there was clear, consistent evidence of a causal
20 relationship between short-term O3 exposure and respiratory health effects (U.S. EPA, 2006,
21 088089). This causal association was substantiated by the coherence of effects observed across
22 controlled human exposure, epidemiologic, and toxicological studies indicating associations of
23 short-term O3 exposures with a range of respiratory health endpoints from respiratory tract
24 inflammation to respiratory hospital admissions and ED visits. Across disciplines, acute O3
25 exposures induced or were associated with statistically significant declines in lung function. An
26 equally strong body of evidence from controlled human exposure and toxicological studies
27 demonstrated O3-induced inflammatory responses, increased epithelial permeability, and airway
28 hyperresponsiveness (both specific and nonspecific). Toxicological studies provided additional
29 evidence for O3-induced impairment of host defenses. Coherent with inflammation and airway
30 hyperresponsiveness, epidemiologic studies consistently demonstrated positive associations of
31 increases in ambient O3 concentrations with increases in respiratory symptoms and asthma
32 medication use in asthmatic children and with respiratory-related hospital admissions and asthma-
33 related ED visits. Although O3 was consistently associated with nonaccidental and cardiopulmonary
34 mortality, the contribution of respiratory causes to these findings was uncertain.
35 Building on the strong body of evidence presented in the 2006 AQCD, recent studies continue
36 to support associations between short-term O3 exposure and respiratory effects. In young healthy
37 adults exposed to O3 for 6.6 h, studies demonstrate mean FEVi decrements of about 3% at 60 ppb
March 2011 6-97 DRAFT - DO NOT CITE OR QUOTE
-------
1 (Adams, 2006, 087681: Brown et al, 2008, 195140): 5% at 70 ppb (Schelegle et al., 2009, 618629):
2 and 6-8% at 80 ppb (Adams, 2003, 042245: Adams, 2006, 087681: Horstman et al., 1990, 042187:
3 McDonnell et al., 1991, 042384). These studies also show considerable intersubject variability in
4 responsiveness to O3, with the percentage of subjects with >10% decrement in FEVi increasing with
5 increasing concentration of O3 exposure. The proportion (uncorrected for FA responses) of
6 individuals with >10% FEVi decrements ranges from 3 to 20% at an average O3 exposure level of
7 60 ppb (Adams, 1998; (Adams, 2006, 087681: Schelegle et al., 2009, 618629) and from 17 to 29% at
8 80 ppb (Adams, 2006, 087681: McDonnell, 1996, 082679).
9 The collective body of epidemiologic evidence supports demonstrates associations between
10 ambient O3 and decrements in lung function, although recent studies contributed more mixed
11 evidence. A notable difference among newer studies is the limited investigation of populations
12 engaged in outdoor recreation, exercise, or work, which contributed to the strength of evidence in
13 previous AQCDs (U.S. EPA, 1996, 017831: U.S. EPA, 2006, 088089). Some recent evidence
14 suggests that public attention to daily AQI may be reducing exposures of some groups. Recent
15 epidemiologic studies contributed insight into susceptibility factors for O3-associated respiratory
16 morbidity. Among subjects with atopy (Khatri et al., 2009, 594282). asthmatics with concurrent
17 respiratory infection (Lewis et al., 2005, 081079). elderly with AHR or obesity (Alexeeff et al.,
18 2007, 195862). or groups with diminished antioxidant enzyme activity (Alexeeff et al., 2008,
19 195864). lung function responses to ambient O3 exposures generally were exacerbated. The
20 susceptibility of these populations is supported by extensive laboratory evidence (human and animal)
21 for O3-induced exacerbation of allergic inflammation, increased susceptibility to bacterial and viral
22 infections, exacerbation of O3-induced AHR by obesity, and modulation of O3 effects by the
23 oxidative stress/antioxidant balance. In recent controlled human exposure studies, lung function
24 responses to O3 are enhanced in subjects with higher BMI (Bennett et al., 2007, 418827: McDonnell
25 etal. 2010. 383972).
26 As with lung function, recent controlled human exposure studies demonstrate increases in
27 respiratory symptoms in healthy, young adults following 5.6- to 6.6-h exposures to O3 at levels
28 <80 ppb (Adams, 2006, 087681: Schelegle et al., 2009, 618629). The collective body of
29 epidemiologic studies strongly demonstrates positive associations of ambient O3 exposure with
30 respiratory symptoms and asthma medication use among asthmatic subjects, especially in
31 populations with additional susceptibility factors such as asthmatics with atopy (Escamilla-Nunez et
32 al., 2008, 594284: Feo Brito et al., 2007, 093259: Khatri et al., 2009, 594282). asthmatics with
33 diminished antioxidant enzyme activity (Romieu et al., 2006, 090969). or infants with asthmatic
34 mothers (Triche et al., 2006, 093274).
35 Recent studies in animals and in vitro models also continue to demonstrate O3-induced lung
36 injury and inflammatory responses. Building on the extensive experimental evidence, new
37 epidemiologic evidence emerged for ambient O3-associated increases in mediators of inflammation
38 measured in upper and lower airway samples, including eNO (Barraza-Villarreal et al., 2008,
39 156254: Khatri et al., 2009, 594282). cytokines such as IL-6 or IL-8 (Barraza-Villarreal et al., 2008,
March 2011 6-98 DRAFT - DO NOT CITE OR QUOTE
-------
1 156254; Sienra-Monge et al., 2004, 196422). and inflammatory cells such as eosinophils (Khatri et
2 al., 2009, 594282). Epidemiologic studies also report associations of increases in ambient O3 with
3 decreased levels of glutathione (Sienra-Monge et al., 2004, 196422) and increased levels of
4 malondialdehyde in airways (Romieu et al., 2008, 179908). At the time of the 2006 O3 AQCD,
5 controlled human studies of dietary antioxidant supplementation had shown some protective effects
6 of alpha-tocopherol and ascorbate on lung function from O3 exposure, but not on the intensity of
7 subjective symptoms and inflammatory response. More recent evidence indicates that diminished
8 activity of oxidant metabolizing enzymes (e.g., GSTM1, GSTP1) or intake of antioxidant vitamins
9 influences inflammatory responses to O3 exposure (Romieu et al., 2009, 548788; Sienra-Monge et
10 al., 2004, 196422). Across all three disciplines, evidence suggests a role antioxidant defenses in
11 modulating respiratory responses to O3.
12 Recent epidemiologic studies build upon the strong of extant body of evidence of consistently
13 positive associations between daily changes in O3 exposure and respiratory-related hospital
14 admissions and ED visits by demonstrating associations in diverse populations across the U.S.,
15 Canada, and Europe. In all-year analyses, recent multicity studies and a multicontinent study
16 (Katsouyanni et al., 2009, 199899) found an approximate 1.6-5.4% increase in all respiratory-related
17 hospital admissions and ED visits for standardized increases in ambient O3 concentrations1. Positive
18 associations persisted in analyses restricted to the summer season, but the magnitude varied
19 depending on the study location (Katsouyanni et al., 2009, 199899). Compared with studies
20 reviewed in the 2006 O3 AQCD, more recent studies examine associations between short-term O3
21 exposure and hospital admissions and ED visits for specific respiratory outcomes. Although still
22 limited in number, both single- and multicity studies found consistent, positive associations of daily
23 changes in O3 concentrations with asthma and COPD hospital admissions and ED visits. Evidence
24 was more limited for pneumonia. Consistent with the conclusions of the 2006 O3 AQCD (U.S. EPA,
25 2006, 088089). in studies that conducted seasonal analyses, larger effects were estimated for the
26 warm season or summer months than for the cold season or for all seasons, particularly for asthma
27 and COPD. Although the current body of evidence did not include detailed age-stratified results, the
28 increased risk of asthma hospital admissions (Dales et al., 2006, 090744; Silverman and Ito, 2010,
29 386252; Strickland et al., 2010, 624878) observed for children provided additional support for the
30 conclusion from the 2006 O3 AQCD that children are particularly susceptible to O3-induced
31 respiratory effects (U.S. EPA, 2006, 088089). Among studies that evaluated the potential
32 confounding effects of co-pollutants, O3 effect estimates for respiratory-related hospital admissions
33 and ED visits remained relatively robust upon the inclusion of PM and gaseous pollutants in two-
34 pollutant models (Medina-Ramon et al., 2006, 087721; Strickland et al., 2010, 624878; Tolbert et al.,
35 2007, 090316). Although the concentration-response relationship between short-term O3 exposure
36 and respiratory-related hospital admissions and ED visits has not been extensively examined,
1 Effect estimates were standardized to a 20-ppb increase for 24-h avg O3, a 30 ppb increase for 8-h max O3, and a 40-ppb increase for 1-h
max O3.
March 2011 6-99 DRAFT - DO NOT CITE OR QUOTE
-------
1 preliminary examinations found no evidence of a threshold between short-term O3 exposure and
2 pediatric asthma ED visits (Silverman and Ito, 2010, 386252: Strickland et al, 2010, 624878).
3 New evidence extends the potential continuum of well-established O3-associated respiratory
4 effects (e.g., airway inflammation; impaired host defense; lung function decrements; and respiratory
5 symptoms, ED visits, and hospital admissions) by demonstrating associations between ambient O3
6 exposure and respiratory-related mortality. The multicontinent APHENA study reported primarily
7 positive associations with respiratory mortality in all-year analyses, with stronger associations
8 observed in analyses restricted to the summer season. These findings were supported by U.S.
9 (Zanobetti and Schwartz, 2008, 101596) and European (Samoli et al., 2009, 195855) multicity
10 studies, in which a majority of respiratory mortality effect estimates ranged from a 2.3 to 6.8%
11 increase per standardized increase in ambient O3 concentrations. Although co-pollutant confounding
12 was not extensively examined, the O3-respiratory mortality relationship was moderately to
13 substantially sensitive (e.g., increased or attenuated) to inclusion of PMi0 in co-pollutant models
14 (Katsouyanni et al., 2009, 199899: Stafoggia et al., 2010, 625034). However, interpretation of these
15 results requires caution due to the limited PM datasets used in these studies.
16 In summary, new studies evaluated in the current review support or expand upon the strong
17 body of evidence presented in the 2006 O3 AQCD that short-term O3 exposure is causally associated
18 with respiratory health effects. Recent controlled human exposure studies demonstrate decreases in
19 FEVi in the range of 2.8 to 3.6% with prolonged O3 exposures (6.6 hours) as low as 60 ppb in
20 concentration. By demonstrating O3-induced airway hyperresponsiveness, activation of neural
21 reflexes, allergic responses, lung injury, impaired host defense, and airway inflammation,
22 toxicological studies have characterized O3 modes of action and provided biological plausibility for
23 epidemiologic observations of associations of ambient O3 exposure with decreases in lung function
24 and increases in respiratory symptoms. The coherence of results across studies for O3-associated
25 changes in lung function, airway inflammation, and respiratory symptoms, in turn, provides the
26 biological plausibility for epidemiologic findings of consistently positive associations of ambient O3
27 exposure with respiratory hospital admissions and ED visits in diverse populations across the U.S.,
28 Europe, and Canada. Additionally, a multicontinent study and several multicity studies reported
29 positive associations between ambient O3 exposures and respiratory mortality. New epidemiologic
30 studies provide evidence for associations of ambient O3 exposure with biological markers of airway
31 inflammation and oxidative stress and indicated that groups with diminished antioxidant capacity or
32 comorbidities such as atopy, AHR, or obesity may have increased susceptibility to respiratory
33 morbidity associated with O3 exposure. This new information is consistent with previously available
34 toxicological and clinical evidence as well as current information on modes of action. A common
35 observation among epidemiologic studies of respiratory morbidity and mortality was stronger
36 associations in analyses restricted to warm seasons compared to cold seasons. Additionally, although
37 co-pollutant confounding was evaluated infrequently, O3 effect estimates generally remained
38 statistically significant in co-pollutant models with PM2 5, PMi0, or NO2. Collectively, the evidence
39 integrated across controlled human exposure, epidemiologic, and toxicological studies as well as
March 2011 6-100 DRAFT - DO NOT CITE OR QUOTE
-------
1 across the spectrum of respiratory health endpoints continues to demonstrate that there is a causal
2 relationship between short-term O3 exposure and respiratory health effects.
6.3. Cardiovascular Effects
6.3.1. Controlled Human Exposure
3 O3 reacts rapidly on contact with respiratory system tissue and is not absorbed or transported
4 to extrapulmonary sites to any significant degree as such. Controlled human exposure studies
5 discussed in the previous AQCDs (U.S. EPA, 1986, 017607: U.S. EPA, 1996, 017831) failed to
6 demonstrate any consistent extrapulmonary effects. Some controlled human exposure studies have
7 attempted to identify specific markers of exposure to O3 in blood. Foster et al. (1996, 079920) found
8 a reduction in the serum levels of the free radical scavenger a-tocopherol after O3 exposure. Liu et al.
9 (1997, 084627: 1999, 012049) used a salicylate metabolite, 2,3, dehydroxybenzoic acid (DHBA), to
10 indicate increased levels of hydroxyl radical which hydroxylates salicylate to DHBA. Increased
11 DHBA levels after exposure to 120 and 400 ppb suggest that O3 increases production of hydroxyl
12 radical. The levels of DHBA were correlated with changes in spirometry.
13 Gong et al. (1998, 029938) observed a small, statistically significant O3-induced increase in
14 the alveolar-to-arterial PO2 gradient in both healthy (n = 6) and hypertensive (n = 10) adult males
15 (aged 41-78 years) exposed for 3 hours with exercise to 300 ppb O3. The mechanism for the decrease
16 in arterial oxygen tension in the Gong et al. (1998, 029938) study could be due to an O3-induced
17 ventilation-perfusion mismatch. Gong et al. (1998, 029938) suggested that by impairing alveolar-
18 arterial oxygen transfer, the O3 exposure could potentially lead to adverse cardiac events by
19 decreasing oxygen supply to the myocardium. The subjects in the Gong et al. (1998, 029938) study
20 had sufficient functional reserve so as to not experience significant ECG changes or myocardial
21 ischemia and/or injury. In studies evaluating the exercise performance of healthy adults, no
22 significant effect of O3 on arterial O2 saturation has been observed (Schelegle and Adams, 1986,
23 040351).
24 More recently, Fakhri et al. (2009, 191914) evaluated changes in HRV among healthy adult
25 volunteers (n=50; 27 ± 7 years) during 2-h exposures to PM2 5 CAPs (127±62 ug/m3) and O3
26 (114±7 ppb), alone and in combination. High frequency HRV was increased following CAPs-only
27 (p=0.046) and O3-only (p=0.051) exposures, but not in combination. Diastolic blood pressure
28 increased by 2 mmHg following the combined O3 + CAPs exposure, but was not altered by either O3
29 or CAPs alone. Urch et al. (2005, 081080) also reported a 6 mmHg increase in diastolic blood
30 pressure following a 2-h resting exposure to O3 (120 ppb) + PM2 5 CAPs (150 ug/m3) in healthy
31 adults (n=23; 32 ± 107 years), which was statistically different from the 1 mmHg increase seen
32 following FA exposure.
March 2011 6-101 DRAFT - DO NOT CITE OR QUOTE
-------
6.3.2. Epidemiology
1 The 2006 O3 AQCD concluded that the "generally limited body of evidence is highly
2 suggestive that O3 directly and/or indirectly contributes to cardiovascular-related morbidity,"
3 including physiologic effects (e.g., release of platelet activating factor [PAF]), HRV, arrhythmias,
4 and myocardial infarctions, although the available body of evidence reviewed during the 2006 O3
5 AQCD does not "fully substantiate links between ambient O3 exposure and adverse cardiovascular
6 outcomes" (U.S. EPA, 2006, 088089). Since the completion of the 2006 O3 AQCD an increasing
7 number of studies have examined the relationship between short-term O3 exposure and
8 cardiovascular morbidity and mortality. These new studies, as well as evidence from the previous
9 AQCDs, are presented within this section.
6.3.2.1. Arrhythmia
10 In the 2006 O3 AQCD, conflicting results were observed when examining the effect of O3 on
11 arrhythmias (Dockery et al, 2005, 078995: Rich et al, 2005, 079620). A study by Dockery et al.
12 (2005, 078995) reported no association between O3 levels and ventricular arrhythmias among
13 patients with implantable cardioverter defibrillators (ICD) living in Boston, MA, although when O3
14 was categorized into quartiles, there was weak evidence of an association. Rich et al. (2005, 079620)
15 performed a re-analysis of this cohort using a case-crossover design, which did detect a positive
16 association. Recent studies were conducted in various locations and each used a different cardiac
17 episode to define an arrhythmic event and a different time period of exposure, which may help
18 explain observed differences across studies. Ozone levels for each new study are reported in Table 6-
19 23.
Table 6-23. Characterization of ozone concentrations (in ppb) from studies of arrhythmias
Reference
Sarnat et al. (2006, 0904891
Richetal. (2006, 0898141
Richetal. (2006, 0884271
Anderson et al. (2010,6250281
Metzgeretal. (2007, 0928561
Location
Steubenville, Ohio
St. Louis, Missouri
Boston, Massachusetts
London, England
Atlanta, Georgia
Averaging Time
24 h
Summer and Fall only
5 days
24 h
1 h
24 h
8-h max
8-h max
Summer only
Mean Concentration (Standard Deviation) Upper Range of Concentration
21.8(12.6)
22.2(9.1)
21*
22.2*
22.6*
8.08
53.9 (23)
75th: 28.5
Max: 74.8
75th: 29.1
Max: 44
75th: 31
75th: 33
Max: 119.5
75th: 30.9
Max: 77.5
75th: 11.5
Max: 148
"Median presented (information on mean not given).
20 Multiple studies examined O3-related effects on individuals with ICDs. One study of 518 ICD
21 patients who had at least 1 tachyarrythmia within a 10-year period (totaling 6287 tachyarrhythmic
22 event-days; 1993-2002) was conducted in Atlanta, Georgia (Metzger et al., 2007, 092856).
March 2011
6-102
DRAFT - DO NOT CITE OR QUOTE
-------
1 Tachyarrhythmic events were defined as any ventricular tachyarrhythmic event, any ventricular
2 tachyarrhythmic event that resulted in electrical therapy, and any ventricular tachyarrhythmic event
3 that resulted in defibrillation. In the primary analysis, no evidence of association was observed for a
4 30-ppb increase in 8-h max O3 concentrations and tachyarrhythmic events (OR: 1.00 [95% CI: 0.92,
5 1.08]; lag 0). Season-specific as well as several sensitivity analyses (including the use of an
6 unconstrained distributed lag model [lags 0-6]) analyses were conducted resulting in similar null
7 associations. A strength of this study is that it incorporated a much larger sample size over a longer
8 time period.
9 In a case-crossover analysis, a population of ICD patients in Boston, previously examined in a
10 similar study (Rich et al., 2005, 079620) was used to assess the association between air pollution and
11 paroxysmal atrial fibrillation (PAF) episodes (Rich et al., 2006, 088427). In addition to ventricular
12 arrhythmias, ICD devices may also detect supraventricular arrhythmias, of which atrial fibrillation is
13 the most common. Although atrial fibrillation is generally not considered lethal, it has been
14 associated with increased premature mortality as well as hospitalization and stroke. Ninety-one
15 electrophysiologist-confirmed episodes of PAF were ascertained among 29 patients. An association
16 (OR: 3.86 [95% CI: 1.44, 10.28] per 40-ppb increase in 1-h max O3 concentrations) was observed
17 between increases in O3 during the concurrent hour and PAF episodes (lag 0). The estimated OR for
18 the 24-h moving average concentration was elevated (OR: 1.81 [95% CI: 0.86, 3.83] per 20 ppb), but
19 weaker than the estimate for the shorter exposure window. The association between PAF and O3 in
20 the concurrent hour during the cold months was comparable to that during the warm months. In
21 addition, no evidence of a deviation from linearity between O3 concentration and the log odds of
22 PAF was observed. Authors report that the difference between O3 exposure and observed effect
23 between this study (PAF and 1 h O3) and their previous study (ventricular arrhythmias and 24-h
24 moving average O3) (Rich et al., 2005, 079620) suggest a more rapid response to air pollution for
25 PAF (Rich et al., 2006, 088427V
26 In an additional study, Rich et al. (2006, 089814) employed a case-crossover design to
27 examine the association between air pollution and 139 confirmed ventricular arrhythmias among 56
28 ICD patients in St Louis, Missouri. The authors observed a positive association with O3 (OR: 1.17
29 [95% CI: 0.58, 2.38] per 20-ppb increase in 24-h moving avg O3 concentrations [lags 0-23 hours]).
30 Although the authors concluded these results were similar to their results from Boston (Rich et al.,
31 2005, 079620). the pollutants responsible for the increased risk in ventricular arrhythmias are
32 different (O3 and PM2 5 in Boston and sulfur dioxide in St Louis).
33 Anderson et al. (2010, 625028) used a case-crossover framework to assess air pollution and
34 activation of ICDs among patients from all 9 ICD clinics in the London National Health Service
35 hospitals. "Activation" was defined as tachycardias for which the defibrillator delivered treatment.
36 Investigators modeled associations using unconstrained distributed lags from 0 to 5 days. The
37 sample consisted of 705 patients with 5,462 activation days (O3 information was for 543 patients and
38 4,092 activation days). Estimates for O3 were consistently positive, although weak (OR: 1.09 [95%
March 2011 6-103 DRAFT - DO NOT CITE OR QUOTE
-------
1 CI: 0.76, 1.55] per 30 ppb for 0-1 day lag; OR: 1.04 [95% CI: 0.60, 1.81] per 30 ppb for 0-5 day lag)
2 (Anderson et al.. 2010. 625028).
3 In contrast to arrhythmia studies conducted among ICD patients, Sarnat et al. (2006, 090489)
4 recruited non-smoking adults (age range: 54-90 years) to participate in a study of air pollution and
5 arrhythmias conducted over two 12-week periods during summer and fall of 2000 in a region
6 characterized by industrial pollution (Steubenville, Ohio). Continuous ECG data acquired on a
7 weekly basis over a 30-minute sampling period were used to assess ectopy, defined as extra cardiac
8 depolarizations within the atria (supraventricular ectopy, SVE) or the ventricles (ventricular ectopy,
9 VE). Increases in the 5-day moving average (days 1-5) of O3 were associated with an increased odds
10 of SVE (OR: 2.17 [95% CI: 0.93, 5.07] per 20-ppb increase in 24-h avg O3 concentrations). A
11 weaker association was observed for VE (OR: 1.62 [95% CI: 0.54, 4.90] per 20-ppb increase in 24-h
12 avg O3 concentrations). The inclusion of SO42" in the model slightly reduced the effect of 5-day O3
13 on SVE [OR: 1.62 (95% CI: 0.54, 4.90)]. The authors indicate that the strong associations observed
14 at the 5-day moving averages, as compared to shorter time periods, suggests a relatively long-acting
15 mechanistic pathways, such as inflammation, may have promoted the ectopic beats in this population
16 (Sarnat et al., 2006, 090489).
6.3.2.2. Heart Rate/Heart Rate Variability
17 In the 2006 O3 AQCD (U.S. EPA, 2006, 088089). two large population-based studies of air
18 pollution and HRV were summarized (Liao et al., 2004, 056590: Park et al., 2005, 057331). In
19 addition, the biological mechanisms and potential importance of HRV were discussed. Briefly, the
20 study of acute adverse effects of air pollution on cardiac autonomic control is based on the
21 hypothesis that increased air pollution levels may stimulate the autonomic nervous system and lead
22 to an imbalance of cardiac autonomic control characterized by sympathetic activation unopposed by
23 parasympathetic control (U.S. EPA, 2006, 088089). Examples of HRV indices include the standard
24 deviation of normal-to-normal intervals (SDNN), the square root of the mean of the sum of the
25 squares of differences between adjacent NN intervals (r-MSSD), high-frequency power (HF), low-
26 frequency power (LF), and the LF/HF ratio. Liao et al. (2004, 056590) examined the association
27 between air pollution and cardiac autonomic control in the fourth cohort examination (1996-1998) of
28 the U.S.-based Atherosclerosis Risk in Communities Study. A decrease in log-transformed HF was
29 associated with an increase in O3 concentration among white study participants. Park et al. (2005,
30 057331) examined the effects of air pollution on indices of HRV in a population-based study among
31 men from the Normative Aging Study in Boston, Massachusetts. Several associations were observed
32 with O3 and HRV outcomes; a reduction in LF was associated with increased O3 concentration,
33 which was robust to inclusion of PM2s. The associations with all HRV indices and O3 were stronger
34 among those with ischemic heart disease and hypertension. In addition to these population-based
35 studies included in the 2006 O3 AQCD was a study by Schwartz et al. (2005, 074317). who
36 conducted a panel study to assess the relationship between exposure to summertime air pollution and
37 HRV. A weak association of O3 during the hour immediately preceding the health measures was
March 2011 6-104 DRAFT - DO NOT CITE OR QUOTE
-------
1 observed with r-MSSD among a study population that consisted of mostly older female participants.
2 In summary, these studies suggest that short-term exposures to O3 are predictors of decreased HRV
3 and that the relationship may be stronger among certain subgroups. The generally consistent
4 (although weak) associations between pollutants and reduced cardiac autonomic control were
5 observed at relatively low pollution concentrations typically experienced by the U.S. general
6 population on a daily basis (U.S. EPA, 2006, 088089). More recent studies of O3 and HRV and are
7 described below. The O3 concentrations for these studies are presented in Table 6-24.
Table 6-24. Characterization of ozone concentrations (in ppb) from studies of heart rate variability
Reference
Chuanaetal. (2007, 0910631
Ruidavetsetal. (2005, 0894431
Chan et al. (2005, 0889881
Zanobetti et al. (2010,5972501
Wheeler etal. (2006, 0884531
Bajaetal. (2010, 626540)
Park et al. (2008, 0930271
Park et al. (2007, 0932681
Wuetal. (2010, 6260331
Location
Taipei, Taiwan
Toulouse, France
Taipei, Taiwan
Boston, Massachusetts
Atlanta, Georgia
Boston, Massachusetts
Boston, Massachusetts
Boston, Massachusetts
Taipei, Taiwan
Averaging Time [^
ard Deviation) Vppw Ran9e of Concentration
24 h 28.4(12.1) Max: 49.3
48 h 33.3 (8.9) Max: 47.8
72 h 33.8(7.1) Max: 48.3
ah ™iu/i^ 75th: 46.9
8h 38'3<148' Max: 80.3
1h 21.9(15.4) Max: 114.9
0.5 h 20.7*
2 h 20.5*
3D 21.9*
5D 22.8*
4h 18.5
24 h 29.4
Olag 23(16
10-hlag 21(15
75th: 30.33
75th: 30.08
75th: 28.33
75th: 29.28
75th: 22.5
24 h 23.4(13)
24 h Range of 17.0-29.1
Working period 24.9(14.0) Max: 59.2
"Median presented (information on mean not given).
8 Several follow-up examinations of HRV were conducted among the participants of the
9 Normative Aging Study in Boston. A trajectory cluster analysis was used to assess whether pollution
10 originating from different locations had varying relationships with HRV (Park et al., 2007, 093268).
11 Subjects who were examined on days when air parcels originated in the west had the strongest
12 associations with O3; however, the O3 concentration in this cluster was low (24-h avg, 17.0 ppb)
13 compared to the other clusters (24-h avg of 21.3-29.1 ppb). LF and SDNN decreased with increases
14 in the 4-h moving average of O3 from the west (LF decreased by 33.4% [95% CI: 0.9, 55.3%] and
15 SDNN decreased by 17.1% [95% CI: -0.3, 31.5%] per 17-ppb increase in 4-h avg O3 concentrations)
16 (Park et al., 2007, 093268). The Boston air mass originating in the west traveled over Illinois,
17 Indiana, and Ohio; states typically characterized by coal-burning power plants. Due to the low O3
18 concentrations observed in the west cluster, the authors hypothesize that O3 on those days could be
19 capturing the effects of other, secondary and/or transported pollutants from the coal belt or that the
20 relationship between ambient O3 and personal exposure to O3 is stronger during that period
21 (supported by a comparatively low apparent temperature which could indicate a likelihood to keep
22 windows open and reduced air conditioning use) (Park et al., 2007, 093268). An additional follow-up
23 evaluation using the Normative Aging Study examined the potential for effect modification by
March 2011
6-105
DRAFT - DO NOT CITE OR QUOTE
-------
1 chronic lead exposure on the relationship between air pollution and HRV (Park et al, 2008, 093027).
2 Authors observed graded reductions in HF and LF of HRV in relation to O3 (and sulfate) across
3 increasing quartiles of tibia and patella lead (HF: %change 16.1 [95% CI: -18.9, 66.2] for the first
4 quartile of tibia Pb and -37.9 [95% CI: -54.6, -14.9] for the fourth quartile of tibia Pb per 16-ppb
5 increase in 4-h avg O3 concentrations; LF: %change 4.2 [95% CI: -21.8, 38.8] for the first quartile of
6 tibia Pb and -38.1 [95% CI: -51.9, -20.4] forthe fourth quartile of tibia Pb per 16-ppb increase in 4-h
7 avg O3 concentrations). In addition, O3 associations were similar when education and cumulative
8 traffic-adjusted bone lead levels were used in analyses. Authors indicate the possibility that O3
9 (which has low indoor concentrations) was acting as a proxy for sulfate (correlation coefficient for
10 O3 and sulfate = 0.57). Investigators of a more recent follow-up to the Normative Aging Study
11 hypothesized that the relationships between short-term air pollution exposures and ventricular
12 repolarization, as measured by changes in the heart-rate corrected QT interval (QTc), would be
13 modified by participant characteristics (e.g., obesity, diabetes, smoking history) and genetic
14 susceptibility to oxidative stress (Baja et al., 2010, 626540). No evidence of an association between
15 O3 (using a quadratic constrained distributed lag model and hourly exposure lag models over a 10-h
16 time window preceding the visit) and QTc was reported (change in mean QTc -0.74 [95% CI: -3.73,
17 2.25]); therefore, potential effect modification of personal and genetic characteristics with O3 was
18 not assessed (Baja et al., 2010, 626540). Collectively, the results from studies that examined the
19 Normative Aging Study cohort found an association between increases in short-term exposures to O3
20 and decreases in HRV (Park et al., 2005, 057331: Park et al., 2007, 093268: Park et al., 2008,
21 093027) although not consistently in all of the studies (Baja et al., 2010, 626540). Further, observed
22 relationships appear to be stronger among those with ischemic heart disease, hypertension, and
23 elevated bone lead levels, as well as when air masses arrive from the west (the coal belt). However,
24 it is not clear if O3 is acting as a proxy for other, secondary particle pollutants (such as sulfate) (Park
25 et al., 2005, 057331: 2007, 093268: 2008, 093027). In addition, since the Normative Aging Study
26 participants were older, predominately white men, results may not be generalizable to women,
27 younger individuals, or those of different racial/ethnic groups (Baja et al., 2010, 626540).
28 A panel study among 18 individuals with COPD and 12 individuals with recent myocardial
29 infarction (MI) was conducted in Atlanta, Georgia (Wheeler et al., 2006, 088453). HRV was assessed
30 for each participant on 7 days in fall 1999 and/or spring 2000. The mean 4-h O3 concentration (time
31 period immediately preceding the HRV measures) was 18.5 ppb; however, O3 concentrations
32 differed substantially within study sites (8.0 - 33.8 ppb). Ozone concentrations were not associated
33 with HRV (SDNN) among all subjects (percent change of 0.75 [95% CI: -3.6, 5.3] per 9.61-ppb 4-h
34 O3 increase) or when stratified by disease type (COPD, recent MI, and baseline FEVi) (Wheeler et
35 al.. 2006. 088453).
36 HRV and air pollution was assessed in a panel study among 46 predominately white male
37 patients (study population: 80.4% male, 93.5% white) aged 43-75 years in Boston, Massachusetts,
38 with coronary artery disease (Zanobetti et al., 2010, 597250). Up to four home visits were made to
39 assess HRV over the year following the index event. Pollution lags used in analyses ranged between
March 2011 6-106 DRAFT - DO NOT CITE OR QUOTE
-------
1 30 minutes to a few hours and up to 5 days prior to the HRV assessments. Decreases in r-MSSD
2 were reported for all averaging times of O3 (percent change of-5.18% [95% CI: -7.89, -2.30] per 20
3 ppb of 5-day moving average of O3 concentration), but no evidence of association between O3 and
4 HF was observed (quantitative results not provided). In two-pollutant models with O3 and either
5 PM2 5 or BC, the independent effects of O3 were observed.
6 A few studies were conducted outside of the U.S. to assess the relationship between air
7 pollution concentrations and heart rate and HRV (Chan et al, 2005, 088988; Chuang et al, 2007,
8 098629: Ruidavets et al., 2005, 089443: Wu et al., 2010, 626033). No associations were reported
9 between O3 and HRV among CHD patients and patients with one or more major CHD risk factors
10 residing in Taipei, Taiwan (Chan et al., 2005, 088988). Another study taking place in Taipei, Taiwan
11 examined mail carriers and reported O3 levels measured using personal monitors. No association was
12 observed between O3 and the measures of HRV (percent change for SDNN: 0.57 [95% CI: -21.27,
13 28.46], r-MSSD: -7.10 [95% CI: -24.24, 13.92], HF: -1.92 [95% CI: -23.68, 26.02], LF: -4.82 [95%
14 CI:
15 -25.34, 21.35] per 40 ppb O3) (Wu et al., 2010, 626033). In addition, no consistent relationships were
16 identified between O3 and resting heart rate among middle-aged (35-64 years) participants residing
17 in Toulouse, France (Ruidavets et al., 2005, 089443). A negative trend was reported for the 3-day
18 cumulative (lag days 1-3) concentration of O3 with heart rate (p for trend = 0.02); however, the
19 individual odds ratios comparing quintiles of exposure showed no association (OR for O3 of 0.93
20 [95% CI: 0.86, 1.01] for the highest quintile of resting heart rate compared to the lowest). When
21 stratified by current smoking status, non-smokers had a decreased trend with increased 3-day
22 cumulative O3 concentrations but none of the quintiles for heart rate were statistically significant. A
23 panel study was conducted in Taiwan to assess the relationship between air pollutants and
24 inflammation, oxidative stress, blood coagulation, and autonomic dysfunction (Chuang et al., 2007,
25 091063: Chuang et al., 2007, 098629). Participants were apparently healthy college students (aged
26 18-25 year) who were living in a university dormitory in metropolitan Taipei. Health endpoints were
27 measured three times from April to June in 2004 or 2005. Ozone was assessed in statistical models
28 using the average of the 24, 48, and 72 hours before the hour of each blood sampling. Decreases in
29 HRV (measured as SDNN, r-MSSD, LF, and HF) were associated with increases in O3
30 concentrations in single-pollutant models (percent change for SDNN: -13.45 [95% CI: -16.26, -
31 10.60], r-MSSD -13.76 [95% CI: -21.62, -5.44], LF -9.16 [95% CI: -13.29, -4.95], HF -10.76 [95%
32 CI: -18.88, -2.32] per 20 ppb 3-day avg O3 concentrations) and remained associated with 3-day O3
33 concentrations in two-pollutant models with sulfate.
6.3.2.3. Stroke
34 The 2006 O3 AQCD (U.S. EPA, 2006, 088089) did not identify any studies that examined the
35 association between short-term O3 exposure and stroke. However, recent studies have attempted to
36 examine this relationship. Lisabeth et al. (2008, 155939) used a time-series approach to assess the
37 relationship between daily counts of ischemic stroke and transient ischemic attack (TIA) with O3
March 2011 6-107 DRAFT - DO NOT CITE OR QUOTE
-------
1 concentrations in a southeast Texas community among residents 45 years and older (2001-2005;
2 median age of cases, 72 years). The median O3 (hourly average per 24-h time-period) concentration
3 was 25.6 ppb (IQR 18.1-33.8). The associations between same-day (RR: 1.03 [95% CI: 0.96, 1.10]
4 per 20-ppb increase in 24-h avg O3 concentrations) and previous-day (RR: 1.05 [95% CI: 0.99, 1.12]
5 per 20-ppb increase in 24-h avg O3 concentrations) O3 concentrations and stroke/TIA risk were
6 positive. Associations were robust to adjustment for PM2 5. The effect of season on the relationship
7 was not assessed.
8 A case-crossover design was used in a study conducted in Dijon, France between March 1994
9 and December 2004, among those 40 years of age and older who presented with first-ever stroke
10 (Henrotin et al., 2007, 093270). The mean O3 concentration, calculated over 8-h daytime periods,
11 was 14.95 ppb (IQR: 6-22 ppb). An association between ischemic stroke occurrence and O3
12 concentrations with a 1-day lag was observed (OR: 1.54 [95% CI: 1.14, 2.09] per 30-ppb increase in
13 8-h max O3 concentrations). The effect of O3 persisted in two-pollutant models with PMi0, SO2, NO2,
14 and CO. This association was stronger among men (OR: 2.12 [95% CI: 1.36, 3.30] per 30-ppb
15 increase in 8 h max O3 concentrations) than among women (OR: 1.17 [95%CI: 0.77, 1.78] per
16 30-ppb increase in 8 h max O3 concentrations). When stroke was examined by subtype among men,
17 an association was observed for ischemic strokes of large arteries and for transient ischemic attacks
18 but not for cardioembolic or lacunar ischaemic strokes. The subtype analysis was not performed for
19 women. Additionally, for men a linear exposure-response was observed when O3 was assessed based
20 on quintiles (p for trend = 0.01) (Figure 6-21). A potential limitation of this study is that 67.4% of the
21 participating men were smokers compared to 9.3% of the women.
3.5
3 -
2.5 -
O
5?
1.5 -
0.5 -
0-8 9-20 21-32 33-48 48-115
O3 concentration (ppb)
Source: Henrotin et al. (2007, 0932701.
Figure 6-21. Odds ratio (95% confidence interval) for stroke by quintiles of ozone.
March 2011
6-108
DRAFT - DO NOT CITE OR QUOTE
-------
6.3.2.4. Biomarkers
1 An increasing number of studies have examined the relationship between air pollution and
2 biomarkers of inflammation and oxidative stress in an attempt to elucidate the biological
3 mechanisms linking air pollution and cardiovascular disease. A wide range of markers assessed as
4 well as different types of study designs and locations chosen make comparisons across studies
5 difficult. Table 6-25 provides an overview of the O3 concentrations reported in each of the studies
6 evaluated.
Table 6-25. Characterization of ozone concentrations (in ppb) from studies of biomarkers
Reference
Rudezetal. (2009, 1937831
Chuangetal. (2007. 091063)
Baccarelli et al. (2007, 0913101
Steinvil et al. (2008, 1888931
Welleniusetal. (2007, 0928301
Liao et al. (2005, 0886771
Goldberg et al. (2008, 1803801
Chen et al. (2007, 1459561
Thompsonetal. (2010, 3868591
Chuangetal. (2010, 3799931
Location
Rotterdam, the Netherlands
Taipei, Taiwan
Lombardia, Italy
Tel-Aviv, Israel
Boston, Massachusetts
3 U.S. counties
Montreal, Quebec
Los Angeles and San Francisco, California
Toronto, Ontario
Taiwan
Averaging Time
24 h
24 h
48 h
72 h
1 h
0.5 h
1 h/24 h
8h
24 h
8 h/2 wk
8 h/1 mon
1 h/1 yr
Mean Concentration (Standard Deviation)
22*
28.4(12.1)
33.3 (8.9)
33.8(7.1)
18.3*
29.2 (9.7)
25.1 (12.9)
40 (20)
NS
30.8*
28.3*
21.94(15.78)
26.83 (9.7)
Upper Range of
Concentration
75th: 31. 5
Max: 90
Max: 49.3
Max: 47.8
Max: 48.3
75th: 35.1
Max: 202.3
75th: 36
Max: 47.9
Max: 43.1
Max: 62.1
"Median presented (information on mean not given).
7 Thompson et al. (2010, 386859) assessed ambient air pollution exposures and measures of
8 systemic inflammatory biomarkers, IL-6 and fibrinogen. This retrospective repeated measures
9 analysis was conducted among 45 adults (18-40 years of age) in Toronto, Canada between the years
10 of 1999 and 2006. Single pollutant models were used to analyze the repeated-measures data using
11 moving averages up to 7 days. A positive association was observed between IL-6 and O3 with the
12 strongest effects observed for the 4-day moving average of O3 (quantitative results not provided). No
13 association was seen for shorter averaging times (<1 day). When examined by season using 2-day
14 moving averages, the association between O3 and IL-6 was positive during only the spring and
15 summer. No evidence of association was observed for O3 and fibrinogen.
16 The association between O3 exposure and markers of lipid peroxidation and antioxidant
17 capacity was examined among 120 nonsmoking healthy college students, aged 18-22 years, from the
18 University of California, Berkeley (February-June 2002) (Chen et al., 2007, 145956). By design,
19 students were chosen that had experienced different geographic concentrations of O3 over their
20 lifetimes and during recent summer vacation in either greater Los Angeles (LA) or the San Francisco
21 Bay Area (SF). Long-term (based on lifetime residential history) and shorter-term (based on the
22 moving averages of 8-h max concentrations 1-30 days prior to the day of blood collection) O3
March 2011 6-109 DRAFT - DO NOT CITE OR QUOTE
-------
1 exposures were estimated (lifetime exposure results presented in the chronic exposure section). A
2 marker of lipid peroxidation, 8-isoprostane (8-iso-PGF), was assessed. This marker is formed
3 continuously under normal physiological conditions but has been found at elevated concentrations in
4 response to environmental exposures. A marker of overall antioxidant capacity, ferric reducing
5 ability of plasma (FRAP), was also measured. Substantial overlap in the more recent O3 exposure
6 estimates (8-h moving averages) was observed between the two geographic areas sampled. Levels of
7 8-iso-PGF were associated with 2-week ((3 = 0.035 [pg/mL]/8-h ppb O3, p = 0.007) and 1-month ((3 =
8 0.031 [pg/mL]/8-h ppb O3, p = 0.006) estimated O3 exposure levels. No evidence of association was
9 observed between O3 and FRAP. A chamber study performed among a subset of study participants
10 supported the primary study results. The concentrations of 8-iso-PGF increased immediately after
11 the 4-h controlled O3 exposure ended (p = 0.10). However, levels returned to near baseline by
12 18 hours without further exposure. The authors note that O3 was highly correlated with PMi0-2.5 and
13 NO2 in this study population; however, inclusion of these pollutants in the O3 models did not
14 substantially modify the magnitude of the associations with O3.
15 A 2-month panel study among 31 congestive heart failure patients (aged 50-85 years) was
16 conducted to assess the relationship between air pollution and oxygen saturation and pulse rate in
17 Montreal, Canada from July 2002 to October 2003 (Goldberg et al, 2008, 180380V All participants
18 had limited physical functioning (New York Heart Association Classification > II) and an ejection
19 fraction (the fraction of blood pumped out of the heart per beat) less than or equal to 35% (normal is
20 above 55%). Daily mean O3 concentrations were calculated based on hourly measures at 10
21 monitoring stations. There was a negative association between O3 (lag-0) and oxygen saturation
22 when adjustment was made for temporal trends (unadjusted mean difference -0.097 [95% CI: -0.178,
23 -0.015] per 11.85 ppb O3). In the models incorporating personal covariates and weather factors, the
24 association remained suggestive although not statistically significant (adjusted mean difference -
25 0.074 [95% CI: -0.157, 0.010] per 11.85 ppb O3). The associations of O3 with a lag of 1 day or a
26 3-day mean were not statistically significant. No evidence of association was observed between O3
27 exposure and pulse rate.
28 A population-based study was conducted to assess the relationship between short-term
29 exposure to air pollution and markers of blood coagulation/systemic inflammation [fibrinogen, factor
30 VIII coagulant activity (VIII-C), von Willebrand factor (vWF), white blood cell count (WBC), and
31 albumin] using the Atherosclerosis Risk in Communities (ARIC) study cohort (Liao et al., 2005,
32 088677). Significant curvilinear associations were observed for O3 (1 day prior to blood draw) and
33 fibrinogen and vWF (quantitative results not provided for regression models although adjusted
34 means [SE] of vWF were given as 118% [0.79%] for O3 concentrations <40 ppb, 117% [0.86%] for
35 O3 concentrations 40-70 ppb, and 124% [1.97%] for O3 concentrations of 70 ppb). The association
36 between O3 and fibrinogen was more pronounced among those with a history of cardiovascular
37 disease (CVD) and was statistically significant among only this subgroup of the population. The
38 curvilinear relationship between exposure and outcome suggested stronger relationships at higher
39 concentrations of O3 which could indicate threshold effects. The authors note that the most
March 2011 6-110 DRAFT-DO NOT CITE OR QUOTE
-------
1 pronounced associations occurred when the pollutants were 2-3 standard deviations above the mean.
2 In addition, the regression coefficients are small, indicating weak associations. The results from this
3 relatively large-scale cross-sectional study suggest weak associations with O3 and fibrinogen (among
4 those with a history of CVD) and vWF.
5 In a repeated-measures study conducted in Boston among 28 patients with congestive heart
6 failure and impaired systolic function, Wellenius et al. (2007, 092830) found no evidence of an
7 association between B-type natriuretic peptide (BNP) and short-term O3 exposures at lags 0-3 days
8 (quantitative results not provided). BNP was chosen because it is directly associated with cardiac
9 hemodynamics and symptom severity among those with heart failure and is, therefore, considered a
10 marker of functional status. However, the authors conclude that the use of BNP may not be useful in
11 studies of the health effects of ambient air pollutants due to the large amount of within-person
12 variability in BNP levels observed in this population.
13 International studies were identified that also examined the association between air pollution
14 and biomarkers of cardiovascular risk (Baccarelli et al., 2007, 091310; Chuang et al., 2007, 091063;
15 Rudez et al., 2009, 193783; Steinvil et al., 2008, 188893). The relationship between pollutant
16 concentrations and one-time measures of inflammatory biomarkers was assessed among 3659
17 apparently healthy individuals in Tel Aviv, Israel (Steinvil et al., 2008, 188893). No evidence of
18 association was observed between O3 and high-sensitivity C-reactive protein (hs-CRP) (expected
19 relative change of -2% [95% CI: -12, 9] and -4% [95% CI: -16, 10] per 15 ppb O3 averaged over the
20 last week for men and women, respectively) or WBC (expected absolute change of-25 cells/(iL
21 [95% CI: -178, 191] and 142 cells/^L [95% CI: -79, 363] per 15 ppb O3 averaged over the last week
22 for men and women, respectively). In single pollutant models, O3 was associated with an increase in
23 fibrinogen at a 4-day lag among men (expected absolute change of 4.2 mg/dL [95% CI: 0.1, 8.3] per
24 15 ppb O3) and a same-day O3 concentration among women (expected absolute change of 6.5 mg/dL
25 [95% CI: 1.4, 11.5] per 15 ppb O3) but results for other lags (0 through 7 days) were mixed (some
26 positive, some negative; none statistically significant). The associations for men with 4-day lag and
27 for women with 0-day lag did not persist in multi-pollutant models and, in fact, several inverse
28 associations were observed between O3 and fibrinogen.
29 The effects of air pollution on fasting and postmethionine-load total homocysteine (tHcy)
30 levels were assessed among 1,213 apparently healthy individuals from Lombardia, Italy from
31 January 1995 to September 2005 (Baccarelli et al., 2007, 091310). tHcy is an independent risk factor
32 for vascular disease and measurement of this marker after oral methionine load is used to identify
33 individuals with mild impairment of homocysteine metabolism. An increase in the 24-h O3
34 concentrations was associated with an increase in fasting tHcy (percent change 6.25 [95% CI: 0.84,
35 11.91] per 20 ppb O3) but no association was observed with postmethionine-load tHcy (percent
36 change 4.16 [95% CI: -1.76, 10.42] per 20 ppb O3). In addition, no evidence of association was
37 observed between 7-day O3 concentrations and tHcy (percent change for fasting tHcy 3.36 [95% CI:
38 -1.30, 8.39] and percent change for postmethionine-load tHcy-0.65 [95% CI: -5.66,4.71] per 20 ppb
March 2011 6-111 DRAFT-DO NOT CITE OR QUOTE
-------
1 O3). No evidence of effect modification by smoking was observed. The authors conclude that their
2 results did not show a consistent pattern of an effect of O3 on tHcy.
3 A panel study (n=76) of healthy individuals was conducted in Taiwan to assess the relationship
4 between air pollutants and inflammation, oxidative stress, blood coagulation, and autonomic
5 dysfunction (Chuang et al., 2007, 091063). Health endpoints were measured three times from April
6 to June in 2004 or 2005. Ozone effects were assessed in statistical models using the average of the
7 24 hours (1 day), 48 hours (2 days), and 72 hours (3 days) before the hour of each blood sampling.
8 Increases in hs-CRP, 8-hydroxy-2'-deoxyguanosine (8-OHdG), fibrinogen, and plasminogen
9 activator fibrinogen inhibitor-1 (PAI-1) were associated with increases in O3 concentrations in
10 single-pollutant models (percent change in hs-CRP: 244.38 [95% CI: 4.54, 585.15] per 20 ppb 3-day
11 avg O3; percent change in 8-OHdG: 2.46 [95% CI: 1.01, 3.92] per 20 ppb 1-day avg O3; percent
12 change in fibrinogen: 11.76 [95% CI: 4.03, 19.71] per 20 ppb 3-day avg O3; percent change in PAI-
13 1: 37.53 [95% CI: 38.91, 84.27] per 20 ppb 3-day avg O3). No association was seen between O3 and
14 tissue-type plasminogen activator (tPA), a fibrinolytic factor (percent change 16.15 [95% CI: -4.62,
15 38.34] per 20 ppb 3-day avg O3). PAI-1 remained statistically significantly associated with 3-day O3
16 concentrations in two-pollutant models with sulfate.
17 A repeated measures study was conducted in 40 healthy individuals living or working in the
18 city center of Rotterdam, the Netherlands to assess the relationship between air pollution and
19 markers of hemostasis and inflammation (platelet aggregation, thrombin generation, fibrinogen, and
20 CRP) (Rudez et al., 2009, 193783). Each participant provided between 11 and 13 blood samples
21 throughout a 1-year period (498 samples on 197 days). Examined lags ranged from 6 hours to 3 days
22 prior to blood sampling. No consistent evidence of association was observed between O3 and any of
23 the biomarkers (percent change of max platelet aggregation: -6.87 [95% CI: -21.46, 7.70] per 20 ppb
24 4-day average O3; percent change of endogenous thrombin potential: 0.95 [95% CI: -3.05, 5.23] per
25 20-ppb 4-day avg O3; percent change of fibrinogen: -0.57 [95% CI: -3.05, 2.00] per 20-ppb lag 1-day
26 O3; percent change of CRP: -0.48 [95% CI: -14.05, 13.10] per 20-ppb lag 1-day O3). Some
27 associations with O3 were in the opposite direction to that hypothesized which may be explained by
28 the negative correlation between O3 and the other pollutants (correlation coefficients ranged from -
29 0.4 to -0.6). The statistically significant inverse effects observed with O3 in single-pollutant models
30 were no longer apparent when PMi0 was included in the models (Rudez et al., 2009, 193783).
31 Chuang et al. (2010, 379993) conducted a population-based cross-sectional analysis of data
32 collected on 7,778 participants during the Taiwanese Survey on Prevalence of Hyperglycemia,
33 Hyperlipidemia, and Hypertension in 2001. Apolipoprotein B (ApoB), the primary apolipoprotein
34 among low-density lipoproteins, was associated with 3-day avg O3 at the p < 0.10 level (change in
35 ApoB: 0.78 mg/dL [95% CI: -0.06, 1.62] per 12.15 ppb O3). The 5-day mean O3 concentration was
36 associated with an increase in triglycerides at p < 0.10 (change intriglycerides: 2.15 mg/dL [95% CI:
37 -0.03, 4.32] per 12.15 ppb O3). In addition, the 1-, 3-, and 5-day mean O3 concentrations were
38 associated with increased HbAlc levels (a marker used to monitor the degree of control of glucose
39 metabolism)(p < 0.05; change inHbAlc: 0.06 % [95% CI: 0.02, 0.10], 0.05% [95% CI: 0.02, 0.08],
March 2011 6-112 DRAFT-DO NOT CITE OR QUOTE
-------
1 and 0.07% [95% CI: 0.04, 0.10] per 12.15 ppb O3, respectively). The 5-day mean O3 was associated
2 with increased fasting glucose levels (p < 0.10) (change in fasting glucose: 0.77 mg/dL [95% CI:
3 -0.05, 1.59] per 12.15 ppb O3). No association was observed between O3 concentration and ApoAl
4 (change inApoAl: -0.24 mg/dL [95% CI: -1.04, 0.56], -0.14 [95% CI: -0.94, 0.66], and 0.01 [95%
5 CI: -0.69, 0.72] per 12.15 ppb for 1-, 3-, and 5-day averaged O3, respectively). Co-pollutant models
6 were not assessed.
6.3.2.5. Myocardial Infarction (M)
7 The 2006 O3 AQCD (U.S. EPA, 2006, 088089) reported mixed results on the association
8 between short-term O3 exposure and MI. One study reported a positive association between current
9 day O3 concentration and acute MI, especially among the oldest age group (55- to 64-year olds)
10 (Ruidavets et al., 2005, 074091). No association was observed in a case-crossover study of O3 during
11 the hours surrounding the event and MI (Peters et al., 2001, 016546). Since the 2006 O3 AQCD, no
12 new epidemiology studies have examined this association for MI, but one study has been published
13 on arterial stiffness. Wu et al. (2010, 626033) examined mail carriers aged 25-46 years and measured
14 exposure to O3 through personal monitors [mean O3 24.9 (SD 14.0) ppb]. Ozone exposure was
15 positively associated with arterial stiffness (percent change 11.24% [95% CI: 3.67, 19.62] per 40-ppb
16 O3) and was robust to adjustment for PM.
6.3.2.6. Blood Pressure
17
18
19
In the 2006 O3 AQCD, no epidemiologic studies examined O3-related effects on blood
pressure (BP). Recent studies have been conducted to evaluate this relationship and the O3
concentrations for these studies are listed in Table 6-26.
Table 6-26. Characterization of ozone concentrations (in ppb) from studies of blood pressure
20
21
22
23
24
25
Reference Location Averaging Time
8h
Phni ot il PHD? nO?10fi^ Inrhonn ^nnth Knrm
(cold season)
Delfinoetal. (2010, 625026) Los Angeles, California 24 h
L. u- t , ™n, no,, on, n 1 h
5 days
Chuang etal. (2010, 379993) Taiwan
Mean Concentration
(Standard Deviation)
26.6(11.8)
17.5(7.3)
27.1 (11.5)
20
24
26.83 (9.7)
Upper Range of Concentration
75th: 34.8
Max: 62.4
75th: 22.9
Max: 33.9
Max: 60.7
Max: 62.1
Zanobetti et al. (2004, 087489) examined the relationship between air pollutants and BP from
May 1999 to January 2001 for 631 repeat visits among 62 Boston residents with CVD. In single-
pollutant models, higher resting diastolic blood pressure (DBP) was associated with the 5 -day (0-
4 days) averages of O3 (RR: 1.03 [95% CI: 1.00, 1.05] per 20-ppb increase in 24-h O3
concentrations). However, this effect was no longer apparent when PM2 5 was included in the model
(data not presented) (Zanobetti et al., 2004, 087489). Delfino et al. (2010, 625026) examined 64
March 2011
6-113
DRAFT - DO NOT CITE OR QUOTE
-------
1 subjects 65 years and older with coronary artery disease, no tobacco smoke exposure, and living in
2 retirement communities in the Los Angeles air basin with hourly (up to 14 h/day) ambulatory BP
3 monitoring for 5 days during a warm period (July-mid-October) and 5 days during a cool period
4 (mid-October-February). Investigators assessed lags of 1, 4, and 8 hours, 1 day, and up to 9 days
5 before each BP measure; no evidence of association was observed for O3 exposures (change in BP
6 associated with a 20-ppb change in 24-h O3 was 0.67 [95% CI: -1.16, 2.51 for systolic BP [SBP] and
7 -0.25 [95% CI: -1.25, 0.75] for DBP) (Delfino et al., 2010, 625026). Choi et al. (2007, 093196)
8 conducted a cross-sectional study to investigate the relationship between air pollutants and BP
9 among 10,459 participants of the Inha University Hospital health examination from 2001 to 2003.
10 These individuals had no medical history of cardiovascular disease or hypertension. Ozone was
11 associated with an increase in SBP for 1-day lag in the warm season and similar effect estimates
12 were observed during the cold season but were not statistically significant (quantitative results not
13 provided). Associations between O3 and DBP were present in the cold season but not the warm
14 season (quantitative results not provided). The interaction term between O3 and season was
15 statistically significant. Chuang et al. (2010, 379993) conducted a similar type of study among 7,778
16 participants of the Taiwanese Survey on Prevalence of Hyperglycemia, Hyperlipidemia, and
17 Hypertension in 2001. Investigators examined 1-, 3-, and 5-day avg O3 concentrations. An increase
18 in DBP was associated with the 3-day mean O3 concentration (change in BP for a 20-ppb increase in
19 O3 was 0.61 [95% CI: 0.07, 1.14]) (Chuang et al., 2010, 379993). Associations were not observed for
20 other days or with DBP.
6.3.2.7. Hos pital Admis s ions and Emergency Department Vis its
21 Upon evaluating the collective evidence for O3-related cardiovascular HAs and ED visits, the
22 2006 O3 AQCD concluded that "a few studies observed positive O3 associations, largely in the warm
23 season. Overall, however, the currently available evidence is inconclusive regarding any association
24 between ambient O3 exposure on cardiovascular hospitalizations" (U.S. EPA, 2006, 088089). Table
25 6-27 below provides information on the O3 concentrations reported in each of the recent HA and ED
26 visit studies evaluated.
March 2011 6-114 DRAFT-DO NOT CITE OR QUOTE
-------
Table 6-27. Characterization of ozone concentrations (in ppb) from studies of HAs and ED visits
Study
Ballesteretal. (2006, 088746)
Bell et al. (2008, 0912681
Buadong et al. (2009, 6020601
Cakmak et al. (2006, 0990681
Chan et al. (2006, 0901931
Halonen et al. (2009, 6257641
Hosseinpooret al. (2005, 0874131
Lanki et al. (2006, 0897881
Larrieu et al. (2007, 0930311
Lee et al. (2003, 0955521
Lee et al. (2007, 1966131
Middletonetal. (2008, 156760)
Peel et al. (2007, 0904421
Stieb et al. (2009, 1958581
Symons et al. (2006, 0912581
Villeneuve et al. (2006, 0901911
Von Klot et al. (2005, 0880701
Welleniusetal. (2005, 0874831
Yang (2008, 1571601
Zanobetti and Schwartz (2006, 0901951
Location
Multicity, Spain
Taipei, Taiwan
Bangkok, Thailand
Multicity, Canada
Taipei, Taiwan
Helsinki, Finland
Tehran, Iran
Multicity, Europe
Multicity France
Seoul, Korea
Kaohsiung, Taiwan
Nicosia, Cyprus
Atlanta, GA
Multicity, Canada
Baltimore, MD
Edmonton, Canada
Multicity, Europe
Allegheny County, PA
Taipei, Taiwan
Boston, MA
Averaging Time
8h
warm season
24 h
1 h
1-h max
1-h max
8-h max
warm season
8-h max
8-h max
warm season
8-h max
warm season
1-h max
24 h
8-h max
8-h
warm season
24 h
8h
warm season
24 h
24 h
warm season
24 h
cold season
8-h max
warm season
24 h
24 h
24 h
Mean Concentration (Standard Deviation)
Ranged from 24.2 to 44.3
21.4
14.4(3.2)
17.4
50.9 (26.4)
35.7*
4.9 (4.8)
Ranged from 31. 7 to 57.2*
Ranged from 34.2 to 53.1
36.0(18.6)
26.5
Ranged from 28.7 to 54.9
55.6 (23.8)
18.4
31.0(20.0)
17(9.1)
21.8(8)
12.2(7.4)
Ranged from 16. 4 to 28.0
24.3(12.2)
21.0
22.4*
Upper Range of Concentration
Max: 53.4
Max: 41. 9
Max: 150.3
75th: 42.1
Max: 79.6
75th: 7.2
Max: 99.0
75th: 44.9
75th: 35.5
Max: 83.0
Max: 120.0
75th: 23.5
75th: 27.0
75th: 17.0
75th: 32.0
75th: 26.3
Max: 62.8
75th: 31.0
"Median presented (information on mean not given).
1 Multiple recent studies of O3 exposure and cardiovascular HAs and ED visits have been
2 conducted in the U.S. and Canada. Peel et al. (2007, 090442) used a case-crossover framework to
3 assess the relationship between air pollutants and cardiovascular disease ED visits among those with
4 and without secondary comorbid conditions (hypertension, diabetes, chronic obstructive pulmonary
5 disease [COPD], congestive heart failure [CHF], and dysrhythmia). Data on over 4 million ED visits
6 from 31 hospitals were collected from January 1993 to August 2000. Ozone was monitored from
7 March to October and 8 h max concentrations were used in case-crossover analyses. This study was
8 a re-analysis of a time series study conducted to assess the main effects of air pollutants on
9 cardiovascular ED visits in Atlanta (Metzger et al., 2004, 044222: Tolbert et al., 2007, 090316). In
10 the initial study, no evidence of associations was observed between O3 and all CVD visits or visits
11 for CVD subgroups, such as dysrhythmia, CHF, ischemic heart disease (IHD), and peripheral
12 vascular and cerebrovascular disease. The relative risk for all CVD visits was 1.01 (95% CI: 0.99,
13 1.02) for a 20-ppb increase in the 3-day moving avg (lags 0-2 days) of 8-h O3 (Metzger et al., 2004,
March 2011
6-115
DRAFT - DO NOT CITE OR QUOTE
-------
1 044222). Similar to the initial investigation using a time-series analysis, no evidence of association
2 was observed for the O3 3-day moving average and CVD visits among the entire population using
3 the case-crossover design (Peel et al, 2007, 090442). However, the relationship between O3 and
4 peripheral and cerebrovascular disease visits was substantially stronger among patients with
5 comorbid COPD (OR: 1.19 [95% CI: 1.03-1.36] per 20 ppb, lag 0-2 days) as compared to patients
6 without COPD (OR: 1.01 [95% CI: 0.97-1.04] per 20 ppb, lag 0-2 days) (comparing O3 regression
7 coefficients for visits with and without comorbid COPD: p < 0.05). The same research group
8 expanded upon the number of Atlanta hospitals providing ED visit data (41 hospitals) as well as the
9 length of the study period (1993-2004) in order to assess multi-pollutant models and updated single-
10 pollutant models (Tolbert et al., 2007, 090316). The mean concentration for 8-h O3 was 53.0 ppb.
11 Similar to the results presented by Metzger et al. (2004, 044222) and Peel et al. (2007, 090442)
12 among the entire study population, no evidence of associations was observed for O3 and CVD visits
13 (Tolbert et al., 2007, 090316). Again, models assessing the health effects of O3 were limited to data
14 collected from March through October.
15 Cakmak et al, (2006, 093272) investigated the relationship between gaseous air pollutants and
16 cardiac hospitalizations in 10 large Canadian cities using a time-series approach. A total of 316,234
17 hospital discharge records for primary diagnosis of congestive heart failure, ischemic heart disease,
18 or dysrhythmia were obtained from April 1993 through March 2000. Lags 0-5 were examined in
19 analyses. Correlations between pollutants varied substantially across cities, which could partially
20 explain discrepancies in effect estimates observed across the cities. In addition, pollutant lags
21 differed across cities; the average lag for O3 was 2.9 days. The pooled effect estimate for a 20-ppb
22 increase in the daily 1-h max O3 concentration and the percent change in hospitalizations among all
23 10 cities was 2.3 (95% CI: 0.11, 4.50), and this estimate was not substantially altered in
24 multi-pollutant analyses. The authors reported no evidence of effect modification by gender,
25 neighborhood-level education, or neighborhood-level income. Seasonal variation was not assessed. A
26 similar multicity time-series study was conducted using nearly 400,000 ED visits to 14 hospitals in
27 seven Canadian cities from 1992 to 2003 (Stieb et al., 2009, 195858). Primary analyses considered
28 daily O3 single day lags of 0-2 days; in addition, sub-daily lags of 3-h avg concentrations up to
29 12 hours before presentation to the ED were considered. Seasonal variation was assessed by
30 stratifying analyses by warm and cold seasons. No evidence of effect of O3 on CVD ED visits was
31 observed. One negative, statistically significant association was reported between a 1-day lag of O3
32 and visits for angina/myocardial infarction. Ozone was negatively correlated with many of the other
33 pollutants, particularly during the cold season.
34 The effect of air pollution on daily ED visits for ischemic stroke (n=10,881 visits) in
35 Edmonton, Canada was assessed from April 1992 through March 2002 (Szyszkowicz, 2008,
36 192128). A 26.37% (95% CI: 3.16-54.5) increase in stroke ED visits was associated with a 20-ppb
37 increase in O3 at lag 1 among men aged 20-64 years in the warm season. No associations among
38 women or among men age 65 and older reached statistical significance (p < 0.1). In addition, no
39 associations were observed for the cold season or for other lags (lag 0 or lag 2). A similar
March 2011 6-116 DRAFT-DO NOT CITE OR QUOTE
-------
1 investigation over the same time period in Edmonton, Canada, assessed the relationship between air
2 pollutants and ED visits for stroke (ischemic stroke, hemorrhagic stroke, and transient ischemic
3 attack) among those 65 years of age and older using a case-crossover framework (Villeneuve et al,
4 2006, 090191). Lags considered for pollution levels were same day, 1-day lag, and 3-day avg (lag
5 days 0-2). Two-pollutant models were assessed. In addition, results were stratified by season, gender,
6 and stroke sub-type. No evidence of association was reported for O3 and stroke hospitalization
7 (Villeneuve et al., 2006, 090191).
8 Three additional studies reported no evidence of association between O3 concentrations and
9 ED visits, hospitalizations, or symptoms leading to hospitalization (Symons et al., 2006, 091258;
10 Wellenius et al., 2005, 087483: Zanobetti and Schwartz, 2006, 090195). Symons et al. (2006,
11 091258) used a case-crossover framework to assess the relationship between air pollutants and the
12 onset of symptoms (dyspnea) severe enough to lead to hospitalization (through the ED) for
13 congestive heart failure. The study was conducted from April to December of 2002 in Baltimore,
14 Maryland. Exposures were assigned using 3 index times: 8-h and 24-h periods prior to symptom
15 onset and date of hospital admission. No evidence of association was reported for O3 concentrations.
16 Although seasonal variation was not assessed, the time frame for the study did not involve an entire
17 year (April to December). Wellenius et al. (2005, 087483) investigated the association between air
18 pollutants and congestive heart failure hospitalization among Medicare beneficiaries in Pittsburgh,
19 Pennsylvania from 1987 to 1999 utilizing a case-crossover framework. Atotal of 55,019 admissions
20 from the emergency room with a primary discharge diagnosis of CHF were collected. Single- and
21 two-pollutant models were assessed. In addition, effect modification by age, gender, and presence of
22 secondary diagnoses was considered, but seasonal variation was not assessed. No evidence of an
23 association was reported for O3 and CHF hospitalization (Wellenius et al., 2005, 087483). Finally,
24 Zanobetti and Schwartz (2006, 090195) assessed the relationship between air pollutants and hospital
25 admissions through the ED for myocardial infarction and pneumonia among patients aged 65 and
26 older residing in the greater Boston area (1995-1999) using a case-crossover framework with control
27 days matched on temperature. Pollution exposures were assigned for the same day and for the mean
28 of the exposure the day of and the day before the admission. Seasonal variation was assessed. Ozone
29 was not associated with MI admissions.
30 Several recent studies have examined the relationship between air pollution and CVD hospital
31 admissions and/or emergency department visits in Asia. In Taiwan, fairly consistent positive
32 associations have been reported for O3 and congestive heart failure hospital admissions (for single-
33 and multi-pollutant models) in Taipei on warm days (Yang, 2008, 157160) and in Kaohsiung (Lee et
34 al., 2007, 196613): cerebrovascular disease ED visits (for lag 0 single- and two-pollutant models but
35 not other lags or 3-pollutant models) in Taipei (Chan et al., 2006, 090193): and arrhythmia ED visits
36 in Taipei among those without comorbid conditions (Chiu et al., 2009, 190249: Lee et al., 2008,
37 192076) and in Taipei on warm days among those with and without comorbid conditions (Jansson et
38 al., 2001, 092076: Lee et al., 2008, 192076). However, one study in Taiwan did not shown an
39 association. Bell et al. (2008, 091268) reported no evidence of an O3 association with hospital
March 2011 6-117 DRAFT-DO NOT CITE OR QUOTE
-------
1 admissions for ischemic heart disease or cerebrovascular disease. Three studies based in Asia but
2 outside Taiwan were performed. First, a Hong Kong-based investigation (Wong et al., 2009, 196722)
3 reported no consistent evidence of a modifying effect of influenza on the relationship between O3
4 and CVD admissions. Second, among elderly populations in Thailand, O3 was associated with CVD
5 visits, but this association was not detected among younger age groups (15-64) (Buadong et al.,
6 2009, 602060). Third, a study performed in Seoul, Korea reported a positive association between O3
7 levels and HAs for ischemic heart disease; the association was slightly greater among those over
8 64 years of age (Lee et al.. 2003. 095552).
9 Positive effects of O3 on CVD hospital admissions and/or ED visits have been reported in
10 other areas of the world as well (Ballester et al., 2006, 088746: De Pablo et al., 2006, 196506:
11 Linares and Diaz, 2010, 383413: Middleton et al., 2008, 156760: Turner et al., 2007, 196637:
12 Von Klot et al., 2005, 088070: Yallop et al., 2007, 090702). although not consistently as some studies
13 reported no association (Barnett et al., 2006, 089770: Halonen et al., 2009, 625764: Hinwood et al.,
14 2006, 088976: Hosseinpoor et al., 2005, 087413: Lanki et al., 2006, 089788: Larrieu et al., 2007,
15 093031: Oudin et al., 2010, 384790: Simpson et al., 2005, 087438).
16 Two studies (U.S. and Australia) have examined cardiac arrests where emergency services
17 attempted treatment/resuscitation. No evidence of an association between O3 and out-of-hospital
18 cardiac arrest was observed (Dennekamp et al., 2010, 626767: Silverman et al., 2010, 647265).
19 An increasing number of air pollution studies have investigated the relationship between O3
20 concentrations and CVD hospital admissions and/or ED visits. As summarized in the 2006 O3
21 AQCD, some, especially those reporting results stratified by season (or temperature) or comorbid
22 conditions have reported positive associations. However, even studies performing these stratified
23 analyses are not consistent and the overall evidence remains inconclusive regarding the effects of O3
24 on CVD HAs and ED visits. These HA and ED visit studies are summarized in Figures 6-22 through
25 6-26, which are forest plots depicting the associations for studies in which numerical associations
26 were presented for an overall study population. These figures are followed by Tables 6-28 through
27 6-32, giving the numerical results displayed in the figures.
March 2011 6-118 DRAFT-DO NOT CITE OR QUOTE
-------
Reference
Peel etal. (2007)
Fung etal. (2005)
Linnet al.(2000)
Buadong et al. (2009)
Chang etal. (2005)
Chang etal. (2005)
Yang etal. (2004)
Yang etal. (2004)
Wonget al.(1999a)
Wonget al.(1999a)
Wongetal.(1999b)
Wongetal.(1999b)
Wongetal.(1999b)
Halonen etal. (2009)
Middletonetal. (2008)
Larrieuet al. (2007)
Ballesteretal. (2005)
Ballester etal. (2001)
Atkinson etal. (1999)
Prescottetal. (1998)
Polonieckietal. (1997)
Petraeschevskyet al. (2001)
Cakmaketal. (2006)
Larrieuet al. (2007)
von Klot etal. (2005)
Ballester etal. (2005)
Ballester etal. (2001)
Morgan et al. (1998)
Peel etal. (2007)
Bell etal. (2008)
Chanet al.(2006)
Wonget al.(1999a)
Wonget al.(1999b)
Wongetal.(1999b)
Wongetal.(1999b)
Ballester etal. (2001)
Polonieckietal. (1997)
Location
Atlanta, GA
Windsor, Canada
Los Angeles, CA
Bangkok, Thailand
Taipei, Taiwan
Taipei, Taiwan
Kaohsiung.Taiwan
Kaohsiung.Taiwan
Hong Kong
Hong Kong
Hong Kong
Hong Kong
Hong Kong
Helsinki, Finland
Nicosia, Cyprus
8 French cities
14 Spanish cities
Valencia, Spain
London, England
Edinburgh, Scotland-
London, England
Brisbane, Australia
10 Canadian cities
8 French cities
5 European cities
14 Spanish cities
Valencia, Spain —
Sydney, Australia
Atlanta, GA
Taipei, Taiwan
Taipei, Taiwan
Hong Kong
Hong Kong
Hong Kong
Hong Kong
Valencia, Spain—
London, England
Cardiovascular
disease
Cardiac disease
Cerebrovascuiar
disease
n 70
0 90 1 00 110 1 20 1 30 1 40 1 50
Note: Increase in O3 standardized to 20 ppb for 24-h avg period, 30 ppb for 8-h avg period, and 40 ppb for 1-h avg period. Ozone
concentrations in ppb. Seasons depicted by colors - black: all year; red: warm season. Age groups of study populations were not
specified or were adults with the exception of Wellenius et al. (2005, 087483). Fung et al. (2005, 074322). Wong et al. 1999 (1999,
009172)b. and Prescott et al. (1998, 084610). which included only individuals aged 65+.
Figure 6-22. Odds ratio (95% Cl) per increment ppb increase in ozone for over all cardiovascular
ED visits or HAs.
March 2011
6-119
DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-28. Odds ratio (95% Cl) per increment ppb increase in ozone for overall cardiovascular ED
visits or HAs in studies presented in Figure 6-22.
Study
Peel et al. (2007, 0904421
Fung et al. (2006, 0990681
Linnetal. (2006, 099068)
Buadong et al. (2009, 6020601
Chang et al. (2005, 0800861
Yang et al. (2005, 0800861
Wong et al. (2005, 0800861a
Wongetal. (1999, 011 4631b
Halonen et al. (2009, 6257641
Middletonetal. (2008, 156760)
Larrieu et al. (2007, 0930311
Ballesteretal. (2006, 088746)
Ballesteretal. (2006, 088746)
Atkinson et al. (2006, 0990681
Prescott et al. (1998, 0846101
Polonieckietal. (2006, Q99Q68)
Petroeschevsky et al. (2001,
0164661
Cakmak et al. (2006, 0990681
Von Klot et al. (2005, 0880701
Morgan etal. (2008, 0912681
Bell et al. (2008, 0912681
Chan et al. (2006, 0901931
Location
Atlanta, GA
Windsor, Canada
Los Angeles, California
Bangkok, Thailand
Taipei, Taiwan
Kaohsiung, Taiwan
Hong Kong
Hong Kong
Helsinki, Finland
Nicosia, Cyprus
Multicity France
Multicity, Spain
Valencia, Spain
London, England
Edinburgh, Scotland
London, England
Brisbane, Australia
Multicity, Canada
Multicity, Europe
Sydney, Australia
Taipei, Taiwan
Taipei, Taiwan
Outcome
Cardiovascular disease
Cerebrovascular disease
Cardiovascular disease
Cardiovascular disease
Cardiovascular disease
Cardiovascular disease
Cardiovascular disease
Cardiovascular disease
Cerebrovascular disease
Cardiovascular disease
Cerebrovascular disease
Cardiovascular disease
Cardiovascular disease
Cardiac disease
Cardiovascular disease
Cardiac disease
Cardiovascular disease
Cerebrovascular disease
Cardiovascular disease
Cardiovascular disease
Cardiovascular disease
Cerebrovascular disease
Cardiovascular disease
Cardiac disease
Cardiac disease
Cardiac disease
Cerebrovascular disease
Cerebrovascular disease
Averaging Time
8-h warm season
8-h warm season
1-h
24-h
1-h
24-h warm season
24-h cold season
24-h warm season
24-h cold season
24-h
24-h cold season
24-h
24-h
24-h warm season
24-h cold season
24-h
24-h warm season
24-h cold season
8-h max warm season
8-h max
8-h max warm season
8-h warm season
8-h warm season
8-h
8-h
8-h
24-h
8-h
8-h
8-h
1-h max
8-h max warm season
1-h max
24-h
1-h max
Standardized Estimate (96% Cl)
1.00(0.98,
1.02(0.98,
1.02(0.92,
0.99 (0.98,
1.01(1.00,
1.42(1.33
1.15(1.04,
1.33(1.26
1.05(0.96,
1.08(1.03
1.15(1.04,
0.95 (0.90,
1.02(1.03
1.01 (0.96,
1.06(1.02,
0.99 (0.95,
0.98 (0.90,
1.02(0.96,
1.05(0.96,
1.09(1.00,
1.01(0.98,
1.04(1.02
1.04(1.01,
0.94 (0.84,
0.86 (0.72,
1.03(1.00,
0.89 (0.78,
0.97 (0.93,
0.98 (0.95,
0.96 (0.92,
1.02(1.00,
1.11 (1.00,
1.02(0.99,
0.94 (0.87,
1.02(1.01,
1.02)
1.05)
1.13)
1.00)
1.02)
, 1.50)
1.27)
, 1.40)
1.15)
, 1.13)
1.26)
1.01)
, 1.06)
1.06)
1.11)
1.04)
1.08)
1.10)
1.14)
1.18)
1.04)
, 1.06)
1.07)
1.06)
1.47)
1.05)
1.00)
1.01)
1.02)
1.01)
1.04)
1.22)
1.05)
1.02)
1.03)
Note: Increase in O3 standardized to 20 ppb for 24-h averaging period, 30 ppb for 8-h averaging period, and 40 ppb for 1-h averaging
period. Ozone concentrations in ppb. Outcomes were all congestive heart failure, with the exception of Symons et al. (2006, 091258).
which examined onset of congestive heart failure symptoms leading to a heart attack. Age groups of study populations were not specified
or were adults with the exception of Wellenius et al. (2005, 087483). Fung et al. (2006, 099068). Wong et al. (1999, 011463)b. and
Prescott et al. (1998, 084610). which included only individuals aged 65+.
Warm season defined as: March-October (Peel et al., 2007, 090442). May-October (Ballester et al., 2005, 600865; Wong et al., 1999,
011463)b. May-September (Halonen et al., 2009, 625764). April-September (Larrieu et al., 2007, 093031; Von Klot et al., 2005, 088070).
> 20°C (Chang et al., 2005, 080086) and> 25°C (Yang et al., 2004, 094376). Cold season defined as: November-April (Wong et al.,
1999, 011463)b. <20°C (Chang et al., 2005, 080086) and <25°C (Yang et al., 2004, 094376). December-March (Wong et al., 1999,
009172)a
March 2011
6-120
DRAFT - DO NOT CITE OR QUOTE
-------
Reference
Peelet al. (2007)
Welleniusetal. (2005)
Symonset al. (2006)
Stiebetal. (2009)
Yang (2008)
Yang (2008)
Leeet al. (2007)
Leeet al. (2007)
Wonget al.(1999a)
Wonget al.(1999b)
Wonget al.(1999b)
Wonget al.(1999b)
Polonieckietal. (1997)
Location
Atlanta.GA
Allegheny county, PA
Baltimore, MD
7 Canadian cities
Taipei,Taiwan
Taipei,Taiwan
Kaohsiung, Taiwan
Kaohsiung, Taiwan
Hong Kong
Hong Kong
Hong Kong
Hong Kong
London, England
04
06
n 8
1 7
1 4
1 fi
1 8
Note: Increase in O3 standardized to 20 ppb for 24-h averaging period, 30 ppb for 8-h averaging period, and 40 ppb for 1-h averaging
period. Ozone concentrations in ppb. Seasons depicted by colors: black: all year; red: warm season; light blue: cold season. Outcomes
were all congestive heart failure, with the exception of Symons et al. (2006, 091258). which examined onset of congestive heart failure
symptoms leading to a heart attack. Age groups of study populations were not specified or were adults with the exception of Wellenius et
al. (2005, 087483) and (Wong et al., 1999, 011463)b. which included only individuals aged 65+.
Figure 6-23. Odds Ratio (95% Cl) per increment ppb increase in ozone for congestive heart
failure ED visits or HAs.
March 2011
6-121
DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-29. Odds Ratio (95% Cl) per increment ppb increase in ozone for congestive heart failure ED
visits or HAs for studies presented in Figure 6-23
Study
Peel et al. (2007, 090442)
Welleniusetal. (2005, 087483)
Symons et al. (2006, 0912581
Stieb et al. (2009, 195858)
Yang (2008, 1571601
Lee et al. (2007, 1966131
Wongetal. (1999, 0091 72)a
Wongetal. (1999, 011 4631b
Polonieckietal. (1997, Q84Q04)
Location
Atlanta, GA
Allegheny county, PA
Baltimore, MD
Multicity, Canada
Taipei, Taiwan
Kaohsiung, Taiwan
Hong Kong
Hong Kong
London, England
Outcome
congestive heart failure
congestive heart failure
onset of congestive heart failure
symptoms leading to a heart attack
congestive heart failure
congestive heart failure
congestive heart failure
congestive heart failure
congestive heart failure
congestive heart failure
congestive heart failure
congestive heart failure
Averaging Time
8-h warm season
24-h
8-h warm season
24-h
24-h warm season
24-h cold season
24-h warm season
24-h cold season
24-h
24-h
24-h hwarm season
24-hcold season
8-h
Standardized Estimate (95% Cl)
0.96(0.93, 1.00)
0.98(0.96, 1.01)
0.83(0.49, 1.41)
1.03(0.98, 1.07)
1.39(1.27,1.51)
0.61 (0.52, 0.73)
1.25(1.15, 1.36)
1.24(1.09, 1.41)
1.25(1.11, 1.41)
1.11 (1.04, 1.80)
1.09(0.96, 1.23)
1.16(1.06,1.27)
0.99(0.95,1.03)
Note: Increase in O3 standardized to 20 ppb for 24-h averaging period, 30 ppb for 8-h averaging period, and 40 ppb for 1-h averaging
period. Ozone concentrations in ppb. Age groups of study populations were not specified or were adults with the exception of Wellenius
et al. (2005, 087483) and Wong et al. (1999, 011463)b. which included only individuals aged 65.
Warm season defined as: March-October (Peel et al., 2007, 090442). April-November (Symons et al., (2006, 091258). May-October
(Wong et al., (1999, 011463)b > 20°C (Yang, (2008, 157160). and >25°C (Lee et al,(2007, 196613). Cold season defined as: November-
April (Wong et al., (1999, 011463)b. <20°C (Yang, (2008, 157160). and <25°C (Lee et al., (2007, 196613).
March 2011
6-122
DRAFT - DO NOT CITE OR QUOTE
-------
Reference
Peel et al. (2007)
Buadong et al. (2009)
Bell etal. (2008)
Leeet al. (2003)
Leeet al. (2003)
Wonget al. (1999a)
Wonget al. (1999b)
Wonget al. (1999b)
Wongetal.(1999b)
Larrieuet al. (2007)
Atkinson etal. (1999)
Halonen etal. (2009)
Zanobetti et al. (2006)
Stiebetal. (2009)
Buadong et al. (2009)
von Klot etal. (2005)
Lankiet al. (2006)
Polonieckiet al. (1997)
von Klot etal. (2005)
Hosseinpooretal. (2005)
Poloniecki et al. (1997)
Location
Atlanta, GA
Bangkok, Thailand
Taipei, Taiwan
Seoul, Korea
Seoul, Korea
Hong Kong
Hong Kong
Hong Kong
Hong Kong
8 French cities
London, England
Helsinki, Finland
Boston, MA
7 Canadian cities
Bangkok, Thailand
5 European cities
5 European cities
London, England
5 European cities
Tehran, Iran —
London, England
0 5
0 7
n Q
Ischemic heart
disease
Coronary heart
disease
Myocardia! infarction
Angina pectoris
1 1
1 3
1 5
Note: Increase in O3 standardized to 20 ppb for 24-h averaging period, 30 ppb for 8-h averaging period, and 40 ppb for 1-h averaging
period. Ozone concentrations in ppb. Seasons depicted by colors: black: all year; red: warm season; light blue: cold season. Age groups
of study populations were not specified or were adults with the exception of Wong et al. ((1999, 011463)b and Atkinson et al. (2006,
099068). which included only individuals aged 65.
Figure 6-24. Odds Ratio (95% confidence interval) per increment ppb increase in ozone for
myocardial infarction, angina, ischemic heart disease, and coronary heart disease
ED visits or HAs.
March 2011
6-123
DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-30. Odds Ratio (95% Cl) per increment ppb increase in ozone for myocardial infarction, angina,
ischemic heart disease, and coronary heart disease ED visits or HAs for studies presented
in Figure 6-24
Study
Peel et al. (2007, 0904421
Zanobetti and Schwartz (2006, 090195)
Stieb et al. (2009, 1958581
Bell et al. (2008, 0912681
Lee et al. (2003, 0955521
Buadong et al. (2009, 6020601
Wong et al. (2008, 091268la
Wong et al. (2009, 6020601b
Hosseinpooret al. (2005, 0874131
Von Klot et al. (2005, 0880701
Lanki et al. (2006, 0897881
Larrieu et al. (2007, 0930311
Halonen et al. (2009, 6257641
Atkinson et al. (1999, 0078821
Polonieckietal. (1997, 0840041
Location
Atlanta, GA
Boston, MA
Multicity, Canada
Taipei, Taiwan
Seoul, Korea
Bangkok, Thailand
Hong Kong
Hong Kong
Tehran, Iran
Multicity, Europe
Multicity, Europe
Multicity France
Helsinki, Finland
London, England
London, England
Outcome
Ischemic heart disease
Myocardial infarction
Myocardial infarction
Ischemic heart disease
Ischemic heart disease
Ischemic heart disease
Ischemic heart disease
Myocardial infarction
Ischemic heart disease
Ischemic heart disease
Angina
Myocardial infarction
Angina
Myocardial infarction
Ischemic heart disease
Coronary heart disease
Ischemic heart disease
Myocardial infarction
Angina
Averaging Time Standardized Estimate (96% Cl)
8-h warm season
24-h
2-h
24-h
1-h max
1-h max warm season
1-h
1-h
24-h
24-h
24-h warm season
24-h cold season
8-h max
8-h max warm season
8-h max warm season
8-h max warm season
8-h max warm season
8-h max warm season
8-h
8-h
8-h
1.00(0.97,
0.98 (0.92,
1.00(0.96,
1.01(0.91,
1.07(1.02,
1.07(1.00,
1.00(0.98,
0.97 (0.94,
1.03(0.98,
1.01 (0.94,
1.02(0.94,
1.02(0.95,
0.80 (0.70,
1.00(0.83,
1.19(1.05,
0.96 (0.92,
1.02(0.98,
0.99 (0.79,
0.97 (0.94,
0.98 (0.94,
0.98 (0.94,
1.03)
1.03)
1.04)
1.12)
1.13)
1.17)
1.02)
1.01)
1.08)
1.06)
1.11)
1.09)
0.92)
1.21)
1.35)
1.01)
1.07)
1.25)
1.01)
1.02)
1.03)
Note: Increase in 03 standardized to 20 ppb for 24-h averaging period, 30 ppb for 8-h averaging period, and 40 ppb for 1-h averaging period. Ozone concentrations
in ppb. Age groups of study populations were not specified or were adults with the exception of Wong et al. ((1999, 011463lb and Atkinson et al. (2006, 0990681,
which included only individuals aged 65.
Warm season defined as: March-October (Peel et al., (2007, 0904421, June-August (Lee et al, (2003, 0955521, May-September (Halonen et al, (2009, 6257641,
May-October (Wong et al, (2009, 6020601b, and April-September (Lanki et al, (2006, 0897881, Larrieu et al, (2007, 0930311, von Klot et al, (2005, 08807011. Cold
season defined as: November-April (Wong et al, (2009, 6020601b
March 2011
6-124
DRAFT - DO NOT CITE OR QUOTE
-------
Reference
Larrieuet al. (2007)
Halonen et al. (2009)
Chanet al. (2006)
Villeneuveet al. (2006)
Villeneuveet al. (2006)
Villeneuveet al. (2006)
Chanet al. (2006)
Villeneuveet al. (2006)
Villeneuveet al. (2006)
Villeneuveet al. (2006)
Chanet al. (2006)
Villeneuveet al. (2006)
Villeneuveet al. (2006)
Villeneuveet al. (2006)
Location
8 French cities
Helsinki, Finland
Taipei. Taiwan
Edmonton, Canada
Edmonton, Canada
Edmonton, Canada
Taipei, Taiwan
Edmonton, Canada
Edmonton, Canada
Edmonton, Canada
Taipei, Taiwan
Edmonton, Canada
Edmonton, Canada
Edmonton, Canada
0.5
0.7
0.9
All
Ischemic
Hemorrhagic
Transient
ischemic
1.1
1.5
Note: Increase in O3 standardized to 20 ppb for 24-h averaging period, 30 ppb for 8-h averaging period, and 40 ppb for 1-h averaging
period. Ozone concentrations in ppb. Seasons depicted by colors: black: all year; red: warm season; light blue: cold season. Age groups of
study populations were not specified or were adults with the exception of Villeneuve et al. (2006, 090191). which included only
individuals aged 65+, and Chan et al. (2006, 090193). which included only individuals aged 50+.
Figure 6-25. Odds Ratio (95% confidence interval) per increment ppb increase in ozone for
stroke ED visits or HAs.
March 2011
6-125
DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-31. Odds Ratio (95% Cl) per increment ppb increase in ozone for stroke ED visits or HAs for
studies presented in Figure 6-25
Study
Villeneuve et al. (2006, 090191;
Chan et al. (2006, 0901931
Larrieu et al. (2007, 0930311
Halonen et al. (2009, 625764)
Location Outcome
Ischemic stroke
Ischemic stroke
Ischemic stroke
Hemorrhagic stroke
I Edmonton, Canada Hemorrhagic stroke
Hemorrhagic stroke
Transient ischemic stroke
Transient ischemic stroke
Transient ischemic stroke
All/non-specified stoke
Taipei, Taiwan Ischemic stroke
Hemorrhagic stroke
Multicity, France All/non-specified stoke
Helsinki, Finland All/non-specified stoke
Averaging Time
24-h
24-h warm season
24-h cold season
24-h
24-h warm season
24-h cold season
24-h
24-h warm season
24-h cold season
1-h max
1-h max
1-h max
8-h max warm season
8-h max warm season
Standardized Estimate (96% Cl)
1.00(0.88,1.13)
1.09(0.91, 1.32)
0.98(0.80, 1.18)
1.02(0.87,1.20)
1.12(0.88,1.43)
0.97(0.76, 1.22)
0.98(0.87, 1.10)
0.85(0.70,1.01)
1.11 (0.93, 1.32)
1.01 (0.99,1.03)
1.03(0.99, 1.07)
0.99(0.92,1.06)
0.98(0.93,1.02)
1.08(0.83, 1.41)
Note: Increase in 03 standardized to 20 ppb for 24-h averaging period, 30 ppb for 8-h averaging period, and 40 ppb for 1-h averaging period. Ozone concentrations
in ppb.
Warm season defined as: May-September (Halonen et al., (2009, 6257641, and April-September (Larrieu etal., 2007, 093031)(Villeneuveetal., 2006, 0901911. Cold
season defined as: October-March (Villeneuve et al, 2006, 0901911.
Reference
Location
Peel et al. (2007) Atlanta. GA
Stieb et al. (2009) 7 Canadian cities
Buadong et al. (2009) Bangkok,Thailand
Wong et al. (1999b) Hong Kong
Wonget al. (1999b) Hong Kong
Wong et al. (1999b) Hong Kong
Halonen et al. (2009) Helsinki, Finland
Poloniecki et al. (1997) London, England
n 70
o an
Dysrhythmia
Arrhythmia
n 90 1 tin 1 10 1 ?o 1 an 14n
Note: Increase in O3 standardized to 20 ppb for 24-h averaging period, 30 ppb for 8-h averaging period, and 40 ppb for 1-h averaging
period. Ozone concentrations in ppb. Seasons depicted by colors: black: all year; red: warm season; light blue: cold season. Age groups
of study populations were not specified or were adults with the exception of Wong et al. (1999, 011463)b. which included only
individuals aged 65.
Figure 6-26. Odds Ratio (95% confidence interval) per increment ppb* increase in ozone for
arrhythmia and dysrhythmia ED visits or HAs.
March 2011
6-126
DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-32. Odds Ratio (95% Cl) per increment ppb* increase in ozone for arrhythmia and dysrhythmia
ED visits or HAs for studies presented in Figure 6-26
Study
Peel et al. (2007, 0904421
Stieb et al. (2009, 195858)
Buadong et al. (2009, 6020601
Wong et al. (2009, 6Q2060)b
Halonen et al. (2009, 6257641
Polonieckietal. (2009, 602060)
Location
Atlanta, GA
Multicity, Canada
Bangkok, Thailand
Hong Kong
Helsinki, Finland
London, England
Outcome
Dysrhythmia
Dysrhythmia
Arrhythmia
Arrhythmia
Arrhythmia
Arrhythmia
Averaging Time
8-h warm season
24-h
1-h
24-h
24-h warm season
24-h cold season
8-h max warm season
8-h
Standardized Estimate (96% Cl)
1.01(0.98,1.04)
1.02(0.95, 1.09)
0.99(0.95, 1.04)
1.06(0.99,1.12)
1.10(0.96,1.26)
1.11 (1.01, 1.23)
1.04(0.80, 1.35)
1.02(0.96,1.07)
Note: Increase in 03 standardized to 20 ppb for 24-h averaging period, 30 ppb for 8-h averaging period, and 40 ppb for 1-h averaging period. Ozone
concentrations in ppb. Age groups of study populations were not specified or were adults with the exception of (Wong et al., 1999, 011463)b. which included
only individuals aged 65. Warm season defined as: March-October (Peel et al., 2007, 090442V May-October (Wong et al., 1999, 011463lb and May-
September (Halonen etal., 2009, 625764V Cold season defined as: November-April (Wong et al., 1999, 011463lb.
6.3.2.8. Cardiovascular Mortality
1 The 2006 O3 AQCD provided evidence, primarily from single-city studies, of consistent
2 positive associations between short-term O3 exposure and cardiovascular mortality. Recent multicity
3 studies conducted in the U.S., Canada, and Europe further confirm the association between short-
4 term O3 exposure and cardiovascular mortality.
5 As discussed in Section 6.2.7.2, the APHENA study (Katsouyanni et al., 2009, 199899) found
6 consistent positive associations for cardiovascular mortality in all-year analyses with associations
7 persisting in analyses restricted to the summer season. Additional multicity studies from the U.S.
8 (Zanobetti and Schwartz, 2008, 101596). Europe (Samoli et al., 2009, 195855). and Italy (Stafoggia
9 et al., 2010, 625034) that conducted summer season analyses provide additional support for an
10 association between short-term O3 exposure and cardiovascular mortality.
11 Of the studies evaluated, only the APHENA study (Katsouyanni et al., 2009, 199899) and the
12 Italian multicity study (Stafoggia et al., 2010, 625034) conducted an analysis of the potential for
13 co-pollutant confounding of the O3-cardiovascular mortality relationship. In the European dataset,
14 when focusing on the natural spline model with 8 df/year (Section 6.2.7.2) and lag 1 results in order
15 to compare results across study locations (Section 6.6.2.1), cardiovascular mortality risk estimates
16 were robust to the inclusion of PMi0 in co-pollutant models in all-year analyses with more variability
17 in the Canadian and U.S. datasets (i.e., cardiovascular O3 mortality risk estimates were reduced or
18 increased in co-pollutant models). In summer season analyses, cardiovascular O3 mortality risk
19 estimates were robust in the European dataset and attenuated but remained positive in the U.S.
20 dataset. Similarly, in the Italian multicity study (Stafoggia et al., 2010, 625034). which was limited
21 to the summer season, cardiovascular mortality risk estimates were robust to the inclusion of PMi0 in
22 co-pollutant models. Based on the APHENA and Italian multicity results, O3 cardiovascular
23 mortality risk estimates appear to be robust to inclusion of PMi0 in co-pollutant models. However, in
24 the U.S. and Canadian datasets there was evidence that O3 cardiovascular mortality risk estimates are
March 2011
6-127
DRAFT - DO NOT CITE OR QUOTE
-------
1 moderately to substantially sensitive (e.g., increased or attenuated) to PMi0. The mostly every-6th-
2 day sampling schedule for PMi0 in the Canadian and U.S. datasets greatly reduced their sample size
3 and limits the interpretation of these results.
6.3.2.9. Summary of Epidemiologic Studies
4 Overall, the available body of evidence examining the relationship between short-term
5 exposures to O3 and cardiovascular morbidity is inconsistent. Differences in exposure metrics and
6 windows of exposure, a wide variety of biomarkers considered, and a lack of consistency among
7 definitions used for specific cardiovascular disease endpoints (e.g. arrhythmias, HRV) make
8 comparisons across studies difficult. In addition, several investigators reporting adverse effects of O3
9 discuss the possibility that O3 may be acting as a proxy for sulfate; differences reported across
10 multicity studies and across studies conducted in specific cities/regions point to the importance of
11 considering multi-pollutant relationships that vary across geographic regions. An association
12 between O3 and cardiovascular mortality has been observed.
6.3.3. Toxicology
6.3.3.1. Summary of Findings from Previous Ozone AQCDs
13 In the previous O3 AQCDs (U.S. EPA, 1996, 017831: U.S. EPA, 2006, 088089) experimental
14 animal studies have reported relatively few cardiovascular system alterations after exposure to O3
15 and other photochemical oxidants. The limited amount of research directed at examining O3-induced
16 cardiovascular effects has primarily found alterations in heart rate (HR) and BP after O3 exposure. A
17 group of studies (Arito et al, 1990, 042285; Arito et al, 1992, 042759; Uchiyama and Yokoyama,
18 1989, 042051; Uchiyama et al., 1986, 040883; Yokoyama et al., 1989, 041741) report O3
19 (0.1-1.0 ppm) exposure in rats decreased core temperature (TCo), HR, and mean arterial pressure
20 (MAP). However, these cardiovascular responses to O3 could be attenuated by increased ambient
21 temperatures and were the result of the rodent hypothermic response (Watkinson et al., 1993,
22 043205; Watkinson et al., 2003, 050547). This hypothermic response could be an attempt to
23 minimize the irritant effects of O3 inhalation, serving as a physiological and behavioral defense
24 mechanism (Arito et al., 1997, 082671; Iwasaki et al., 1998, 086165). As decreased HR, MAP, and
25 TOO have not been observed in humans except at very high O3 exposures, caution must be used in
26 extrapolating the results of these animal studies to humans (Section 6.3.1).
27 Other studies have shown that O3 can increase BP in multiple animal models. Dogs treated
28 with 1.0 ppm O3 daily for 17 months developed excessive systolic pressure and pulmonary arterial
29 hypertension (Bloch et al., 1971, 015914). Additionally, rats exposed to 0.6 ppm O3 for 33 days had
30 increased systolic pressure and HR (Revis et al., 1981, 040017). Increased BP triggers the release of
31 atrial natriuretic factor (ANF), which has been found in increased levels in the heart, lungs, and
32 circulation of O3 exposed (0.5 ppm) rats (Vesely et al., 1994, 076015; Vesely et al., 1994, 028877;
March 2011 6-128 DRAFT - DO NOT CITE OR QUOTE
-------
1 Vesely et al., 1994, 076228). High concentration O3 exposure has also been found to lead to heart
2 and lung edema (Friedman et al., 1983, 040547). which could be the result of increased ANF levels.
3 Thus, O3 may increase blood pressure and HR, leading to increased ANF and tissue edema.
4 The toxicological studies that have examined the effect of O3 on the cardiovascular system
5 clearly demonstrate O3-induced responses, but it remains unclear if the mechanism is through a
6 reflex response or due to O3 reaction products, which have been sparsely studied. Oxysterols derived
7 from cholesterol ozonation, such as (3-epoxide and 5|3,6(3-epoxycholesterol (and its metabolite
8 cholestan-6-oxo-3,5-diol), have been implicated in inflammation associated with cardiovascular
9 disease (Pulfer and Murphy, 2004, 076673: Pulfer et al., 2005, 076663). Two other cholesterol
10 ozonolysis products, atheronal-A and -B (e.g. cholesterol secoaldehyde), have been found in human
11 atherosclerotic plaques and shown in vitro to induce foam cell formation and induce cardiomyocyte
12 apoptosis and necrosis (Sathishkumar et al., 2005, 087958; Wentworth et al., 2003, 052486);
13 however, these products have not been found in the lung compartment or systemically after O3
14 exposure. The ability to form these cholesterol ozonation products in the circulation in the absence of
15 O3 exposure complicates their implication in O3 induced cardiovascular disease.
16 Although it has been proposed that O3 reaction products released after the interaction of O3
17 with RTLF constituents (See Section 5.1.2 on O3 interaction with RTLF) are responsible for systemic
18 effects, it is not known whether they gain access to the vascular space. Alternatively, extrapulmonary
19 release of diffusible mediators, such as cytokines or endothelins, may initiate or propagate
20 inflammatory responses in the vascular or systemic compartments (Cole and Freeman, 2009,
21 597507) (Section 5.1.9.1). Ozone reacts within the lung to amplify ROS production, induce
22 pulmonary inflammation, and activate inflammatory cells, resulting in a cascading proinflammatory
23 state and extrapulmonary release of diffusible mediators that could lead to cardiovascular injury.
6.3.3.2. Recent Ozone-induced Cardiovascular Effects
24 According to recent short-term O3 exposure animal toxicology studies, O3 plays a role in
25 inducing vascular oxidative stress and proinflammatory mediators, altering HR and HRV, and
26 regulating the pulmonary endothelin system. A number of these effects were variable between strains
27 examined, suggesting a genetic component to development of O3 induced cardiovascular effects.
28 Further, new studies provide evidence that extended O3 exposure enhances susceptibility to
29 ischemia-reperfusion (I/R) injury and atherosclerotic lesion development. Still, few studies have
30 investigated the role of O3 reaction products in these processes, but more evidence is provided for
31 elevated inflammatory and reduction-oxidation (redox) cascades known to initiate these
32 cardiovascular pathologies.
33 A recent study in young mice (C57B1/6, 6 week old) and rhesus monkeys (Macaca mulatto,
34 180 days old) examined the effects of 1 or 5 days (8 h/day) of O3 (0.5 ppm) exposure on a number of
35 cardiovascular endpoints (Chuang et al., 2009, 197202). Mice exposed to O3 for 5 days had
36 increased heart rate (HR) as well as mean and diastolic blood pressure. Increased blood pressure
37 could be explained by the inhibition in endothelial-dependent (acetylcholine) vasorelaxation from
March 2011 6-129 DRAFT - DO NOT CITE OR QUOTE
-------
1 decreased bioavailability of aortic nitric oxide (-NO). Ozone caused a decrease in aortic NOX (nitrite
2 and nitrate levels) and a decrease in total, but not phosphorylated, endothelial nitric oxide synthase
3 (eNOS). Ozone also increased vascular oxidative stress in the form of increased aortic and lung lipid
4 peroxidation (F2-isoprostane), increased aortic protein nitration (3-nitrotyrosine), decreased aortic
5 superoxide dismutase (SOD2) protein and activity, and decreased aortic aconitase activity, indicating
6 specific inactivation by O2~ and ONOO". Mitochondrial DNA (mtDNA) damage was also used as a
7 measure of oxidative and nitrative stress in mice and infant rhesus monkeys exposed to O3 (0.5 ppm)
8 for 5 days (8 h/day). Chuang et al. (2009, 197202) observed that MtDNA damage accumulated in the
9 lung and aorta of mice after 1 and 5 days of O3 exposure and in the proximal and distal aorta of O3
10 treated nonhuman primates. Additionally, ApoE"'" mice (6-14 weeks old) exposed to O3 (0.5 ppm) for
11 8 weeks (5 days/week, 8 h/day) had increased aortic atherosclerotic lesion area (Section 7.3.1),
12 which may be associated with the short-term exposure changes discussed. Overall, this study
13 suggests that O3 initiates an oxidative environment by increasing O2~ production, which leads to
14 mtDNA damage and -NO consumption, known to perturb endothelial function (Chuang et al., 2009,
15 197202). Endothelial dysfunction is characteristic of early and advanced atherosclerosis and
16 coincides with impaired vasodilation and blood pressure regulation.
17 Vascular occlusion resulting from atherosclerosis can block blood flow causing ischemia. The
18 restoration of blood flow in the vessel or reperfusion can cause injury to the tissue from subsequent
19 inflammation and oxidative damage. Perepu et al. (2010, 385020) observed that O3 exposure
20 (0.8 ppm for 28 or 56 days) enhanced the sensitivity to myocardial ischemia-reperfusion (I/R) injury
21 in Sprague-Dawley rats while increasing oxidative stress levels and pro-inflammatory mediators and
22 decreasing production of anti-inflammatory proteins. Ozone was also found to decrease the left
23 ventricular developed pressure, rate of change of pressure development, and rate of change of
24 pressure decay while increasing left ventricular end diastolic pressure in isolated perfused hearts. In
25 this ex vivo heart model, O3 induced oxidative stress by decreasing SOD enzyme activity and
26 increasing malondialdehyde levels. Ozone also elicited a proinflammatory state which was evident
27 by an increase in TNF-a and a decrease in the anti-inflammatory cytokine IL-10. Perepu et al. (2010,
28 385020) concluded that O3 exposure may result in a greater I/R injury.
Heart Rate and Heart Rate Variability
29 Strain differences in HR and HRV have been observed in response to a 2-h O3 (0.584 ppm)
30 pretreatment followed by a 3-h exposure to particulate matter (carbon black (CB), 536 ug/m3) in 18-
31 to 20-week-old mice (C3H/HeJ [HeJ], C57BL/6J [B6], and C3H/HeOuJ [OuJ]) (Hamade and
32 Tankersley, 2009, 596386; Hamade et al., 2008, 156515). These mice were chosen from prior studies
33 on lung inflammatory and hyperpermeability responses to be susceptible (B6 and OuJ) and resistant
34 (HeJ) to O3-induced health effects (Kleeberger et al., 2000, 014895). HR decreased during O3 pre-
35 exposure for all strains, but recovered during the CB exposure (Hamade et al., 2008, 156515). This is
36 contrary to the tachycardia that was reported in 6-week-old B6 mice treated on 1 or 5 days with O3,
37 as described above (Chuang et al., 2009, 197202). Percent change in HRV parameters, SDNN
March 2011 6-130 DRAFT - DO NOT CITE OR QUOTE
-------
1 (indicating total HRV) and rMSSD (indicating beat-to-beat HRV), were increased in both C3H mice
2 strains, but not B6 mice, during O3 pre-exposure and recovered during CB exposure when compared
3 to the filtered air group. The two C3H strains differ by a mutation in the toll-like receptor 4 (TLR4)
4 gene, but these effects did not seem to be related to this mutation since similar responses were
5 observed. Hamade et al. (2008, 156515) speculate that the B6 and C3H strains differ in mechanisms
6 of HR response after O3 exposure between withdrawal of sympathetic tone and increase of
7 parasympathetic tone; however, no direct evidence for this conclusion was reported. The strain
8 differences observed in HR and HRV suggest that genetic variability affects cardiac responses after
9 acute air pollutant exposures.
10 Hamade and Tankersley (2009, 596386) continued this investigation of gene-environment
11 interactions on cardiopulmonary adaptation of O3 and CB induced changes in HR and HRV using the
12 prior daily exposure scheme for 3 consecutive days. By comparing day-1 interim values it is possible
13 to observe that O3 exposure increased SDNN and rMSSD, but decreased HR in all strains. Measures
14 of HR and HRV in B6 and HeJ mice recovered to levels consistent with filtered air treated mice by
15 day 3; however, these responses in OuJ mice remained suppressed. B6 mice had no change in
16 respiratory rate (RR) after O3 treatment, whereas HeJ mice on days 1 and 2 had increased RR and
17 OuJ mice on days 2 and 3 exhibited increased RR. VT did not change with treatment among the
18 strains. Overall, B6 mice were mildly responsive with rapid adaptation, whereas C3 mice were
19 highly responsive with adaptation only in HeJ mice with regards to changes in cardiac and
20 respiratory responses. HR and HRV parameters were not equally correlated with VT and RR between
21 the three mice strains, which suggest that strains vary in the integration of the cardiac and respiratory
22 systems. These complex interactions could help explain variability in interindividual susceptibility to
23 adverse health effects of air pollution.
24 Hamade et al. (2010, 666324) expanded their investigation to explore the variation of these
25 strain dependent cardiopulmonary responses with age. As was observed previously, all experimental
26 mouse strains (B6, HeJ, and OuJ) exhibited decreased HR and increased HRV after O3 (0.58 ppm,
27 2 hours) exposure. Younger O3 exposed mice (5 months) had a significantly lower HR compared to
28 older exposed mice (12 months), indicating an attenuation of the bradycardic effect of O3 with age.
29 Younger mice also had a greater increase in rMSSD in HeJ and OuJ strains and SDNN in HeJ mice.
30 Conversely, B6 mice had a slightly greater increase in SDNN compared to the aged mice. No change
31 was observed in the magnitude of the O3 induced increase of SDNN in OuJ mice or rMSSD in B6
32 mice. The B6 and HeJ mice genetically vary in respect to the nuclear factor erythroid 2-related factor
33 2 (Nrf-2). The authors propose that the genetic differences between the mice strains could be altering
34 the formation of ROS, which tends to increase with age, thus modulating the changes in
35 cardiopulmonary physiology after O3 exposure.
36 Strain and age differences in HR and heart function were further investigated in B6 and
37 12981/SvlmJ (129) mice in response to a sequential O3 (2 hours, 0.576 ppm) and filtered air or CB
38 (3 hours, 556 ug/m3) exposure (Tankersley et al., 2010, 628062). Young (5 months) 129 mice
39 showed a decrease in HR after O3 or O3 and CB exposure. This bradycardia was not observed in B6
March 2011 6-131 DRAFT - DO NOT CITE OR QUOTE
-------
1 or older animals (18 months) in this study, suggesting a possible alteration or adaptation of the
2 autonomic nervous system activity with age. However, these authors did previously report
3 bradycardia in similarly aged young B6 mice (Hamade and Tankersley, 2009, 596386; Hamade et al,
4 2008, 156515; Hamade et al., 2010, 666324). Ozone exposure in 129 mice also resulted in an
5 increase in left ventricular chamber dimensions at end diastole (LVEDD) in young and old mice and
6 a decrease in left ventricular posterior wall thickness at end systole (PWTES) in older mice. The
7 increase in LVEDD caused a decrease in fractional shortening, which can be used as a rough
8 indicator of left ventricular function. Regression analysis revealed a significant interaction between
9 age and strain on HR and PWTES, which implies that aging affects the HR and function in response
10 to O3 differently between mouse strains.
Ozone-Induced Effects on Cardiovascular-Related Proteins
11 Increased BP, changes in HRV, and increased atherosclerosis may be related to increases in the
12 vasoconstrictor peptide, endothelin-1 (amino acids 1-21, ET-l[i_2i]). Regulation of the pulmonary
13 endothelin system can be affected in rats (Fischer 344) by inhalation (4 hours) of PM (0, 5,
14 50 mg/m3, EHC-93) and O3 (0, 0.4, or 0.8 ppm) (Thomson et al., 2005, 087554; Thomson et al.,
15 2006, 097483). Exposure to either O3 (0.8 ppm) or PM increased plasma ET-l[i_2i], ET-3[i_2i], and the
16 ET-1 precursor peptide, bigET-1. Increases in circulating ET-l[i_2i] could be a result of a transient
17 increase in the gene expression of lung preproET-1 and endothelin converting enzyme-1 (ECE-1)
18 immediately following inhalation of O3 or PM. These latter gene expression changes (e.g. preproET-
19 1 and ECE-1) were additive with co-exposure to O3 and PM. Conversely, preproET-3 decreased
20 immediately after O3 exposure, suggesting the increase in ET-3[i_2i] was not through de novo
21 production. A recent study also found increased ET-1 gene expression in the aorta of acutely exposed
22 rats (O3, 1.0 ppm, 5 h/day, 2 days) (Kodavanti et al., In Press, 666323). These rats also exhibited an
23 increase in ETBR after O3 exposure; however, they did not demonstrate increased biomarkers for
24 vascular inflammation, thrombosis, or oxidation.
25 O3 can oxidize protein functional groups and disturb the affected protein. For example, the
26 soluble plasma protein fibrinogen is oxidized by O3 (0.01-0.03 ppm) in vitro, creating fibrinogen and
27 fibrin aggregates, characteristically similar to defective fibrinogen (Rosenfeld et al., 2009, 201546;
28 Rozenfeld et al., 2008, 596413). In these studies, oxidized fibrinogen retained the ability to form
29 fibrin gels that are involved in coagulation, however the aggregation time increased and the gels
30 were rougher than normal with thicker fibers. Oxidized fibrinogen also developed the ability to self
31 assemble creating fibrinogen aggregates that may play a role in thrombosis. Since O3 does not
32 readily translocate past the ELF and pulmonary epithelium and fibrinogen is primarily a plasma
33 protein, it is uncertain if O3 would have the opportunity to react with plasma fibrinogen. However,
34 fibrinogen can be released from the basolateral face of pulmonary epithelial cells during
35 inflammation, where the deposition of fibrinogen could lead to lung injury (Lawrence and Simpson-
36 Haidaris, 2004, 627654).
March 2011 6-132 DRAFT - DO NOT CITE OR QUOTE
-------
Cardiovascular Effects due to Ozone Reaction Products
1 Although recent toxicological studies have demonstrated O3-induced effects on the
2 cardiovascular system, as concluded in previous O3 AQCDs, it remains unclear if the mechanism is
3 through a reflex response or the result of effects from O3 reaction products (U.S. EPA, 1996, 017831;
4 U.S. EPA, 2006, 088089). A new study that examined O3 reaction byproducts has shown that
5 cholesterol secoaldehyde (e.g., atheronal A) induces apoptosis in vitro in mouse macrophages (Gao
6 et al, 2009, 200764) and cardiomyocytes (Sathishkumar et al, 2009, 201549). Additionally,
7 atheronal-A and -B has been found to induce in vitro macrophage and endothelial cell
8 proinflammatory events involved in the initiation of atherosclerosis (Takeuchi et al., 2006, 197793).
9 These O3 reaction products when complexed with low density lipoprotein upregulate scavenger
10 receptor class A and induce dose-dependent macrophage chemotaxis. Atheronal-A increases
11 expression of the adhesion molecule, E-selectin, in endothelial cells, while atheronal-B induces
12 monocyte differentiation. These events contribute to both monocyte recruitment and foam cell
13 formation in atherosclerotic vessels. It is unknown whether these O3 reaction products gain access to
14 the vascular space from the lungs. Alternative explanations include the extrapulmonary release of
15 diffusible mediators that may initiate or propagate inflammatory responses in the vascular or
16 systemic compartments.
Summary of Toxicological Studies
17 Overall, animal studies suggest that O3 exposure may disrupt both the -NO and endothelin
18 systems, which can result in an increase in HR, HRV, and ANF, as is observed after O3 exposure.
19 Studies in rodents also exhibit O3 induced bradycardia, but it is uncertain if this effect is also
20 observed in humans. Additionally, O3 may increase oxidative stress and vascular inflammation
21 promoting the progression of atherosclerosis and leading to increased susceptibility to I/R injury. As
22 O3 reacts quickly with the ELF and does not translocate to the heart and large vessels, studies
23 suggest that the cardiovascular effects exhibited could be caused by reaction byproducts of O3
24 exposure. However, direct evidence of translocation of O3 reaction products to the cardiovascular
25 system has not been demonstrated in vivo. Alternatively, extrapulmonary release of diffusible
26 mediators, such as cytokines or endothelins, may initiate or propagate inflammatory responses in the
27 vascular or systemic compartments leading to the reported cardiovascular pathologies.
6.3.4. Summary and Causal Determination
28 In past O3 AQCDs the effects of O3 to the cardiovascular system did not receive much
29 attention due to the paucity of information available. However, in recent years, investigation of O3-
30 induced cardiovascular events has advanced. In general, compared with the epidemiologic evidence,
31 the toxicological evidence is more supportive of an O3-induced cardiovascular effects.
32 Epidemiologic evidence does not consistently demonstrate a positive relationship between short-
33 term O3 exposure and cardiovascular-related morbidity. However, most epidemiologic studies have
March 2011 6-133 DRAFT - DO NOT CITE OR QUOTE
-------
1 not extensively investigated the cardiovascular effects of O3 exposure in susceptible populations,
2 which may further support the toxicological findings. Although the epidemiologic evidence of
3 cardiovascular morbidity is limited, single-city studies reviewed in the 2006 O3 AQCD (U.S. EPA,
4 2006, 088089). recent multicity studies, and the multicontinent APHENA study provide evidence of
5 consistently positive associations between short-term O3 exposure and cardiovascular mortality.
6 However, in contrast with respiratory effects, there is weak coherence between associations for
7 cardiovascular morbidity and mortality. Further, there is no apparent biological mechanism to
8 explain the association observed for short-term O3 exposure with cardiovascular mortality but not
9 with cardiovascular morbidity.
10 Animal toxicological studies provide evidence for O3-induced cardiovascular effects,
11 specifically enhanced I/R injury, disrupted NO-induced vascular reactivity, decreased cardiac
12 function, and increased HRV. The observed increase in HRV is supported by a recent controlled
13 human exposure study that also finds increased high frequency HRV, but not altered blood pressure,
14 following O3 exposure. Toxicological studies investigating the role of O3 in heart rate regulation are
15 mixed with both bradycardic and tachycardic responses observed. These changes in cardiac function
16 provide evidence for O3-induced alterations in the autonomic nervous system leading to
17 cardiovascular complications. Epidemiological studies showing positive association between O3 and
18 arrhythmias confirm the development of autonomic dysfunction following O3 exposure. It is still
19 uncertain how O3 inhalation may cause systemic toxicity; however the cardiovascular effects of O3
20 found in animals correspond to the development and maintenance of an extrapulmonary oxidative,
21 proinflammatory environment.
22 In conclusion, animal toxicological studies provide stronger evidence for O3 exposure leading
23 to cardiovascular morbidity than do epidemiologic studies, among which there is a lack of coherence
24 among endpoints. Based on the relatively strong body of toxicological evidence, and the consistent
25 evidence of an association between O3 and cardiovascular mortality, but weak coherence and
26 biological plausibility for O3-induced cardiovascular morbidity, the generally limited body of
27 evidence is suggestive of a causal relationship between relevant short-term exposures to O3 and
28 cardiovascular effects.
6.4. Central Nervous System Effects
29 The 2006 O3 AQCD (U.S. EPA, 2006, 088089) included toxicological evidence that acute
30 exposures to O3 are associated with alterations in neurotransmitters, motor activity, short and long
31 term memory, and sleep patterns. Additionally, histological signs of neurodegeneration have been
32 observed. Research in the area of O3-induced neurotoxicity has notably increased over the past few
33 years, with the majority of the evidence coming from toxicological studies that examined the
34 association between O3 exposure and neurobehavioral effects, and more limited evidence from
35 epidemiologic studies. In an epidemiologic study conducted by Chen and Schwartz (2009, 179945).
36 data from the NHANES III cohort was utilized to study the relationship between long-term O3
March 2011 6-134 DRAFT - DO NOT CITE OR QUOTE
-------
1 exposure (mean annual O3 concentration of 26.5 ppb) and neurobehavioral effects among adults aged
2 20-59 years. The authors observed an association between annual exposure to O3 and tests measuring
3 coding ability and attention/short-term memory. Each 10-ppb increase in annual O3 levels
4 corresponded to an aging-related cognitive performance decline of 3.5 years for coding ability and
5 5.3 years for attention/short-term memory. These associations persisted in both crude and adjusted
6 models. There was no association between annual O3 concentrations and reaction time tests. The
7 authors conclude that overall there is a positive association between O3 exposure and reduced
8 performance on neurobehavioral tests. Although Chen and Schwartz (2009, 179945) is a long-term
9 exposure study, it is included in this section because it is the first epidemiologic study to demonstrate
10 that exposure to ambient O3 is associated with decrements in neurocognitive tests related to memory
11 and attention in humans. This epidemiologic evidence of an adverse effect on the CNS due to
12 exposure to ambient concentrations of O3 is coherent with animal studies demonstrating that
13 exposure to O3 can produce a variety of CNS effects including behavioral deficits, morphological
14 changes, and oxidative stress in the brains of rodents. In these rodent studies, interestingly, CNS
15 effects were reported at O3 concentrations that were generally lower than those concentrations
16 commonly observed to produce adverse pulmonary or cardiac effects in rats.
17 A number of new studies demonstrate various perturbations in neurologic function or
18 histology, including changes consistent with Parkinson's and Alzheimer's disease pathologies.
19 In a subchronic study, rats were exposed to 0.25 ppm O3 for 4 h/day for 15-90 days (Rivas-
20 Arancibia et al, 2010, 201544). The exposures caused a complex array of responses, including a
21 time-dependent increase in lipid peroxidation products and immunohistochemical changes in the
22 hippocampus, a region of the brain which is important for higher cognitive function including
23 memory acquisition, that were correlated with decrements in passive avoidance behavioral tests. The
24 study showed progressive neurodegeneration, and statistically significant decreases in both short and
25 long-term memory after 15 days of exposure to 0.25 ppm O3. Oxidative stress has also been
26 observed in the striatum and substantia nigra of rats after 15 days of exposure to 0.25 ppm O3 for
27 4 h/day (Pereyra-Munoz et al., 2006, 596408). Perturbed oxidative balance has been observed in
28 multiple regions of the brains after 15 days of exposure to 0.75 ppm O3 (Calderon Guzman et al.,
29 2006, 596371). and these changes were dependent on the nutritional status of the rats (high versus
30 low protein diet). For example, O3 produced an increase in glutathione in rats fed the high protein
31 diet but decreases in glutathione in rats fed low protein chow.
32 Martinez-Canabal et al. (2008, 194376) showed exposure of rats to 0.25 ppm, 4h/day, for 7,
33 15, or 30 days increased lipoperoxides in the hippocampus. This effect was observed at day 7 and
34 continued to increase with time, indicating cumulative oxidative damage. The study also observed a
35 loss of neurons and increased expression of COX-2, which has a role in neurodegenerative disease
36 and is observed in the tissues of Alzheimer's patients. Consistent with Alzheimer's incidence in the
37 elderly, O3-induced changes in lipid peroxidation and COX-2 positive cells in the hippocampus
38 could be significantly inhibited by daily treatment with growth hormone (GH). The protective effect
39 of GH on O3-induced oxidative stress was greatest on COX-2 after 15 days of O3 exposure.
March 2011 6-135 DRAFT - DO NOT CITE OR QUOTE
-------
1 Consistent with these findings, lipid peroxidation in the hippocampus of rats was observed to
2 increase significantly after a 30-day exposure to 0.25 ppm O3, but not after a single 4-h exposure to
3 the same concentration (Mokoena et al., 2010, 677667). However, 4 hours of exposure was sufficient
4 to cause significant increases in lipid peroxidation when the concentration was increased to 0.7 ppm
5 O3. Acute exposure to 0.7 ppm O3 and prolonged exposure (30 days) to 0.25 ppm O3 resulted in
6 reduced efficacy of an antidepressant (imipramine).
7 A protective effect of estradiol has been observed in ovariectomized female rats exposed to
8 0.25 ppm O3 (4 h/day) for 30 or 60 days (Guevara-Guzman et al., 2009, 596385). In the olfactory
9 bulb, lipid peroxidation was significantly less in rats exposed to O3 and treated daily with estradiol.
10 This protective effect of estradiol was also demonstrated for O3-induced decrements in a selective
11 olfactory recognition memory test and an olfactory-dependent reward test. Similarly, estradiol
12 protected against O3-induced changes in nigral cell morphology and loss of dopamine neurons in rats
13 exposed to O3 for 30 days (Angoa-Perez et al., 2006, 596366). Thus, repeated exposure of rats to O3
14 produces lipid peroxidation at multiple sites in the brain and this oxidative stress is accompanied by
15 gene expression changes and decrements in behavioral tests. Olfactory changes and loss of
16 substantia nigra neurons are associated with Parkinson's disease in humans. Inhibition of these
17 effects with estradiol treatment is consistent with the higher incidence of Parkinson's disease in men
18 and the amelioration of Parkinsonian symptoms by estrogen therapy.
19 A number of rodent studies have also demonstrated CNS effects after single exposures to O3.
20 Lipid peroxidation, as evidenced by increases in TEARS, occurred in multiple regions of the brain
21 after a 1- to 9-h exposure to 1 ppm O3 (Escalante-Membrillo et al., 2005, 596378). Ozone has also
22 been shown to alter gene expression of endothelin-1 (pituitary) and inducible nitric oxide synthase
23 (cerebral hemisphere) after a single 4-h exposure to 0.8 ppm O3, indicating potential cerebrovascular
24 effects. This dose-dependent effect was not observed at 0.4 ppm O3 (Thomson et al., 2007, 196635).
25 Vascular endothelial growth factor was upregulated in astroglial cells in the central respiratory areas
26 of the brain of rats exposed to 0.5 ppm O3 for 3 hours (Araneda et al., 2008, 596367). The
27 persistence of CNS changes after a single exposure was also examined and the increase in vascular
28 endothelial growth factor was present after a short (3 hours) recovery period. Evidence for more
29 persistent oxidative stress-related changes in the CNS have been studied and morphological changes
30 in the olfactory bulb of rats exposed to 1 ppm O3 for 4 hours were observed at 2 hours, and 1 and
31 10 days, but not 15 days, after exposure (Colin-Barenque et al., 2005, 180458). Thus, there is
32 evidence that O3-induced CNS effects are both concentration- and time-dependent.
33 Because O3 can produce a disruption of the sleep-wake cycle (U.S. EPA, 2006, 088089).
34 Alfaro-Rodriguez et al. (2005, 596365) examined whether acetylcholine in a region of the brain
35 involved in sleep regulation was altered by O3. After a 24-h exposure to 0.5 ppm O3, the
36 acetylcholine concentration in the medial preoptic area was decreased by 58% and strongly
37 correlated with a disruption in paradoxical sleep. Such behavioral-biochemical effects of O3 are
38 confirmed by a number of studies which have demonstrated morphological and biochemical changes
39 in rats.
March 2011 6-136 DRAFT - DO NOT CITE OR QUOTE
-------
1 Adverse CNS effects have also been demonstrated in newborn and adult rats whose only
2 exposure to O3 occurred in utero. Several neurotransmitters were assessed in male offspring of dams
3 exposed to 1-ppm O3 during the entire pregnancy (Gonzalez-Pina et al, 2008, 475317). The data
4 showed that catecholamine neurotransmitters were affected to a greater degree than indole-amine
5 neurotransmitters in the cerebellum. Adverse CNS changes, including behavioral, cellular, and
6 biochemical effects, have also been observed after in utero exposure to 0.5 ppm O3 for 12 h/day from
7 gestational days 5-20 (Boussouar et al., 2009, 596368). Tyrosine hydroxylase labeling in the nucleus
8 tractus solatarius was increased after in utero exposure to O3 whereas Fos protein labeling did not
9 change. When these offspring were challenged by immobilization stress, neuroplasticity pathways,
10 which were activated in air-exposed offspring, were inhibited in O3-exposed offspring. Although an
11 O3 exposure concentration-response was not studied in these two in utero studies, it has been
12 examined in one study. Santucci et al. (2006, 596414) investigated behavioral effects and gene
13 expression after in utero exposure of mice to as little as 0.3 ppm O3. Increased defensive/submissive
14 behavior and reduced social investigation were observed in both the 0.3 and 0.6 ppm O3 groups.
15 Changes in gene expression of brain-derived neurotrophic factor (BDNF, increased in striatum) and
16 nerve growth factor (NGF, decreased in hippocampus) accompanied these behavioral changes. Thus,
17 these three studies demonstrate that CNS effects can occur as a result of in utero exposure to O3, and
18 although the mode of action of these effects is not known, it has been suggested that circulating lipid
19 peroxidation products may play a role (Boussouar et al., 2009, 596368). Importantly, these adverse
20 CNS effects occurred in rodent models after in utero only exposure to relevant concentrations of O3.
6.4.1. Neuroendocrine Effects
21 According to the 2006 O3 AQCD (U.S. EPA, 2006, 088089). early studies suggested an
22 interaction of O3 with the pituitary-thyroid-adrenal axis, because thyroidectomy, hypophysectomy,
23 and adrenalectomy protected against the lethal effects of O3. Concentrations of 0.7-1.0 ppm O3 for a
24 1-day exposure in male rats caused changes in the parathyroid, thymic atrophy, decreased serum
25 levels of thyroid hormones and protein binding, and increased prolactin. Increased toxicity to O3 was
26 reported in hyperthyroid rats and T3 supplementation was shown to increase metabolic rate and
27 pulmonary injury in the lungs of O3-treated animals. The mechanisms by which O3 affects
28 neuroendocrine function are not well understood, but previous work suggests that high ambient
29 levels of O3 can produce marked neural disturbances in structures involved in the integration of
30 chemosensory inputs, arousal, and motor control, effects that may be responsible for some of the
31 behavioral effects seen with O3 exposure. However, no recent studies have become available to add
32 to the limited evidence regarding neuroendocrine effects presented in the 2006 O3 AQCD.
6.4.2. Summary and Causal Determination
33 In rodents, O3 exposure has been shown to cause physicochemical changes in the brain
34 indicative of oxidative stress and inflammation. Newer toxicological studies add to earlier evidence
March 2011 6-137 DRAFT - DO NOT CITE OR QUOTE
-------
1 that acute exposures to O3 can produce a range of effects on the central nervous system and behavior.
2 Previously observed effects, including neurodegeneration, alterations in neurotransmitters, short and
3 long term memory, and sleep patterns, have been further supported by recent studies. In instances
4 where pathology and behavior are both examined, animals exhibit decrements in behaviors tied to
5 the brain regions or chemicals found to be affected or damaged. For example, damage in the
6 hippocampus, which is important for memory acquisition, was correlated with impaired performance
7 in tests designed to assess memory. Thus the brain is functionally affected by O3 exposure. The
8 single epidemiology study conducted showed that O3 affects memory in humans as well, albeit on a
9 long-term exposure basis. Notably, exposure to O3 levels as low as 0.25 ppm has resulted in
10 progressive neurodegeneration and deficits in both short and long-term memory in rodents.
11 Additionally, changes in the CNS, including biochemical, cellular, and behavioral effects, have been
12 observed in animals whose sole exposure occurred in utero, at levels as a low as 0.3 ppm. Although
13 evidence from epidemiologic and controlled human exposure studies is lacking, the toxicological
14 evidence for ozone's impact on the brain and behavior is strong, and at least is suggestive of a
15 causal relationship between O3 exposure and adverse CNS effects.
16
6.5. Effects on Other Organ Systems
6.5.1. Effects on the Liver and Xenobiotic Metabolism
17 Early investigations of the effects of O3 on the liver centered on xenobiotic metabolism, and
18 the prolongation of sleeping time, which was observed at 0.1 ppm O3 (Graham et al., 1981, 039415).
19 In some species, only adults and especially females were affected. In rats, high (1.0-2.0 ppm for
20 3 hours) acute O3 exposures caused increased production of NO by hepatocytes and enhanced
21 protein synthesis (Laskin et al., 1994, 076154: Laskin et al., 1996, 015771). The O3-associated
22 effects shown in the liver are thought to be mediated by inflammatory cytokines or other cytotoxic
23 mediators released by activated macrophages in the lungs (Laskin and Laskin, 2001, 016158; Laskin
24 et al., 1998, 015425: Vincent et al., 1996, 080777). Except for the earlier work on xenobiotic
25 metabolism, the responses occurred only after very high acute O3 exposures. One study, conducted at
26 1 ppm O3 exposure, has been identified (Last et al., 2005, 596400) in which alterations in gene
27 expression underlying O3-induced cachexia and downregulation of xenobiotic metabolism were
28 examined. A number of the down-regulated genes are known to be interferon (IFN) dependent,
29 suggesting a role for circulating IFN. A more recent study by Aibo et al. (2010, 378559)
30 demonstrates exacerbation of acetaminophen-induced liver injury in mice after a single 6-h exposure
31 to 0.25 or 0.5 ppm O3. Data indicate that O3 may worsen drug-induced liver injury by inhibiting
32 hepatic repair.
33 In summary, mediators generated by O3 exposure may cause effects on the liver in laboratory
34 rodents. Ozone exposures as low as 0.1 ppm have been shown to affect drug induced sleeping time,
March 2011 6-138 DRAFT - DO NOT CITE OR QUOTE
-------
1 and exposure to 0.25 ppm can exacerbate liver injury induced by a common analgesic. However,
2 very few studies at relevant concentrations have been conducted, and no data from controlled human
3 exposure or epidemiologic studies are currently available. Therefore the collective evidence is
4 inadequate to determine if a causal relationship exists between short-term O3 exposure and
5 effects on the liver and metabolism.
6.5.2. Effects on Cutaneous and Ocular Tissues
6 In addition to the lungs, the skin is highly exposed to O3 and contains O3 reactive targets
7 (polyunsaturated fatty acids) that can produce lipid peroxides. The 2006 O3 AQCD reported that
8 although there is evidence of oxidative stress at near ambient O3 concentrations, skin and eyes are
9 only affected at high concentrations (greater than 1-5 ppm). Ozone exposure (0.8 ppm for 7 days)
10 induces oxidative stress in the skin of hairless mice, along with proinflammatory cytokines (Valacchi
11 et al, 2009, 201554). A recent study demonstrated that 0.25 ppm O3 differentially alters expression
12 of metalloproteinases in the skin of young and aged mice, indicating age-related susceptibility to
13 oxidative stress (Fortino et al., 2007, 596382). In young mice, healing of skin wounds is not
14 significantly affected by O3 exposure (Lim et al., 2006, 670834). However, exposure to 0.5 ppm O3
15 for 6 h/day significantly delays wound closure in aged mice. As with effects on the liver described
16 above, the effects of O3 on the skin and eyes have not been widely studied, and information from
17 controlled human exposure or epidemiologic studies is not currently available. Therefore the
18 collective evidence is inadequate to determine if a causal relationship exists between short-term
19 O3 exposure and effects on cutaneous and ocular tissues.
6.6. Mortality
6.6.1. Summary of Findings from 2006 Ozone AQCD
20 The 2006 O3 AQCD (U.S. EPA, 2006, 088089) reviewed a large number of time-series studies
21 consisting of single- and multicity studies, and meta-analyses. In the large U.S. multicity studies that
22 examined all-year data, summary effect estimates corresponding to single-day lags ranged from a
23 0.5-1% increase in all-cause (nonaccidental) mortality per the standardized unit increase1 in O3. The
24 association between short-term O3 exposure and mortality was substantiated by a collection of meta-
25 analyses and international multicity studies. The studies evaluated found some evidence for
26 heterogeneity in O3 mortality risk estimates across cities and studies. Although more limited in
27 number, studies that conducted seasonal analyses reported larger O3 mortality risk estimates during
28 the warm or summer season. Overall, the 2006 O3 AQCD (U.S. EPA, 2006, 088089) identified
29 robust associations between various measures of daily ambient O3 concentrations and all-cause
1 In the 2006 O3 AQCD and throughout this document to compare across studies that used the same exposure metric, effect estimates were
standardized to 40 ppb for 1-h maximum, 30 ppb for 8-h maximum, and 20 ppb for 24-h average O3 concentrations.
March 2011 6-139 DRAFT - DO NOT CITE OR QUOTE
-------
1 mortality, with additional evidence for associations with cardiovascular mortality, which could not be
2 readily explained by confounding due to time, weather, or co-pollutants. However, it was noted that
3 multiple uncertainties remain regarding the O3-mortality relationship including: the extent of residual
4 confounding by co-pollutants; factors that modify the O3-mortality association; the appropriate lag
5 structure for identifying O3-mortality effects (e.g., single-day lags versus distributed lag model); the
6 shape of the O3-mortality C-R function and whether a threshold exists; and the identification of
7 susceptible populations. Collectively, the 2006 O3 AQCD (U.S. EPA, 2006, 088089) concluded that
8 "the overall body of evidence is highly suggestive that O3 directly or indirectly contributes to non-
9 accidental and cardiopulmonary-related mortality."
6.6.2. Associations of Mortality and Short-Term Ozone Exposure
10
11 The recent literature that examined the association between short-term O3 exposure and
12 mortality further confirmed the associations reported in the 2006 O3 AQCD (U.S. EPA, 2006,
13 088089). New multicontinent and multicity studies reported consistent positive associations between
14 short-term O3 exposure and all-cause mortality in all-year analyses, with additional evidence for
15 larger mortality risk estimates during the warm or summer months (Figure 6-27; Table 6-33). These
16 associations were reported across a range of ambient O3 concentrations that were in some cases quite
17 low (Table 6-34).
March 2011 6-140 DRAFT - DO NOT CITE OR QUOTE
-------
Study
Gryparisetal. (2004;57276)
Belletal. (2007;93256)
Schwartz (2005;57333)
Bell and Dominici (2008; 193828)
Belletal. (2004;94417)a
Levy etal. (2005; 74347)a
Katsouyanni etal. (2009; 1
Belletal. (2005;74345)a
Ito etal. (2005; 743 46)a
Katsouyanni etal. (2009; 199899)
Katsouyanni etal. (2009; 199899)
Katsouyanni et al. (2009; 199899)b
Samoli etal. (2009; 195855)
Belletal. (2004;94417)a
Schwartz (2005;57333)
Zanobetti and Schwartz (2008; 195755)
Zanobetti and Schwartz (2008; 101596)
Franklin and Schwartz (2008; 156448)
Gryparisetal. (2004;57276)
Medina-Ramon and Schwartz (2008)
Katsouyanni etal. (2009; 199899)
Belletal. (2005;74345)a
Katsouyanni etal. (2009; 199899)
Katsouyanni et al. (2009; 199899)b
Levy etal.(2005; 7434 7)a
Ito etal. (2005;74346)a
Katsouyanni etal. (2009; 199899)
Stafoggia et al. (2010; 625034)
Location
Lag
APHEA2 (23 cities)
98 U.S. communities
14 U.S. cities
98 U.S. communities
95 U.S. communities
U.S. and Non-U.S.
APHENA-Europe
U.S. and Non-llS.
U.S. and Non-U.S.
APHENA-U.S.
APHENA-Canada
APHENA-Canada
21 European cities
95 U.S. communities
14 U.S. cities
48 U.S. cities
48 U.S. cities
18 U.S. communities
APHEA2 21 cities)
48 U.S. cities
APHENA-Europe
U.S. and Non-U.S.
APHENA-Canada
APHENA-Canada
U.S. and Non-U.S.
U.S. and Non-U.S.
APHENA-U.S.
10 Italian cities
0-1
0-1
0
0-6
0-6
DL(0-2)
DL 0-2
DL 0-2
DL 0-2
0-1
0-6
0
0
0-3
0
0-1
0-2
DL(0-2)
DLIO-2
DLJO-2
DLIO-2
DLJO-5
All-Year
Summer
357
% Increase
11
Figure 6-27. Summary of mortality risk estimates for short-term ozone exposure and all-cause
(nonaccidental) mortality from all-year and summer season analyses. [Effect
estimates are for a 40-ppb increase in 1-h max, 30-ppb increase in 8-h max, and
20-ppb increase in 24-h avg ozone concentrations. An "a" represent multicity
studies and meta-analyses from the 2006 ozone AQCD. Bell et al. (2005, 074345). Ito
et al. (2005, 074346). and Levy et al. (2005, 074347) used a range of lag days in the
meta-analysis: Lag 0,1, 2, or average 0-1 or 1-2; single-day lags from 0 to 3; and lag
0 and 1-2; respectively. A "b" represents risk estimates from APHENA-Canada
standardized to an approximate IQR of 5.1 ppb for a 1-h max increase in ozone
concentrations (see explanation in Section 6.2.7.2).
March 2011
6-141
DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-33. Corresponding effect estimates for Figure 6-27
Study
Location
Lag
Avg Time
% Increase (95% Cl)
All-year
Gryparisetal. (2004, 057276)
Bell et al. (2007, 0932561
Schwartz (2005, 0573331
Bell and Dominici (2008, 1938281
Bell et al. (2004, 09441 71a
Levy et al. (2005, Q74347)a
Katsouyanni et al. (2009, 1998991
Bell et al. (2005, 0743451a
Ito et al. (2005, 0743461a
Katsouyanni et al. (2009, 1998991
Katsouyanni et al. (2009, 1998991
Katsouyanni et al. (2009, 199899lb
APHEA2 (23 cities)
98 U.S. communities
14 U.S. cities
98 U.S. communities
95 U.S. communities
U.S. andNon-U.S.
APHENA-Europe
U.S. andNon-U.S.
U.S. andNon-U.S.
APHENA-U.S.
APHENA-Canada
APHENA-Canada
0-1
0-1
0
0-6
0-6
...
DL(0-2)
...
...
DL(0-2)
DL(0-2)
DL(0-2)
1-h max
24-h avg
1-h max
24-h avg
24-h avg
24-h avg
1-h max
24-h avg
24-h avg
1-h max
1-h max
1-h max
0.24 (-0.86, 1.98)
0.64 (0.34, 0.92)
0.76(0.13, 1.40)
1.04(0.56, 1.55)
1.04(0.54, 1.55)
1.64(1.25,2.03)
1.66(0.47,2.94)
1.75(1.10,2.37)
2.20 (0.80, 3.60)
3.02(1.10,4.89)
5.87(1.82,9.81)
0.73(0.23,1.20)
Summer
Samoli et al. (2009, 1958551
Bell et al. (2004, 09441 71a
Schwartz (2005, 0573331
Zanobetti and Schwartz (2008, 1957551
Zanobetti and Schwartz (2008, 1015961
Franklin and Schwartz (2008, 1564481
Gryparisetal. (2004, 057276)
Medina-Ramon and Schwartz (2008, 1938291
Katsouyanni et al. (2009, 1998991
Bell et al. (2005, 0743451a
Katsouyanni et al. (2009, 1998991
Katsouyanni et al. (2009, 1998991
Levy et al. (2005, 0743471a
Ito et al. (2005, 0743461a
Katsouyanni et al. (2009, 1998991
Stafoggia et al. (2010, 625034)
21 European cities
95 U.S. communities
14 U.S. cities
48 U.S. cities
48 U.S. cities
18 U.S. communities
APHEA2 (21 cities)
48 U.S. cities
APHENA-Europe
U.S. andNon-U.S.
APHENA-Canada
APHENA-Canada
U.S. andNon-U.S.
U.S. andNon-U.S.
APHENA-U.S.
10 Italian cities
0-1
0-6
0
0
0-3
0
0-1
0-2
DL(0-2)
...
DL(0-2)
DL(0-2)
...
...
DL(0-2)
DL(O-S)
8-h max
24-h avg
1-h max
8-h max
8-h max
24-h avg
8-h max
8-h max
1-h max
24-h avg
1-h max
1-h max
24-h avg
24-h avg
1-h max
8-h max
0.66(0.24, 1.05)
0.78(0.26, 1.30)
1.00(0.30,1.80)
1.51 (1.14, 1.87)
1.60(0.84,2.33)
1.79(0.90,2.68)
1.80(0.99,3.06)
1.96(1.14,2.82)
2.38(0.87,3.91)
3.02(1.45,4.63)
3.34(1.26,5.38)
0.42(0.16,0.67)
3.38 (2.27, 4.42)
3.50(2.10,4.90)
3.83(1.90,5.79)
9.15(5.41,13.0)
aMulticity studies and meta-analyses from the 2006 03AQCD. Belletal. (2005, 074345)a, Itoetal. (2005, 074346)a, and Levyetal. (2005, 0743471a used a range
of lag days in the meta-analysis: Lag 0,1, 2, or average 0-1 or 1 -2; Single-day lags from 0-3; and Lag 0 and 1 -2; respectively.
bRisk estimates from APHENA-Canada standardized to an approximate IQR of 5.1 ppb for a 1 -h max increase in 03 concentrations (see explanation in Section
6.2.7.2).
March 2011
6-142
DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-34. Range of mean and upper percentile ozone concentrations in previous and recent multicity
studies
Study
Gryparis et al. (2004,
057276)"
Schwartz (2005, 057333)"
Bell et al. (2004, 094417)
Bell et al. (2007, 093256)
Bell and Dominici (2008,
193828)
Franklin and Schwartz
(Franklin and Schwartz,
2008, 1 56448)
Katsouyanni et al. (2009,
Medina-Ramon and
Schwartz (2008, 193829)"
Samoli et al. (2009,
195855)"
Stafoggia et al. (2010,
625034)
Zanobetti and Schwartz
(2008, 1 01 596)
Zanobetti and Schwartz
(2008, 1 95755)
Location
23 European
cities (APHEA2)
14 U.S. cities
95 U.S.
communities
(NMMAPS)
98 U.S.
communities
(NMMAPS)
98 U.S.
communities
(NMMAPS)
18 U.S.
communities
NMMAPS
12 Canadian
cities
(APHEA2)
48 U.S. cities
21 European
cities (APHEA2)
10 Italian cities
48 U.S. cities
48 U.S. cities0
Years
1990-1997
1986-1993
1987-2000
1987-2000
1987-2000
(All year and
May-September)
2000-2005
(May-
September)
1987-1996
(Canada and
U.S.) varied by
city for Europe
1989-2000
(May-
September)
1990-1997
(June-August)
2001-2005
(April-
September)
1989-2000
(June-August)
1989-2000
(Winter:
December-
February)
(Spring: March-
May)
(Summer: June-
August)
(Autumn:
September-
November)
Metric
1-h max
8-h max
1-h max
24-h avg
24- h avg
24-h avg
24-h avg
1-h max
8-h max
8-h max
8-h max
8-h max
8-h max
Mean
Concentration (ppb)a
Summer:
1-h max: 44-117
8-h max: 30-99
Winter:
1-h max: 11-57
8-h max: 8-49
35.1-60
26.0
26.0 d
All year: 26.8
May-September: 30.0
21 .4-48.7
U.S.: 13.3-38.4
Canada: 6.7-8.4
Europe:18.3-41.9
16.1-58.8
20.0-62.8
41.2-58.9
15.1-62.8
Winter: 16.5
Spring: 41.6
Summer: 47.8
Autumn: 33.5
Middle/Upper Percentile
Concentrations (ppb)a
Summer:
1-h max: 62-173
8-h max: 57-154
Winter:
1-h max: 40-88
8-h max: 25-78
25th: 26.5-52
75th: 46.3-69
NR
NR
Maximum:
All year: 37.3
May-September: 47.2
NR
75th:
U.S.: 21.0-52.0
Canada: 8.7-12.5
Europe: 24.0-65.8
NR
75th: 27.2-74.8
75th: 47.0-71 .6
Max: 34.3-1 46.2
75th: 19.8-75.9
Max:
Winter: 40.6
Spring: 91.4
Summer: 103.0
Autumn: 91.2
a03 concentrations were converted to ppb if the study presented them as pg/m3 by using the conversion factor of 0.51 assuming standard temperature (25° C) and
pressure (1 atm).
Study only reported median 03 concentrations.
"Cities with less than 75% observations in a season excluded. As a result, 29 cities examined in winter, 32 in spring, 33 in autumn, and all 48 in the summer.
"Bell et al. (2007, Q93256)did not report mean 03 concentrations, however, it used a similar dataset as Bell et al. (2004, 0944171 which consisted of 95 U.S.
communities for 1987-2000. For comparison purposes the 24-h avg 03 concentrations for the 95 communities from Belletal. (2004, 0944171 are reported here.
'Study did not present air quality data for the summer months.
1 In addition to examining the relationship between short-term O3 exposure and all-cause
2 mortality, recent studies attempted to address the uncertainties that remained upon the completion of
3 the 2006 O3 AQCD. As a result, given the robust associations between short-term O3 exposure and
4 mortality presented across studies in the 2006 O3 AQCD and supported in the new multicity studies,
5 the following sections primarily focus on the examination of previously identified uncertainties in
6 the O3-mortality relationship, specifically: confounding, lag structure (e.g., multiday effects and
March 2011
6-143
DRAFT - DO NOT CITE OR QUOTE
-------
1 mortality displacement), effect modification (i.e., sources of heterogeneity in risk estimates across
2 cities); the O3-mortality C-R relationship; and O3 associations with cause-specific mortality.
3 Focusing specifically on these uncertainties allows for a more detailed characterization of the
4 relationship between short-term O3 exposure and mortality.
6.6.2.1. Confounding
Confounding by PM and PM Constituents
5 An important question in the evaluation of the association between short-term O3 exposure and
6 mortality is whether the relationship is confounded by particulate matter, particularly the PM
7 chemical components that are found in the "summer haze" mixture which also contains O3.
8 However, because of the temporal correlation among these PM components and O3, and their
9 possible interactions, the interpretation of results from multi-pollutant models that attempt to
10 disentangle the health effects associated with each pollutant is limited.
11 The potential confounding effects of PMi0 and PM25 on the O3-mortality relationship were
12 examined by Bell et al. (2007, 093256) using data on 98 U.S. urban communities for the years
13 1987-2000 from the National Morbidity, Mortality, and Air Pollution Study (NMMAPS). In this
14 analysis the authors included PM as a covariate in time-series models, and also examined
15 O3-mortality associations on days when O3 concentrations were below a specified value. This
16 analysis was limited by the small fraction of days when both PM and O3 data were available, due to
17 the every-3rd- or 6th-day sampling schedule for the PM indices, and the limited amount of city-
18 specific data for PM25 because it was only collected in most cities since 1999. As a result, of the 91
19 communities with PM25 data, only 9.2% of days in the study period had data for both O3 and PM25,
20 resulting in the use of only 62 communities in the PM2 5 analysis. An examination of the correlation
21 between PM and O3 found that neither PM size fraction was highly correlated with various levels of
22 daily concentrations of O3 or PM (e.g., PMi0 and PM25). These results were also observed when
23 using 8-h max and 1-h max O3 exposure metrics. National and community-specific effect estimates
24 of the association between short-term O3 exposure and mortality were robust to inclusion of PMi0 or
25 PM25 in time-series models through the range of O3 concentrations (i.e., <10 ppb, 10-20, 20-40,
26 40-60, 60-80, and >80 ppb). For example, the percent increase in nonaccidental deaths per 10 ppb
27 increase 24-h avg O3 concentrations at lag 0-1 day were 0.22% (95% CI: -0.22, 0.65) without PM25
28 and 0.21% (95% CI: -0.22, 0.64) with PM25 in 62 communities.
29 Although no strong correlations between PM and O3 were reported by Bell et al. (2007,
30 093256) the patterns observed suggest regional differences in their correlation. (Table 6-35). Both
31 PMio and PM2 5 show positive correlations with O3 in the Industrial Midwest, Northeast, Urban
32 Midwest, and Southeast, especially in the summer months, presumably, because of the summer
33 peaking sulfate. However, the mostly negative or weak correlations between PM and O3 in the
34 summer in the Southwest, Northwest, and southern California could be due to winter-peaking nitrate.
March 2011 6-144 DRAFT - DO NOT CITE OR QUOTE
-------
1 Thus, the potential confounding effect of PM on the O3-mortality relationship could be influenced by
2 the relative contribution of sulfate and nitrate, which varies regionally and seasonally.
Table 6-35. Correlations
between PM
No. of
Communities
and ozone by season and
Winter
Spring
region
Summer
Fall
Yearly
PM10
Industrial Midwest
Northeast
Urban Midwest
Southwest
Northwest
southern California
Southeast
U.S.
19
15
6
9
11
7
25
93
0.37
0.34
0.24
0.00
-0.17
0.19
0.33
0.23
0.44
0.44
0.25
0.02
-0.20
0.08
0.35
0.26
0.44
0.36
0.22
-0.02
-0.13
0.12
0.31
0.24
0.39
0.44
0.26
0.10
-0.11
0.19
0.31
0.26
0.41
0.40
0.24
0.03
-0.16
0.14
0.32
0.25
PM2.5
Industrial Midwest
Northeast
Urban Midwest
Southwest
Northwest
southern California
Southeast
U.S.
19
13
4
9
11
7
26
90
0.18
0.05
0.22
-0.15
-0.32
-0.25
0.38
0.09
0.39
0.26
0.31
-0.08
-0.34
-0.22
0.47
0.21
0.43
0.16
0.15
-0.17
-0.39
-0.25
0.30
0.12
0.44
0.43
0.32
-0.15
-0.24
-0.15
0.37
0.22
0.36
0.25
0.20
-0.14
-0.31
-0.23
0.39
0.16
Source: Bell et al. (2007, 0932561
3 In an attempt to reassess a number of issues associated with the O3-mortality relationship,
4 including confounding, Smith et al. (2009, 199750) re-analyzed the publicly available NMMAPS
5 database forthe years 1987-2000. The authors conducted a number of analyses using constrained
6 distributed lag models and the average of 0- and 1-day lags. In addition, Smith et al. (2009, 199750)
7 examined the effect of different averaging times (24-h, 8-h, and 1-h max) on O3-mortality regression
8 coefficients, and whether PMi0 confounded the O3-mortality relationship. The authors reported that,
9 in most cases, O3 mortality risk estimates were reduced by between 22% and 33% in co-pollutant
10 models with PMi0. This is further highlighted in Figure 6-28, which shows scatter plots of
11 O3-mortality risk estimates with adjustment for PMi0 versus without adjustment for PMi0. Smith et
12 al. (2009, 199750) point out that a larger fraction (89 out of 93) of the posterior estimates lie below
13 the diagonal line (i.e., estimates are smaller with PMi0 adjustment) compared to the raw estimates
14 (56 out of 93). This observation could be attributed to both sets of posterior estimates being
15 calculated by "shrinking towards the mean." However, the most prominent feature of these plots is
16 that the variation of O3-mortality risk estimates across cites is much larger than the impact of PMi0
17 adjustment on the O3-mortality relationship.
March 2011
6-145
DRAFT - DO NOT CITE OR QUOTE
-------
Raw Estimates
15-
10-
0-
-5-
-5
0 5
Without PM10
Posterior Estimates
10
1.5-i
1.0-
o
i 0.5-
CL
| 0.0-
-0.5-
-1.0
-0.5
0.0 0.5
Without PM 10
1.0
1.5
Source: Used with permission from Informa UK Ltd., Smith et al. (2009, 1997501.
Figure 6-28. Scatter plots of ozone mortality risk estimates with versus without adjustment for
in NMMAPS cities. [The diagonal line indicates 1:1 ratio.
1 Franklin and Schwartz (2008, 156448) examined the sensitivity of O3 mortality risk estimates
2 to the inclusion of PM25 or PM chemical components associated with secondary aerosols (e.g.,
3 sulfate [SO42~], organic carbon [OC], and nitrate [NO3-]) in co-pollutant models. This analysis
4 consisted of between 3 and 6 years of data from May through September 2000-2005 from 18 U.S.
5 communities. The association between O3 and non-accidental mortality was examined in single-
6 pollutant models and after adjustment for PM2 5, sulfate, organic carbon, or nitrate concentrations.
7 The single-city effect estimates were combined into an overall estimate using a random-effects
8 model. In the single-pollutant model, the authors found a 0.89% (95% CI: 0.45, 1.33%) increase in
9 nonaccidental mortality with a 10-ppb increase in same-day 24-h summertime O3 concentrations
10 across the 18 U.S. communities. Adjustment for PM2 5 mass, which was available for 84% of the
1 1 days, decreased the O3-mortality risk estimate only slightly (from 0.88% to 0.79%), but the inclusion
12 of sulfate in the model reduced the risk estimate by 3 1% (from 0.85% to 0.58%). However, sulfate
13 data was only available for 18% of the days. Therefore, a limitation of this study is the limited
14 amount of data for PM2 5 chemical components due to the every-3rd-day or every-6th-day sampling
15 schedule. For example, when using a subset of days when organic carbon measurements were
16 available (i.e., 17% of the available days), O3 mortality risk estimates were reduced to 0.51% (95%
17 CI: -0.36 to 1.36) in a single-pollutant model.
March 2011
6-146
DRAFT - DO NOT CITE OR QUOTE
-------
1 Consistent with the studies previously discussed, the results from Franklin and Schwartz
2 (2008, 156448) also demonstrate that the interpretation of the potential confounding effects of
3 co-pollutants on O3 mortality risk estimates is not straightforward. As presented in Figure 6-29, the
4 regional and city-to-city variations in O3 mortality risk estimates appear greater than the impact of
5 adjusting for co-pollutants. In addition, in some cases, a negative O3 mortality risk estimate becomes
6 even more negative with the inclusion of sulfate (e.g., Seattle) in a co-pollutant model, or a null O3
7 mortality risk estimate becomes negative when sulfate is included (e.g., Dallas and Detroit). Thus,
8 the reduction in the overall O3 mortality risk estimate (i.e., across cities) needs to be assessed in the
9 context of the heterogeneity in the single-city estimates.
Seattle
Fresno
raso
Dallas
Houston
Beaumont
Kansas City
St. Louis
Detroit
Pittsburgh
Buffalo
Rochester
Philadelphia
Boston
1 D =H
1 ' W
^Q^
^c.
rt
1 ^^
h1 '•&-
\ .
n Ozone with sulfate I
x Ozone alone |
D ' 1
n i
J*— '
p^
=!)
< 1
^
dx . I
-5 0 5
Percent increase in mortality
with 10 ppb increase in ozone
Source: Used with permission of Franklin and Schwartz (2008, 1564481.
Figure 6-29. Community-specific ozone-mortality risk estimates for nonaccidental mortality per
10-ppb increase in same-day 24-h avg summertime ozone concentrations in single-
pollutant models and co-pollutant models with sulfate.
10 In the APHENA study, the investigators from the U.S. (NMMAPS), Canadian, and European
11 (APHEA2) multicity studies collaborated and conducted a joint analysis of PMi0 and O3 using each
12 of these datasets (Katsouyanni et al, 2009, 199899). For mortality, each dataset consisted of a
13 different number of cities and years of air quality data: U.S. encompassed 90 cities with daily O3
14 data from 1987-1996 of which 36 cities had summer only O3 measurements; Europe included 23
15 cities with 3-7 years of daily O3 data during 1990-1997; and Canada consisted of 12 cities with daily
March 2011
6-147
DRAFT - DO NOT CITE OR QUOTE
-------
1 O3 data from 1987 to 1996. As discussed in Section 6.2.7.2, the APHENA study conducted extensive
2 sensitivity analyses, of which the 8 df/year results for both the penalized spline (PS) and natural
3 spline (NS) models are presented in the text for comparison purposes, but only the NS results are
4 presented in figures because alternative spline models have previously been shown to result in
5 similar effect estimates (Health Effects Institute, 2003, 042829). Additionally, for the Canadian
6 results, figures contain risk estimates standardized to both a 40-ppb increment for 1-h max O3
7 concentrations, consistent with the rest of the ISA, but also the approximate IQR across the Canadian
8 cities as discussed previously (Section 6.2.7.2).
9 In the three datasets, the authors found generally positive associations between short-term O3
10 exposure and all-cause, cardiovascular, and respiratory mortality. The estimated excess risks for O3
11 were larger for the Canadian cities than for the U.S. and European cities. When examining the
12 potential confounding effects of PMi0 on O3 mortality risk estimates, the sensitivity of the estimates
13 varied across the data sets and age groups. In the Canadian dataset, adjusting for PMi0 modestly
14 reduced O3 risk estimates for all-cause mortality for all ages in the PS (4.5% [95% CI: 2.2, 6.7%])
15 and NS (4.2% [95% CI: 1.9, 6.5%]) models to 3.8% (95% CI: -1.4, 9.8%) and 3.2% (95% CI: -2.2,
16 9.0%), respectively, at lag 1 for a 40-ppb increase in 1-h max O3 concentrations (Figure 6-30; Table
17 6-36). However, adjusting for PMi0 reduced O3 mortality risk estimates in the^ 75-year age group,
18 but increased the risk estimates in the <75-year age group. For cardiovascular and respiratory
19 mortality more variable results were observed with O3 risk estimates being reduced and increased,
20 respectively, in co-pollutant models with PMi0 (Figure 6-30; Table 6-36). Unlike the European and
21 U.S. datasets, the Canadian dataset only conducted co-pollutant analyses at lag 1; as a result, to
22 provide a comparison across study locations only the lag 1 results are presented for the European and
23 U.S. datasets in this section.
24 In the European data, O3 risk estimates were robust when adjusting for PMi0 in the year-round
25 data for all-cause, cardiovascular and respiratory mortality. When restricting the analysis to the
26 summer months moderate reductions were observed in O3 risk estimates for all-cause mortality (e.g.,
27 lag 1 in a PS (0.29% [95% CI: 0.19, 0.39]) andNS (0.26% [95% CI: 0.14, 0.37]) model was reduced
28 to 0.19% (95% CI: 0.07, 0.32%) and 0.16% (95% CI: 0.02, 0.29%), respectively) with more
29 pronounced reductions in respiratory mortality. In the U.S. data, adjusting for PMi0 moderately
30 reduced O3 risk estimates for all-cause mortality in a year-round analysis at lag 1 (e.g., both the PS
31 andNS models were reduced from 0.18% to 0.13%) (Figure 6-30; Table 6-36). Similar to the
32 European data, when restricting the analysis to the summer months, adjusting for PMi0 moderately
33 reduced O3 mortality risk estimates in the U.S. For example, the O3 risk estimate for all-cause
34 mortality for all ages at lag 1 day in the PS (3.9% [95% CI: 2.3, 5.5%]) and NS (4.3% [95% CI: 2.2,
35 6.5%]) models was reduced to 2.1% (95% CI: -0.55, 4.9%) and 1.9% (95% CI: -0.78, 4.6%),
36 respectively. However, when examining cause-specific mortality risk estimates, consistent with the
37 results from the Canadian dataset, which employed a similar PM sampling strategy (i.e., every-6th-
38 day sampling), O3 risk estimates for cardiovascular and respiratory mortality were more variable;
39 reduced or increased in all-year and summer analyses. Overall, the estimated O3 risks appeared to be
March 2011 6-148 DRAFT - DO NOT CITE OR QUOTE
-------
1 moderately to substantially sensitive to inclusion of PMi0 in co-pollutant models. Despite the
2 multicity approach, the mostly every-6th-day sampling schedule for PMi0 in the Canadian and U.S.
3 datasets greatly reduced the sample size and limits the interpretation of these results.
Location
APHENA-U.S.
APHENA-Canada
a
a
a
a
a
APHENA-Europe
Ages
All
>75
<75
>75
<75
All
All
>75
<75
All
All
>75
<75
>75
<75
All
All-Cause
Cardiovascular
Respiratory
All-Cause
Cardiovascular
Respiratory
All-Cause
Cardiovascular
Respiratory
All-Year
Summer
All-Year
Summer
All-Year
Summer
All-Year
All-Year
Summer
All-Year
Summer
All-Year
Summer
-10
5 10 15
% Increase
20
25
30
Figure 6-30. Percent increase in all-cause (nonaccidental) and cause-specific mortality from the
APHENA study for single- and co-pollutant models. [Effect estimates are for a 40-
ppb increase in 1-h max ozone concentrations at lag 1. All estimates are for the
8 df/year model with natural splines. Circles represent all-year analysis results while
diamonds represent summer season analysis results. Open circles and diamonds
represent co-pollutant models with PM10. Black = all-cause mortality; red =
cardiovascular mortality; and blue = respiratory mortality. An "a" represents risk
estimates from APHENA-Canada standardized to an approximate IQR of 5.1 ppb for
a 1-h max increase in ozone concentrations (see explanation in Section 6.2.7.2).
March 2011
6-149
DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-36. Corresponding Effect Estimates for Figure 6-30
Location Mortality Ages
APHENA-U.S. All-Cause All
Cardiovascular a 75
<75
>75
<75
Respiratory All
APHENA-Canada All-Cause All
Cardiovascular a 75
<75
Respiratory All
APHENA-Europe All-Cause All
Cardiovascular > 75
<75
>75
<75
Respiratory All
Season
All-year
Summer
All-year
Summer
All-year
Summer
All-year
All-year
Summer
All-year
Summer
All-year
Summer
Co-pollutant
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
% Increase (95% Cl)
1.42(0.08,2.78)
1.02 (-1.40, 3.50)
4.31 (2.22, 6.45)
1.90 (-0.78, 4.64)
1.10 (-1.33, 3.67)
0.47 (-4.61, 5.79)
-0.1 6 (-3.02, 2.86)
1.34 (-3.63, 6.61)
3.58 (0.87, 6.37)
-1.1 7 (-6. 18, 4.07)
3.18(0.31,6.12)
1.26 (-4.46, 7.28)
2.46 (-1.87, 6.86)
3.50 (-4.23, 11.8)
6.04(1.18, 11.1)
7.03 (-3.48, 18.5)
4.15(1.90,6.45)
0.52 (0.24, 0.80)a
3. 18 (-2. 18, 8.96)
0.40 (-0.28, 1.1 0)a
5.62(0.95, 10.7)
0.70(0.12, 1.30)a
1.90 (-9.03, 14.1)
0.24 (-1.20,1.70)3
1.10 (-4.08, 6.61)
0.1 4 (-0.53, 0.82)a
-2.64 (-14.7, 11.5)
-0.34 (-2.00, 1.40)a
0.87 (-6.40, 8.96)
0.11 (-0.84, 1.1 0)a
22.3 (-12.6, 71.3)
2.60 (-1.70, 7.1 0)a
1.02(0.39, 1.66)
1.26(0.47,1.98)
2.06(1.10,2.94)
1.26(0.16,2.30)
1.10 (-0.47, 2.70)
1.1 8 (-0.55, 2.94)
1.34 (-0.24, 2.94)
1.74 (-0.31, 3.75)
2.54 (0.39, 4.80)
1.58 (-0.70, 3.99)
1.66 (-0.70, 4.15)
1.66 (-1.02, 4.40)
1.42 (-1.02, 3.83)
1.42 (-1.02, 3.83)
4.31 (1.66,7.11)
1.18 (-1.79, 4.31)
aRisk estimates from APHENA-Canada standardized to an approximate IQR of 5.1 ppb for a 1 -h max increase in 03 concentrations (see explanation in Section
6.2.7.2).
March 2011
6-150
DRAFT - DO NOT CITE OR QUOTE
-------
1 Stafoggia et al. (2010, 625034) examined the potential confounding effects of PMi0 on the
2 O3-mortality relationship in individuals 35 years of age and older in 10 Italian cities from 2001 to
3 2005. In a time-stratified case-crossover analysis, using data for the summer months (i.e., April-
4 September), the authors examined O3-mortality associations across each city, and then obtained a
5 pooled estimate through a random-effects meta-analysis. Stafoggia et al. (2010, 625034) found a
6 strong association with nonaccidental mortality (9.2% [95% CI: 5.4, 13.0%] for a 30-ppb increase in
7 8-h max O3 concentrations) in an unconstrained distributed lag model (lag 0-5) that persisted in
8 co-pollutant models with PM10 (9.2% [95% CI: 5.4, 13.7%]). Additionally, when examining cause-
9 specific mortality, the authors found positive associations between short-term O3 exposure and
10 cardiovascular (14.3% [95% CI: 6.7, 22.4%]), cerebrovascular (8.5% [95% CI: 0.1, 16.3%]), and
11 respiratory (17.6% [95% CI: 1.8, 35.6%]) mortality in single-pollutant models. In co-pollutant
12 models, O3-mortality effect estimates for cardiovascular and cerebrovascular mortality were robust
13 to the inclusion of PM10 (9.2% [95% CI: 5.4, 13.7%]) and 7.3% [95% CI: -1.2, 16.3%],
14 respectively), and attenuated, but remained positive, for respiratory mortality (9.2% [95% CI: -6.9,
15 28.8%]). Of note, the correlations between O3 and PMi0 across cities were found to be generally low,
16 ranging from (-0.03 to 0.49). Unlike the other studies mentioned above that used every-3rd-day
17 sampling for PMi0, the authors do not specify the sampling strategy used for PMi0 in this analysis
Confounding by Seasonal Trend
18 The APHENA study (Katsouyanni et al., 2009, 199899). mentioned above, also conducted
19 extensive sensitivity analyses to identify the appropriate: smoothing method and basis functions to
20 estimate smooth functions of time in city-specific models; and degrees of freedom to be used in
21 smooth functions of time, to adjust for seasonal trends. Because O3 peaks in the summer and
22 mortality peaks in the winter, not adjusting or not sufficiently adjusting for the seasonal trend would
23 result in an apparent negative association between the O3 and mortality time-series. Katsouyanni et
24 al. (2009, 199899) examined the effect of the extent of smoothing for seasonal trends by using
25 models with 3 df/year, 8 df/year (the choice for their main model), 12 df/year, and df/year selected
26 using the sum of absolute values of partial autocorrelation function of the model residuals (PACF)
27 (i.e., choosing the degrees of freedom that minimizes positive and negative autocorrelations in the
28 residuals). Table 6-37 presents the results of the degrees of freedom analysis using alternative
29 methods to calculate a combined estimate: the Berkey et al. (1998, 684190) meta-regression and the
30 two-level normal independent sampling estimation (TLNISE) hierarchical method. The results show
31 that the methods used to combine single-city estimates did not influence the overall results, and that
32 neither 3 df/year nor choosing the df/year by minimizing the sum of absolute values of PACF of
33 regression residuals was sufficient to adjust for the seasonal negative relationship between O3 and
34 mortality. However, it should be noted, the majority of studies in the literature that examined the
35 mortality effects of short-term O3 exposure, particularly the multicity studies, used 7 or 8 df/year to
36 adjust for seasonal trends, and in both methods a positive association was observed between O3
37 exposure and mortality when using 8 df/year to adjust for seasonal trends.
March 2011 6-151 DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-37. Sensitivity of ozone risk estimates per 10 ug/m3 increase in 24-h avg ozone concentrations
at lag 0-1 to alternative methods for adjustment of seasonal trend, for all-cause mortality
using Berkey MLE and TLNSE Hierarchical Models
Seasonality Control
3 df/year
8 df/year
12 df/year
PACF
Berkey
-0.54 (-0.88, 0.20)
0.30(0.11,0.50)
0.34(0.15,0.53)
-0.62 (-1.01, -0.22)
TLNISE
-0.55 (-0.88, -0.22)
0.31 (0.09, 0.52)
0.33(0.12,0.54)
-0.62 (-0.98, -0.27)
Source: Used with permission from Health Effects Institute, Katsouyanni et al. (2009, 199899V
6.6.2.2. Effect Modification
1 There have been several multicity studies that examined potential effect modifiers, or time-
2 invariant factors, that may modify O3 mortality risk estimates. These effect modifiers can be
3 categorized into either individual-level or community-level characteristics, which are traditionally,
4 examined in second stage regression models. In addition to potentially modifying the association
5 between short-term O3 exposure and mortality, both individual-level and community-level
6 characteristics may also contribute to the apparent geographic pattern of spatial heterogeneity in O3
7 mortality risk estimates. As a result, the geographic pattern of O3 mortality risk estimates is also
8 evaluated in this section.
Individual-Level Characteristics
9 Medina-Ramon and Schwartz (2008, 193829) conducted a case-only study in 48 U.S. cities to
10 identify populations particularly susceptible to O3-related mortality for the period 1989-2000 (May
11 through September of each year [i.e., warm season]). A case-only design predicts the occurrence of
12 time-invariant characteristics among cases as a function of the exposure level (Armstrong, 2003,
13 153211). For each potential effect modifier (time-invariant individual-level characteristics), city-
14 specific logistic regression models were fitted, and the estimates were pooled across all cities.
15 Furthermore, the authors examined potential differences in individual effect modifiers according to
16 several city characteristics (e.g., mean O3 level, mean temperature, households with central air
17 conditioning, and population density) in a meta-regression. Across cities the authors found a 1.96%
18 (95% CI: 1.14-2.82%) increase in mortality at lag 0-2 for a 30-ppb increase in 8-h max O3
19 concentrations. Additionally, Medina-Ramon and Schwartz (2008, 193829) examined a number of
20 individual-level characteristics (e.g., age, race) and chronic conditions (e.g., secondary causes of
21 death) as effect modifiers of the association between short-term O3 exposure and mortality
22 (Table 6-38). The authors found that older adults (i.e., > 65) (1.10% [95% CI: 0.44, 1.77%]), women
23 >60 years of age (0.58% [95% CI: 0.18, 0.98%]), black race (0.53% [95% CI: 0.19, 0.87%]), and
24 secondary atrial fibrillation (1.66% [95% CI: 0.03, 3.32%]) showed the greatest additional percent
March 2011 6-152 DRAFT - DO NOT CITE OR QUOTE
-------
1 change in O3-related mortality.: In addition, when examining city-level characteristics, the authors
2 found that older adults, black race, and secondary atrial fibrillation had a larger effect on O3
3 mortality risk estimates in cities with lower O3 levels. Of note, is a similar case-only study
4 (Schwartz, 2005, 667864) that examined potential effect modifiers of the association between
5 temperature and mortality, which would be expected to find results consistent with the Medina-
6 Ramon and Schwartz (2008, 193829) study due to the high correlation between temperature and O3.
7 However, when stratifying days by temperature Schwartz (2005, 667864) found strong evidence that
8 diabetes increased the temperature-mortality association on hot days, which was not as evident when
9 examining the O3-mortality association in Medina-Ramon and Schwartz (2008, 193829). This
10 difference could be due to the study design and populations included in both studies, a multicity
11 study including all ages (Medina-Ramon and Schwartz, 2008, 193829) compared to a single-city
12 study of individuals^ 65 years of age (Schwartz, 2005, 667864). However, when examining results
13 stratified by race, nonwhites were found to have higher mortality risks on both hot and cold days,
14 which provide some support for the additional risk found for black race in Medina-Ramon and
15 Schwartz (2008,193829).
16 Individual-level factors that may result in susceptibility to O3-related mortality were also
17 examined by Stafoggia et al. (2010, 625034). As discussed above, using a time-stratified case-
18 crossover analysis, the authors found an association between short-term O3 exposure and
19 nonaccidental mortality in an unconstrained distributed lag model in 10 Italian cities (9.2% [95% CI:
20 5.4, 13.0%; lag 0-5 for a 30-ppb increase in 8-h max O3 concentrations). Stafoggia et al. (2010,
21 625034) conducted additional analyses to examine whether age, sex, income level, location of death,
22 and underlying chronic conditions increased the risk of O3-related mortality, but data for only nine of
23 the cities was available for these analyses. Of the individual-level factors examined, the authors
24 found the strongest evidence for increased risk of O3-related mortality in individuals 85 years of
25 age (22.4% [95% CI: 15.0, 30.2%]), women (13.7% [95% CI: 8.5, 19.7%]), and out-of-hospital
26 deaths (13.0% [95% CI: 6.0, 20.4%]). When focusing specifically on out-of hospital deaths and the
27 subset of individuals with chronic conditions, Stafoggia et al. (2010, 625034) found the strongest
28 association for individuals with diabetes, which is consistent with the potentially increased
29 susceptibility of diabetics on hot days observed in Schwartz (2005, 667864).
30 Overall, uncertainties exist in the interpretation of the potential effect modifiers, identified in
31 Medina-Ramon and Schwartz (2008, 193829) and Stafoggia et al. (2010, 625034). of the O3.mortality
32 relationship due to the expected heterogeneity in O3 mortality risk estimates across cities as
33 highlighted in Smith et al. (2009, 199750) (Figure 6-28) and Franklin and Schwartz (2008, 156448)
34 (Figure 6-29). For example, it is difficult to determine the relative importance of a susceptibility
35 factor that results in an additional percent increase in mortality in a multicity analysis when analyses
36 of the individual cities within the study did not indicate associations between O3 and mortality. It
1 These estimates represent the additional percent change in mortality for persons who had the characteristic being examined compared to
persons who did not have the characteristic, when the mean O3 level of the previous 3 days increased 10 ppb. These values were not
standardized because they do not represent the actual effect estimate for the characteristic being examined, but instead the difference
between effect estimates for persons with versus without the condition.
March 2011 6-153 DRAFT - DO NOT CITE OR QUOTE
-------
1 also remains unclear if the individual-level susceptibility factors identified in Medina-Ramon and
2 Schwartz (2008, 193829) and Stafoggia et al. (2010, 625034) only modify the O3-mortality
3 relationship. More than likely, the factors identified span pollutants as is evident by older adults (i.e.,
4 > 65) often being identified as an effect modifier of PM mortality risk estimates (U.S. EPA, 2009,
5 179916V
Table 6-38. Additional percent change in ozone-related mortality for individual-level susceptibility
factors
Percent3
(95% Cl)
Socio-demographic characteristics
Age 65 yr or older
Women
Women <60 yr old"
Women > 60 yr old"
Black race
Low education
1.10
0.58
-0.09
0.60
0.53
-0.29
(0.44 to 1.77)
(0.1 8 to 0.98)
(-0.76 to 0.58)
(0.25 to 0.96)
(0.1 9 to 0.87)
(-0.81 to 0.23)
Chronic conditions (listed as secondary cause)
Respiratory system diseases
Asthma
COPD
1.35
0.01
(-0.31 to 3.03)
(-0.49 to 0.52)
Circulatory system diseases
Atherosclerosis
Atherosclerotic CVD
Atherosclerotic heart disease
Congestive heart disease
Atrial fibrillation
Stroke
-0.72
0.74
-0.38
-0.04
1.66
0.17
(-1 .89 to 0.45)
(-0.86 to 2.37)
(-1 .70 to 0.96)
(-0.39 to 0.30)
(0.03 to 3.32)
(-0.28 to 0.62)
Other diseases
Diabetes
Inflammatory diseases
0.19
0.18
(-0.46 to 0.84)
(-1 .09 to 1 .46)
"These estimates represent the additional percent change in mortality for persons who had the
characteristic being examined compared to persons who did not have the characteristic, when the mean
O3 level of the previous 3 days increased 10 ppb. These values were not standardized because they do
not represent the actual effect estimate for the characteristic being evaluated, but instead, the difference
between effect estimates for persons with versus without the condition.
bCompared with males in the same age group.
Source: Used with permission from Lippincott Williams & Wilkins, Medina-Ramon and Schwartz (2008,
Community-level Characteristics
6 Rather than using individual-level characteristics, several studies examined city-level (i.e.,
7 ecological) variables to explain city-to-city variation in estimated O3 mortality risk estimates. Bell
8 and Dominici (2008, 193828) investigated whether community-level characteristics, such as race,
9 income, education, urbanization, transportation use, PM and O3 levels, number of O3 monitors,
10 weather, and air conditioning use could explain the heterogeneity in O3-mortality risk estimates
11 across cities. The authors analyzed 98 U.S. urban communities from NMMAPS for the period
March 2011
6-154
DRAFT - DO NOT CITE OR QUOTE
-------
1 1987-2000. In the all-year regression model that included no community-level variables, a 20-ppb
2 increase in 24-h avg O3 concentrations during the previous week was associated with a 1.04% (95%
3 CI: 0.56, 1.55) increase in mortality. Bell and Dominci (2008, 193828) found that higher
4 Os-mortality effect estimates were associated with higher: percent unemployment, fraction of the
5 population Black/African-American, percent of the population that take public transportation to
6 work; and with lower: temperatures and percent of households with central air conditioning
7 (Figure 6-31). The negative percent change in O3-mortality risk estimates reported for city-specific
8 temperature and prevalence of central air conditioning in this analysis confirm the result from the
9 meta-analyses reviewed in the 2006 O3 AQCD.
10 The APHENA project (Katsouyanni et al, 2009, 199899) examined potential effect
11 modification of O3 risk estimates in the Canadian, European, and U.S. data sets using a consistent set
12 of city-specific variables. Table 6-39 presents the results from all age analyses for all-cause mortality
13 using all-year O3 data for the average of lag 0-1 day. While there are several significant effect
14 modifiers in the U.S. data, the results are mostly inconsistent with the results from the Canadian and
15 European data sets. The positive effect modification by percentage unemployed and the negative
16 effect modification by mean temperature (i.e., a surrogate for air conditioning rate) are consistent
17 with the results reported by Bell and Dominici (2008, 193828) discussed above. However, the lack
18 of consistency across the data sets, even between the Canadian and U.S. data, makes it difficult to
19 interpret the results. Some of these associations may be due to coincidental correlations with other
20 unmeasured factors that vary regionally (e.g., mean SO2 tend to be higher in the eastern U.S.).
March 2011 6-155 DRAFT - DO NOT CITE OR QUOTE
-------
345678
Percentage of population unemployed
0 10 20 30 40 50 60
Percentage of population
Black/African American
50 55 60 65 70 75
Long-term temperature (°F)
10
20
30
40
Percentage of population taking
public transportation to work
50
£ 8
i «5
E o
s-1
1 o-
-2
,O* n ®
0 20 40 60 80
Percentage of households with central AC
Source: Used with permission from Johns Hopkins Bloomberg School of Public Health, Bell and Dominici (2008,1938281.
Figure 6-31. Ozone mortality risk estimates and community-specific characteristics, U.S.,
1987-2000. [The size of each circle corresponds to the inverse of the standard error
of the community's maximum likelihood estimate. Risk estimates are fora 10-ppb
increase in 24-h avg ozone concentrations during the previous week.
March 2011
6-156
DRAFT - DO NOT CITE OR QUOTE
-------
Table 6-39. Percent change in all-cause mortality, for all ages, associated with a 40-ppb increase in 1-h
max ozone concentrations at Lag 0-1 at the 25th and 75th percentile of the center-specific
distribution of selected effect modifiers
Effect
Modifier
N02CV
Mean S02
03CV
Mean
N02/PM10
Mean
Temperature
% > 75 yr
Age
standardized
Mortality
%
Unemployed
26th
Percentile
Estimate
(96% Cl)
3.10
(1.90,4.40)
2.22
(0.71,3.83)
2.86
(0.79, 5.05)
3.91
(2.54, 5.29)
2.86
(0.95, 4.72)
2.22
(0.79, 3.58)
2.62
(0.79, 4.48)
2.78
(1.42,4.07)
Canada
76th
Percentile
Estimate
(96% Cl)
3.99
(2.38, 5.62)
4.72
(2.94,6.61)
3.50
(2.14,4.89)
2.54
(0.95,4.15)
3.50
(2.22, 4.89)
4.23
(3.02, 5.54)
4.07
(2.22, 5.87)
3.75
(2.54, 4.89)
26th
Percentile
Estimate
(Value (96% Cl)
133 1'66
^ (0.71,2.62)
91R 1'58
Z1b (0.47,2.62)
nRn Z62
°'60 (1.50,3.75)
ms 174
"'•°° (0.87,2.70)
n „, 1.58
u'w (0.39, 2.86)
9RR 1'50
^•D0 (0.55, 2.46)
114 1.10
Llq (-0.16,2.38)
1RR 1'42
'•°° (-0.47,3.34)
Europe
76th
Percentile
Estimate
(96% Cl) (Value
(-0.08, 2.78) •°'49
1.66 n,R
(0.39,2.86) u'1b
1.10 ,Rc
(0.24,1.98) ~^DO
(0.47, 2.62) ~°'43
(0.31,2.78) "°'°4
1'82 052
(0.55,3.10) u'°^
1.98 10?
(0.79,3.26) LU'
1.34 ooy
(-0.47,3.18) u'u'
26th
Percentile
Estimate
(96% Cl)
1.26
(0.47,1.98)
0.47
(-0.47,1.42)
0.16
(-0.70, 1.10)
-0.08
(-1.02,0.95)
2.14
(1.34,2.94)
1.02
(0.24, 1.90)
0.00
(-0.94, 0.87)
0.16
(-0.78, 1.18)
U.S.
76th
Percentile
Estimate
(96% Cl)
0.08
(-0.78, 0.95)
1.98
(1.10,2.94)
1.50
(0.71,2.22)
1.26
(0.47, 2.06)
0.00
(-0.78, 0.79)
1.02
(0.31, 1.74)
1.58
(0.87, 2.38)
1.50
(0.71,2.30)
t Value
-2.87
2.79
2.68
2.64
-4.40
-0.02
3.81
2.45
Source: Adapted with permission from Health Effects Institute, Katsouyanni et al. (2009, 1998991
Regional Pattern of Ozone-Mortality Risk Estimates
1 In addition to examining whether individual- and community-level factors modify the
2 O3-mortality association, studies also examined whether these associations varied regionally within
3 the U.S. Bell and Dominici (2008, 193828). in the study discussed above, also noted that
4 O3-mortality risk estimates were higher in the Northeast (1.44% [95% Cl: 0.78, 2.10%]) and
5 Industrial Midwest (0.73% [95% Cl: 0.11, 1.35%]), while null associations were observed in the
6 Southwest and Urban Midwest (Table 6-40). The regional heterogeneity in O3-mortality risk
7 estimates was further reflected by Bell and Dominici (2008, 193828) in a map of community-
8 specific Bayesian O3-mortality risk estimates (Figure 6-32). It is worth noting that in the analysis of
9 PMio using the same data set, Peng et al. (2005, 087463) also found that both the Northeast and
10 Industrial Midwest showed particularly elevated effects, especially during the summer months. As
11 mentioned above, although no evidence for confounding of O3 mortality risk estimates by PMi0 was
12 observed, Bell et al. (2007, 093256) did find regional differences in the correlation between O3 and
13 PMio. Thus, the heterogeneity in O3 mortality risk estimates may need to be examined as a function
14 of the correlation between PM and O3.
March 2011
6-157
DRAFT - DO NOT CITE OR QUOTE
-------
1 Smith et al. (2009, 199750). as discussed earlier, also examined the regional difference in O3
2 mortality risk estimates across the same seven regions and similarly found evidence for regional
3 heterogeneity. In addition, Smith et al. (2009, 199750) constructed spatial maps of the risk estimates
4 by an extension of a hierarchical model that allows for spatial auto-correlation among the city-
5 specific random effects. Figure 6-33 presents the spatial map of O3 mortality coefficients from the
6 Smith et al. (2009, 199750) analysis that used 8-h max O3 concentrations during the summer. The
7 results from the Bell and Dominici (2008, 193828) analysis (Figure 6-32) shows much stronger
8 apparent heterogeneity in O3-mortality risk estimates across cities than the smoothed map from
9 Smith et al. (2009, 199750) (Figure 6-33), but both maps generally show larger risk estimates in the
10 eastern region of the U.S.
Table 640. Percentage increase in daily mortality fora 10-ppb increase in 24-h avg ozone
concentrations during the previous week by geographic region in the U.S., 1987-2000
No. of Communities
Regional Estimate
95% PI*
Regional results
Industrial Midwest
Northeast
Northwest
southern California
Southeast
Southwest
Urban Midwest
20
16
12
7
26
9
7
0.73
1.44
0.08
0.21
0.38
-0.06
-0.05
0.11, 1.35
0.78, 2.10
-0.92, 1.09
-0.46, 0.88
-0.07, 0.85
-0.92, 0.81
-1.28, 1.19
National results
All continental communities
All communities
97
98
0.51
0.52
0.27, 076
0.28, 0.77
Source: Used with permission from Johns Hopkins Bloomberg School of Public Health, Bell and Dominici (2008,;
March 2011
6-158
DRAFT - DO NOT CITE OR QUOTE
-------
<0.0
Source: Used with permission from Johns Hopkins Bloomberg School of Public Health, Bell and Dominici, online supplement (2008, 1938281.
Figure 6-32. Community-specific Bayesian ozone-mortality risk estimates in 98 U.S.
communities.
8H: summer
Source: Used with permission from Informa UK Ltd., Smith et al. (2009, 1997501.
Figure 6-33. Map of spatially dependent ozone-mortality coefficients for 8-h max ozone
concentrations using summer data.
March 2011
6-159
DRAFT - DO NOT CITE OR QUOTE
-------
6.6.2.3. Interaction
1 The terms effect modification and interaction are often used interchangeably, but theoretically
2 they represent different concepts. Although interactions can lead to either antagonistic or synergistic
3 effects, most studies attempt to identify potential factors that interact synergistically with O3 to
4 increase the risk of mortality. Within this section, interactive effects are defined as time-varying
5 covariates, such as temperature and co-pollutants that are included in 1st stage time-series regression
6 models. To date, only a few time-series studies have investigated the potential interaction between
7 O3 exposure and co-pollutants or weather variables. This can be attributed to the moderate to high
8 correlation between O3 and these covariates, which makes such investigations methodologically
9 challenging.
10 Ren et al. (2008, 093281) examined the possible synergistic effect between O3 and temperature
11 on mortality in the 60 largest eastern U.S. communities from the NMMAPS data during the warm
12 months (i.e., April to October) from 1987-2000. This analysis was restricted to the eastern areas of
13 the U.S. (i.e., Northeast, Industrial Midwest and Southeast) because a previous study which focused
14 specifically on the eastern U.S. found that temperature-mortality patterns differ between the
15 northeast and southeast regions possibly due to climatic differences (Curriero et al., 2002, 055878).
16 To examine possible geographic differences in the interaction between temperature and O3, Ren et al.
17 (2008, 093281) further divided the NMMAPS regions into the Northeast, which included the
18 Northeast and Industrial Midwest regions (34 cities), and the Southeast, which included the
19 Southeast region (26 cities). The potential synergistic effects between O3 and temperature were
20 examined using two different models. Model 1 included an interaction term in a Generalized
21 Additive Model (GAM) for O3 and maximum temperature (3-day avg values were used for both
22 terms) to examine the bivariate response surface and the pattern of interaction between the two
23 variables in each community. Model 2 consisted of a Generalized Linear Model (GLM) that used
24 interaction terms to stratify by "low," "moderate," and "high" temperature days using the first and
25 third quartiles of temperature as cut-offs to examine the percent increase in mortality in each
26 community. Furthermore, a two-stage Bayesian hierarchical model was used to estimate the overall
27 percent increase in all-cause mortality associated with short-term O3 exposure across temperature
28 levels and each region using model 2. The same covariates were used in both model 1 and 2. The
29 bivariate response surfaces from model 1 suggest possible interactive effects between O3 and
30 temperature although the interpretation of these results is not straightforward due to the high
31 correlation between these terms. The apparent interaction between temperature and O3 as evaluated
32 in model 2 varied across geographic regions. In the northeast region, a 20-ppb increase in 24-h avg
33 O3 concentrations at lag 0-2 was associated with an increase of 4.49% (95% posterior interval [PI]:
34 2.39, 6.36%), 6.21% (95% PI: 4.47, 7.66%) and 12.8% (95% PI: 9.77, 15.7%) in mortality at low,
35 moderate and high temperature levels, respectively. The corresponding percent increases in mortality
36 in the southeast region were 2.27% (95% PI: -2.23, 6.46%) for low temperature, 3.02% (95% PI:
37 0.44, 5.70%) for moderate temperature, and 2.60% (95% PI: -0.66, 6.01%) for high temperature.
March 2011 6-160 DRAFT - DO NOT CITE OR QUOTE
-------
1 When examining the relationship between temperature and O3-related mortality, the results
2 reported by Ren et al. (2008, 093281) (i.e., higher O3-mortality risks on days with higher
3 temperatures) may appear to contradict the results of Bell and Dominici (2008, 193828) described
4 earlier (i.e., communities with higher temperature have lower O3-mortality risk estimates). However,
5 the observed difference in results can be attributed to the interpretation of effect modification in a
6 2nd stage regression which uses long-term average temperatures, as was performed by Bell and
7 Dominici (2008, 193828). compared to a first-stage regression that examines the interaction between
8 daily temperature and O3-related mortality. In this case, the second-stage regression results from Bell
9 and Dominici (2008, 193828) indicate that a city with lower temperatures, on average, tend to show
10 a stronger O3 mortality effect, whereas, in the first-stage regression performed by Ren et al. (2008,
11 093281). the days with higher temperature tend to show a larger O3-mortality effect. This observed
12 difference may in part reflect the higher air conditioning use in communities with higher long-term
13 average temperatures. Therefore, the findings from Ren et al. (2008, 093281) indicating generally
14 lower O3 risk estimates in the southeast region where the average temperature is higher than in the
15 northeast region is consistent with the regional results reported by Bell and Dominici (2008,
16 193828). As demonstrated by the results from both Ren et al. (2008, 093281) and Bell and Dominici
17 (2008, 193828) caution is required when interpreting results from studies that examined interactive
18 effects using two different approaches because potential effect modification as suggested in a
19 SECOND stage regression generally does not provide evidence for a short-term interaction
20 examined in a first-stage regression. Overall, further examination of the potential interactive
21 (synergistic) effects of O3 and covariates in time-series regression models is required to more clearly
22 understand the factors that may influence O3 mortality risk estimates.
6.6.2.4. Evaluation of the Ozone-Mortality C-R Relations hip and Related Issues
23 Evaluation of the O3-mortality concentration-response relationship is not straightforward
24 because the evidence from multicity studies (using log-linear models) suggests that O3-mortality
25 associations are highly heterogeneous across regions. In addition, there are numerous issues that may
26 influence the shape of the O3-mortality concentration-response relationship that warrant examination
27 including: multi-day effects (distributed lags), potential adaptation, mortality displacement (i.e.,
28 hastening of death by a short period), and the exposure metric used to compute risks (e.g., 1-h daily
29 max versus 24-h avg). The following section presents the recent studies identified that conducted an
30 initial examination of these issues.
Multiday Effects, Mortality Displacement, and Adaptation
31 The pattern of positive lagged associations followed by negative associations in a distributed
32 lag model may be considered an indication of "mortality displacement" (i.e., deaths are occurring in
33 frail individuals and exposure is only moving the day of death to a day slightly earlier). Zanobetti
34 and Schwartz (2008, 101596) examined this issue in 48 U.S. cities during the warm season (i.e.,
35 June-August) for the years 1989-2000. In an initial analysis, the authors applied a GLM to examine
March 2011 6-161 DRAFT - DO NOT CITE OR QUOTE
-------
1 same-day O3-mortality effects, and in the model included an unconstrained distributed lag for
2 apparent temperature to take into account the effect of temperature today and the previous 7 days. To
3 examine mortality displacement Zanobetti and Schwartz (2008, 101596) refit models using two
4 approaches: an unconstrained and a smooth distributed lag each with 21-day lags for O3. In this
5 study, all-cause mortality as well as cause-specific mortality (i.e., cardiovascular, respiratory, and
6 stroke) were examined for evidence of mortality displacement. The authors found a 0.96% (95% CI:
7 0.60, 1.30%) increase in all-cause mortality across all 48 cities for a 30-ppb increase in 8-h max O3
8 concentrations at lag 0 whereas the combined estimate of the unconstrained distributed lag model
9 (lag 0-20) was 1.54% (95% CI: 0.15, 2.91%). Similarly, when examining the cause-specific
10 mortality results (Table 6-41), larger risk estimates were observed for the distributed lag model
11 compared to the lag 0 day estimates. However, for stroke a slightly larger effect was observed at lags
12 4-20 compared to lags 0-3 suggesting a larger window for O3-induced stroke mortality. This is
13 further supported by the sum of lags 0 through 20 days showing the greatest effect. Overall, these
14 results suggest that estimating the mortality risk using a single day of O3 exposure may
15 underestimate the public health impact, but the extent of multi-day effects appear to be limited to a
16 few days. This is further supported by the shape of the combined smooth distributed lag
17 (Figure 6-34). It should be noted that the proportion of total variation in the effect estimates due to
18 the between-cities heterogeneity, as measured by 12 statistic, was relatively low (4% for the lag 0
19 estimates and 21% for the distributed lag), but 21 out of the 48 cities exhibited null or negative
20 estimates. As a result, the estimated shape of the distributed lag cannot be interpreted as a general
21 form of lag structure of associations applicable to all the cities included in this analysis.
March 2011 6-162 DRAFT - DO NOT CITE OR QUOTE
-------
Table 641. Estimated effect of a 10-ppb increase in 8-h max ozone concentrations on mortality during
the summer months for single-day and distributed lag models
% 95% Cl
Total mortality
LagO
Sum lags 0-20
Sum lags 0-3
Sum lags 4-20
0.32
0.51
0.53
-0.02
0.20
0.05
0.28
-0.35
0.43
0.96
0.77
0.31
Cardiovascular mortality
LagO
Sum lags 0-20
Sum lags 0-3
Sum lags 4-20
0.47
0.49
0.80
-0.23
0.30
-0.01
0.48
-0.67
0.64
1.00
1.13
0.22
Respiratory mortality
LagO
Sum lags 0-20
Sum lags 0-3
Sum lags 4-20
0.54
0.61
0.83
-0.24
0.26
-0.41
0.38
-1.08
0.81
1.65
1.28
0.60
Stroke
LagO
Sum lags 0-20
Sum lags 0-3
Sum lags 4-20
0.37
2.20
0.92
1.26
0.01
0.76
0.26
0.05
0.74
3.67
1.59
2.49
Source: Used with permission from American Thoracic Society, Zanobetti and Schwartz (2008, 101596V
March 2011 6-163 DRAFT - DO NOT CITE OR QUOTE
-------
<*?
o '
0)
I
CO
tu
O
O '
10
20
Day Lag
Source: Used with permission from American Thoracic Society, Zanobetti and Schwartz (2008, 1015961.
Figure 6-34. Estimated combined smooth distributed lag for 48 U.S. cities during the summer
months. [The triangles represent the percent increase in all-cause mortality for a
10-ppb increase in 8-h max ozone concentrations at each lag while the shaded areas
are the 95% point-wise confidence intervals.
1 Samoli et al. (2009, 195855) also investigated the temporal pattern of mortality effects in
2 response to short-term exposure to O3 in 21 European cities that were included in the APHEA2
3 project. Using a method similar to Zanobetti and Schwartz (2008, 101596). the authors applied
4 unconstrained distributed lag models with lags up to 21 days in each city during the summer months
5 (i.e., June through August) to examine the effect of O3 on all-cause, cardiovascular, and respiratory
6 mortality. They also applied a generalized additive distributed lag model to obtain smoothed
7 distributed lag coefficients. However, unlike Zanobetti and Schwartz (2008, 101596). Samoli et al.
8 (2009, 195855) controlled for temperature using a linear term for humidity and an unconstrained
9 distributed lag model of temperature at lags 0-3 days. The choice of 0- through 3-day lags of
10 temperature was based on a previous European multicity study (Baccini et al., 2008, 633196). which
11 suggested that summer temperature effects last only a few days. Upon combining the individual city
12 estimates across cities in a second stage regression, Samoli et al. (2009, 195855) found that the
13 estimated effects on respiratory mortality were extended for a period of two weeks. However, for all-
14 cause and cardiovascular mortality, the 21-day distributed lag models yielded null or (non-
15 significant) negative estimates (Table 6-42). Figure 6-35 shows the distributed lag coefficients for
16 all-cause mortality, which exhibit a declining trend and negative coefficients beyond 5-day lags. The
17 authors' interpretation of these results was that "using single-day exposures may have overestimated
March 2011
6-164
DRAFT - DO NOT CITE OR QUOTE
-------
1 the effects on all-cause and cardiovascular mortality, but underestimated the effects on respiratory
2 mortality." Thus, the results in part suggest evidence of mortality displacement for all-cause and
3 cardiovascular mortality.
Table 642. Estimated percent increase in cause-specific mortality (and 95% CIs) for a 10-|jg/m3
increase in maximum 8-h ozone during June-August, for the same day (lag 0), the average
of the same and previous day (lag 0-1), the unconstrained distributed lag model for the sum
of 0-20 days and the penalized distributed lag model (lag 0-20)
Fixed effects
Percent increase
(95% Cl)
Random effects
Percent increase
(95% Cl)
Total mortality
LagO
Average lags 0-1
Sum lags 0-20, unconstrained
Sum lags 0-20, penalized
0.28(0.11 to 0.45)
0.24(0.15to0.34)
0.01 (-0.40 to 0.41)
0.01 (-0.41 to 0.42)
0.28 (0.07 to 0.48)
0.22 (0.08 to 0.35)
-0.54 (-1.28 to 0.20)
-0.56 (-1.30 to 0.1 9)
Cardiovascular mortality
LagO
Average lags 0-1
Sum lags 0-20, unconstrained
Sum lags 0-20, penalized
0.43(0.18(00.69)
0.33(0.19(00.48)
-0.33 (-0.93 to 0.29)
-0.32 (-0.92 to 0.28)
0.37 (0.05 to 0.69)
0.25 (0.03 to 0.47)
-0.62 (-1 .47 to 0.24)
-0.57 (-1.39 to 0.26)
Respiratory mortality
LagO
Average lags 0-1
Sum lags 0-20, unconstrained
Sum lags 0-20, penalized
0.36 (-0.21 to 0.94)
0.40(0.11 to 0.70)
3.35(1.90to4.83)
3.66 (2.25 to 5.08)
0.36 (-0.21 to 0.94)
0.40(0.11 to 0.70)
3.35(1.90to4.83)
3.66 (2.25 to 5.08)
Source: Used with permission from BMJ Group, Samoli et al. (2009, 195855V
4 Although the APHENA project (Katsouyanni et al., 2009, 199899) did not specifically
5 investigate mortality displacement and therefore did not consider longer lags (e.g., lag > 3 days), the
6 study did present O3 risk estimates for lag 0-1, lag 1, and a distributed lag model of 0-2 days in the
7 Canadian, European, and U.S. datasets. Katsouyanni et al. (2009, 199899) found that the results
8 somewhat vary across the regions, but, in general, there was no indication that the distributed lag
9 model with up to a 2-day lag yielded meaningfully larger O3 mortality risk estimates than the lag 0-1
10 and lag 1 results. For example, for all-cause mortality, using the model with natural splines and
11 8 df/year to adjust for seasonal trends, a reported percent excess risk for mortality for a 40-ppb
12 increase in 1-h max O3 concentrations for lag 0-1, lag 1, and the distributed lag model (lag 0-2) was
13 2.70% (95% Cl: 1.02, 4.40%), 1.42% (95% Cl: 0.08, 2.78%), and 3.02% (95% Cl: 1.10, 4.89%),
14 respectively. Thus, the observed associations appear to occur over a short time period, (i.e., a
15 few days).
16 When comparing the studies that explicitly examined the potential for mortality displacement
17 in the O3-mortality relationship, the results from Samoli et al. (2009, 195855). which provide
18 evidence that suggests mortality displacement, are not consistent with those reported by Zanobetti
March 2011
6-165
DRAFT - DO NOT CITE OR QUOTE
-------
1 and Schwartz (2008, 101596). However, the shapes of the estimated smooth distributed lag
2 associations are similar (Figure 6-34 versus Figure 6-35). A closer examination of these figures
3 shows that in the European data beyond a lag of 5 days the estimates remain negative whereas in the
4 U.S. data the results remain near zero for the corresponding lags. These observed difference could be
5 due the differences in the model specification between the 2 studies, specifically the use of: an
6 unconstrained distributed lag model for apparent temperature up to 7 previous days (Zanobetti and
7 Schwartz, 2008, 101596) versus a linear term for humidity and an unconstrained distributed lag
8 model of temperature up to 3 previous days (Samoli et al., 2009, 195855): and natural cubic splines
9 with 2 df per season (Zanobetti and Schwartz, 2008, 101596) versus dummy variables per month per
10 year to adjust for season (Samoli et al., 2009, 195855). It is important to note, that these differences
11 in model specification may have also influenced the city-to-city variation in risk estimates observed
12 in these two studies (i.e., homogenous estimates across cities in Zanobetti and Schwartz (2008,
13 101596) and heterogeneous estimates across cities in Samoli et al. (2009, 195855). Overall, the
14 evidence of mortality displacement remains unclear, but Samoli et al. (2009, 195855). Zanobetti and
15 Schwartz (2008, 101596). and Katsouyanni et al. (2009, 199899) all suggest that the positive
16 associations between O3 and mortality are observed mainly in the first few days after exposure.
i _
o
'////„„
15
20
10
Day Lag
Source: Used with permission from BMJ Group, Samoli et al. (2009,1958551.
Figure 6-35. Estimated combined smooth distributed lag in 21 European cities during the
summer (June-August) months. [The triangles represent the percent increase in all-
cause mortality for a 10-ug/m3 increase in 8-h max ozone concentrations at each lag;
the shaded area represents the 95% CIs.
March 2011
6-166
DRAFT - DO NOT CITE OR QUOTE
-------
Adaptation
1 Controlled human exposure studies have demonstrated an adaptive response to O3 exposure
2 for respiratory effects, such as lung function decrements, but this issue has not been examined in the
3 epidemiologic investigation of mortality effects of O3. Zanobetti and Schwartz (2008, 195755)
4 examined if there was evidence of an adaptive response in the O3-mortality relationship in 48 U.S.
5 cities from 1989 to 2000 (i.e., the same data analyzed in Zanobetti and Schwartz (2008, 101596)).
6 The authors examined all-cause mortality using a case-crossover design to estimate the same-day
7 (i.e., lag 0) effect of O3, matched on referent days from every-3rd-day in the same month and year as
8 the case. Zanobetti and Schwartz (2008, 195755) examined O3-mortality associations by: season,
9 month in the summer season (i.e., May through September), and age categories in the summer
10 season (Table 6-43). The estimated O3 mortality risk estimate at lag 0 was found to be highest in the
11 summer (1.51% [95% CI: 1.14, 1.87%]; lag 0 fora 30-ppb increase in 8-h max O3 concentrations),
12 and, within the warm months, the association was highest in July (1.96% [95% CI: 1.42, 2.48%];
13 lag 0) (Table 6-43). Upon further examination of the summer months, the authors also observed
14 diminished effects in August (0.84% [95% CI: 0.33, 1.39%]; lag 0). Based on these results, the
15 authors concluded that the mortality effects of O3 appear diminished later in the O3 season.
16 To further evaluate the potential adaptive response observed in Zanobetti and Schwartz (2008,
17 195755) the distribution of the O3 concentrations across the 48 U.S. cities during July and August
18 was examined. Both July and August were found to have comparable means of 48.6 and 47.9 ppb
19 with a reported maximum value of 97.9 and 96.0 ppb, respectively. Thus, the observed reduction in
20 O3-related mortality effect estimates in August (0.84%) compared to July (1.96%) appears to support
21 the existence of an adaptive response. However, unlike an individual's adaptive response to
22 decrements in lung function from short-term O3 exposure, an examination of mortality prevents a
23 direct observation of adaptation. Rather, for mortality the adaptation hypothesis is tested with a tacit
24 assumption that, whatever the mechanism for O3-induced mortality, the risk of death from short-term
25 O3 exposure is reduced over the course of the summer months through repeated exposures. This idea
26 would translate to a smaller population that would die from O3 exposure towards the end of summer.
27 This may complicate the interpretation of the distributed lag coefficients with long lag periods
28 because the decreased coefficients may reflect diminished effects of the late summer, rather than
29 diminished effects that are constant across the summer. These inter-twined issues need to be
30 investigated together in future research.
March 2011 6-167 DRAFT - DO NOT CITE OR QUOTE
-------
Table. 643. Percent excess all-cause
day, by season, month, and
mortality perl 0-ppb
age groups
increase in daily 8-h
max ozone on the same
% 95% Cl
By Season
Winter
Spring
Summer
Fall
-0.13
0.35
0.50
0.05
-0.56
0.16
0.38
-0.14
0.29
0.54
0.62
0.24
By Month
May
June
July
August
September
0.48
0.46
0.65
0.28
-0.09
0.28
0.24
0.47
0.11
-0.35
0.68
0.68
0.82
0.46
0.16
By Age Group
0-20
21-30
31-40
41-50
51-60
61-70
71-80
80
0.08
0.10
0.07
0.08
0.54
0.38
0.50
0.29
-0.42
-0.67
-0.38
-0.27
0.19
0.16
0.32
0.13
0.57
0.87
0.52
0.43
0.89
0.61
0.67
0.44
Source: Used with permission from BioMed Central Ltd., Zanobetti and Schwartz (2008, 1957551
Ozone-Mortality Concentration-Response Relationship and Threshold Analyses
1 Several of the recent studies evaluated have applied a variety of statistical approaches to
2 examine the shape of the O3-mortality C-R relationship and whether a threshold exists. The approach
3 used by Bell et al. (2006, 087680) consisted of applying four statistical models to the NMMAPS
4 data, which included 98 U.S. communities for the period 1987-2000. These models included: a linear
5 analysis (i.e., any change in O3 concentration can be associated with mortality) (Model 1); a subset
6 analysis (i.e., examining O3-mortality relationship below a specific concentration, ranging from 5 to
7 60 ppb) (Model 2); a threshold analysis (i.e., assuming that an association between O3 and mortality
8 is observed above a specific concentration and not below it, using the threshold values set at an
9 increment of 5 ppb between 0 to 60 ppb and evaluating a presence of a local minima in AICs
10 computed at each increment) (Model 3); and nonlinear models using natural cubic splines with
11 boundary knots placed at 0 and 80 ppb, and interior knots placed at 20 and 40 ppb (Model 4). Atwo-
12 stage Bayesian hierarchical model was used to examine these models and O3-mortality risk estimates
13 at the city-level in the first stage analysis and aggregate estimates across cities in the 2nd stage
14 analysis using the average of 0- and 1-day lagged 24-h avg O3 concentrations. The results from all of
15 these models suggest that if a threshold exists it does so well below the current O3 NAAQS. When
16 restricting the analysis to all days when the current 8 h standard (i.e., 84 ppb daily 8-h max) is met in
March 2011 6-168 DRAFT - DO NOT CITE OR QUOTE
-------
1 each community, Bell et al. (2006, 087680) found there was still a 0.60% (95% PI: 0.30, 0.90%)
2 increase in mortality per 20-ppb increase in 24-h avg O3 concentrations at lag 0-1. Figure 6-36 shows
3 the combined C-R curve obtained using the nonlinear model (Model 4). Although these results
4 suggest the lack of threshold in the O3-mortality relationship, it is difficult to interpret such a curve
5 because it does not take into consideration the heterogeneity in O3-mortality risk estimates across
6 cities.
JS 4
o
3
CD •)
0» *-
u
Z 1
0>
o
Central estimate
95% posterior interval
0 20 40 60 80
Average of same and previous days' 03 (ppb)
Source: Bell et al. (2006,
Figure 6-36. Estimated combined C-R curve for ozone and nonaccidental mortality using the
nonlinear (spline) model.
7 The APHENA project (Katsouyanni et al., 2009, 199899) also analyzed the Canadian and
8 European datasets (the U.S. data were analyzed for PMi0 only) for evidence of a threshold, using the
9 threshold analysis method (Model 3) applied in Bell et al.'s (2006, 087680) study described above.
10 There was no evidence of a threshold in the Canadian data (i.e., the pattern of AIC values for each
11 increment of a potential threshold value varied across cities, most of which showed no local
12 minima). Likewise, the threshold analysis conducted using the European data also showed no
13 evidence of a threshold.
14 Additional threshold analyses were conducted using NMMAPS data, by Xia and Tong (2006,
15 623157) and Stylianou and Nicolich (2009, 620299). Both studies used a new statistical approach
16 developed by Xia and Tong (2006, 623157) to examine thresholds in the O3 mortality C-R
17 relationship. The approach consisted of an extended GAM model, which accounted for the
March 2011
6-169
DRAFT - DO NOT CITE OR QUOTE
-------
1 cumulative and nonlinear effects of air pollution using a weighted cumulative sum for each pollutant,
2 with the weights (non-increasing further into the past) derived by a restricted minimization method.
3 The authors did not use the term distributed lag model, but their model has the form of distributed
4 lag model, except that it allows nonlinear functional forms. Using NMMAPS data for 1987-1994 for
5 3 U.S. cities (Chicago, Pittsburgh, and El Paso), Xia and Tong (2006, 623157) found that the extent
6 of cumulative effects of O3 on mortality were relatively short. While the authors also note that there
7 was evidence of a threshold effect around 24-h avg concentrations of 25 ppb, the threshold values
8 estimated in the analysis were sometimes in the range where data density was low. Thus, this
9 threshold analysis needs to be replicated in a larger number of cities. It should be noted that the
10 model used in this analysis did not include a smooth function of days to adjust for unmeasured
11 temporal confounders, and instead adjusted for season using a temperature term. As a result, these
12 results need to be viewed with caution because some potential temporal confounders (e.g., influenza)
13 do not always follow seasonal patterns of temperature.
14 Stylianou and Nicolich (2009, 620299) examined the existence of thresholds following an
15 approach similar to Xia and Tong (2006, 623157) for all-cause, cardiovascular, and respiratory
16 mortality using data from NMMAPS for nine major U.S. cities (i.e., Baltimore, Chicago, Dallas/Fort
17 Worth, Los Angeles, Miami, New York, Philadelphia, Pittsburgh, and Seattle) for the years
18 1987-2000. The authors found that PMi0 and O3 were the two important predictors of mortality.
19 Stylianou and Nicolich (2009, 620299) found that the estimated O3-mortality risks varied across the
20 nine cities with the models exhibiting apparent thresholds, in the 10-45 ppb range for O3. However,
21 given the city-to-city variation in risk estimates, combining the city-specific estimates into an overall
22 estimate complicates the interpretation of a threshold. Unlike the Xia and Tong (2006, 623157)
23 analysis, Stylianou and Nicolich (2009, 620299) included a smooth function of time to adjust for
24 seasonal/temporal confounding, which could explain the difference in results between the two
25 studies.
26 In conclusion, the evaluation of the O3-mortality C-R relationship did not find any evidence
27 that supports a threshold for the association between short-term exposure to O3 and mortality. It was
28 also demonstrated that the heterogeneity in the O3-mortality relationship across cities (or regions)
29 complicates the interpretation of a combined C-R curve and threshold analysis. Additionally, given
30 the effect modifiers identified in the mortality analyses that are also expected to vary regionally (e.g.,
31 temperature, air conditioning prevalence), a national or combined analysis may not be appropriate to
32 identify whether a threshold exists in the O3-mortality C-R relationship.
6.6.2.5. Associations of Cause-Specific Mortality and Short-term Ozone Exposure
33 In the 2006 O3 AQCD, an evaluation of studies that examined cause-specific mortality found
34 consistent positive associations between short-term O3 exposure and cardiovascular mortality, with
35 less consistent evidence for associations with respiratory mortality. The majority of the evidence for
36 associations between O3 exposure and cause-specific mortality were from single-city studies, which
37 had small daily mortality counts and subsequently limited statistical power to detect associations.
March 2011 6-170 DRAFT - DO NOT CITE OR QUOTE
-------
1 New multicity studies evaluated in this review build upon and confirm the associations
2 between short-term O3 exposure and cause-specific mortality identified in the 2006 O3 AQCD
3 (Figure 6-37; Table 6-44). In APHENA, a multicontinent study that consisted of the NMMAPS,
4 APHEA2 and Canadian multicity datasets, consistent positive associations were reported for both
5 cardiovascular and respiratory mortality in all-year analyses when focusing on the natural spline
6 model with 8 df/year (Section 6.6.2.1). Cardiovascular mortality associations persisted in analyses
7 restricted to the summer season with evidence for stronger respiratory mortality associations
8 compared to the all-year analysis results (Figure 6-37; Table 6-44). Additional multicity studies from
9 the U.S. (Zanobetti and Schwartz, 2008, 101596) and Europe (Samoli et al, 2009, 195855: Stafoggia
10 et al., 2010, 625034) that conducted summer season analyses also found strong associations between
11 O3 exposure and cardiovascular and respiratory mortality.
12 Of the studies evaluated, only the APHENA study (Katsouyanni et al., 2009, 199899) and an
13 Italian multicity study (Stafoggia et al., 2010, 625034) conducted an analysis of the potential for
14 co-pollutant confounding of the O3 cause-specific mortality relationship. When focusing on the
15 natural spline model with 8 df/year and lag 1 results (as discussed in Section 6.6.2.1), the APHENA
16 study found that O3 cause-specific mortality risk estimates were fairly robust to the inclusion of
17 PMio in co-pollutant models in the European dataset with more variability in the U.S. and Canadian
18 datasets (i.e., co-pollutant risk estimates increased and decreased for respiratory and cardiovascular
19 mortality). In summer season analyses in the U.S. and Europe, the Canadian dataset did not examine
20 co-pollutant models during the summer season, cardiovascular O3 mortality risk estimates were
21 robust in the European dataset and attenuated but remained positive in the U.S. datasets; whereas,
22 respiratory O3 mortality risk estimates were attenuated in the European dataset and robust in the U.S.
23 dataset (Figure 6-37; Table 6-44). Interpretation of these results requires caution; however, due to the
24 different PM sampling schedules employed in each of these study locations (i.e., primarily every-6th
25 day in the U.S. and Canadian datasets and every-day in the European dataset). The results of the
26 summer season analyses from the APHENA study (Katsouyanni et al., 2009, 199899) are consistent
27 with those from a study of 10 Italian cities during the summer months (Stafoggia et al., 2010,
28 625034V Stafoggia et al. (2010, 625034) found that cardiovascular (14.3% [95% CI: 6.7, 22.4%])
29 and cerebrovascular (8.5% [95% CI: 0.06, 16.3%]) mortality O3 effect estimates were robust to the
30 inclusion of PM10 in co-pollutant models (14.3% [95% CI: 6.7, 23.1%] and 7.3% [95% CI: -1.2,
31 16.3], respectively), while respiratory mortality O3 effects estimates (17.6% [95% CI: 1.8, 35.5%])
32 were attenuated, but remained positive (9.2% [95% CI: -6.9, 28.8%]).
March 2011 6-171 DRAFT - DO NOT CITE OR QUOTE
-------
Study
Bell etal. (2005; 74345)a
Katsouyanni etal. (2009; 199899)
Gryparisetal. (2004;57276)a
Samoli etal. (2009; 195855)
Zanobetti and Schwartz (2008; 101596)
Stafoggia et al. (2010; 625034)
Katsouyanni etal. (2009; 199899)
Bell etal. (2005; 74345)a
Katsouyanni etal. (2009; 199899)
Gryparisetal. (2004;57276)a
Zanobetti and Schwartz (2008; 101596)
Katsouyanni etal. (2009; 199899)
Samoli etal. (2009; 195855)
Stafoggia et al. (2010; 625034)
Katsouyanni etal. (2009; 199899)
Location
U.S. and non-U.S.
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
21 European cities
21 European cities
48 U.S. cities
10 Italian cities
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
U.S. and non-U.S.
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
21 European cities
48 U.S. cities
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
21 European cities
10 Italian cities
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
Ages
All
275
235
275
235
275
NR
DL(0-2
DL(0-2
DL(0-2)
DL 0-2)
DL 0-2)
DL 0-2)
DL(0-2)b
DL(0-2)
0-1
0-1
0-3
DLO-5
DLO-2
DL(0-2)
DL(0-2)b
DL(0-2)
DL(0-2)
DL(0-2)
DL(0-2)b
DL(0-2)
NR
DL(0-2
DL(0-2
DL(0-2)
DL 0-2)
DL 0-2)
DL 0-2)
DL(0-2)b
DL(0-2)
0-1
0-3
DL(0-2)
DL(0-2)
DL(0-2)b
DL(0-2)
0-1
DL(0-5)
DL(0-2)
DL(0-2)
DL(0-2)b
DL(0-2)
Cardiovascular
Respiratory
Ail-Year
5 10 15
% Increase
25 30
Figure 6-37. Percent increase in cause-specific mortality. [Effect estimates are for a 20-ppb
increase in 24-h avg; 30 in 8-h max; and 40-ppb increase in 1-h max ozone
concentrations. Red = cardiovascular; blue = respiratory; closed circles = all-year
analysis; and open circles = summer-only analysis. An "a" represents studies from
the 2006 ozone AQCD. A "b" represents risk estimates from APHENA-Canada
standardized to an approximate IQR of 5.1 ppb for a 1-h max increase in ozone
concentrations (Section 6.2.7.2).
March 2011
6-172
DRAFT - DO NOT CITE OR QUOTE
-------
Table 644. Corresponding effect estimates for Figure 6-37
Study
Location Ages
Lag
Avg Time
% Increase (95% Cl)
Cardiovascular
All-year
Bell et al. (2005, 074345)3
Katsouyanni et al. (2009, 1 99899)
U.S. andnon-U.S. All
APHENA-U.S. >75
APHENA-Canada
APHENA-Canada
APHENA-Europe
APHENA-U.S. <75
APHENA-Canada
APHENA-Canada
APHENA-Europe
NR
DL(0-2)
DL(0-2)
DL(0-2)b
DL(0-2)
DL(0-2)
DL(0-2)
DL(0-2)b
DL(0-2)
24- h avg
1-h max
2.23(1.36,3.08)
2.30 (-1.33, 6.04)
8.96(0.75,18.6)
1.1 (0.10,2.20)
2.06 (-0.24, 4.31)
3.83 (-0.1 6, 7.95)
7.03 (-2.71, 17.7)
0.87 (-0.35, 2.10)
1.98 (-1.09, 5.13)
Summer
Gryparis et al. (2004, 057276)3
Samoli et al. (2009, 195855)
Zanobetti and Schwartz (2008, 101596)
Stafoggia et al. (2010, 625034)
Katsouyanni et al. (2009, 1 99899)
21 European cities All
21 European cities
48 U.S. cities
10 Italian cities > 35
APHENA-U.S. >75
APHENA-Canada
APHENA-Canada
APHENA-Europe
APHENA-U.S. <75
APHENA-Canada
APHENA-Canada
APHENA-Europe
0-1
0-1
0-3
DL(0-5)
DL(0-2)
DL(0-2)
DL(0-2)b
DL(0-2)
DL(0-2)
DL(0-2)
DL(0-2)b
DL(0-2)
8-h max
8-h max
8-h max
8-h max
1-h max
2.7(1.29,4.32)
1.48(0.18, 2.80)
2.42(1.45, 3.43)
14.3(6.65,22.4)
3.1 8 (-0.47, 6.95)
1.50 (-2.79, 5.95)
0.1 9 (-0.36, 0.74)
3.67 (0.95, 6.53)
6.78(2.70, 11.0)
-1.02 (-4.23, 2.30)
-0.1 3 (-0.55, 0.29)
2.22 (-1.48, 6.04)
Respiratory
All-year
Bell et al. (2005, 074345)3
Katsouyanni et al. (2009, 199899)
U.S. andnon-U.S. All
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
APHENA-U.S. >75
APHENA-Canada
APHENA-Canada
APHENA-Europe
NR
DL(0-2)
DL(0-2)
DL(0-2)b
DL(0-2)
DL(0-2)
DL(0-2)
DL(0-2)b
DL(0-2)
24- h avg
1-h max
0.94 (-1.02, 2.96)
2.54 (-3.32, 8.79)
1.02 (-11. 9, 15.9)
0.1 3 (-1.60, 1.90)
1.82 (-2. 18, 6.04)
1.10 (-6.48, 9.21)
-4.61 (-19.3, 13.3)
-0.60 (-2.70, 1 .60)
1.10 (-3.48, 5.95)
Summer
Gryparis et al. (2004, 057276)3
Zanobetti and Schwartz (2008, 101596)
Katsouyanni et al. (2009, 199899)
Samoli et al. (2009, 1 95855)
Stafoggia et al. (2010, 625034)
Katsouyanni et al. (2009, 199899)
21 European cities All
48 U.S. cities
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
21 European cities
10 Italian cities > 35
APHENA-U.S. >75
APHENA-Canada
APHENA-Canada
APHENA-Europe
0-1
0-3
DL(0-2)
DL(0-2)
DL(0-2)b
DL(0-2)
0-1
DL(0-5)
DL(0-2)
DL(0-2)
DL(0-2)b
DL(0-2)
8-h max
8-h max
1-h max
8-h max
8-h max
1-h max
6.75(4.38,9.10)
2.51 (1.14,3.89)
4.40 (-2. 10, 11.3)
26.1 (13.3, 41.2)
3.00(1.60, 4.50)
3.83 (-1.33, 9.21)
2.38(0.65,4.19)
17.6(1.78,35.5)
4.07 (-4.23, 13.0)
19.5(2.22, 40.2)
2.30 (0.28, 4.40)
2.46 (-3.40, 8.62)
'Studies from the 2006 03 AQCD.
bRisk estimates from APHENA-Canada standardized to an approximate IQR of 5.1 ppbfora 1-h max increase in 03 concentrations (Section 6.2.7.2).
1 Collectively, the results from the new multicity studies provide evidence of associations
2 between short-term O3 exposure and cardiovascular and respiratory mortality with additional
3 evidence indicating these associations persist, and in the case of respiratory mortality are
4 strengthened, in the summer season. Although co-pollutant analyses of cause-specific mortality are
March 2011
6-173
DRAFT - DO NOT CITE OR QUOTE
-------
1 limited, the APHENA study found that O3 cause-specific mortality risk estimates were fairly robust
2 to the inclusion of PMi0 in co-pollutant models in the European dataset, which is supported by the
3 results from Stafoggia et al. (2010, 625034). Additionally, APHENA found that O3 cause-specific
4 mortality risk estimates were moderately to substantially sensitive (e.g., increased or attenuated) to
5 inclusion of PMi0 in the U.S. and Canadian datasets. However, the mostly every-6th-day sampling
6 schedule for PMi0 in the U.S. and Canadian datasets greatly reduced their sample size and limits the
7 interpretation of these results.
6.6.3. Summary and Causal Determination
8 The evaluation of new multicity studies that examined the association between short-term O3
9 exposure and mortality found evidence which supports the conclusions of the 2006 O3 AQCD. These
10 new studies reported consistent positive associations between short-term O3 exposure and all-cause
11 (nonaccidental) mortality, with associations being stronger during the warm season, as well as
12 additional support for associations between O3 exposure and cardiovascular and respiratory
13 mortality.
14 New studies further examined potential confounders (e.g., co-pollutants and seasonality) of the
15 O3-mortality relationship. Because the PM-O3 correlation varies across regions, due to the difference
16 in PM chemical constituents, interpretation of the combined effect of PM on the relationship
17 between O3 and mortality is not straightforward. Unlike previous studies that were limited to
18 primarily examining the confounding effects of PMi0, the new studies expanded their analyses to
19 include multiple PM indices (e.g., PMi0, PM25, and PM components). An examination of co-
20 pollutant models found evidence that associations between O3 and all-cause mortality were robust to
21 the inclusion of PM10 or PM25 (Bell et al., 2007, 093256: Katsouyanni et al., 2009, 199899:
22 Stafoggia et al., 2010, 625034). while other studies found evidence for a modest reduction
23 (-20-30%) when examining PM10 (Smith et al. (2009, 199750). Additional evidence suggests
24 potential sensitivity (e.g., increases and attenuation) of O3 mortality risk estimates to co-pollutants
25 by age group or cause-specific mortality (e.g., respiratory and cardiovascular) (Katsouyanni et al.,
26 2009, 199899: Stafoggia et al., 2010, 625034). An examination of PM components, specifically
27 sulfate, found evidence for reductions in O3-mortality risk estimates in co-pollutant models (Franklin
28 and Schwartz, 2008, 156448). Overall, across studies, the potential impact of PM indices on
29 O3-mortality risk estimates tended to be much smaller than the variation in O3-mortality risk
30 estimates across cities. Although some studies suggest that O3-mortality risk estimates may be
31 confounded by PM or its chemical components the interpretation of these results requires caution
32 due to the limited PM datasets used as a result of the every-3rd- and 6th-day PM sampling schedule.
33 When examining the potential for seasonal confounding of the O3-mortality relationship it was
34 observed that the extent of smoothing or the methods used for adjustment can influence O3 risk
35 estimates because of the opposing seasonal trends of O3 and mortality when not instituting enough
36 degrees of freedom to control for temporal/seasonal trends (Katsouyanni et al., 2009, 199899).
March 2011 6-174 DRAFT - DO NOT CITE OR QUOTE
-------
1 The multicity studies evaluated in this review also examined the regional heterogeneity
2 observed in O3-mortality risk estimates. These studies provide evidence which suggests generally
3 higher O3-mortality risk estimates in northeastern U.S. cities with some regions showing no
4 associations between O3 exposure and mortality (e.g., Southwest, Urban midwest) (Bell and
5 Dominici, 2008, 193828: Smith et al, 2009, 199750). Multicity studies that examined individual-
6 and community-level characteristics identified characteristics that may explain the observed regional
7 heterogeneity in O3-mortality risk estimates as well as characteristics of populations potentially
8 susceptible to O3-related health effects. An examination of community-level characteristics found an
9 increase in the O3-mortality risk estimates in cities with higher unemployment, percentage of the
10 population Black/African-American, percentage of the working population that uses public
11 transportation, lower temperatures, and lower prevalence of central air conditioning (Medina-Ramon
12 and Schwartz, 2008, 193829). Additionally, a potential interactive, or synergistic, effect on the
13 O3-mortality relationship was observed when examining differences in the O3-mortality association
14 across temperature levels (Ren et al. (2008, 093281). An examination of individual-level
15 characteristics found evidence that older age, female sex, Black race, having atrial fibrillation, and
16 out-of hospital deaths, specifically in those individuals with diabetes, are significant effect modifiers
17 of O3-mortality associations (Medina-Ramon and Schwartz, 2008, 193829; Stafoggia et al., 2010,
18 625034). and may increase susceptibility to O3-related health effects. Overall, additional research is
19 needed to further confirm whether these characteristics, individually or in combination, can explain
20 the observed regional heterogeneity.
21 Additional studies were evaluated that examined factors, such as multi-day effects, mortality
22 displacement, adaptation, and whether a threshold exists in the O3-mortality relationship, which may
23 influence the shape of the O3-mortality C-R curve. An examination of multiday effects in a U.S. and
24 European multicity study found conflicting evidence for mortality displacement, but both studies
25 suggest that the positive associations between O3 and mortality are observed mainly in the first
26 few days after exposure (Samoli et al., 2009, 195855: Zanobetti and Schwartz, 2008, 101596). A
27 U.S. multicity study found evidence of an adaptive response to O3 exposure, with the highest risk
28 estimates earlier in the O3 season (i.e., July) and diminished effects later (i.e., August) (Zanobetti and
29 Schwartz, 2008, 195755). However, the evidence of adaptive effects has an implication for the
30 interpretation of multi-day effects, and requires further analysis. Analyses that specifically focused
31 on the O3-mortality C-R relationship found no evidence of a threshold, but did observe evidence for
32 potential differences in the C-R relationship across cities (Bell et al., 2006, 087680; Katsouyanni et
33 al., 2009, 199899: Stylianou and Nicolich, 2009, 620299). Collectively, these studies support the
34 conclusions of the 2006 O3 AQCD that "if a population threshold level exists in O3 health effects, it
35 is likely near the lower limit of ambient O3 concentrations in the U.S."
36 In conclusion, the new epidemiologic studies build upon and confirm the associations reported
37 in the 2006 O3 AQCD resulting in a body of evidence that is sufficient to conclude that there is
38 likely to be a causal relationship between short-term O3 exposure and all-cause mortality.
March 2011 6-175 DRAFT - DO NOT CITE OR QUOTE
-------
6.7. Overall Summary
1 The evidence reviewed in this chapter describes the recent findings regarding the health effects
2 of short-term exposure to ambient O3 concentrations. Table 6-45 provides an overview of the causal
3 determinations for each of the health categories evaluated.
Table 645. Summary of causal determinations for short-term exposures to ozone
Health Category Causal Determination
Respiratory Effects Causal relationship
Cardiovascular Effects Suggestive of a causal relationship
Central Nervous System Effects Suggestive of a causal relationship
Effects on Liver and Xenobiotic Metabolism Inadequate to infer a causal relationship
Effects on Cutaneous and Ocular Tissues Inadequate to infer a causal relationship
Mortality Likely to be a causal relationship
March 2011 6-176 DRAFT - DO NOT CITE OR QUOTE
-------
References
A list of all epidemiologic references considered for inclusion in this chapter can be found at
http://hero.epa. gov/index.cfm?action=search.do&submit=Search&group_id=403
A list of all toxicological references considered for inclusion in this chapter can be found at
http://hero.epa. gov/index.cfm?action=search.do&submit=Search&group_id=401
A list of all controlled human exposure references considered for inclusion in this chapter can be found at
http://hero.epa. gov/index.cfm?action=search.do&submit=Search&group_id=477
Adamkiewicz, G; Ebelt, S.; Syring, M.; Slater, I; Speizer, F. E.; Schwartz, I; Suh, H.; Gold, D. R. (2004). Association
between air pollution exposure and exhaled nitric oxide in an elderly population. Thorax, 59: 204-209.
http://dx.doi.org/10.1136/thorax.2003.006445087925
Adams, W. C. (1998). Dose-response effect of varied equivalent minute ventilation rates on pulmonary function responses
during exposure to ozone. Washington, DC: American Petroleum Institute. 670457
Adams, W. C. (2002). Comparison of chamber and face-mask 6.6-hour exposures to ozone on pulmonary function and
symptoms responses. Inhal Toxicol, 14: 745-764. 093690
Adams, W. C. (2003). Comparison of chamber and face mask 6.6-hour exposure to 0.08 ppm ozone via square-wave and
triangular profiles on pulmonary responses. Inhal Toxicol, 15: 265-281. 042245
Adams, W. C. (2006). Comparison of chamber 6.6-h exposures to 0.04-0.08 PPM ozone via square-wave and triangular
profiles on pulmonary responses. Inhal Toxicol, 18: 127-136. http://dx.doi.org/10.1080/08958370500306107
087681
Adams, W. C. (2006). Human pulmonary responses with 30-minute time intervals of exercise and rest when exposed for 8
hours to 0.12 ppm ozone via square-wave and acute triangular profiles. Inhal Toxicol, 18: 413-22.
http://dx.doi.org/10.1080/08958370600563599 196494
Adams, W. C.; Schelegle, E. S. (1983). Ozone and high ventilation effects on pulmonary function and endurance
performance. J Appl Physiol, 55: 805-812. 039822
Aibo, D. I.; Birmingham, N. P.; Lewandowski, R.; Maddox, J. F.; Roth, R. A.; Ganey, P. E.; Wagner, J. G; Harkema, J. R.
(2010). Acute Exposure to Ozone Exacerbates Acetaminophen-Induced Liver Injury in Mice. Toxicol Sci, 115: 267-
285. http://dx.doi.org/10.1093/toxsci/kfq034 378559
Alexeeff, S. E.; Litonjua, A. A.; Suh, H.; Sparrow, D.; Vokonas, P. S.; Schwartz, J. (2007). Ozone exposure and lung
function: effect modified by obesity and airways hyperresponsiveness in the VA normative aging study. Chest, 132:
1890-1897. http://dx.doi.org/10.1378/chest.07-1126 195862
Alexeeff, S. E.; Litonjua, A. A.; Wright, R. O.; Baccarelli, A.; Suh, H.; Sparrow, D.; Vokonas, P. S.; Schwartz, J. (2008).
Ozone exposure, antioxidant genes, and lung function in an elderly cohort: VA normative aging study. Occup
Environ Med, 65: 736-742. http://dx.doi.org/10.1136/oem.2007.035253 195864
Alexis, N. E.; Lay, J. C.; Hazucha, M.; Harris, B.; Hernandez, M. L.; Bromberg, P. A.; Kehrl, H.; Diaz-Sanchez, D.; Kim,
C.; Devlin, R. B.; Peden, D. B. (2010). Low-level ozone exposure induces airways inflammation and modifies cell
surface phenotypes in healthy humans. Inhal Toxicol, 22: 593-600. http://dx.doi.org/10.3109/08958371003596587
628538
Alexis, N. E.; Zhou, H.; Lay, J. C.; Harris, B.; Hernandez, M. L.; Lu, T S.; Bromberg, P. A.; Diaz-Sanchez, D.; Devlin, R.
B.; Kleeberger, S. R.; Peden, D. B. (2009). The glutathione-S-transferase Mu 1 null genotype modulates ozone-
induced airway inflammation in human subjects. J Allergy Clin Immunol, 124: 1222-1228.
http://dx.doi.0rg/10.1016/i.iaci.2009.07.036628542
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments such as the Integrated Science Assessments (ISAs) and the Integrated Risk Information System (IRIS).
March 2011 6-177 DRAFT - DO NOT CITE OR QUOTE
-------
Alexis, N.; Urch, B.; Tarlo, S.; Corey, P.; Pengelly, D.; O'Byrne, P.; Silverman, F. (2000). Cyclooxygenase metabolites play
a different role in ozone-induced pulmonary function decline in asthmatics compared to normals. Inhal Toxicol, 12:
1205-1224. 013072
Alfaro-Rodriguez, A.; Gonzalez-Pina, R. (2005). Ozone-induced paradoxical sleep decrease is related to diminished
acetylcholine levels in the medial preoptic area in rats. Chem Biol Interact, 151: 151-158.
http://dx.doi.0rg/10.1016/j.cbi.2004.10.001 596365
American Thoracic Society. (2000). Guidelines for methacholine and exercise challenge testing-1999. Am J Respir Crit
CareMed, 161: 309-329. 090799
American Thoracic Society (1991). Lung function testing: selection of reference values and interpretative strategies. Am J
Respir Crit Care Med, 144: 1202-1218. 044889
Anderson, H. R.; Armstrong, B.; Hajat, S.; Harrison, R.; Monk, V; Poloniecki, J.; Timmis, A.; Wilkinson, P. (2010). Air
pollution and activation of implantable cardioverter defibrillators in London. Epidemiology, 21: 405-413.
http://dx.doi.org/10.1097/EDE.Ob013e3181d61600625028
Angoa-Perez, M.; Jiang, H.; Rodriguez, A. I.; Lemini, C.; Levine, R. A.; Rivas-Arancibia, S. (2006). Estrogen counteracts
ozone-induced oxidative stress and nigral neuronal death. Neuroreport, 17: 629-633.
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids= 16603924.
596366
Apte, M. G; Buchanan, I. S.; Mendell, M. J. (2008). Outdoor ozone and building-related symptoms in the BASE study.
Indoor Air, 18: 156-170. http://dx.doi.Org/10.llll/i.1600-0668.2008.00521.xl95865
Araneda, S.; Commin, L.; Atlagich, M.; Kitahama, K.; Parraguez, V. H.; Pequignot, J. M.; Dalmaz, Y. (2008). VEGF
overexpression in the astroglial cells of rat brainstem following ozone exposure. Neurotoxicology, 29: 920-927.
http://dx.doi.0rg/10.1016/i.neuro.2008.09.006596367
Aranyi, C.; Vana, S. C.; Thomas, P. T; Bradof, J. N.; Fenters, J. D.; Graham, J. A.; Miller, F. J. (1983). Effects of
subchronic exposure to a mixture of O3, SO2, and (NH4)2SO4 on host defenses of mice. J Toxicol Environ Health,
12: 55-71.040512
Arbex, A. M.; de Souza Conceicao, G. M.; Perez Cendon, S.; Arbex, F. F.; Lopes, A. C.; Providello Moyses, E.; Santiago,
S. L.; Saldiva, P. H. N.; Pereira, L. A. A.; Ferreira Braga, A. L. (2009). Urban air pollution and COPD-related
emergency room visits. J Epidemiol Community Health, 966: 777-783. http://dx.doi.org/10.1136/jech.2008.078360
184334
Aris, R. M.; Tager, I.; Christian, D.; Kelly, T; Balmes, J. R. (1995). Methacholine responsiveness is not associated with
OS-induced decreases in FEV1. Chest, 107: 621-628. 075945
Arito, H.; Takahashi, M.; Iwasaki, T; Uchiyama, I. (1997). Age-related changes in ventilatory and heart rate responses to
acute ozone exposure in the conscious rat. Ind Health, 35: 78-86. 082671
Arito, H.; Uchiyama, L; Arakawa, H.; Yokoyama, E. (1990). Ozone-induced bradycardia and arrhythmia and their relation
to sleep-wakefulness in rats. Toxicol Lett, 52: 169-178. http://dx.doi.org/10.1016/0378-4274(90)90151-6 042285
Arito, H.; Uchiyama, L; Yokoyama, E. (1992). Acute effects of ozone on EEG activity, sleep-wakefulness and heart rate in
rats. Ind Health, 30: 23-34. 042759
Armstrong, B. G. (2003). Fixed factors that modify the effects of time-varying factors: Applying the case-only approach.
Epidemiology, 14: 467-472. 153211
Atkinson, R. W.; Bremner, S. A.; Anderson, H. R.; Strachan, D. P.; Bland, J. M.; Ponce de Leon, A. (1999). Short-term
associations between emergency hospital admissions for respiratory and cardiovascular disease and outdoor air
pollution in London. Arch Environ Occup Health, 54: 398-411. 007882
Avissar, N. E.; Reed, C. K.; Cox, C.; Frampton, M. W.; Finkelstein, J. N. (2000). Ozone, but not nitrogen dioxide, exposure
decreases glutathione peroxidases in epithelial lining fluid of human lung. Am J Respir Crit Care Med, 162: 1342-
1347. 012528
Avol, E. L.; Linn, W. S.; Venet, T. G; Shamoo, D. A.; Hackney, J. D. (1984). Comparative respiratory effects of ozone and
ambient oxidant pollution exposure during heavy exercise. J Air Waste Manag Assoc, 34: 804-809. 040221
Avol, E. L.; Navidi, W. C.; Rappaport, E. B.; Peters, J. M. (1998). Acute effects of ambient ozone on asthmatic, wheezy,
and healthy children (Report No. 82). Topsfield, MA: Health Effects Institute; Flagship Press. 086365
March 2011 6-178 DRAFT - DO NOT CITE OR QUOTE
-------
Avol, E. L.; Trim, S. C.; Little, D. E.; Spier, C. E.; Smith, M. N.; Peng, R. C.; Linn, W. S.; Hackney, J. D. (1991). Ozone
exposure and lung function: a southern California summer camp study. In Berglund, R. L.; Lawson, D. R.; McKee,
D. J. (Eds.), Tropospheric ozone and the environment: papers from an international conference; March 1990; Los
Angeles, CA (pp. 90-99). Pittsburgh, PA: Air & Waste Management Association. 042613
Avol, E. L.; Trim, S. C.; Little, D. E.; Spier, C. E.; Smith, M. N.; Peng, R.-C.; Linn, W. S.; Hackney, J. D.; Gross, K. B.;
DArcy, J. B.; Gibbons, D.; Higgins, I. T T (1990). Ozone exposure and lung function in children attending a
southern California summer camp. In Proceedings of the 83rd A&WMA Annual Meeting (pp. 90-150.3).
Pittsburgh, PA: Air & Waste Management Association. 042366
Baccarelli, A.; Zanobetti, A.; Martinelli, I.; Grillo, P.; Hou, L.; Lanzani, G; Mannucci, P. M.; Bertazzi, P. A.; Schwartz, J.
(2007). Air pollution, smoking, and plasma homocysteine. Environ Health Perspect, 115:176-181. 091310
Baccini, M.; Biggeri, A.; Accetta, G; Kosatsky, T; Katsouyanni, K.; Analitis, A.; Anderson, H. R.; Bisanti, L.; DTppoliti,
D.; Danova, J.; Forsberg, B.; Medina, S.; Paldy, A.; Rabczenko, D.; Schindler, C.; Michelozzi, P. (2008). Heat
effects on mortality in 15 European cities. Epidemiology, 19: 711-719.
http://dx.doi.org/10.1097/EDE.Ob013e318176bfcd633196
Baja, E. S.; Schwartz, J. D.; Wellenius, G. A.; Coull, B. A.; Zanobetti, A.; Vokonas, P. S.; Suh, H. H. (2010). Traffic-related
air pollution and QT interval: Modification by diabetes, obesity, and oxidative stress gene polymorphisms in the
Normative Aging Study. Environ Health Perspect, 118: 840-846. http://dx.doi.org/10.1289/ehp.0901396 626540
Balbi, B.; Pignatti, P.; Corradi, M.; Baiardi, P.; Bianchi, L.; Brunetti, G; Radaeli, A.; Moscato, G; Mutti, A.; Spanevello,
A.; Malerba, M. (2007). Bronchoalveolar lavage, sputum and exhaled clinically relevant inflammatory markers:
values in healthy adults. Eur Respir J, 30: 769-781. http://dx.doi.org/10.1183/09031936.00112306 625784
Ballester, F.; Rodriguez, P.; Iniguez, C.; Saez, M.; Daponte, A.; Galan, L; Taracido, M.; Arribas, F.; Bellido, J.; Cirarda, F.
B.; Canada, A.; Guillen, J. J.; Guillen-Grima, F.; Lopez, E.; Perez-Hoyos, S.; Lertxundi, A.; Toro, S. (2006). Air
pollution and cardiovascular admisisons association in Spain: results within the EMECAS project. J Epidemiol
Community Health, 60: 328-336. 088746
Ballester, F.; Saez, M.; Daponte, A.; Ordonez, J. M.; Taracido, M.; Cambra, K.; Arribas, F.; Bellido, J. B.; Guillen, J. J.;
Aguinaga, L; Canada, A.; Lopez, E.; Iniguez, C.; Rodriguez, P.; Perez-Hoyos, S.; Barcelo, M. A.; Ocana, R.;
Aranguez, E. (2005). The EMECAS Project: Spanish multicentre study on short-term health effects of air pollution.
Rev Esp Salud Publica, 79: 229-242.
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids= 15913057.
600865
Balmes, J. R.; Aris, R. M.; Chen, L. L.; Scannell, C.; Tager, I. B.; Finkbeiner, W.; Christian, D.; Kelly, T; Hearne, P. Q.;
Ferrando, R.; Welch, B. (1997). Effects of ozone on normal and potentially sensitive human subjects part I: Airway
inflammation and responsiveness to ozone in normal and asthmatic subjects. Boston, MA: Health Effects Institute.
086092
Balmes, J. R.; Chen, L. L.; Scannell, C.; Tager, I.; Christian, D.; Hearne, P. Q.; Kelly, T; Aris, R. M. (1996). Ozone-
induced decrements in FEV1 and FVC do not correlate with measures of inflammation. Am J Respir Crit Care
Med, 153: 904-909. 080830
Baraldi, E.; Carraro, S.; Alinovi, R.; Pesci, A.; Ghiro, L.; Bodini, A.; Piacentini, G; Zacchello, F.; Zanconato, S. (2003).
Cysteinyl leukotrienes and 8-isoprostane in exhaled breath condensate of children with asthma exacerbations.
Thorax, 58: 505-509. http://dx.doi.Org/10.1136/thorax.58.6.505 625802
Barnes, P. J.; Liew, F. Y. (1995). Nitric oxide and asthmatic inflammation. Immunol Today, 16: 128-130.
http://dx.doi.org/10.1016/0167-5699(95)80128-6 083814
Barnett, A. G; Williams, G. M.; Schwartz, J.; Best, T. L.; Neller, A. H.; Petroeschevsky, A. L.; Simpson, R. W. (2006). The
effects of air pollution on hospitalizations for cardiovascular disease in elderly people in Australian and New
Zealand cities. Environ Health Perspect, 114: 1018-1023. 089770
Barraza-Villarreal, A.; Sunyer, J.; Hernandez-Cadena, L.; Escamilla-Nunez, M. C.; Sienra-Monge, J. J.; Ramirez-Aguilar,
M.; Cortez-Lugo, M.; Holguin, F.; Diaz-Sanchez, D.; Olin, A. C.; Romieu, I. (2008). Air pollution, airway
inflammation, and lung function in a cohort study of Mexico City schoolchildren. Environ Health Perspect, 116:
832-838. http://dx.doi.org/10.1289/ehp.10926 156254
March 2011 6-179 DRAFT - DO NOT CITE OR QUOTE
-------
Basha, M. A.; Gross, K. B.; Gwizdala, C. I; Haidar, A. H.; Popovich, J. Jr (1994). Bronchoalveolar lavage neutrophilia in
asthmatic and healthy volunteers after controlled exposure to ozone and filtered purified air. Chest, 106: 1757-1765.
075950
Bell, M. L.; Dominici, F. (2008). Effect modification by community characteristics on the short-term effects of ozone
exposure and mortality in 98 US communities. Am J Epidemiol, 167: 986-997.
http://dx.doi.org/10.1093/aje/kwm396 193828
Bell, M. L.; Dominici, F.; Samet, J. M. (2005). A meta-analysis of time-series studies of ozone and mortality with
comparison to the national morbidity, mortality, and air pollution study. Epidemiology, 16: 436-445. 074345
Bell, M. L.; Kim, J. Y; Dominici, F. (2007). Potential confounding of particulate matter on the short-term association
between ozone and mortality in multisite time-series studies. Environ Health Perspect, 115: 1591-1595.
http://dx.doi.org/10.1289/ehp.10108093256
Bell, M. L.; Levy, J. K.; Lin, Z. (2008). The effect of sandstorms and air pollution on cause-specific hospital admissions in
Taipei, Taiwan. Occup Environ Med, 65: 104-111. http://dx.doi.org/10.1136/oem.2006.031500 091268
Bell, M. L.; McDermott, A.; Zeger, S. L.; Samet, J. M.; Dominici, F. (2004). Ozone and short-term mortality in 95 US
urban communities, 1987-2000. JAMA, 292: 2372-2378. http://dx.doi.org/10.1001/jama.292.19.2372 094417
Bell, M. L.; Peng, R. D.; Dominici, F. (2006). The exposure-response curve for ozone and risk of mortality and the
adequacy of current ozone regulations. Environ Health Perspect, 114: 532-536. 087680
Bennett, W. D.; Hazucha, M. J.; Folinsbee, L. J.; Bromberg, P. A.; Kissling, G. E.; London, S. J. (2007). Acute pulmonary
function response to ozone in young adults as a function of body mass index. Inhal Toxicol, 19:1147-1154.
http://dx.doi.org/10.1080/08958370701665475418827
Bergamaschi, E.; De Palma, G; Mozzoni, P.; Vanni, S.; Vettori, M. V; Broeckaert, F.; Bernard, A.; Mutti, A. (2001).
Polymorphism of quinone-metabolizing enzymes and susceptibility to ozone-induced acute effects. Am J Respir
Crit Care Med, 163: 1426-1431. 052670
Berkey, C. S.; Hoaglin, D. C.; Antczak-Bouckoms, A.; Mosteller, F.; Colditz, G. A. (1998). Meta-analysis of multiple
outcomes by regression with random effects. Stat Med, 17: 2537-2550. 684190
Berry, M.; Lioy, P. J.; Gelperin, K.; Buckler, G; Klotz, J. (1991). Accumulated exposure to ozone and measurement of
health effects in children and counselors at two summer camps. Environ Res, 54: 135-150. 042377
Biggeri, A.; Baccini, M.; Bellini, P.; Terracini, B. (2005). Meta-analysis of the Italian studies of short-term effects of air
pollution (MISA), 1990-1999. Int J Occup Environ Health, 11: 107-122. 087395
Bloch, W. N. Jr; Miller, F. J.; Lewis, T R. (1971). Pulmonary hypertension in dogs exposed to ozone. Cincinnati, OH: EPA.
015914
Blomberg, A.; Mudway, I. S.; Nordenhall, C.; Hedenstrom, H.; Kelly, F. J.; Frew, A. J.; Holgate, S. T; Sandstrom, T.
(1999). Ozone-induced lung function decrements do not correlate with early airway inflammatory or antioxidant
responses. Eur Respir J, 13: 1418-1428. 001267
Bosson, J.; Stenfors, N.; Bucht, A.; Helleday, R.; Pourazar, J.; Holgate, S. T; Kelly, F. J.; Sandstrom, T; Wilson, S.; Frew,
A. J.; Blomberg, A. (2003). Ozone-induced bronchial epithelial cytokine expression differs between healthy and
asthmatic subjects. Clin Exp Allergy, 33: 777-782. 051687
Boussouar, A.; Araneda, S.; Hamelin, C.; Soulage, C.; Kitahama, K.; Dalmaz, Y. (2009). Prenatal ozone exposure abolishes
stress activation of Fos and tyrosine hydroxylase in the nucleus tractus solitarius of adult rat. Neurosci Lett, 452:
75-78.
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids= 19444957.
596368
Brauer, M.; Blair, J.; Vedal, S. (1996). Effect of ambient ozone exposure on lung function in farm workers. Am J Respir
Crit Care Med, 154: 981-987. 080754
Braun-Fahrlander Ch; Kunzli, N.; Domenighetti, G; Carell, C. F.; Ackermann-Liebrich, U. (1994). Acute effects of ambient
ozone on respiratory function of Swiss schoolchildren after a 10-minute heavy exercise. Pediatr Pulmonol, 17: 169-
177. http://dx.doi.org/10.1002/ppul.1950170306 038665
March 2011 6-180 DRAFT - DO NOT CITE OR QUOTE
-------
Brown, J. S.; Bateson, T. F.; McDonnell, W. R (2008). Effects of exposure to 0.06 ppm ozone on FEV1 in humans: a
secondary analysis of existing data. Environ Health Perspect, 116: 1023-1026. http://dx.doi.org/10.1289/ehp.11396
195140
Brunekreef, B.; Hoek, G; Breugelmans, O.; Leentvaar, M. (1994). Respiratory effects of low-level photochemical air
pollution in amateur cyclists. Am J Respir Crit Care Med, 150: 962-966. 045161
Buadong, D.; Jinsart, W.; Funatagawa, I.; Karita, K.; Yano, E. (2009). Association between PM10 and O-3 levels and
hospital visits for cardiovascular diseases in Bangkok, Thailand. J Epidemiol, 19: 182-188.
http://dx.doi.org/10.2188/iea.JE20080047602060
Burleson, G. R.; Keyes, L. L.; Stutzman, J. D. (1989). Immunosuppression of pulmonary natural killer activity by exposure
to ozone. Immunopharmacol Immunotoxicol, 11: 715-735. 042214
Burnett, R.; Raizenne, M.; Krewski, D. (1990). Acute health effects of transported air pollution: A study of children
attending a residential summer camp. Can J Stat, 18: 367-373. http://dx.doi.org/10.2307/3315843 670386
Burra, T. A.; Moineddin, R.; Agha, M. M.; Glazier, R. H. (2009). Social disadvantage, air pollution, and asthma physician
visits in Toronto, Canada. Environ Res, 109: 567-574. http://dx.doi.Org/10.1016/j.envres.2009.03.004 195868
Bush, M. L.; Asplund, P. T; Miles, K. A.; Ben-Jebria, A.; Ultman, J. S. (1996). Longitudinal distribution of O3 absorption
in the lung: gender differences and intersubject variability. J Appl Physiol, 81: 1651-1657. 080763
Cakmak, S.; Dales, R. E.; Judek, S. (2006). Do gender, education, and income modify the effect of air pollution gases on
cardiac disease? J Occup Environ Med, 48: 89-94. http://dx.doi.org/10.1097/01.jom.0000184878.11956.4b 099068
Cakmak, S.; Dales, R. E.; Judek, S. (2006). Respiratory health effects of air pollution gases: Modification by education and
income. Arch Environ Occup Health, 61: 5-10. 093272
Calderon Guzman, D.; Barragan Mejia, G; Hernandez Garcia, E.; Juarez Olguin, H. (2006). Effect of nutritional status and
ozone exposure on some biomarkers of oxidative stress in rat brain regions. Nutr Cancer, 55: 195-200.
http://dx.doi.org/10.1207/sl5327914nc5502_ll 596371
Carey, S. A.; Minard, K. R.; Trease, L. L.; Wagner, J. G; Garcia, G. J.; Ballinger, C. A.; Kimbell, J. S.; Plopper, C. G;
Corley, R. A.; Postlethwait, E. M.; Harkema, J. R.; Einstein, D. R. (2007). Three-dimensional mapping of ozone-
induced injury in the nasal airways of monkeys using magnetic resonance imaging and morphometric techniques.
Toxicol Pathol, 35: 27-40. http://dx.doi.org/10.1080/01926230601072343 195752
Carpagnano, G. E.; Foschino Barbara, M. P.; Cagnazzo, M.; Di Gioia, G; Giliberti, T; Di Matteo, C.; Resta, O. (2005). Use
of exhaled breath condensate in the study of airway inflammation after hypertonic saline solution challenge. Chest,
128: 3159-3166. http://dx.doi.Org/10.1378/chest.128.5.3159625789
Castagna, R.; Davis, P. A.; Vasu, V. T; Soucek, K.; Cross, C. E.; Greci, L.; Valacchi, G. (2009). Nitroxide radical TEMPO
reduces ozone-induced chemokine IL-8 production in lung epithelial cells. Toxicol In Vitro, 23: 365-370.
http://dx.doi.0rg/10.1016/i.tiv.2008.12.016596372
Castillejos, M.; Gold, D. R.; Damokosh, A. L; Serrano, P.; Allen, G; McDonnell, W. F.; Dockery, D.; Velasco, S. R.;
Hernandez, M.; Hayes, C. (1995). Acute effects of ozone on the pulmonary function of exercising schoolchildren
from Mexico City. Am J Respir Crit Care Med, 152: 1501-1507. 078485
Chan, C.-C.; Chuang, K.-J.; Chien, L.-C.; Chen, W.-J.; Chang, W.-T (2006). Urban air pollution and emergency admissions
for cerebrovascular diseases in Taipei, Taiwan. Eur Heart J, 27: 1238-1244. 090193
Chan, C.-C.; Chuang, K.-J.; Su, T.-C.; Lin, L.-Y. (2005). Association between nitrogen dioxide and heart rate variability in
a susceptible population. Eur J Cardiovasc Prev Rehabil, 12: 580-586. 088988
Chang MM-J; Wu, R.; Plopper, C. G; Hyde, D. M. (1998). IL-8 is one of the major chemokines produced by monkey
airway epithelium after ozone-induced injury. Am J Physiol, 275: L524-L532. 011983
Chang, C.-C.; Tsai, S.-S.; Ho, S.-C.; Yang, C.-Y (2005). Air pollution and hospital admissions for cardiovascular disease in
Taipei, Taiwan. Environ Res, 98: 114-119. 080086
Chen, C.; Arjomandi, M.; Balmes, J.; Tager, I.; Holland N (2007). Effects of Chronic and Acute Ozone Exposure on Lipid
Peroxidation and Antioxidant Capacity in Healthy Young Adults. Environ Health Perspect, 115: 1732-1737.
http://dx.doi.org/10.1289/ehp.10294 145956
March 2011 6-181 DRAFT - DO NOT CITE OR QUOTE
-------
Chen, J.-C.; Schwartz, J. (2009). Neurobehavioral effects of ambient air pollution on cognitive performance in US adults.
Neurotoxicology, 30: 231-239. http://dx.doi.Org/10.1016/j.neuro.2008.12.011 179945
Chen, L.; Jennison, B. L.; Yang, W.; Omaye, S. T. (2000). Elementary school absenteeism and air pollution. Inhal Toxicol,
12: 997-1016.011931
Chen, P.-C.; Lai, Y.-M.; Chan, C.-C.; Hwang, J.-S.; Yang, C.-Y; Wang J.-D. (1999). Short-term effect of ozone on the
pulmonary function of children in primary school. Environ Health Perspect, 107: 921-925.
http://dx.doi.org/10.1289/ehp.99107921 011149
Chhabra, S. K.; Yasir, A.; Chaudhry, K.; Shah, B. (2010). Effect of ozone on response to ovalbumin & its modulation by
vitamins C & E in sensitized guinea pigs. Indian J Med Res, 132: 87-93. 677665
Chimenti, L.; Morici, G; Paterno, A.; Bonanno, A.; Vultaggio, M.; Bellia, V; Bonsignore, M. R. (2009). Environmental
conditions, air pollutants, and airway cells in runners: A longitudinal field study. J Sports Sci, 27: 925-935.
http://dx.doi.org/10.1080/02640410902946493418828
Chiu, H. F.; Cheng, M. H.; Yang, C. Y (2009). Air pollution and hospital admissions for pneumonia in a subtropical city:
Taipei, Taiwan. Inhal Toxicol, 21: 32-37. 190249
Cho, H. Y; Morgan, D. L.; Bauer, A. K.; Kleeberger, S. R. (2007). Signal transduction pathways of turn or necrosis factor-
mediated lung injury induced by ozone in mice. Am J Respir Crit Care Med, 175: 829-839.
http://dx.doi.org/10.1164/rccm.200509-1527OC596373
Choi, J. H.; Xu, Q. S.; Park, S. Y; Kim, J. H.; Hwang, S. S.; Lee, K. H.; Lee, H. J.; Hong, Y. C. (2007). Seasonal variation
of effect of air pollution on blood pressure. J Epidemiol Community Health, 61: 314-318. 093196
Christian, D. L.; Chen, L. L.; Scannell, C. H.; Ferrando, R. E.; Welch, B. S.; Balmes, J. R. (1998). Ozone-induced
inflammation is attenuated with multiday exposure. Am J Respir Crit Care Med, 158: 532-537. 029925
Chuang, G. C.; Yang, Z.; Westbrook, D. G; Pompilius, M.; Ballinger, C. A.; White, R. C.; Krzywanski, D. M.; Postlethwait,
E. M.; Ballinger, S. W. (2009). Pulmonary ozone exposure induces vascular dysfunction, mitochondria! damage,
and atherogenesis. Am J Physiol Lung Cell Mol Physiol, 297: L209-L216.
http://dx.doi.org/10.1152/aiplung.00102.2009 197202
Chuang, K. J.; Chan, C. C.; Su, T. C.; Lin, L. Y; Lee, C. T. (2007). Associations between particulate sulfate and organic
carbon exposures and heart rate variability in patients with or at risk for cardiovascular diseases. J Occup Environ
Med, 49: 610-617. 098629
Chuang, K.-L; Chan, C.-C.; Su, T.-C.; Lee, C.-T; Tang, C.-S. (2007). The effect of urban air pollution on inflammation,
oxidative stress, coagulation, and autonomic dysfunction in young adults. Am J Respir Crit Care Med, 176: 370-
376. 091063
Chuang, K.-J.; Yan, Y.-H.; Cheng, T.-J. (2010). Effect of air pollution on blood pressure, blood lipids, and blood sugar: a
population-based approach. J Occup Environ Med, 52: 258-262. http://dx.doi.org/10.1097/JOM.Ob013e3181ceff7a
379993
Cockcroft, D. W.; Davis, B. E.; Todd, D. C.; Smycniuk, A. J. (2005). Methacholine challenge: Comparison of two methods.
Chest, 127: 839-844. 090805
Cole, M. P.; Freeman, B. A. (2009). Promotion of cardiovascular disease by exposure to the air pollutant ozone. Am J
Physiol Lung Cell Mol Physiol, 297: L209-L216. 597507
Colin-Barenque, L.; Dorado-Martinez, C.; Rivas-Arancibia, S.; Avila-Costa, M. R.; Fortoul, T. I. (2005). Morphological
recovery of the granule cells from the olfactory bulb after the cessation of acute ozone exposure. Int J Neurosci,
115: 411-421. http://dx.doi.org/10.1080/00207450590521028 180458
Corradi, M.; Folesani, G; Andreoli, R.; Manini, P.; Bodini, A.; Piacentini, G; Carraro, S.; Zanconato, S.; Baraldi, E. (2003).
Aldehydes and glutathione in exhaled breath condensate of children with asthma exacerbation. Am J Respir Crit
Care Med, 167: 395-399. http://dx.doi.org/10.1164/rccm.200206-507OC 625790
Cremillieux, Y; Servais, S.; Berthezene, Y; Dupuich, D.; Boussouar, A.; Stupar, V.; Pequignot, J. M. (2008). Effects of
ozone exposure in rat lungs investigated with hyperpolarized 3He MRI. J Magn Reson Imaging, 27: 771-776.
http://dx.doi.org/10.1002/imri.21216 180454
Curriero, F. C.; Heiner, K. S.; Samet, J. M.; Zeger, S. L.; Strug, L.; Patz, J. A. (2002). Temperature and mortality in 11
cities of the eastern United States. Am J Epidemiol, 155: 80-87. 055878
March 2011 6-182 DRAFT - DO NOT CITE OR QUOTE
-------
Dahl, M.; Bauer, A. K.; Arredouani, M.; Soininen, R.; Tryggvason, K.; Kleeberger, S. R.; Kobzik, L. (2007). Protection
against inhaled oxidants through scavenging of oxidized lipids by macrophage receptors MARCO and SR-AI/II. J
Clin Invest, 117: 757-764. http://dx.doi.org/10.1172/JCI29968 196986
Dales, R. E.; Cakmak, S.; Doiron, M. S. (2006). Gaseous air pollutants and hospitalization for respiratory disease in the
neonatal period. Environ Health Perspect, 114: 1751-1754. http://dx.doi.org/10.1289/ehp.9044 090744
Dales, R.; Chen, L.; Frescura, A. M.; Liu, L.; Villeneuve, P. J. (2009). Acute effects of outdoor air pollution on forced
expiratory volume in 1 s: A panel study of schoolchildren with asthma. Eur Respir J, 34: 316-323.
http://dx.doi.org/10.1183/09031936.00138908594285
Damera, G; Jester William, R; Jiang, M.; Zhao, H.; Fogle Homer, W.; Mittelman, M.; Haczku, A.; Murphy, E.; Parikh, L;
Panettieri Reynold, A. (2010). Inhibition of myristoylated alanine-rich C kinase substrate (MARCKS) protein
inhibits ozone-induced airway neutrophilia and inflammation. Exp Lung Res, 36: 75-84.
http://dx.doi.org/10.3109/01902140903131200380255
Damera, G; Zhao, H.; Wang, M.; Smith, M.; Kirby, C.; Jester, W. R; Lawson, J. A.; Panettieri RA, Jr (2009). Ozone
modulates IL-6 secretion in human airway epithelial and smooth muscle cells. Am J Physiol Lung Cell Mol
Physiol, 296: L674-L683. http://dx.doi.org/10.1152/ajplung.90585.2008 596375
Darrow, L. A.; Klein, M.; Sarnat, J. A.; Mulholland, J. A.; Strickland, M. J.; Sarnat, S. E.; Russell, A. G; Tolbert, P. E.
(2011). The use of alternative pollutant metrics in time-series studies of ambient air pollution and respiratory
emergency department visits. J Expo Sci Environ Epidemiol, 21: 10-19. http://dx.doi.org/10.1038/jes.2009.49
202800
Delfino, R. J.; Gone, H.; Linn, W. S.; Pellizzari, E. D.; Hu, Y. (2003). Asthma symptoms in Hispanic children and daily
ambient exposures to toxic and criteria air pollutants. Environ Health Perspect, 111: 647-656.
http://dx.doi.org/10.1289/ehp.5992050460
Delfino, R. J.; Staimer, N.; Tjoa, T.; Arhami, M.; Polidori, A.; Gillen, D. L.; George, S. C.; Shafer, M. M.; Schauer, J. J.;
Sioutas, C. (2010). Associations of primary and secondary organic aerosols with airway and aystemic inflammation
in an elderly panel cohort. Epidemiology, 21: 892-902. http://dx.doi.org/10.1097/EDE.Ob013e3181f20e6c 647222
Delfino, R. J.; Tjoa, T; Gillen, D. L.; Staimer, N.; Polidori, A.; Arhami, M.; Jamner, L.; Sioutas, C.; Longhurst, J. (2010).
Traffic-related air pollution and blood pressure in elderly subjects with coronary artery disease. Epidemiology, 21:
396-404. http://dx.doi.org/10.1097/EDE.Ob013e3181d5el9b 625026
Delfino, R. J.; Zeiger, R. S.; Seltzer, J. M.; Street, D. H.; Matteucci, R. M.; Anderson, P. R.; Koutrakis, P. (1997). The effect
of outdoor fungal spore concentrations on daily asthma severity. Environ Health Perspect, 105: 622-635. 084531
Delfino, R. J.; Zeiger, R. S.; Seltzer, J. M.; Street, D. H.; McLaren, C. E. (2002). Association of asthma symptoms with
peak particulate air pollution and effect modification by anti-inflammatory medication use. Environ Health
Perspect, 110: A607-A617. 093740
DeLucia, A. J.; Adams, W. C. (1977). Effects of O3 inhalation during exercise on pulmonary function and blood
biochemistry. J Appl Physiol, 43: 75-81. 038281
Dennekamp, M.; Akram, M.; Abramson, M. J.; Tonkin, A.; Sim, M. R.; Fridman, M.; Erbas, B. (2010). Outdoor air
pollution as a trigger for out-of-hospital cardiac arrests. Epidemiology, 21: 494-500.
http://dx.doi.org/10.1097/EDE.Ob013e3181e093db626767
Depuydt, P.; Joos, G. F.; Pauwels, R. A. (1999). Ambient ozone concentrations induce airway hyperresponsiveness in some
rat strains. Eur Respir J, 14: 125-131. 011995
Devlin, R. B.; Folinsbee, L. J.; Biscardi, F.; Hatch, G; Becker, S.; Madden, M. C.; Robbins, M.; Koren, H. S. (1997).
Inflammation and cell damage induced by repeated exposure of humans to ozone. Inhal Toxicol, 9: 211-235.
083577
Devlin, R. B.; McDonnell, W. R; Becker, S.; Madden, M. C.; McGee, M. P.; Perez, R.; Hatch, G; House, D. E.; Koren, H.
S. (1996). Time-dependent changes of inflammatory mediators in the lungs of humans exposed to 0.4 ppm ozone
for 2 hr: a comparison of mediators found in bronchoalveolar lavage fluid 1 and 18 hr after exposure. Toxicol Appl
Pharmacol, 138: 176-185. 042840
Devlin, R. B.; McDonnell, W. R; Mann, R.; Becker, S.; House, D. E.; Schreinemachers, D.; Koren, H. S. (1991). Exposure
of humans to ambient levels of ozone for 6.6 hours causes cellular and biochemical changes in the lung. Am J
Respir Cell Mol Biol, 4: 72-81. 040359
March 2011 6-183 DRAFT - DO NOT CITE OR QUOTE
-------
De Pablo, R; Lopez, A.; Soriano, L. R.; Tomas, C.; Diego, L.; Gonzalez, M.; Barrueco, M. (2006). Relationships of daily
mortality and hospital admissions to air pollution in Castilla-Leon, Spain. Atmosfera, 19: 23-39.
http://www.ejournal.unam.mx/atm/voll9-l/ATM19103.pdf. 196506
Diaz, G; Barrantes, O.; Honrubia, M.; Gracia, C. (1996). Effect of ozone and sulphur dioxide on mycorrhizae of Pinus
halepensis Miller. Annals of Forest Science, 53: 849-856. 026481
Dimeo, M. I; Glenn, M. G; Holtzman, M. I; Sheller, J. R.; Nadel, J. A.; Boushey, H. A. (1981). Threshold concentration
of ozone causing an increase in bronchial reactivity in humans and adaptation with repeated exposures. Am Rev
Respir Dis, 124: 245-248. 039662
Dockery, D. W.; Luttmann-Gibson, H.; Rich, D. Q.; Link, M. S.; Mittleman, M. A.; Gold, D. R.; Koutrakis, P.; Schwartz, J.
D.; Verrier, R. L. (2005). Association of air pollution with increased incidence of ventricular tachyarrhythmias
recorded by implanted cardioverter defibrillators. Environ Health Perspect, 113: 670-674. 078995
Dohm, M. R.; Mautz, W. J.; Andrade, J. A.; Gellert, K. S.; Salas-Ferguson, L. J.; Nicolaisen, N.; Fujie, N. (2005). Effects of
ozone exposure on nonspecific phagocytic capacity of pulmonary macrophages from an amphibian, Bufo marinus.
Environ Toxicol Chem, 24: 205-210. 180452
Driscoll, K. E.; Vollmuth, T A.; Schlesinger, R. B. (1987). Acute and subchronic ozone inhalation in the rabbit: response of
alveolar macrophages. J Toxicol Environ Health, 21: 27-43. 040803
Duramad, P.; Tager, I. B.; Holland, N. T. (2007). Cytokines and other immunological biomarkers in children's
environmental health studies. Toxicol Lett, 172: 48-59. 625792
Eiswerth, M. E.; Douglass Shaw, W.; Yen, S. T. (2005). Impacts of ozone on the activities of asthmatics: Revisiting the
data. J Environ Manage, 77: 56-63. http://dx.doi.Org/10.1016/i.ienvman.2005.02.010 196443
Escalante-Membrillo, C.; Gonzalez-Maciel, A.; Reynoso-Robles, R.; Gonzalez-Pina, R. (2005). Brain thiobarbituric acid-
reactive substances in rats after short periods of ozone exposure. Environ Res, 99: 68-71.
http://dx.doi.0rg/10.1016/i.envres.2005.02.006596378
Escamilla-Nunez, M. C.; Barraza-Villarreal, A.; Hernandez-Cadena, L.; Moreno-Macias, H.; Ramirez-Aguilar, M.; Sienra-
Monge, J. J.; Cortez-Lugo, M.; Texcalac, J. L.; del Rio-Navarro, B.; Romieu, I. (2008). Traffic-related air pollution
and respiratory symptoms among asthmatic children, resident in Mexico City: The EVA cohort study. Respir Res, 9:
74. http://dx.doi.org/10.1186/1465-9921-9-74 594284
Fakhri, A. A.; Ilic, L. M.; Wellenius, G. A.; Urch, B.; Silverman, R; Gold, D. R.; Mittleman, M. A. (2009). Autonomic
effects of controlled fine particulate exposure in young healthy adults: Effect modification by ozone. Environ
Health Perspect, 117: 1287-1292. http://dx.doi.org/10.1289/ehp.0900541 191914
Fakhrzadeh, L.; Laskin, J. D.; Laskin, D. L. (2008). Regulation of caveolin-1 expression, nitric oxide production and tissue
injury by tumor necrosis factor-alpha following ozone inhalation. Toxicol Appl Pharmacol, 227: 380-389.
http://dx.doi.0rg/10.1016/i.taap.2007.ll.012596380
Farraj, A. K.; Boykin, E.; Ledbetter, A.; Andrews, D.; Gavett, S. H. (2010). Increased lung resistance after diesel particulate
and ozone co-exposure not associated with enhanced lung inflammation in allergic mice. Inhal Toxicol, 22: 33-41.
http://dx.doi.org/10.3109/08958370902862434380846
Feng, R.; He, W.; Ochi, H.; Castranova, V. (2006). Ozone exposure impairs antigen-specific immunity but activates IL-7-
induced proliferation of CD4-CD8-thymocytes in BALB/c mice. J Toxicol Environ Health A, 69: 1511-1526.
http://dx.doi.org/10.1080/15287390500468696596381
Feo Brito, R; Mur Gimeno, P.; Martinez, C.; Tobias, A.; Suarez, L.; Guerra, R; Borja, J. M.; Alonso, A. M. (2007). Air
pollution and seasonal asthma during the pollen season: A cohort study in Puertollano and Ciudad Real (Spain).
Allergy, 62: 1152-1157. 093259
Ferdinands, J. M.; Crawford, C. A.; Greenwald, R.; Van Sickle, D.; Hunter, E.; Teague, W. G. (2008). Breath acidification
in adolescent runners exposed to atmospheric pollution: A prospective, repeated measures observational study.
Environ Health, 7: 11. http://dx.doi.org/10.1186/1476-069X-7-10 156433
Folinsbee, L. J.; Bedi, J. R; Horvath, S. M. (1980). Respiratory responses in humans repeatedly exposed to low
concentrations of ozone. Am Rev Respir Dis, 121: 431-439. 038880
Folinsbee, L. J.; Bedi, J. R; Horvath, S. M. (1984). Pulmonary function changes after 1 h continuous heavy exercise in 0.21
ppm ozone. J Appl Physiol, 57: 984-988. 040065
March 2011 6-184 DRAFT - DO NOT CITE OR QUOTE
-------
Folinsbee, L. I; Devlin, R. B.; Abdul-Salaam, S.; Koren, H. S. (1993). Repeated severe ozone exposure causes depressed
baseline spirometry. Am Rev Respir Dis, 147: A638. 043781
Folinsbee, L. I; Devlin, R. B.; Robbins, M. K.; Biscardi, F. H.; Abdul-Salaam, S.; Koren, H. S. (1998). Repeated exposure
of humans to ozone: Pulmonary function and symptom responses. Research Triangle Park, NC: U.S. Environmental
Protection Agency. 038663
Folinsbee, L. J.; Drinkwater, B. L.; Bedi, J. F.; Horvath, S. M. (1978). The influence of exercise on the pulmonary function
changes due to exposure to low concentrations of ozone. In LJ Folinsbee; JA Wagner; JF Borgia; BL Drinkwater;
JA Gliner; JF Bedi (Eds.), Environmental stress: individual human adaptations (pp. 125-145). New York, NY:
Academic Press. 039060
Folinsbee, L. J.; Hazucha, M. J. (1989). Persistence of ozone-induced changes in lung function and airway responsiveness.
In Atmospheric ozone research and its policy implications (pp. 483-492). Amsterdam, The Netherlands: Elsevier.
041732
Folinsbee, L. J.; Hazucha, M. J. (2000). Time course of response to ozone exposure in healthy adult females. Inhal Toxicol,
12: 151-167.001701
Folinsbee, L. J.; Horstman, D. H.; Kehrl, H. R.; Harder, S.; Abdul-Salaam, S.; Ives, P. J. (1994). Respiratory responses to
repeated prolonged exposure to 0.12 ppm ozone. Am J Respir Crit Care Med, 149: 98-105. 044189
Folinsbee, L. J.; McDonnell, W. F.; Horstman, D. H. (1988). Pulmonary function and symptom responses after 6.6-hour
exposure to 012 ppm ozone with moderate exercise. J Air Waste Manag Assoc, 38: 28-35. 040898
Folinsbee, L. J.; Silverman, F.; Shephard, R. J. (1977). Decrease of maximum work performance following ozone
exposure. J Appl Physiol, 42: 531-536. 038283
Fortino, V; Maioli, E.; Torricelli, C.; Davis, P.; Valacchi, G. (2007). Cutaneous MMPs are differently modulated by
environmental stressors in old and young mice. Toxicol Lett, 173: 73-79.
http://dx.doi.0rg/10.1016/i.toxlet.2007.06.004596382
Foster, W. M.; Stetkiewicz, P. T (1996). Regional clearance of solute from the respiratory epithelia: 18—20 h postexposure
to ozone. J Appl Physiol, 81:1143-1149. 079920
Fox, S. D.; Adams, W. C.; Brookes, K. A.; Lasley, B. L. (1993). Enhanced response to ozone exposure during the follicular
phase of the menstrual cycle. Environ Health Perspect, 101: 242-244. 043906
Foxcroft, W. J.; Adams, W. C. (1986). Effects of ozone exposure on four consecutive days on work performance and
V02max. J Appl Physiol, 61: 960-966. 040463
Frampton, M. W.; Morrow, P. E.; Torres, A.; Cox, C.; Voter, K. Z.; Utell, M. J.; Gibb, F. R.; Speers, D. M. (1997). Ozone
responsiveness in smokers and nonsmokers. Am J Respir Crit Care Med, 155: 116-121. 082692
Frampton, M. W.; Pryor, W. A.; Cueto, R.; Cox, C.; Morrow, P. E.; Utell, M. J. (1999). Ozone exposure increases aldehydes
in epithelial lining fluid in human lung. Am J Respir Crit Care Med, 159: 1134-1137. 040757
Franklin, M.; Schwartz, J. (2008). The impact of secondary particles on the association between ambient ozone and
mortality. Environ Health Perspect, 116: 453-458. http://dx.doi.org/10.1289/ehp.10777156448
Franze, T; Weller, M. G; Niessner, R.; Poschl, U. (2005). Protein nitration by polluted air. Environ Sci Technol, 39: 1673-
1678. 066088
Friedman, M.; Gallo, J. M.; Nichols, H. P.; Bromberg, P. A. (1983). Changes in inert gas rebreathing parameters after ozone
exposure in dogs. Am Rev Respir Dis, 128: 851-856. 040547
Fung, K. Y; Luginaah, L; Gorey, K. M.; Webster, G. (2005). Air pollution and daily hospital admissions for cardiovascular
diseases in Windsor, Ontario. Can J Public Health, 96: 29-33. 074322
Gao, X.; Raghavamenon, A. C.; DAuvergne, O.; Uppu, R. M. (2009). Cholesterol secoaldehyde induces apoptosis in J774
macrophages via mitochondrial pathway but not involving reactive oxygen species as mediators. Biochem Biophys
Res Commun, 389: 382-387. http://dx.doi.Org/10.1016/j.bbrc.2009.09.005 200764
Garantziotis, S.; Li, Z.; Potts, E. N.; Lindsey, J. Y; Stober, V. P.; Polosukhin, V. V; Blackwell, T. S.; Schwartz, D. A.;
Foster, W. M.; Hollingsworth, J. W. (2010). TLR4 is necessary for hyaluronan-mediated airway
hyperresponsiveness after ozone inhalation. Am J Respir Crit Care Med, 181: 666-675.
http://dx.doi.org/10.1164/rccm.200903-0381OC624947
March 2011 6-185 DRAFT - DO NOT CITE OR QUOTE
-------
Gent, J. R; Triche, E. W.; Holford, T. R.; Belanger, K.; Bracken, M. B.; Beckett, W. S.; Leaderer, B. P. (2003). Association
of low-level ozone and fine particles with respiratory symptoms in children with asthma. JAMA, 290: 1859-1867.
http://dx.doi.org/10.1001/iama.290.14.1859052885
Gershwin, L. J.; Osebold, J. W.; Zee, Y. C. (1981). Immunoglobulin E-containing cells in mouse lung following allergen
inhalation and ozone exposure. Int Arch Allergy Immunol, 65: 266-277. 039729
Gielen, M. H.; Van Der Zee, S. C.; Van Wijnen, J. H.; Van Steen, C. J.; Brunekreef, B. (1997). Acute effects of summer air
pollution on respiratory health of asthmatic children. Am J Respir Crit Care Med, 155: 2105-2108. 083592
Gilliland, R D.; Berhane, K.; Rappaport, E. B.; Thomas, D. C.; Avol, E.; Gauderman, W. J.; London, S. J.; Margolis, H. G;
McConnell, R.; Islam, K. T.; Peters, J. M. (2001). The effects of ambient air pollution on school absenteeism due to
respiratory illnesses. Epidemiology, 12: 43-54. 013232
Gilmour, M. L; Jakab, G. J. (1991). Modulation of immune function in mice exposed to 08 ppm ozone. Inhal Toxicol, 3:
293-308. 042391
Girardot, S. P.; Ryan, P. B.; Smith, S. M.; Davis, W. T; Hamilton, C. B.; Obenour, R. A.; Renfro, J. R.; Tromatore, K. A.;
Reed, G. D. (2006). Ozone and PM25 exposure and acute pulmonary health effects: A study of hikers in the Great
Smoky Mountains National Park. Environ Health Perspect, 113: 612-617. http://dx.doi.org/10.1289/ehp.8637
088271
Gold, D. R.; Damokosh, A. L; Pope CA 3rd; Dockery, D. W.; McDonnell, W. R; Serrano, P.; Retama, A.; Castillejos, M.
(1999). Particulate and ozone pollutant effects on the respiratory function of children in southwest Mexico City.
Epidemiology, 10: 8-16. 086919
Goldberg, M. S.; Giannetti, N.; Burnett, R. T; Mayo, N. E.; Valois, M. R; Brophy, J. M. (2008). A panel study in
congestive heart failure to estimate the short-term effects from personal factors and environmental conditions on
oxygen saturation and pulse rate. Occup Environ Med, 65: 659-666. http://dx.doi.org/10.1136/oem.2007.034934
180380
Gong, H. Jr; Bradley, P. W.; Simmons, M. S.; Tashkin, D. P. (1986). Impaired exercise performance and pulmonary
function in elite cyclists during low-level ozone exposure in a hot environment. Am J Respir Crit Care Med, 134:
726-733. 040465
Gong, H. Jr; McManus, M. S.; Linn, W. S. (1997). Attenuated response to repeated daily ozone exposures in asthmatic
subjects. Arch Environ Occup Health, 52: 34-41. 082696
Gong, H. Jr; Shamoo, D. A.; Anderson, K. R.; Linn, W. S. (1997). Responses of older men with and without chronic
obstructive pulmonary disease to prolonged ozone exposure. Arch Environ Occup Health, 52: 18-25. 083593
Gong, H. Jr; Wong, R.; Sarma, R. J.; Linn, W. S.; Sullivan, E. D.; Shamoo, D. A.; Anderson, K. R.; Prasad, S. B. (1998).
Cardiovascular effects of ozone exposure in human volunteers. Am J Respir Crit Care Med, 158: 538-546. 029938
Gonzalez-Pina, R.; Escalante-Membrillo, C.; Alfaro-Rodriguez, A.; Gonzalez-Maciel, A. (2008). Prenatal exposure to
ozone disrupts cerebellar monoamine contents in newborn rats. Neurochem Res, 33: 912-918.
http://dx.doi.org/10.1007/sll064-007-9534-3475317
Graham, J. A.; Menzel, D. B.; Miller, R J.; Illing, J. W.; Gardner, D. E. (1981). Influence of ozone on pentobarbital-induced
sleeping time in mice, rats, and hamsters. Toxicol Appl Pharmacol, 61: 64-73. http://dx.doi.org/10.1016/0041-
008X(81)90008-9 039415
Grunewald, J.; Eklund, A. (2007). Role of CD4+ T cells in sarcoidosis. Proc Am Thorac Soc, 4: 461-464. 187925
Gryparis, A.; Forsberg, B.; Katsouyanni, K.; Analitis, A.; Touloumi, G; Schwartz, J.; Samoli, E.; Medina, S.; Anderson, H.
R.; Niciu, E. M.; Wichmann, H. E.; Kriz, B.; Kosnik, M.; Skorkovsky, J.; Vonk, J. M.; Dortbudak, Z. (2004). Acute
effects of ozone on mortality from the "Air pollution and health: A European approach" project. Am J Respir Crit
Care Med, 170: 1080-1087. http://dx.doi.org/10.1164/rccm.200403-333OC057276
Guevara-Guzman, R.; Arriaga, V; Kendrick, K. M.; Bernal, C.; Vega, X.; Mercado-Gomez, O. R; Rivas-Arancibia, S.
(2009). Estradiol prevents ozone-induced increases in brain lipid peroxidation and impaired social recognition
memory in female rats. Neuroscience, 159: 940-950. http://dx.doi.Org/10.1016/j.neuroscience.2009.01.047 596385
Hackney, J. D.; Linn, W. S.; Mohler, J. G; Pedersen, E. E.; Breisacher, P.; Russo, A. (1975). Experimental studies on
human health effects of air pollutants: II four-hour exposure to ozone alone and in combination with other pollutant
gases. Arch Environ Occup Health, 30: 379-384. 039208
March 2011 6-186 DRAFT - DO NOT CITE OR QUOTE
-------
Halonen, J. I.; Lanki, T.; Tiittanen, P.; Niemi, J. V; Loh, M.; Pekkanen J (2009). Ozone and cause-specific
cardiorespiratory morbidity and mortality. J Epidemiol Community Health, TBD: TBD.
http://dx.doi.org/10.1136/iech.2009.087106625764
Hamade, A. K.; Misra, V; Rabold, R.; Tankersley, C. G. (2010). Age-related changes in cardiac and respiratory adaptation
to acute ozone and carbon black exposures: Interstrain variation in mice. Inhal Toxicol, 22: 84-94.
http://dx.doi.org/10.3109/08958378.2010.503974666324
Hamade, A. K.; Rabold, R.; Tankersley, C. G. (2008). Adverse cardiovascular effects with acute particulate matter and
ozone exposures: Interstrain variation in mice. Environ Health Perspect, 116: 1033-1039. 156515
Hamade, A. K.; Tankersley, C. G. (2009). Interstrain variation in cardiac and respiratory adaptation to repeated ozone and
particulate matter exposures. Am J Physiol Regul Integr Comp Physiol, 296: R1202-R1215.
http://dx.doi.org/10.1152/aipregu.90808.2008596386
Han, S. G; Andrews, R.; Gairola, C. G; Bhalla, D. K. (2008). Acute pulmonary effects of combined exposure to carbon
nanotubes and ozone in mice. Inhal Toxicol, 20: 391-398. http://dx.doi.org/10.1080/08958370801904014 596387
Hazucha, M. J.; Folinsbee, L. J.; Bromberg, P. A. (2003). Distribution and reproducibility of spirometric response to ozone
by gender and age. J Appl Physiol, 95: 1917-1925. 048168
Hazucha, M. J.; Folinsbee, L. J.; Seal, E. Jr (1992). Effects of steady-state and variable ozone concentration profiles on
pulmonary function. Am J Respir Crit Care Med, 146: 1487-1493. 042789
Hazucha, M. J.; Madden, M.; Pape, G; Becker, S.; Devlin, R.; Koren, H. S.; Kehrl, H.; Bromberg, P. A. (1996). Effects of
cyclo-oxygenase inhibition on ozone-induced respiratory inflammation and lung function changes. Eur J Appl
Physiol, 73: 17-27. 043923
Health Effects Institute (2003). Revised analyses of the National Morbidity, Mortality, and Air Pollution Study
(NMMAPS), part II. Boston, MA: Health Effects Institute. 042829
Heidenfelder, B. L.; Reif, D. M.; Harkema, J. R.; Cohen Hubal, E. A.; Hudgens, E. E.; Bramble, L. A.; Wagner, J. G;
Morishita, M.; Keeler, G. J.; Edwards, S. W.; Gallagher, J. E. (2009). Comparative microarray analysis and
pulmonary changes in brown Norway rats exposed to ovalbumin and concentrated air particulates. Toxicol Sci, 108:
207-221. 190026
Hemmingsen, A.; Fryer, A. A.; Hepple, M.; Strange, R. C.; Spiteri, M. A. (2001). Simultaneous identification of GSTP1
Ilel05-^Vall05 and Alall4^Vall 14 substitutions using an amplification refractory mutation systempolymerase
chain reactionassay: studies in patients with asthma. Respir Res, 2: 255-260. http://dx.doi.org/10.1186/rr64 670827
Henrotin, J. B.; Besancenot, J. P.; Bejot, Y; Giroud, M. (2007). Short-term effects of ozone air pollution on ischaemic
stroke occurrence: A case-crossover analysis from a 10-year population-based study in Dijon, France. Occup
Environ Med, 64: 439-445. 093270
Hernandez-Cadena, L.; Holguin, F.; Barraza-Villarreal, A.; Del Rio-Navarro, B. E.; Sienra-Monge, J. J.; Romieu, I. (2009).
Increased levels of outdoor air pollutants are associated with reduced bronchodilation in children with asthma.
Chest, 136: 1529-1536. http://dx.doi.org/10.1378/chest.08-1463 594283
Hicks, A.; Goodnow, R. Jr; Cavallo, G; Tannu, S. A.; Ventre, J. D.; Lavelle, D.; Lora, J. M.; Satjawatcharaphong, J.;
Brovarney, M.; Dabbagh, K.; Tare, N. S.; Oh, H.; Lamb, M.; Sidduri, A.; Dominique, R.; Qiao, Q.; Lou, J. P.;
Gillespie, P.; Fotouhi, N.; Kowalczyk, A.; Kurylko, G; Hamid, R.; Wright, M. B.; Pamidimukkala, A.; Egan, T;
Gubler, U.; Hoffman, A. F.; Wei, X.; Li, Y. L.; O'Neil, J.; Marcano, R.; Pozzani, K.; Molinaro, T; Santiago, J.;
Singer, L.; Hargaden, M.; Moore, D.; Catala, A. R.; Chao, L. C.; Benson, J.; March, T; Venkat, R.; Mancebo, H.;
Renzetti, L. M. (2010). Effects of LTB4 receptor antagonism on pulmonary inflammation in rodents and non-
human primates. Prostaglandins Other Lipid Medial, TBD: TBD.
http://dx.doi.0rg/10.1016/i.prostaglandins.2010.02.003624932
Higgins, I. T. T; DArcy, J. B.; Gibbons, D. L; Avol, E. L.; Gross, K. B. (1990). Effect of exposures to ambient ozone on
ventilatory lung function in children. Am J Respir Crit Care Med, 141: 1136-1146. 042195
Hiltermann, J. T. N.; Lapperre, T. S.; Van Bree, L.; Steerenberg, P. A.; Brahim, J. J.; Sont, J. K.; Sterk, P. J.; Hiemstra, P. S.;
Stolk, J. (1999). Ozone-induced inflammation assessed in sputum and bronchial lavage fluid from asthmatics: a
new noninvasive tool in epidemiologic studies on air pollution and asthma. Free Radic Biol Med, 27: 1448-1454.
013196
March 2011 6-187 DRAFT - DO NOT CITE OR QUOTE
-------
Hiltermann, T. J. N.; de Bruijne, C. R.; Stolk, J.; Zwinderman, A. H.; Spieksma FThM; Roemer, W.; Steerenberg, P. A.;
Fischer, P. H.; van Bree, L.; Hiemstra, P. S. (1997). Effects of photochemical air pollution and allergen exposure on
upper respiratory tract inflammation in asthmatics. Am J Respir Crit Care Med, 156: 1765-1772. 084979
Hiltermann, T. J. N.; Peters, E. A.; Alberts, B.; Kwikkers, K.; Borggreven, P. A.; Hiemstra, P. S.; Dijkman, J. H.; van Bree,
L. A.; Stolk, J. (1998). Ozone-induced airway hyperresponsiveness in patients with asthma: role of neutrophil-
derived serine proteinases. Free Radic Biol Med, 24: 952-958. 086158
Hiltermann, T. J. N.; Stolk, J.; Hiemstra, P. S.; Fokkens, P. H. B.; Rombout, P. J. A.; Sont, J. K.; Sterk, P. J.; Dijkman, J. H.
(1995). Effect of ozone exposure on maximal airway narrowing in non-asthmatic and asthmatic subjects. Clin Sci
(Lond), 89: 619-624. 078494
Hinwood, A. L.; De Klerk, N.; Rodriguez, C.; Jacoby, P.; Runnion, T.; Rye, P.; Landau, L.; Murray, F.; Feldwick, M.;
Spickett, J. (2006). The relationship between changes in daily air pollution and hospitalizations in Perth, Australia
1992-1998: A case-crossover study. Int J Environ Health Res, 16: 27-46.
http://dx.doi.org/10.1080/09603120500397680088976
Hoek, G; Brunekreef, B. (1995). Effect of photochemical air pollution on acute respiratory symptoms in children. Am J
Respir Crit Care Med, 151: 27-32. 046184
Hoek, G; Brunekreef, B.; Kosterink, P.; Van den Berg, R.; Hofschreuder, P. (1993). Effect of ambient ozone on peak
expiratory flow of exercising children in the Netherlands. Arch Environ Occup Health, 48: 27-32.
http://dx.doi.org/10.1080/00039896.1993.9938390043009
Holz, O.; Jorres, R. A.; Timm, P.; Mucke, M.; Richter, K.; Koschyk, S.; Magnussen, H. (1999). Ozone-induced airway
inflammatory changes differ between individuals and are reproducible. Am J Respir Crit Care Med, 159: 776-784.
058731
Holz, O.; Mucke, M.; Paasch, K.; Bohme, S.; Timm, P.; Richter, K.; Magnussen, H.; Jorres, R. A. (2002). Repeated ozone
exposures enhance bronchial allergen responses in subjects with rhinitis or asthma. Clin Exp Allergy, 32: 681-689.
041632
Holz, O.; Tal-Singer, R.; Kanniess, F.; Simpson, K. J.; Gibson, A.; Vessey, R. S. J.; Janicki, S.; Magnussen, H.; Jorres, R.
A.; Richter, K. (2005). Validation of the human ozone challenge model as a tool for assessing anti-inflammatory
drugs in early development. J Clin Pharmacol, 45: 498-503. 077170
Hoppe, P.; Peters, A.; Rabe, G; Praml, G; Lindner, J.; Jakobi, G; Fruhmann, G; Nowak, D. (2003). Environmental ozone
effects in different population subgroups. Int J Hyg Environ Health, 206: 505-516. http://dx.doi.org/10.1078/1438-
4639-00250 055618
Horstman, D. H.; Ball, B. A.; Brown, J.; Gerrity, T; Folinsbee, L. J. (1995). Comparison of pulmonary responses of
asthmatic and nonasthmatic subjects performing light exercise while exposed to a low level of ozone. Toxicol Ind
Health, 11: 369-385. 075834
Horstman, D. H.; Folinsbee, L. J.; Ives, P. J.; Abdul-Salaam, S.; McDonnell, W. F. (1990). Ozone concentration and
pulmonary response relationships for 6.6-hour exposures with five hours of moderate exercise to 0.08, 0.10, and
0.12 ppm. Am J Respir Crit Care Med, 142: 1158-1163. 042187
Horvath, S. M.; Gliner, J. A.; Folinsbee, L. J. (1981). Adaptation to ozone: duration of effect. Am Rev Respir Dis, 123:
496-499. 039221
Horvath, S. M.; Gliner, J. A.; Matsen-Twisdale, J. A. (1979). Pulmonary function and maximum exercise responses
following acute ozone exposure. Aviat Space Environ Med, 50: 901-905. 039222
Hosseinpoor, A. R.; Forouzanfar, M. H.; Yunesian, M.; Asghari, F.; Naieni, K. H.; Farhood, D. (2005). Air pollution and
hospitalization due to angina pectoris in Tehran, Iran: A time-series study. Environ Res, 99: 126-131. 087413
Housley, D. G; Eccles, R.; Richards, R. J. (1996). Gender difference in the concentration of the antioxidant uric acid in
human nasal lavage. Acta Otolaryngol, 116: 751-754. 080811
Howarth, P. H.; Persson, C. G; Meltzer, E. O.; Jacobson, M. R.; Durham, S. R.; Silkoff, P. E. (2005). Objective monitoring
of nasal airway inflammation in rhinitis. J Allergy Clin Immunol, 115: S414-S441. 625805
Huffman, L. J.; Beighley, C. M.; Frazer, D. G; McKinney, W. G; Porter, D. W. (2006). Increased susceptibility of the lungs
of hyperthyroid rats to oxidant injury: Specificity of effects. Toxicology, 225: 119-127.
http://dx.doi.0rg/10.1016/i.tox.2006.05.008596388
March 2011 6-188 DRAFT - DO NOT CITE OR QUOTE
-------
Hunt, J. (2002). Exhaled breath condensate: An evolving tool for noninvasive evaluation of lung disease. J Allergy Clin
Immunol, 110: 28-34. http://dx.doi.org/10.1067/mai.2002.124966625808
Hunt, J. E; Fang, K.; Malik, R.; Snyder, A.; Malhotra, N.; Platts-Mills, T. A. E.; Gaston, B. (2000). Endogenous airway
acidification: implications for asthma pathophysiology. Am J Respir Crit Care Med, 161: 694-699.
http://ajrccm.atsjournals.Org/cgi/content/full/161/3/694. 002173
Hurst, D. J.; Gardner, D. E.; Coffin, D. L. (1970). Effect of ozone on acid hydrolases of the pulmonary alveolar
macrophage. J Leukoc Biol, 8: 288-300. 015591
lijima, M. K.; Kobayashi, T. (2004). Nasal allergy-like symptoms aggravated by ozone exposure in a concentration-
dependent manner in guinea pigs. Toxicology, 199: 73-83. http://dx.doi.Org/10.1016/i.tox.2004.01.008 596389
Inoue, K.; Takano, H.; Kaewamatawong, T.; Shimada, A.; Suzuki, J.; Yanagisawa, R.; Tasaka, S.; Ishizaka, A.; Satoh, M.
(2008). Role of metallothionein in lung inflammation induced by ozone exposure in mice. Free Radic Biol Med, 45:
1714-1722. http://dx.doi.0rg/10.1016/i.freeradbiomed.2008.09.008 197803
Ito, K.; De Leon, S. F.; Lippmann, M. (2005). Associations between ozone and daily mortality, analysis and meta-analysis.
Epidemiology, 16: 446-457. 074346
Ito, K.; Thurston, G D.; Silverman, R. A. (2007). Characterization of PM2.5, gaseous pollutants, and meteorological
interactions in the context of time-series health effects models. J Expo Sci Environ Epidemiol, 17: S45-S60. 156594
Iwasaki, T.; Takahashi, M.; Saito, H.; Arito, H. (1998). Adaptation of extrapulmonary responses to ozone exposure in
conscious rats. Ind Health, 36: 57-60. 086165
Jakab, G. J.; Hmieleski, R. R. (1988). Reduction of influenza virus pathogenesis by exposure to 05 ppm ozone. J Toxicol
Environ Health, 23: 455-472. 041806
Jalaludin, B. B.; Chey, T.; O'Toole, B. L; Smith, W. T.; Capon, A. G; Leeder, S. R. (2000). Acute effects of low levels of
ambient ozone on peak expiratory flow rate in a cohort of Australian children. Int J Epidemiol, 29: 549-557.
http://dx.doi.0rg/10.1093/iie/29.3.549011929
Jalaludin, B. B.; O'Toole, B. L; Leeder, S. R. (2004). Acute effects of urban ambient air pollution on respiratory symptoms,
asthma medication use, and doctor visits for asthma in a cohort of Australian children. Environ Res, 95: 32-42.
http://dx.doi.org/10.1016/S0013-9351(03)00038-0 056595
Janero, D. R. (1990). Malondialdehyde and thiobarbituric acid-reactivity as diagnostic indices of lipid peroxidation and
peroxidative tissue injury. Free Radic Biol Med, 9: 515-540. http://dx.doi.org/10.1016/0891-5849(90)90131-2
625809
Jang, A. S.; Choi, I. S.; Yang, S. Y; Kim, Y G; Lee, J. H.; Park, S. W.; Park, C. S. (2005). Antioxidant responsiveness in
BALB/c mice exposed to ozone. Respiration, 72: 79-84. http://dx.doi.org/10.1159/000083405 195638
Janic, B.; Umstead, T. M.; Phelps, D. S.; Floras, J. (2005). Modulatory effects of ozone on THP-1 cells in response to SP-A
stimulation. Am J Physiol Lung Cell Mol Physiol, 288: L317-L325. http://dx.doi.org/10.1152/ajplung.00125.2004
483658
Jansson, M.; Bergstrom, A.-K.; Drakare, S.; Blomqvist, P. (2001). Nutrient limitation of bacterioplankton and
phytoplankton in humic lakes in northern Sweden. Freshw Biol, 14: 76-85. 092076
Johnston, C.; Holm, B.; Gelein, R.; Finkelstein, J. (2006). Postnatal lung development: Immediate-early gene responses
post ozone and LPS exposure. Inhal Toxicol, 18: 875-883. http://dx.doi.org/10.1080/08958370600822466 097439
Johnston, R. A.; Mizgerd, J. P.; Flynt, L.; Quinton, L. J.; Williams, E. S.; Shore, S. A. (2007). Type I interleukin-1 receptor
is required for pulmonary responses to subacute ozone exposure in mice. Am J Respir Cell Mol Biol, 37: 477-484.
http://dx.doi.org/10.1165/rcmb.2006-0315OC596392
Johnston, R. A.; Mizgerd, J. P.; Shore, S. A. (2005). CXCR2 is essential for maximal neutrophil recruitment and
methacholine responsiveness after ozone exposure. Am J Physiol Lung Cell Mol Physiol, 288: L61-L67.
http://dx.doi.org/10.1152/ajplung.00101.2004 00101.2004 596393
Johnston, R. A.; Schwartzman, I. N.; Flynt, L.; Shore, S. A. (2005). Role of interleukin-6 in murine airway responses to
ozone. Am J Physiol Lung Cell Mol Physiol, 288: L390-L397. http://dx.doi.org/10.1152/ajplung.00007.2004
596394
March 2011 6-189 DRAFT - DO NOT CITE OR QUOTE
-------
Jones, S. L.; Kittelson, I; Cowan, J. O.; Flannery, E. M.; Hancox, R. I; McLachlan, C. R.; Taylor, D. R. (2001). The
predictive value of exhaled nitric oxide measurements in assessing changes in asthma control. Am J Respir Crit
CareMed, 164: 738-743. http://ajrccm.atsjournals.Org/cgi/content/abstract/164/5/738. 625816
Jorres, R. A.; Holz, O.; Zachgo, W.; Timm, P.; Koschyk, S.; Muller, B.; Grimminger, R; Seeger, W.; Kelly, F. J.; Dunster,
C.; Frischer, T; Lubec, G; Waschewski, M.; Niendorf, A.; Magnussen, H. (2000). The effect of repeated ozone
exposures on inflammatory markers in bronchoalveolar lavage fluid and mucosal biopsies. Am J Respir Crit Care
Med, 161: 1855-1861.005654
Jorres, R.; Nowak, D.; Magnussen, H.; Speckin, R; Koschyk, S. (1996). The effect of ozone exposure on allergen
responsiveness in subjects with asthma or rhinitis. Am J Respir Crit Care Med, 153: 56-64. 078122
Just, J.; Segala, C.; Sahraoui, F.; Priol, G; Grimfeld, A.; Neukirch, F. (2002). Short-term health effects of particulate and
photochemical air pollution in asthmatic children. Eur Respir J, 20: 899-906.
http://dx.doi.org/10.1183/09031936.02.00236902035429
Katsouyanni, K.; Samet, J. M.; Anderson, H. R.; Atkinson, R.; Le Tertre, A.; Medina, S.; Samoli, E.; Touloumi, G; Burnett,
R. T; Krewski, D.; Ramsay, T; Dominici, F.; Peng, R. D.; Schwartz, J.; Zanobetti, A. (2009). Air pollution and
health: A European and North American approach (APHENA) (Report No. Research Report 142). Boston, MA:
Health Effects Institute. http://pubs.healtheffects.org/view.php?id=327. 199899
Katsouyanni, K.; Touloumi, G; Samoli, E.; Gryparis, A.; Le Tertre, A.; Monopolis, Y; Rossi, G; Zmirou, D.; Ballester, F.;
Boumghar, A.; Anderson, H. R.; Wojtyniak, B.; Paldy, A.; Braunstein, R.; Pekkanen, J.; Schindler, C.; Schwartz, J.
(2001). Confounding and effect modification in the short-term effects of ambient particles on total mortality: results
from 29 European cities within the APHEA2 project. Epidemiology, 12: 521-531. 019008
Kehrl, H. R.; Hazucha, M. J.; Solic, J. J.; Bromberg, P. A. (1985). Responses of subjects with chronic obstructive
pulmonary disease after exposures to 03 ppm ozone. Am Rev Respir Dis, 131: 719-724. 040294
Kehrl, H. R.; Peden, D. B.; Ball, B. A.; Folinsbee, L. J.; Horstman, D. H. (1999). Increased specific airway reactivity of
persons with mild allergic asthma after 7.6 hours of exposure to 0.16 ppm ozone. J Allergy Clin Immunol, 104:
1198-1204.022101
Kehrl, H. R.; Vincent, L. M.; Kowalsky, R. J.; Horstman, D. H.; O'Neil, J. J.; McCartney, W. H.; Bromberg, P. A. (1987).
Ozone exposure increases respiratory epithelial permeability in humans. Am Rev Respir Dis, 135: 1124-1128.
040824
Kenyon, N. J.; Last, M. S.; Eiserich, J. P.; Morrissey, B. M.; Temple, L. M.; Last, J. A. (2006). Differentiation of the roles
of NO from airway epithelium and inflammatory cells in ozone-induced lung inflammation. Toxicol Appl
Pharmacol, 215: 250-259. http://dx.doi.Org/10.1016/j.taap.2006.03.005 596396
Kharitonov, S. A.; Barnes, P. J. (2000). Clinical aspects of exhaled nitric oxide. Eur Respir J, 16: 781-792. 625817
Khatri, S. B.; Holguin, F. C.; Ryan, P. B.; Mannino, D.; Erzurum, S. C.; Teague, W. G. (2009). Association of ambient
ozone exposure with airway inflammation and allergy in adults with asthma. J Asthma, 46: 777-785.
http://dx.doi.org/10.1080/02770900902779284594282
Kim, C. S.; Alexis, N. E.; Rappold, A. G; Kehrl, H.; Hazucha, M. J.; Lay, J. C.; Schmitt, M. T; Case, M.; Devlin, R. B.;
Peden, D. B.; Diaz-Sanchez, D. (In Press). Lung function and inflammatory responses in healthy young adults
exposed to 0.06 ppm ozone for 6.6 hours. Am J Respir Crit CareMed. http://dx.doi.org/10.1164/rccm.201011-
1813OC 674869
Kinney, P. L.; Thurston, G. D.; Raizenne, M. (1996). The effects of ambient ozone on lung function in children: A
reanalysis of six summer camp studies. Environ Health Perspect, 104: 170-174. 079203
Kleeberger, S. R.; Reddy, S.; Zhang, L.-Y.; Jedlicka, A. E. (2000). Genetic susceptibility to ozone-induced lung
hyperpermeability: role of toll-like receptor 4. Am J Respir Cell Mol Biol, 22: 620-627. 014895
Klestadt, D.; Laval-Gilly, P.; Foucaud, L.; Falla, J. (2005). Influences of ozone exposure upon macrophage responsivity to
N-formyl-methionyl-leucyl-phenylalanine: Mobility and metabolic changes. Toxicol In Vitro, 19: 199-206.
http://dx.doi.0rg/10.1016/i.tiv.2004.08.004 130425
Kodavanti, U. P.; Thomas, R.; Ledbetter, A. D.; Schladweiler, M. C.; Shannahan, J. H.; Wallenborn, J. G; Lund, A. K.;
Campen, M. J.; Butler, E. O.; Gottipolu, R. R.; Nyska, A.; Richards, J. E.; Andrews, D.; Jaskot, R. H.; McKee, J.;
Kotha, S. R.; Patel, R. B.; Parianandi, N. L. (In Press). Vascular and cardiac impairments in rats Inhaling ozone and
diesel exhaust particles. Environ Health Perspect. http://dx.doi.org/10.1289/ehp.1002386 666323
March 2011 6-190 DRAFT - DO NOT CITE OR QUOTE
-------
Kooter, I. M.; Pennings, J. L.; Fokkens, P. H.; Leseman, D. L.; Boere, A. J.; Gerlofs-Nijland, M. E.; Cassee, F. R.; Schalk,
J. A.; Orzechowski, T. J.; Schaap, M. M.; Breit, T. M.; Dormans, J. A.; van Oostrom, C. T.; de Vries, A.; van Steeg,
H. (2007). Ozone induces clear cellular and molecular responses in the mouse lung independently of the
transcription-coupled repair status. J Appl Physiol, 102: 1185-1192.
http://dx.doi.org/10.1152/iapplphvsiol.00796.2006596397
Korrick, S. A.; Neas, L. M.; Dockery, D. W.; Gold, D. R.; Allen, G. A.; Hill, L. B.; Kimball, K. D.; Rosner, B. A.; Speizer,
F. E. (1998). Effects of ozone and other pollutants on the pulmonary function of adult hikers. Environ Health
Perspect, 106: 93-99. http://dx.doi.org/10.1289/ehp.9810693 026841
Kostikas, K.; Papatheodorou, G; Ganas, K.; Psathakis, K.; Panagou, P.; Loukides, S. (2002). pH in expired breath
condensate of patients with inflammatory airway diseases. Am J Respir Crit Care Med, 165: 1364-1370.
http://ajrccm.atsjournals.org/cgi/content/abstract/165/10/1364. 625821
Kreit, J. W.; Gross, K. B.; Moore, T. B.; Lorenzen, T. J.; DArcy, J.; Eschenbacher, W. L. (1989). Ozone-induced changes in
pulmonary function and bronchial responsiveness in asthmatics. J Appl Physiol, 66: 217-222. 041817
Krevans, J. R.; Asper, S. P. Jr; Rienhoff, W. F. Jr (1962). Fatal aplastic anemia following use of potassium perchlorate in
thyrotoxicosis. JAMA, 181: 162-164. 014710
Kulle, T. J.; Sauder, L. R.; Hebel, J. R.; Chatham, M. D. (1985). Ozone response relationships in healthy nonsmokers. Am
Rev Respir Dis, 132: 36-41. 040311
Kulle, T. J.; Sauder, L. R.; Kerr, H. D.; Farrell, B. P.; Bermel, M. S.; Smith, D. M. (1982). Duration of pulmonary function
adaptation to ozone in humans. Am Ind Hyg Assoc J, 43: 832-837. 040668
Lagorio, S.; Forastiere, F.; Pistelli, R.; lavarone, L; Michelozzi, P.; Fano, V; Marconi, A.; Ziemacki, G; Ostro, B. D.
(2006). Air pollution and lung function among susceptible adult subjects: A panel study. Environ Health, 5:11.
http://dx.doi.org/10.1186/1476-069X-5-ll 089800
Lanki, T.; Pekkanen, J.; Aalto, P.; Elosua, R.; Berglind, N.; D'Ippoliti, D.; Kulmala, M.; Nyberg, F.; Peters, A.; Picciotto, S.;
Salomaa, V; Sunyer, J.; Tiittanen, P.; Von Klot, S.; Forastiere, F.; for the HEAPSS Study Group (2006).
Associations of traffic-related air pollutants with hospitalisation for first acute myocardial infarction: the HEAPSS
study. Occup Environ Med, 63: 844-851.089788
Larrieu, S.; Jusot, J.-F.; Blanchard, M.; Prouvost, H.; Declercq, C.; Fabre, P.; Pascal, L.; Le Tertre, A.; Wagner, V; Riviere,
S.; Chardon, B.; Borelli, D.; Cassadou, S.; Eilstein, D.; Lefranc, A. (2007). Short term effects of air pollution on
hospitalizations for cardiovascular diseases in eight French cities: The PSAS program. Sci Total Environ, 387: 105-
112.093031
Larsen, S. T; Matsubara, S.; McConville, G; Poulsen, S. S.; Gelfand, E. W. (2010). Ozone increases airway hyperreactivity
and mucus hyperproduction in mice previously exposed to allergen. J Toxicol Environ Health A, 73: 738-747.
http://dx.doi.org/10.1080/15287391003614034628560
Laskin, D. L.; Heck, D. E.; Laskin, J. D. (1998). Role of inflammatory cytokines and nitric oxide in hepatic and pulmonary
toxicity. Toxicol Lett, 102-103: 289-293. 015425
Laskin, D. L.; Laskin, J. D. (2001). Role of macrophages and inflammatory mediators in chemically induced toxicity.
Toxicology, 160: 111-118. 016158
Laskin, D. L.; Pendino, K. J.; Punjabi, C. J.; del Valle, M. R.; Laskin, J. D. (1994). Pulmonary and hepatic effects of
inhaled ozone in rats. Environ Health Perspect, 10: 61-64. 076154
Laskin, J. D.; Heck, D. E.; Laskin, D. L. (1996). Nitric oxide production in the lung and liver following inhalation of the
pulmonary irritant ozone. Adv Exp Med Biol, 387: 141-146. 015771
Last, J. A.; Gohil, K.; Mathrani, V. C.; Kenyon, N. J. (2005). Systemic responses to inhaled ozone in mice: cachexia and
down-regulation of liver xenobiotic metabolizing genes. Toxicol Appl Pharmacol, 208: 117-126.
http://dx.doi.0rg/10.1016/i.taap.2005.02.001 596400
Lawrence, S. O.; Simpson-Haidaris, P. J. (2004). Regulated de novo biosynthesis of fibrinogen in extrahepatic epithelial
cells in response to inflammation. Thromb Haemostasis, 92: 234-243. http://dx.doi.org/10.1160/TH04-01-0024
627654
March 2011 6-191 DRAFT - DO NOT CITE OR QUOTE
-------
Lee, I. M.; Tsai, S. S.; Ho, C. K.; Chiu, H. R; Wu, T. N.; Yang, C. Y. (2008). Air pollution and hospital admissions for
congestive heart failure: are there potentially sensitive groups? Environ Res, 108: 348-353.
http://dx.doi.0rg/10.1016/i.envres.2008.07.024 192076
Lee, I. M.; Tsai, S. S.; Ho, C. K.; Chiu, H. F.; Yang, C. Y. (2007). Air pollution and hospital admissions for congestive heart
failure in a tropical city: Kaohsiung, Taiwan. Inhal Toxicol, 19: 899-904.
http://dx.doi.org/10.1080/08958370701479406 196613
Lee, J. T; Kim, H.; Cho, Y. S.; Hong, Y. C.; Ha, E. H.; Park, H. (2003). Air pollution and hospital admissions for ischemic
heart diseases among individuals 64+ years of age residing in Seoul, Korea. Arch Environ Health, 58: 617-623.
095552
Lee, Y.-L.; Lin, Y.-C.; Lee, Y.-C.; Wang, J.-Y; Hsiue, T.-R.; Guo, Y. L. (2004). Glutathione S-transferase PI gene
polymorphism and air pollution as interactive risk factors for childhood asthma. Clin Exp Allergy, 34: 1707-1713.
090971
Levy, J. I.; Chemerynski, S. M.; Sarnat, J. A. (2005). Ozone exposure and mortality, an empiric Bayes metaregression
analysis. Epidemiology, 16: 458-468. 074347
Lewis, T. C.; Robins, T. G; Dvonch, J. T; Keeler, G J.; Yip, F. Y; Mentz, G B.; Lin, X.; Parker, E. A.; Israel, B. A.;
Gonzalez, L.; Hill, Y. (2005). Air pollution-associated changes in lung function among asthmatic children in
Detroit. Environ Health Perspect, 113: 1068-1075. 081079
Liao, D.; Duan, Y; Whitsel, E. A.; Zheng, Z.-l; Heiss, G; Chinchilli, V. M.; Lin, H.-M. (2004). Association of higher
levels of ambient criteria pollutants with impaired cardiac autonomic control: a population-based study. Am J
Epidemiol, 159: 768-777. 056590
Liao, D.; Heiss, G; Chinchilli, V. M.; Duan, Y; Folsom, A. R.; Lin, H. M.; Salomaa, V. (2005). Association of criteria
pollutants with plasma hemostatic/inflammatory markers: a population-based study. J Expo Sci Environ Epidemiol,
15: 319-328. 088677
Lim, Y; Phung, A. D.; Corbacho, A. M.; Aung, H. H.; Maioli, E.; Reznick, A. Z.; Cross, C. E.; Davis, P. A.; Valacchi, G.
(2006). Modulation of cutaneous wound healing by ozone: Differences between young and aged mice. Toxicol Lett,
160: 127-134. http://dx.doi.Org/10.1016/j.toxlet.2005.06.013 670834
Lin, S.; Bell, E. M.; Liu, W.; Walker, R. J.; Kim, N. K.; Hwang, S. A. (2008). Ambient ozone concentration and hospital
admissions due to childhood respiratory diseases in New York State, 1991-2001. Environ Res, 108: 42-47.
http://dx.doi.0rg/10.1016/i.envres.2008.06.007 195856
Linares, C.; Diaz, J. (2010). Short-term effect of concentrations of fine particulate matter on hospital admissions due to
cardiovascular and respiratory causes among the over-75 age group in Madrid, Spain. Public Health, 124: 28-36.
http://dx.doi.0rg/10.1016/i.puhe.2009.ll.007383413
Linn, W. S.; Avol, E. L.; Shamoo, D. A.; Spier, C. E.; Valencia, L. M.; Venet, T. G; Fischer, D. A.; Hackney, J. D. (1986). A
dose-response study of healthy, heavily exercising men exposed to ozone at concentrations near the ambient air
quality standard. Toxicol Ind Health, 2: 99-112. 040481
Linn, W. S.; Fischer, D. A.; Medway, D. A.; Anzar, U. T; Spier, C. E.; Valencia, L. M.; Venet, T. G; Hackney, J. D. (1982).
Short-term respiratory effects of 0.12 ppm ozone exposure in volunteers with chronic obstructive pulmonary
disease. Am Rev Respir Dis, 125: 658-663. 039645
Linn, W. S.; Medway, D. A.; Anzar, U. T; Valencia, L. M.; Spier, C. E.; Tsao FS-D; Fischer, D. A.; Hackney, J. D. (1982).
Persistence of adaptation to ozone in volunteers exposed repeatedly for six weeks. Am Rev Respir Dis, 125: 491-
495. 039646
Linn, W. S.; Shamoo, D. A.; Anderson, K. R.; Peng, R.-C.; Avol, E. L.; Hackney, J. D.; Gong, H. Jr (1996). Short-term air
pollution exposures and responses in Los Angeles area schoolchildren. J Expo Sci Environ Epidemiol, 6: 449-472.
082508
Linn, W. S.; Shamoo, D. A.; Venet, T. G; Spier, C. E.; Valencia, L. M.; Anzar, U. T; Hackney, J. D. (1983). Response to
ozone in volunteers with chronic obstructive pulmonary disease. Arch Environ Occup Health, 38: 278-283. 040672
Lisabeth, L. D.; Escobar, J. D.; Dvonch, J. T; Sanchez, B. N.; Majersik, J. J.; Brown, D. L.; Smith, M. A.; Morgenstern, L.
B. (2008). Ambient air pollution and risk for ischemic stroke and transient ischemic attack. Ann Neurol, 64: 53-59.
http://dx.doi.org/10.1002/ana.21403 155939
March 2011 6-192 DRAFT - DO NOT CITE OR QUOTE
-------
Liu, L.; Leech, J. A.; Urch, R. B.; Poon, R.; Zimmerman, B.; Kubay, J. M.; Silverman, F. S. (1999). A comparison of
biomarkers of ozone exposure in human plasma, nasal lavage, and sputum. Inhal Toxicol, 11: 657-674. 012049
Liu, L.; Leech, J. A.; Urch, R. B.; Silverman, F. S. (1997). In vivo salicylate hyroxylation: Apotential biomarker for
assessing acute ozone exposure and effects in humans. Am J Respir Crit Care Med, 156: 1405-1412. 084627
Liu, L.; Poon, R.; Chen, L.; Frescura, A. M.; Montuschi, P.; Ciabattoni, G; Wheeler, A.; Dales, R. (2009). Acute effects of
air pollution on pulmonary function, airway inflammation, and oxidative stress in asthmatic children. Environ
Health Perspect, 117: 668-674. http://dx.doi.org/10.1289/ehpll813 192003
Lu, F. L.; Johnston, R. A.; Flynt, L.; Theman, T. A.; Terry, R. D.; Schwartzman, I. N.; Lee, A.; Shore, S. A. (2006).
Increased pulmonary responses to acute ozone exposure in obese db/db mice. Am J Physiol Lung Cell Mol Physiol,
290: L856-L865. http://dx.doi.org/10.1152/ajplung.00386.2005 597955
Mann, J. K.; Balmes, J. R.; Bruckner, T. A.; Mortimer, K. M.; Margolis, H. G; Pratt, B.; Hammond, S. K.; Lurmann, F.;
Tager, I. B. (2010). Short-term effects of air pollution on wheeze in asthmatic children in Fresno, California.
Environ Health Perspect, 118: 1497-1502. http://dx.doi.org/10.1289/ehp.0901292635827
Manzer, R.; Wang, J.; Nishina, K.; McConville, G; Mason, R. J. (2006). Alveolar epithelial cells secrete chemokines in
response to IL-lbeta and lipopolysaccharide but not to ozone. Am J Respir Cell Mol Biol, 34: 158-166.
http://dx.doi.org/10.1165/rcmb.2005-0205OC596404
Mapp, C. E.; Fryer, A. A.; De Marzo, N.; Pozzato, V.; Padoan, M.; Boschetto, P.; Strange, R. C.; Hemmingsen, A.; Spiteri,
M. A. (2002). Glutathione S-transferase GSTP1 is a susceptibility gene for occupational asthma induced by
isocyanates. J Allergy Clin Immunol, 109: 867-872. http://dx.doi.org/10.1067/mai.2002.123234670826
Martinez-Canabal, A.; Angora-Perez, M. (2008). Effect of growth hormone on cyclooxygenase-2 expression in the
hippocampus of rats chronically exposed to ozone. Int J Neurosci, 118: 455-469.
http://dx.doi.org/10.1080/00207450701593160 194376
McBride, D. E.; Koenig, J. Q.; Luchtel, D. L.; Williams, P. V; Henderson, W R. Jr (1994). Inflammatory effects of ozone
in the upper airways of subjects with asthma. Am J Respir Crit Care Med, 149: 1192-1197. 043912
McDonnell WF 3rd; Horstman, D. H.; Abdul-Salaam, S.; House, D. E. (1985). Reproducibility of individual responses to
ozone exposure. Am Rev Respir Dis, 131: 36-40. 040283
McDonnell, W. F. (1996). Individual variability in human lung function responses to ozone exposure. Environ Toxicol
Pharmacol, 2: 171-175. 082679
McDonnell, W. F.; Horstman, D. H.; Hazucha, M. J.; Seal, E. Jr; Haak, E. D.; Salaam, S. A.; House, D. E. (1983).
Pulmonary effects of ozone exposure during exercise: Dose-response characteristics. J Appl Physiol, 54: 1345-
1352.040680
McDonnell, W. F.; Kehrl, H. R.; Abdul-Salaam, S.; Ives, P. J.; Folinsbee, L. J.; Devlin, R. B.; O'Neil, J. J.; Horstman, D. H.
(1991). Respiratory response of humans exposed to low levels of ozone for 66 hours. Arch Environ Occup Health,
46: 145-150. 042384
McDonnell, W. F.; Stewart, P. W.; Andreoni, S.; Seal, E. Jr; Kehrl, H. R.; Horstman, D. H.; Folinsbee, L. J.; Smith, M. V.
(1997). Prediction of ozone-induced FEVI changes: effects of concentration, duration, and ventilation. Am J Respir
Crit Care Med, 156: 715-722. 084266
McDonnell, W. F.; Stewart, P. W.; Smith, M. V. (2007). The temporal dynamics of ozone-induced FEV1 changes in
humans: An exposure-response model. Inhal Toxicol, 19: 483-494. 093104
McDonnell, W. F.; Stewart, P. W.; Smith, M. V. (2010). Prediction of ozone-induced lung function responses in humans.
Inhal Toxicol, 22: 160-168. http://dx.doi.org/10.1080/08958370903089557 383972
McDonnell, W. F.; Stewart, P. W.; Smith, M. V; Pan, W. K.; Pan, J. (1999). Ozone-induced respiratory symptoms:
Exposure-response models and association with lung function. Eur Respir J, 14: 845-853. 010939
Medina-Ramon, M.; Schwartz, J. (2008). Who is more vulnerable to die from ozone air pollution? Epidemiology, 19: 672-
679. 193829
Medina-Ramon, M.; Zanobetti, A.; Schwartz, J. (2006). The effect of ozone and PM10 on hospital admissions for
pneumonia and chronic obstructive pulmonary disease: A national multicity study. Am J Epidemiol, 163: 579-588.
http://dx.doi.org/10.1093/aje/kwj078 087721
March 2011 6-193 DRAFT - DO NOT CITE OR QUOTE
-------
Metzger, K. B.; Klein, M.; Flanders, W. D.; Peel, J. L.; Mulholland, J. A.; Langberg, J. I; Tolbert, P. E. (2007). Ambient air
pollution and cardiac arrhythmias in patients with implantable deiibrillators. Epidemiology, 18: 585-592.
http://dx.doi.org/10.1097/EDE.Ob013e318124froe092856
Metzger, K. B.; Tolbert, P. E.; Klein, M.; Peel, J. L.; Flanders, W. D.; Todd, K. H.; Mulholland, J. A.; Ryan, P. B.; Frumkin
H (2004). Ambient air pollution and cardiovascular emergency department visits. Epidemiology, 15: 46-56. 044222
Michelson, P. H.; Dailey, L.; Devlin, R. B.; Peden, D. B. (1999). Ozone effects on the immediate-phase response to allergen
in the nasal airways of allergic asthmatic subjects. Otolaryngol Head Neck Surg, 120: 225-232. 001147
Middleton, N.; Yiallouros, P.; Kleanthous, S.; Kolokotroni, O.; Schwartz, J.; Dockery, D. W.; Demokritou, P.; Koutrakis, P.
(2008). A 10-year time-series analysis of respiratory and cardiovascular morbidity in Nicosia, Cyprus: The effect of
short-term changes in air pollution and dust storms. Environ Health, 7: 39. 156760
Mikerov, A. N.; Gan, X.; Umstead, T M.; Miller, L.; Chinchilli, V. M.; Phelps, D. S.; Floras, J. (2008). Sex differences in
the impact of ozone on survival and alveolar macrophage function of mice after Klebsiella pneumoniae infection.
RespirRes, 9: 24. http://dx.doi.org/10.1186/1465-9921-9-24 597493
Mikerov, A. N.; Haque, R.; Gan, X.; Guo, X.; Phelps, D. S.; Floras, J. (2008). Ablation of SP-A has a negative impact on
the susceptibility of mice to Klebsiella pneumoniae infection after ozone exposure: Sex differences. Respir Res, 9:
77. http://dx.doi.org/10.1186/1465-9921-9-77 201537
Mikerov, A. N.; Umstead, T. M.; Gan, X.; Huang, W.; Guo, X.; Wang, G; Phelps, D. S.; Floras, J. (2008). Impact of ozone
exposure on the phagocytic activity of human surfactant protein A (SP-A) and SP-A variants. Am J Physiol Lung
Cell Mol Physiol, 294: L121-L130. http://dx.doi.org/10.1152/ajplung.00288.2007 596405
Mokoena, M. L.; Harvey, B. H.; Oliver, D. W.; Brink, C. B. (2010). Ozone modulates the effects of imipramine on
immobility in the forced swim test, and nonspecific parameters of hippocampal oxidative stress in the rat. Metab
Brain Dis, 25: 125-133. http://dx.doi.org/10.1007/sll011-010-9189-7 677667
Molfmo, N. A.; Wright, S. C.; Katz, L; Tarlo, S.; Silverman, R; McClean, P. A.; Szalai, J. P.; Raizenne, M.; Slutsky, A. S.;
Zamel, N. (1991). Effect of low concentrations of ozone on inhaled allergen responses in asthmatic subjects.
Lancet, 338: 199-203. 042379
Moon, J. S.; Kim, Y. S.; Kim, J. H.; Son, B. S.; Kim, D. S.; Yang, W. (2009). Respiratory health effects among
schoolchildren and their relationship to air pollutants in Korea. Int J Environ Health Res, 19: 31-48.
http://dx.doi.org/10.1080/09603120802272201 190297
Morrow, J. D.; Hill, K. E.; Burk, R. F.; Nammour, T. M.; Badr, K. F.; Roberts LJ 2nd (1990). A series of prostaglandin F2-
like compounds are produced in vivo in humans by a non-cyclooxygenase, free radical-catalyzed mechanism.
PNAS, 87: 9383-9387. http://www.pnas.org/content/87/23/9383.abstract. 625835
Mortimer, K. M.; Neas, L. M.; Dockery, D. W.; Redline, S.; Tager, I. B. (2002). The effect of air pollution on inner-city
children with asthma. Eur Respir J, 19: 699-705. http://dx.doi.org/10.1183/09031936.02.00247102030281
Mortimer, K. M.; Tager, I. B.; Dockery, D. W.; Neas, L. M.; Redline, S. (2000). The effect of ozone on inner-city children
with asthma: Identification of susceptible subgroups. Am J Respir Crit Care Med, 162: 1838-1845. 013255
Mudway, I. S.; Blomberg, A.; Frew, A. J.; Holgate, S. T; Sandstrom, T; Kelly, F. J. (1999). Antioxidant consumption and
repletion kinetics in nasal lavage fluid following exposure of healthy human volunteers to ozone. Eur Respir J, 13:
1429-1438. 001270
Mudway, I. S.; Kelly, F. J. (2000). Ozone and the lung: a sensitive issue. Mol Aspects Med, 21: 1-48. 010452
Mudway, I. S.; Kelly, F. J. (2004). An investigation of inhaled ozone dose and the magnitude of airway inflammation in
healthy adults. Am J Respir Crit Care Med, 169: 1089-1095. 057299
Mudway, I. S.; Stenfors, N.; Blomberg, A.; Helleday, R.; Dunster, C.; Marklund, S. L.; Frew, A. J.; Sandstrom, T; Kelly, F.
J. (2001). Differences in basal airway antioxidant concentrations are not predictive of individual responsiveness to
ozone: A comparison of healthy and mild asthmatic subjects. Free Radic Biol Med, 31: 962-974. 025327
Murugan, A.; Prys-Picard, C.; Calhoun, W. J. (2009). Biomarkers in asthma. Curr Opin Pulm Med, 15: 12-18.
http://dx.doi.org/10.1097/MCP.Ob013e32831de235625839
Naeher, L. P.; Holford, T. R.; Beckett, W. S.; Belanger, K; Triche, E. W.; Bracken, M. B.; Leaderer, B. P. (1999). Healthy
women's PEF variations with ambient summer concentrations of PM10, PM2.5, SO42-, H+, and O3. Am J Respir
Crit Care Med, 160: 117-125. 033568
March 2011 6-194 DRAFT - DO NOT CITE OR QUOTE
-------
Nakamura, K.; Matsunaga, K. (1998). Susceptibility of natural killer (NK) cells to reactive oxygen species (ROS) and their
restoration by the mimics of superoxide dismutase (SOD). Cancer Biother Radiopharm, 13: 275-290.
http://www.liebertonline.com/doi/abs/10.1089/cbr.1998.13.275. 625133
Neas, L. M.; Dockery, D. W.; Koutrakis, P.; Speizer, F. E. (1999). Fine particles and peak flow in children: Acidity versus
mass. Epidemiology, 10: 550-553. 003466
Neuberger, M.; Schimek, M. G; Horak, F. Jr; Moshammer, H.; Kundi, M.; Frischer, T; Gomiscek, B.; Puxbaum, H.;
Hauck, H.; AUPHEP-Team (2004). Acute effects of particulate matter on respiratory diseases, symptoms and
functions: epidemiological results of the Austrian Projects on Health Effects of Particulate Matter (AUPHEP).
Atmos Environ, 38: 3971-3981. 093249
Newson, E. I; Krishna, M. T.; Lau, L. C. K.; Howarth, P. H.; Holgate, S. T.; Frew, A. J. (2000). Effects of short-term
exposure to 0.2 ppm ozone on biomarkers of inflammation in sputum, exhaled nitric oxide, and lung function in
subjects with mild atopic asthma. J Occup Environ Med, 42: 270-277. 000853
Nickmilder, M.; De Burbure, C.; Sylviane, C.; Xavier, D.; Alfred, B.; Alain, D. (2007). Increase of exhaled nitric oxide in
children exposed to low levels of ambient ozone. J Toxicol Environ Health A, 70: 270-274. 090710
Nightingale, J. A.; Rogers, D. F.; Chung, K. F.; Barnes, P. J. (2000). No effect of inhaled budesonide on the response to
inhaled ozone in normal subjects. Am J Respir Crit Care Med, 161: 479-486. 000796
O'Connor, G. T.; Neas, L.; Vaughn, B.; Kattan, M.; Mitchell, H.; Crain, E. F.; Evans R 3rd; Gruchalla, R.; Morgan, W.;
Stout, J.; Adams, G. K.; Lippmann, M. (2008). Acute respiratory health effects of air pollution on children with
asthma in US inner cities. J Allergy Clin Immunol, 121: 1133-1139. http://dx.doi.Org/10.1016/j.jaci.2008.02.020
156818
Orazzo, F.; Nespoli, L.; Ito, K.; Tassinari, D.; Giardina, D.; Funis, M.; Cecchi, A.; Trapani, C.; Forgeschi, G; Vignini, M.;
Nosetti, L.; Pigna, S.; Zanobetti, A. (2009). Air pollution, aeroallergens, and emergency room visits for acute
respiratory diseases and gastroenteric disorders among young children in six Italian cities. Environ Health Perspect,
117: 1780-1785. http://dx.doi.org/10.1289/ehp.0900599202801
Oslund, K. L.; Hyde, D. M.; Putney, L. F.; Alfaro, M. F.; Walby, W. F.; Tyler, N. K.; Schelegle, E. S. (2008). Activation of
neurokinin-1 receptors during ozone inhalation contributes to epithelial injury and repair. Am J Respir Cell Mol
Biol, 39: 279-288. http://dx.doi.org/10.1165/rcmb.2008-0009OC 195654
Oslund, K. L.; Hyde, D. M.; Putney, L. F.; Alfaro, M. F.; Walby, W. F.; Tyler, N. K.; Schelegle, E. S. (2009). Activation of
calcitonin gene-related peptide receptor during ozone inhalation contributes to airway epithelial injury and repair.
Toxicol Pathol, 37: 805-813. http://dx.doi.org/10.1177/0192623309345691 201539
Ostro, B.; Broadwin, R.; Green, S.; Feng, W.-Y.; Lipsett, M. (2006). Fine particulate air pollution and mortality in nine
California counties: results from CALFINE. Environ Health Perspect, 114: 29-33. 087991
Ostro, B.; Lipsett, M.; Mann, J.; Braxton-Owens, H.; White, M. (2001). Air pollution and exacerbation of asthma in
African-American children in Los Angeles. Epidemiology, 12: 200-208. 016702
Oudin, A.; Stromberg, U.; Jakobsson, K.; Stroh, E.; Bjork, J. (2010). Estimation of short-term effects of air pollution on
stroke hospital admissions in southern Sweden. Neuroepidemiology, 34: 131-142.
http://dx.doi.org/10.1159/000274807384790
Oyarzun, M.; Dussaubat, N.; Gonzalez, S. (2005). Effect of 0.25 ppm ozone exposure on pulmonary damage induced by
bleomycin. Biol Res, 38: 353-358. 596407
Park, J. W.; Lim, Y. H.; Kyung, S. Y; An, C. H.; Lee, S. P.; Jeong, S. H; Ju, S.-Y (2005). Effects of ambient particulate
matter on peak expiratory flow rates and respiratory symptoms of asthmatics during Asian dust periods in Korea.
Respirology, 10: 470-476. http://dx.doi.Org/10.llll/i.1440-1843.2005.00728.x088673
Park, S. K.; O'Neill, M. S.; Stunder, B. J. B.; Vokonas, P. S.; Sparrow, D.; Koutrakis, P.; Schwartz, J. (2007). Source
location of air pollution and cardiac autonomic function: Trajectory cluster analysis for exposure assessment. J
Expo Sci Environ Epidemiol, 17: 488-497. 093268
Park, S. K.; O'Neill, M. S.; Vokonas, P. S.; Sparrow, D.; Schwartz, J. (2005). Effects of air pollution on heart rate
variability: The VA normative aging study. Environ Health Perspect, 113: 304-309. 057331
March 2011 6-195 DRAFT - DO NOT CITE OR QUOTE
-------
Park, S. K.; O'Neill, M. S.; Vokonas, P. S.; Sparrow, D.; Wright, R. O.; Coull, B.; Nie, H.; Hu, H.; Schwartz, J. (2008). Air
pollution and heart rate variability: Effect modification by chronic lead exposure. Epidemiology, 19: 111-120.
093027
Passannante, A. N.; Hazucha, M. J.; Bromberg, P. A.; Seal, E.; Folinsbee, L.; Koch, G. (1998). Nociceptive mechanisms
modulate ozone-induced human lung function decrements. J Appl Physiol, 85: 1863-1870. 030114
Peden, D. B. (2001). Air pollution in asthma: Effect of pollutants on airway inflammation. Ann Allergy Asthma Immunol,
3: 12-17. 025355
Peden, D. B.; Boehlecke, B.; Horstman, D.; Devlin, R. (1997). Prolonged acute exposure to 0.16 ppm ozone induces
eosinophilic airway inflammation in asthmatic subjects with allergies. J Allergy Clin Immunol, 100: 802-808.
085842
Peden, D. B.; Setzer, R. W. Jr; Devlin, R. B. (1995). Ozone exposure has both a priming effect on allergen-induced
responses and an intrinsic inflammatory action in the nasal airways of perennially allergic asthmatics. Am J Respir
CritCareMed, 151: 1336-1345. 076189
Peel, J. L.; Metzger, K. B.; Klein, M.; Flanders, W. D.; Mulholland, J. A.; Tolbert, P. E. (2007). Ambient air pollution and
cardiovascular emergency department visits in potentially sensitive groups. Am J Epidemiol, 165: 625-633. 090442
Peel, J. L.; Tolbert, P. E.; Klein, M.; Metzger, K. B.; Flanders, W. D.; Knox, T.; Mulholland, J. A.; Ryan, P. B.; Frumkin, H.
(2005). Ambient air pollution and respiratory emergency department visits. Epidemiology, 16: 164-174. 056305
Pellegrino, R.; Viegi, G; Brusasco, V; Crapo, R. O.; Burgos, R; Casaburi, R.; Coates, A.; van der Grinten, C. P.;
Gustafsson, P.; Hankinson, J.; Jensen, R.; Johnson, D. C.; Maclntyre, N.; McKay, R.; Miller, M. R.; Navajas, D.;
Pedersen, O. R; Wanger, J. (2005). Interpretative strategies for lung function tests. Eur Respir J, 26: 948-968.
http://dx.doi.org/10.1183/09031936.05.00035205626521
Peng, R. D.; Dominici, R; Pastor-Barriuso, R.; Zeger, S. L.; Samet, J. M. (2005). Seasonal analyses of air pollution and
mortality in 100 US cities. Am J Epidemiol, 161: 585-594. 087463
Perepu, R. S.; Garcia, C.; Dostal, D.; Sethi, R. (2010). Enhanced death signaling in ozone-exposed ischemic-reperfused
hearts. Mol Cell Biochem, 336: 55-64. http://dx.doi.org/10.1007/sll010-009-0265-4 385020
Pereyra-Munoz, N.; Rugerio-Vargas, C.; Angoa-Perez, M.; Borgonio-Perez, G; Rivas-Arancibia, S. (2006). Oxidative
damage in substantia nigra and striatum of rats chronically exposed to ozone. J Chem Neuroanat, 31: 114-123.
http://dx.doi.0rg/10.1016/i.ichemneu.2005.09.006 596408
Peters, A.; Dockery, D. W.; Muller, J. E.; Mittleman, M. A. (2001). Increased particulate air pollution and the triggering of
myocardial infarction. Circulation, 103: 2810-2815. 016546
Petroeschevsky, A.; Simpson, R. W.; Thalib, L.; Rutherford, S. (2001). Associations between outdoor air pollution and
hospital admissions in Brisbane, Australia. Arch Environ Occup Health, 56: 37-52. 016466
Pichavant, M.; Goya, S.; Meyer, E. H.; Johnston, R. A.; Kim, H. Y; Matangkasombut, P.; Zhu, M.; Iwakura, Y; Savage, P.
B.; DeKruyff, R. H.; Shore, S. A.; Umetsu, D. T. (2008). Ozone exposure in a mouse model induces airway
hyperreactivity that requires the presence of natural killer T cells and IL-17. J Exp Med, 205: 385-393.
http://dx.doi.org/10.1084/iem.20071507596409
Pinkerton, K. E.; Brody, A. R.; Miller, F. J.; Crapo, J. D. (1989). Exposure to low levels of ozone results in enhanced
pulmonary retention of inhaled asbestos fibers. Am J Respir Crit Care Med, 140: 1075-1081. 042102
Plopper, C. G; Mango, G. W.; Hatch, G. E.; Wong, V. J.; Toskala, E.; Reynolds, S. D.; Tarkington, B. K.; Stripp, B. R.
(2006). Elevation of susceptibility to ozone-induced acute tracheobronchial injury in transgenic mice deficient in
Clara cell secretory protein. Toxicol Appl Pharmacol, 213: 74-85. http://dx.doi.Org/10.1016/j.taap.2005.09.003
596410
Plopper, C. G; Smiley-Jewell, S. M.; Miller, L. A.; Fanucchi, M. V; Evans, M. J.; Buckpitt, A. R.; Avdalovic, M.;
Gershwin, L. J.; Joad, J. P.; Kajekar, R.; Larson, S.; Pinkerton, K. E.; Van Winkle, L. S.; Schelegle, E. S.; Pieczarka,
E. M.; Wu, R.; Hyde, D. M. (2007). Asthma/allergic airways disease: Does postnatal exposure to environmental
toxicants promote airway pathobiology? Toxicol Pathol, 35: 97-110. http://dx.doi.org/10.1080/01926230601132030
596412
Poloniecki, J. D.; Atkinson, R. W.; Ponce de Leon, A.; Anderson, H. R. (1997). Daily time series for cardiovascular hospital
admissions and previous day's air pollution in London, UK. Occup Environ Med, 54: 535-540. 084004
March 2011 6-196 DRAFT - DO NOT CITE OR QUOTE
-------
Prescott, G. I; Cohen, G. R.; Elton, R. A.; Fowkes, F. G R.; Agius, R. M. (1998). Urban air pollution and cardiopulmonary
ill health: A 145 year time series study. Occup Environ Med, 55: 697-704. 084610
Pulfer, M. K.; Murphy, R. C. (2004). Formation of biologically active oxysterols during ozonolysis of cholesterol present in
lung surfactant. J Biol Chem, 279: 26331-26338. 076673
Pulfer, M. K.; Taube, C.; Gelfand, E.; Murphy, R. C. (2005). Ozone exposure in vivo and formation of biologically active
oxysterols in the lung. J Pharmacol Exp Ther, 312: 256-264. 076663
Qian, Z.; Lin, H. M.; Chinchilli, V. M.; Lehman, E. B.; Duan, Y; Craig, T. J.; Wilson, W. E.; Liao, D.; Lazarus, S. C.;
Bascom, R. (2009). Interaction of ambient air pollution with asthma medication on exhaled nitric oxide among
asthmatics. Arch Environ Occup Health, 64: 168-176. http://dx.doi.org/10.1080/19338240903240616 548793
Rabinovitch, N.; Zhang, L. N.; Murphy, J. R.; Vedal, S.; Button, S. J.; Gelfand, E. W. (2004). Effects of wintertime ambient
air pollutants on asthma exacerbations in urban minority children with moderate to severe disease. J Allergy Clin
Immunol, 114: 1131-1137. http://dx.doi.Org/10.1016/j.jaci.2004.08.026 096753
Raizenne, M. E.; Burnett, R. T.; Stern, B.; Franklin, C. A.; Spengler, J. D. (1989). Acute lung function responses to ambient
acid aerosol exposures in children. Environ Health Perspect, 79: 179-185. 041700
Raizenne, M.; Stern, B.; Burnett, R.; Spengler, J. (1987). Acute respiratory function and transported air pollutants:
Observational studies (paper no. 87-32.6). In Proceedings of the 80th Annual Meeting of the Air Pollution Control
Association (p. 18). Pittsburgh, PA: Air Pollution Control Association. 040903
Ren, C.; Williams, G. M.; Mengersen, K.; Morawska, L.; long, S. (2008). Does temperature modify short-term effects of
ozone on total mortality in 60 large eastern US communities? An assessment using the NMMAPS data. Environ Int,
34: 451-458. 093281
Revis, N. W.; Major, T.; Dalbey, W. E. (1981). Cardiovascular effects of ozone and cadmium inhalation in the rat. Research
Triangle Park, NC: EPA. 040017
Rich, D. Q.; Kim, M. H.; Turner, J. R.; Mittleman, M. A.; Schwartz, J.; Catalano, P. J.; Dockery, D. W. (2006). Association
of ventricular arrhythmias detected by implantable cardioverier defibrillator and ambient air pollutants in the St
Louis, Missouri metropolitan area. Occup Environ Med, 63: 591-596. 089814
Rich, D. Q.; Mittleman, M. A.; Link, M. S.; Schwartz, J.; Luttmann-Gibson, H.; Catalano, P. J.; Speizer, F. E.; Gold, D. R.;
Dockery, D. W. (2006). Increased risk of paroxysmal atrial fibrillation episodes associated with acute increases in
ambient air pollution. Environ Health Perspect, 114: 120-123. 088427
Rich, D. Q.; Schwartz, J.; Mittleman, M. A.; Link, M.; Luttmann-Gibson, H.; Catalano, P. J.; Speizer, F. E.; Dockery, D. W.
(2005). Association of short-term ambient air pollution concentrations and ventricular arrhythmias. Am J
Epidemiol, 161: 1123-1132. 079620
Riediker, M.; Monn, C.; Roller, T; Stahel, W. A.; Wuthrich, B. (2001). Air pollutants enhance rhinoconjunctivitis
symptoms in pollen-allergic individuals. Ann Allergy Asthma Immunol, 87: 311-318. 051776
Rivas-Arancibia, S.; Guevara-Guzman, R.; Lopez-Vidal, Y; Rodriguez-Martinez, E.; Gomes, M. Z.; Angoa-Perez, M.;
Raisman-Voz ari, R. (2010). Oxidative stress caused by ozone exposure induces loss of brain repair in the
hippocampus of adult rats. Toxicol Sci, 113: 187-197. http://dx.doi.org/10.1093/toxsci/kfp252 201544
Rodriguez, C.; Tonkin, R.; Heyworth, J.; Kusel, M.; De Klerk, N.; Sly, P. D.; Franklin, P.; Runnion, T; Blockley, A.;
Landau, L.; Hinwood, A. L. (2007). The relationship between outdoor air quality and respiratory symptoms in
young children. Int J Environ Health Res, 17: 351-360 . http://dx.doi.org/10.1080/09603120701628669 092842
Romieu, I.; Barraza-Villarreal, A.; Escamilla-Nunez, C.; Texcalac-Sangrador, J. L.; Hernandez-Cadena, L.; Diaz-Sanchez,
D.; De Batlle, J.; Del Rio-Navarro, B. E. (2009). Dietary intake, lung function and airway inflammation in Mexico
City school children exposed to air pollutants. Respir Res, 10: 122. 548788
Romieu, I.; Barraza-Villarreal, A.; Escamilla-Nunez, C.;Almstrand, A. C.; Diaz-Sanchez, D.; Sly, P. D.; Olin, A. C. (2008).
Exhaled breath malondialdehyde as a marker of effect of exposure to air pollution in children with asthma. J
Allergy Clin Immunol, 121: 903-909. http://dx.doi.Org/10.1016/j.jaci.2007.12.004 179908
Romieu, I.; Meneses, F.; Ramirez, M.; Ruiz, S.; Padilla, R. P.; Sienra, J. J.; Gerber, M.; Grievink, L.; Dekker, R.; Walda, L;
Brunekreef, B. (1998). Antioxidant supplementation and respiratory functions among workers exposed to high
levels of ozone. Am J Respir Crit Care Med, 158: 226-232. 086756
March 2011 6-197 DRAFT - DO NOT CITE OR QUOTE
-------
Romieu, I.; Meneses, R; Ruiz, S.; Huerta, J.; Sienra, J. I; White, M.; Etzel, R.; Hernandez, M. (1997). Effects of
intermittent ozone exposure on peak expiratory flow and respiratory symptoms among asthmatic children in
Mexico City. Arch Environ Occup Health, 52: 368-376. 085807
Romieu, I.; Meneses, R; Ruiz, S.; Sienra, J. J.; Huerta, J.; White, M. C.; Etzel, R. A. (1996). Effects of air pollution on the
respiratory health of asthmatic children living in Mexico City. Am J Respir Crit Care Med, 154: 300-307. 080748
Romieu, I.; Ramirez-Aguilar, M.; Sienra-Monge, J. J.; Moreno-Macias, H.; Del Rio-Navarro, B. E.; David, G; Marzec, J.;
Hernandez-Avila, M.; London, S. (2006). GSTM1 and GSTP1 and respiratory health in asthmatic children exposed
to ozone. Eur Respir J, 28: 953-959. http://dx.doi.org/10.1183/09031936.06.00114905 090969
Romieu, I.; Sienra-Monge, J. J.; Ramirez-Aguilar, M.; Moreno-Macias, H.; Reyes-Ruiz, N. I.; Estela del Rio-Navarro, B.;
Hernandez-Avila, M.; London, S. J. (2004). Genetic polymorphism of GSTM1 and antioxidant supplementation
influence lung function in relation to ozone exposure in asthmatic children in Mexico City. Thorax, 59: 8-10.
056796
Romieu, I.; Sienra-Monge, J. J.; Ramirez-Aguilar, M.; Tellez-Rojo, M. M.; Moreno-Macias, H.; Reyes-Ruiz, N. I.; Del
Rio-Navarro, B. E.; Ruiz-Navarro, M. X.; Hatch, G; Slade, R.; Hernandez-Avila, M. (2002). Antioxidant
supplementation and lung functions among children with asthma exposed to high levels of air pollutants. Am J
Respir Crit Care Med, 166: 703-709. 034711
Rosenfeld, M. A.; Leonova, V. B.; Konstantinova, M. L.; Razumovskii, S. D. (2009). Self-assembly of fibrin monomers
and fibrinogen aggregation during ozone oxidation. Biochemistry (Mosc), 74: 41-46.
http://dx.doi.org/10.1134/S0006297909010064201546
Ross, M. A.; Persky, V. W.; Scheff, P. A.; Chung, J.; Curtis, L.; Ramakrishnan, V; Wadden, R. A.; Hryhorczuk, D. O.
(2002). Effect of ozone and aeroallergens on the respiratory health of asthmatics. Arch Environ Occup Health, 57:
568-578. 042749
Rozenfeld, M. A.; Leonova, V. B.; Konstantinova, M. L.; Razumovskii, S. D.; Makarov, V. A.; Nevedrova, O. E.;
Belozerskaja, G. G. (2008). Disturbance of functional properties of fibrinogen under ozone oxidation. Dokl
Biochem Biophys, 422: 315-318. http://dx.doi.org/10.1134/S1607672908050165 596413
Rudez, G; Janssen, N. A.; Kilinc, E.; Leebeek, R W; Gerlofs-Nijland, M. E.; Spronk, H. M.; ten Cate, H.; Cassee, R R.; de
Maat, M. P. (2009). Effects of ambient air pollution on hemostasis and inflammation. Environ Health Perspect, 117:
995-1001. 193783
Ruidavets, J.-B.; Cassadou, S.; Cournot, M.; Bataille, V; Meybeck, M.; Ferrieres, J. (2005). Increased resting heart rate
with pollutants in a population based study. J Epidemiol Community Health, 59: 685-693. 089443
Ruidavets, J.-B.; Cournot, M.; Cassadou, S.; Giroux, M.; Meybeck, M.; Ferrieres, J. (2005). Ozone air pollution is
associated with acute myocardial infarction. Circulation, 111: 563-569. 074091
Samet, J. M.; Hatch, G. E.; Horstman, D.; Steck-Scott, S.; Arab, L.; Bromberg, P. A.; Levine, M.; McDonnell, W. F.;
Devlin, R. B. (2001). Effect of antioxidant supplementation on ozone-induced lung injury in human subjects. Am J
Respir Crit Care Med, 164: 819-825. 019034
Samet, J. M.; Zeger, S. L.; Dominici, R; Curriero, R; Coursac, L; Dockery, D. W.; Schwartz, J.; Zanobetti, A. (2000). The
national morbidity, mortality, and air pollution study. Part II: Morbidity, mortality, and air pollution in the United
States. Cambridge, MA: Health Effects Institute. 010269
Samoli, E.; Zanobetti, A.; Schwartz, J.; Atkinson, R.; Le Tertre, A.; Schindler, C.; Perez, L.; Cadum, E.; Pekkanen, J.;
Paldy, A.; Touloumi, G; Katsouyanni, K. (2009). The temporal pattern of mortality responses to ambient ozone in
the APHEA project. J Epidemiol Community Health, 63: 960-966. http://dx.doi.org/10.1136/jech.2008.084012
195855
Santucci, D.; Sorace, A.; Francia, N.; Aloe, L.; Alleva, E. (2006). Prolonged prenatal exposure to low-level ozone affects
aggressive behaviour as well as NGF and BDNF levels in the central nervous system of CD-I mice. Behav Brain
Res, 166: 124-130. http://dx.doi.Org/10.1016/j.bbr.2005.07.032 596414
Sarnat, S. E.; Suh, H. H.; Coull, B. A.; Schwartz, J.; Stone, P. H.; Gold, D. R. (2006). Ambient particulate air pollution and
cardiac arrhythmia in a panel of older adults in Steubenville, Ohio. Occup Environ Med, 63: 700-706. 090489
March 2011 6-198 DRAFT - DO NOT CITE OR QUOTE
-------
Sathishkumar, K.; Gao, X.; Raghavamenon, A. C.; Parinandi, N.; Pryor, W. A.; Uppu, R. M. (2009). Cholesterol
secoaldehyde induces apoptosis in H9c2 cardiomyoblasts through reactive oxygen species involving mitochondrial
and death receptor pathways. Free Radic Biol Med, 47: 548-558.
http://dx.doi.0rg/10.1016/i.freeradbiomed.2009.05.020201549
Sathishkumar, K.; Haque, M.; Perumal, T. E.; Francis, I; Uppu, R. M. (2005). A major ozonation product of cholesterol,
3beta-hydroxy-5-oxo-5,6-secocholestan-6-al, induces apoptosis in H9c2 cardiomyoblasts. FEES Lett, 579: 6444-
6450. 087958
Scannell, C.; Chen, L.; Aris, R. M.; Tager, I.; Christian, D.; Ferrando, R.; Welch, B.; Kelly, T.; Balmes, J. R. (1996).
Greater ozone-induced inflammatory responses in subjects with asthma. Am J Respir Crit Care Med, 154: 24-29.
080755
Scarlett, J. F.; Abbott, K. J.; Peacock, J. L.; Strachan, D. P.; Anderson, H. R. (1996). Acute effects of summer air pollution
on respiratory function in primary school children in southern England. Thorax, 51: 1109-1114. 081158
Schelegle, E. S.; Adams, W. C. (1986). Reduced exercise time in competitive simulations consequent to low level ozone
exposure. Med Sci Sports Exerc, 18: 408-414. 040351
Schelegle, E. S.; Morales, C. A.; Walby, W. F.; Marion, S.; Allen, R. P. (2009). 6.6-hour inhalation of ozone concentrations
from 60 to 87 parts per billion in healthy humans. Am J Respir Crit Care Med, 180: 265-272.
http://dx.doi.org/10.1164/rccm.200809-1484OC618629
Schelegle, E. S.; Siefkin,A. D.; McDonald, R. J. (1991). Time course of ozone-induced neutrophilia in normal humans. Am
J Respir Crit Care Med, 143: 1353-1358. 042491
Schildcrout, J. S.; Sheppard, L.; Lumley, T; Slaughter, J. C.; Koenig, J. Q.; Shapiro, G G (2006). Ambient air pollution
and asthma exacerbations in children: An eight-city analysis. Am J Epidemiol, 164: 505-517.
http://dx.doi.org/10.1093/aje/kwj225 089812
Schmekel, B.; Ahlner, J.; Malmstrom, M.; Venge, P. (2001). Eosinophil cationic protein (ECP) in saliva: A new marker of
disease activity in bronchial asthma. Respir Med, 98: 670-675. http://dx.doi.org/10.1053/rmed.2001.1123 625849
Schwartz, J. (2005). How sensitive is the association between ozone and daily deaths to control for temperature? Am J
Respir Crit Care Med, 171: 627-631.057333
Schwartz, J. (2005). Who is sensitive to extremes of temperature? A case-only analysis. Epidemiology, 16: 67-72.
http://dx.doi.org/10.1097/01.ede.0000147114.25957.71 667864
Schwartz, J.; Litonjua, A.; Suh, H.; Verrier, M.; Zanobetti, A.; Syring, M.; Nearing, B.; Verrier, R.; Stone, P.; MacCallum,
G; Speizer, F. E.; Gold, D. R. (2005). Traffic related pollution and heart rate variability in a panel of elderly
subjects. Thorax, 60: 455-461. 074317
Seal, E. Jr; McDonnell, W. F.; House, D. E. (1996). Effects of age, socioeconomic status, and menstrual cycle on
pulmonary response to ozone. Arch Environ Occup Health, 51:132-137. 044251
Seal, E. Jr; McDonnell, W. F.; House, D. E.; Salaam, S. A.; Dewitt, P. J.; Butler, S. O.; Green, J.; Raggio, L. (1993). The
pulmonary response of white and black adults to six concentrations of ozone. Am J Respir Crit Care Med, 147: 804-
810.039357
Selgrade, M. K.; Daniels, M. J.; Grose, E. C. (1990). Acute, subchronic, and chronic exposure to a simulated urban profile
of ozone: Effects on extrapulmonary natural killer cell activity and lymphocyte mitogenic responses. Inhal Toxicol,
2: 375-389. 042363
Selwyn, B. J.; Stock, T. H.; Hardy, R. J.; Chan, F. A.; Jenkins, D. E.; Kotchmar, D. J.; Chapman, R. S. (1985). Health
effects of ambient ozone exposure in vigorously exercising adults. In Evaluation of the Scientific Basis for
Ozone/Oxidants Standards: Proceedings of an APCA International Specialty Conference (pp. 281-296). Pittsburgh,
PA: Air Pollution Control Association. 041356
Servais, S.; Boussouar, A.; Molnar, A.; Douki, T; Pequignot, J. M.; Favier, R. (2005). Age-related sensitivity to lung
oxidative stress during ozone exposure. Free Radic Res, 39: 305-316.
http://dx.doi.org/10.1080/10715760400011098 195667
Shore, S. A.; Rivera-Sanchez, Y. M.; Schwartzman, I. N.; Johnston, R. A. (2003). Responses to ozone are increased in
obese mice. J Appl Physiol, 95: 938-945. 057302
March 2011 6-199 DRAFT - DO NOT CITE OR QUOTE
-------
Sienra-Monge, J. I; Ramirez-Aguilar, M.; Moreno-Macias, H.; Reyes-Ruiz, N. I.; Del Rio-Navarro, B. E.; Ruiz-Navarro,
M. X.; Hatch, G; Crissman, K.; Slade, R.; Devlin, R. B.; Romieu, I. (2004). Antioxidant supplementation and nasal
inflammatory responses among young asthmatics exposed to high levels of ozone. Clin Exp Immunol, 138: 317-
322. http://dx.doi.0rg/10.llll/i.1365-2249.2004.02606.x 196422
Silverman, R. A.; Ito, K. (2010). Age-related association of fine particles and ozone with severe acute asthma in New York
City. J Allergy Clin Immunol, 125: 367-373.e5. http://dx.doi.Org/10.1016/j.jaci.2009.10.061 386252
Silverman, R. A.; Ito, K.; Freese, J.; Kaufman, B. J.; De Claro, D.; Braun, J.; Prezant, D. J. (2010). Association of ambient
fine particles with out-of-hospital cardiac arrests in New York City. Am J Epidemiol, 172: 917-923.
http://dx.doi.org/10.1093/aje/kwq217 647265
Simpson, R.; Williams, G; Petroeschevsky, A.; Best, T; Morgan, G; Denison, L.; Hinwood, A.; Neville, G. (2005). The
short-term effects of air pollution on hospital admissions in four Australian cities. Aust N Z J Public Health, 29:
213-221. 087438
Sinclair, A. H.; Edgerton, E. S.; Wyzga, R.; Tolsma, D. (2010). A two-time-period comparison of the effects of ambient air
pollution on outpatient visits for acute respiratory illnesses. J Air Waste Manag Assoc, 60: 163-175.
http://dx.doi.0rg/10.3155/1047-3289.60.2.163 386271
Sinclair, A. H.; Tolsma, D. (2004). Associations and lags between air pollution and acute respiratory visits in an ambulatory
care setting: 25-month results from the aerosol research and inhalation epidemiological study. J Air Waste Manag
Assoc, 54: 1212-1218. 088696
Smith, R. L.; Xu, B.; Switzer, P. (2009). Reassessing the relationship between ozone and short-term mortality in U.S. urban
communities. Inhal Toxicol, 21: 37-61. http://dx.doi.org/10.1080/08958370903161612 199750
Solic, J. J.; Hazucha, M. J.; Bromberg, P. A. (1982). The acute effects of 0.2 ppm ozone in patients with chronic obstructive
pulmonary disease. Am Rev Respir Dis, 125: 664-669. 039610
Son, J. Y; Bell, M. L.; Lee, J. T. (2010). Individual exposure to air pollution and lung function in Korea: Spatial analysis
using multiple exposure approaches. Environ Res, 110: 739-49. http://dx.doi.Org/10.1016/j.envres.2010.08.003
646655
Spannhake, E. W.; Reddy, S. P. M.; Jacoby, D. B.; Yu, X.-Y; Saatian, B.; Tian, J. (2002). Synergism between rhinovirus
infection and oxidant pollutant exposure enhances airway epithelial cell cytokine production. Environ Health
Perspect, 110: 665-670. 030637
Spektor, D. M.; Lippmann, M. (1991). Health effects of ambient ozone on healthy children at a summer camp. In R. L.
Berglund; D. R. Lawson; D. J. McKee (Eds.), Tropospheric Ozone and the Environment: Papers from an
International Conference; March 1990; Los Angeles, CA(pp. 83-89). Pittsburgh, PA: Air & Waste Management
Association. 042612
Spektor, D. M.; Lippmann, M.; Lioy, P. J.; Thurston, G. D.; Citak, K.; James, D. J.; Bock, N.; Speizer, F. E.; Hayes, C.
(1988). Effects of ambient ozone on respiratory function in active, normal children. Am Rev Respir Dis, 137: 313-
320. 040904
Spektor, D. M.; Lippmann, M.; Thurston, G. D.; Lioy, P. J.; Stecko, J.; O'Connor, G; Garshick, E.; Speizer, F. E.; Hayes, C.
(1988). Effects of ambient ozone on respiratory function in healthy adults exercising outdoors. Am Rev Respir Dis,
138: 821-828.041710
Spektor, D. M.; Thurston, G. D.; Mao, J.; He, D.; Hayes, C.; Lippmann, M. (1991). Effects of single- and multiday ozone
exposures on respiratory function in active normal children. Environ Res, 55: 107-122. 042383
Stafoggia, M.; Forastiere, F.; Faustini, A.; Biggeri, A.; Bisanti, L.; Cadum, E.; Cernigliaro, A.; Mallone, S.; Pandolfi, P.;
Serinelli, M.; Tessari, R.; Vigotti, M. A.; Perucci, C. A. (2010). Susceptibility factors to ozone-related mortality: A
population-based case-crossover analysis. Am J Respir Crit Care Med, 182: 376-384.
http://dx.doi.org/10.1164/rccm.200908-1269OC625034
Steinvil, A.; Fireman, E.; Kordova-Biezuner, L.; Cohen, M.; Shapira, L; Berliner, S.; Rogowski, O. (2009). Environmental
air pollution has decremental effects on pulmonary function test parameters up to one week after exposure. Am J
Med Sci, 338: 273-279. http://dx.doi.org/10.1097/MAJ.Ob013e3181adb3ed548780
Steinvil, A.; Kordova-Biezuner, L.; Shapira, L; Berliner, S.; Rogowski, O. (2008). Short-term exposure to air pollution and
inflammation-sensitive biomarkers. Environ Res, 106: 51-61. 188893
March 2011 6-200 DRAFT - DO NOT CITE OR QUOTE
-------
Stenfors, N.; Bosson, J.; Helleday, R.; Behndig, A. F.; Pourazar, I; Tornqvist, H.; Kelly, R I; Frew, A. I; Sandstrom, T.;
Mudway, I. S.; Blomber, A. (2010). Ozone exposure enhances mast-cell inflammation in asthmatic airways despite
inhaled corticosteroid therapy. Inhal Toxicol, 22: 133-139. http://dx.doi.org/10.3109/08958370903005736 386512
Stenfors, N.; Pourazar, J.; Blomberg, A.; Krishna, M. T.; Mudway, I.; Helleday, R.; Kelly, F. J.; Frew, A. J.; Sandstrom, T.
(2002). Effect of ozone on bronchial mucosal inflammation in asthmatic and healthy subjects. Respir Med, 96: 352-
358. 030473
Stieb, D. M.; Szyszkowicz, M.; Rowe, B. H.; Leech, J. A. (2009). Air pollution and emergency department visits for
cardiac and respiratory conditions: A multi-city time-series analysis. Environ Health, 8: 25.
http://dx.doi.org/10.1186/1476-069X-8-25 195858
Strickland, M. J.; Darrow, L. A.; Klein, M.; Flanders, W. D.; Sarnat, J. A.; Waller, L. A.; Sarnat, S. E.; Mulholland, J. A.;
Tolbert, P. E. (2010). Short-term associations between ambient air pollutants and pediatric asthma emergency
department visits. Am J Respir Crit Care Med, 182: 307-316. http://dx.doi.org/10.1164/rccm.200908-1201OC
624878
Stylianou, M.; Nicolich, M. J. (2009). Cumulative effects and threshold levels in air pollution mortality: Data analysis of
nine large US cities using the NMMAPS dataset. Environ Pollut, 157: 2216-2223.
http://dx.doi.0rg/10.1016/j.envpol.2009.04.011 620299
Symons, J. M.; Wang, L.; Guallar, E.; Howell, E.; Dominici, F.; Schwab, M.; Ange, B. A.; Samet, J.; Ondov, J.; Harrison,
D.; Geyh, A. (2006). A case-crossover study of fine particulate matter air pollution and onset of congestive heart
failure symptom exacerbation leading to hospitalization. Am J Epidemiol, 164: 421-433. 091258
Szyszkowicz, M. (2008). Ambient air pollution and daily emergency department visits for ischemic stroke in Edmonton,
Canada. Int J Occup Med Environ Health, 21: 295-300. 192128
Takeuchi, C.; Galve, R.; Nieva, J.; Witter, D. P.; Wentworth, A. D.; Troseth, R. P.; Lerner, R. A.; Wentworth P, Jr (2006).
Proatherogenic effects of the cholesterol ozonolysis products, atheronal-A and atheronal-B. Biochemistry, 45: 7162-
7170. http://dx.doi.org/10.1021/bi0604330 197793
Tamer, L.; Caliko?lu, M.; Ates, N. A.; Yildirim, H.; Ercan, B.; Saritas, E.; Unlii, A.; Atik, U. (2004). Glutathione-S-
transferase gene polymorphisms (GSTT1, GSTM1, GSTP1) as increased risk factors for asthma. Respirology, 9:
493-498. 199914
Tankersley, C. G; Peng, R. D.; Bedga, D.; Gabrielson, K.; Champion, H. C. (2010). Variation in echocardiographic and
cardiac hemodynamic effects of PM and ozone inhalation exposure in strains related to Nppa and Nprl gene knock-
out mice. Inhal Toxicol, 22: 695-707. http://dx.doi.org/10.3109/08958378.2010.487549 628062
Thaller, E. L; Petronella, S. A.; Hochman, D.; Howard, S.; Chhikara, R. S.; Brooks, E. G. (2008). Moderate increases in
ambient PM2.5 and ozone are associated with lung function decreases in beach lifeguards. J Occup Environ Med,
50: 202-211. http://dx.doi.org/10.1097/JOM.Ob013e31816386b4 195869
Thompson, A. M.; Zanobetti, A.; Silverman, F.; Schwartz, J.; Coull, B.; Urch, B.; Speck, M.; Brook, J. R.; Manno, M.;
Gold, D. R. (2010). Baseline Repeated Measures from Controlled Human Exposure Studies: Associations between
Ambient Air Pollution Exposure and the Systemic Inflammatory Biomarkers IL-6 and Fibrinogen. Environ Health
Perspect, 118: 120-124. http://dx.doi.org/10.1289/ehp.0900550 386859
Thomson, E. M.; Kumarathasan, P.; Calderon-Garciduenas, L.; Vincent, R. (2007). Air pollution alters brain and pituitary
endothelin-1 and inducible nitric oxide synthase gene expression. Environ Res, 105: 224-233.
http://dx.doi.0rg/10.1016/i.envres.2007.06.005 196635
Thomson, E.; Kumarathasan, P.; Goegan, P.; Aubin, R. A.; Vincent, R. (2005). Differential regulation of the lung endothelin
system by urban particulate matter and ozone. Toxicol Sci, 88: 103-113. 087554
Thomson, E.; Kumarathasan, P.; Vincent, R. (2006). Pulmonary expression of preproET-1 and preproET-3 mRNAs is
altered reciprocally in rats after inhalation of air pollutants. Exp Biol Med, 231: 979-984.
http://ebm.rsmjournals.com/cgi/content/abstract/231/6/979. 097483
Thurston, G. D.; Lippmann, M.; Scott, M. B.; Fine, J. M. (1997). Summertime haze air pollution and children with asthma.
Am J Respir Crit Care Med, 155: 654-660. 077645
Tolbert, P. E.; Klein, M.; Peel, J. L.; Sarnat, S. E.; Sarnat, J. A. (2007). Multipollutant modeling issues in a study of ambient
air quality and emergency department visits in Atlanta. J Expo Sci Environ Epidemiol, 17: S29-S35.
http://dx.doi.org/10.1038/si.ies.7500625090316
March 2011 6-201 DRAFT - DO NOT CITE OR QUOTE
-------
Torres, A.; Utell, M. I; Morow, P. E.; Voter, K. Z.; Whitin, J. C.; Cox, C.; Looney, R. I; Speers, D. M.; Tsai, Y; Frampton,
M. W. (1997). Airway inflammation in smokers and nonsmokers with varying responsiveness to ozone. Am J
Respir Crit Care Med, 156: 728-736. 084265
Trenga, C. A.; Koenig, J. Q.; Williams, P. V. (2001). Dietary antioxidants and ozone-induced bronchial
hyperresponsiveness in adults with asthma. Arch Environ Occup Health, 56: 242-249. 019845
Triche, E. W.; Gent, J. R; Holford, T. R.; Belanger, K.; Bracken, M. B.; Beckett, W. S.; Naeher, L.; McSharry, J.-E.;
Leaderer, B. P. (2006). Low-level ozone exposure and respiratory symptoms in infants. Environ Health Perspect,
114: 911-916. http://dx.doi.org/10.1289/ehp.8559 093274
Turner, R. M.; Muscatello, D. J.; Zheng, W.; Willmore, A.; Arendts, G. (2007). An outbreak of cardiovascular syndromes
requiring urgent medical treatment and its association with environmental factors: an ecological study. Environ
Health, 6: 37. http://dx.doi.org/10.1186/1476-069X-6-37 196637
U.S. EPA (1986). Air quality criteria for ozone and other photochemical oxidants (Report No. EPA-600/8-84-020aF - EPA-
600/8-84-020eF). Research Triangle Park, NC: U.S. Environmental Protection Agency. 017607
U.S. EPA (1996). Air quality criteria for ozone and related photochemical oxidants (Report No. EPA/600/P-93/004AF).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development. (NTIS
No. PB94173127) 017831
U.S. EPA (2006). Air quality criteria for ozone and related photochemical oxidants (Report No. EPA/600/R-05/004AF).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid= 149923. 088089
U.S. EPA (2009). Integrated science assessment for particulate matter (Report No. EPA/600/R-08/139F). Research Triangle
Park, NC: U.S. Environmental Protection Agency, Office of Research and Development.
http://cfpub.epa. gov/ncea/cfm/recordisplay.cfm?deid=216546. 179916
Uchiyama, L; Simomura, Y; Yokoyama, E. (1986). Effects of acute exposure to ozone on heart rate and blood pressure of
the conscious rat. Environ Res, 41: 529-537. 040883
Uchiyama, L; Yokoyama, E. (1989). Effects of short- and long-term exposure to ozone on heart rate and blood pressure of
emphysematous rats. Environ Res, 48: 76-86. 042051
Ulmer, C.; Kopp, M.; Ihorst, G; Frischer, T; Forster, J.; Kuehr, J. (1997). Effects of ambient ozone exposures during the
spring and summer of 1994 on pulmonary function of schoolchildren. Pediatr Pulmonol, 23: 344-353.
http://dx.doi.org/10.1002/(SICI)1099-0496(199705)23:5<344::AID-PPUL6>3.0.CO;2-K 083625.
Ultman, J. S.; Ben-Jebria, A.; Arnold, S. F. (2004). Uptake distribution of ozone in human lungs: intersubject variability in
physiologic response. University Park, PA: Health Effects Institute. http://pubs.healtheffects.org/view.php?id=70.
057197
Urch, B.; Silverman, F.; Corey, P.; Brook, J. R.; Lukic, K. Z.; Rajagopalan, S.; Brook, R. D. (2005). Acute blood pressure
responses in healthy adults during controlled air pollution exposures. Environ Health Perspect, 113: 1052-1055.
081080
Vagaggini, B.; Bartoli, M. L.; Cianchetti, S.; Costa, F.; Bacci, E.; Dente, F. L.; Di Franco, A.; Malagrino, L.; Paggiaro, P.
(2010). Increase in markers of airway inflammation after ozone exposure can be observed also in stable treated
asthmatics with minimal functional response to ozone. Respir Res, 11:5. http://dx.doi.org/10.1186/1465-9921-ll-5
387127
Vagaggini, B.; Carnevali, S.; Macchioni, P.; Taccola, M.; Fornai, E.; Bacci, E.; Bartoli, M. L.; Cianchetti, S.; Dente, F. L.;
Di Franco, A.; Giannini, D.; PaggiaroPL (1999). Airway inflammatory response to ozone in subjects with different
asthma severity. Eur Respir J, 13: 274-280. 001210
Vagaggini, B.; Cianchetti, S.; Bartoli, M.; Ricci, M.; Bacci, E.; Dente, F. L.; Di Franco, A.; Paggiaro, P. (2007). Prednisone
blunts airway neutrophilic inflammatory response due to ozone exposure in asthmatic subjects. Respiration, 74: 61-
58. http://dx.doi.org/10.1159/000096078 196638
Vagaggini, B.; Taccola, M.; Cianchetti, S.; Carnevali, S.; Bartoli, M. L.; Bacci, E.; Dente, F. L.; Di Franco, A.; Giannini,
D.; Paggiaro, P. L. (2002). Ozone exposure increases eosinophilic airway response induced by previous allergen
challenge. Am J Respir Crit Care Med, 166: 1073-1077. 035191
March 2011 6-202 DRAFT - DO NOT CITE OR QUOTE
-------
Vagaggini, B.; Taccola, M.; Conti, I.; Carnevali, S.; Cianchetti, S.; Bartoli, M. L.; Bacci, E.; Dente, F. L.; Di Franco, A.;
Giannini, D.; Paggiaro, P. L. (2001). Budesonide reduces neutrophilic but not functional airway response to ozone
in mild asthmatics. Am J Respir Crit Care Med, 164: 2172-2176. 025343
Valacchi, G; Pecorelli, A.; Mencarelli, M.; Maioli, E.; Davis, P. A. (2009). Beta-carotene prevents ozone-induced
proinflammatory markers in murine skin. Toxicol Ind Health, 25: 241-247.
http://dx.doi.org/10.1177/0748233709103030201554
Van Loveren, H.; Krajnc, E. L; Rombout, P. J.; Blommaert, F. A.; Vos, J. G. (1990). Effects of ozone, hexachlorobenzene,
and bis(tri-n-butyltin)oxide on natural killer activity in the rat lung. Toxicol Appl Pharmacol, 102: 21-33. 042264
Vancza, E. M.; Galdanes, K.; Gunnison, A.; Hatch, G; Gordon, T. (2009). Age, strain, and gender as factors for increased
sensitivity of the mouse lung to inhaled ozone. Toxicol Sci, 107: 535-543. http://dx.doi.org/10.1093/toxsci/kfn253
596419
van Loveren, H.; Rombout, P. J. A.; Wagenaar, S. S.; Walvoort, H. C.; Vos, J. G. (1988). Effects of ozone on the defense to
a respiratory Listeria monocytogenes infection in the rat: suppression of macrophage function and cellular
immunity and aggravation of histopathology in lung and liver during infection. Toxicol Appl Pharmacol, 94: 374-
393.041847
Vesely, D. L.; Giordano, A. T; Raska-Emery, P.; Montgomery, M. R. (1994). Increase in atrial natriuretic factor in the
lungs, heart, and circulatory system owing to ozone. Chest, 105: 1551-1554. 076015
Vesely, D. L.; Giordano, A. T; Raska-Emery, P.; Montgomery, M. R. (1994). Ozone increases amino- and carboxy-terminal
atrial natriuretic factor prohormone peptides in lung, heart, and circulation. J Biochem Mol Toxicol, 9: 107-112.
028877
Vesely, D. L.; Giordano, A. T; Raska-Emery, P.; Montgomery, M. R. (1994). Ozone increases atrial natriuretic peptides in
heart, lung and circulation of aged vs adult animals. Gerontology, 40: 227-236.
http://dx.doi.org/10.1159/000213590076228
Villeneuve, P. J.; Chen, L.; Stieb, D.; Rowe, B. H. (2006). Associations between outdoor air pollution and emergency
department visits for stroke in Edmonton, Canada. Eur J Epidemiol, 21: 689-700. 090191
Villeneuve, P. J.; Doiron, M.-S.; Stieb, D.; Dales, R.; Burnett, R. T; Dugandzic, R. (2006). Is outdoor air pollution
associated with physician visits for allergic rhinitis among the elderly in Toronto, Canada? Allergy, 61: 750-758.
http://dx.doi.0rg/10.llll/i.1398-9995.2006.01070.x091179
Vincent, R.; Janzen, E. G; Chen, G; Kumarathasan, P.; Haire, D. L.; Guenette, J.; Chen, J. Z.; Bray, T. M. (1996). Spin
trapping study in the lungs and liver of F344 rats after exposure to ozone. Free Radic Res, 25: 475-488. 080777
Von Klot, S.; Peters, A.; Aalto, P.; Bellander, T; Berglind, N.; DTppoliti, D.; Elosua, R.; Hermann, A.; Kulmala, M.; Lanki,
T; Lowel, H.; Pekkanen, J.; Picciotto, S.; Sunyer, J.; Forastiere, F.; Health Effects of Particles on Susceptible
Subpopulations (HEAPSS) Study Group (2005). Ambient air pollution is associated with increased risk of hospital
cardiac readmissions of myocardial infarction survivors in five European cities. Circulation, 112: 3073-3079.
http://dx.doi.org/10.1161/CIRCULATIONAHA.105.548743088070
Voynow, J. A.; Fischer, B. M.; Zheng, S.; Potts, E. N.; Grover, A. R.; Jaiswal, A. K.; Ohio, A. J.; Foster, W. M. (2009).
NAD(P)H quinone oxidoreductase 1 is essential for ozone-induced oxidative stress in mice and humans. Am J
Respir Cell Mol Biol, 41: 107-113. 194311
Wagner, J. G; Harkema, J. R.; Jiang, Q.; Illek, B.; Ames, B. N.; Peden, D. B. (2009). Gamma-tocopherol attenuates ozone-
induced exacerbation of allergic rhinosinusitis in rats. Toxicol Pathol, 37: 481-491.
http://dx.doi.org/10.1177/0192623309335630201574
Wagner, J. G; Jiang, Q.; Harkema, J. R.; Illek, B.; Patel, D. D.; Ames, B. N.; Peden, D. B. (2007). Ozone enhancement of
lower airway allergic inflammation is prevented by gamma-tocopherol. Free Radic Biol Med, 43:1176-1188.
http://dx.doi.0rg/10.1016/i.freeradbiomed.2007.07.013596420
Wang, D. J.; Zhou, W. D.; Dai, X. J.; Yan, Y. (2007). Study on effect and mechanism of sodium ferulate in preventing and
treating ozone induced lung injury in mice. Chin J Integr Med, 13: 211-214. http://dx.doi.org/10.1007/sll655-007-
0211-9596421
Ward, D. J.; Roberts, K. T; Jones, N.; Harrison, R. M.; Ayres, J. G; Hussain, S.; Walters, S. (2002). Effects of daily
variation in outdoor particulates and ambient acid species in normal and asthmatic children. Thorax, 57: 489-502.
http://dx.doi.0rg/10.1136/thorax.57.6.489025839
March 2011 6-203 DRAFT - DO NOT CITE OR QUOTE
-------
Watkinson, W. P.; Aileru, A. A.; Dowd, S. M.; Doerfler, D. L.; Tepper, J. S.; Costa, D. L. (1993). Acute effects of ozone on
heart rate and body temperature in the unanesthetized, unrestrained rat maintained at different ambient
temperatures. Inhal Toxicol, 5: 129-147. 043205
Watkinson, W. P.; Campen, M. J.; Wichers, L. B.; Nolan, J. P.; Costa, D. L. (2003). Cardiac and thermoregulatory responses
to inhaled pollutants in healthy and compromised rodents: modulation via interaction with environmental factors.
Environ Res, 92: 35-47. 050547
Weinmann, G. G; Bowes, S. M.; Gerbase, M. W.; Kimball, A. W.; Frank, R. (1995). Response to acute ozone exposure in
healthy men: results of a screening procedure. Am J Respir Crit Care Med, 151: 33-40. 076022
Weissbecker, L.; Carpenter, R. D.; Luchsinger, P. C.; Osdene, T. S. (1969). In vitro alveolar macrophage viability: effect of
gases. Arch Environ Occup Health, 18: 756-759. 015806
Wellenius, G. A.; Bateson, T. F.; Mittleman, M. A.; Schwartz, J. (2005). Particulate air pollution and the rate of
hospitalization for congestive heart failure among medicare beneficiaries in Pittsburgh, Pennsylvania. Am J
Epidemiol, 161: 1030-1036. 087483
Wellenius, G. A.; Yeh, G. Y; Coull, B. A.; Suh, H. H.; Phillips, R. S.; Mittleman, M. A. (2007). Effects of ambient air
pollution on functional status in patients with chronic congestive heart failure: a repeated-measures study. Environ
Health, 6: 1-7. 092830
Wentworth, P. Jr; Nieva, J.; Takeuchi, C.; Galve, R.; Wentworth, A. D.; Dilley, R. B.; DeLaria, G. A.; Saven, A.; Babior, B.
M.; Janda, K. D.; Eschenmoser, A.; Lerner, R. A. (2003). Evidence for ozone formation in human atherosclerotic
arteries. Science, 302: 1053-1056. 052486
Wheeler, A.; Zanobetti, A.; Gold, D. R.; Schwartz, J.; Stone, P.; Suh, H. H. (2006). The relationship between ambient air
pollution and heart rate variability differs for individuals with heart and pulmonary disease. Environ Health
Perspect, 114: 560-566. 088453
Williams, A. S.; Leung, S. Y; Nath, P.; Khorasani, N. M.; Bhavsar, P.; Issa, R.; Mitchell, J. A.; Adcock, I. M.; Chung, K. F.
(2007). Role of TLR2, TLR4, and MyD88 in murine ozone-induced airway hyperresponsiveness and neutrophilia. J
ApplPhysiol, 103: 1189-1195. http://dx.doi.org/10.1152/iapplphvsiol.00172.2007597545
Wiwatanadate, P.; Trakultivakorn, M. (2010). Air pollution-related peak expiratory flow rates among asthmatic children in
Chiang Mai, Thailand. Inhal Toxicol, 22: 301-308. http://dx.doi.org/10.3109/08958370903300327 387706
Wong, C. M.; Yang, L.; Thach, T. Q.; Chau, P. Y; Chan, K. P.; Thomas, G. N.; Lam, T. H.; Wong, T. W.; Hedley, A. J.;
Peiris, J. S. (2009). Modification by influenza on health effects of air pollution in Hong Kong. Environ Health
Perspect, 117: 248-253. http://dx.doi.org/10.1289/ehp.11605 196722
Wong, C.-M.; Ma, S.; Hedley AJLam, T.-H. (1999). Does ozone have any effect on daily hospital admissions for
circulatory diseases? J Epidemiol Community Health, 53: 580-581. 011463
Wong, T. W.; Lau, T. S.; Yu, T. S.; Neller, A.; Wong, S. L.; Tarn, W.; Pang, S. W. (1999). Air pollution and hospital
admissions for respiratory and cardiovascular diseases in Hong Kong. Occup Environ Med, 56: 679-683. 009172
Wu, C. F.; Kuo, I. C.; Su, T. C.; Li, Y. R.; Lin, L. Y; Chan, C. C.; Hsu, S. C. (2010). Effects of personal exposure to
particulate matter and ozone on arterial stiffness and heart rate variability in healthy adults. Am J Epidemiol, 171:
1299-1309. http://dx.doi.org/10.1093/aje/kwq060 626033
Xia, Y; Tong, H. (2006). Cumulative effects of air pollution on public health. Stat Med, 25: 3548-3559.
http://dx.doi.org/10.1002/sim.2446 623157
Yallop, D.; Duncan, E. R.; Norris, E.; Fuller, G. W.; Thomas, N.; Walters, J.; Dick, M. C.; Height, S. E.; Thein, S. L.; Rees,
D. C. (2007). The associations between air quality and the number of hospital admissions for acute pain and sickle-
cell disease in an urban environment. Br JHaematol, 136: 844-848. http://dx.doi.Org/10.llll/j.1365-
2141.2007.06493.x 090702
Yang, C. Y. (2008). Air pollution and hospital admissions for congestive heart failure in a subtropical city: Taipei, Taiwan. J
Toxicol Environ Health A, 71: 1085-1090. 157160
Yang, C.-Y; Chen, Y.-S.; Yang, C.-H.; Ho, S.-C. (2004). Relationship between ambient air pollution and hospital
admissions for cardiovascular diseases in Kaohsiung, Taiwan. J Toxicol Environ Health A, 67: 483-493. 094376
March 2011 6-204 DRAFT - DO NOT CITE OR QUOTE
-------
Yang, Q.; Chen, Y; Krewski, D.; Burnett, R. T.; Shi, Y; McGrail, K. M. (2005). Effect of short-term exposure to low levels
of gaseous pollutants on chronic obstructive pulmonary disease hospitalizations. Environ Res, 99: 99-105.
http://dx.doi.0rg/10.1016/i.envres.2004.09.014090184
Ying, R. L.; Gross, K. B.; Terzo, T. S.; Eschenbacher, W. L. (1990). Indomethacin does not inhibit the ozone-induced
increase in bronchial responsiveness in human subjects. Am J Respir Crit Care Med, 142: 817-821. 042334
Yokoyama, E.; Uchiyama, L; Arito, H. (1989). Extrapulmonary effects of low level ozone exposure. In Schneider, T.; Lee,
S. D.; Wolters, G. J. R.; Grant, L. D. (Eds.), Atmospheric ozone research and its policy implications: proceedings of
the 3rd US-Dutch international symposium; May 1988; Nijmegen, The Netherlands Amsterdam, The Netherlands:
Elsevier. 041741
Yoon, H. K.; Cho, H. Y; Kleeberger, S. R. (2007). Protective role of matrix metalloproteinase-9 in ozone-induced airway
inflammation. Environ Health Perspect, 115: 1557-1563. http://dx.doi.org/10.1289/ehp.10289 596422
Zanobetti, A.; Canner, M. J.; Stone, P. H.; Schwartz, J.; Sher, D.; Eagan-Bengston, E.; Gates, K. A.; Hartley, L. H.; Suh, H.;
Gold, D. R. (2004). Ambient pollution and blood pressure in cardiac rehabilitation patients. Circulation, 110: 2184-
2189. http://dx.doi.org/10.1161/01.cir.0000143831.33243.d8087489
Zanobetti, A.; Gold, D. R.; Stone, P. H.; Suh, H. H.; Schwartz, J.; Coull, B. A.; Speizer, F. E. (2010). Reduction in heart rate
variability with traffic and air pollution in patients with coronary artery disease. Environ Health Perspect, 118: 324-
330. http://ehsehplp03.niehs.nih.gov/article/info%3Adoi%2F10.1289%2Fehp.0901003. 597250
Zanobetti, A.; Schwartz, J. (2006). Air pollution and emergency admissions in Boston, MA. J Epidemiol Community
Health, 60: 890-895. 090195
Zanobetti, A.; Schwartz, J. (2008). Is there adaptation in the ozone mortality relationship: A multi-city case-crossover
analysis. Environ Health, 7: 22. http://dx.doi.org/10.1186/1476-069X-7-22 195755
Zanobetti, A.; Schwartz, J. (2008). Mortality displacement in the association of ozone with mortality: An analysis of 48
cities in the United States. Am J Respir Crit Care Med, 177: 184-189. http://dx.doi.org/10.1164/rccm.200706-
8230C 101596
March 2011 6-205 DRAFT - DO NOT CITE OR QUOTE
-------
Chapter 7. Integrated Health Effects of
Long-Term Ozone Exposure
7.1. Introduction
1 This chapter reviews, summarizes, and integrates the evidence on relationships between health
2 effects and long-term exposures to O3. Both epidemiologic and toxicological studies provide a basis
3 for examining long-term O3 exposure health effects for respiratory effects, cardiovascular effects,
4 reproductive and developmental effects, central nervous system effects, cancer outcomes, and
5 mortality.
6 Conclusions from the 2006 O3 AQCD are summarized briefly at the beginning of each section,
7 and the evaluation of evidence from recent studies builds upon what was available during the
8 previous review. For each health outcome (e.g., respiratory disease, lung function), results are
9 summarized for studies from the specific scientific discipline, i.e., epidemiologic and toxicological
10 studies. The major sections (i.e. respiratory, cardiovascular, mortality, reproductive/developmental,
11 cancer) conclude with summaries of the evidence for the various health outcomes within that
12 category and integration of the findings that lead to conclusions regarding causality based upon the
13 framework described in Chapter 1. Determination of causality is made for the overall health effect
14 category, such as respiratory effects, with coherence and plausibility being based on evidence from
15 across disciplines and also across the suite of related health outcomes, including cause-specific
16 mortality.
7.2. Respiratory Effects
17 Studies reviewed in the 2006 O3 AQCD (U.S. EPA, 2006, 088089) examined evidence for
18 relationships between long-term O3 exposure and effects on respiratory health outcomes including
19 seasonal declines in lung function, increases in inflammation, and development of asthma in children
20 and adults. The term seasonal was used in these studies as a measure of a long-term exposure of
21 several months. Animal toxicology data provided a clearer picture indicating that long-term O3
22 exposure may have lasting effects. Chronic exposure studies in animals have reported biochemical
23 and morphological changes suggestive of irreversible long-term O3 impacts on the lung. In contrast
24 to supportive evidence from chronic animal studies, the epidemiologic studies on longer-term lung
25 function declines, inflammation, and new asthma development remained inconclusive. Several
26 studies (e.g., Frischer et al, 1999, 001037: Horak et al, 2002, 034792) collectively indicated that
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments such as the Integrated Science Assessments (ISAs) and the Integrated Risk Information System (IRIS).
March 2011 7-1 DRAFT - DO NOT CITE OR QUOTE
-------
1 seasonal O3 exposure was associated with smaller increases in lung function growth in children. For
2 longer time periods, the definitive analysis in the Child Health Study (CHS) reported by Gauderman
3 et al. (2004, 056569) provided little evidence that long-term exposure to ambient O3 at current levels
4 was associated with significant deficits in the growth rate of lung function in children in contrast to
5 the effects observed with other pollutants such as acid vapor, NO2, and PM2 5. Asthmatic children
6 with GSTM1 null genotype were found to be more susceptible to the impact of O3 exposure on small
7 airways function in Mexico (Romieu et al., 2004, 056796). Limited epidemiologic research
8 examined the relationship between long-term O3 exposures and inflammation. Inflammatory effects
9 consistent with known effects of O3 such as increased eosinophil levels were observed in an Austrian
10 study (Frischer et al., 2001, 019683). The cross-sectional surveys available for the 2006 O3 AQCD
11 detected no associations between long-term O3 exposures and asthma prevalence, asthma-related
12 symptoms or allergy to common aeroallergens in children after controlling for covariates.
13 New evidence presented below reports consistent associations between long-term O3 exposure
14 and new-onset asthma related to genotype in U.S. cohorts in multi-community studies. Related
15 studies report coherent relationships between respiratory symptoms among asthmatics and long-term
16 O3 exposure. Anew line of evidence reports a positive exposure response relationship between first
17 asthma hospitalization and long-term O3 exposure. Results from recent studies examining pulmonary
18 function, inflammation, and allergic responses are also presented.
7.2.1. New Onset Asthma
19 Risk for new-onset asthma is related in part to genetic susceptibility, behavioral factors and
20 environmental exposure (Gilliland et al., 1999, 155792). Complex chronic diseases, such as asthma,
21 are partially the result of a sequence of biochemical reactions involving exposures to various
22 environmental agents metabolized by a number of different genes (Conti et al., 2003, 626696).
23 Understanding the relation between genetic polymorphisms and environmental exposure can help
24 identify high-risk subgroups in the population and provide better insight into pathway mechanisms
25 for these complex diseases. Oxidative stress likely underlies these mechanistic hypotheses (Gilliland
26 et al., 1999, 155792). Susceptibility genes act through modification of disease risk associated with
27 environmental factors. Epidemiologic investigation of hypotheses of possible mechanisms involving
28 the gene-environmental (GxE) interaction involves statistical analysis of these interactions for the
29 risk of new-onset asthma in children being influenced by exposure to air pollution (Gauderman,
30 2001, 625862: Gauderman, 2002, 626945: Gilliland et al., 1999, 155792V
31 Evidence for the potential importance of genetic susceptibility and behavioral factors on new
32 onset asthma are provided by several recent studies (Ercan et al., 2006, 595172: Gilliland et al.,
33 2002, 090970: Hanene et al., 2007, 595428: Himes et al., 2009, 480112: Islam et al., 2008, 097348:
34 Li et al., 2006, 596447: Li et al., 2008, 596449: Tamer et al., 2004, 199914). Evidence for a gene-
35 pollution interaction in the pathogenesis of asthma are supported by recent study findings (Gilliland
36 et al., 2002, 090970: Islam et al., 2008, 097348: Islam et al., 2009, 196715: Lee et al., 2004, 090971:
37 Oryszczyn et al., 2007, 596460).
March 2011 7-2 DRAFT - DO NOT CITE OR QUOTE
-------
1 Evidence for associations between long-term exposureto O3 and new-onset asthma is provided
2 by new studies from the CHS. Initiated in the early 1990's, the CHS was originally designed to
3 examine whether ambient pollutants were related to chronic respiratory outcomes in children (Peters
4 et al, 1999, 087243: Peters et al., 1999, 087237V About 10 years later, the CHS inaugurated a series
5 of genetic studies (Gilliland et al., 1999, 155792) nested within the CHS cohort by obtaining
6 biological samples from the study subjects (buccal cells). These new studies examined the
7 relationship between health outcomes, genetic susceptibility, behavioral factors and environmental
8 exposure.
9 First, the hypothesis that the functional polymorphisms of HMOX-1 [(GT)n repeat], CAT
10 (-262C > T -844C > TO, and MNSOD (Ala-9Val) are associated with new-onset asthma was
11 evaluated, and then whether the effects of these variants varied by exposure to O3 (Islam et al., 2008,
12 097348). HMOX1 [heme oxygenase (decycling) 1] is a human gene that encodes for the enzyme
13 heme oxygenase. Heme oxygenase 1 (HO-1) is an enzyme that catalyzes the metabolism of heme.
14 The heme iron serves as a source or sink of electrons during electron transfer or redox chemistry, so
15 the presence of the HMOX1 gene, and therefore the generation of heme oxygenase, protects against
16 oxidative stress in the body. The authors observed that functional promoter variants in CAT and
17 HMOX-1 showed ethnicity-specific associations with new-onset asthma and that oxidant gene
18 protection was restricted to children living in low-O3 communities.
19 The subjects were obtained from the CHS from 12 communities in southern California.
20 Children with a history of asthma or wheeze were excluded from this analysis. Analyses were
21 restricted to children of Hispanic (n = 576) or non-Hispanic white ethnicity (n = 1,125). New-onset
22 asthma was classified as such for children with no prior history of asthma at study entry who
23 subsequently reported physician-diagnosed asthma at annual follow-up with the date of onset
24 assigned to be the midpoint of the interval between the interview date when asthma diagnosis was
25 first reported and the previous interview date. As a sensitivity analysis, the asthma definition was
26 restricted to those new-onset asthma cases who also used an inhaler (n = 121). Long-term pollutant
27 levels were calculated from 1994 through 2003. The effect of ambient air pollution on the
28 relationship between genetic polymorphism and new-onset asthma was assessed using models where
29 the community specific average air pollution levels were fitted as a continuous variable together with
30 the appropriate interaction terms for genes and air pollutants (Berhane et al., 2004, 626732). Cox
31 proportional hazard regression models were fitted to the data. A stratified analysis for the two
32 independent fourth-grade cohorts of the study population recruited in 1993 and 1996 were conducted
33 to assess whether the results could be replicated in independent groups of children.
34 Over the follow-up period, 160 new cases of asthma were diagnosed (Islam et al., 2008,
35 097348). The evidence indicated that the effect of variation in the HMOX-1 gene on risk of new-
36 onset asthma differed by ambient O3 level. An interaction P value was reported of 0.003 from the
37 hierarchical two stage Cox proportional hazard model fitting the community-specific O3 and PMi0
38 levels (continuous) and controlling for random effect of the communities. The annual O3 levels
39 (10:00 a.m. - 6:00 p.m.) ranged from 46.5 to 64.9 ppb in the six higher O3 communities (mean =
March 2011 7-3 DRAFT - DO NOT CITE OR QUOTE
-------
1 55.2 ppb) and 28.6 to 45.5 ppb in the six lower O3 communities (mean = 38.4 ppb). Average O3
2 levels showed low correlation with the other monitored pollutants. The interaction indicated a greater
3 effect (association) of community O3 level among children with the gene than with children without
4 the gene. Alleles with 23 or fewer (GT)n repeats are categorized as short (S). The S-allele variant of
5 this protective enzyme is more readily induced than those with more numerous repeats. The largest
6 protective effect of the (GT)n repeat polymorphism of HMOX-1 was observed for children who
7 were S-allele carriers and resided in low-O3 communities with Hazard Ratio (HR) of 0.44 (95% CI:
8 0.23, 0.83). The ratio of HR of S-allele carriers who resided in high O3 communities (HR 0.88; [95%
9 CI: 0.33, 2.34]) was twofold greater than in those who resided in the low-O3 communities (HR 0.44).
10 The non-parallelism of the two lines in Figure 7-1 illustrates the interaction: Children with the S-
11 allele have protection against the onset of asthma; however, in high- O3 communities, this protection
12 is attenuated. The results from sensitivity analyses on the two fourth-grade cohorts, and the inhaler
13 definition for asthma were both consistent with the main results. An analysis related to children's
14 participation in sports or time spent outdoors produced the same outcome. No significant interactions
15 were observed between PMi0 or other pollutants and the HMOX -1 gene. A potential concern for not
16 adjusting for multiple testing was considered by the authors as not a factor in this analysis because
17 the selection of the genes was based on a priori hypotheses defined by a well-studied biological
18 pathway. Thus in this cohort in southern California, Islam et al. (2008, 097348) related new-onset
19 asthma to O3 exposure in genetically susceptible children.
20 Related to the findings in Islam et al. (2008, 097348) discussed above, Islam et al. (2009,
21 196715) examined whether GSTP1, GSTM1, exercise and O3 exposure have interrelated effects on
22 the pathogenesis of asthma. A modifying role of air pollution on the association between Ilel05Val
23 and asthma in a cohort of children had been observed (Lee et al., 2004, 090971). but the study did
24 not examine O3 specifically or consider exercise. A primary conclusion that the authors (Islam et al.,
25 2009, 196715) reported was that the common functional variants of GSTP1 and GSTM1 null
26 genotypes modulate the risk of new onset asthma during adolescence. Children who had the GSTM1
27 null genotype were at 1.6-fold (95% CI: 1.2, 2.2) increased risk of developing new onset asthma
28 compared with those without the null genotype. Further, the CHS investigators examined the
29 complex interrelationship of antioxidant defenses with asthma risk with increasing doses of O3,
30 resulting from increasing ventilation associated with vigorous exercise characterized by the number
31 of team sports played. In an earlier analysis, McConnell et al. (2002, 023150) had reported that the
32 risk of new onset asthma was associated with outdoor exercise, especially in high O3 communities
33 but did not consider genetic variants. The plausibility of a causal association is strengthen by the
34 observation by Islam et al. (2009, 196715) that the risk of participation in team sports was related to
35 increased genetic susceptibility to oxidative stress. The sixfold increased risk of asthma (HR 6.15,
36 [95% CI: 2.2, 7.4]) for children who were homozygous for Ilel05, participated in three or more team
37 sports and lived in high-O3 communities demonstrates the potential importance of a combination of
38 genetic variability, O3 exposure and behavior on asthma risk.
March 2011 7-4 DRAFT - DO NOT CITE OR QUOTE
-------
Interaction of Gene presence and Ozone Level on the
Hazard Ratio of New Onset Asthma (P-value of 0.003)
2.5 -
2 -
1.5 -
1 -
0.5 -
0 -
(2.43)
Children with no S-Allele
IS
OL
0.44 —
(0.28)
•(0.83) _
li ^^ ^^™ f liili-lrnn iivil-fl-i C All^l^
Children with S-Allele
(2.34)
0.88
(0.36) (0.33)
Community Mean Ozone Level High
<38-4 PPb) (55.2 ppb)
(Confidence limits based on comparison with reference group)
Source:Usedwith permission from American Thoracic Society, Islam etal. (2008, 0973481.
Figure 7-1. Interaction of gene presence and ozone level on the Hazard Ratio (HR) of new-onset
asthma in the 12 Children's Health Study communities. An interaction P-value of
0.003 was obtained from the hierarchical two stage Cox proportional hazard model
fitting the community specific ozone and controlling for random effect of the
communities. The interaction indicates there is a greater effect (association) of
community ozone level on children with the gene than with children without the
gene. The HRs are off-set as opposed to overlapping in the figure to allow clearer
presentation of the results.
1 Epidemiologic evidence of associations of arginase variants with asthma are limited (Li et al,
2 2006, 596447). Asthmatic subjects have higher arginase activity than nonasthmatic subjects (Morris
3 et al., 2004, 674145). NO is a mediator of nitrosative stress synthesized from L-arginine by nitric
4 oxide synthases. In the CHS, Salam et al. (2009, 596644) examined whether arginase variants
5 (ARG1 and ARG2 genes) were associated with asthma and whether atopy and exposures to smoking
6 and air pollution influence the associations. The modifying effect of O3 and atopy on the association
7 between haplotypes and asthma were evaluated using likelihood ratio tests with appropriate
8 interaction terms. They found that both ARG1 and ARG2 genetic loci were associated with
9 childhood-onset asthma. The effect of the ARG1 haplotype varied by the child's history of atopy and
10 ambient O3. Among atopic children living in high O3 communities, those carrying the ARG1
11 halotype had reduced asthma risk (OR per ARGlh4 halotype copy: 0.12; [95% CI: 0.04, 0.43];
12 P heterogeneity across atopy/O3 categories = 0.008).
13 Further, the CHS presents results examining the relationship of new onset asthma with traffic-
14 related pollution near homes and schools (McConnell et al., 2010, 625501). Asthma risk increased
15 with modeled traffic-related pollution exposure from roadways near homes and near schools. The
16 HR was 0.76 (95% CI: 0.38, 1.54) across the range of ambient O3 exposure in the communities. With
March 2011
7-5
DRAFT - DO NOT CITE OR QUOTE
-------
1 adjustment for school and residential non-freeway traffic-related exposure, the estimated HR for O3
2 was 1.01 (95% CI: 0.49, 2.11). Gene variants were not evaluated in this study.
3 Some cross-sectional studies reviewed in the 2006 O3 AQCD observed positive relationships
4 between chronic exposure to O3 and prevalence of asthma and asthmatic symptoms in school
5 children (Ramadour et al., 2000, 013259: Wang et al, 1999, 008105) while others (Charpin et al.,
6 1999, 015152: Kuo et al., 2002, 036310) did not. Recent studies provide additional evidence.
7 In a cross-sectional nationwide study of 32,672 Taiwanese school children, Hwang et al.
8 (2005, 089454) assessed the effects of air pollutants on the risk of asthma. The study population was
9 recruited from elementary and middle schools within 1 km of air monitoring stations. The risk of
10 asthma was related to O3 in the one-pollutant model. The addition of other pollutants, in two-
11 pollutant and three-pollutant models, increased the O3 risk estimates. The prevalence of childhood
12 asthma was assessed in Portugal by contrasting the risk of asthma between a high O3 rural area and
13 an area with low O3 levels (Sousa et al., 2008, 619959: Sousa et al., 2009, 619956: Sousa et al.,
14 2011, 676712). The locations were selected to provide a difference in O3 levels without the
15 confounding effects of other pollutants. Both evaluation for asthma symptoms and FEVi suggested
16 that O3 increased asthma prevalence. Clark et al. (2010, 594440) investigated the effect of exposure
17 to ambient air pollution in utero and during the first year of life on risk of subsequent incidence
18 asthma diagnosis up to 3-4 years of age in a population-based nested case-control study for all
19 children born in southwestern British Columbia in 1999 and 2000 (n=37,401; including 3,482 [9.3%]
20 with asthma). Air pollution exposure for each subject was estimated based on their residential
21 address history using regulatory monitoring data, land use regression modeling, and proximity to
22 stationary pollutant sources. Daily values from the three closest monitors within 50 km were used to
23 calculate exposures. Traffic-related pollutants were associated with the highest risk. Ozone was
24 inversely correlated with the primary traffic-related pollutants (r = -0.7 to -0.9). The low reliability of
25 asthma diagnosis in infants makes this study difficult to interpret (Martinez et al., 1995, 046150). In
26 a cross-sectional analysis, Akinbamia et al. (2010, 378580) examined the association between
27 chronic exposure to outdoor pollutants (12-month average levels by county) and asthma outcomes in
28 a national sample of children ages 3-17 years living in U.S. metropolitan areas (National Health
29 Interview Survey, N = 34,073). A 5-ppb increase in estimated 8-h max O3 concentration (annual
30 average) yielded a positive association for both currently having asthma and for having at least 1
31 asthma attack in the previous year; while the adjusted odds ratios for other pollutants were not
32 statistically significant. Models in which pollutant value ranges were divided into quartiles produced
33 comparable results. Multi-pollutant models produced similar results. The median value for 12-month
34 average O3 levels was 39.5 ppb and the IQR was 35.9-43.7 ppb. The adjusted odds for current
35 asthma for the highest quartile (49.9-59.5 ppb) of estimated O3 exposure was 1.56 (95% CI: 1.15,
36 2.10) with a positive dose-response relationship apparent from the lowest quartile to the highest.
37 Thus, this cross-sectional analysis and Hwang et al. (2005, 089454) provides further evidence
38 relating O3 exposure and the risk of asthma.
March 2011 7-6 DRAFT - DO NOT CITE OR QUOTE
-------
1 The occurrence of bronchitic symptoms among children with asthma was investigated in the
2 CHS examining the role of gene-environment interactions and long-term O3 exposure. Lee et al.
3 (2009, 199915) studied associations of TNF-308 genotype with bronchitis symptoms among
4 asthmatic children and investigated whether associations vary with ambient O3 exposure since
5 increased airway TNF may be related to inflammation. Asthmatic children with the GG genotype
6 had a lower prevalence of bronchitic symptoms compared with children carrying at least one A-allele
7 (e.g., GA or AA). Low-versus high-O3 strata were defined as less than or greater than 50- ppb O3
8 average. Asthmatic children with TNF-308 GG genotype had a significantly reduced risk of
9 bronchitic symptoms with low-O3 exposure (OR: 0.53; [95% CI: 0.31, 0.91]). The risk was not
10 reduced in children living in high-O3 communities (OR: 1.42; [95% CI: 0.75, 2.70]). The difference
11 in genotypic effects between low- and high-O3 environments was statistically significant among
12 asthmatics (P for interaction = 0.01), but insignificant among non-asthmatic children. By using
13 dummy variables in each community, Lee et al. (2009, 199915) calculated the effect of TNF-308 GG
14 genotype on the occurrence of bronchitic symptoms among children with asthma. Figure 7-2
15 presents adjusted O3 community-specific beta-coefficients plotted against ambient O3 concentration,
16 using weights proportional to the inverse variance. They further report that they found no substantial
17 differences in the effect of the GG genotype in asthmatic children in relation to exposure to PMi0,
18 PM2s, NO2, acid vapor or second-hand smoke exposure. These results suggest a role of gene-
19 environment interactions such as long-term O3 exposure on the occurrence of bronchitic symptoms
20 among children with asthma.
03
Q_
C3 o
O H
co E
O >s
co w
LL O
S -1
_Q
-2
20 30 40 50 60 70
Average ozone from 10 a.m. to 6 p.m. in communities (ppb)
Source: Used with permission from John Wiley & Sons A/S, Lee et al. (2009, 1999151.
Figure 7-2. Ozone modifies the effect of TNF G-308A genotype on bronchitic symptoms among
children with asthma in the CHS. Using dummy variables in each of the 12
communities, betas were calculated of TNF-308 GG genotype on the occurrence of
bronchitic symptoms among children with asthma.
March 2011
7-7
DRAFT - DO NOT CITE OR QUOTE
-------
1 The French Epidemiology study on Genetics and Environment of Asthma (EGEA)
2 investigated the relationship between ambient air pollution and asthma severity in a cohort in five
3 French cities (Paris, Lyon, Marseille, Montpellier, and Grenoble) (Rage et al., 2009, 196720). In this
4 cross-sectional study, asthma severity over the past 12 months was assessed among 328 adult
5 asthmatics using two methods: (1) a four-class severity score that integrated clinical events and type
6 of treatment; and (2) a five-level asthma score based only on symptoms. Two measures of exposure
7 were also assessed: (1 [first method]) closest monitor data from 1991 to 1995 where a total of 93%
8 of the subjects lived within 10 km of a monitor, but where 70% of the O3 concentrations were back-
9 extrapolated values; and (2 [second method]) a validated spatial model that used geostatistical
10 interpolations and then assigned air pollutants to the geocoded residential addresses of all
11 participants and individually assigned exposure to ambient air pollution estimates. Higher asthma
12 severity scores were significantly related to both the 8-h avg O3 during April-September and the
13 number of days with 8-h O3 averages above 55 ppb. Both exposure assessment methods and severity
14 score methods resulted in very similar findings. Effect estimates of O3 were similar in three-pollutant
15 models. No PM data were available. Since these estimates were not sensitive to the inclusion of
16 ambient NO2 in the three-pollutant models, the authors viewed the findings not to be explained by
17 particles which usually have substantial correlations between PM and NO2. Ozone concentrations by
18 the first method for annual levels, nearest monitor were 8-hours (n = 210); mean (+SD) 30.25 (+9.7);
19 IQR: 21-36.5 (+15.5) ppb. The second exposure approach yielded summer (n = 308); mean (+SD)
20 levels of 31.5 (+5.2); IQR of 28.5-33.9 (+5.5) ppb. Effect estimates for O3 in three-pollutant models
21 including O3, SO2, and NO2 yielded OR for O3-days of 2.74 (95% CI: 1.68, 4.48) per IQR days of
22 10-28 (+18) ppb. The effect estimates for SO2 and NO2 in the three-pollutant model were 1.33 (95%
23 CI: 0.85, 2.11) and 0.94 (95% CI: 0.68, 1.29) respectfully. Taking into account duration of residence
24 did not change the result. This study suggests that a higher asthma severity score is related to long-
25 term O3 exposure.
26 The interrelationships between variants in catalase (CAT) and myeloperoxidase (MPO) genes,
27 ambient pollutants, and acute respiratory illness were investigated in a national U.S. cohort (Wenten
28 et al., 2009, 597084). Health information, air pollution, and incident respiratory-related school
29 absences were ascertained in January-June 1996 for 1,136 Hispanic and non-Hispanic white U.S.
30 elementary schoolchildren as part of the prospective Air Pollution and Absence Study, a population
31 based cohort study conducted as part of the CHS. A related earlier study (Gilliland et al., 2001,
32 013232). which was discussed in the 2006 O3 AQCD, examined the effects of ambient air pollution
33 on school absenteeism due to respiratory illness without a genetic aspect to the study. In a new study
34 Wenten et al. (2009, 597084) hypothesized that variation in the level or function of these enzymes
35 would modulate respiratory illness risk, especially under high levels of oxidative stress. The joint
36 effect of these two genes on respiratory illness was examined. Risk of respiratory-related school
37 absences was elevated for children with the CAT (G/G) and MPO (G/A or A/A) genes (relative risk =
38 1.35, [95% CI: 1.03, 1.77]; P-interaction = 0.005). To assess effects of long-term average levels of
39 O3 on acute effects, communities were divided into high and low exposure groups by median levels
March 2011 7-8 DRAFT - DO NOT CITE OR QUOTE
-------
1 (46.9 ppb O3). The epistatic effect of CAT and MPO variants was evident in communities exhibiting
2 high ambient O3 levels (P-interaction = 0.03). The association of respiratory-illness absences with
3 functional variants in CAT and MPO that differ by air pollution levels illustrates the need to consider
4 genetic epistasis in assessing gene-environment interactions. In high O3 communities, CAT/MPO
5 genotypes that resulted in decreased oxidative stress were associated with a decreased risk of
6 respiratory related school absences compared with the CAT/MPO wild-type genotype (RR = 0.42,
7 [95% CI: 0.20, 0.89]).
7.2.2. Asthma Hospital Admissions and ED Vis its
8 The studies on O3-related hospital discharges and emergency department (ED) visits for
9 asthma and respiratory disease that were available in the 2006 O3 AQCD mainly looked at the daily
10 time metric. New studies evaluated long-term O3 exposure metrics providing a new line of evidence
11 that suggests a positive exposure-response relationship between first asthma hospital admission and
12 long-term O3 exposure.
13 An ecologic study (Moore et al, 2008, 196685) evaluated time trends in associations between
14 declining warm-season O3 concentrations and hospitalization for asthma in children in California's
15 South Coast Air Basin who ranged in age from birth to 19 years. Quarterly average concentrations
16 from 195 spatial grids, 10x10 km, were used. Ozone was the only pollutant associated with
17 increased hospital admissions over the study period. A linear relation was observed for asthma
18 hospital discharges (Moore et al., 2008, 196685). A matched case-control study (Karr et al., 2007,
19 090719) was conducted of infant bronchiolitis (ICD 9, code 466.1) hospitalization and two measures
20 of long-term pollutant exposure (the month prior to hospitalization and the lifetime average) for O3
21 in the South Coast Air Basin of southern California among 18,595 infants born between 1995 and
22 2000. Ozone was associated with reduced risk in the single-pollutant model, but this relation did not
23 persist in multi-pollutant models.
24 In a cross-sectional study, Meng et al. (2010, 594252) examined associations between air
25 pollution and asthma morbidity in the San Joaquin Valley in California by using the 2001 California
26 Health Interview Survey data from subjects ages 1 to 65+ who reported physician-diagnosed asthma
27 (n = 1502). Subjects were assigned annual average concentrations for O3 based on residential ZIP
28 code and the closet air monitoring station within 8 km but did not have data on duration of residence.
29 Multi-pollutant models for O3 and PM did not differ substantially from single-pollutant estimates,
30 indicating that pollutant multi-collinearity is not a problem in these analyses. The authors reported
31 increased asthma-related ED visits or hospitalizations for O3 (OR 1.49; [95% CI: 1.05, 2.11] per
32 10 ppb) for all ages. Positive associations were obtained for symptoms but 95% CIs included null
33 values. Associations for symptoms for adults (ages 18 +) were observed (OR 1.40; [95% CI: 1.02,
34 1.91} per 10 ppb).
35 Associations between air pollution and poorly controlled asthma among adults in Los Angeles
36 and San Diego Counties, were investigated using the California Health Interview Survey data
37 collected between November 2000 and September 2001 (Meng et al., 2007, 093275). Each
March 2011 7-9 DRAFT - DO NOT CITE OR QUOTE
-------
1 respondent was assigned an annual average concentration measured at the nearest station within
2 5 miles of the residential cross-street intersection. Poorly controlled asthma was defined as having
3 daily or weekly asthma symptoms or at least one ED visit or hospitalization because of asthma
4 during the past 12 months. This cross-sectional study reports an OR of 3.34 (95% CI: 1.01, 11.09)
5 for poorly controlled asthma when comparing those 65 years of age and older above the 90th
6 percentile (28.7 ppb) level to those below that level. Multi-pollutant analysis produced similar
7 results.
8 Evidence associating long-term O3 exposure to first asthma hospital admission in a
9 concentration-response relationship is provided in a retrospective cohort study (Lin et al, 2008,
10 196680). This study investigated the association between chronic exposure to O3 and childhood
11 asthma admissions (defined as a principal diagnosis of ICD9, code 493) by following a birth cohort
12 of 1,204,396 eligible births born in New York State during 1995-1999 to first asthma admission or
13 until 31 December 2000. There were 10,429 (0.87%) children admitted to the hospital for asthma
14 between 1 and 6 years of age. The asthma hospitalization rate in New York State in 1993 was 2.87
15 per 1,000 (Lin et al., 1999, 377917). Three indicators (all 8-h max from 10:00 a.m. to 6:00 p.m.)
16 were used to define chronic O3 exposure: (1) mean concentration during the follow-up period
17 (41.06 ppb); (2) mean concentration during the O3 season (50.62 ppb); and (3) proportion of follow-
18 up days with O3 levels >70 ppb. In this study the authors aimed to predict the risk of having asthma
19 admissions in a birth cohort, but the time to the first admission in children that is usually analyzed in
20 survival models was not their primary interest. The effects of co-pollutants were assessed and
21 controlled for using the Air Quality Index (AQI). Interaction terms were used to assess potential
22 effect modifications. A positive association between chronic exposure to O3 and childhood asthma
23 hospital admissions was observed indicating that children exposed to high O3 levels over time are
24 more likely to develop asthma severe enough to be admitted to the hospital. The various factors were
25 examined and differences were found for younger children (1-2 years), poor neighborhoods,
26 Medicaid/self-paid births, geographic region and others. As shown in Figure 7-3, positive
27 concentration-response relationships were observed. Asthma admissions were significantly
28 associated with increased O3 levels for all chronic exposure indicators (ORs, 1.16-1.68). When
29 estimating the O3 effect using the exceedance proportion, an increase was observed (OR 1.68; [95%
30 CI: 1.64, 1.73]) in hospital admissions with an IQR (2.51%) increase in O3. A proportional hazards
31 model for the New York City data was run as a sensitivity analysis and it yielded similar results
32 between asthma admissions and chronic exposure to O3 ( HR from the Cox model: HR: 1.14, [95%
33 CI: 1.124, 1.155] is similar to OR from the logistic model: 1.16 (95% CI: 1.15, 1.17) (Lin, personal
34 communication, 2010, 676733). Thus, this study provides evidence associating long-term O3
35 exposure to first asthma hospital admission in a concentration-response relationship.
March 2011 7-10 DRAFT - DO NOT CITE OR QUOTE
-------
i i Low exposure 0-33%
3.0
2.5
3 2.0
Hi
r^
g 1.0
0.5
n
' i Medium exposure 34-66%
^M High exposure ;> 67%
2.06
(1.87 2.27)
1.69 .64
1.43 (1.52-1.80) (1.48-1.82)
(1.29-1.58)
T~
1.00
(ref)
1
T
1
r^
1.00
(ref) |
T
1
New York City
Other NYS regions
Regions
Source: Lin etal. (2008, 196680):fLin. personal communication, 2010, 6767331.
Figure 7-3. Ozone-asthma concentration-response relationship using the mean concentration
during the entire follow-up period adjusted for child's sex, age, birth weight, and
gestational age; maternal race, ethnicity, age, education, insurance, and smoking
status during pregnancy; and regional poverty level and temperature.
OR's by low, medium, and high exposure are shown for New York City (NYC: low [37.3 ppb],
medium [37.3 - 38.11] ppb, high [38.11 + ppb]) and other New York State regions
(Other NYS regions: low [42.58 ppb], medium [42.58-45.06 ppb], high [45.06+ ppb])
for first asthma hospital admission.
7.2.3. Pulmonary Structure and Function
1 The definitive 8-year follow-up analysis of the first cohort of the CHS (Gauderman et al,
2 2004, 056569) provided little evidence that long-term exposure to ambient O3 at current levels was
3 associated with significant deficits in the growth rate of lung function in children. A later CHS study
4 (Islam et al., 2007, 090697) examined relationships between air pollution, lung function, and new
5 onset asthma and reported no substantial differences in the effect of lung function between "high-"
6 and "low-" O3 communities. Ozone concentrations from the least to most polluted communities
7 (mean annual average of 8-h avg O3) ranged from 30 to 65 ppb, whereas the ranges observed for the
8 other pollutants had four- to eightfold differences in concentrations. In a more recent CHS study,
9 Breton et al. (2011, 687660) hypothesized that genetic variation in genes on the glutathione
10 metabolic pathway may influence the association between ambient air pollutant exposures and lung
11 function growth in children. They investigated whether genetic variation in glutathione genes GSS,
12 GSR, GCLC, and GCLM was associated with lung function growth in healthy children using data
13 collected on 2,106 children over an 8-year time-period as part of the Children's Health Study. Breton
14 et al. (2011, 687660) found that variation in the GSS locus was associated with differences in
15 susceptibility of children for lung function growth deficits associated with NO2, PMio, PM2s,
March 2011
7-11
DRAFT - DO NOT CITE OR QUOTE
-------
1 elemental carbon, organic carbon, and O3. The negative effects of air pollutants were largely
2 observed within participants who had a particular GSS haplotype. The effects ranged from -124.2 to
3 -149.1 mL for FEVi, -92.9 to -126.7 mL for FVC and -193.9 to -277.9 mL/s for MMEF for all
4 pollutants except O3, for which some positive associations were reported: 25.9 mL for FEVi; 0.1 mL
5 for FVC, and 166.5 mL/s for MMEF. Ozone did show larger decreases in lung function in children
6 without this haplotype, when compared to the other pollutants with values of -76.6 mL for FEVi,
7 -17.2 mL for FVC, and -200.3 mL/s for MMEF, but only MMEF was statistically significant.
8 As discussed in the 2006 O3 AQCD, a study of freshman students at the University of
9 California, Berkeley reported that lifetime exposure to O3 was associated with decreased measures of
10 small airways (<2 mm) function (FEF75 and FEF25_75) (Tager et al., 2005, 087538). There was an
11 interaction with the FEF25-75/FVC ratio, a measure of intrinsic airway size. Subjects with a large ratio
12 were less likely to have decreases in FEF75 and FEF25-75 for a given estimated lifetime exposure to
13 O3. Kinney and Lippmann (2000, 011913) examined 72 nonsmoking adults (mean age 20 years)
14 from the second-year class of students at the U.S. Military Academy in West Point, NY, and reported
15 results that appear to be consistent with a seasonal decline in lung function that may in part be due to
16 O3 exposures. Ilhorst et al. (2004, 055608) examined 2,153 children with a median age of 7.6 years
17 and reported summer pulmonary function results which indicated that significantly lower FVC and
18 FEVi increases were associated with higher O3 exposures in the summer, but not in the winter. Semi-
19 annual mean O3 concentrations ranged from 22 to 54 ppb during the summer and 4 to 36 ppb during
20 the winter. However, over a 3.5-year period Ilhorst et al. (2004, 055608) found no associations
21 between increases in lung function and mean summer O3 levels for FVC and FEVi, in contrast to the
22 significant seasonal effects. Frischer et al. (1999, 001037) showed results similar to the Ilhorst et al.
23 (2004, 055608) study.
24 Mortimer et al. (2008, 122163; 2008, 187280) examined the association of prenatal and
25 lifetime exposures to air pollutants with pulmonary function and allergen sensitization in a subset of
26 asthmatic children (ages 6-11) included in the Fresno Asthmatic Children's Environment Study
27 (FACES). Monthly means of pollutant levels for the years 1989-2000 were created and averaged
28 separately across several important developmental time-periods, including: the entire pregnancy,
29 each trimester, the first 3 years of life, the first 6 years of life, and the entire lifetime. In the first
30 analysis (Mortimer et al., 2008, 122163). negative effects on pulmonary function were found for
31 exposure to PMi0, NO2, and CO during key neonatal and early life developmental periods. The
32 authors did not find a negative effect of exposure to O3 within this cohort. In the second analysis
33 (Mortimer et al., 2008, 187280). sensitization to at least one allergen was associated, in general, with
34 higher levels of CO and PMi0 during the entire pregnancy and second trimester, and higher PMi0
35 during the first 2 years of life. Lower exposure to O3 during the entire pregnancy or second trimester
36 was associated with an increased risk of allergen sensitization. Although the pollutant metrics across
37 time periods were correlated, the strongest associations with the outcomes were observed for
38 prenatal exposures. Though it may be difficult to disentangle the effect of prenatal and postnatal
39 exposures, the models from this group of studies suggest that each time period of exposure may
March 2011 7-12 DRAFT - DO NOT CITE OR QUOTE
-------
1 contribute independently to different dimensions of school-aged children's pulmonary function. For
2 4 of the 8 pulmonary-function measures (FVC, FEVi, PEF, FEF25_75), prenatal exposures were more
3 influential on pulmonary function than early-lifetime metrics, while, in contrast, the ratio of
4 measures (FEVi/FVC and FEF25-75/FVC) were most influenced by postnatal exposures. When
5 lifetime metrics were considered alone, or in combination with the prenatal metrics, the lifetime
6 measures were not associated with any of the outcomes. This suggests that the timing of the O3
7 exposure may be more important than the overall dose, and prenatal exposures are not just markers
8 for lifetime or current exposures.
9 Latzin et al. (2009, 195721) examined whether prenatal exposure to air pollution was
10 associated with lung function changes in the newborn. Tidal breathing, lung volume, ventilation
11 inhomogeneity and eNO were measured in 241 unsedated, sleeping neonates (age = 5 weeks).
12 Consistent with the previous studies, no association was found for prenatal exposure to O3 and lung
13 function.
14 In a cross-sectional study of adults, Qian et al. (2005, 093283) examined the association of
15 long-term exposure to O3 and PMi0 with pulmonary function from data of 10,240 middle-aged
16 subjects who participated in the Atherosclerosis Risk in Communities (ARIC) study in four U.S.
17 communities. A surrogate for long-term O3 exposure from daily data was determined at the
18 individual level. Ozone was significantly and negatively associated with measures of pulmonary
19 function.
20 To determine the extent to which long-term exposure to outdoor air pollution accelerates adult
21 decline in lung function, Forbes et al. (2009, 595425) studied the association between chronic
22 exposure to outdoor air pollution and lung function in approximately 42,000 adults aged 16 and
23 older who were representatively sampled cross-sectionally from participants in the Health Survey for
24 England (1995, 1996, 1997, and 2001). FEVi was not associated with O3 concentrations. In contrast
25 to the results for PMi0, NO2, and SO2; combining the results of all the survey years showed that a
26 5-ppb difference in O3 was counter-intuitively associated with a higher FEVi by 22 mL.
27 In a prospective cohort study consisting of school-age, non-asthmatic children in Mexico City
28 (n = 3,170) who were 8 years of age at the beginning of the study, Roj as-Martinez et al. (2007,
29 091064) evaluated the association between long-term exposure to O3, PMi0 and NO2 and lung
30 function growth every 6 months from April 1996 through May 1999. Exposure data were provided
31 by 10 air quality monitor stations located within 2 km of each child's school. Over the study period,
32 8-h O3 concentrations ranged from 60 ppb (SD, ±25) in the northeast area of Mexico City to 90 ppb
33 (SD, ±34) in the southwest, with an overall mean of 69.8 ppb. In multi-pollutant models, an IQR
34 increase in mean O3 concentration of 11.3 ppb was associated with an annual deficit in FEVi of
35 12 mL in girls and 4 mL in boys. Single-pollutant models showed an association between ambient
36 pollutants (O3, PMi0 and NO2) and deficits in lung function growth. While the estimates from
37 co-pollutant models were not substantially different than single pollutant models, independent effects
38 for pollutants could not be estimated accurately because the traffic-related pollutants were correlated.
39 To reduce exposure misclassification, microenvironmental and personal exposure assessments were
March 2011 7-13 DRAFT - DO NOT CITE OR QUOTE
-------
1 conducted in a randomly selected subsample of 60 children using passive O3 samplers. Ozone
2 concentrations were correlated (p < 0.05) with the measurements obtained from the fixed-site air
3 monitoring stations.
4 In the 2006 O3 AQCD, few studies had investigated the effect of chronic O3 exposure on
5 pulmonary function. The strongest evidence was for seasonal effects of extended O3 exposures on
6 lung function in children, i.e., reduced lung function growth being associated with higher ambient O3
7 levels. Longer-term studies, investigating the association of chronic O3 exposure on yearly lung
8 function such as the CHS, were inconclusive. Thus for new studies for pulmonary function, in one
9 study where O3 and other pollutant levels were higher (90 ppb at high end of the range) than those in
10 the CHS, a relationship between O3 concentration and pulmonary function declines was observed in
11 school-aged children. Two studies of adult cohorts provide mixed results where long-term exposures
12 were at the high end of the range with levels of 49.5 ppb in one study and 27 ppb IQR in the other.
13 Thus there is little new evidence to build upon the very limited studies from the 2006 O3 AQCD.
7.2.3.1. Evidence from Toxicological Studies
14 As reviewed in the 1996 and 2006 O3 AQCDs (U.S. EPA, (1996, 017831). (2006, 088089)).
15 considerable controversy surrounds the extrapolation of data generated by rodent toxicology studies
16 to the understanding of adverse health effects observed in humans, as documented by epidemiology
17 and controlled exposure studies. Chief among these data extrapolation issues are the differences
18 between rodent and human respiratory physiology, cellular makeup, dosimetry, and morphometry.
19 Unique among the six NAAQS criteria pollutants, however, O3-inhalation studies have been
20 performed in non-human primates whose respiratory system most closely resembles that of the
21 human. A long series of studies have used non-human primates to examine the effect of O3 alone or
22 in combination with an inhaled allergen, house dust mite antigen, on morphology and lung function.
23 These studies, by Plopper and colleagues, have demonstrated changes in pulmonary function and
24 airway morphology in adult and infant non-human primates repeatedly exposed to environmentally
25 relevant concentrations of O3 (Carey et al, 2007, 195752: Chang MM-J; Wu et al, 1998, 011983:
26 Chen et al., 2003, 035576: Duan et al., 1993, 086326: Duan et al., 1996, 080791: Evans et al., 2003,
27 048167: Evans et al., 2004, 596379: Fanucchi et al., 2000, 012284: Fanucchi et al., 2006, 096491:
28 Fujinaka et al., 1985, 040278: Harkema et al., 1987, 041496: Harkema et al., 1987, 040816:
29 Harkema et al., 1993, 039794: Hatch et al., 1994, 076120: Hyde et al., 1989, 094057: Hyde et al.,
30 1999, 015124: Joad et al., 2000, 012984: Joad et al., 2006, 596390: Joad et al., 2008, 596391: Larson
31 et al., 2004, 057062: Lee et al., 1998, 054473: Moffatt et al., 1987, 040841: Plopper and Schelegle,
32 1997, 656713: Plopper et al., 1991, 042617: Plopper et al., 1998, 087203: Schelegle et al., 2003,
33 053778: Tran et al., 2004, 628626: Tucker et al., 1993, 056406: Wilson et al., 1984, 040044: Wu et
34 al.. 1999.012089V
35 Since the 1996 and 2006 O3 AQCDs, the initial observations in adult non-human primates
36 have been expanded in a series of experiments using infant rhesus monkeys repeatedly exposed to
37 0.5 ppm O3 starting at 1 month of age (Plopper et al., 2007, 596412). Many of the observations
March 2011 7-14 DRAFT - DO NOT CITE OR QUOTE
-------
1 found in adult monkeys have also been noted in infant rhesus monkeys, although a direct comparison
2 of the degree of adverse effects between adult and infant monkeys has not been reported. In terms of
3 pulmonary function changes, after several episodic exposures of infant monkeys to O3 (each cycle:
4 5 days of 0.5 ppm O3 at 8 h/day, followed by 9 days of filtered air exposures), they observed more
5 than a doubling in the baseline airway resistance, which was accompanied by a small increase in
6 airway responsiveness to inhaled histamine (Schelegle et al, 2003, 053778). although neither
7 measurement was statistically different from filtered air control values. Exposure of animals to
8 inhaled house dust mite antigen alone also produced small but not statistically significant changes in
9 baseline airway resistance and airway responsiveness, whereas the combined exposure to both (O3 +
10 antigen) produced statistically significant and greater than additive changes in both functional
11 measurements. This non-human primate evidence, of an O3-induced change in airway
12 responsiveness, supports the biologic plausibility of long_term exposure to O3 contributing to the
13 adverse effects of asthma in children. To understand which conducting airways and inflammatory
14 mechanisms are involved in O3-induced airway hyperresponsiveness in the infant rhesus monkey, a
15 follow-up study examined airway responsiveness ex vivo in lung slices (Joad et al., 2006, 596390).
16 Using video microscopy to morphometrically evaluate the response of bronchi and respiratory
17 bronchioles to methacholine, (a bronchoconstricting agent commonly used to evaluate airway
18 responsiveness in asthmatics), the investigators observed differential effects for the two airway sizes.
19 While episodic exposure to O3 alone (0.5 ppm) had little effect on ex vivo airway responsiveness in
20 bronchi and respiratory bronchioles, exposure to dust mite antigen alone produced airway
21 hyperresponsiveness in the large bronchi, whereas O3 + antigen produced significant increases in
22 airway hyperresponsiveness only in the respiratory bronchioles. These results suggest that ozone's
23 effect on airway responsiveness occurs predominantly in the smaller bronchioles.
24 The functional changes in the conducting airways of infant rhesus monkeys exposed to either
25 O3 alone or O3 + antigen were accompanied by a number of cellular and morphological changes,
26 including a significant fourfold increase in eosinophils, (a cell type important in allergic asthma), in
27 the bronchoalveolar lavage of infant monkeys exposed to O3 alone. Thus, these studies demonstrate
28 both functional and cellular changes in the lung of infant monkeys after cyclic exposure to 0.5 ppm
29 O3. This concentration, while higher than those used in controlled human exposure studies, provides
30 relevant information to understanding the adverse effects of ambient O3 exposure on the respiratory
31 tract of humans. No concentration-response data, however, are available from these non-human
32 primate studies.
33 In addition to these functional and cellular changes, significant structural changes in the
34 respiratory tract have been observed in infant rhesus monkeys exposed to O3. During normal
35 respiratory tract development, conducting airways increase in diameter and length in the infant
36 rhesus monkey. Exposure to O3 alone (5 days of 0.5 ppm O3 at 8 h/day, followed by 9 days of
37 filtered air exposures for 11 cycles), however, markedly affected the growth pattern of distal
38 conducting airways (Fanucchi et al., 2006, 096491). Whereas the first alveolar outpocketing
39 occurred at airway generation 13 or 14 in filtered air-control infant monkeys, the most proximal
March 2011 7-15 DRAFT - DO NOT CITE OR QUOTE
-------
1 alveolarized airways occurred at an average of 10 airway generations in O3-exposed monkeys.
2 Similarly, the diameter and length of the terminal and respiratory bronchioles were significantly
3 decreased in O3-exposed monkeys. Importantly, the O3-induced structural pathway changes persisted
4 after recovery in filtered air for 6 months after cessation of the O3 exposures. These structural effects
5 were accompanied by significant increases in mucus goblet cell mass, alterations in smooth muscle
6 orientation in the respiratory bronchioles, epithelial nerve fiber distribution, and basement membrane
7 zone morphometry. These latter effects are significant because of their potential contribution to
8 airway obstruction and airway hyperresponsiveness which are central features of asthma.
9 As noted above, a significant increase in airway responsiveness to inhaled histamine occurred
10 in infant rhesus monkeys exposed to O3 + house dust mite antigen, but not to O3 alone (Schelegle et
11 al, 2003, 053778). To study the underlying mechanisms of this airway hyperresponsiveness, these
12 investigators evaluated the effect of exposure to O3 alone and in combination with (+) antigen on
13 non-specific airway responsiveness to methacholine at different airway generations. After exposure
14 to filtered air, O3, antigen, or O3 + antigen, the bronchi and respiratory bronchioles of 6-month-old
15 monkeys were challenged ex vivo with methacholine. Exposure to O3 alone had no significant effect
16 on airway responsiveness to methacholine in either airway, whereas O3 + antigen produced a
17 significant increase in airway responsiveness in the respiratory bronchioles but not the larger
18 bronchi.
19 Because many cellular and biochemical factors are known to contribute to allergic asthma, the
20 effect of exposure to O3 alone or O3 + antigen on immune system parameters was also examined in
21 infant rhesus monkeys. Mast cells, which contribute to asthma via the release of potent proteases,
22 were elevated in animals exposed to antigen alone but O3 alone had little effect on mast cell numbers
23 and the response of animals exposed to O3 + antigen was not different from that of animals exposed
24 to antigen alone; thus suggesting that mast cells played little role in the interaction between O3 and
25 antigen in this model of allergic asthma (Van Winkle et al., 2010, 670301). Increases in CD4+ and
26 CD8+ lymphocytes were observed at 6 months of age in the blood and bronchoalveolar lavage fluid
27 of infant rhesus monkeys exposed to O3 + antigen but not in monkeys exposed to either agent alone
28 (Miller et al., 2009, 596406). Activated lymphocytes (i.e., CD25+ cells) were morphometrically
29 evaluated in the airway mucosa and significantly increased in infant monkeys exposed to antigen
30 alone or O3 + antigen. Although O3 alone had no effect on CD25+ cells, it did alter the anatomic
31 distribution of CD25+ cells within the airways. Ozone had only a small effect on these sets of
32 immune cells and did not produce a strong interaction with an inhaled allergen in this non-human
33 primate model; more mechanistic studies are necessary to understand a concentration-response effect
34 of O3 on allergic asthma.
35 In addition to alterations in the immune system, nervous system interactions with epithelial
36 cells are thought to play a contributing role to airway hyperresponsiveness. As noted in the 2006 O3
37 AQCD, exposure of infant rhesus monkeys altered the normal development of neural innervation in
38 the epithelium of the conducting airways (Larson et al., 2004, 057062). Whereas, a significant
39 reduction in airway innervation occurred after exposure to O3 alone, a significantly greater reduction
March 2011 7-16 DRAFT - DO NOT CITE OR QUOTE
-------
1 occurred in monkeys exposed to O3 + antigen. This reduction in overall airway innervation was
2 accompanied, however, by an increase in the abundance of protein gene product 9.5, a nonspecific
3 neural marker. Significant increases in protein gene product 9.5 were still observed in O3 alone- and
4 O3 + antigen-exposed infant monkeys after a 6-month recovery protocol (Kajekar et al, 2007,
5 567661). Thus, in addition to structural, immune, and inflammatory effects, exposure to O3 produces
6 alterations in airway innervation which may contribute to O3-induced exacerbation of asthma.
7 While the infant rhesus monkey studies examined the effect of long-term O3 exposure on
8 functional and morphologic development of the lung during early life, a small number of rodent
9 studies have examined the role of age in the response to O3. In mice, age-related differences in
10 O3-induced inflammation and the immediate-early gene response were observed. Johnston and
11 colleagues (2006, 097439) demonstrated that the lung damage produced by O3 occurred through
12 distinct (compared to inhaled endotoxin), early gene expression responses. Whereas c-fos and c-jun
13 mRNA levels were elevated in a concentration-dependent manner (1 and 2.5 ppm O3 for 4 hours) in
14 the lungs of C57BL/6 mice at 4, 10, and 56 days of age, the relative abundance of mRNA for TLR-4
15 (which has been shown to play a role in the pulmonary response to inhaled O3 (Hollingsworth et al.,
16 2010, 635786: Kleeberger et al., 2001, 016163)). was induced in the lungs of 10- and 56-day old but
17 not 4-day-old mice. Similar age-related differences in response were observed with inhaled
18 endotoxin, thus suggesting that the murine lung responds differently throughout the postnatal stage
19 of development. A study by Vancza et al. (2009, 596419) also demonstrated age-related differences
20 in the pulmonary response of mice to O3. Significantly greater inflammatory changes were observed
21 in neonatal (15 to 16 days old) compared to adult (15 week old) mice. Because this increase in
22 neonatal response was seen only in a subset of the 8 inbred mouse strains exposed to 0.8 ppm O3 for
23 5 hours, this strain-dependency suggests that genetic host factors play a role in age-related
24 differences in response to O3. Thus, these rodent studies suggest that the response to O3 in the
25 neonatal period is dependent on which postnatal day(s) the exposure occurs, as would be expected in
26 a rapidly developing mammalian lung.
27 Collectively, evidence from animal studies strongly suggests that chronic O3 exposure is
28 capable of damaging the distal airways and proximal alveoli, resulting in lung tissue remodeling -
29 leading to apparent irreversible changes. Compromised pulmonary function and structural changes
30 due to persistent inflammation may exacerbate the progression and development of chronic lung
31 disease. These findings offer some insight into potential biological mechanisms for the suggested
32 association between seasonal O3 exposure and reduced lung function development in children as
33 observed in epidemiologic studies.
7.2.4. Pulmonary Inflammation, Injury, and Oxidative Stress
34 The 2006 O3 AQCD stated that the extensive human clinical and animal toxicological
35 evidence, together with the limited epidemiologic evidence available, suggests a causal role for O3 in
36 inflammatory responses in the airways. Though the majority of recent studies focus on short-term
March 2011 7-17 DRAFT - DO NOT CITE OR QUOTE
-------
1 exposures, several epidemiologic and toxicology studies of long-term exposure add to observations
2 of O3-induced inflammation and injury.
3 Inflammatory markers and peak expiratory pulmonary function were examined in 37 allergic
4 children with physician-diagnosed mild persistent asthma in a highly polluted urban area in Italy and
5 then again 7 days after relocation to a rural location with significantly lower pollutant levels
6 (Renzetti et al., 2009, 199834). The authors observed a fourfold decrease in nasal eosinophils and a
7 statistically significant decrease in fractional exhaled nitric oxide along with an improvement in
8 lower airway function. Several pollutants were examined, including PMi0, NO2, and O3, though
9 pollutant-specific results were not presented. These results are consistent with studies showing that
10 traffic-related exposures are associated with increased airway inflammation and reduced lung
11 function in children with asthma and contribute to the notion that this negative influence may be
12 rapidly reversible. Exhaled NO (eNO) has been shown to be a useful biomarker for airway
13 inflammation in large population-based studies (Linn et al., 2009, 597363). Thus, while the time
14 scale of 7 days between examinations for eNO needs to be evaluated for appropriateness, the results
15 suggest that inflammatory responses are reduced when O3 levels are decreased.
16 Chest radiographs (CXR) of 249 children in Mexico City who were chronically exposed to O3
17 and PM2 5 were analyzed by Calderon-Garciduenas et al. (2006, 091253). They reported an
18 association between chronic exposures to O3 and other pollutants and a significant increase in
19 abnormal CXR's and lung CTs suggestive of a bronchiolar, peribronchiolar, and/or alveolar duct
20 inflammatory process, in clinically healthy children with no risk factors for lung disease. These CXR
21 and CT results should be viewed with caution because it is difficult to attribute effects to air
22 pollution exposure.
23 In a cross-sectional study, Wood et al. (2009, 597085) examined the association of outdoor air
24 pollution with respiratory phenotype (PiZZ type) in alpha 1-Antitrypsin deficiency (a-ATD) from the
25 U.K. a-ATD registry. In total, 304 PiZZ subjects underwent full lung function testing and
26 quantitative high-resolution computed tomography to identify the presence and severity of COPD -
27 emphysema. Mean annual air pollution data for 2006 was matched to the location of patients' houses
28 and used in regression models to identify phenotypic associations with pollution controlling for
29 covariates. Relative trends in O3 levels were assessed to validate use of a single year's data to
30 indicate long-term exposure and validation; data showed good correlations between modeled and
31 measured data (Stedman and Kent, 2008, 110057). Regression models showed that estimated higher
32 exposure to O3 exposure was associated with worse gas transfer and more severe emphysema, albeit
33 accounting for only a small proportion of the lung function variability. This suggests that a gene-
34 specific group demonstrates a long-term O3 exposure effect.
3 5 The similarities of non-human primates to humans make them attractive models in which to
36 study the effects of O3 on the respiratory tract. The nasal mucous membranes, which protect the more
37 distal regions of the respiratory tract, are susceptible to injury from O3. Carey et al. (2007, 195752)
38 conducted a study of O3 exposure in infant rhesus macaques, whose nasal airways closely resemble
39 that of humans. Monkeys were exposed either acutely for 5 days (8 h/day) to 0.5 ppm O3, or
March 2011 7-18 DRAFT - DO NOT CITE OR QUOTE
-------
1 episodically for several biweekly cycles alternating 5 days of 0.5 ppm O3 with 9 days of filtered air
2 (0 ppm O3), designed to mimic human exposure (70 days total). All monkeys acutely exposed to O3
3 had moderate to marked necrotizing rhinitis, with focal regions of epitheliar exfoliation, numerous
4 infiltrating neutrophils, and some eosinophils. The distribution, character, and severity of lesions in
5 episodically exposed monkeys were similar to that of acutely exposed animals. Neither group
6 exhibited mucous cell metaplasia proximal to the lesions, a protective adaptation observed in adult
7 monkeys exposed continuously to 0.3 ppm O3 in another study (Harkema et al., 1987, 040816). A
8 90-day exposure of rats to 0.8 ppm O3 (8 h/day) elicited significantly elevated pro-inflammatory
9 eicosanoids PGE2 and 12-HETE in BAL, but cytokine profiles did not differ from those of filtered
10 air-exposed rats (Schmelzer et al., 2006, 112994).
7.2.5. Allergic Responses
11 The association of air pollutants with childhood respiratory allergies was examined in the U.S.
12 using the 1999-2005 National Health Interview Survey of approximately 70,000 children, and
13 ambient air pollution data from the U.S. EPA, with monitors within 20 miles of each child's
14 residential block (Parker et al., 2009, 192359). The authors examined the associations between the
15 reporting of respiratory allergy or hay fever and summer exposure to O3, controlling for
16 demographic and geographic factors. Increased respiratory allergy/hay fever was associated with
17 increased O3 levels (adjusted OR per 10 ppb = 1.20; [95% CI: 1.15, 1.26]). These associations
18 persisted after stratification by urban-rural status, inclusion of multiple pollutants, and definition of
19 exposure by differing exposure radii; smaller samples within 5 miles of monitors were remarkably
20 similar to the primary results. No associations between the other pollutants and the reporting of
21 respiratory allergy/hay fever were apparent. Ramadour et al. (2000, 013259) reported no relationship
22 between O3 levels and rhinitis symptoms and hay fever. Hwang et al. (2006, 088971) report the
23 prevalence of allergic rhinitis (adjusted OR per 10 ppb = 1.05; [95% CI: 0.98, 1.12]) in a large cross-
24 sectional study in Taiwan. In a large cross-sectional study in France, Penard-Morand et al. (2005,
25 087951) reported a positive relationship between lifetime allergic rhinitis and O3 exposure in a two-
26 pollutant model with NO2. These studies related positive outcomes of allergic response and O3
27 exposure but with variable strength for the effect estimates. Nasal eosinophils, which participate in
28 allergic disease, were observed to decrease by fourfold in 37 atopic, mildly asthmatic children 7 days
29 after relocation from a highly polluted urban area in Italy to a rural location with significantly lower
30 pollutant levels (Renzetti et al., 2009, 199834).
31 Total IgE levels were related to air pollution levels in 369 adult asthmatics in five French
32 centers using generalized estimated equations (GEE) as part of the EGEA study described earlier
33 (Rage et al., 2009, 196719). Geostatistical models were performed on 4x4 km grids to assess
34 individual outdoor air pollution exposure that was assigned to subject's home address. Ozone
35 concentrations were positively related to total IgE levels and an increase of 5 ppb of O3 resulted in an
36 increase of 20.4% (95% CI: 3.0, 40.7) in total IgE levels. Nearly 75% of the subjects were atopic.
37 Two-pollutant models for O3 with NO2 were decreased by 25% while NO2 was decreased by 57%.
March 2011 7-19 DRAFT - DO NOT CITE OR QUOTE
-------
1 Associations were not sensitive to adjustment for covariates or the season of IgE measurements.
2 These cross-sectional results suggest that exposure to O3 may increase total IgE in adult asthmatics.
3 No toxicological studies of long-term exposure are available, but short-term exposure studies
4 in rodents and non-human primates demonstrate allergic skewing of immune responses and
5 enhanced IgE production. Due to the persistent nature of these responses, the short-term
6 toxicological evidence lends biological plausibility to the limited epidemiologic findings of an
7 association between long-term O3 exposure and allergic outcomes.
7.2.6. Host Defense
8 Short-term exposures to O3 cause decreases in host defenses against infectious lung disease in
9 animal models. However, acute O3-induced suppression of alveolar phagocytosis and immune
10 functions observed in animals appears to be transient and attenuated with continuous or repeated
1 1 exposures. Chronic exposures (weeks, months) of 0. 1 ppm do not cause greater effects on infectivity
12 than short exposures, due to defense parameters becoming reestablished with prolonged exposures,
13 although chronic exposure has been shown to slow alveolar clearance. In an older study (Jakab and
14 Bassett, 1990, 042196). no detrimental effects were seen with a 120-day exposure to 0.5 ppm O3 on
15 acute lung injury from influenza virus administered immediately before O3 exposure started. But
16 there were O3-enhanced postinfluenzal alveolitis and lung parenchymal changes. No new evidence
17 has become available to address the effects of long-term exposure on host defense mechanisms.
7.2.7. Respiratory Mortality
18 A limited number of epidemiologic studies have assessed the relationship between long-term
19 exposure to O3 and mortality. The 2006 O3 AQCD concluded that an insufficient amount of evidence
20 existed "to suggest a causal relationship between chronic O3 exposure and increased risk for
21 mortality in humans" (U.S. EPA, 2006, 088089). Though total and cardio-pulmonary mortality were
22 considered in these studies, respiratory mortality was not specifically considered. In the most recent
23 follow-up analysis of the ACS cohort (Jerrett et al, 2009, 194160). cardiopulmonary deaths were
24 subdivided into respiratory and cardiovascular, separately, as opposed to combined in the Pope et al.
25 (2002, 024689) work. A 10-ppb increment in exposure to O3 elevated the risk of death from
26 respiratory causes and this effect was robust to the inclusion of PM2 5. The association between
27 increased O3 concentrations and increased risk of death from respiratory causes was insensitive to
28 the use of a random-effects survival model allowing for spatial clustering within the metropolitan
29 area and state of residence, and to adjustment for several ecologic variables considered individually.
7.2.8. Summary and Causal Determination
30 The epidemiologic studies reviewed in the 2006 O3 AQCD detected no associations between
3 1 long-term O3 exposures and asthma-related symptoms, asthma prevalence, or allergy to common
32 aeroallergens among children after controlling for covariates. Little evidence was available to relate
March 201 1 7-20 DRAFT - DO NOT CITE OR QUOTE
-------
1 long-term exposure to current ambient O3 concentrations to deficits in the growth rate of lung
2 function in children. Additionally, limited evidence was available evaluating the relationship
3 between long-term O3 levels and pulmonary inflammation and other endpoints. From toxicological
4 studies, it appeared that O3-induced inflammation tapered off during long-term exposures, but that
5 hyperplastic and fibrotic changes remained elevated and in some cases even worsened after a
6 postexposure period in clean air. Episodic exposures were also known to cause more severe
7 pulmonary morphologic changes than continuous exposure (U.S. EPA, 2006, 088089).
8 The new epidemiologic evidence base consists of studies using a variety of designs and
9 analysis methods evaluating the relationship between long-term measures of exposure to ambient O3
10 and measures of respiratory morbidity conducted by different research groups in different locations.
11 See Table 7-1 for O3 concentrations associated with selected studies. The positive results from
12 various designs and locations support an association between long-term O3 concentrations and
13 respiratory morbidity.
14 New studies examined the relationship between long-term O3 exposure and new onset asthma
15 in children. Studies have provided evidence for a relationship between different genetic variants
16 (HMOX, GST, ARG) that, in combination with O3 exposure, are related to new onset asthma (Islam
17 et al, 2008, 097348: Islam et al, 2009, 196715: Salam et al, 2009, 596644). These studies involve
18 two separate cohorts in the CHS. These prospective cohort studies represent strong evidence because
19 they are methodologically rigorous epidemiology studies. The studies were conducted in 12
20 California communities. The stratified analysis for the two independent fourth-grade cohorts of the
21 study population recruited in 1993 and 1996 yielded consistent results and provides replication in
22 independent groups of children. Also, no meaningful interactions were observed between other air
23 pollutants such as PMi0 and genes.
Table 7-1. Summary of selected key new studies examining annual ozone exposure and respiratory
health effects
Study; Health Effect; Location
Islam et al.(2008, 097348); new-onset asthma;
CHS
Islam et al. (2009, 196715): new-onset asthma; CHS
Salam et al. (2009, 596644): childhood onset asthma; CHS
Lin et al. (2008, 196680): first asthma hospital admission;
New York State - 10 regions
Moore et al. (2008, 196685): asthma
hospital admissions; South Coast Basin
Meng et al. (2010, 594252):
asthma ED visits or hospitalizations;
San Joaquin Valley, CA
Lee et al. (2009, 199915):
bronchitic symptoms in asthmatic children; CHS
Rage et al. (2009, 196719):
asthma severity; five French cities
Mean Annual O3 Concentration (ppb)
55.2 high vs. 38.4 low communities
10:00 a.m. to 6:00 p.m.
55.2 high vs. 38.4 low communities
10:00 a.m. to 6:00 p.m.
O3 greater than or less than 50 ppb
Range of mean O3 concentrations over the
10 New York Regions 37.51 to 47.78
Median 87.8 ppb
Median 30.3 ppb
Above and below 50 ppb
Mean 30 ppb
O3 Range (ppb)
Percentiles
See left
See left
See left
See left
Range 28.6 to
199.9 ppb
25-75% range
27.1 to 34.0
See left
25th-75th
21 -36 ppb
March 2011
7-21
DRAFT - DO NOT CITE OR QUOTE
-------
Median«.W0:OOa.m.-6:OOp.m.
1 Studies using a cross-sectional design provide support for a relationship between long-term O3
2 exposure and health effects in asthmatics. A long-term O3 exposure study relates bronchitic
3 symptoms to TNF-308 genotype asthmatic children with ambient O3 exposure in the CHS (Lee et al.,
4 2009, 199915). A study relating asthma severity to long-term O3 exposure in five French cities
5 provides additional support to the notion that effects on asthma are related to long-term O3 exposure
6 (Rage et al., 2009, 196720). For the respiratory health of the general U.S. population, risk of
7 respiratory-related school absences was elevated for children with the CAT and MPO variant genes
8 related to communities with high ambient O3 levels (Wenten et al., 2009, 597084).
9 Chronic O3 exposure was related to first childhood asthma hospital admissions in a positive
10 concentration-response relationship in a New York State birth cohort (Lin et al., 2008, 196680). A
11 separate hospitalization cross-sectional study in San Joaquin Valley in California reports similar
12 findings (Meng et al., 2010, 594252). Another study relates asthma hospital admissions to quarterly
13 average O3 in the South Coast Air Basin of California (Moore et al., 2008, 196685).
14 Information from toxicological studies indicates that long term exposure to O3 during gestation
15 or development can result in irreversible morphological changes in the lung, which in turn can
16 influence pulmonary function. Studies by Plopper and colleagues have demonstrated changes in
17 pulmonary function and airway morphology in adult and infant non-human primates repeatedly
18 exposed to environmentally relevant concentrations of O3 (Fanucchi et al., 2006, 096491; Harkema
19 et al., 1987, 041496: Joad et al., 2006, 596390: Schelegle et al., 2003, 053778). This non-human
20 primate evidence of an O3-induced change in airway responsiveness supports the biologic
21 plausibility of long term exposure to O3 contributing to the adverse effects of asthma in children.
22 Results from epidemiologic studies examining long-term O3 exposure and pulmonary function
23 effects are inconclusive with some new studies relating effects at higher exposure levels. The results
24 from the CHS still remain as the definitive line of evidence. Other cross-sectional studies provide
25 mixed results.
26 The 2006 O3 AQCD states that the extensive human clinical and animal toxicological
27 evidence, together with the limited epidemiologic evidence available, suggests a causal role for O3 in
28 inflammatory responses in the airways. Though the majority of recent studies focus on short-term
29 exposures, several epidemiologic and toxicology studies of long-term exposure add to observations
30 of O3-induced inflammation and injury. Toxicological studies in rodents and non-human primates
31 indicate that chronic O3 exposure causes structural changes in the respiratory tract, and simulated
32 seasonal exposure studies suggest that such exposures might have cumulative impacts. The strongest
33 epidemiologic evidence for a relationship between long-term O3 exposure and respiratory morbidity
34 is provided by new studies that demonstrate associations between long-term measures of O3
35 exposure and new-onset asthma in children and increased respiratory symptom effects in asthmatics.
36 While there are currently a limited number of studies in this data base, these U.S. multi-community
37 prospective cohort studies are methodologically rigorous epidemiology studies. Asthma risk is
March 2011 7-22 DRAFT - DO NOT CITE OR QUOTE
-------
1 related to the important relationships between genetic variability, environmental O3 exposure, and
2 behavior. These relationships are complex. The genes, evaluated in these studies, are both key
3 candidates in the oxidative stress pathway and have been shown to play an important role in asthma
4 development. Reduced risk for asthma development is reported in some studies in children living in
5 low- O3 communities. Ozone levels in the studies (10:00 a.m. to 6:00 p.m.) ranged from 28.6 to
6 45.5 ppb in low O3 communities (mean = 38.4 ppb) and from 46.5 to 64.9 ppb in high O3
7 communities (mean = 55.2 ppb). Other studies in the new data base provide coherent evidence for
8 long-term O3 exposure and respiratory morbidity effects such as first asthma hospitalization and
9 respiratory symptoms in asthmatics. Studies considering other pollutants provide data suggesting
10 that the effects related to O3 are independent from potential effects of the other pollutants. Some
11 studies provide evidence for a positive concentration-response relationship. The above discussion of
12 the recent epidemiologic and toxicological data base provides a compelling case to support the
13 hypothesis that a relationship exists between long-term exposure to ambient O3 and measures of
14 respiratory morbidity. The 2006 O3 AQCD concluded the evidence was suggestive but inconclusive
15 at that time. The new epidemiological data base, combined with toxicological studies in rodents and
16 non-human primates, provides biologically plausible evidence that there is likely to be causal
17 relationship between long-term exposure to O3 and respiratory morbidity.
7.3. Cardiovascular Effects
7.3.1. Cardiovascular Disease
7.3.1.1. Cardiovas cular Epidemiology
18 Long-term exposure to O3 and its effects on cardiovascular morbidity were not considered in
19 the 2006 O3 AQCD (U.S. EPA, 2006, 088089). However, recent studies have assessed the chronic
20 effects of O3 exposure on cardiovascular morbidity (Chen et al., 2007, 145956; Chuang et al., 2011,
21 670846; Forbes et al., 2009, 190351). The association between O3 exposure and markers of lipid
22 peroxidation and antioxidant capacity was examined among 120 nonsmoking healthy college
23 students, aged 18-22 years, from the University of California, Berkeley (Feb-Jun 2002) (Chen et al.,
24 2007, 145956). By design, students were chosen that had experienced different geographic levels of
25 O3 over their lifetimes and during recent summer vacation in either greater Los Angeles (LA) or the
26 San Francisco Bay Area (SF). A marker of lipid peroxidation, 8-isoprostane (8-iso-PGF) in plasma,
27 was assessed. This marker is formed continuously under normal physiological conditions but has
28 been found at elevated concentrations in response to environmental exposures. A marker of overall
29 antioxidant capacity, ferric reducing ability of plasma (FRAP), was also measured. The lifetime O3
30 exposure estimates (estimated monthly average, ppb) did not show much overlap between the two
31 geographic areas [median (range): LA, 42.9 (28.5-65.3); SF, 26.9 (17.6-33.5)]. Estimated lifetime O3
32 exposure was related to 8-iso-PGF [(3 = 0.025 (pg/mL)/8-h ppb O3, p = 0.0007]. For the 17-ppb
March 2011 7-23 DRAFT - DO NOT CITE OR QUOTE
-------
1 cumulative lifetime O3 exposure difference between LA and SF participants, there was a
2 17.41-pg/mL (95% CI: 15.43, 19.39) increase in 8-iso-PGF. No evidence of association was
3 observed between lifetime O3 exposure and FRAP [(3 = -2.21 (pg/mL)/8-h ppb O3, p = 0.45]. The
4 authors note that O3 was highly correlated with PMi0-2.5 and NO2 in this study population; however,
5 their inclusion in the O3 models did not substantially modify the magnitude of the associations with
6 O3. Because the lifetime exposure results were supported by shorter-term exposure results from
7 analyses considering O3 concentrations up to 30 days prior to sampling, the authors conclude that
8 persistent exposure to O3 can lead to sustained oxidative stress and increased lipid peroxidation.
9 However, because there was not much overlap in lifetime O3 exposure estimates between LA and SF,
10 it is possible that the risk estimates involving the lifetime O3 exposures could be confounded by
11 unmeasured factors related to other differences between the two cities.
12 Forbes et al. (2009, 190351) used the annual average exposures to assess the relationship
13 between chronic ambient air pollution and levels of fibrinogen and C-reactive protein (CRP) in a
14 cross-sectional study conducted in England. Data were collected from the Health Survey of England
15 for 1994, 1998, and 2003. The sampling strategy was designed to obtain a representative sample of
16 the English population; however, due to small group sizes, only data from white ethnic groups were
17 analyzed. For analyses, the annual concentrations of O3 were averaged for the year of data collection
18 and the previous year with the exception of 1994 (because pollutant data were not available for
19 1993). Median O3 concentrations were 26.7 ppb, 25.4 ppb, and 28 ppb for 1994, 1998, and 2003,
20 respectively. Year specific adjusted effect estimates were created and combined in a meta-analysis.
21 No evidence of association was observed for O3 and levels of fibrinogen or CRP (e.g., the combined
22 estimates for the percent change in fibrinogen and CRP for a 10 ppb increase in O3 were -0.28 [95%
23 CI: -2.43, 1.92] and -3.05 [95% CI: -16.10, 12.02], respectively). Further research will be important
24 for understanding the effects, if any, of chronic O3 exposure on cardiovascular morbidity risk.
25 A study was performed in Taiwan to examine the association between long-term O3
26 concentrations and blood pressure and blood markers using the Social Environment and Biomarkers
27 of Aging Study (SEBAS) (Chuang et al., 2011, 670846). Individuals included in the study were
28 54 years of age and older. The mean annual O3 concentration during the study period was 22.95 ppb
29 (SD 6.76 ppb). Positive associations were observed between O3 concentrations and both systolic and
30 diastolic blood pressure [changes in systolic and diastolic blood pressure were 21.51mmHg (95% CI:
31 16.90, 26.13) and 20.56 mmHg (95% CI: 18.14, 22.97) per 8.95 ppb increase in O3, respectively).
32 Increased O3 concentrations were also associated with increased levels of total cholesterol, fasting
33 glucose, hemoglobin Ale, and neutrophils. No associations were observed between O3
34 concentrations and triglyceride and IL-6 levels. The observed associations were reduced when other
35 pollutants were added to the models.
7.3.1.2. Long-Term Cardiovas cular Toxicology
36 Three new studies have investigated the cardiovascular effects of long-term exposure to O3 in
37 animal models. In addition to the short-term effects described in Section 6.3.1, a recent study found
March 2011 7-24 DRAFT - DO NOT CITE OR QUOTE
-------
1 that ApoE"7" mice (6-14 weeks old) exposed to O3 (0.5 ppm) for 8 weeks (5 days/week, 8 h/day) had
2 enhanced aortic atherosclerotic lesion area compared to air exposed controls (Chuang et al., 2009,
3 197202). Chuang et al. (2009, 197202) not only provided evidence for increased atherogenesis in
4 susceptible mice, but also reported an elevated vascular inflammatory and redox state in wild-type
5 mice and infant primates (Section 6.3.1). This study is compelling in that it identifies biochemical
6 and cellular events responsible for transducing the airway epithelial reactions of O3 into
7 proinflammatory responses that are apparent in the extrapulmonary vasculature (Cole and Freeman,
8 2009, 597507).
9 Another recent study provides further evidence for increased vascular inflammation and
10 oxidation and long term effects in the extrapulmonary space. Rats episodically exposed to O3
11 (0.4 ppm) for 16 weeks (5 h/day, 1 day/week) presented marked increases in gene expression of
12 biomarkers of oxidative stress, thrombosis, vasoconstriction, and proteolysis (Kodavanti et al., In
13 Press, 666323). Ozone exposure upregulated aortic mRNA expression of heme oxygenase-1 (HO-1),
14 tissue plasminogen activator (tPA), plasminogen activator inhibitor-1 (PAI-1), von Willebrand factor
15 (vWf), thrombomodulin, endothelial nitric oxide synthase (eNOS), endothelin-1 (ET-1), matrix
16 metalloprotease-2 (MMP-2), matrix metalloprotease-3 (MMP-3), and tissue inhibitor of matrix
17 metalloprotease-2 (TIMP-2). In addition, O3 exposure depleted some cardiac mitochondrial
18 phospholipid fatty acids (C16:0 and C18:1), which may be the result of oxidative modifications. The
19 authors speculate that oxidatively modified lipids and proteins produced in the lung and heart
20 promote vascular pathology through activation of lectin-like oxidized-low density lipoprotein
21 receptor-1 (LOX-1). Activated LOX-1 induces expression of a number of the biomarkers induced by
22 O3 exposure and is considered pro-atherogenic. Both LOX-1 mRNA and protein were increased in
23 mouse aorta after O3 exposure. This study provides a possible pathway and further support to the
24 observed O3 induced atherosclerosis.
25 Vascular occlusion resulting from atherosclerosis can block blood flow through vessels
26 causing ischemia. The restoration of blood flow or reperfusion can cause injury to the tissue from
27 subsequent inflammation and oxidative damage. Ozone exposure (0.8 ppm for 28 or 56 days)
28 enhanced the sensitivity to myocardial ischemia-reperfusion (I/R) injury in Sprague-Dawley rats
29 while increasing oxidative stress levels and pro-inflammatory mediators and decreasing production
30 of anti-inflammatory proteins (Perepu et al., 2010, 385020). Both long- and short-term O3 exposure
31 decreased the left ventricular developed pressure, rate of change of pressure development, and rate
32 of change of pressure decay and increased left ventricular end diastolic pressure in isolated perfused
33 hearts. In this ex vivo heart model, O3 induced oxidative stress by decreasing SOD enzyme activity
34 and increasing malondialdehyde levels. Ozone also elicited a proinflammatory state evident by an
35 increase in TNF-a and a decrease in the anti-inflammatory cytokine IL-10. The authors conclude that
36 O3 exposure will result in a greater I/R injury.
March 2011 7-25 DRAFT - DO NOT CITE OR QUOTE
-------
7.3.2. Cardiac Mortality
1 A limited number of epidemiologic studies have assessed the relationship between long-term
2 exposure to O3 and mortality. The 2006 O3 AQCD concluded that an insufficient amount of evidence
3 existed "to suggest a causal relationship between chronic O3 exposure and increased risk for
4 mortality in humans" (U.S. EPA, 2006, 088089). Though total and cardio-pulmonary mortality were
5 considered in these studies, cardiovascular mortality was not specifically considered. In the most
6 recent follow-up analysis of the ACS cohort (Jerrett et al, 2009, 194160). cardiopulmonary deaths
7 were subdivided into respiratory and cardiovascular, separately, as opposed to combined in the Pope
8 et al. (2002, 024689) work. A 10-ppb increment in exposure to O3 elevated the risk of death from the
9 cardiopulmonary, cardiovascular, and ischemic heart disease. Inclusion of PM2s as a co-pollutant
10 attenuated the association with exposure to O3 for all of the cardiovascular endpoints to become null.
7.3.3. Summary and Causal Determination
11 Previous AQCDs did not address the cardiovascular effects of long-term O3 exposure due to
12 limited data availability. The evidence remains limited; however the emerging data is supportive of a
13 role for O3 in chronic cardiovascular diseases. Few epidemiologic studies have investigated
14 cardiovascular morbidity after long-term O3 exposure, and the majority only assessed cardiovascular
15 disease related biomarkers. A study on O3 and cardiovascular mortality reported no association after
16 adjustment for PM2s levels. Further epidemiologic studies on cardiovascular morbidity and mortality
17 after long-term exposure have not been published.
18 Toxicological evidence on long-term O3 exposure is also limited but three strong toxicological
19 studies have been published since the previous AQCD. These studies provide evidence for O3
20 enhanced atherosclerosis and I/R injury, corresponding with development of a systemic oxidative,
21 proinflammatory environment. Although questions exist for how O3 inhalation causes systemic
22 effects, a recent study proposes a mechanism for development of vascular pathology that involves
23 activation of LOX-1 by O3 oxidized lipids and proteins. This activation may also be responsible for
24 O3 induced changes in genes involved in proteolysis, thrombosis, and vasoconstriction. Taking into
25 consideration the positive toxicological studies reported, the generally limited body of evidence is
26 suggestive of a causal relationship between relevant long-term exposures to O3 and
27 cardiovascular effects.
7.4. Reproductive and Developmental Effects
28 Although the body of literature is growing, the research focusing on adverse birth outcomes is
29 limited when compared to the numerous studies that have examined the more well-established health
30 effects of air pollution. Among this small number of studies, various measures of birth weight and
31 fetal growth, such as low birth weight (LEW), small for gestational age (SGA), and intrauterine
March 2011 7-26 DRAFT - DO NOT CITE OR QUOTE
-------
1 growth restriction (IUGR), and preterm birth (<37-week gestation; [PTB]) have received more
2 attention in air pollution research, while congenital malformations are less studied.
3 Infants and fetal development processes may be particularly vulnerable to O3 exposure, and
4 although the physical mechanisms are not fully understood, several hypotheses have been proposed
5 involving direct effects on fetal health, altered placenta function, or indirect effects on the mother's
6 health (Bracken et al, 2003, 156288: Clifton et al, 2001, 156360: Maisonet et al., 2004, 156725:
7 Schatz et al., 1990, 156073: Sram et al., 2005, 087442). Study of these outcomes can be difficult
8 given the need for detailed exposure data and potential residential movement of mothers during
9 pregnancy. Air pollution epidemiologic studies reviewed in the 2006 O3 AQCD examined impacts on
10 birth-related endpoints, including intrauterine, perinatal, postneonatal, and infant deaths; premature
11 births; intrauterine growth retardation; very low birth weight (weight <1,500 grams) and low birth
12 weight (weight <2,500 grams); and birth defects. However, in the limited number of studies that
13 investigated O3, no associations were found between O3 and birth outcomes, with the possible
14 exception of birth defects.
15 Two recent articles have reviewed methodological issues relating to the study of outdoor air
16 pollution and adverse birth outcomes (Ritz and Wilhelm, 2008, 156914: Slama et al., 2008, 156985).
17 Some of the key challenges to interpretation of these study results include the difficulty in assessing
18 exposure as most studies use existing monitoring networks to estimate individual exposure to
19 ambient air pollution; the inability to control for potential confounders such as other risk factors that
20 affect birth outcomes (e.g., smoking); evaluating the exposure window (e.g., trimester) of
21 importance; and limited evidence on the physiological mechanism of these effects (Ritz and
22 Wilhelm, 2008, 156914: Slama et al., 2008, 156985). Although early animal studies (Kavlock et al.,
23 1980, 094043) found that exposure to O3 in the late gestation of pregnancy in rats led to some
24 abnormal reproductive performances for neonates, to date human studies have reported inconsistent
25 results for the association of ambient O3 on birth outcomes.
7.4.1. Effects on Sperm
26 A limited amount of research has been conducted to examine the association between air
27 pollution and male reproductive outcomes, specifically semen quality. To date, the epidemiologic
28 studies have considered various exposure durations before semen collection that encompass either
29 the entire period of spermatogenesis (i.e., 90 days) or key periods of sperm development that
30 correspond to epididymal storage, development of sperm motility, and spermatogenesis. In an
31 analysis conducted as part of the Teplice Program, 18-year-old men residing in the heavily polluted
32 district of Teplice in the Czech Republic were found to be at greater risk of having abnormalities in
33 sperm morphology and chromatin integrity than men of similar age residing in Prachatice, a less
34 polluted district (Selevan et al., 2000, 012578: Sram et al., 1999, 078127). A follow-up longitudinal
35 study conducted on a subset of the same men from Teplice revealed associations between total
36 episodic air pollution and abnormalities in sperm chromatin (Rubes et al., 2005, 078091). A
37 limitation of these studies is that they did not identify specific pollutants and their concentrations.
March 2011 7-27 DRAFT - DO NOT CITE OR QUOTE
-------
1 More recent epidemiologic studies conducted in the U.S. have also reported associations
2 between ambient air pollution and sperm quality for individual air pollutants, including O3 and
3 PM2 5. In a repeated measures study in Los Angeles, CA, Sokol et al. (2006, 098539) reported a
4 reduction in average sperm concentration during three exposure windows (0-9, 10-14, and
5 70-90 days before semen collection) associated with high ambient levels of O3 in healthy sperm
6 donors. This effect persisted under a joint additive model for O3, CO, NO2 and PMi0. The authors did
7 not detect a reduction in sperm count. Hansen et al. (2010, 594438) investigated the effect of
8 exposure to O3 and PM2 5 on sperm quality in three southeastern counties (Wake County, NC; Shelby
9 County, TN; Galveston County, TX). Outcomes included sperm concentration and count,
10 morphology, DNA integrity and chromatin maturity. Overall, the authors found both protective and
11 adverse effects, although some results suggested adverse effects on sperm concentration, count and
12 morphology. There was evidence of an association between decreased sperm concentration and
13 count with O3, though these associations were not statistically significant.
14 The biological mechanisms linking ambient air pollution to decreased sperm quality have yet
15 to be determined, though O3-induced oxidative stress, inflammatory reactions, and the induction of
16 the formation of circulating toxic species have been suggested as possible mechanisms (Sokol et al.,
17 2006, 098539). Decremental effects on testicular morphology have been demonstrated in
18 toxicological studies with histological evidence of O3-induced depletion of germ cells in testicular
19 tissue and decreased seminiferous tubule epithelial layer. Jedlinska-Krakowska et al. (2006, 195640)
20 demonstrated histopathological evidence of impaired spermatogenesis (round spermatids/
21 spermatocytes, giant spermatid cells, and focal epithelial desquamation with denudation to the
22 basement membrane). The exposure protocol used five month old adult rats exposed to O3 as adults
23 (0.5 ppm, 5 h/day for 50 days). This degeneration could be rescued by vitamin E administration,
24 indicating an antioxidant effect. Vitamin C administration had no effect at low doses of ascorbic acid
25 and exacerbated the O3-dependent damage at high doses, as would be expected as vitamin C can be a
26 radical generator instead of an antioxidant at higher doses. In summary, this study provided
27 toxicological evidence of impaired spermatogenesis with O3 exposure that was rescued with certain
28 antioxidant supplementation.
29 Overall, there is limited epidemiologic evidence for an association with O3 concentration and
30 decreased sperm concentration. A recent toxicological study provides limited evidence for a possible
31 biological mechanism (histopathology showing impaired spermatogenesis) for such an association.
7.4.2. Effects on Reproduction
32 Evidence suggests that exposure to air pollutants during pregnancy is associated with adverse
33 birth outcomes, which has been attributed to the increased susceptibility of the fetus due to
34 physiologic immaturity. Gametes (i.e., ova and sperm) may be even more vulnerable, especially
35 outside of the human body, as occurs with assisted reproduction. Smokers require twice the number
36 of in vitro fertilization (IVF) attempts to conceive as non-smokers (Feichtinger et al., 1997, 625251).
37 suggesting that a preconception exposure can be harmful to pregnancy. A recent study used an
March 2011 7-28 DRAFT - DO NOT CITE OR QUOTE
-------
1 established national-scale, log-normal kriging method to spatially estimate daily mean
2 concentrations of criteria pollutants at addresses of women undergoing their first IVF cycle and at
3 their IVF labs from 2000 to 2007 in the northeastern U.S. (Legro et al, 2010, 597377). Increasing O3
4 concentration at the patient's address was significantly associated with an increased chance of live
5 birth during ovulation induction (OR 1.13, [95% CI: 1.05, 1.22] per 10 ppb increase), but with
6 decreased odds of live birth when exposed from embryo transfer to live birth (OR 0.79, [95% CI:
7 0.69, 0.90] per 10 ppb increase). After controlling for NO2 in a co-pollutant model, however, O3 was
8 no longer significantly associated with IVF failure. The results of this study suggest that exposure to
9 O3 during ovulation was beneficial (perhaps due to early conditioning to O3), whereas later exposure
10 to O3 (e.g., during gestation) was detrimental, and reduced the likelihood of a live birth.
11 In toxicological studies, reproductive success in rats appears unaffected by O3 exposure.
12 Ozone administration (continuous 0.4, 0.8 or 1.2 ppm O3) to CD-I mouse dams during the majority
13 of pregnancy (PD7-17, which excludes the pre-implantation period), led to no adverse effects on
14 reproductive success (proportion of successful pregnancies, litter size, sex ratio, frequency of still
15 birth, or neonatal mortality) (Bignami et al., 1994, 076063). There was a nearly statistically
16 significant increase in pregnancy duration (0.8 and 1.2 ppm O3). Initially, dam body weight (0.8 and
17 1.2 ppm), water consumption (0.4, 0.8 and 1.2 ppm O3) and food consumption (0.4, 0.8 and 1.2 ppm)
18 during pregnancy were decreased with O3 exposure but these deficits dissipated a week or two after
19 the initial exposure (Bignami et al., 1994, 076063). The anorexigenic effect of O3 exposure on the
20 pregnant dam appears to dissipate with time; the dams seem to adapt to the O3 exposure. In males,
21 data exist showing morphological evidence of altered spermatogenesis in O3 exposed animals
22 (Jedlinska-Krakowska et al. (2006, 195640). Some evidence suggests that O3 may affect
23 reproductive success when combined with other chemicals. Kavlock et al. (1979, 039228) showed
24 that O3 acted synergistically with sodium salicylate to increase the rate of pup resorptions after
25 midgestational exposure (1.0 ppm O3, GD9-12). At low doses of O3 exposure, toxicological studies
26 show reproductive effects to include a transient anorexigenic effect of O3 on gestational weight gain,
27 and a synergistic effect of O3 on salicylate-induced pup resorptions; other fecundity, pregnancy and
28 gestation related outcomes appear unaffected by O3 exposure. Collectively, there is very little
29 epidemiologic evidence for the effect of O3 on reproductive success, and the reproductive success in
30 rats appears to be unaffected in toxicological studies of O3 exposure.
7.4.3. Birth Weight
31 With birth weight routinely collected in vital statistics and being a powerful predictor of infant
32 mortality, it is the most studied outcome within air pollution-birth outcome research. Air pollution
33 researchers have analyzed birth weight as a continuous variable and/or as a dichotomized variable in
34 the form of LEW (<2,500 g [5 Ibs, 8 oz]).
35 Birth weight is primarily determined by gestational age and intrauterine growth, but also
36 depends on maternal, placental and fetal factors as well as on environmental influences. In both
37 developed and developing countries, LEW is the most important predictor for neonatal mortality and
March 2011 7-29 DRAFT - DO NOT CITE OR QUOTE
-------
1 is a significant determinant of postneonatal mortality and morbidity. Recent studies report that
2 infants who are smallest at birth have a higher incidence of diseases and disabilities, which continue
3 into adulthood (Hack and Fanaroff, 1999, 625952).
4 The strongest evidence for an effect of O3 on birth weight comes from the Children's Health
5 Study conducted in southern California. In this study, Salam et al. (2005, 087885) report that
6 maternal exposure to O3 averaged over the entire pregnancy was associated with reduced birth
7 weight for 24-h avg (39.3 g decrease [95% CI: -55.8, -22.8] in birth weight per 10 ppb and 8-h avg
8 (19.2-g decrease [95% CI: -27.7, -10.7] in birth weight per 10 ppb) O3 concentrations. This effect
9 was stronger for concentrations averaged over the second and third trimesters. PMi0, NO2 and CO
10 concentrations averaged over the entire pregnancy were not statistically significantly associated with
11 birth weight, though CO concentrations in the first trimester and PMi0 concentrations in the third
12 trimester were associated with a decrease in birth weight. Additionally, the authors observed a
13 concentration-response relationship of birth weight with 24-h avg O3 concentrations averaged over
14 the entire pregnancy that was clearest above the 30-ppb level (see Figure 7-4). Relative to the lowest
15 decile of 24-h avg O3, estimates for the next 5 lowest deciles were approximately -40 g to -50 g, with
16 no clear trend and with 95% confidence bounds that included zero. The highest four deciles of O3
17 exposure showed an approximately linear decrease in birth weight, and all four 95% CIs excluded
18 zero, and ranged from mean decreases of 74 grams to decreases of 148 grams.
50
0
-50
-100
-150
-200
-250
O3
03
a
Q
0 0
o
0
20 30 40
24-hr 03 (ppb)
50
Source: Salam et al. (2005, 0878851
Figure 7-4. Birthweight deficit by decile of 24-h avg ozone concentration averaged over the
entire pregnancy compared with the decile group with the lowest ozone exposure. D
Deficits are plotted against the decile-group-specific median ozone exposure. Error
bars represent 95% CIs. Indicator variables for each decile of ozone exposure
(except the least-exposed group) were included in a mixed model.
March 2011
7-30
DRAFT - DO NOT CITE OR QUOTE
-------
1 Several additional studies conducted in the U.S. and Canada also investigated the association
2 between ambient O3 concentrations and birth weight and found little evidence for an association.
3 Morello-Frosch et al. (2010, 670076) estimated ambient O3 concentrations throughout pregnancy
4 and for each trimester in the neighborhoods of women who delivered term singleton births between
5 1996 and 2006 in California. A 10-ppb increase in O3 averaged across the entire pregnancy was
6 associated with a 5.7-g decrease (95% CI: -6.6, -4.9) in birth weight when exposures were calculated
7 using monitors within 10 km of the maternal address at date of birth. When the distance from the
8 monitor was restricted to 3 km, the decrease in birth weight associated with a 10-ppb increase in O3
9 increased to 8.9 g (95% CI: -10.6, -7.1). These results persisted in co-pollutant models and in models
10 that stratified by trimester of exposure, SES, and race. Chen et al. (2002, 024945) used 8-h avg O3
11 concentrations to create exposure variables based on average maternal exposure for each trimester.
12 Ozone was not found to be related to birth weight in single-pollutant models, though the O3 effect
13 during the third trimester was borderline significant in a co-pollutant model with PMi0. Wilhelm and
14 Ritz (2005, 088668) extended previous analyses of term LEW (Ritz and Yu, 1999, 086976; Ritz et
15 al., 2000, 012068) to include the period 1994-2000. The authors examined varying residential
16 distances from monitoring stations to see if the distance affected risk estimation, exploring the
17 possibility that effect attenuation may result from local pollutant heterogeneity inadequately captured
18 by ambient monitors. As in their previous studies, the authors observed associations between
19 elevated concentrations of CO and PMi0 both early and late in pregnancy and risk of term LEW.
20 After adjusting for CO and/or PMi0 the authors did not observe associations between O3 and term
21 LEW in any of their models. Brauer et al. (2008, 156292) evaluated the impacts of air pollution (CO,
22 NO2, NO, O3, SO2, PM2 5, PM10) on birth weight for the period 1999-2002 using spatiotemporal
23 residential exposure metrics by month of pregnancy in Vancouver, BC. Quantitative results were not
24 presented for the association between O3 and LEW, though the authors observed associations that
25 were largely protective. Dugandzic et al. (2006, 088681) examined the association between LEW
26 and ambient levels of air pollutants by trimester of exposure among a cohort of term singleton births
27 from 1988-2000. Though there was some indication of an association with SO2 and PMi0, there were
28 no effects for O3.
29 Similarly, studies conducted in Australia, Latin America, and Asia report limited evidence for
30 an association between ambient O3 and measures of birth weight. In Sydney, Australia, Mannes et al.
31 (2005, 087895) found that O3 concentrations in the second trimester of pregnancy had small adverse
32 effects on birth weight (7.5-g decrease; [95 % CI: -13.8, 1.2] per 10 ppb), though this effect
33 disappeared when the analysis was limited to births with a maternal address within 5 km of a
34 monitoring station (87.7-g increase; [95% CI: 10.5, 164.9] per 10 ppb). Hansen et al. (2007, 090703)
35 reported that trimester and monthly specific exposures to all pollutants were not statistically
36 significantly associated with a reduction in birth weight in Brisbane, Australia. In Sao Paulo, Brazil,
37 Gouveia et al. (2004, 055613) found that O3 exhibited a small inverse relation with birth weight over
38 the third trimester (6.0-g decrease; [95% CI: -30.8, 18.8] per 10 ppb). Lin et al. (2004, 089503)
39 reported a positive, though not statistically significant, exposure-response relationship for O3 during
March 2011 7-31 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
the entire pregnancy in a Taiwanese study. In a study performed in Korea, Ha et al. (2001, 019390)
reported no O3 effect during the first trimester of pregnancy, but they found that during the third
trimester of pregnancy O3 was associated with LEW (RR 1.05 [95% CI: 1.02, 1.08] per 10 ppb).
Table 7-2. Brief summary of epidemiologic studies of birth weight
Study
Salam et al. (2005, 087885)
Morello-Frosch et al. (2010,
670076)
Chen et al. (2002, 024945)
Wilhelm and Ritz (2005,
088668)
Braueret al. (2008,
Dugandzic et al. (2006,
088681)
Mannes et al. (2005,
Hansen et al. (2007,
090703)
Gouveia et al. (2004,
Lin et al. (2004, 089503)
Ha et al. (2001 , 01 9390)
Location .. /-> / u\
Sample Size Mean O3 (ppb)
?srs- ^§:27-3
California, U.S. ~, . ~, c
(n=3,545,177) 24-h avg. 23.5
northern Nevada, US „ , . 07 0
(n=36,305) °"n' ^'"^
Los Angeles County, CA < h- 01 1 oo o
(n=136,134) -\-n.z-\.-\-zz.z
Vancouver, BC, Canada ~, . ,„_. , .
(n=70,249) 24-h avg. 14
(Nn=V74S2C804)a' ****** ** ^ 21
Sydney, Australia ,. h „,,,.,. •,< R
(n=1 38,056) 1-hmax.31.6
(rv^efyf UStral'a 8 h max: 26-7
Sao Paulo, Brazil , . _„,. ~. c
(n=1 79,460) 1 -Umax. 31 .5
Tfi?^"9 and Ta'Pei' 24-h av9: 1 5-86-
Taiwan 47 7R a
(n=92,288) ^'-'°
Seoul, Korea 8-h avg: 22.4-
(n=276,763) 23.3*
Exposure assessment
ZIP code level
Nearest Monitor
(within 10, 5, 3 km)
County level
Varying distances from
monitor
Nearest Monitor
(within 10 km)
Inverse Distance
Weighting (IDW)
Nearest Monitor
(within 25 km)
City-wide avg and
<5 km from monitor
City-wide avg
City-wide avg
Nearest monitor
(within 3 km)
City-wide avg
Effect Estimate (95% CI)
Entire pregnancy: -39.3 g
(-55.8, -22.8)
T1:-6.1 g(-16.8, 4.8)
T2: -20.0 g (-31 .7, -8.4)
T3: -20.7 g (-32.1, -9.3)
Entire pregnancy: -5.7 g
(-6.6, -4.9)
T1:-2.1 g (-2.9, -1.4)
T2: -2.3 g (-3.1, -1.5)
T3:-1.3g(-2.1,-0.6)
Entire pregnancy: 20.9 g (6.3,
35.5)
T1 : 23.4 g (-35.6 , 82.4)
T2: -1 9.4 g (-77.0, 38.2)
T3: 7.7 g (-50.9, 66.3)
T1: NR
T3:NR
6 weeks before birth: NR
Entire pregnancy: NR
First 30 days of pregnancy:
NR
Last 30 days of pregnancy:
NR
T1:NR
T3: NR
T1: 0.97 (0.81, 1.18)
T2: 1 .06 (0.87, 1 .27)
T3: 1.01 (0.83-1.24)
T1:-0.9g(-6.6, 4.8)
T2: -7.5 g (-13.8, 1.2)
T3: -4.5 g (-10.8, 1.8)
Last 30 days: -1.1 g (-5.6,
3.4)
T1 : 2.8 g (-10.5, 16.0)
T2:4.4g(-11.4, 20.1)
T3: 11. 3 g (-4.4, 27.1)
T1: -3.2 g (-25.6, 19)
T2: -0.2 g (-23.8, 23.4)
T3: -6.0 g (-30.8, -18.8)
Entire pregnancy: 1.13 (0.92,
1 .38)A
T1: 1.02(0.85, 1.22)A
T2: 0.93 (0.78, 1.1 2)A
T3: 1.05(0.87, 1.26)A
T1: 0.87 (0.81, 0.94)+
T3: 1.05 (1.02, 1.08)+
'Median
# Change in birthweight per 10 ppb change in O3
AOdds ratios of LBW; Highest quartile of exposure compared to lowest quartile of exposure
+Relative risk of LBW per 10 ppb change in O3
T1 = First Trimester, T2 = Second Trimester, T3 = Third Trimester
NR: No quantitative results reported
Table 7-2 provides a brief overview of the epidemiologic studies of birth weight. In summary,
only the Children's Health Study conducted in southern California (Salam et al., 2005, 087885)
provides strong evidence for an effect of ambient O3 on birth weight. The study by Morello-Frosch
March 2011
7-32
DRAFT - DO NOT CITE OR QUOTE
-------
1 et al. (2010, 670076). also conducted in California, provides support for the results of the Children's
2 Health Study. Additional studies conducted in the U.S., Canada, Australia, Latin America, and Asia
3 provide limited and inconsistent evidence to support the effect reported in the Children's Health
4 Study. The toxicological literature on the effect of O3 on birth weight is sparse. In some studies, the
5 reporting of birth weight may be avoided because birth weight can be confounded by decreased litter
6 size resulting from an increased rate of pup resorption (aborted pups) in O3 exposed dams. In one
7 toxicological study by Haro and Paz (1993, 044194). no differences in litter size were observed and
8 decreased birth weight in pups from dams who were exposed to Ippm O3 during pregnancy was
9 reported.
7.4.4. Preterm Birth
10 Preterm birth (PTB) is a syndrome (Romero et al., 2006, 625253) that is characterized by
11 multiple etiologies. It is therefore unusual to be able to identify an exact cause for each PTB. In
12 addition, PTB is not an adverse outcome in itself, but an important determinant of health status (i.e.,
13 neonatal morbidity and mortality). Although some overlap exists for common risk factors, different
14 etiologic entities related to distinct risk factor profiles and leading to different neonatal and
15 postneonatal complications are attributed to PTB and measures of fetal growth. Although both
16 restricted fetal growth and PTB can result in LEW, prematurity does not have to result in LEW or
17 growth restricted babies.
18 A major issue in studying environmental exposures and preterm birth is selecting the relevant
19 exposure period, since the biological mechanisms leading to preterm birth and the critical periods of
20 vulnerability are poorly understood (Bobak, 2000, 011448). Exposures proximate to the birth may be
21 most relevant if exposure causes an acute effect. However, exposure occurring in early gestation
22 might affect placentation, with results observable later in pregnancy, or cumulative exposure during
23 pregnancy may be the most important determinant. The studies reviewed have dealt with this issue in
24 different ways. Many have considered several exposure metrics based on different periods of
25 exposure. Often the time periods used are the first month (or first trimester) of pregnancy and the
26 last month (or 6 weeks) prior to delivery. Using a time interval prior to delivery introduces an
27 additional problem since cases and controls are not in the same stage of development when they are
28 compared. For example, a preterm infant delivered at 36 weeks is a 32-week fetus 4 weeks prior to
29 birth, while an infant born at term (40 weeks) is a 36-week fetus 4 weeks prior to birth.
30 Recently, investigators have examined the association of PTB with both short- and long-term
31 exposure periods. Time-series studies have been used to examine the association between air
32 pollution concentrations during the days immediately preceding birth. An advantage of these time-
33 series studies is that this approach can remove the influence of covariates that vary across individuals
34 over a short period of time. Retrospective cohort and case-control studies have been used to examine
35 long-term exposure periods, often averaging air pollution concentrations over months or trimesters
36 of pregnancy.
March 2011 7-33 DRAFT - DO NOT CITE OR QUOTE
-------
1 Reported studies fail to show consistency in pollutants and periods during pregnancy where an
2 effect occurs. For example, while some studies find the strongest effects associated with exposures
3 early in pregnancy, others report effects when the exposure is limited to the second or third trimester.
4 However, the effect of air pollutant exposure during pregnancy on PTB has a biological basis. There
5 is an expanding list of possible mechanisms that may explain the association between O3 exposure
6 and PTB. These include: decreased in utero oxygen supply leading to a reduction of oxygen carrying
7 capacity; changes in blood viscosity and disturbances of uterine blood flow; genetic mutations in
8 first trimester leading to placental abnormalities; complex vascular alterations leading to placental
9 abnormalities; disrupted implantation and placentation and suboptimal placental function; acute or
10 sustained inflammatory response; disturbances of the pituitary-adrenocortico-placental system; and
11 increased maternal susceptibility to infections.
12 Many studies of PTB compare exposure in quartiles, using the lowest quartile as the reference
13 (or control) group. No studies use a truly unexposed control group. If exposure in the lowest quartile
14 confers risk, than it may be difficult to demonstrate additional risk associated with a higher quartile.
15 Thus negative studies must be interpreted with caution.
16 Preterm birth occurs both naturally (idiopathic preterm), and as a result of medical
17 intervention (iatrogenicpreterm). Ritz et al. (2000, 012068; 2007, 096146) excluded all births by
18 Cesarean section to limit their studies to idiopathic preterm. No other studies attempted to
19 distinguish the type of preterm birth, although air pollution exposure maybe associated with only one
20 type. This is a source of potential effect misclassification.
21 A number of air pollution-birth outcome studies have investigated the possible association
22 between PTB and maternal exposure to O3. Most recently, Darrow et al. (2009, 195818) used vital
23 record data to construct a retrospective cohort of 476,489 births occurring between 1994 and 2004 in
24 5 central counties of metropolitan Atlanta. Using a time-series approach, the authors examined
25 aggregated daily counts of preterm birth in relation to ambient levels of CO, NO2, SO2, O3, PMi0,
26 PM25 and speciated PM measurements. This study investigated 3 gestational windows of exposure:
27 the first month of gestation, the final week of gestation, and the final 6 weeks of gestation. The
28 authors did not observe associations of preterm birth with O3.
29 A number of U.S. studies were conducted in southern California, and report somewhat
30 inconsistent results. Ritz et al. (2000, 012068) evaluated the effect of air pollution (CO, NO2, O3,
31 PMio) exposure during pregnancy on the occurrence of PTB in a cohort of 97,518 neonates born in
32 southern California between 1989 and 1993. The authors averaged pollutant measures taken at the
33 closest air-monitoring station over distinct periods, such as 1,2, 4, 6, 8, 12, and 26 weeks before
34 birth and the whole pregnancy period. Additionally, they calculated average exposures for the first
35 and second months of pregnancy. The authors found no consistent effects for O3 over any of the
36 pregnancy periods in single or multi-pollutant models. Wilhelm and Ritz (2005, 088668) extended
37 previous analyses of PTB (Ritz and Yu, 1999, 086976: Ritz et al., 2000, 012068) in California to
38 include 1994-2000. The authors examined varying residential distances from monitoring stations to
39 see if the distance affected risk estimation, because effect attenuation may result from local pollutant
March 2011 7-34 DRAFT - DO NOT CITE OR QUOTE
-------
1 heterogeneity inadequately captured by ambient monitors. The authors analyzed the association
2 between O3 exposure during varying periods of pregnancy and PTB, finding a positive association
3 between O3 levels in both the first trimester of pregnancy (RR 1.23 [95% CI: 1.06, 1.42] per 10 ppb
4 increase in 24-h avg O3) and the first month of pregnancy (results for first trimester exposure were
5 similar, but slightly smaller, quantitative results not presented) in models containing all pollutants.
6 No association was observed between O3 in the 6 weeks before birth and preterm delivery. Finally,
7 Ritz et al. (2007, 096146) conducted a case-control survey nested within a birth cohort and assessed
8 the extent to which residual confounding and exposure misclassification impacted air pollution effect
9 estimates. The authors calculated mean exposure levels for three gestational periods: the entire
10 pregnancy, the first trimester, and the last 6 weeks before delivery. Though positive associations
11 were observed for CO and PM25, no consistent patterns of increase in the odds of preterm birth for
12 O3 or NO2 were observed.
13 One study conducted in Canada evaluated the impacts of air pollution (including CO, NO2,
14 NO, O3, SO2, PM25, and PM10) on preterm births (1999-2002) using spatiotemporal residential
15 exposure metrics by month of pregnancy in Vancouver, BC (Brauer et al., 2008, 156292). The
16 authors did not observe consistent associations with any of the pregnancy average exposure metrics
17 except for PM2 5 for PTB. The O3 associations were largely protective, and no quantitative results
18 were presented for O3. Additionally, Lee et al. (2008, 195720) used time-series techniques to
19 investigate the short-term associations of O3 and PTB in London, England. In addition to exposure
20 on the day of birth, cumulative exposure up to 1 week before birth was investigated. The risk of
21 preterm birth did not increase with exposure to the levels of ambient air pollution experienced by
22 this population.
23 Conversely, two studies conducted in Australia and one from China do provide evidence for an
24 association between ambient O3 and PTB. Hansen et al. (2006, 089818) reported that exposure to O3
25 during the first trimester was associated with an increased risk of PTB (OR 1.38, [95% CI:
26 1.14, 1.69] per 10 ppb increase). Although the test for trend was significant due to the strong effect
27 in the highest quartile, there was not an obvious exposure-response pattern across the quartiles of O3
28 during the first trimester. The effect estimate was diminished and lost statistical significance when
29 PM10 was included in the model (OR 1.23, [95% CI: 0.97, 1.59] per 10 ppb increase). Maternal
30 exposure to O3 during the 90 days prior to birth showed a weak, positive association with PTB (OR
31 1.09, [95% CI: 0.85, 1.39] per 10 ppb increase). Jalaludin et al. (2007, 156601) found that O3 levels
32 in the month and three months preceding birth had a statistically significant association with PTB.
33 Ozone levels in the first trimester of pregnancy were associated with increased risks for PTBs (OR
34 1.15 [95% CI: 1.05, 1.24] per 10 ppb increase in 1-h max O3 concentration), and remained a
35 significant predictor of preterm birth in co-pollutant models (ORs between 1.07 and 1.10). ORs
36 increased for first month of pregnancy when restricted to within 5 km of a monitoring station (OR
37 1.60, [95% CI: 1.27, 2.03]), but did not show a cumulative effect for first 3 months of pregnancy
38 (OR 0.81, [95% CI: 0.67, 0.98]). Jiang et al. (2007, 093029) examined the acute effect of air
39 pollution on preterm birth, including risk in relation to levels of pollutants for a single day exposure
March 2011 7-35 DRAFT - DO NOT CITE OR QUOTE
-------
1 window with lags from 0 to 6 days before birth. An increase of 10 ppb of the 8-week average of O3
2 corresponded to 9.47 % (95% CI: 0.70, 18.7%) increase in PTBs. Increases in PTB were also
3 observed for PMi0, SO2, and NO2. The authors did not observe any significant acute effect of
4 outdoor air pollution on PTB among the 1-day acute time windows examined in the week before
5 birth.
6 Little data is available from toxicological studies; one study reported a nearly statistically
7 significant increase in pregnancy duration in mice when exposed to 0.8 or 1.2 ppm O3. This
8 phenomenon was most likely due to the anorexigenic effect of relatively high O3 concentrations
9 (Bignami et al, 1994, 076063).
Table 7-3. Brief summary of epidemiologic studies of PTB
Study
Location
Sample Size
Mean O3
(PPb)
Exposure
assessment
Effect Estimates (95% CI)
Darrow et al. (2009, Atlanta, GA
195818) (n=476,489)
Population-weighted
Q h ™^- /i/i 1 spatial averages
8-h max. 44.1 Barest Monlfor
(within 4 miles)
First month: 0.98 (0.97, 1.00)
Last week: 0.99 (0.98, 1.00)
Last 6 weeks: 1.00 (0.98, 1.02)
Ritz et al. (2000,
012068)
California, US
(n=97,158)
8 h: 36.9
<2 mi of monitor
First month: NR
Last 6 weeks: NR
Wilhelm and Ritz
(2005, 088668)
Los Angeles, CA
(n=106,483)
Varying distances to
monitor
First month: 1.23 (1.06, 1.42)
T1:NR
T2: 1.38(1.14, 1.66)
Last 6 weeks: NR
Ritz et al.
0961 46)
(2007,
Los Angeles,
(n=58,316)
CA
24-h
22.5
avg:
Nearest monitor to
ZIP code
Entire pregnancy: NR
T1 : 0.93 (0.82, 1 .06)
Last 6 weeks: NR
Brauer et al. (2008,
156292)
Vancouver, BC,
Canada
(n=70,249)
Nearest Monitor
24 haver 14 (within 10 km)
^4 n avg. 14 |nverse Distance
Weighting (IDW)
Entire pregnancy: NR
First 30 days of pregnancy: NR
Last 30 days of pregnancy: NR
T1: NR
T3: NR
Lee et al. (2008,
London, UK
24-h avg: NR 1 monitor
LagO: 1.00(1.00, 1.01)
Hansen et al. (2006,
089818)
Brisbane,
Australia
(n=28,200)
8-h max: 26.7 City-wide avg
T1: 1.39(1.15, 1.70)
T3: 1.09 (0.88, 1.39)
First month: 1.604 (1.268, 2.030)*
Sydney, Australia , h m=v. ,n Q City-wide avg and <5 T1: 0.807 (0.668, 0.976)*
(n=123,840) '~n max- JU-a km from monitor T3'1011 (0910 1124)*
Last month: 0.984 (0.906, 1.069)*
Jalaludin et al. (2007,
156601)
Jiang et al. (2007,
093029)
Shanghai, China
(n=3,346 preterm
births)
8-h avg: 32.7 City-wide avg
4 wks before birth: 1.06 (1.00, 1.12)
6 wks before birth: 1.06 (0.99, 1.13)
8 wks before birth: 1.09 (1.01, 1.19)
LO: NR (results presented in figure)
L1: NR (results presented in figure)
L2: NR (results presented in figure)
L3: NR (results presented in figure)
L4: NR (results presented in figure)
L5: NR (results presented in figure)
L6: NR (results presented in figure)
March 2011
7-36
DRAFT - DO NOT CITE OR QUOTE
-------
#Relative risk of PTB per 10 ppb change in O3.
'Relative risk of PTB per 1 ppb change in O3.
T1 = First Trimester, T2 = Second Trimester, T3 = Third Trimester
LO = Lag 0, LI = Lag 1, L2 = Lag 2, L3 = Lag 3, L4 = Lag 4, L5 = Lag 5, L6 = Lag 6
NR: No quantitative results reported
1 Table 7-3 provides a brief overview of the epidemiologic studies of PTB. In summary, the
2 evidence is consistent when examining shorter-term, late-pregnancy exposure to O3 and reports no
3 association with PTB. However when long-term exposure to O3 early in pregnancy is examined the
4 results are inconsistent. Studies conducted in the U.S., Canada, and England find no association with
5 O3 and PTB, while studies conducted in Australia and China report an O3 effect on PTB.
6
7.4.5. Fetal Growth
7 Low birth weight has often been used as an outcome measure because it is easily available and
8 accurately recorded on birth certificates. However, LEW may result from either short gestation, or
9 inadequate growth in utero. Most of the studies investigating air pollution exposure and LEW
10 limited their analyses to term infants to focus on inadequate growth. A number of studies were
11 identified that specifically addressed growth restriction in utero by identifying infants who failed to
12 meet specific growth standards. Usually these infants had birth weight less than the 10th percentile
13 for gestational age, using an external standard. Many of these studies have been previously
14 discussed, since they also examined other reproductive outcomes (i.e., LEW or PTB).
15 A limitation of environmental studies that use birth weight as a proxy measure of fetal growth
16 is that patterns of fetal growth during pregnancy cannot be assessed. This is particularly important
17 when investigating pollutant exposures during early pregnancy as birth weight is recorded
18 many months after the exposure period. The insult of air pollution may have a transient effect on
19 fetal growth, where growth is hindered at one point in time but catches up at a later point. For
20 example, maternal smoking during pregnancy can alter the growth rate of individual body segments
21 of the fetus at variable developmental stages, as the fetus experiences selective growth restriction
22 and augmentation (Lampl and Jeanty, 2003,
23 Fetal growth is influenced by maternal, placental, and fetal factors. The biological mechanisms
24 by which air pollutants may influence the developing fetus remain largely unknown. Several
25 mechanisms have been proposed, including maternal susceptibility to infection, oxidative stress,
26 hematological factors such as blood viscosity, and the direct effect of specific pollutants on fetal
27 development or on DNA and its transcription. Air pollution may affect maternal respiratory function
28 or general health, which may in turn impair uteroplacental and umbilical blood flow, transplacental
29 glucose, and total insulin, all of which are important determinants of fetal growth. Additionally,
30 certain changes resulting in fetal growth retardation may occur in early pregnancy (around the time
31 of implantation) caused by an abnormal reaction between the trophoblast and uterine tissues. A
32 defective trophoblast invasion, resulting in suboptimal placentation and maternal hemodynamic
33 maladaptation can alter growth and development of the fetus. Inhalation of air pollution can cause
34 inflammatory responses and oxidative stress, and both of these reactions can interfere with normal
March 2011 7-37 DRAFT - DO NOT CITE OR QUOTE
-------
1 intrauterine growth via vascular dysfunction in the placenta and damaged DNA. Also, pro-
2 inflammatory cytokines can limit trophoblast invasion during the early stages of pregnancy,
3 restricting fetal growth. Poor placental vascularity is caused partly by dysregulation of gene
4 expression in key angiogenic factors in early pregnancy, and if ambient air pollution is associated
5 with poor placental function it may partly be caused by perturbed DNA transcription early in
6 pregnancy.
7 The terms small-for-gestational-age (SGA), which is defined as a birth weight <10th percentile
8 for gestational age (and often sex and/or race), and intrauterine growth retardation (IUGR) are often
9 used interchangeably. However, this definition of SGA does have limitations. For example, using it
10 for IUGR may overestimate the percentage of "growth-restricted" neonates as it is unlikely that 10%
11 of neonates have growth restriction (Wollmann, 1998, 193812). On the other hand, when the 10th
12 percentile is based on the distribution of live births at a population level, the percentage of SGA
13 among PTB is most likely underestimated (Hutcheon and Platt, 2008, 193795). Nevertheless, SGA
14 represents a statistical description of a small neonate, whereas the term IUGR is reserved for those
15 with clinical evidence of abnormal growth. Thus all IUGR neonates will be SGA, but not all SGA
16 neonates with be IUGR (Wollmann, 1998, 193812). In the following section the terms SGA and
17 IUGR are referred to as each cited study used the terms.
18 Over the past decade a number of studies examined various metrics of fetal growth restriction.
19 Salam et al. (2005, 087885) assessed the effect of increasing O3 concentrations on IUGR in a
20 population of infants born in California from 1975-1987 as part of the Children's Health Study. The
21 authors reported that maternal O3 exposures averaged over the entire pregnancy and during the third
22 trimester were associated with increased risk of IUGR. A 10-ppb difference in 24-h maternal O3
23 exposure during the third trimester increased the risk of IUGR by 11% (95% CI: 0, 20%). Brauer et
24 al. (2008, 156292) evaluated the impacts of air pollution (CO, NO2, NO, O3, SO2, PM2 5, PM10) on
25 SGA (1999-2002) using spatiotemporal residential exposure metrics by month of pregnancy in
26 Vancouver, BC. The O3 associations were largely protective (OR= 0.87, [95% CI: 0.81, 0.93] for a
27 10 ppb increase in inverse distance weighted SGA), and no additional quantitative results were
28 presented for O3. Liu et al. (2007, 090429) examined the association between IUGR among singleton
29 term live births and SO2, NO2, CO, O3, and PM2 5 in 3 Canadian cities for the period 1985-2000. No
30 increase in the risk of IUGR in relation to exposure to O3 averaged over each month and trimester of
31 pregnancy was noted.
32 Three studies conducted in Australia provide evidence for an association between ambient O3
33 and fetal growth restriction. Hansen et al. (2007, 090703) examined SGA among singleton, full-term
34 births in Brisbane, Australia in relation to ambient air pollution (bsp, PMi0, NO2, O3) during
35 pregnancy. They also examined head circumference and crown-heel length in a subsample of term
36 neonates. Trimester specific exposures to all pollutants were not statistically significantly associated
37 with a reduction in head circumference or an increased risk of SGA. When monthly specific
38 exposures were examined, the authors observed an increased risk of SGA associated with exposure
39 to O3 during month 4 (OR 1.11 [95% CI: 1.00, 1.24] per 10 ppb increase). In a subsequent study,
March 2011 7-38 DRAFT - DO NOT CITE OR QUOTE
-------
1 Hansen et al. (2008, 190273) examined the possible associations between fetal ultrasonic
2 measurements and ambient air pollution (PMi0, O3, NO2, SO2) during early pregnancy. This study
3 had two strengths: (1) fetal growth was assessed during pregnancy as opposed to at birth; and (2)
4 there was little delay between exposures and fetal growth measurements, which reduces potential
5 confounding and uses exposures that are concurrent with the observed growth pattern of the fetus.
6 Fetal ultrasound biometric measurements were recorded for biparietal diameter (BPD), femur length,
7 abdominal circumference, and head circumference. To further improve exposure assessment, the
8 authors restricted the samples to include only scans from women for whom the centroid of their
9 postcode was within 14 km of an air pollution monitoring site. Ozone during days 31-60 was
10 associated with decreases in all of the fetal growth measurements, and a 1.78 mm reduction in
11 abdomen circumference per 10 ppb increase in O3 concentration, though this effect did not persist in
12 co-pollutant models. The change in ultrasound measurements associated with O3 during days 31-60
13 of gestation indicated that increasing O3 concentration decreased the magnitude of ultrasound
14 measurements for women living within 2 km of the monitoring site. The relationship decreased
15 toward the null as the distance from the monitoring sites increased. When assessing effect
16 modification due to SES, there was some evidence of effect modification for most of the
17 associations, with the effects of air pollution stronger in the highest SES quartile. In the third study,
18 Mannes et al. (2005, 087895) estimated the effects of pollutant (PM10, PM2 5, NO2, CO and O3)
19 exposure in the first, second and third trimesters of pregnancy and risk of SGA in Sydney, Australia.
20 Citywide average air pollutant concentrations in the last month, third trimester, and first trimester of
21 pregnancy had no effect on SGA. Concentrations of O3 in the second trimester of pregnancy had
22 small but adverse effects on SGA (OR 1.10 [95% CI: 1.00, 1.14] per 10 ppb increment). This effect
23 disappeared when the analysis was limited to births with a maternal address within 5 km of a
24 monitoring station (OR 1.00 [95% CI: 0.60, 1.79] per 10 ppb increment).
25 Very little information from toxicological studies is available to address effects on fetal
26 growth. However, there is evidence to suggest that prenatal exposure to O3 can affect postnatal
27 growth. A few studies reported that mice or rats exposed developmentally (gestationally ±
28 lactationally) to O3 had deficits in body weight gain in the postpartum period (Bignami et al., 1994,
29 076063: Haro and Paz, 1993, 044194: Kavlock et al., 1980, 094043).
30 Table 7-4 provides a brief overview of the epidemiologic studies of fetal growth restriction. In
31 summary, the evidence is inconsistent when examining exposure to O3 and fetal growth restriction.
32 Similar to PTB, studies conducted in Australia have reported an effect of O3 on fetal growth, whereas
33 studies conducted in other areas have not found such an effect. This may be due to the restriction of
34 births to those within 2-14 km of a monitoring station, as was done in the Australian studies.
March 2011 7-39 DRAFT - DO NOT CITE OR QUOTE
-------
Table 74. Brief summary of epidemiologic studies of fetal growth
Study
Salam et al. (2005,
Braueret al. (2008,
Liu et al. (2007,
090429)
Hansen et al. (2007,
090703)
Hansen et al. (2008,
190273)
Mannes et al. (2005,
Location .. /-> / u\
(Sample Size) Mean O3 (ppb)
California, U.S. 24-h avg: 27.3
(n=3901) 8 h: 50.6
Vancouver, BC, Canada ~, . ,„_. , .
(n=70,249) 24-h avg. 14
Calgary, Edmonton, and 0/. h „,,„. 1C c
Montreal, Canada t4?™*!??
(n= 16,430) 1 -Umax. 31 .2
(Bn^2b6a 617? UStraNa 8'h max: 26'7
Brisbane, Australia R h ...... ~,A K
(n=1 5,623) 8-h avg. 24.8
Sydney, Australia , . ,, -
(n=1 38,056) 1-h max. 31.6
Exposure
assessment
ZIP code level
Nearest Monitor (within
10km)
Inverse Distance
Weighting (IDW)
Census Subdivision
avg
City-wide avg
Within 2 km of monitor
City-wide avg and
<5 km from monitor
Effect Estimate (95% Cl)
Entire pregnancy: 1.16 (1.00, 1.32)
11:1.00(0.94, 1.11)
12:1.06(1.00, 1.12)
T3:1.11 (1.00, 1.17)
Entire pregnancy: NR
First 30 days of pregnancy: NR
Last 30 days of pregnancy: NR
T1:NR
T3:NR
Entire pregnancy: NR (results
presented in figure)
T1 : NR (results presented in figure)
T2: NR (results presented in figure)
T3: NR (results presented in figure)
T1:1.01 (0.89, 1.15)
12:1.00(0.86, 1.17)
13:0.83(0.71, 0.97)
M1: -0.32 (-1.56, 0.91)*
M2: -0.58 (-1.97, 0.80)*
M3: 0.26 (-1.07, 1.59)*
M4:0.11 (-0.98, 1.21)*
T1 : 0.90 (0.48, 1 .34)
T2: 1 .00 (0.60, 1 .79)
13:1.10(0.66, 1.97)
Last 30 days of pregnancy: 1.10
(0.74, 1.79)
#Relative risk of fetal growth restriction per 10 ppb change in O3.
*Mean change in fetal ultrasonic measure of head circumference recorded between 13 and 26 weeks gestation for a 10-ppb
increase in maternal exposure to O3 during early pregnancy
T1 = First Trimester, T2 = Second Trimester, T3 = Third Trimester
M1 = Month 1, M2 = Month 2, M3 = Month 3, M4 = Month 4
NR: No quantitative results reported
7.4.6. Birth Defects
1 Despite the growing body of literature evaluating the association between ambient air
2 pollution and various adverse birth outcomes, relatively few studies have investigated the effect of
3 temporal variations in ambient air pollution on birth defects. Heart defects and oral clefts have been
4 the focus of the majority of these recent studies, given the higher prevalence than other birth defects
5 and associated mortality.
6 Mechanistically, air pollutants could be involved in the etiology of birth defects via
7 hemodynamic events, anoxic events, oxidative stress, and toxicity to certain cell populations during
8 development. One potential etiologic pathway may include the neural crest cell population. Normal
9 migration and differentiation of neural crest cells are important for heart development and are
10 particularly sensitive to toxic insults. They respond by undergoing apoptosis, in part because they
11 lack antioxidative stress proteins. Ozone is a very reactive molecule and a strong oxidizing agent that
12 can generate superoxides, hydrogen peroxide, and hydroxyl radicals, contributing to oxidative stress
March 2011
7-40
DRAFT - DO NOT CITE OR QUOTE
-------
1 Several studies have been conducted examining the relationship between O3 exposure during
2 pregnancy and birth defects. The earliest of these studies was conducted in southern California (Ritz
3 et al, 2002, 023227). This study evaluated the effect of air pollution on the occurrence of cardiac
4 and orofacial birth defects in neonates and fetuses delivered in southern California in 1987-1993.
5 Maternal exposure estimates were based on data from the fixed site closest to the mother's ZIP code
6 area. When using a case-control design where cases were matched to 10 randomly selected controls,
7 results showed increased risks for aortic artery and valve defects (OR 1.56 [95% CI: 1.16, 2.09] per
8 10 ppb O3), pulmonary artery and valve anomalies (OR 1.34 [95% CI: 0.96, 1.87] per 10 ppb O3),
9 and conotruncal defects (OR 1.36 [95% CI: 0.91, 2.03] per 10 ppb O3) in a dose-response manner
10 with second-month O3 exposure. These associations were not observed for orofacial defects. The
11 average effect sizes and patterns of second-month O3 exposure were similar for these defects and
12 varied only slightly from single- to multi-pollutant models, or when the models were adjusted for
13 other potential confounding factors. Additionally, the authors reported an OR of 1.13 (95% CI: 0.90,
14 1.40) per 10 ppb during the second trimester for cleft lip with or without cleft palate.
15 A study conducted in Texas (Gilboa et al., 2005, 087892) looked at a similar period of
16 exposure but reported no association with most of the birth defects studied (O3 concentration was
17 studied using quartiles with the lowest representing <18 ppb and the highest representing 31 ppb).
18 The authors found slightly elevated odds ratios for pulmonary artery and valve defects. They also
19 detected an inverse association between O3 exposure and isolated ventricular septal defects. For cleft
20 lip with or without cleft palate, the authors reported and OR of 1.09 (95% CI: 0.70, 1.69) for the
21 fourth quartile contrasted with the first quartile of exposure during 3-8 weeks of pregnancy. Overall,
22 this study did not provide strong evidence that air pollution increases the risk of cardiac defects or
23 oral clefts.
24 A recent study conducted in Atlanta, GA examined O3 exposure during the third through
25 seventh week of pregnancy and reported no association with risk of cardiovascular malformations
26 (mean long-term average of 8-h O3 concentrations excluding November through February ranged by
27 5-year groups from 39.8 to 43.3 ppb) (Strickland et al., 2009, 190324).
28 Hwang and Jaakola (2008, 193794) conducted a population-based case-control study to
29 investigate exposure to ambient air pollution and the risk of cleft lip with or without cleft palate in
30 Taiwan. The risk of cleft lip with or without cleft palate was increased in relation to O3 levels in the
31 first gestational month (OR 1.17 [95% CI: 1.01, 1.36] per 10 ppb) and second gestational month
32 (OR 1.22 [95% CI: 1.03, 1.46] per 10 ppb), but was not related to any of the other pollutants. In
33 three-pollutant models, the effect estimates for O3 exposure were stable for the four different
34 combinations of pollutants and were all statistically significant.
35 Marshall et al. (2010, 597374) compared estimated exposure to ambient pollutants during
36 early pregnancy among mothers of children with oral cleft defects to that among mothers of controls.
37 The authors observed no consistent elevated associations between any of the air pollutants examined
38 and cleft malformations, though there was a weak association between cases of cleft palate only and
39 increasing O3 concentrations. This association increased when cases and controls were limited to
March 2011 7-41 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
those with residences within 10 km of the closest O3 monitor (OR 2.2 [95% CI: 1.0, 4.9], comparing
highest quartile [>33 ppb] to lowest quartile [<15 ppb]).
A limited number of toxicological studies have examined birth defects in animals exposed
gestationally to O3. Kavlock et al. (1979, 039228) exposed pregnant rats to O3 for precise periods
during organogenesis. No significant teratogenic effects were found in rats exposed 8 hr/day to
concentrations of O3 varying from 0.44 to 1.97 ppm during early (days 6-9), mid (days 9-12), or late
(days 17 to 20) gestation, or the entire period of organogenesis (days 6-15). Earlier research found
eyelid malformation following gestational and postnatal exposure to 0.2 ppm O3 (Veninga, 1967,
040746).
Table 7-5 provides a brief overview of the epidemiologic studies of birth defects. Results from
these studies are not entirely consistent. This inconsistency could be due to the absence of true
associations between O3 and risks of cardiovascular malformations and oral cleft defects; it could
also be due to differences in populations, pollution levels, outcome definitions, or analytical
approaches. The lack of consistency of associations between O3 and cardiovascular malformations or
oral cleft defects might be due to issues relating to statistical power or measurement error.
Table 7-5. Brief summary of epidemiologic studies of birth defects
Study
Ritz et al. (2002,
Gilboa et al. (2005,
087892)
Hwang and Jaakola
(2008, 1 93794)
Strickland et al.
(2009, 1 90324)
Marshall et al. (2010,
Outcomes
Examined
Cardiac and Cleft
Defects
Cardiac and Cleft
Defects
Oral Cleft Defects
Cardiac Defects
Oral Cleft Defects
J-Samp^Size, Mean O3 (ppb,
southern California
(n=3,549 cases; 24-h avg: NR
10, 649 controls)
7 Counties in TX
(n=5,338 cases; 24-h avg: NR
4,580 controls)
Taiwan
(n=653 cases; 24-h avg: 27.31
6,530 controls)
Atlanta, GA R . m_v. ,Q K A-, ,
(n=3,338 cases) 8'h max' 39-8-43-3
New Jersey
(n=71 7 cases; 24-h avg: 25
12,925 controls)
Exposure
Assessment
Nearest Monitor
(within 10 mi)
Nearest Monitor
Inverse Distance
Weighting (IDW)
Weighted City-wide
avg
Nearest Monitor
(within 40 km)
Exposure Window
Month 1,2,3
Trimester 2,3
3-mo period prior to
conception
Weeks 3-8 of
gestation
Months 1,2,3
Weeks 3-7 of
gestation
Weeks 5-1 0 of
gestation
7.4.7. Developmental Respiratory Effects
The issue of prenatal exposure has assumed increasing importance, since ambient air pollution
exposures of pregnant women have been shown to lead to adverse pregnancy outcomes, as well as to
respiratory morbidity and mortality in the first year of life. Growth and development of the
respiratory system take place mainly during the prenatal and early postnatal periods. This early
developmental phase is thought to be very important in determining long-term lung growth. Studies
have recently examined this emerging issue, and several were included in Sections 7.2.1 and 7.2.3,
and are included here because they included both prenatal and post-natal exposure periods.
Mortimer et al. (2008, 122163; 2008, 187280) examined the association of prenatal and
lifetime exposures to air pollutants with pulmonary function and allergen sensitization in a subset of
March 2011
7-42
DRAFT - DO NOT CITE OR QUOTE
-------
1 asthmatic children (ages 6-11) included in the Fresno Asthmatic Children's Environment Study
2 (FACES). Monthly means of pollutant levels for the years 1989-2000 were created and averaged
3 separately across several important developmental time-periods, including the entire pregnancy, each
4 trimester, the first 3 years of life, the first 6 years of life, and the entire lifetime. The 8-h avg O3
5 concentrations were approximately 50 ppb for each of the exposure metrics (estimated from figure).
6 In the first analysis (Mortimer et al., 2008, 122163). negative effects on pulmonary function were
7 found for exposure to PMi0, NO2, and CO during key neonatal and early life developmental periods.
8 The authors did not find a negative effect of exposure to O3 among this cohort. In the second
9 analysis (Mortimer et al., 2008, 187280). sensitization to at least one allergen was associated, in
10 general, with higher levels of CO and PMi0 during the entire pregnancy and second trimester and
11 higher PMi0 during the first 2 years of life. Lower exposure to O3 during the entire pregnancy or
12 second trimester was associated with an increased risk of allergen sensitization. Although the
13 pollutant metrics across time periods are correlated, the strongest associations with the outcomes
14 were observed for prenatal exposures. Though it may be difficult to disentangle the effect of prenatal
15 and postnatal exposures, the models from this group of studies suggest that each time period of
16 exposure may contribute independently to different dimensions of school-aged children's pulmonary
17 function. For 4 of the 8 pulmonary-function measures (FVC, FEVi, PEF, FEF25-75), prenatal
18 exposures were more influential on pulmonary function than early-lifetime metrics, while, in
19 contrast, the ratio of measures (FEVi/FVC and FEF25-75/FVC) were most influenced by postnatal
20 exposures. When lifetime metrics were considered alone, or in combination with the prenatal
21 metrics, the lifetime measures were not associated with any of the outcomes, suggesting the timing
22 of the exposure may be more important than the overall dose and prenatal exposures are not just
23 markers for lifetime or current exposures.
24 Clark et al. (2010, 594440) investigated the effect of exposure to ambient air pollution in utero
25 and during the first year of life on risk of subsequent asthma diagnosis (incident asthma diagnosis up
26 to age 3-4) in a population-based nested case-control study. Air pollution exposure for each subject
27 based on their residential address history was estimated using regulatory monitoring data, land use
28 regression modeling, and proximity to stationary pollution sources. An average exposure was
29 calculated for the duration of pregnancy (~ 15 ppb; transformed from (ig/m3) and the first year of life
30 (-14 ppb; transformed from (ig/m3). In contrast to the Mortimer et al. studies (2008, 122163; 2008,
31 187280). the effect estimates for first-year exposure were generally larger than for in utero
32 exposures. However, similar to the Mortimer et al. studies, the observed associations with O3 were
33 largely protective. Because of the relatively high correlation between in utero and first-year
34 exposures for many pollutants, it was difficult to discern the relative importance of the individual
35 exposure periods.
36 Latzin et al. (2009, 195721) examined whether prenatal exposure to air pollution was
37 associated with lung function changes in the newborn. Tidal breathing, lung volume, ventilation
38 inhomogeneity and eNO were measured in 241 unsedated, sleeping neonates (age= 5 weeks). The
39 median of the 24-h avg O3 concentrations averaged across the post-natal period was ~44 ppb
March 2011 7-43 DRAFT - DO NOT CITE OR QUOTE
-------
1 (transformed from (ig/m3). Consistent with the previous studies, no association was found for
2 prenatal exposure to O3 and lung function.
3 The new toxicological literature since the 2006 O3 AQCD (U.S. EPA, 2006, 088089). covering
4 respiratory changes related to developmental O3 exposure, reports ultrastructural changes in
5 bronchiole development, alterations in placental and pup cytokines, and increased pup airway hyper-
6 reactivity. These studies are detailed below. Older studies are discussed where new information is not
7 available.
8 Fetal rat lung bronchiole development is triphasic, comprised of the glandular phase
9 (measured at GDI8), the canalicular phase (GD20), and the saccular phase (GD21). The
10 ultrastructural lung development in fetuses of pregnant rats exposed to 1-ppm O3 (12 h/day, out to
11 either GDI8, GD20 or GD21) was examined by electron microscopy during these three phases. In
12 the glandular phase, bronchiolar columnar epithelial cells in fetuses of dams exposed to O3 had
13 cytoplasmic damage and swollen mitochondria. Bronchial epithelium at the canalicular phase in O3
14 exposed pups had delayed maturation in differentiation, i.e., glycogen abundance in secretory cells
15 had not diminished as it should with this phase of development. Congruent with this finding, delayed
16 maturation of tracheal epithelium following early neonatal O3 exposure (1 ppm, 4-5 h/day for
17 first week of life) in lambs has been previously reported (Mariassy et al., 1989, 042246; Mariassy et
18 al., 1990, 042311). Also at the canalicular phase, atypical cells were seen in the bronchiolar lumen of
19 O3 exposed rat fetuses. Finally, in the saccular phase, mitochondrial degradation was present in the
20 non-ciliated bronchiolar cells of rats exposed in utero to O3. In conclusion, O3 exposure of pregnant
21 rats produced ultra-structural damage to near-term fetal bronchiolar epithelium (Lopez et al., 2008,
22 197786).
23 Exposure of laboratory animals to multiple airborne pollutants can differentially affect pup
24 physiology. One study showed that exposure of C57BL/6 mouse dams to 0.48 mg PM intratracheally
25 twice weekly for 3 weeks during pregnancy augmented O3-induced airway hyper-reactivity in
26 juvenile offspring. Maternal PM exposure also significantly increased placental cytokines above
27 vehicle-instilled controls. Pup postnatal O3 exposure (1 ppm 3 h/day, every other day, thrice weekly
28 for 4 weeks) induced significantly increased cytokine levels (IL-lp, TNF-a, KC, and IL-6) in whole
29 lung versus postnatal air exposed groups; this was further exacerbated with gestational PM exposure
30 (Auten et al., 2009, 200760).
31 A series of experiments using infant rhesus monkeys repeatedly exposed to 0.5 ppm O3
32 starting at one-month of age have examined the effect of O3 alone or in combination with an inhaled
33 allergen on morphology and lung function (Plopper et al., 2007, 596412). Exposure to O3 alone or
34 allergen alone produced small but not statistically significant changes in baseline airway resistance
35 and airway responsiveness, but the combined exposure to both O3 + antigen produced statistically
36 significant and greater than additive changes in both functional measurements. Additionally, cellular
37 changes and significant structural changes in the respiratory tract have been observed in infant
38 rhesus monkeys exposed to O3 (Fanucchi et al., 2006, 096491). A more detailed description of these
39 studies can be found in Section 7.2.3 (Pulmonary Structure and Function).
March 2011 7-44 DRAFT - DO NOT CITE OR QUOTE
-------
7.4.8. Developmental Central Nervous System Effects
7.4.8.1. Laterally
1 Two reports of laterality changes in mice developmentally exposed to O3 have been reported in
2 the literature. Mice developmentally exposed to 0.6 ppm O3 (6 days before breeding to weaning at
3 PND21) showed a turning preference (left turns) distinct from air exposed controls (clockwise turns)
4 (Dell'Omo et al., 1995, 080787): in previous studies this behavior in mice has been found to
5 correlate with specific structural asymmetries of the hippocampal mossy fiber projections (Schopke
6 et al., 1991, 684254). The 2006 AQCD evidence forthe effect of O3 on laterality or handedness
7 demonstrated that rats exposed to O3 during fetal and neonatal life showed limited, gender-specific
8 changes in handedness after exposure to the intermediate dose of O3 (only seen in female mice
9 exposed to 0.6 ppm O3, and not in males at 0.6 ppm or in either sex of 0.3 or 0.9 ppm O3 with
10 exposure from 6 days before breeding to PND26) (Petruzzi et al., 1999, 012066).
7.4.8.2. Brain Morphology and Neurochemical Changes
11 The nucleus tractus solitarius (NTS), a medullary area of respiratory control, of adult animals
12 exposed prenatally to 0.5 ppm O3 (12h/day, ED5-ED20) had significantly less tyrosine hydroxylase
13 staining versus control (Boussouar et al., 2009, 596368). Tyrosine hydroxylase is the rate-limiting
14 enzyme for dopamine synthesis and serves as a precursor for catecholamine synthesis; thus,
15 decreased staining is used as a marker of dopaminergic or catecholaminergic cell or activity loss in
16 these regions and thus functions in neuronal plasticity. After physical restraint stress, control animals
17 respond at the histological level with Fos activation, a marker of neuronal activity, and tyrosine
18 hydroxylase activation in the NTS, a response which is absent or attenuated in adult animals exposed
19 prenatally to 0.5 ppm O3 (Boussouar et al., 2009, 596368) when compared to control air exposed
20 animals who also were restrained. The O3-exposed offspring in this study were cross-fostered to
21 control air exposed dams to avoid O3-dependent dam related neonatal effects on offspring outcomes
22 (i.e., dam behavioral or lactational contributions to pup outcomes) (Boussouar et al., 2009, 596368).
23 Developmental exposure to 0.3 or 0.6 ppm O3 prior to mating pair formation through GD17
24 induced significant increased levels of BDNF in the striatum of adult (PND140) O3 exposed
25 offspring as compared to control air exposed animals; these O3-exposed animals also had
26 significantly decreased level of NGF in the hippocampus versus control (Santucci et al., 2006,
27 596414V
28 Changes in the pup cerebellum with prenatal 1 ppm O3 exposure include altered morphology
29 (Rivas-Manzano and Paz, 1999, 012072: Romero-Velazquez et al., 2002, 035575). decreased total
30 area (Romero-Velazquez et al., 2002, 035575). decreased number of Purkinje cells (Romero-
31 Velazquez et al., 2002, 035575). and altered monoamine neurotransmitter content with the
32 catecholamine system affected and the indoleamine system unaffected by O3 (Gonzalez-Pina et al.,
33 2008.475317).
March 2011 7-45 DRAFT - DO NOT CITE OR QUOTE
-------
7.4.8.3. Neurobehavioral Outcomes
1
2 O3 administration to dams during pregnancy with or without early neonatal exposure has been
3 shown to contribute to multiple neurobehavioral outcomes in offspring that are described in further
4 detail below.
5 O3 administration (0.4, 0.8 or 1.2 ppm O3) during the majority of pregnancy (PD7-17) of CD-I
6 mice did not affect pup behavioral outcomes including early behavioral ultrasonic vocalizations and
7 more permanent later measurements (PND60 or 61) including pup activity, habituation and
8 exploration and d-amphetamine-induced hyperactivity (Bignami et al., 1994, 076063); these pups
9 were all cross-fostered or reared on non- O3 exposed dams.
10 Testing for aggressive behavior in mice continuously exposed to O3 (0.3 or 0.6 ppm from
11 30 days prior to mating to GDI7) revealed that mice had significantly increased defensive/
12 submissive behavior (increased freezing posturing on the first day only of a multiple-day exam)
13 versus air exposed controls (Santucci et al., 2006, 596414). Similar to this and as reported in
14 previous AQCDs, continuous exposure of adult animals to O3 induced significant increases in fear
15 behavior and decreased aggression as measured by significantly decreased freezing behavior
16 (Petruzzi et al., 1995, 077448).
17 Developmentally exposed animals also had significantly decreased amount of time spent nose
18 sniffing other mice (Santucci et al., 2006, 596414); this social behavior deficit, decreased sniffing
19 time, was not found in an earlier study with similar exposures (Petruzzi et al., 1995, 077448). but
20 sniffing of specific body areas was measured in Santucci et al. (2006, 596414) and total number of
21 sniffs of the entire body was measured in Petruzzi et al. (1995, 077448). The two toxicology studies
22 exploring social behavior (sniffing) employ different study designs and find opposite effects in
23 animals exposed to O3
7.4.8.4. Sleep Aberrations after Developmental Ozone Exposure
24 The effect of gestational O3 exposure (1 ppm O3, 12h/day, during dark period) on sleep
25 patterns in rat offspring was followed using 24 h polysomnographic recordings at 30, 60 and 90 days
26 of age (Haro and Paz, 1993, 044194). Ozone-exposed pups manifested with inverted sleep-wake
27 patterns or circadian rhythm phase-shift. Rat vigilance was characterized in wakefulness, slow wave
28 sleep (SWS), and paradoxical sleep (PS) using previously characterized criteria. The O3 exposed
29 offspring spent longer time in the wakefulness state during the light period, more time in SWS
30 during the period of darkness, and showed significant decrements in PS. Chronic O3 inhalation
31 significantly decreased the duration of PS during both the light and dark periods (Haro and Paz,
32 1993, 044194). These effects were consistent at all time periods measured (30, 60 and 90 days of
33 age). These sleep effects reported after developmental exposures expand upon the existing literature
34 on sleep aberrations in adult animals exposed to O3 [rodents: (Arito et al., 1992, 042759; Paz and
35 Huitron-Resendiz, 1996, 082684); and cats: (Paz and Bazan-Perkins, 1992, 036436)1. A role for
36 inhibition of cyclooxygenase-2 and the interleukins and prostaglandins in the O3-dependent sleep
March 2011 7-46 DRAFT - DO NOT CITE OR QUOTE
-------
1 changes potentially exists with evidence from a publication on indomethacin pretreatment
2 attenuating O3-induced sleep aberrations in adult male animals (Rubio and Paz, 2003, 053541).
7.4.9. Early Life Mortality
3 Infants may be particularly susceptible to the adverse effects of air pollution. The lung is not
4 well developed at birth, with 80% of alveoli being formed postnatally. An important question
5 regarding the association between PM and infant mortality is the critical window of exposure during
6 development for which infants are susceptible. Several age intervals have been explored: neonatal
7 (<1 month); postneonatal (1 month to 1 year); and an overall interval for infants that includes both
8 the neonatal and postneonatal periods (<1 year). Within these various age categories, multiple causes
9 of deaths have been investigated, particularly total deaths and respiratory-related deaths. The studies
10 reflect a variety of study designs, exposure periods, regions, and adjustment for confounders. Within
11 the first year of life, infants develop rapidly; therefore their susceptibility may change within weeks
12 or months. During the neonatal and post-neonatal periods, the developing lung is highly susceptible
13 to environmental toxicants. As discussed below, a handful of studies have examined the effect of
14 ambient air pollution on neonatal and postneonatal mortality, with the former the least studied. These
15 studies varied somewhat with regard to the outcomes and exposure periods examined and study
16 designs employed.
17 The results of these infant mortality studies are presented here, and in Table 7-6, with the other
18 reproductive and developmental outcomes because it is likely that in utero exposures contribute to
19 this outcome. Both long-term and short-term exposure studies of infant mortality are included in this
20 section. A major issue in studying environmental exposures and infant mortality is selecting the
21 relevant exposure period, since the biological mechanisms leading to death and the critical periods of
22 vulnerability are poorly understood. Exposures proximate to the death may be most relevant if
23 exposure causes an acute effect. However, exposure occurring in early life might affect critical
24 growth and development, with results observable later in the first year of life, or cumulative
25 exposure during the first year of life may be the most important determinant. The studies reviewed
26 below have dealt with this issue in different ways. Many have considered several exposure metrics
27 based on different periods of exposure.
7.4.9.1. Stillbirth
28 Pereira et al. (1998, 007264) investigated the association among daily counts of intrauterine
29 mortality (over 28 weeks of gestation) and air pollutant concentrations in Sao Paulo, Brazil from
30 1991 through 1992. The association was strong for NO2, but lesser for SO2 and CO. These
31 associations exhibited a short lag time, less than 5 days. No significant association was detected
32 between O3 and intrauterine mortality.
March 2011 7-47 DRAFT - DO NOT CITE OR QUOTE
-------
7.4.9.2. Infant Mortality, Less than 1 Year
1 Ritz et al. (2006, 089819) linked birth and death certificates for infants who died between
2 1989 and 2000 to evaluate the influence of outdoor air pollution on infant death in the South Coast
3 Air Basin of California. The authors examined exposure periods 2 weeks, 1 month, 2 months, and
4 6 months before each case subject's death and reported no association between ambient levels of O3
5 and infant mortality. Similarly, Diaz et al. (2004, 089894) analyzed the effects of extreme
6 temperatures and air pollutants on daily mortality in children less than 1 year of age in Madrid,
7 Spain, from 1986 to 1997 and observed no statistically significant association between mortality and
8 O3 concentrations. Hajat et al. (2007, 093276) analyzed time-series data of daily infant mortality
9 counts in 10 major cities in the UK to quantify any associations with short-term changes in air
10 pollution. When the results from the 10 cities were combined there was no relationship between O3
11 and infant mortality, even after restricting the analysis to just the summer months.
12 Conversely, a time-series study of infant mortality conducted in the southwestern part of
13 Mexico City in the years 1993-1995 found that infant mortality was associated with the levels of
14 NO2 and O3 3-5 days before death, but not as consistently as with PM. A 10-ppb increase in 24-h avg
15 O3 was associated with a 2.78% increase (95% CI: 0.29, 5.26%) in infant mortality (lag 3) (Loomis
16 et al., 1999, 087288). This increase was attenuated, though still positive when evaluated in a two-
17 pollutant model with PM25. One-hour max concentrations of O3 exceeded prevailing Mexican and
18 international standards nearly every day.
7.4.9.3. Neonatal Mortality, Less than 1 Month
19 Three studies have evaluated ambient O3 concentrations and neonatal mortality and observed
20 no association. Ritz et al. (2006, 089819) linked birth and death certificates for infants who died
21 between 1989 and 2000 to evaluate the influence of outdoor air pollution on infant death in the South
22 Coast Air Basin of California. The authors examined exposure periods 2 weeks, 1 month, 2 months,
23 and 6 months before each case subject's death and reported no association between ambient levels of
24 O3 and neonatal mortality. Hajat et al. (2007, 093276) analyzed time-series data of daily infant
25 mortality counts in 10 major cities in the UK to quantify any associations with short-term changes in
26 air pollution. When the results from the 10 cities were combined there was no relationship between
27 O3 and neonatal mortality, even after restricting the analysis to just the summer months. Lin et al.
28 (2004, 095787) assessed the impact of daily changes in air pollutants on the number of daily
29 neonatal deaths in Sao Paulo, Brazil. The authors observed no association between ambient levels of
30 O3 and neonatal mortality.
7.4.9.4. Postneonatal Mortality, 1 Month to 1 Year
31 A number of studies focused on the postneonatal period when examining the effects of O3 on
32 infant mortality. Ritz et al. (2006, 089819) linked birth and death certificates for infants who died
33 between 1989 and 2000 to evaluate the influence of outdoor air pollution on infant death in the South
March 2011 7-48 DRAFT - DO NOT CITE OR QUOTE
-------
1 Coast Air Basin of California. The authors examined exposure periods 2 weeks, 1 month, 2 months,
2 and 6 months before each case subject's death and reported no association between ambient levels of
3 O3 and postneonatal mortality. Woodruff et al. (2008, 098386) evaluated the county-level
4 relationship between cause-specific postneonatal infant mortality and chronic early-life (first
5 2 months of life) exposure to air pollutants across the U.S. Similarly, they found no association
6 between O3 exposure and deaths from respiratory causes. In the U.K., Hajat et al. (2007, 093276)
7 analyzed time-series data of daily infant mortality counts in 10 major cities to quantify any
8 associations with short-term changes in air pollution. When the results from the 10 cities were
9 combined there was no relationship between O3 and postneonatal mortality, even after restricting the
10 analysis to just the summer months. In Ciudad Juarez, Mexico, Romieu et al. (2004, 093074)
11 examined the daily number of deaths between 1997 and 2001, estimating the modifying effect of
12 SES on the risk of postneonatal mortality. Ambient O3 concentrations were not related to infant
13 mortality overall, or in any of the SES groups. In a follow-up study, Carbajal-Arroyo (In Press,
14 667773) evaluated the relationship of 1-h daily max O3 levels with postneonatal infant mortality in
15 the Mexico City Metropolitan Area between 1997 and 2005. Generally, O3 was not significantly
16 related to infant mortality. However, upon estimating the modifying effect of SES on the risk of
17 postneonatal mortality, the authors found that O3 was significantly related to respiratory mortality
18 among those with low SES. In a separate analysis, the effect of PMi0 was evaluated with O3 level
19 quartiles. PMi0 alone was related to a significant increase in all-cause mortality. The magnitude of
20 this effect remained the same when only the days when O3 was in the lowest quartile were included
21 in the analyses. However, when only the days when O3 was in the highest quartile were included in
22 the analyses, the magnitude of the PMi0 effect increased dramatically (OR=1.06 [0.909, 1.241] for
23 PM10 on days with O3 in lowest quartile; OR=1.26 [1.08, 1.47] for PM10 on days with O3 in the
24 highest quartile. These results suggest that while O3 alone may not have an effect on infant mortality,
25 it may serve to potentiate the observed effect of PMi0 on infant mortality.
26 Tsai et al. (2006, 090709) used a case-crossover analysis to examine the relationship between
27 air pollution and postneonatal mortality in Kaohsiung, Taiwan during the period 1994-2000. The risk
28 of postneonatal deaths was 1.023 (95% CI: 0.564, 1.858) per 10-ppb increase in 24-h avg O3. The
29 confidence interval for this effect estimate is very wide, likely due to the small number of infants that
30 died each day, making it difficult to interpret this result. Several other studies conducted in Asia did
31 not find any association between O3 concentrations and infant mortality in the postneonatal period.
32 Ha et al. (2003, 042552) conducted a daily time-series study in Seoul, Korea to evaluate the effect of
33 short-term changes in ambient 8-h O3 concentrations on postneonatal mortality. Son et al. (2008,
34 190323) examined the relationship between air pollution and postneonatal mortality from all causes
35 among firstborn infants in Seoul, Korea during 1999-2003. Yang et al. (2006, 090760) used a case-
36 crossover analysis to examine the relationship between air pollution exposure and postneonatal
37 mortality in Taipei, Taiwan for the period 1994-2000. The authors observed no associations between
38 ambient levels of O3 and postneonatal mortality.
March 2011 7-49 DRAFT - DO NOT CITE OR QUOTE
-------
7.4.9.5. Sudden Infant Death Syndrome
1 The strongest evidence for an association between ambient O3 concentrations and SIDS comes
2 from a study that evaluated the county-level relationship between SIDS and chronic early-life (first
3 2 months of life) exposure to air pollutants across the U.S.(Woodruff et al., 2008, 098386). The
4 authors observed a 1.20 (95% CI: 1.09, 1.32) odds ratio for a 10-ppb increase in O3 and deaths from
5 SIDS. There was a monotonic increase in odds of SIDS for each quartile of O3 exposure compared
6 with the lowest quartile (highest quartile OR = 1.51; [95% CI: 1.17, 1.96]). In a multi-pollutant
7 model including PMi0 or PM2 5, CO and SO2, the OR for SIDS and O3 was not substantially lower
8 than that found in the single-pollutant model. When examined by season, the relationship between
9 SIDS deaths and O3 was generally consistent across seasons with a slight increase for those babies
10 born in the summer. When stratified by birth weight, the OR for LEW babies was 1.27 (95% CI:
11 0.95, 1.69) per 10-ppb increase in O3 and the OR for normal weight babies was 1.16 (95% CI: 1.01,
12 1.32) per 10-ppb increase in O3.
13 Conversely, two additional studies reported no association between ambient levels of O3 and
14 SIDS. Ritz et al. (2006, 089819) linked birth and death certificates for infants who died between
15 1989 and 2000 to evaluate the influence of outdoor air pollution on infant death in the South Coast
16 Air Basin of California. The authors examined exposure periods 2 weeks, 1 month, 2 months, and
17 6 months before each case subject's death and reported no association between ambient levels of O3
18 and SIDS. Dales et al. (2004, 087342) used time-series analyses to compare the daily mortality rates
19 for SIDS and the daily air pollution concentrations in 12 Canadian cities during the period of
20 1984-1999. Increased daily rates of SIDS were associated with previous day increases in the levels
21 of SO2, NO2, and CO, but not O3 or PM2 5.
March 2011 7-50 DRAFT - DO NOT CITE OR QUOTE
-------
Table 7-6. Brief summary of infant mortality studies
Study
Pereiraet al. (1998,
007264)
Diaz et al. (2004,
089894)
Loomiset al. (1999,
087288)
Ritz et al. (2006,
089819)
Hajat et al. (2007,
093276)
Lin et al. (2004,
Ha et al. (2003,
Romieu et al. (2004,
093074)
Carbajal-Arroyo et al.
(In Press, 667773)
Son et al. (2008,
190323)
Tsai et al. (2006,
090709)
Woodruff etal. (2008,
098386)
Yang et al. (2006,
090760)
Dales et al. (2004,
087342)
Location
Sao Paulo, Brazil
Madrid, Spain
Mexico City,
Mexico
southern California
10 Cities in the UK
Sao Paulo, Brazil
Seoul, South
Korea
Ciudad Juarez,
Mexico
Mexico City,
Mexico
Seoul, South
Korea
Kaohsiung,
Taiwan
Nationwide, US
Taipei, Taiwan
12 Canadian cities
Mean O3 (ppb)
1-h max: 33.8
24-havg: 11.4
24-h avg: 44.1
1-h max: 163.5
24-havg: 21.9-22.1
24-h avg: 20.5-42.6
24-h avg: 38.06
8-havg: 21.2
8-h avg: 43.43-55. 12
1-h max: 103.0
8-ha avg: 25.61
24-h avg: 23.60
24-h avg: 26.6
24-havg: 18.14
24-h: 31 .77
Exposure
Assessment
Citywide avg
Citywide avg
1 monitor
Nearest Monitor
Citywide avg
Citywide avg
Citywide avg
Citywide avg
Citywide avg
Citywide avg
Citywide avg
County wide avg
Citywide avg
Citywide avg
Effect Estimate (95% Cl):
LO-2: 1.00(0.99, 1.01)
NR
LO: 0.99 (0.97, 1 .02)
L1: 0.99(0.96, 1.01)
L2: 1 .00 (0.98, 1 .03)
L3: 1 .03 (1 .00, 1 .05)
L4: 1 .01 (0.98, 1 .03)
L5: 1 .02 (0.99, 1 .04)
LO-2: 1 .02 (0.99, 1 .05)
2 wk before death: 1 .03 (0.93,
1 mo before death: NR
2 mo before death: 0.93 (0.89,
6 mo before death: NR
LO-2: 1 .00 (0.96, 1 .06)
LO: 1.00(0.99, 1.01)
LO: 0.93 (0.90, 0.96)
L1: 0.96(0.90, 1.03)
L2: 0.97(0.91, 1.04)
LO-1 cum: 0.96(0.89, 1.04)
LO-2 cum: 0.94 (0.87, 1 .02)
LO: 1 .00 (0.99, 1 .00)
L1: 0.99(0.99, 0.99)
L2: 0.99 (0.99, 1 .00)
LO-2: 0.99 (0.99, 1 .00)
L(NR): 0.984 (0.976, 0.992)*
LO-2 cum: 1 .02 (0.56, 1 .86)
First 2 mo of life: 1 .04 (0.98, 1 .
LO-2 cum: 1.00 (0.62, 1.61)
LO: NR
L1: NR
L2: NR
L3: NR
L4: NR
L5: NR
Multiday lags of 2-6 days: NR
1.14)
0.97)
,10)
#Relative risk of infant mortality per 10 ppb change in O3
* No increment provided
LO = Lag 0, L1 = Lag 1, L2 = Lag 2, L3 = Lag 3, L4 = Lag 4, L5 = Lag 5, L6 = Lag 6
NR: No quantitative results reported
March 2011
7-51
DRAFT - DO NOT CITE OR QUOTE
-------
7.4.10. Summary and Causal Determination
1 The 2006 O3 AQCD concluded that the limited number of studies that investigated O3
2 demonstrated no associations between O3 and birth outcomes, with the possible exception of birth
3 defects. The current review included an expanded body of evidence on the associations between O3
4 and reproductive and developmental effects. Recent epidemiologic and toxicological studies provide
5 evidence for an effect of prenatal exposure to O3 on pulmonary structure and function, including
6 lung function changes in the newborn, incident asthma, ultrastructural changes in bronchiole
7 development, alterations in placental and pup cytokines, and increased pup airway hyper-reactivity.
8 Also, there is limited toxicological evidence for an effect of prenatal and early life exposure on
9 central nervous system effects, including laterality, brain morphology, neurobehavioral
10 abnormalities, and sleep aberration. Recent epidemiologic studies have begun to explore the effects
11 of O3 on sperm quality, and provide limited evidence for decrements in sperm concentration, while
12 there is limited toxicological evidence for testicular degeneration associated with O3.
13 While the collective evidence for many of the birth outcomes examined is generally
14 inconsistent (including birth defects), there are several well-designed, well-conducted studies that
15 indicate an association between O3 and adverse outcomes. For example, as part of the southern
16 California Children's Health Study, Salam et al. (2005, 087885) observed a concentration-response
17 relationship of decreasing birth weight with increasing O3 concentrations averaged over the entire
18 pregnancy that was clearest above the 30-ppb level (see Figure 7-4). Similarly, Hansen et al. (2008,
19 190273) utilized fetal ultrasonic measurements and found a change in ultrasound measurements
20 associated with O3 during days 31-60 of gestation indicated that increasing O3 concentration
21 decreased an ultrasound measurement for women living within 2 km of the monitoring site.
22 There is no evidence that prenatal or early life O3 concentrations are associated with infant
23 mortality. Collectively, there is limited though positive toxicological evidence for O3-induced
24 developmental effects, including effects on pulmonary structure and function and central nervous
25 system effects. Limited epidemiologic evidence for an effect on prenatal O3 exposure on respiratory
26 development provides coherence with the effects observed in toxicological studies. There is also
27 limited epidemiologic evidence for an association with O3 concentration and decreased sperm
28 concentration. A recent toxicological study provides limited evidence for a possible biological
29 mechanism (histopathology showing impaired spermatogenesis) for such an association.
30 Additionally, though the evidence for an association between O3 concentrations and adverse birth
31 outcomes is generally inconsistent, there are several influential studies that indicate an association
32 with reduced birth weight and restricted fetal growth. Taking into consideration the positive evidence
33 for developmental and reproductive outcomes from toxicological and epidemiological studies, and
34 the few influential birth outcome studies, the evidence is suggestive of a causal relationship between
35 long-term exposures to O3 and reproductive and developmental effects.
March 2011 7-52 DRAFT - DO NOT CITE OR QUOTE
-------
7.5. Central Nervous System Effects
7.5.1. Effects on the Brain and Behavior
1 An epidemiologic study has recently been published examining the association between O3
2 exposure and neurobehavioral effects. Chen et al. (2009, 179945) utilized data from the NHANES
3 III cohort to study the relationship between O3 levels (mean annual O3 concentration 26.5 ppb) and
4 neurobehavioral effects among adults aged 20-59 years. The authors observed an association
5 between annual exposure to O3 and tests measuring coding ability (symbol-digit substitution test)
6 and attention/short-term memory (serial-digit learning test). Each 10-ppb increase in annual O3
7 levels corresponded to an aging-related cognitive performance decline of 3.5 yr for coding ability
8 and 5.3 years for attention/short-term memory. These associations persisted in both crude and
9 adjusted models. There was no association between O3 levels and reaction time tests. The authors
10 conclude that overall, there is an association between long-term O3 exposure and reduced
11 performance on neurobehavioral tests.
12 In a subchronic study, rats were exposed to 0.25 ppm O3 for 4 h/day for 15-90 days (Rivas-
13 Arancibia et al., 2010, 201544). The exposures caused a complex array of responses, including a
14 time-dependent increase in lipid peroxidation products and immunohistochemical changes in the
15 hippocampus, that were correlated with decrements in passive avoidance behavioral tests.
16 A protective effect of estradiol has been observed in ovariectomized female rats exposed to
17 0.25 ppm O3 (4 h/day) for 30 or 60 days (Guevara-Guzman et al., 2009, 596385). In the olfactory
18 bulb, lipid peroxidation was significantly less in rats exposed to O3 and treated daily with estradiol.
19 This protective effect of estradiol was also demonstrated for O3-induced decrements in a selective
20 olfactory recognition memory test and an olfactory-dependent reward test. Similarly, estradiol
21 protected against O3-induced changes in nigral cell morphology and loss of dopamine neurons in rats
22 exposed to O3 for 30 days (Angoa-Perez et al., 2006, 596366). Thus, repeated exposure of rats to O3
23 produces lipid peroxidation at multiple sites in the brain and this oxidative stress is accompanied by
24 gene expression changes and decrements in behavioral tests. Olfactory changes and loss of
25 substantia nigra neurons are associated with Parkinson's disease in humans. Inhibition of these
26 effects with estradiol treatment is consistent with the higher incidence of Parkinson's disease in men
27 and the amelioration of Parkinsonian symptoms by estrogen therapy.
28 Adverse CNS effects have also been demonstrated in newborn and adult rats whose only
29 exposure to O3 occurred in utero. Several neurotransmitters were assessed in male offspring of dams
30 exposed to 1 ppm O3 during the entire pregnancy (Gonzalez-Pina et al., 2008, 475317). The data
31 showed that catecholamine neurotransmitters were affected to a greater degree than indole-amine
32 neurotransmitters in the cerebellum. Adverse CNS changes, including behavioral, cellular, and
33 biochemical effects, have also been observed after in utero exposure to 0.5 ppm O3 for 12 h/day from
34 GD5 to GD20 (Boussouar et al., 2009, 596368). Tyrosine hydroxylase labeling in the nucleus tractus
35 solitarius was increased after in utero exposure to O3 whereas Fos protein labeling did not change.
March 2011 7-53 DRAFT - DO NOT CITE OR QUOTE
-------
1 When these offspring were challenged by immobilization stress, neuroplasticity pathways, which
2 were activated in air offspring, were inhibited in O3 offspring. Although the effect of O3 exposure
3 concentration was not studied in these two in utero studies, it has been examined in one study.
4 Santucci et al. (2006, 596414) investigated behavioral effects and gene expression after in utero
5 exposure of mice to as little as 0.3 ppm O3. Increased defensive/submissive behavior and reduced
6 social investigation were observed in both the 0.3- and 0.6-ppm O3 groups. Changes in gene
7 expression of brain-derived neurotrophic factor (BDNF, increased in striatum) and nerve growth
8 factor (NGF, decreased in hippocampus) accompanied these behavioral changes. Thus, these three
9 studies demonstrate that CNS effects can occur as a result of in utero exposure to O3, and although
10 the mode of action of these effects is not known, it has been suggested that circulating lipid
11 peroxidation products may play a role (Boussouar et al., 2009, 596368). Importantly, these adverse
12 CNS effects occurred in rodent models after in utero only exposure to (semi-) relevant
13 concentrations of O3.
14
7.5.2. Summary and Causal Determination
15 The 2006 O3 AQCD (U.S. EPA, 2006, 088089) included toxicological evidence that acute
16 exposures to O3 are associated with alterations in neurotransmitters, motor activity, short and long
17 term memory, and sleep patterns. Additionally, histological signs of neurodegeneration have been
18 observed. However, evidence regarding chronic exposure and neurobehavioral effects was not
19 available. Recent research in the area of O3-induced neurotoxicity has included several long-term
20 exposure studies. Notably, the first epidemiologic study to examine the relationship between O3
21 exposure and neurobehavioral effects observed an association between annual O3 levels and an
22 aging-related cognitive performance decline in tests measuring coding ability and attention/short-
23 term memory. This observation is supported by studies in rodents which demonstrate oxidative stress
24 in the brain and associated decrements in behavioral tests, including those measuring memory, after
25 subchronic exposure to 0.25 ppm O3. Additionally, neurobehavioral changes are evident in animals
26 whose only exposure to O3 occurred in utero. Collectively, the limited epidemiologic and
27 toxicological evidence is coherent and suggestive of a causal relationship between O3 exposure
28 and adverse CNS effects.
7.6. Carcinogenic and Genotoxic Potential of Ozone
7.6.1. Introduction
29 The radiomimetic and clastogenic qualities of O3, combined with its ability to stimulate
30 proliferation of cells in the respiratory tract, have suggested that O3 could act as a carcinogen.
31 However, toxicological studies of tumorigenesis in the rodent lung have yielded mixed and often
32 confusing results, and the epidemiologic evidence is equally conflicted. The 2006 O3 AQCD
March 2011 7-54 DRAFT - DO NOT CITE OR QUOTE
-------
1 concluded that, "the weight of evidence from recent animal toxicological studies and a very limited
2 number of epidemiologic studies do not support ambient O3 as a pulmonary carcinogen" (U.S. EPA,
3 2006, 088089).
4 Multiple epidemiologic studies reported in the 2006 O3 AQCD examined the direct association
5 between O3 exposure and cancer. The largest of these studies, by Pope et al. (2002, 024689).
6 included 500,000 adults from the American Cancer Society's (ACS) Cancer Prevention II study. In
7 this study, no association was observed between O3 and lung cancer mortality. The Adventist Health
8 Study of Smog (AHSMOG) also examined the association between O3 and lung cancer mortality
9 (Abbey et al., 1999, 047559). There was a positive association between O3 levels and lung cancer
10 mortality among men. No association was reported for women. Another study using the AHSMOG
11 cohort assessed the risk of incident lung cancer (Beeson et al., 1998, 048890). Among males, an
12 association with incidence of lung cancer was observed with increasing O3 concentrations. When
13 stratified by smoking status, the association persisted among never smokers but was null for former
14 smokers. No association was detected for females. The Six Cities Study examined various air
15 pollutants and mortality but did not specifically explore the association between O3 concentrations
16 and lung cancer mortality due to low variability in O3 levels across the cities (Dockery et al., 1993,
17 044457). An ecologic study performed in Sao Paulo City, Brazil examined the correlations between
18 O3 levels in four of the city districts and incident cancer of the larynx and lung reported in 1997
19 (Pereira et al., 2005, 073851). A correlation between the average number of days O3 levels exceeded
20 air quality standards from 1981 to 1990 and cancer incidence was present for larynx cancer but not
21 for lung cancer.
22 Early toxicological research demonstrated lung adenoma1 acceleration in mice with daily
23 exposure to 1 ppm over 15 months (Stokinger, 1962, 015101). Later work demonstrated a significant
24 increase in lung tumor numbers in one strain of mouse (A/J) but not another after exposure to
25 0.3-0.8 ppm O3 (Hassett et al., 1985, 040704: Last et al., 1987, 040830). The A/J mouse strain is
26 known to have a high incidence of spontaneous adenomas, and further studies using this strain found
27 a statistically significant increase in lung tumor incidence after a 9-month exposure to 0.5 ppm and
28 incidence and multiplicity after a 5 month exposure to 0.12 ppm with a 4-month recovery period
29 (Witschi et al., 1999, 011602). However, these findings were discounted by the study authors due to
30 the lack of a clear dose response, and results from the Hassett et al. and Last et al. studies were
31 retrospectively deemed spurious based on what appeared to be unusually low spontaneous tumor
32 incidences in the control groups (Witschi, 1991, 042509). A study of carcinogenicity of O3 by the
33 National Toxicology Program (NTP, 1994, 011143) reported increased incidences of
34 alveolar/bronchiolar adenoma or carcinoma (combined) in female B6C3Fi mice exposed over
35 2 years or a lifetime to 1.0 ppm and marginally increased incidences in male mice exposed to 0.5 and
36 1.0 ppm. Thus there was equivocal evidence of carcinogenic activity in male mice and some
1 NOTE: Although adenomas are benign, over time they may progress to become malignant, at which point they are called
adenocarcinomas. Adenocarcinoma is the predominant lung cancer subtype in most countries, and is the only lung cancer found in
nonsmokers. From page 8-33 of the 1970 O3 AQCD: "No true lung cancers have been reported, however, from experimental exposures
to either O3 alone or any other combination or ingredient of photochemical oxidants."
March 2011 7-55 DRAFT - DO NOT CITE OR QUOTE
-------
1 evidence of carcinogenic activity of O3 in females. Some semblance of a dose-response relationship
2 was also evident in this study.
3 In Fischer-344/N rats (50 of each sex per group), neither a 2-year nor lifetime exposure to O3
4 ranging from 0.12 to 1.0 ppm was found to be carcinogenic (Boorman et al., 1994, 038195).
5 However, a marginally significant carcinogenic effect of 0.2 ppm O3 was reported in a study of male
6 Sprague-Dawley rats exposed for 6 months (n = 50) (Monchaux et al., 1996, 086197). These two
7 studies also examined co-carcinogenicity of O3 with NNK1 (Boorman et al., 1994, 038195) or a
8 relatively high dose of radon (Monchaux et al., 1996, 086197). finding no enhancement of NNK
9 related tumors and a slight non-significant increase in tumor incidence after combined exposure with
10 radon, respectively. Another study exploring co-carcinogenicity was conducted in hamsters. Not only
11 was there no enhancement of chemically induced tumors in the peripheral lung or nasal cavity, but
12 results suggested that O3 could potentially delay or inhibit tumor development (Witschi et al., 1993,
13 043206). Thus there is no concrete evidence that O3 can act as a co-carcinogen.
14 Immune surveillance is an important defense against cancer, and it should be noted that natural
15 killer (NK) cells, which destroy tumor cells in the lung, appear to be inhibited by higher doses of O3
16 and either unaffected or stimulated at lower doses (Section 6.2.5.4, Infection and Adaptive
17 Immunity). This aspect of tumorigenesis adds yet another layer of complexity which may be
18 reflected by conflicting results across studies.
19 The following sections will examine epidemiologic studies of cancer incidence and mortality
20 that have been published since the 2006 O3 AQCD. One study has been published with cancer as the
21 outcome; most studies examine markers of exposure or susceptibility and will be reported on later in
22 this section. Recent toxicological studies are also described.
7.6.2. Lung Cancer Incidence and Mortality
23 A recent re-analysis of the full ACS CPSII cohort by the Health Effects Institute is the only
24 epidemiologic study that has explored the association between O3 and cancer mortality since the last
25 O3 AQCD. Krewski et al. (2009, 191193) conducted an extended follow-up of the cohort
26 (1982-2000). Mean O3 levels [obtained from the Aerometric Information Retrieval System (AIRS)
27 for 1980] were 22.91 ppb for the full year and 30.15 ppb for the summer months (April-September).
28 No association was reported between lung cancer mortality and O3 (HR 1.00 [95% CI: 0.96-1.04]
29 per 10 ppb O3). Additionally, no association was observed when O3 was restricted to the summer
30 months. There was also no association present in a sub-analysis of the cohort examining the
31 relationship between O3 and lung cancer mortality in the Los Angeles area.
32 Since the 2006 O3 AQCD, two toxicological studies have examined potential carcinogenicity
33 of O3 (Kim and Cho, 2009, 200775: Kim and Cho, 2009, 200773). Looking across both studies,
34 which used the same mouse strain as the National Toxicology Program study described above,
35 0.5 ppm O3 alone or in conjunction with chemical tumor inducers did not enhance lung tumor
1 4-(N-nitrosomethylamino)-1 -(3-pyridyl)-1 -butanone
March 2011
7-56
DRAFT - DO NOT CITE OR QUOTE
-------
1 incidence in males or females. However, a 10% incidence of oviductal carcinoma was observed in
2 mice exposed to 0.5 ppm O3 for 16 weeks. The implications of this observation are unclear,
3 particularly in light of the lack of statistical information reported. Additionally, there is no mention
4 of oviductal carcinoma after 32 weeks of exposure, and no oviductal carcinoma was observed after
5 one year of exposure.
7.6.3. DNA Damage
6 The potential for genotoxic effects relating to O3 exposure was predicted from the
7 radiomimetic properties of O3. The decomposition of O3 in water produces OH and HO2 radicals, the
8 same species that are generally considered to be the biologically active products of ionizing
9 radiation. Ozone has been observed to cause degradation of DNA in a number of different models
10 and bacterial strains. Until the reports of Zelac et al.(1971, 039743; 1971, 039752). the toxic effects
11 of O3 were generally assumed to be confined to the tissues directly in contact with the gas, such as
12 the respiratory epithelium. Due to the highly reactive nature of O3, little systemic absorption was
13 predicted. Zelac et al, however, reported a significant increase in chromosome aberrations in
14 peripheral blood lymphocytes from Chinese hamsters exposed to 0.2 ppm for 5 hours. Other in vivo
15 exposure studies found increased DNA strand breaks in respiratory cells from guinea pigs (Ferng et
16 al., 1997, 083578) and mice (Bornholdt et al., 2002, 036677) but only with exposure to higher doses
17 of O3 (1 ppm for 72 hours and 1 or 2 ppm for 90 minutes, respectively). In other studies there were
18 no observations of chromosomal aberrations in germ cells, but mutagenic effects have been seen in
19 offspring of mice exposed to 0.2 ppm during gestation (blepharophimosis or dysplasia of the
20 eyelids). The overall evidence for mutagenic activity from in vitro studies is positive, and in the
21 National Toxicology Program report described above, O3 was found to be mutagenic in Salmonella,
22 with and without S9 metabolic activation. No new toxicological studies of DNA damage have
23 become available since the 2006 O3 AQCD.
24 A number of epidemiologic studies looked at the association between O3 and DNA and cellular
25 level damages. These changes may be relevant to mechanisms leading to cancers development and
26 serve as early indicators of elevated risk of mutagenicity.
27 Two studies performed in California examined cytogenetic damage in relation to O3
28 exposures. Huen et al. (2006, 089035) examined cytogenetic damage among African American
29 children and their mothers in Oakland, CA. Increased O3 (mean monthly 8-h O3 concentrations
30 ranged from about 30 ppb in April to 14 ppb in November) was associated with increased
31 cytogenetic damage (micronuclei frequency among lymphocytes and buccal cells) even after
32 adjustment for household/personal smoking status and distance-weighted traffic density. Chen et al.
33 (2006, 196504) recruited college students at the University or California, Berkeley who reported
34 never smoking and compared their levels of cytogenetic damage (micronuclei frequency from buccal
35 cells) in the spring and fall. Cytogenetic damage was greater in the fall, which the authors attributed
36 to the increase in O3 over the summer. However, O3 levels over 2, 7, 10, 14, or 30 days
37 (concentrations not given) before collection of buccal cells did not correlate with cytogenetic
March 2011 7-57 DRAFT - DO NOT CITE OR QUOTE
-------
1 damage. Estimated lifetime O3 exposure was also not correlated with cytogenetic damage.
2 Additionally, the authors exposed a subset of the students (n=15) to 200 ppb O3 for 4 hours while the
3 students exercised intermittently. Ozone was found to be associated with an increase in cytogenetic
4 damage in degenerated cells but not in normal cells 9-10 days after exposure. Increased cytogenetic
5 damage was also noted in peripheral blood lymphocytes collected 18 hours after exposure.
6 A study performed in Mexico recruited 55 male workers working indoors (n=27) or outdoors
7 (n=28) in Mexico City or Puebla, Mexico in order to study the relationship between O3 and DNA
8 damage (detected from peripheral blood samples using the Comet assay) (Tovalin et al., 2006,
9 091322). The median estimated daily O3 concentrations were estimated to be 28.5 ppb for outdoor
10 workers and 5.1 ppb for indoor workers in Mexico City and 36.1 ppb for outdoor workers and
11 19.5 ppb for indoor workers in Puebla. Overall, a positive correlation between O3 levels and DNA
12 damage was observed. However, when examining the relationship by city and workplace, only DNA
13 damage in outdoor workers in Mexico City remained correlated with O3 levels.
14 Three studies examining the relationship between O3 and DNA-level damage have been
15 performed in Europe. The largest of these was the GenAir case-control study, which was nested
16 within the European Prospective Investigation into Cancer and Nutrition (EPIC) study, and included
17 individuals recruited between 1993 and 1998 from ten European countries. Only non-smokers (must
18 not have smoked for at least 10 years prior to enrollment) were enrolled in the study. The researchers
19 examined DNA adduct levels (DNA bonded to cancer-causing chemicals) and their relationship with
20 O3 concentrations (concentrations not given) (Peluso M Hainaut et al., 2005, 089057). A positive
21 association was seen between DNA adduct levels and O3 concentrations from 1990-1994 but not O3
22 levels from 1995-1999. In adjusted analyses with DNA adduct levels dichotomized as high and low
23 (detectable versus non-detectable), the OR was 1.97 (95% CI: 1.08, 3.58) when comparing the upper
24 tertile of O3 concentration to the lower two tertiles. Two other European studies were conducted in
25 Florence, Italy. The most recent of these enrolled individuals from the EPIC study into a separate
26 study between March and September of 1999 (Palli et al., 2009, 196688). The purpose of the study
27 was to examine oxidative DNA damage (determined by Comet assay using blood lymphocytes) in
28 association with varying periods of O3 exposure. The researchers observed that longer periods of
29 high O3 exposure (concentrations not given) were more strongly correlated with oxidative DNA
30 damage than shorter exposures (i.e., the rho [p-value] was 0.26 [0.03] for 0-10 days and 0.35 [0.002]
31 for 0-90 days). This correlation was stronger among men compared to women. The correlations for
32 all time periods had p-values <0.05 for ex- and never-smokers. For current smokers, the correlation
33 was only observed among time periods25 days. When adjusted for age, gender, smoking history,
34 traffic pollution exposure, period of blood draw, and area of residence, the association between O3
35 levels and oxidative DNA damage was positive for O3 levels 0-60 days, 0-75 days, and 0-90 days
36 prior to blood draw. Positive, statistically significant associations were not observed among shorter
37 time periods. The other study performed in Florence recruited healthy volunteers who reported being
38 non-smokers or light smokers (Giovannelli et al., 2006, 199894). The estimated O3 levels during the
39 study ranged from approximately 4-40 ppb for 3-day averages, 5-35 ppb for 7-day averages, and
March 2011 7-58 DRAFT - DO NOT CITE OR QUOTE
-------
1 7.5-32.5 ppb for 30-day averages. Ozone concentrations were correlated with DNA strand breaks
2 (measured from blood lymphocytes) over longer exposure periods (p-value: 0.002 at 30 days,
3 p-value: 0.04 at 7 days; p-value: 0.17 at 3 days). This association was robust to control for
4 temperature, solar radiation, gender, and age. No association was seen between O3 concentrations
5 and measures of oxidative DNA damage at 3, 7, or 30 days.
7.6.4. Summary and Causal Determination
6 The 2006 O3 AQCD (U.S. EPA, 2006, 088089) reported that evidence did not support ambient
7 O3 as a pulmonary carcinogen. Since the 2006 O3 AQCD, very few epidemiologic and toxicological
8 studies have been published that examine O3 as a carcinogen, but collectively, study results indicate
9 that O3 may contribute to DNA damage. Overall, the evidence is inadequate to determine if a
10 causal relationship exists between ambient O3 exposures and cancer.
7.7. Mortality
11 A limited number of epidemiologic studies have assessed the relationship between long-term
12 exposure to O3 and mortality in adults. The 2006 O3 AQCD concluded that an insufficient amount of
13 evidence existed "to suggest a causal relationship between chronic O3 exposure and increased risk
14 for mortality in humans" (U.S. EPA, 2006, 088089). In addition to the infant mortality studies
15 discussed in Section 7.4.9, two additional studies have been conducted among adults since the last
16 review; an ecologic study that finds no association between mortality and O3, and a reanalysis of the
17 ACS cohort that specifically points to a relationship between long-term O3 exposure and an
18 increased risk of respiratory mortality. These studies supplement the evidence from long-term cohort
19 studies characterized in previous reviews of O3, (U.S. EPA, 1996, 017831: U.S. EPA, 2006, 088089)
20 and are summarized here briefly.
21 In the Harvard Six Cities Study (Dockery et al., 1993, 044457). adjusted mortality rate ratios
22 were examined in relation to long-term mean O3 concentrations in six cities: Topeka, KS; St. Louis,
23 MO; Portage, WI; Harriman, TN; Steubenville, OH; and Watertown, MA. Mean O3 concentrations
24 from 1977 to 1985 ranged from 19.7 ppb in Watertown to 28.0 ppb in Portage. Long-term mean O3
25 concentrations were not found to be associated with mortality in the six cities. However, the authors
26 noted that "The small differences in O3 levels among the (six) cities limited the power of the study to
27 detect associations between mortality and O3 levels." In addition, while total and cardio-pulmonary
28 mortality were considered in this study, respiratory mortality was not specifically considered.
29 In a subsequent large prospective cohort study of approximately 500,000 U.S. adults, Pope et
30 al. (2002, 024689) examined the effects of long-term exposure to air pollutants on mortality
31 (American Cancer Society, Cancer Prevention Study II). All-cause, cardiopulmonary, lung cancer
32 and other mortality risk estimates for long-term O3 exposure are shown in Figure 7-5. While no
33 consistently significant positive associations were observed between O3 and mortality, the mortality
34 risk estimates were larger when analyses considered more accurate exposure metrics, rising when the
March 2011 7-59 DRAFT - DO NOT CITE OR QUOTE
-------
1 entire period was considered (versus just at the start of the study) and becoming marginally
2 significant when the exposure estimate was restricted to the summer months (July to September),
3 especially when considering cardio-pulmonary deaths. In contrast, consistent positive and significant
4 effects of PM2 5 were observed for both lung cancer and cardio-pulmonary mortality.
t^-
cc
A
B
C
1.50-1
1.40-
1.30-
1.20-
1.10-
1.00-
0.90-
0.80-
All Cause
Mortality
I I
§ 1 1
A B C
Years of Data
Collection
1980-1981
1982-1998
1982-1 998 (July -Sept)
Cardiovascular
Mortality Lung Cancer All Other Causes
' T T ' Mortality Mortality
1 J, I.I 1 1
I * [ o j I 1 ^
ABC ABC ABC
Number of Number of Participants H ., .. /-.s .. /ori.
Metropolitan Areas (in thousands) 1-hMaxO Mean (SD)
134 559 47.9(11.0)
119 525 45.5(7.3)
134 557 59.7(12.8)
Source: Derived with permission from American Medical Association, Pope etal. (2002, 0246891.
Figure 7-5. Adjusted ozone-mortality relative risk estimates (95% Cl) by time period of analysis
per subject-weighted mean ozone concentration in the Cancer Prevention Study II
by the American Cancer Society.
5 A study by Abbey et al. (1999, 047559) examined the effects of long-term air pollution
6 exposure, including O3, on all-cause (n = 1,575), cardiopulmonary (n = 1,029), nonmalignant
7 respiratory (n = 410), and lung cancer (n = 30) mortality in the long-term prospective Adventist
8 Health Study of Smog (AHSMOG) of 6,338 nonsmoking, non-Hispanic white individuals living in
9 California. A particular strength of this study was the extensive effort devoted to assessing long-term
10 air pollution exposures, including interpolation to residential and work locations from monitoring
11 sites over time and space. No associations with long-term O3 exposure were observed for all cause,
12 cardiopulmonary, and nonmalignant respiratory mortality. In a follow-up, Chen et al. (2005, 087942)
13 utilized data from the AHSMOG study and reported no significant associations between long-term
14 O3 exposure (mean O3 concentration 26.2 ppb) and fatal coronary heart disease. Thus, no association
15 of chronic O3 exposure with mortality outcomes has been detected in this study.
16 Lipfert et al. (2000, 004087; 2003, 052250) reported positive effects on all-cause mortality for
17 peak O3 exposures (95th percentile levels) in the U.S. Veterans Cohort study of approximately
18 50,000 middle-aged men recruited with a diagnosis of hypertension. The actual analysis involved
19 smaller subcohorts based on exposure and mortality follow-up periods. Four separate exposure
March 2011
7-60
DRAFT - DO NOT CITE OR QUOTE
-------
1 periods were associated with three mortality follow-up periods. For concurrent exposure periods,
2 peak O3 was positively associated with all-cause mortality, with a 9.4% (95% CI: 0.4, 18.4) excess
3 risk per mean 95th percentile O3 less estimated background level (not stated). "Peak" refers, in this
4 case, to the 95th percentile of the hourly measurements, averaged by year and county. In a further
5 analysis, Lipfert et al. (2003, 052250) reported the strongest positive association for concurrent
6 exposure to peak O3 for the subset of subjects with low diastolic blood pressure during the 1982 to
7 1988 period. Two more recent studies of this cohort focused specifically on traffic density (Lipfert et
8 al., 2006, 088756: 2006, 088218). Lipfert (2006, 088218) concluded that: "Traffic density is seen to
9 be a significant and robust predictor of survival in this cohort, more so than ambient air quality, with
10 the possible exception of O3," reporting a significant O3 effect even with traffic density included in
11 the model: RR=1.080 per 40 ppb peak O3 (95% CI: 1.019, 1.146). In Lipfert (2006, 088756X which
12 considers only the EPA Speciation Trends Network (STN) sites, O3 drops to non-significant predictor
13 of total mortality for this cohort. However, the authors acknowledge that: "Peak O3 has been
14 important in analyses of this cohort for previous periods, but in the STN data set, this variable has
15 limited range and somewhat lower values and its small coefficient of variation results in a relatively
16 large standard error." The restriction to subjects near STN sites likely reduced the power of this
17 analysis, though the size of the remaining subjects considered was not reported in this paper. In
18 addition, these various Veteran's Cohort studies considered only total mortality, and did not consider
19 mortality on a by-cause basis.
20 An ecological study in Brisbane, Australia used a geospatial approach to analyze the
21 association of long-term exposure to gaseous air pollution with cardio-respiratory mortality, in the
22 period 1996-2004 (Wang et al., 2009, 199990). A generalized estimating equations model was
23 employed to investigate the impact of NO2, O3 and SO2, but PM was not addressed. The results
24 indicated that long-term exposure to SO2 was associated with cardio-respiratory mortality, but the
25 fact that this study considered only one city, and that the range of O3 exposure across that city
26 (23.7-35.6 ppb) was low and slight in variation in comparison to the range of other pollutants across
27 the city, limited study power. In addition, confounding factors (e.g., smoking) could not be addressed
28 at the individual level in this ecological study. Respiratory mortality was not evaluated separately.
29 In the most recent follow-up analysis of the ACS cohort (Jerrett et al., 2009, 194160).
30 cardiopulmonary deaths were subdivided into respiratory and cardiovascular, separately, as opposed
31 to combined in the Pope et al. (2002, 024689) work. This analysis utilized the ACS cohort with data
32 from 1977 through 2000 (mean O3 concentration ranged from 33.3 to 104.0 ppb). A 10-ppb
33 increment in exposure to O3 elevated the risk of death from the cardiopulmonary, cardiovascular,
34 ischemic heart disease, and respiratory causes. Inclusion of the concentration of PM25 measured in
35 1999-2000 as a co-pollutant attenuated the association with exposure to O3 for all end points except
36 death from respiratory causes, for which a significant association persisted (Table 7.7). The
37 association between increased O3 concentrations and increased risk of death from respiratory causes
38 was insensitive to the use of a random-effects survival model allowing for spatial clustering within
39 the metropolitan area and state of residence, and to adjustment for several ecologic variables
March 2011 7-61 DRAFT - DO NOT CITE OR QUOTE
-------
1 considered individually. Subgroup analyses showed that temperature and region of country, but not
2 sex, age at enrollment, body-mass index, education, or PM2 5 concentration, modified the effects of
3 O3 on the risk of death from respiratory causes (i.e., risks were higher at higher temperature, and in
4 the Southeast, Southwest, and Upper Midwest). Ozone threshold analyses indicated that the
5 threshold model was not a better fit to the data (p > 0.05) than a linear representation of the overall
6 Os-mortality association. Overall, this new analysis indicates that long-term exposure to PM2 5
7 increases risk of cardiac death, while long-term exposure to O3 is specifically associated with an
8 increased risk of respiratory death.
9
10
11
12
13
14
15
16
17
18
19
20
Table 7-7. Relative risk (and 95% Cl) of death attributable to a 10-ppb change in the ambient ozone
concentration
Cause of Death
Any Cause
Cardiopulmonary
Respiratory
Cardiovascular
Ischemic Heart Disease
O3 (96 MSAs)
1.001
1.014
1.029
1.011
1.015
(0,
(1
(1
(1
(1
.996,
.007,
.010,
.003,
.003,
1 .007)
1 .022)
1 .048)
1 .023)
1 .026)
O3 (86 MSAs)
1.001
1.016
1.027
1.014
1.017
(0,
(1
(1
(1
(1
.996,
.008,
.007,
.005,
.006,
1 .007)
1 .024)
1 .046)
1 .023)
1 .029)
03+PM2.5(86
0.989
0.992
1.040
0.983
0.973
(0.981,
(0.982,
(1.013,
(0.971,
(0.958,
MSAs)
0.996)
1 .003)
1 .067)
0.994)
0.988)
Source: Used with permission from Massachusetts Medical Society, Jerrett et al. (2009, 194160)
7.7.1. Summary and Caus al Determination
The 2006 O3 AQCD concluded that an insufficient amount of evidence existed "to suggest a
causal relationship between chronic O3 exposure and increased risk for mortality in humans"
(U.S. EPA, 2006, 088089). Two additional studies have been conducted since the last review, an
ecologic study that finds no association between mortality and O3 (Wang et al., 2009, 199990). and a
re-analysis of the ACS cohort that specifically points to a relationship between long-term O3
exposure and an increased risk of respiratory mortality (Jerrett et al., 2009, 194160). The findings
from the Jerrett et al. (2009, 194160) study are consistent and coherent with the evidence from
epidemiologic, controlled human exposure, and animal toxicological studies for the effects of short-
and long-term exposure to O3 on respiratory effects. Additionally, the evidence for short- and long-
term respiratory morbidity provides biological plausibility for mortality due to respiratory disease.
Collectively, the evidence is suggestive of a causal relationship between long-term Os exposures
and all-cause mortality.
March 2011
7-62
DRAFT - DO NOT CITE OR QUOTE
-------
References
A list of all epidemiologic references considered for inclusion in this chapter can be found at
http://hero.epa. gov/index.cfm?action=search.do&submit=Search&group_id=403
A list of all toxicological references considered for inclusion in this chapter can be found at
http://hero.epa. gov/index.cfm?action=search.do&submit=Search&group_id=401
A list of all controlled human exposure references considered for inclusion in this chapter can be found at
http://hero.epa. gov/index.cfm?action=search.do&submit=Search&group_id=477
Abbey, D. E.; Nishino, N.; McDonnell, W. R; Burchette, R. I; Knutsen, S. R; Beeson, W. L.; Yang, J. X. (1999). Long-
term inhalable particles and other air pollutants related to mortality in nonsmokers. Am J Respir Crit Care Med,
159: 373-382. 047559
Akinbami, L. J.; Lynch, C. D.; Parker, J. D.; Woodruff TJ (2010). The association between childhood asthma prevalence
and monitored air pollutants in metropolitan areas, United States, 2001-2004. Environ Res, 110: 294-301.
http://dx.doi.0rg/10.1016/j.envres.2010.01.001 378580
Angoa-Perez, M.; Jiang, H.; Rodriguez, A. L; Lemini, C.; Levine, R. A.; Rivas-Arancibia, S. (2006). Estrogen counteracts
ozone-induced oxidative stress and nigral neuronal death. Neuroreport, 17: 629-633.
http://www.ncbi. nlm.nih.gov/entrez/query. fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids= 16603 924.
596366
Arito, H.; Uchiyama, L; Yokoyama, E. (1992). Acute effects of ozone on EEG activity, sleep-wakefulness and heart rate in
rats. Ind Health, 30: 23-34. 042759
Auten, R. L.; Potts, E. N.; Mason, S. N.; Fischer, B.; Huang, Y; Foster, W. M. (2009). Maternal exposure to particulate
matter increases postnatal ozone-induced airway hyperreactivity in juvenile mice. Am J Respir Crit Care Med, 180:
1218-1226. http://dx.doi.org/10.1164/rccm.200901-01160C 200760
Beeson, W. L.; Abbey, D. E.; Knutsen, S. F. (1998). Long-term concentrations of ambient air pollutants and incident lung
cancer in California adults: results from the AHSMOG study. Environ Health Perspect, 106: 813-823. 048890
Berhane, K.; Gauderman, W. J.; Stram, D. O.; Thomas, D. C. (2004). Statistical issues in studies of the long-term effects of
air pollution: The southern California children's health study. Stat Sci, 19: 414-449.
http://dx.doi.org/10.1214/088342304000000413626732
Bignami, G; Musi, B.; Dell'Omo, G; Laviola, G; Alleva, E. (1994). Limited effects of ozone exposure during pregnancy
on physical and neurobehavioral development of CD-I mice. Toxicol Appl Pharmacol, 129: 264-271.
http://dx.doi.org/10.1006/taap.1994.1251 076063
Bobak, M. (2000). Outdoor air pollution, low birth weight, and prematurity. Environ Health Perspect, 108: 173-176.
011448
Boorman, G. A.; Hailey, R.; Grumbein, S.; Chou, B. J.; Herbert, R. A.; Goehl, T; Mellick, P. W.; Roycroft, J. H.; Haseman,
J. K.; Sills, R. (1994). Toxicology and carcinogenesis studies of ozone and ozone 4-(N-nitrosomethylamino)-l-(3-
pyridyl)-l-butanone in Fischer-344/N rats. Toxicol Pathol, 22: 545-554. 038195
Bornholdt, J.; Dybdahl, M.; Vogel, U.; Hansen, M.; Loft, S.; Wallin, H. (2002). Inhalation of ozone induces DNA strand
breaks and inflammation in mice. DNA Repair, 520: 63-71. 036677
Boussouar, A.; Araneda, S.; Hamelin, C.; Soulage, C.; Kitahama, K.; Dalmaz, Y. (2009). Prenatal ozone exposure abolishes
stress activation of Fos and tyrosine hydroxylase in the nucleus tractus solitarius of adult rat. Neurosci Lett, 452:
75-78.
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids= 19444957.
596368
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments such as the Integrated Science Assessments (ISAs) and the Integrated Risk Information System (IRIS).
March 2011 7-63 DRAFT - DO NOT CITE OR QUOTE
-------
Bracken, M. B.; Triche, E. W.; Belanger, K.; Saftlas, A.; Beckett, W. S.; Leaderer, B. P. (2003). Asthma symptoms, severity,
and drug therapy: a prospective study of effects on 2205 pregnancies. Obstet Gynecol, 102: 739-753. 156288
Brauer, M.; Lencar, C.; Tamburic, L.; Koehoorn, M.; Demers, P.; Karr, C. (2008). A cohort study of traffic-related air
pollution impacts on birth outcomes. Environ Health Perspect, 116: 680-686. 156292
Breton, C. V.; Salam, M. T; Vora, H.; Gauderman, W. J.; Gilliland, F. D. (2011). Genetic variation in the glutathione
synthesis pathway, air pollution, and children's lung function growth. Am J Respir Crit Care Med, 183: 243-248.
http://dx.doi.org/10.1164/rccm.201006-0849OC687660
Calderon-Garciduenas, L.; Mora-Tiscareno, A.; Fordham, L. A.; Chung, C. J.; Valencia-Salazar, G; Flores-Gomez, S.; Solt,
A. C.; Gomez-del Campo, A.; Jardon-Torres, R.; Henriquez-Roldan, C.; Hazucha, M. J.; Reed, W. (2006). Lung
radiology and pulmonary function of children chronically exposed to air pollution. Environ Health Perspect, 114:
1432-1437.091253
Carbajal-Arroyo, L.; Miranda-Soberanis, V; Medina-Ramon, M.; Rojas-Bracho, L.; Tzintzun, G; Solis-Gutierrez, P.;
Mendez-Ramirez, L; Hurtado-Diaz, M.; Schwartz, J.; Romieu, I. (In Press). Effect of PM10 and O3 on infant
mortality among residents in the Mexico City Metropolitan Area: A case-crossover analysis, 1997-2005. J
Epidemiol Community Health, http://dx.doi.org/10.1136/jech.2009.101212 667773
Carey, S. A.; Minard, K. R.; Trease, L. L.; Wagner, J. G; Garcia, G. J.; Ballinger, C. A.; Kimbell, J. S.; Plopper, C. G;
Corley, R. A.; Postlethwait, E. M.; Harkema, J. R.; Einstein, D. R. (2007). Three-dimensional mapping of ozone-
induced injury in the nasal airways of monkeys using magnetic resonance imaging and morphometric techniques.
Toxicol Pathol, 35: 27-40. http://dx.doi.org/10.1080/01926230601072343 195752
Chang MM-J; Wu, R.; Plopper, C. G; Hyde, D. M. (1998). IL-8 is one of the major chemokines produced by monkey
airway epithelium after ozone-induced injury. Am J Physiol, 275: L524-L532. 011983
Charpin, D.; Pascal, L.; Birnbaum, J.; Armengaud, A.; Sambuc, R.; Lanteaume, A.; Vervloet, D. (1999). Gaseous air
pollution and atopy. Clin Exp Allergy, 29: 1474-1480. 015152
Chen, C.-Y; Bonham, A. C.; Plopper, C. G; Joad, J. P. (2003). Plasticity in respiratory motor control: selected contribution:
neuroplasticity in nucleus tractus solitarius neurons following episodic ozone exposure in infant primates. J Appl
Physiol, 94: 819-827. 035576
Chen, C.; Arjomandi, M.; Balmes, J.; Tager, L; Holland N (2007). Effects of Chronic and Acute Ozone Exposure on Lipid
Peroxidation and Antioxidant Capacity in Healthy Young Adults. Environ Health Perspect, 115: 1732-1737.
http://dx.doi.org/10.1289/ehp.10294 145956
Chen, C.; Arjomandi, M.; Qin, H.; Balmes, J.; Tager, L; Holland, N. (2006). Cytogenetic damage in buccal epithelia and
peripheral lymphocytes of young healthy individuals exposed to ozone. Mutagenesis, 21: 131-137.
http://dx.doi.org/10.1093/mutage/gel007 196504
Chen, J.-C.; Schwartz, J. (2009). Neurobehavioral effects of ambient air pollution on cognitive performance in US adults.
Neurotoxicology, 30: 231-239. http://dx.doi.Org/10.1016/j.neuro.2008.12.011 179945
Chen, L. H.; Knutsen, S. F.; Shavlik, D.; Beeson, W. L.; Petersen, F.; Ghamsary, M.; Abbey, D. (2005). The association
between fatal coronary heart disease and ambient particulate air pollution: Are females at greater risk? Environ
Health Perspect, 113: 1723-1729. 087942
Chen, L.; Yang, W.; Jennison, B. L.; Goodrich, A.; Omaye, S. T. (2002). Air pollution and birth weight in northern Nevada,
1991-1999. Inhal Toxicol, 14: 141-157. 024945
Chuang, G. C.; Yang, Z.; Westbrook, D. G; Pompilius, M.; Ballinger, C. A.; White, R. C.; Krzywanski, D. M.; Postlethwait,
E. M.; Ballinger, S. W. (2009). Pulmonary ozone exposure induces vascular dysfunction, mitochondrial damage,
and atherogenesis. Am J Physiol Lung Cell Mol Physiol, 297: L209-L216.
http://dx.doi.org/10.1152/aiplung.00102.2009 197202
Chuang, K. J.; Yan, Y H.; Chiu, S. Y; Cheng, T. J. (2011). Long-term air pollution exposure and risk factors for
cardiovascular diseases among the elderly in Taiwan. Occup Environ Med, 68: 64-68.
http://dx.doi.org/10.1136/oem.2009.052704670846
Clark, N. A.; Demers, P. A.; Karr, C. J.; Koehoorn, M.; Lencar, C.; Tamburic, L.; Brauer, M. (2010). Effect of early life
exposure to air pollution on development of childhood asthma. Environ Health Perspect, 118: 284-290.
http://dx.doi.org/10.1289/ehp.0900916594440
March 2011 7-64 DRAFT - DO NOT CITE OR QUOTE
-------
Clifton, V. L.; Giles, W. B.; Smith, R.; Bisits, A. T.; Hempenstall, P. A.; Kessell, C. G; Gibson, P. G (2001). Alterations of
placental vascular function in asthmatic pregnancies. Am J Respir Crit Care Med, 164: 546-554. 156360
Cole, M. P.; Freeman, B. A. (2009). Promotion of cardiovascular disease by exposure to the air pollutant ozone. Am J
Physiol Lung Cell Mol Physiol, 297: L209-L216. 597507
Conti, D. V.; Cortessis, V.; Molitor, J.; Thomas, D. C. (2003). Bayesian modeling of complex metabolic pathways. Hum
Hered, 56: 83-93. http://dx.doi.org/10.1159/000073736 626696
Dales, R.; Burnett, R. T.; Smith-Doiron, M.; Stieb, D. M.; Brook, J. R. (2004). Air pollution and sudden infant death
syndrome. Pediatrics, 113: 628-631. 087342
Darrow, L. A.; Klein, M.; Flanders, W. D.; Waller, L. A.; Correa, A.; Marcus, M.; Mulholland, J. A.; Russell, A. G; Tolbert,
P. E. (2009). Ambient air pollution and preterm birth: a time-series analysis. Epidemiology, 20: 689-698. 195818
Dell'Omo, G; Wolfer, D.; Alleva, E.; Lipp, H.-P (1995). Developmental exposure to ozone induces subtle changes in
swimming navigation of adult mice. Toxicol Lett, 81: 91-99. 080787
Diaz, J.; Linares, C.; Garcia-Herrera, R.; Lopez, C.; Trigo, R. (2004). Impact of temperature and air pollution on the
mortality of children in Madrid. J Occup Environ Med, 46: 768-774. 089894
Dockery, D. W; Pope CA 3rd; Xu, X.; Spengler, J. D.; Ware, J. H.; Fay, M. E.; Ferris, B. G. Jr; Speizer, F. E. (1993). An
association between air pollution and mortality in six US cities. N Engl J Med, 329: 1753-1759. 044457
Duan, X.; Buckpitt, A. R.; Pinkerton, K. E.; Ji, C.; Plopper, C. G. (1996). Ozone-induced alterations in glutathione in lung
subcompartments of rats and monkeys. Am J Respir Cell Mol Biol, 14: 70-75. 080791
Duan, X.; Buckpitt, A. R.; Plopper, C. G. (1993). Variation in antioxidant enzyme activities in anatomic subcompartments
within rat and rhesus monkey lung. Toxicol Appl Pharmacol, 123: 73-82. 086326
Dugandzic, R.; Dodds, L.; Stieb, D.; Smith-Doiron, M. (2006). The association between low level exposures to ambient air
pollution and term low birth weight: A retrospective cohort study. Environ Health, 5:3.
http://dx.doi.org/10.1186/1476-069X-5-3088681
Ercan, H.; Birben, E.; Dizdar, E. A.; Keskin, O.; Karaaslan, C.; Soyer, O. U.; Out, R.; Sackesen, C.; Besler, T; Kalayci, O.
(2006). Oxidative stress and genetic and epidemiologic determinants of oxidant injury in childhood asthma. J
Allergy Clin Immunol, 118: 1097-1104. 595172
Evans, M. J.; Fanucchi, M. V; Baker, G. L.; Van Winkle, L. S.; Pantle, L. M.; Nishio, S. J.; Schelegle, E. S.; Gershwhin, L.
J.; Miller, L. A.; Hyde, D. M.; Sannes, P. L.; Plopper, C. G. (2003). Atypical development of the tracheal basement
membrane zone of infant rhesus monkeys exposed to ozone and allergen. Am J Physiol, 285: L931-L939. 048167
Evans, M. J.; Fanucchi, M. V; Baker, G. L.; Van Winkle, L. S.; Pantle, L. M.; Nishio, S. J.; Schelegle, E. S.; Gershwin, L.
J.; Miller, L. A.; Hyde, D. M.; Plopper, C. G. (2004). The remodelled tracheal basement membrane zone of infant
rhesus monkeys after 6 months of recovery. Clin Exp Allergy, 34: 1131-1136. http://dx.doi.Org/10.llll/j.1365-
2222.2004.02004.x CEA2004 596379
Fanucchi, M. V; Plopper, C. G; Evans, M. J.; Hyde, D. M.; Van Winkle, L. S.; Gershwin, L. J.; Schelegle, E. S. (2006).
Cyclic exposure to ozone alters distal airway development in infant rhesus monkeys. Am J Physiol Lung Cell Mol
Physiol, 291: L644-L650. http://dx.doi.org/10.1152/ajplung.00027.2006 096491
Fanucchi, M. V; Wong, V J.; Hinds, D.; Tarkington, B. K.; Van Winkle, L. S.; Evans, M. J.; Plopper, C. G. (2000).
Repeated episodes of exposure to ozone alters postnatal development of distal conducting airways in infant rhesus
monkeys [Abstract]. Am J Respir Crit Care Med, 161: A615. 012284
Feichtinger, W.; Papalambrou, K.; Poehl, M.; Krischker, U.; Neumann, K. (1997). Smoking and in vitro fertilization: A
meta-analysis. J Assist Reprod Genet, 14: 596-599. http://dx.doi.Org/10.1023/A:1022584802711 625251
Ferng, S.-F.; Castro, C. E.; Afifi, A. A.; Bermudez, E.; Mustafa, M. G. (1997). Ozone-induced DNA strand breaks in guinea
pig tracheobronchial epithelial cells. J Toxicol Environ Health, 51: 353-367. 083578
Forbes, L. J. L.; Kapetanakis, V; Rudnicka, A. R.; Cook, D. G; Bush, T; Stedman, J. R.; Whincup, P. H.; Strachan, D. P.;
Anderson, H. R. (2009). Chronic exposure to outdoor air pollution and lung function in adults. Thorax, 64: 657-
663. 595425
March 2011 7-65 DRAFT - DO NOT CITE OR QUOTE
-------
Forbes, L. I; Patel, M. D.; Rudnicka, A. R.; Cook, D. G; Bush, T.; Stedman, J. R.; Whincup, P. H.; Strachan, D. P.;
Anderson, R. H. (2009). Chronic exposure to outdoor air pollution and markers of systemic inflammation.
Epidemiology, 20: 245-253. http://dx.doi.org/10.1097/EDE.Ob013e318190ea3fl90351
Frischer, T.; Studnicka, M.; Gartner, C.; Tauber, E.; Horak, R; Veiter, A.; Spengler, J.; Ruhr, J.; Urbanek, R. (1999). Lung
function growth and ambient ozone: a three-year population study in school children. Am J Respir Crit Care Med,
160: 390-396. 001037
Frischer, T.; Studnicka, M.; Halmerbauer, G; Horak, R; Gartner, C.; Tauber, E.; Roller, D. Y. (2001). Ambient ozone
exposure is associated with eosinophil activation in healthy children. Clin Exp Allergy, 31: 1213-1219. 019683
Fujinaka, L. E.; Hyde, D. M.; Plopper, C. G; Tyler, W. S.; Enmgworth, D. L.; Lollini, L. O. (1985). Respiratory
bronchiolitis following long-term ozone exposure in bonnet monkeys: a morphometric study. Exp Lung Res, 8:
167-190. 040278
Gauderman, W. J. (2001). Sample size requirements for matched case-control studies of gene-environment interaction. Stat
Med, 21: 35-50. http://dx.doi.org/10.1002/sim.973 625862
Gauderman, W. J. (2002). Sample size requirements for association studies of gene-gene interaction. Am J Epidemiol, 155:
478-484. 626945
Gauderman, W. J.; Avol, E.; Gilliland, R; Vora, H.; Thomas, D.; Berhane, K.; McConnell, R.; Kuenzli, N.; Lurmann, R;
Rappaport, E.; Margolis, H.; Bates, D.; Peters, J. (2004). The effect of air pollution on lung development from 10 to
18 years of age. N Engl J Med, 351: 1057-1067. http://nejm.highwire.org/cgi/content/abstract/351/ll/1057. 056569
Gilboa, S. M.; Mendola, P.; Olshan, A. R; Langlois, P. H.; Savitz, D. A.; Loomis, D.; Herring, A. H.; Fixler, D. E. (2005).
Relation between ambient air quality and selected birth defects, seven county study, Texas, 1997-2000. Am J
Epidemiol, 162: 238-252. 087892
Gilliland, F. D.; Berhane, K.; Rappaport, E. B.; Thomas, D. C.; Avol, E.; Gauderman, W. J.; London, S. J.; Margolis, H. G;
McConnell, R.; Islam, K. T; Peters, J. M. (2001). The effects of ambient air pollution on school absenteeism due to
respiratory illnesses. Epidemiology, 12: 43-54. 013232
Gilliland, F. D.; McConnell, R.; Peters, J.; Gong Jr, H. (1999). A Theoretical Basis for Investigating Ambient Air Pollution
and Children's Respiratory Health. Environ Health Perspect, 107: 403-407. 155792
Gilliland, F. D.; Rappaport, E. B.; Berhane, K.; Islam, T; Dubeau, L.; Gauderman, W. J.; McConnell, R. (2002). Effects of
glutathione S-Transferase PI, Ml, and Tl on acute respiratory illness in school children. Am J Respir Crit Care
Med, 166: 346-351. 090970
Giovannelli, L.; Pitozzi, V; Moretti, S.; Boddi, V; Dolara, P. (2006). Seasonal variations of DNA damage in human
lymphocytes: Correlation with different environmental variables. Mutat Res-Fundam Mol Mech Mutagen, 593:
143-152. http://dx.doi.0rg/10.1016/j.mrfmmm.2005.07.002 199894
Gonzalez-Pina, R.; Escalante-Membrillo, C.; Alfaro-Rodriguez, A.; Gonzalez-Maciel, A. (2008). Prenatal exposure to
ozone disrupts cerebellar monoamine contents in newborn rats. Neurochem Res, 33: 912-918.
http://dx.doi.org/10.1007/sll064-007-9534-3475317
Gouveia, N.; Bremner, S. A.; Novaes, H. M. D. (2004). Association between ambient air pollution and birth weight in Sao
Paulo, Brazil. J Epidemiol Community Health, 58: 11-17. 055613
Guevara-Guzman, R.; Arriaga, V.; Kendrick, K. M.; Bernal, C.; Vega, X.; Mercado-Gomez, O. F.; Rivas-Arancibia, S.
(2009). Estradiol prevents ozone-induced increases in brain lipid peroxidation and impaired social recognition
memory in female rats. Neuroscience, 159: 940-950. http://dx.doi.Org/10.1016/j.neuroscience.2009.01.047 596385
Ha, E.-H.; Hong, Y.-C.; Lee, B.-E.; Woo, B.-H.; Schwartz, J.; Christiani, D. C. (2001). Is air pollution a risk factor for low
birth weight in Seoul? Epidemiology, 12: 643-648. 019390
Ha, E.-H.; Lee, J.-T; Kim, H.; Hong, Y.-C.; Lee (2003). Infant susceptibility of mortality to air pollution in Seoul, South
Korea. Pediatrics, 111: 284-290. 042552
Hack, M.; Fanaroff, A. A. (1999). Outcomes of children of extremely low birth weight and gestational age in the 1990s.
Early Hum Dev, 53: 193-218. http://dx.doi.org/10.1016/80378-3782(98)00052-8 625952
Hajat, S.; Armstrong, B.; Wilkinson, P.; Busby, A.; Dolk, H. (2007). Outdoor air pollution and infant mortality: Analysis of
daily time-series data in 10 English cities. J Epidemiol Community Health, 61: 719-722. 093276
March 2011 7-66 DRAFT - DO NOT CITE OR QUOTE
-------
Hanene, C.; Jihene, L.; Jame, A.; Kamel, H.; Agnes, H. (2007). Association of GST Genes Polymorphisms with Asthma in
Tunisian Children. Mediators Inflamm, 19564: 6. 595428
Hansen, C. A.; Barnett, A. G; Pritchard, G (2008). The effect of ambient air pollution during early pregnancy on fetal
ultrasonic measurements during mid-pregnancy. Environ Health Perspect, 116: 362-369. 190273
Hansen, C.; Luben, T I; Sacks, J. D.; Olshan, A.; Jeffay, S.; Strader, L.; Perreault, S. D. (2010). The effect of ambient air
pollution on sperm quality. Environ Health Perspect, 118: 203-209. http://dx.doi.org/10.1289/ehp.0901022 594438
Hansen, C.; Neller, A.; Williams, G; Simpson, R. (2006). Maternal exposure to low levels of ambient air pollution and
preterm birth in Brisbane, Australia. BJOG, 113: 935-941. 089818
Hansen, C.; Neller, A.; Williams, G; Simpson, R. (2007). Low levels of ambient air pollution during pregnancy and fetal
growth among term neonates in Brisbane, Australia. Environ Res, 103: 383-389. 090703
Harkema, J. R.; Plopper, C. G; Hyde, D. M.; StGeorge, J. A.; Dungworth, D. L. (1987). Effects of an ambient level of
ozone on primate nasal epithelial mucosubstances: quantitative histochemistry. Am J Pathol, 127: 90-96.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMCl899606A 040816
Harkema, J. R.; Plopper, C. G; Hyde, D. M.; StGeorge, J. A.; Wilson, D. W.; Dungworth, D. L. (1987). Response of the
macaque nasal epithelium to ambient levels of ozone: a morphologic and morphometric study of the transitional
and respiratory epithelium. Am J Pathol, 128: 29-44. 041496
Harkema, J. R.; Plopper, C. G; Hyde, D. M.; StGeorge, J. A.; Wilson, D. W.; Dungworth, D. L. (1993). Response of
macaque bronchiolar epithelium to ambient concentrations of ozone. Am J Pathol, 143: 857-866. 039794
Haro, R.; Paz, C. (1993). Effects of ozone exposure during pregnancy on ontogeny of sleep in rats. Neurosci Lett, 164: 67-
70. http://dx.doi.org/10.1016/0304-3940(93)90859-1 044194
Hassett, C.; Mustafa, M. G; Coulson, W. F.; Elashoff, R. M. (1985). Murine lung carcinogenesis following exposure to
ambient ozone concentrations. J Natl Cancer Inst, 75: Ill-Ill'. 040704
Hatch, G. E.; Harkema, J. R.; Plopper, C. G; Harris, L. (1994). Ozone dosimetry studies in the nose and eye using oxygen-
18. Inhal Toxicol, 6(suppl): 125-134. 076120
Himes, B. E.; Hunninghake, G. M.; Baurley, J. W.; Rafaels, N. M.; Sleiman, P.; Strachan, D. P.; Wilk, J. B.; Willis-Owen, S.
A. G; Klanderman, B.; Lasky-Su, J.; Lazarus, R.; Murphy, A. J.; Soto-Quiros, M. E.; Avila, L.; Beaty, T; Mathias,
R. A.; Ruczinski, L; Barnes, K. C.; Celedon, J. C.; Cookson, W. O. C.; Gauderman, W. J.; Gilliland, F. D.;
Hakonarson, H.; Lange, C.; Moffatt, M. F.; O'Connor, G. T; Raby, B. A.; Silverman, E. K.; Weiss, S. T. (2009).
Genome-wide Association Analysis Identifies PDE4D as an Asthma-Susceptibility Gene. Am J Hum Genet, 84:
581-593. http://dx.doi.0rg/10.1016/j.ajhg.2009.04.006 480112
Hollingsworth, J. W.; Free, M. E.; Li, Z.; Andrews, L. N.; Nakano, H.; Cook, D. N. (2010). Ozone activates pulmonary
dendritic cells and promotes allergic sensitization through a Toll-like receptor 4-dependent mechanism. J Allergy
Clinlmmunol, 125: 1167-1170. 635786
Horak, F. Jr; Studnicka, M.; Gartner, C.; Spengler, J. D.; Tauber, E.; Urbanek, R.; Veiter, A.; Frischer, T. (2002). Particulate
matter and lung function growth in children: a 3-yr follow-up study in Austrian schoolchildren. Eur Respir J, 19:
838-845. 034792
Huen, K.; Gunn, L.; Duramad, P.; Jeng, M.; Scalf, R.; Holland, N. (2006). Application of a geographic information system
to explore associations between air pollution and micronucleus frequencies in African American children and
adults. Environ Mol Mutagen, 47: 236-246. http://dx.doi.org/10.1002/em.20193 089035
Hutcheon, J. A.; Platt, R. W. (2008). The missing data problem in birth weight percentiles and thresholds for "small-for-
gestational-age". Am J Epidemiol, 167: 786-792. 193795
Hwang, B. F.; Jaakkola, J. J. (2008). Ozone and other air pollutants and the risk of oral clefts. Environ Health Perspect,
116: 1411-1415. 193794
Hwang, B.-F.; Jaakkola, J. J. K.; Lee, Y.-L.; Lin, Y.-C.; Guo Y-LL (2006). Relation between air pollution and allergic
rhinitis in Taiwanese schoolchildren. Respir Res, 7: 23. 088971
Hwang, B.-F.; Lee, Y.-L.; Lin, Y.-C.; Jaakkola, J. J. K.; Guo, Y L. (2005). Traffic related air pollution as a determinant of
asthma among Taiwanese school children. Thorax, 60: 467-473. 089454
March 2011 7-67 DRAFT - DO NOT CITE OR QUOTE
-------
Hyde, D. M.; Miller, L. A.; McDonald, R. I; Stovall, M. Y; Wong, V; Pinkerton, K. E.; Wegner, C. D.; Rothlein, R.;
Plopper, C. G. (1999). Neutrophils enhance clearance of necrotic epithelial cells in ozone-induced lung injury in
rhesus monkeys. Am J Physiol, 277: L1190-L1198. 015124
Hyde, D. M.; Plopper, C. G; Harkema, J. R.; St. George, J. A.; Tyler, W. S.; Dungworth, D. L. (1989). Ozone-induced
structural changes in monkey respiratory system. In T. Schneider; S. D. Lee; G. J. R. Wolters; L. D. Grant (Eds.),
Atmospheric Ozone Research and its Policy Implications; Proceedings of the 3rd US-Dutch International
Symposium, Nijmegen, the Netherlands. May 9-13, 1988 (pp. 523-532). Amsterdam, the Netherlands: Elsevier
Science Publishers. 094057
Ihorst, G; Frischer, T.; Horak, F.; Schumacher, M.; Kopp, M.; Forster, J.; Mattes, J.; Kuehr, J. (2004). Long- and medium-
term ozone effects on lung growth including a broad spectrum of exposure. Eur Respir J, 23: 292-299. 055608
Islam, T.; Berhane, K.; McConnell, R.; Gauderman, W. L; Avol, E.; Peters, J. M.; Gilliland, F. D. (2009). Glutathione-S-
transferase (GST) PI, GSTM1, exercise, ozone and asthma incidence in school children. Thorax, 64: 197-202.
http://dx.doi.org/10.1136/thx.2008.099366 196715
Islam, T; Gauderman, W. J.; Berhane, K.; McConnell, R.; Avol, E.; Peters, J. M.; Gilliland, F. D. (2007). The relationship
between air pollution, lung function and asthma in adolescents. Thorax, 62: 957-963. 090697
Islam, T; McConnell, R.; Gauderman, W. J.; Avol, E.; Peters, J. M.; Gilliland, F. D. (2008). Ozone, oxidant defense genes
and risk of asthma during adolescence. Am J Respir Crit Care Med, 177: 388-395.
http://dx.doi.org/10.1164/rccm.200706-863OC097348
Jakab, G. J.; Bassett, D. J. P. (1990). Influenza virus infection, ozone exposure, and fibrogenesis. Am J Respir Crit Care
Med, 141: 1307-1315. 042196
Jalaludin, B.; Mannes, T; Morgan, G; Lincoln, D.; Sheppeard, V.; Corbett, S. (2007). Impact of ambient air pollution on
gestational age is modified by season in Sydney, Australia. Environ Health, 6: 16. 156601
Jedlinska-Krakowska, M.; Bomba, G; Jakubowski, K.; Rotkiewicz, T; Jana, B.; Penkowskii, A. (2006). Impact of
oxidative stress and supplementation with vitamins E and C on testes morphology in rats. J Reprod Dev, 52: 203-
209. 195640
Jerrett, M.; Burnett, R. T; Pope, C. A.; Ito, K.; Thurston, G; Krewski, D.; Shi, Y; Calle, E.; Thun, M. (2009). Long-term
ozone exposure and mortality. N Engl J Med, 360: 1085-1095. http://dx.doi.org/10.1056/NEJMoa0803894 194160
Jiang L- L; Zhang Y- H; Song G- X; Chen G- H; Chen B- H; Zhao N- Q; Kan H- D (2007). A time series analysis of
outdoor air pollution and preterm birth in Shanghai, China. Biomed Environ Sci, 20: 426-431. 093029
Joad, J. P.; Brie, J. M.; Weir, A. J.; Putney, L.; Hyde, D. M.; Postlewait, E. M.; Plopper, C. G. (2000). Effect of respiratory
pattern on ozone injury to the airways of isolated rat lungs. Toxicol Appl Pharmacol, 169: 26-32. 012984
Joad, J. P.; Kott, K. S.; Brie, J. M.; Peake, J. L.; Plopper, C. G; Schelegle, E. S.; Gershwin, L. J.; Pinkerton, K. E. (2006).
Structural and functional localization of airway effects from episodic exposure of infant monkeys to allergen and/or
ozone. Toxicol Appl Pharmacol, 214: 237-243. http://dx.doi.Org/10.1016/j.taap.2005.12.012 596390
Joad, J. P.; Kott, K. S.; Brie, J. M.; Schelegle, E. S.; Gershwin, L. J.; Plopper, C. G; Peake, J. L.; Pinkerton, K. E. (2008).
The effects of inhaled corticosteroids on intrinsic responsiveness and histology of airways from infant monkeys
exposed to house dust mite allergen and ozone. Toxicol Appl Pharmacol, 226: 153-160.
http://dx.doi.0rg/10.1016/i.taap.2007.09.005596391
Johnston, C.; Holm, B.; Gelein, R.; Finkelstein, J. (2006). Postnatal lung development: Immediate-early gene responses
post ozone and LPS exposure. Inhal Toxicol, 18: 875-883. http://dx.doi.org/10.1080/08958370600822466 097439
Kajekar, R.; Pieczarka, E. M.; Smiley-Jewell, S. M.; Schelegle, E. S.; Fanucchi, M. V; Plopper, C. G. (2007). Early
postnatal exposure to allergen and ozone leads to hyperinnervation of the pulmonary epithelium. Respir Physiol
Neurobiol, 155: 55-63. http://dx.doi.Org/10.1016/j.resp.2006.03.002 567661
Karr, C.; Lumley, T; Schreuder, A.; Davis, R.; Larson, T; Ritz, B.; Kaufman, J. (2007). Effects of subchronic and chronic
exposure to ambient air pollutants on infant bronchiolitis. Am J Epidemiol, 165: 553-560. 090719
Kavlock, R. J.; Meyer, E.; Grabowski, C. T. (1980). Studies on the developmental toxicity of ozone: postnatal effects.
Toxicol Lett, 5: 3-9. 094043
Kavlock, R.; Daston, G; Grabowski, C. T. (1979). Studies on the developmental toxicity of ozone I Prenatal effects.
Toxicol Appl Pharmacol, 48: 19-28. http://dx.doi.org/10.1016/S0041-008X(79)80004-6 039228
March 2011 7-68 DRAFT - DO NOT CITE OR QUOTE
-------
Kim, M. Y; Cho, M. Y. (2009). Toxicity and carcinogenicity of ozone in combination with 4-(N-methyl-N-nitrosamino)-l-
(3-pyridyl)-l-butanone and dibutyl phthalate in B6C3F1 mice for 16 and 32 weeks. Biomed Environ Sci, 22: 216-
222. 200775
Kim, M. Y; Cho, M. Y. (2009). Tumorigenesis in B6C3F1 mice exposed to ozone in combination with 4-(N-methyl-N-
nitrosamino)-l-(3-pyridyl)-l-butanone and dietary dibutyl phthalate. Toxicol Ind Health, 25: 189-195.
http://dx.doi.org/10.1177/0748233709106185200773
Kinney, P. L.; Lippmann, M. (2000). Respiratory effects of seasonal exposures to ozone and particles. Arch Environ Occup
Health, 55: 210-216. 011913
Kleeberger, S. R.; Reddy, S. P.; Zhang, L.-Y; Cho, H.-Y; Jedlicka, A. E. (2001). Toll-like receptor 4 mediates ozone-
induced murine lung hyperpermeability via inducible nitric oxide synthase. Am J Physiol, 280: L326-L333. 016163
Kodavanti, U. P.; Thomas, R.; Ledbetter, A. D.; Schladweiler, M. C.; Shannahan, J. H.; Wallenborn, J. G; Lund, A. K.;
Campen, M. J.; Butler, E. O.; Gottipolu, R. R.; Nyska, A.; Richards, J. E.; Andrews, D.; Jaskot, R. H.; McKee, J.;
Kotha, S. R.; Patel, R. B.; Parianandi, N. L. (In Press). Vascular and cardiac impairments in rats Inhaling ozone and
diesel exhaust particles. Environ Health Perspect. http://dx.doi.org/10.1289/ehp.1002386 666323
Krewski, D.; Jerrett, M.; Burnett, R. T; Ma, R.; Hughes, E.; Shi, Y; Turner, M. C.; Pope AC 3rd; Thurston, G; Calle, E. E.;
Thun, M. J. (2009). Extended follow-up and spatial analysis of the American Cancer Society study linking
particulate air pollution and mortality (Report No. Report Nr. 140). Cambridge, MA: Health Effects Institute.
191193
Kuo, H. W.; Lai, J. S.; Lee, M. C.; Tai, R. C.; Lee, M. C. (2002). Respiratory effects of air pollutants among asthmatics in
central Taiwan. Arch Environ Occup Health, 57: 194-200. 036310
Lampl, M.; Jeanty, P. (2003). Timing is everything: A reconsideration of fetal growth velocity patterns identifies the
importance of individual and sex differences. Am J Hum Biol, 15: 667-680. http://dx.doi.org/10.1002/ajhb.10204
625252
Larson, S. D.; Schelegle, E. S.; Walby, W. R; Gershwin, L. J.; Fanuccihi, M. V; Evans, M. J.; Joad, J. P.; Tarkington, B. K.;
Hyde, D. M.; Plopper, C. G. (2004). Postnatal remodeling of the neural components of the epithelial-me senchymal
trophic unit in the proximal airways of infant rhesus monkeys exposed to ozone and allergen. Toxicol Appl
Pharmacol, 194: 211-220. 057062
Last, J. A.; Warren, D. L.; Pecquet-Goad, E.; Witschi, H. (1987). Modification by ozone of lung tumor development in
mice. JNatl Cancer Inst, 78: 149-154. 040830
Latzin, P.; Roosli, M.; Huss, A.; Kuehni, C. E.; Frey, U. (2009). Air pollution during pregnancy and lung function in
newborns: A birth cohort study. Eur Respir J, 33: 594-603. 195721
Lee, C.; Watt, K. C.; Chang, A. M.; Plopper, C. G; Buckpitt, A. R.; Pinkerton, K. E. (1998). Site-selective differences in
cytochrome P450 isoform activities: comparison of expression in rat and rhesus monkey lung and induction in rats.
Drug Metab Dispos, 26: 396-400. 054473
Lee, S. J.; Hajat, S.; Steer, P. J.; Filippi, V. (2008). Atime-series analysis of any short-term effects of meteorological and air
pollution factors on preterm births in London, UK. Environ Res, 106: 185-194. 195720
Lee, Y. L.; McConnell, R.; Berhane, K.; Gilliland, F. D. (2009). Ambient ozone modifies the effect of tumor necrosis factor
G-308A on bronchitic symptoms among children with asthma. Allergy, 64: 1342-1348.
http://dx.doi.0rg/10.llll/i.1398-9995.2009.02014.xl99915
Lee, Y.-L.; Lin, Y.-C.; Lee, Y.-C.; Wang, J.-Y; Hsiue, T.-R.; Guo, Y. L. (2004). Glutathione S-transferase PI gene
polymorphism and air pollution as interactive risk factors for childhood asthma. Clin Exp Allergy, 34: 1707-1713.
090971
Legro, R. S.; Sauer, M. V; Mottla, G. L.; Richter, K. S.; Li, X.; Dodson, W. C.; Liao, D. (2010). Effect of air quality on
assisted human reproduction [in press]. Hum Reprod, TBD: TBD. http://dx.doi.org/10.1093/humrep/deq021 597377
Li, H.; Romieu, L; Sienra-Monge, J. J.; Ramirez-Aguilar, M.; Estela del Rio-Navarro, B.; Kistner, E. O.; Gjessing, H.
K.;Irma del Carmen, L. S.; Chiu, G. Y; London, S. J. (2006). Genetic polymorphisms in arginase I and II and
childhood asthma and atopy. J Allergy Clin Immunol, 117: 119-126. 596447
March 2011 7-69 DRAFT - DO NOT CITE OR QUOTE
-------
Li, Y. R; Gauderman, W. I; Conti, D. V; Lin, P. C.; Avol, E.; Gilliland, F. D. (2008). Glutathione S-Transferase PI,
Maternal Smoking, and Asthma in Children: A Haplotype-Based Analysis. Environ Health Perspect, 116: 409-415.
596449
Lin, C. A.; Pereira, L. A. A.; Nishioka, D. C.; Conceicao, G M. S.; Graga, A. L. R; Saldiva, P. H. N. (2004). Air pollution
and neonatal deaths in Sao Paulo, Brazil. Braz J Med Biol Res, 37: 765-770. 095787
Lin, C.-M.; Li, C.-Y; Yang G.-Y; Mao, I.-R (2004). Association between maternal exposure to elevated ambient sulfur
dioxide during pregnancy and term low birth weight. Environ Res, 96: 41-50. 089503
Lin, S.; Fitzgerald, E.; Hwang, S. A.; Munsie, J. P.; Stark, A. (1999). Asthma hospitalization rates and socioeconomic status
in New York State (1987-1993). JAsthma, 36: 239-251.
http://informahealthcare.com/doi/abs/10.3109/02770909909075407. 377917
Lin, S.; Liu, X.; Le, L. H.; Hwang, S. A. (2008). Chronic exposure to ambient ozone and asthma hospital admissions
among children. Environ Health Perspect, 116: 1725-1730. http://dx.doi.org/10.1289/ehp.11184 196680
Linn, W. S.; Rappaport, E. B.; Berhane, K. T.; Bastain, T. M.; Avol, E. L.; Gilliland, F. D. (2009). Exhaled nitric oxide in a
population-based study of southernCalifornia schoolchildren. Respir Res, 10: 28. 597363
Lipfert, F. W.; Baty, J. D.; Miller, J. P.; Wyzga, R. E. (2006). PM2.5 constituents and related air quality variables as
predictors of survival in a cohort of U.S. military veterans. Inhal Toxicol, 18: 645-657. 088756
Lipfert, F. W.; Perry, H. M. Jr; Miller, J. P.; Baty, J. D.; Wyzga, R. E.; Carmody, S. E. (2000). The Washington University-
EPRI veterans' cohort mortality study: Preliminary results. Inhal Toxicol, 4: 41-73. 004087
Lipfert, F. W.; Perry, H. M. Jr; Miller, J. P.; Baty, J. D.; Wyzga, R. E.; Carmody, S. E. (2003). Air pollution, blood pressure,
and their long-term associations with mortality. Inhal Toxicol, 15: 493-512. 052250
Lipfert, F. W.; Wyzga, R. E.; Baty, J. D.; Miller, J. P. (2006). Traffic density as a surrogate measure of environmental
exposures in studies of air pollution health effects: Long-term mortality in a cohort of US veterans. Atmos Environ,
40: 154-169. 088218
Liu, S.; Krewski, D.; Shi, Y; Chen, Y; Burnett, R. (2007). Association between maternal exposure to ambient air pollutants
during pregnancy and fetal growth restriction. J Expo Sci Environ Epidemiol, 17: 426-432. 090429
Loomis, D.; Castillejos, M.; Gold, D. R.; McDonnell, W.; Borja-Aburto, V. H. (1999). Air pollution and infant mortality in
Mexico City. Epidemiology, 10: 118-123. 087288
Lopez, L; Sanchez, L; Bizarro, P.; Acevedo, S.; Ustarroz, M.; Fortoul, T. (2008). Ultrastructural alterations during
embryonic rats' lung development caused by ozone. J Electron Microsc (Tokyo), 57: 19-23.
http://dx.doi.org/10.1093/jmicro/dfm033 197786
Maisonet, M.; Correa, A.; Misra, D.; Jaakkola, J. J. (2004). A review of the literature on the effects of ambient air pollution
on fetal growth. Environ Res, 95: 106-115. 156725
Mannes, T; Jalaludin, B.; Morgan, G; Lincoln, D.; Sheppeard, V.; Corbett, S. (2005). Impact of ambient air pollution on
birth weight in Sydney, Australia. Occup Environ Med, 62: 524-530. 087895
Mariassy, A. T; Abraham, W. M.; Phipps, R. J.; Sielczak, M. W.; Wanner, A. (1990). Effect of ozone on the postnatal
development of lamb mucociliary apparatus. J Appl Physiol, 68: 2504-2510. 042311
Mariassy, A. T; Sielczak, M. W.; McCray, M. N.; Abraham, W. M.; Wanner, A. (1989). Effects of ozone on lamb tracheal
mucosa: Quantitative glycoconjugate histochemistry. Am J Pathol, 135: 871-879. 042246
Marshall, E.; Harris, G; Wartenberg, D. (2010). Oral cleft defects and maternal exposure to ambient air pollutants in New
Jersey. Birth Defects Res A Clin Mol Teratol, 88: 205-215. http://dx.doi.org/10.1002/bdra.20650 597374
Martinez, F. D.; Wright, A. L.; Taussig, L. M.; Holberg, C. J.; Halonen, M.; Morgan, W. J.; Group Health Medical
Associates (1995). Asthma and wheezing in the first six years of life. N Engl J Med, 332: 133-138. 046150
McConnell, R.; Berhane, K.; Gilliland, R; London, S. J.; Islam, T; Gauderman, W. J.; Avol, E.; Margolis, H. G; Peters, J.
M. (2002). Asthma in exercising children exposed to ozone: A cohort study. Lancet, 359: 386-391. 023150
McConnell, R.; Islam, T; Shankardass, K.; Jerrett, M.; Lurmann, F.; Gilliland, F.; Gauderman, J.; Avol, E.; Kuenzli, N.;
Yao, L.; Peters, J.; Berhane, K. (2010). Childhood incident asthma and traffic-related air pollution at home and
school. Environ Health Perspect, TBD: TBD. http://dx.doi.org/10.1289/ehp.0901232 625501
March 2011 7-70 DRAFT - DO NOT CITE OR QUOTE
-------
Meng, Y. Y; Rull, R. P.; Wilhelm, M.; Lombard!, C.; Balmes, I; Ritz, B. (2010). Outdoor air pollution and uncontrolled
asthma in the San Joaquin Valley, California. J Epidemiol Community Health, 64: 142-147.
http://dx.doi.org/10.1136/iech.2008.083576594252
Meng, Y. Y; Wilhelm, M.; Rull, R. P.; English, P.; Ritz, B. (2007). Traffic and outdoor air pollution levels near residences
and poorly controlled asthma in adults. Ann Allergy Asthma Immunol, 98: 455-463. 093275
Miller, L. A.; Gerriets, J. E.; Tyler, N. K.; Abel, K.; Schelegle, E. S.; Plopper, C. G; Hyde, D. M. (2009). Ozone and
allergen exposure during postnatal development alters the frequency and airway distribution of CD25+ cells in
infant rhesus monkeys. Toxicol Appl Pharmacol, 236: 39-48. http://dx.doi.Org/10.1016/j.taap.2008.12.031 596406
Moffatt, R. K.; Hyde, D. M.; Plopper, C. G; Tyler, W. S.; Putney, L. F. (1987). Ozone-induced adaptive and reactive
cellular changes in respiratory bronchioles of Bonnet monkeys. Exp Lung Res, 12: 57-74. 040841
Monchaux, G; Morlier, J. P.; Morin, M.; Rochefort, P.; Maximilien, R.; Tredaniel, J. (1996). Co-carcinogenic effects in rats
of combined exposure to radon and ozone. Environ Int, 221: S909-S915. 086197
Moore, K.; Neugebauer, R.; Lurmann, F.; Hall, J.; Brajer, V.; Alcorn, S.; Tager, I. (2008). Ambient ozone concentrations
cause increased hospitalizations for asthma in children: An 18-year study in Southern California. Environ Health
Perspect, 116: 1063-1070. http://dx.doi.org/10.1289/ehp.10497196685
Morello-Frosch, R.; Jesdale, B. M.; Sadd, J. L.; Pastor, M. (2010). Ambient air pollution exposure and full-term birth
weight in California. Environ Health, 9: 44. http://dx.doi.org/10.1186/1476-069X-9-44 670076
Morris, C. R.; Poljakovic, M.; Lavrisha, L.; Machado, L.; Kuypers, F. A.; Morris, S. M. Jr (2004). Decreased arginine
bioavailability and increased serum arginase activity in asthma. Am J Respir Crit Care Med, 170: 148-153.
http://dx.doi.org/10.1164/rccm.200309-1304OC674145
Mortimer, K.; Neugebauer, R.; Lurmann, F.; Alcorn, S.; Balmes, J.; Tager, I. (2008). Air pollution and pulmonary function
in asthmatic children: Effects of prenatal and lifetime exposures. Epidemiology, 19: 550-557.
http://dx.doi.org/10.1097/EDE.Ob013e31816a9dcbl22163
Mortimer, K.; Neugebauer, R.; Lurmann, F.; Alcorn, S.; Balmes, J.; Tager, I. (2008). Early-Lifetime exposure to air
pollution and allergic sensitization in children with asthma. J Asthma, 45: 874-881.
http://dx.doi.org/10.1080/02770900802195722 187280
NTP (1994). Toxicology and carcinogenesis: studies of ozone (CAS No 10028-15-6) and ozone/NNK (CAS No 10028-15-
6/64091-91-4) in F344/N rats and B6C3F1 mice (Report No. Technical Report No. 440). Research Triangle Park,
NC: National Toxicology Program. http://ntp.niehs.nih.gov/index.cfm?objectid=070AOEBD-081E-B501-
E38F640803C3542C.011143
Oryszczyn, M. P.; Bouzigon, E.; Maccario, J.; Siroux, V; Nadif, R.; Wright, A.; Kauffmann, F. (2007). Interrelationships of
quantitative asthma-related phenotypes in the Epidemiological Study on the Genetics and Environment of Asthma,
Bronchial Hyperresponsiveness, and Atopy. J Allergy Clin Immunol, 119: 57-63. 596460
Palli, D.; Sera, F.; Giovannelli, L.; Masala, G; Grechi, D.; Bendinelli, B.; Caini, S.; Dolara, P.; Saieva, C. (2009).
Environmental ozone exposure and oxidative DNA damage in adult residents of Florence, Italy. Environ Pollut,
157: 152171525. http://dx.doi.0rg/10.1016/j.envpol.2008.09.Oil 196688
Parker, J. D.; Akinbami, L. J.; Woodruff, T J. (2009). Air pollution and childhood respiratory allergies in the United States.
Environ Health Perspect, 117: 140-147. http://dx.doi.org/10.1289/ehp.11497 192359
Paz, C.; Bazan-Perkins, B. (1992). Sleep-wake disorganization in cats exposed to ozone. Neurosci Lett, 140: 270-272.
036436
Paz, C.; Huitron-Resendiz, S. (1996). The effects of ozone exposure on the sleep-wake cycle and serotonin contents in the
pons of the rat. Neurosci Lett, 204: 49-52. 082684
Peluso M Hainaut, P.; Airoldi, L.; Autrup, H.; Dunning, A.; Garte, S.; Gormally, E.; Malaveille, C.; Matullo, G; Munniaa,
A.; Riboli, E.; Vineis P EPIC investigators (2005). Methodology of laboratory measurements in prospective studies
on gene-environment interactions: The experience of GenAir. DNA Repair, 574: 92-104. 089057
Penard-Morand, C.; Charpin, D.; Raherison, C.; Kopferschmitt, C.; Caillaud, D.; Lavaud, F.; Annesi-Maesano, I. (2005).
Long-term exposure to background air pollution related to respiratory and allergic health in schoolchildren. Clin
Exp Allergy, 35: 1279-1287. 087951
March 2011 7-71 DRAFT - DO NOT CITE OR QUOTE
-------
Pereira, F. A. C.; De Assuncao, J. V; Saldiva, P. H. N.; Pereira, L. A. A.; Mirra, A. P.; Braga, A. L. F. (2005). Influence of
air pollution on the incidence of respiratory tract neoplasm. J Air Waste Manag Assoc, 55: 83-87.
http://www.ncbi.nlm.nih.gov/pubmed/=l5704542. 073851
Pereira, L. A. A.; Loomis, D.; Conceicao, G. M. S.; Braga, A. L. F.; Areas, R. M.; Kishi, H. S.; Singer, J. M.; Bohm, G. M.;
Saldiva, P. H. N. (1998). Association between air pollution and intrauterine mortality in Sao Paulo, Brazil. Environ
Health Perspect, 106: 325-329. 007264
Perepu, R. S.; Garcia, C.; Dostal, D.; Sethi, R. (2010). Enhanced death signaling in ozone-exposed ischemic-reperfused
hearts. Mol Cell Biochem, 336: 55-64. http://dx.doi.org/10.1007/sll010-009-0265-4 385020
Peters, J. M.; Avol, E.; Gauderman, W. J.; Linn, W. S.; Navidi, W.; London, S. J.; Margolis, H.; Rappaport, E.; Vora, H.;
Gong, H. Jr; Thomas, D. C. (1999). A study of twelve southern California communities with differing levels and
types of air pollution II Effects on pulmonary function. Am J Respir Crit Care Med, 159: 768-775. 087237
Peters, J. M.; Avol, E.; Navidi, W.; London, S. J.; Gauderman, W. J.; Lurmann, F.; Linn, W. S.; Margolis, H.; Rappaport, E.;
Gong, H. Jr; Thomas, D. C. (1999). A study of twelve southern California communities with differing levels and
types of air pollution I Prevalence of respiratory morbidity. Am J Respir Crit Care Med, 159: 760-767. 087243
Petruzzi, S.; De Acetis, L.; Chiarotti, F.; Sorace, A.; Alleva, E. (1999). Limited changes in handedness and morphine
reactivity in CD-I mice after pre- and postnatal ozone exposure. Acta Neurobiol Exp (Wars), 59: 115-122. 012066
Petruzzi, S.; Fiore, M.; Dell'Omo, G; Bignami, G; Alleva, E. (1995). Medium and long-term behavioral effects in mice of
extended gestational exposure to ozone. Neurotoxicol Teratol, 17: 463-470. 077448
Plopper, C. G; Harkema, J. R.; Last, J. A.; Pinkerton, K. E.; Tyler, W. S.; StGeorge, J. A.; Wong, V. J.; Nishio, S. J.; Weir,
A. S.; Dungworth, D. L.; Barry, B. E.; Hyde, D. M. (1991). The respiratory system of nonhuman primates responds
more to ambient concentrations of ozone than does that of rats. In Tropospheric Ozone and the Environment (pp.
137-150). Pittsburgh, PA: Air and Waste Management Association. 042617
Plopper, C. G; Hatch, G. E.; Wong, V; Duan, X.; Weir, A. J.; Tarkington, B. K.; Devlin, R. B.; Becker, S.; Buckpitt, A. R.
(1998). Relationship of inhaled ozone concentration to acute tracheobronchial epithelial injury, site-specific ozone
dose and glutathione depletion in rhesus monkeys. Am J Respir Cell Mol Biol, 19: 387-399. 087203
Plopper, C. G; Schelegle, E. S. (1997). Report of on-going and future research on ozone at the Davis campus of the
University of California. Washington, DC: U.S. Environmental Protection Agency. 656713
Plopper, C. G; Smiley-Jewell, S. M.; Miller, L. A.; Fanucchi, M. V; Evans, M. J.; Buckpitt, A. R.; Avdalovic, M.;
Gershwin, L. J.; Joad, J. P.; Kajekar, R.; Larson, S.; Pinkerton, K. E.; Van Winkle, L. S.; Schelegle, E. S.; Pieczarka,
E. M.; Wu, R.; Hyde, D. M. (2007). Asthma/allergic airways disease: Does postnatal exposure to environmental
toxicants promote airway pathobiology? Toxicol Pathol, 35: 97-110. http://dx.doi.org/10.1080/01926230601132030
596412
Pope CA 3rd; Burnett, R. T; Thun, M. J.; Calle, E. E.; Krewski, D.; Ito, K.; Thurston, G. D. (2002). Lung cancer,
cardiopulmonary mortality, and long-term exposure to fine particulate air pollution. JAMA, 287: 1132-1141.
024689
Qian, Z.; Liao, D.; Lin, H.-M.; Whitsel, E. A.; Rose, K. M.; Duan, Y. (2005). Lung function and long-term exposure to air
pollutants in middle-aged American adults. Arch Environ Occup Health, 60: 156-163. 093283
Rage, E.; Jacquemin, B.; Nadif, R.; Oryszczyn, M. P.; Siroux, V.; Aguilera, L; Kauffmann, F.; Kunzli, N. (2009). Total
serum IgE levels are associated with ambient ozone concentration in asthmatic adults. Allergy, 64: 40-46. 196719
Rage, E.; Siroux, V.; Kunzli, N.; Pin, I.; Kauffmann, F. (2009). Air pollution and asthma severity in adults. Occup Environ
Med, 66: 182-188. http://dx.doi.org/10.1136/oem.2007.038349 196720
Ramadour, M.; Burel, C.; Lanteaume, A.; Vervloet, D.; Charpin, D.; Brisse, F.; Dutau, H.; Charpin, D. (2000). Prevalence
of asthma and rhinitis in relation to long-term exposure to gaseous air pollutants. Allergy, 55: 1163-1169. 013259
Renzetti, G; Silvestre, G; DAmario, C.; Bottini, E.; Gloria-Bottini, F.; Bottini, N.; Auais, A.; Perez, M. K.; Piedimonte, G.
(2009). Less air pollution leads to rapid reduction of airway inflammation and improved airway function in
asthmatic children. Pediatrics, 123: 1051-1058. http://dx.doi.org/10.1542/peds.2008-1153 199834
Ritz, B.; Wilhelm, M. (2008). Ambient air pollution and adverse birth outcomes: Methodologic issues in an emerging field.
Basic Appl Ecol, 102: 182-190. 156914
March 2011 7-72 DRAFT - DO NOT CITE OR QUOTE
-------
Ritz, B.; Wilhelm, M.; Hoggatt, K. I; Ghosh, J. K. (2007). Ambient air pollution and preterm birth in the environment and
pregnancy outcomes study at the University of California, Los Angeles. Am J Epidemiol, 166: 1045-1052. 096146
Ritz, B.; Wilhelm, M.; Zhao, Y. (2006). Air pollution and infant death in southern California, 1989-2000. Pediatrics, 118:
493-502. 089819
Ritz, B.; Yu, F. (1999). The effect of ambient carbon monoxide on low birth weight among children born in southern
California between 1989 and 1993. Environ Health Perspect, 107: 17-25. 086976
Ritz, B.; Yu, F.; Chapa, G; Fruin, S. (2000). Effect of air pollution on preterm birth among children born in Southern
California between 1989 and 1993. Epidemiology, 11: 502-511. 012068
Ritz, B.; Yu, F.; Fruin, S.; Chapa, G; Shaw, G. M.; Harris, J. A. (2002). Ambient air pollution and risk of birth defects in
Southern California. Am J Epidemiol, 155: 17-25. 023227
Rivas-Arancibia, S.; Guevara-Guzman, R.; Lopez-Vidal, Y; Rodriguez-Martinez, E.; Gomes, M. Z.; Angoa-Perez, M.;
Raisman-Vozari, R. (2010). Oxidative stress caused by ozone exposure induces loss of brain repair in the
hippocampus of adult rats. Toxicol Sci, 113: 187-197. http://dx.doi.org/10.1093/toxsci/kfp252 201544
Rivas-Manzano, P.; Paz, C. (1999). Cerebellar morphological alterations in rats induced by prenatal ozone exposure.
Neurosci Lett, 276: 37-40. 012072
Rojas-Martinez, R.; Perez-Padilla, R.; Olaiz-Fernandez, G; Mendoza-Alvarado, L.; Moreno-Macias, H.; Fortoul, T.;
McDonnell, W.; Loomis, D.; Romieu, I. (2007). Lung function growth in children with long-term exposure to air
pollutants in Mexico City. Am J Respir Crit Care Med, 176: 377-384. 091064
Romero, R.; Espinoza, J.; Kusanovic, J. P.; Gotsch, F.; Hassan, S.; Erez, O.; Chaiworapongsa, T.; Mazor, M. (2006). The
preterm parturition syndrome. BJOG, 113: 17-42. http://dx.doi.Org/10.llll/i.1471-0528.2006.01120.x625253
Romero-Velazquez, R. M.; Alfaro-Rodriguez, A.; Gonzalez-Pina, R.; Gonzalez-Maciel, A. (2002). Effect of ozone prenatal
exposure on postnatal development of cerebellum. Proc West Pharmacol Soc, 45: 65-67. 035575
Romieu, I.; Ramirez-Aguilar, M.; Moreno-Macias, H.; Barraza-Villarreal, A.; Miller, P.; Hernandez-Cadena, L.; Carbajal-
Arroyo, L. A.; Hernandez-Avila, M. (2004). Infant mortality and air pollution: Modifying effect by social class. J
Occup Environ Hyg, 46: 1210-1216. 093074
Romieu, I.; Sienra-Monge, J. J.; Ramirez-Aguilar, M.; Moreno-Macias, H.; Reyes-Ruiz, N. L; Estela del Rio-Navarro, B.;
Hernandez-Avila, M.; London, S. J. (2004). Genetic polymorphism of GSTM1 and antioxidant supplementation
influence lung function in relation to ozone exposure in asthmatic children in Mexico City. Thorax, 59: 8-10.
056796
Rubes, J.; Selevan, S. G; Evenson, D. P.; Zudova, D.; Vozdova, M.; Zudova, Z.; Robbins, W. A.; Perreault, S. D. (2005).
Episodic air pollution is associated with increased DNA fragmentation in human sperm without other changes in
semen quality. Hum Reprod, 20: 2776-2783. 078091
Rubio, C.; Paz, C. (2003). Indomethacin reverts sleep disorders produced by ozone exposure in rats. Toxicology, 191: 89-
96. http://dx.doi.org/10.1016/S0300-483X(03)00245-2 053541
Salam, M. T; Islam, T; Gauderman, W. J.; Gilliland, F. D. (2009). Roles of arginase variants, atopy, and ozone in
childhood asthma. J Allergy Clin Immunol, 123: 596-602. http://dx.doi.org/10.1016/jjaci.2008.12.020 596644
Salam, M. T; Millstein, J.; Li, Y.-F.; Lurmann, F. W.; Margolis, H. G; Gilliland, F. D. (2005). Birth outcomes and prenatal
exposure to ozone, carbon monoxide, and particulate matter: Results from the Children's Health Study. Environ
Health Perspect, 113: 1638-1644. 087885
Santucci, D.; Sorace, A.; Francia, N.; Aloe, L.; Alleva, E. (2006). Prolonged prenatal exposure to low-level ozone affects
aggressive behaviour as well as NGF and BDNF levels in the central nervous system of CD-I mice. Behav Brain
Res, 166: 124-130. http://dx.doi.Org/10.1016/j.bbr.2005.07.032 596414
Schatz, M.; Zeiger, R. S.; Hoffman, C. P. (1990). Intrauterine growth is related to gestational pulmonary function in
pregnant asthmatic women. Kaiser-Permanente Asthma and Pregnancy Study Group. Chest, 98: 389-392. 156073
Schelegle, E. S.; Miller, L. A.; Gershwin, L. J.; Fanucchi, M. V; Van Winkle, L. S.; Gerriets, J. E.; Walby, W. F.; Mitchell,
V; Tarkington, B. K.; Wong, V J.; Baker, G. L.; Pantle, L. M.; Joad, J. P.; Pinkerton, K. E.; Wu, R.; Evans, M. J.;
Hyde, D. M.; Plopper, C. G. (2003). Repeated episodes of ozone inhalation amplifies the effects of allergen
sensitization and inhalation on airway immune and structural development in Rhesus monkeys. Toxicol Appl
Pharmacol, 191: 74-85. 053778
March 2011 7-73 DRAFT - DO NOT CITE OR QUOTE
-------
Schmelzer, K. R.; Wheelock, A. M.; Dettmer, K.; Morin, D.; Hammock, B. D. (2006). The role of inflammatory mediators
in the synergistic toxicity of ozone and 1-nitronaphthalene in rat airways. Environ Health Perspect, 114: 1354-1360.
112994
Schopke, R.; Wolfer, D. P.; Lipp, H. P.; Leisinger-Trigona, M. C. (1991). Swimming navigation and structural variations of
the infrapyramidal mossy fibers in the hippocampus of the mouse. Hippocampus, 1: 315-328.
http://dx.doi.org/10.1002/hipo.450010322684254
Selevan, S. G; Borkovec, L.; Slott, V. L.; Zudova, Z.; Rubes, I; Evenson, D. P.; Perreault, S. D. (2000). Semen quality and
reproductive health of young Czech men exposed to seasonal air pollution. Environ Health Perspect, 108: 887-894.
012578
Slama, R.; Darrow, L.; Parker, J.; Woodruff, T. J.; Strickland, M.; Nieuwenhuijsen, M.; Glinianaia, S.; Hoggatt, K. J.;
Kannan, S.; Hurley, F.; Kalinka, J.; Sram, R.; Brauer, M.; Wilhelm, M.; Heinrich, J.; Ritz, B. (2008). Meeting
report: Atmospheric pollution and human reproduction. Environ Health Perspect, 116: 791-798. 156985
Sokol, R. Z.; Kraft, P.; Fowler, I. M.; Mamet, R.; Kim, E.; Berhane, K. T. (2006). Exposure to environmental ozone alters
semen quality. Environ Health Perspect, 114: 360-365. 098539
Son, J. Y; Cho, Y. S.; Lee, J. T. (2008). Effects of air pollution on postneonatal infant mortality among firstborn infants in
Seoul, Korea: Case-crossover and time-series analyses. Arch Environ Occup Health, 63: 108-13. 190323
Sousa, S. I. V.; Alvim-Ferraz, M. C. M.; Martins, F. G; Pereira, M. C. (2009). Ozone exposure and its influence on the
worsening of childhood asthma. Allergy, 64: 1046-1055. http://dx.doi.Org/10.llll/j.1398-9995.2009.01946.x
619956
Sousa, S. I. V; Pereira, M. M. C.; Martins, F. G; Alvim-Ferraz, C. M. (2008). Identification of regions with high ozone
concentrations aiming the impact assessment on childhood asthma. Hum Ecol Risk Assess, 14: 610-622.
http://dx.doi.org/10.1080/10807030802074147619959
Sousa, S. L; Ferraz, C.; Alvim-Ferraz, M. C.; Martins, F. G; Vaz, L. G; Pereira, M. C. (2011). Spirometric tests to assess
the prevalence of childhood asthma at Portuguese rural areas: Influence of exposure to high ozone levels. Environ
Int, 37: 474-478. http://dx.doi.Org/10.1016/j.envint.2010.ll.014 676712
Sram, R. J.; Binkova, B.; Dejmek, J.; Bobak, M. (2005). Ambient air pollution and pregnancy outcomes: a review of the
literature. Environ Health Perspect, 113: 375-382. 087442
Sram, R. J.; Binkova, B.; Rossner, P.; Rubes, J.; Topinka, J.; Dejmek, J. (1999). Adverse reproductive outcomes from
exposure to environmental mutagens. Mutat Res-Fundam Mol Mech Mutagen, 428: 203-215.
http://dx.doi.org/10.1016/81383-5742(99)00048-4 078127
Stedman, J. R.; Kent, A. J. (2008). An analysis of the spatial patterns of human health related surface ozone metrics across
the UK in 1995, 2003 and 2005. Atmos Environ, 42: 1702-1716. 110057
Stokinger, H. E. (1962). Effects of air pollution in animals. In AC Stern (Ed.), Air pollution (pp. 282-334). New York, NY:
Academic Press. 015101
Strickland, M. J.; Klein, M.; Correa, A.; Reller, M. D.; Mahle, W. T.; Riehle-Colarusso, T. J.; Botto, L. D.; Flanders, W. D.;
Mulholland, J. A.; Siffel, C.; Marcus, M.; Tolbert, P. E. (2009). Ambient air pollution and cardiovascular
malformations in Atlanta, Georgia, 1986-2003. Am J Epidemiol, 169: 1004-14. 190324
Tager, I. B.; Balmes, J.; Lurmann, F.; Ngo, L.; Alcorn, S.; Kunzli, N. (2005). Chronic exposure to ambient ozone and lung
function in young adults. Epidemiology, 16: 751-759. http://dx.doi.org/10.1097/01.ede.0000183166.68809.bO
087538
Tamer, L.; Caliko?lu, M.; Ates, N. A.; Yildirim, H.; Ercan, B.; Saritas, E.; Unlii, A.; Atik, U. (2004). Glutathione-S-
transferase gene polymorphisms (GSTT1, GSTM1, GSTP1) as increased risk factors for asthma. Respirology, 9:
493-498. 199914
Tovalin, H.; Valverde, M.; Morandi, M. T; Blanco, S.; Whitehead, L.; Rojas, E. (2006). DNA damage in outdoor workers
occupationally exposed to environmental air pollutants. Occup Environ Med, 63: 230-236. 091322
Tran, M. U.; Weir, A. J.; Fanucchi, M. V; Rodriguez, A. E.; Pantle, L. M.; Smiley-Jewell, S. M.; Van Winkle, L. S.; Evans,
M. J.; Miller, L. A.; Schelegle, E. S.; Gershwin, L. J.; Hyde, D. M.; Plopper, C. G. (2004). Smooth muscle
hypertrophy in distal airways of sensitized infant rhesus monkeys exposed to house dust mite allergen. Clin Exp
Allergy, 34: 1627-1633. http://dx.doi.Org/10.llll/i.1365-2222.2004.02057.x628626
March 2011 7-74 DRAFT - DO NOT CITE OR QUOTE
-------
Tsai, S.-S.; Chen, C.-C.; Hsieh, H.-I; Chang, C.-C.; Yang, C.-Y. (2006). Air pollution and postneonatal mortality in a
tropical city: Kaohsiung, Taiwan. Inhal Toxicol, 18: 185-189. 090709
Tucker, J. D.; Auletta, A.; Cimino, M. C.; Dearfield, K. L.; Jacobson-Kram, D.; Tice, R. R.; Carrano, A. V. (1993). Sister-
chromatid exchange: second report of the GENE-TOX program. DNA Repair, 297: 101-180.
http://dx.doi.org/10.1016/0165-1110(93)90001-4 056406
U.S. EPA (1996). Air quality criteria for ozone and related photochemical oxidants (Report No. EPA/600/P-93/004AF).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development. (NTIS
No. PB94173127) 017831
U.S. EPA (2006). Air quality criteria for ozone and related photochemical oxidants (Report No. EPA/600/R-05/004AF).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid= 149923. 088089
Vancza, E. M.; Galdanes, K.; Gunnison, A.; Hatch, G; Gordon, T (2009). Age, strain, and gender as factors for increased
sensitivity of the mouse lung to inhaled ozone. Toxicol Sci, 107: 535-543. http://dx.doi.org/10.1093/toxsci/kfn253
596419
Van Winkle, L. S.; Baker, G. L.; Chan, J. K.; Schelegle, E. S.; Plopper, C. G. (2010). Airway mast cells in a rhesus model of
childhood allergic airways disease. Toxicol Sci, 116: 313-322. http://dx.doi.org/10.1093/toxsci/kfqll9 670301
Veninga, T. S. (1967). Toxicity of ozone in comparison with ionizing radiation. Strahlentherapie, 134: 469-477. 040746
Wang, T.-N.; Ko, Y.-C.; Chao, Y.-Y; Huang, C.-C.; Lin, R.-S. (1999). Association between indoor and outdoor air pollution
and adolescent asthma from 1995 to 1996 in Taiwan. Environ Res, 81: 239-247. 008105
Wang, X. Y; Hu, W.; Tong S (2009). Long-term exposure to gaseous air pollutants and cardio-respiratory mortality in
Brisbane, Australia. Geospat Health, 3: 257-63. http://www.geospatialhealth.unina.it/fulltext.php?ida=72. 199990
Wenten, M.; Gauderman, W. J.; Berhane, K.; Lin, P. C.; Peters, J.; Gilliland, F. D. (2009). Functional variants in the
catalase and myeloperoxidase genes, ambient air pollution, and respiratory-related school absences: An example of
epistasis in gene-environment interactions. Am J Epidemiol, 170: 1494-1501. http://dx.doi.org/10.1093/aje/kwp310
597084
Wilhelm, M.; Ritz, B. (2005). Local variations in CO and particulate air pollution and adverse birth outcomes in Los
Angeles County, California, USA. Environ Health Perspect, 113: 1212-1221. 088668
Wilson, D. W.; Plopper, C. G; Dungworth, D. L. (1984). The response of the macaque tracheobronchial epithelium to acute
ozone injury: a quantitative ultrastructural and autoradiographic study. Am J Pathol, 116: 193-206. 040044
Witschi, H. (1991). Effects of oxygen and ozone on mouse lung tumorigenesis. Exp Lung Res, 17: 473-483. 042509
Witschi, H.; Espiritu, L; Pinkerton, K. E.; Murphy, K.; Maronpot, R. R. (1999). Ozone carcinogenesis revisited. Toxicol
Sci, 52: 162-167. 011602
Witschi, H.; Wilson, D. W.; Plopper, C. G. (1993). Modulation of N-nitrosodiethylamine-induced hamster lung tumors by
ozone. Toxicology, 77: 193-202. 043206
Wollmann, H. A. (1998). Intrauterine growth restriction: definition and etiology. Horm Res, 49: 1-6. 193812
Wood, A. M.; Harrison, R. M.; Semple, S.; Ayres, J. G; Stockley, R. A. (2009). Outdoor air pollution is associated
withdisease severity in al-antitrypsin deficiency. Eur Respir J, 34: 346-353. 597085
Woodruff, T. J.; Darrow, L. A.; Parker, J. D. (2008). Air pollution and postneonatal infant mortality in the United States,
1999-2002. Environ Health Perspect, 116: 110-115. 098386
Wu, R.; Zhao, Y. H.; Plopper, C. G; Chang, M. M.; (1999). Differential expression of stress proteins in nonhuman primate
lung and conducting airway after ozone exposure. Am J Physiol, 277: L511-L522. 012089
Yang, C.-Y; Hsieh, H.-l; Tsai, S.-S.; Wu, T.-N.; Chiu, H.-F. (2006). Correlation between air pollution and postneonatal
mortality in a subtropical city: Taipei, Taiwan. J Toxicol Environ Health A, 69: 2033-2040. 090760
Zelac, R. E.; Cromroy, H. L.; Bolch, W. E. Jr; Dunavant, B. G; Bevis, H. A. (1971). Inhaled ozone as a mutagen: I
chromosome aberrations induced in Chinese hamster lymphocytes. Environ Res, 4: 262-282. 039743
March 2011 7-75 DRAFT - DO NOT CITE OR QUOTE
-------
Zelac, R. E.; Cromroy, H. L.; Bolch, W. E. Jr; Dunavant, B. G; Bevis, H. A. (1971). Inhaled ozone as a mutagen: II effect
on the frequency of chromosome aberrations observed in irradiated Chinese hamsters. Environ Res, 4: 325-342.
039752
March 2011 7-76 DRAFT - DO NOT CITE OR QUOTE
-------
Chapter 8. Populations Susceptible to
Ozone-related Health Effects
1 Interindividual variation in human responses to air pollution exposure suggests that some
2 populations are at increased risk for detrimental effects of ambient exposure to an air pollutant. The
3 NAAQS are intended to provide an adequate margin of safety for both the population as a whole and
4 those individuals potentially at increased risk for health effects in response to ambient air pollution
5 (Section 1.1). To facilitate the identification of populations at greater risk for O3-related health
6 effects, studies have evaluated factors that contribute to the susceptibility and/or vulnerability of an
7 individual to O3. The definition for both of these terms has been found to vary across studies, but in
8 most instances susceptibility refers to biological or intrinsic factors (e.g., lifestage, sex) while
9 vulnerability refers to non-biological or extrinsic factors (e.g., SES) (U.S. EPA, 2009, 179916;
10 U.S. EPA, 2010, 626035). Additionally, in some cases, the terms "at-risk" and sensitive populations
11 have been used to encompass these concepts more generally. However, in many cases, a factor
12 identified that increases an individual's risk for morbidity or mortality effects from exposure to an air
13 pollutant cannot be easily categorized as either a susceptibility or vulnerability factor.
14 As developed in previous ISAs and reviews (Sacks et al., In Press, 664486; U.S. EPA, 2009,
15 179916; U.S. EPA, 2010, 626035). an all encompassing definition for "susceptible population" is
16 used to circumvent the need to distinguish between susceptible and vulnerable, and to identify the
17 populations at greater risk for O3-induced heath effects. This definition identifies susceptible
18 populations as the following:
Individual- and population-level characteristics that increase the risk of O3-related health
effects in a population including, but not limited to: genetic background, birth outcomes (e.g.,
low birth weight, birth defects), race, sex, lifestage, lifestyle (e.g., smoking status, nutrition),
preexisting disease, SES (e.g., educational attainment, reduced access to health care), and
characteristics that may modify exposure to O3 (e.g., time spent outdoors).
19 To examine whether O3 differentially affects certain populations, epidemiologic studies
20 conduct stratified analyses to identify the presence or absence of effect measure modification. A
21 thorough evaluation of potential effect measure modifiers may help identify populations that are
22 more susceptible to O3. Although the design of toxicological and controlled human exposure studies
23 do not allow for the examination of effect measure modifiers, both can provide support and
24 biological plausibility for factors that may lead to increased susceptibility for O3-related health
25 effects through the study of animal models of disease or individuals with underlying disease or
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments such as the Integrated Science Assessments (ISAs) and the Integrated Risk Information System (IRIS).
March 2011 8-1 DRAFT - DO NOT CITE OR QUOTE
-------
1 genetic polymorphisms that allow for comparisons between subgroups. Therefore, the results from
2 these studies, combined with those results obtained through stratified analyses in epidemiologic
3 studies, comprise the overall weight of evidence for the increased susceptibility of specific
4 populations to O3-related health effects.
5 This chapter discusses the epidemiologic, controlled human exposure, and toxicological
6 studies evaluated in Chapters 5, 6, and 7 that provide information on potentially susceptible
7 populations. The studies in this chapter include only those epidemiologic studies that presented
8 stratified results (e.g., males versus females or <65 years of age versus > 65 years of age). This
9 approach allowed for a comparison between populations exposed to similar O3 concentrations and
10 within the same study design. Thus, numerous studies that focus on only one potentially susceptible
11 population are described in previous chapters, but these studies are not discussed in detail in this
12 chapter because of the lack of an adequate comparison group within the study. Included controlled
13 human exposure studies are those that consisted of individuals with an underlying disease or genetic
14 polymorphism, or studies that categorized the study population by age, race, etc. Included
15 toxicological studies were those with animal models of disease.
16 Factors examined for possible susceptibility to O3-related health effects based on the overall
17 evidence integrated across disciplines are described in greater detail in the following sections.
8.1. Pre-existing Disease/Conditions
18 Individuals with pre-existing disease are likely to constitute a susceptible population. Recent
19 studies that examined whether pre-existing diseases and conditions lead to increased susceptibility to
20 O3 were identified and are summarized below. Table 8-1 displays the prevalence rates of these
21 conditions by age and region among adults in the U.S. population. Substantial proportions of the
22 U.S. population are affected by these conditions and therefore may represent a potentially large
23 susceptible population.
March 2011 8-2 DRAFT - DO NOT CITE OR QUOTE
-------
Table 8-1. Prevalence of respiratory diseases, cardiovascular diseases, and diabetes by age and
region among individuals 18 years and older in the U.S.
N (in thousands)
Chronic Disease/Condition
Age
18-44
45-64
65-74
75+
Northeast
Region
Midwest
South
West
Respiratory Diseases
Asthma 28,260
13.5
12.0
12.0
10.0
12.8
13.4
11.2
13.9
COPD
Chronic Bronchitis 9,832
Emphysema 3,789
3.2
0.2
5.5
2.0
5.9
5.7
5.3
5.0
3.4
1.2
4.8
1.9
5.2
1.9
2.9
1.3
Cardiovascular Diseases
All Heart Disease 26,628
Coronary Heart Disease 14,428
Hypertension 56,159
Diabetes 18,651
4.6
1.1
8.7
2.3
12.3
6.7
32.5
12.1
26.7
16.9
54.4
20.4
39.2
26.7
61.1
17.3
11.3
5.7
22.9
4.5
12.7
6.5
24.1
7.6
12.2
7.3
27.1
9.0
9.9
4.9
20.6
7.7
Source: Pleisetal. (2009,
8.1.1. Influenza/Infections
1 Recent studies have indicated that underlying infections may increase susceptibility of
2 individuals to O3-related health effects, although there are only a limited number of studies. A study
3 of hospitalizations in Hong Kong reported that increased levels of influenza intensity resulted in
4 increased excess risk of respiratory disease hospitalizations related to O3 exposure (Wong et al.,
5 2009, 196722). In addition, a study of lung function in asthmatic children reported decreases in lung
6 function with increased short-term O3 exposure for those with upper respiratory infections but not
7 those without infections (Lewis et al., 2005, 081079).
8.1.2. Asthma/Corticosteroid Use
8 Previous O3 AQCDs (U.S. EPA, 1996, 080828: U.S. EPA, 2006, 088089) identified asthmatic
9 individuals as a population susceptible to O3-related health effects, and approximately 12% of adults
10 have reported ever having asthma (Pleis et al., 2009, 629608).
11 Multiple epidemiologic studies included in this ISA have evaluated the potential for the
12 susceptibility of asthmatics to O3-related health effects. No evidence for increased susceptibility was
13 found in a study that examined the effect of short-term O3 exposure on respiratory health (Barraza-
14 Villarreal et al., 2008, 156254). A positive association was reported for airway inflammation among
15 asthmatic children, but the observed association was similar in magnitude to that of non-asthmatics.
16 However, several studies have indicated some evidence for increased susceptibility of asthmatics
17 related to O3 exposure. A study of lifeguards in Texas reported lung function decreases with short-
18 term O3 exposure among both asthmatics and non-asthmatics, however, the decrease was greater
19 among asthmatics (Thaller et al., 2008, 195869). A Mexican study of children ages 6-14 detected an
March 2011
8-3
DRAFT - DO NOT CITE OR QUOTE
-------
1 association between short-term O3 and wheeze, cough, and bronchodilator use among asthmatics but
2 not non-asthmatics although this may have been the result of a small non-asthmatic population
3 (Escamilla-Nunez et al., 2008, 594284). A study of the modification of the effect of greater O3
4 associated decreases in short-term O3 exposure on lung function by airway hyperresponsiveness
5 (AHR) (a condition common among asthmatics) reported greater O3-associated decreases in lung
6 function in elderly individuals with airway hyperresponsiveness, especially among those who were
7 obese (Alexeeff et al., 2007, 195862). Finally, a study of O3 exposure and airway inflammation
8 (FENO) among elderly individuals stratified the study population by asthmatics and non-asthmatics
9 (Delfino et al., 2010, 647222). The regression coefficient was higher among asthmatics but the 95%
10 CIs were wide due to the small number of asthmatics in the study (n=4), resulting in the inability to
11 clearly identify a difference between the populations. Finally, some studies have reported null results
12 for both asthmatics and nonasthmatics. Khatri et al. (2009, 594282) found no association between
13 short-term O3 exposure and lung function for either asthmatic or non-asthmatic adults, but did note a
14 decrease in lung function among individuals with allergies5)594282Khatri et al., 2009,
15 (5)594282Khatri et al., 2009, (. One study compared individuals with asthma to individuals with
16 other diseases/conditions (chronic obstructive pulmonary disease [COPD] or ischemic heart disease
17 [IHD]) (Lagorio et al., 2006, 089800). No association was observed between O3 exposure and
18 decrease in lung function among the asthmatic group or the other groups.
19 Additional evidence for difference in effects among asthmatics has been observed in studies
20 that examined the association between O3 exposure and lung function by asthma medication use. A
21 study of asthmatic children living in Detroit reported a greater association between short-term O3 and
22 lung function for corticosteroid users compared with non-corticosteroid users (Lewis et al., 2005,
23 081079). Conversely, another study found decreased lung function among non-corticosteroid users
24 compared to users, although in this study, a large proportion of non-users were considered to be
25 persistent asthmatics (Hernandez-Cadena et al., 2009, 594283). Lung function was not related to
26 short-term O3 exposure for corticosteroid users and non-users in a study taking place during the
27 winter months in Canada (Liu et al., 2009, 192003). Additionally, a study of airway inflammation
28 reported a counterintuitive inverse association with O3 of similar magnitude for all groups of
29 corticosteroid users and non-users (Qian et al., 2009, 548793).
30 Controlled human exposures studies that have examined the effects of O3 on both asthmatic
31 and healthy controls are limited. Based on studies reviewed in the 1996 and 2006 O3 AQCD
32 (U.S. EPA, 1996, 017831: U.S. EPA, 2006, 088089). asthmatic subjects appear to be more sensitive
33 to acute effects of O3 in terms of FEVi and inflammatory responses than healthy nonasthmatic
34 subjects. For instance, Horstman et al. (1995, 075834) observed mild-to-moderate asthmatics to, on
35 average, experiences double the O3-induced FEVi decrement of healthy subjects (19% versus 10%,
36 respectively, p = 0.04). Moreover, a statistically significant positive correlation between FEVi
37 responses to O3 and baseline lung function was observed in the asthmatic individuals, i.e., responses
38 increased with severity of disease. Only study reported a tendency for asthmatics to have smaller O3-
39 induced FEVi decrements than healthy subjects (3% versus 8%, respectively) (Mudway et al., 2001,
March 2011 8-4 DRAFT - DO NOT CITE OR QUOTE
-------
1 025327). However, the asthmatics in that study also tended to be older than the healthy subjects,
2 which could partially explain their lesser response since FEVi responses to O3 diminish with age.
3 Asthmatics also show significantly more neutrophils in the BALF (18 hours postexposure) than
4 similarly exposed healthy individuals (Basha et al, 1994, 075950; Peden et al, 1997, 085842;
5 Scannell et al., 1996, 080755).
6 Toxicological studies provide support for greater effects of O3 among those with asthma or
7 airway hyperresponsiveness. In animal toxicological studies, an asthmatic phenotype is modeled by
8 allergic sensitization of the respiratory tract. Many of the studies that provide evidence that O3 is an
9 inducer of airway hyperresponsiveness and remodeling utilize these types of animal models. For
10 example, a series of experiments in infant rhesus monkeys show these effects, but only in monkeys
11 sensitized to house dust mite allergen (Fanucchi et al., 2006, 096491; Joad et al., 2006, 596390;
12 Schelegle et al., 2003, 053778). Similarly, Funabashi et al. (2004, 596384) demonstrated adverse
13 changes in pulmonary function in mice exposed to O3, and Wagner et al. (2007, 596420)
14 demonstrated enhanced inflammatory responses in rats exposed to O3, but only in animals sensitized
15 to allergen. In general, it is the combined effects of O3 and allergic sensitization which result in
16 measurable effects on pulmonary function. In a bleomycin induced pulmonary fibrosis model,
17 exposure to 250 ppb O3 for 5 days increased pulmonary inflammation and fibrosis, along with the
18 frequency of bronchopneumonia in rats. Thus short-term exposure to O3 may enhance damage in a
19 previously injured lung (Oyarzun et al., 2005, 596407).
20 In the 2006 O3 AQCD, the potential for asthmatics to have greater susceptibility to O3-related
21 effects was supported by a number of controlled human exposure studies, evidence from
22 toxicological studies, and a limited number of epidemiologic studies. Overall, in the recent
23 epidemiologic literature some, but not all, studies report greater effects among individuals with
24 asthma. Studies examining effect measure modification of the relationship between short-term O3
25 and lung function by corticosteroid use had mixed results. Inconsistent findings of epidemiologic
26 studies may be due to the differences in O3 concentration across the studies. In addition, recent
27 studies of behavioral responses have found studies that do not take into account individuals
28 behavioral adaptation to forecasted air pollution levels (such as avoidance and reduced time
29 outdoors) may be biased towards the null (Neidell and Kinney, 2010, 384492). Evidence from
30 controlled human exposure studies support greater FEVi and inflammatory responses to O3 in
31 asthmatics than in healthy individuals without a history of asthma. Information from recent
32 toxicological studies adds to the evidence for heightened susceptibility of asthmatics to effects of O3
33 exposure.
8.1.3. Chronic obstructive pulmonary disease (COPD)
34 Although not extensively examined in the literature, initial evidence suggests that pre-existing
35 COPD may modify of the association between short-term O3 exposure and cardiovascular-related
36 health effects. In the U.S. over 4% of adults report having chronic bronchitis and almost 2% report
37 having emphysema, both of which are classified as COPD.
March 2011 8-5 DRAFT - DO NOT CITE OR QUOTE
-------
1 In a recent study, Peel et al. (2007, 090442) found that individuals with COPD were more
2 susceptible to effects of short-term O3 exposure on cardiovascular ED visits compared to healthy
3 individuals in Atlanta, GA. They reported that short-term O3 exposure was associated with higher
4 odds of an ED visit for peripheral and cerebrovascular disease among individuals with COPD
5 compared to individuals without COPD. However, pre-existing COPD did not increase the odds of
6 hospitalization for all CVD outcomes (i.e. IHD, dysrhythmia, or congestive heart failure). In an
7 additional study performed in Taiwan, both individuals with and without COPD had higher odds of
8 congestive heart failure associated with O3 exposure on warm days (Lee et al., 2008, 192076).
9 In a study of elderly individuals, although O3 was associated with greater increases in airway
10 inflammation (FENO) among elderly individuals with than without COPD, the number of individuals
11 with COPD was small (n=5) and the 95% CI for the association in COPD subjects was wide
12 (Delfino et al., 2010, 647222). An additional study also found no association between O3 exposure
13 and lung function regardless of whether the study participant had COPD or other health issues
14 (asthma or IHD) (Lagorio et al., 2006, 089800).
15 Recent epidemiologic evidence does not indicate that COPD modifies the association between
16 O3 exposure and respiratory effects, but COPD may affect O3-related cardiovascular effects.
8.1.4. Cardiovascular Disease
17 Cardiovascular disease (CVD) has become increasingly prevalent in the U.S., with about 12%
18 of adults reporting being diagnosed with heart disease. Additionally, a high prevalence of other
19 cardiovascular-related conditions has also been observed, such as hypertension, which is prevalent
20 among approximately 24% of adults. In the 2006 AQCD, little evidence was available regarding pre-
21 existing CVD as a susceptibility factor. Recent epidemiologic studies have examined cardiovascular-
22 related diseases as modifiers of the O3-outcome associations; however, no recent evidence is
23 available from controlled human studies or toxicological studies.
24 Peel et al. (2007, 090442) compared the associations between short-term O3 exposure and
25 cardiovascular ED visits in Atlanta, GA among multiple comorbid conditions. The authors found no
26 evidence of increased risk of cardiovascular ED visits in individuals previously diagnosed with
27 dysrhythmia, congestive heart failure, or hypertension compared to healthy individuals. In Taiwan, a
28 positive association was observed for O3 on warm days and congestive heart failure hospital
29 admissions (HAs), but the association did not differ between individuals with/without hypertension
30 or with/without dysrhythmia (Lee et al., 2008, 192076). Another study in Taiwan reported that the
31 association between O3 levels and ED visits for arrhythmias were greater on warm days among those
32 with congestive heart failure compared to those without congestive heart failure; however, the
33 estimate and 95% CI for those without congestive heart failure is completely contained within the
34 95% CI of those with congestive heart failure (Chiu and Yang, 2009, 603236).
35 Among individuals with a history of CVD there was a greater association between O3
36 exposure and certain, but not all, blood inflammatory markers. Liao et al. (2005, 088677) found that
37 fibrinogen was positively associated with short-term O3 exposure but this association was present
March 2011 8-6 DRAFT - DO NOT CITE OR QUOTE
-------
1 only among individuals with a history of CVD. Those without CVD showed a null association.
2 However, for another biomarker (vWF), CVD status did not modify the positive association with
3 short-term O3 exposureO)088677Liao et al, 2005, (0)088677Liao et al, 2005, (.
4 Some modification by pre-existing atrial fibrillation and atherosclerosis were noted in
5 mortality studies. A study of 48 U.S. cities reported increased risk of mortality with short-term O3
6 exposure among only individuals with secondary atrial fibrillation (Medina-Ramon and Schwartz,
7 2008, 193829). No association was observed for short-term O3 exposure and mortality in a study of
8 diabetics with or without CVD prior to death; however, there was some evidence of increased effects
9 during the warm season if individuals had diabetes and atherosclerosis compared to having only
10 diabetes (Goldberg et al., 2006, 088641).
11 Finally, a study of O3 exposure and lung function was performed among participants with
12 IHD, asthma, or COPD (Lagorio et al., 2006, 089800). No association was observed regardless of
13 whether the participant had IHD.
14 Overall, most studies do not report increased O3-related health effects for individuals with
15 CVD with the possible exception of O3 exposure and mortality. Future research among those with
16 CVD compared to those without will increase the understanding of potential susceptibility of
17 O3-related health effects among this group.
8.1.5. Diabetes
18 Recent literature has not extensively examined whether individuals with diabetes (about 8% of
19 U.S. adults) are potentially susceptible to O3-related health effects. In a study of short-term O3
20 exposure and cardiovascular ED visits in Atlanta, GA, no association was seen for individuals with
21 or without diabetes (Peel et al., 2007, 090442). A similar study conducted in Taiwan reported a
22 positive association between O3 exposure on warm days and HAs for congestive heart failure but
23 again no modification by diabetes was observed (Lee et al., 2008, 192076). Finally, in a study of ED
24 visits for arrhythmia in Taiwan, there was no effect measure modification by diabetes on warm or
25 cool days (Chiu and Yang, 2009, 603236).
8.2. Lifestage
26 The 2006 AQCD (U.S. EPA, 2006, 088089) identifies children, especially those with asthma,
27 and older adults as susceptible populations. New evidence, summarized below, further supports these
28 findings.
8.2.1. Children
29 The 2000 Census reports that 28.6% of the U.S. population is under 20 years of age, with
30 14.1% under the age of 10 (Social Science Data Analysis Network; CensusScope and CensusScope,
31 2010, 647298). Children are considered to be more susceptible to O3-related health effects compared
32 to adults because they spend more time outside and are more highly active, especially during the
March 2011 8-7 DRAFT - DO NOT CITE OR QUOTE
-------
1 summer when O3 concentrations are the highest (U.S. EPA, 2006, 088089). Moreover, children's
2 respiratory systems are undergoing development until about age 18-20 and are therefore thought to
3 be more sensitive to O3-induced damage (U.S. EPA, 2006, 088089).
4 Multiple studies have been performed examining different age groups and their susceptibility
5 to O3-related respiratory HAs and emergency department (ED) visits. A study in Cyprus of short-
6 term O3 concentrations and respiratory HA detected possible effect measure modification by age
7 with a larger association among individuals less than 15 years of age compared with those over
8 15 years of age. However, this difference was only apparent with a 2-day lag (Middleton et al., 2008,
9 156760). Similarly, a Canadian study of asthma-ED visits reported a positive association among 5-
10 to 14-year olds but no association in any of the other age groups (ages examined 0-75+) (Villeneuve
11 et al., 2007, 195859). A study in Finland reported a greater O3-associated change in asthma-related
12 ED visits among children (<15 year) as compared to adults (15-64 years) (Halonen et al., 2009,
13 625764). A study of New York City HAs demonstrated an increase in the association between O3
14 exposure and asthma-related hospitalizations for 6- to 18-year olds compared to those less than 6 and
15 those older than 18 year of age (Silverman and Ito, 2010, 386252). A study of long-term O3 exposure
16 and asthma HA among children reported larger associations among children 1- to 2-year old
17 compared to children 2- to 6-year old (Lin et al., 2008, 196680). A few studies reported positive
18 associations among both children and adults and no modification of the effect by age. A study
19 performed in Hong Kong examined O3 and asthma-related HAs for ages 0-14, 15-65, and >65 (Ko et
20 al., 2007, 092844). The researchers reported that the association was greater among the 0-14 and
21 14-65 age groups compared to the >65 age group. Another study looking at asthma-related ED visits
22 in Maine reported positive associations for all age groups (ages 2-65) (Paulu and Smith, 2008,
23 180168). A study performed in Washington found effects of O3 on asthma hospitalizations among
24 both children and adults (<18 an218 years old) but reported that only children had statistically
25 significant results at lag day 0, which the authors write, "suggests that children are more immediately
26 responsive to adverse effects of O3 exposure." (Mar and Koenig, 2009, 594410). Additionally, a
27 study examining asthma physician visits reported consistently negative effects in all age groups
28 (1-17 and 18-64) (Burra et al., 2009, 195868).
29 The 1996 O3 AQCD, reported clinical evidence that children, adolescents, and young adults
30 (<18 years of age) appear, on average, to have nearly equivalent spirometric responses to O3, but
31 have greater responses than middle-aged and older adults when exposed to comparable O3 doses
32 (U.S. EPA, 1996, 017831). Symptomatic responses (e.g., cough, shortness of breath, pain on deep
33 inspiration) to O3 exposure, however, appear to increase with age until early adulthood and then
34 gradually decrease with increasing age (U.S. EPA, 1996, 017831). For subjects aged 18-36 years,
35 McDonnell et al. (1999, 010939) reported that symptom responses from O3 exposure also decrease
36 with increasing age. Lung growth and development is not achieved until 18-20 years of age in
37 females and the early 20s for males; pulmonary function is at its maximum during this time as well.
38 Additionally, PBPK modeling reported regional extraction of O3 to be higher in infants compared to
39 adults. This is thought to be due to the smaller nasal and pulmonary regions surface area in children
March 2011 8-8 DRAFT - DO NOT CITE OR QUOTE
-------
1 under the age of 5 compared to the total airway surface area observed in adults (Sarangapani et al,
2 2003, 054581).
3 Recent toxicological studies support previous findings of greater susceptibility in immature
4 animals. Early life exposures of multiple species of laboratory animals, including infant monkeys,
5 resulted in changes in conducting airways at the cellular, functional, ultra-structural, and
6 morphological levels as is detailed below. Carey et al. (2007, 195752) conducted a study of O3
7 exposure in infant rhesus macaques, whose nasal airways closely resemble that of humans. Monkeys
8 were exposed either acutely for 5 days to 0.5 ppm O3, or episodically for 5 biweekly cycles
9 alternating 5 days of 0.5 ppm O3 with 9 days of filtered air, designed to mimic human exposure
10 (70 days total). All monkeys acutely exposed to O3 had moderate to marked necrotizing rhinitis, with
11 focal regions of epitheliar exfoliation, numerous infiltrating neutrophils, and some eosinophils. The
12 distribution, character, and severity of lesions in episodically exposed monkeys were similar to that
13 of acutely exposed animals. Neither group exhibited mucous cell metaplasia proximal to the lesions,
14 a protective adaptation observed in adult monkeys exposed continuously to 0.3 ppm O3 in another
15 study (Harkema et al., 1987, 040816). Functional (increased airway resistance and responsiveness
16 with antigen + O3 co-exposure) and cellular changes in conducting airways (increased numbers of
17 inflammatory eosinophils) also manifested among the infant monkeys (Plopper et al., 2007, 596412).
18 In addition, the lung structure of the conducting airways was significantly stunted or altered versus
19 control animals and this aberrant development was persistent 6 months postexposure (Fanucchi et
20 al., 2006, 096491).
21 Similarly, rat fetuses exposed to O3 in utero had significant ultra-structural changes in
22 bronchiolar epithelium when examined near the end of gestation ((Lopez et al., 2008, 197786). In
23 addition, exposure of mice to mixtures of air pollutants early in development affected pup lung
24 cytokine levels (TNF, IL-1, KC, IL-6, and MCP-1). In utero exposure of animals to PM augmented
25 O3-induced airway hyper-reactivity in these pups as juveniles (Auten et al., 2009, 200760).
26 Age may affect the immune response to O3 exposure. In comparing neonatal mice to adults,
27 increased bronchoalveolar lavage (BAL) neutrophils were observed in four strains of neonates 24 h
28 after exposure to 0.8 ppm O3 for 5 hours (Vancza et al., 2009, 596419). Three of these strains also
29 exhibited increased BAL protein, although the two endpoints were not necessarily consistently
30 correlated in a given strain. In some strains, however, adults were more sensitive, indicating a strain-
31 age interaction. In young mice, healing of skin wounds is not significantly affected by O3 exposure
32 (Lim et al., 2006, 670834). However, exposure to 0.5 ppm O3 for 6 h/day significantly delays wound
33 closure in aged mice.
34 Increased susceptibility found in the younger lifestage may be due to age-related changes in
35 endogenous antioxidants and sensitivity to oxidative stress. A recent study demonstrated that
36 0.25 ppm O3 differentially alters expression of metalloproteinases in the skin of young (8 weeks) and
37 aged (18 months) mice, indicating age-related susceptibility to oxidative stress (Fortino et al., 2007,
38 596382). Valacchi et al. (2007, 596418) found that aged mice had more vitamin E in their plasma but
39 less in their lungs compared to young mice, which may affect their pulmonary antioxidant defenses.
March 2011 8-9 DRAFT - DO NOT CITE OR QUOTE
-------
1 Servais et al. (2005, 195667) found higher levels of oxidative damage indicators in three week-old
2 (immature) and 20 month-old (aged) rats compared to adult rats, which were relatively resistant to an
3 intermittent 7-day exposure to 0.5 ppm O3. Immature rats exhibited a higher ventilation rate, which
4 may have increased exposure. Senescent rats had similar ventilatory rates as adults, but their
5 antioxidant enzyme responses had a different profile from those of adult rats.
6 Additionally, a series of toxicological studies reported an association between O3 exposure and
7 bradycardia that was present among young mice but not among older mice (Hamade and Tankersley,
8 2009, 596386: Hamade et al., 2008, 156515: Hamade et al., 2010, 666324: Tankersley et al., 2010,
9 628062). Regression analysis revealed a significant interaction between age and strain on heart rate,
10 which implies that aging may affect heart rate differently between mouse strains (Tankersley et al.,
11 2010, 628062). The authors propose that the genetic differences between the mice strains could be
12 altering the formation of ROS, which tends to increase with age, thus modulating the changes in
13 cardiopulmonary physiology after O3 exposure.
14 The human clinical and toxicological studies reported evidence of increased susceptibility for
15 younger ages. Studies of respiratory HA and ED visits observed mixed findings for associations
16 among children and young adults, although generally studies reported positive associations among
17 both children and adults or just among children. For other outcomes, inconsistent findings regarding
18 susceptibility to O3-related health effects. The interpretation of these studies is limited by the lack of
19 consistency in comparison age groups and outcomes examined.
8.2.2. Older Adults
20 The gradual decline in physiological processes that occur with aging may lead to increased
21 susceptibility to O3-related health effects (U.S. EPA, 2006, 192082). Diminished symptomatic
22 responses may also put the elderly at increased risk for continued O3 exposure. In addition, older
23 adults have a higher prevalence of pre-existing diseases compared to younger age groups and this
24 may also lead to increased susceptibility to O3-related health effects (see Table 8-1 that gives pre-
25 existing rates by age). Moreover, with the number of older Americans increasing in upcoming years
26 (estimated to increase from 12.4% of the U.S. population to 19.7% between 2000 to 2030, which is
27 approximately 35 million and 71.5 million individuals, respectively) this group represents a large
28 population potentially susceptible to O3-related health effects (Social Science Data Analysis
29 Network; CensusScope and CensusScope, 2010, 647298: U.S. Census Bureau, 2010, 647312).
30 A positive association was reported between O3 levels and respiratory HAs for adults 65 and
31 older but not for those adults aged 15-64 (Halonen et al., 2009, 625764). In the same study, no
32 association was observed between O3 levels and respiratory mortality among those 65 years and
33 older or those 15-64 years; however, an inverse association between O3 levels and cardiovascular
34 mortality was present among those 65 years and older but not among those under 65 years of age.
35 This inverse association among those 65 and older persisted when examining hospitalizations for
36 coronary heart disease. A study of CVD-related hospital visits in Bangkok reported an increase in
37 percent change for hospital visits with previous day and cumulative 2-day O3 levels among those
March 2011 8-10 DRAFT - DO NOT CITE OR QUOTE
-------
1 65 years and older, whereas no association was present for individuals less than 65 years of age. No
2 association was observed for current day or cumulative three-day averages in any age group
3 (Buadong et al, 2009, 602060). A study examining O3 and HAs for CVD-related health effects
4 reported no association for individuals aged 15-64 or individuals aged 65 and older, although one
5 lag-time did show an inverse effect for coronary heart disease among elderly that was not present
6 among 15- to 64-year olds (Halonen et al., 2009, 625764). No modification by age (40-64 versus
7 >64) was observed in a study from Brazil examining O3 levels and COPD ED visits (Arbex et al.,
8 2009, 184334).
9 The majority of studies reported greater effects of short-term O3 exposure and mortality
10 among older adults, which is consistent with the findings of the 2006 O3 AQCD (U.S. EPA, 2006,
11 088089). A study conducted in 48 cities across the U.S. reported larger effects among adults 65 and
12 older compared to those younger than 65 years (Medina-Ramon and Schwartz, 2008, 193829).
13 Further investigation of this study population revealed no association between O3 and mortality until
14 age 50 and a reduced effect after age 80 (Zanobetti and Schwartz, 2008, 195755). A study of 7 urban
15 centers in Chile reported similar results, with greater effects in adults 65 and older, however the
16 effects were smaller among those 85 year of age and older compared to those in the 75-84 years old
17 age range (Cakmak et al., 2007, 091170). A study performed in China reported greater effects in
18 populations 45 years and older (compared to 5-44 year olds), with statistically significant effects
19 present only among those 65 years and older (Kan et al., 2008, 156621). An Italian study reported
20 higher risk of all-cause mortality associated with increased O3 concentrations among individuals
21 85 year and older as compared to those 35-84 years old. Those 65-74 and 75-84 years did not show a
22 greater increase in risk compared to those age 35-64 years (Stafoggia et al., 2010, 625034). The Air
23 Pollution and Health: A European and North American Approach (APHENA) project examined the
24 association between O3 exposure and mortality for those <75 and^ 75 years of age. In Canada, the
25 associations for all-cause and cardiovascular mortality were greater among those 75 years and older
26 in the summer-only and all-year analyses. Age groups were not compared in the analysis for
27 respiratory mortality in Canada. In the U.S., the association for all-cause mortality was slightly
28 greater for those younger than 75 years of age compared to those 75 and older in summer-only
29 analyses. No consistent pattern was observed for CVD mortality. In Europe, slightly larger
30 associations for all-cause mortality were observed in those younger than 75 in all-year and summer-
31 only analyses. Larger associations were reported among those <75 for CVD mortality in all-year
32 analyses, but the reverse was true for summer-only analyses (Katsouyanni et al., 2009, 199899).
33 Biological plausibility for increased susceptibility among older adults is provided by clinical
34 and toxicological studies. Respiratory symptom responses to O3 exposure appears to increase with
35 age until early adulthood and then gradually decrease with increasing age (U.S. EPA, 1996, 017831).
36 The decrease in symptomatic responses with age was observed by McDonnell et al. (1999, 010939)
37 in subjects aged 18 to 36 years. In contrast to young adults, the diminished symptomatic responses in
38 older adults may put them at increased risk for continued O3 exposure. Regarding cardiac outcomes,
39 O3 exposure resulted in an increase in left ventricular chamber dimensions at end diastole (LVEDD)
March 2011 8-11 DRAFT-DO NOT CITE OR QUOTE
-------
1 in young and old mice, whereas decreases in left ventricular posterior wall thickness at end systole
2 (PWTES) were only observed among older mice (Tankersley et al., 2010, 628062).
8.3. Sex
3 The distribution of males and females in the U.S. is similar. In 2000, 49.1% of the U.S.
4 population was male and 50.9% were female. The distribution did vary by age with a greater
5 prevalence of females: 65 years old compared to males (Social Science Data Analysis Network;
6 CensusScope and CensusScope, 2010, 647298). Recent epidemiologic studies have evaluated the
7 effects of short-term and long-term exposure to O3 on multiple health endpoints stratified by sex.
8 A study in Maine on short-term O3 concentrations and asthma ED visits detected greater
9 effects among males ages 2-14 and among females ages 15-34 compared to males and females in the
10 same age groups (no difference was detected for males and females aged 35-64) (Paulu and Smith,
11 2008, 180168). A New York study found no effect measure modification of the association between
12 long-term O3 exposure and asthma HA among males and females between 1 and 6 years old (Lin et
13 al., 2008, 196680). Additionally, a Canadian study reported no associations between short-term O3
14 and respiratory infection HAs for either boys or girls under the age of 15 (Lin et al., 2005, 087828).
15 whereas another Canadian study reported a slightly higher but non-statistically significant increase in
16 respiratory HA for males (mean ages 47.6-69.0 years) (Cakmak et al., 2006, 093272). A recent study
17 from Hong Kong examining individuals of all ages reported no effect measure modification by sex
18 for overall respiratory disease HAs, but did detect a greater excess risk of HAs for COPD among
19 females compared to males (Wong et al., 2009, 196722). Similarly a study in Brazil found higher
20 effect estimates for COPD ED visits among females compared to males (Arbex et al., 2009, 184334).
21 Higher levels of respiratory HA with greater O3 concentrations was also observed for females in a
22 study of individuals living in Cyprus (Middleton et al., 2008, 156760). A study of lung function
23 unrelated to HA and ED visits was conducted among lifeguards in Texas and reported decreased lung
24 function with increased O3 exposure among females but not males (Thaller et al., 2008, 195869).
25 This study included individuals aged 16-27, and the majority of participants were male.
26 In addition to examining the potential modification of O3 associations with respiratory
27 outcomes by sex, studies also examined cardiovascular-related outcomes specifically HAs and ED
28 visits. All of these studies reported no effect modification by sex with some studies reporting null
29 associations for both males and females (Middleton et al., 2008, 156760; Villeneuve et al., 2006,
30 090191; Wong et al., 2009, 196722) and one study reporting a positive associations for both sexes
31 (Cakmak et al., 2006, 099068). A French study examining the associations between O3
32 concentrations and risk of ischemic strokes (not limited to ED visits or HAs) reported no association
33 for either males or females with lags of 0, 2, or 3 days (Henrotin et al., 2007, 093270). A positive
34 association was reported for males with a lag of 1 day, but this association was null for females. The
3 5 authors note that men in the study had much higher rates of current and former smoking than women
36 (67.4% versus 9.3%).
March 2011 8-12 DRAFT - DO NOT CITE OR QUOTE
-------
1 A biomarker study investigating the effects of O3 concentrations on high-sensitivity C-reactive
2 protein (hs-CRP), fibrinogen, and white blood cell count (WBC), reported observations for various
3 lag times ranging from 0 to 7 days (Steinvil et al., 2008, 188893). Most of the associations were null
4 for males and females although one association between O3 and fibrinogen was positive for males
5 and null for females (lag day 4); however, this positive association was null or negative when other
6 pollutants were included in the model. Only one study examining correlations between O3 levels and
7 oxidative DNA damage examined results stratified by sex. In this study Palli et al. (2009, 196688)
8 reported stronger correlations for males than females, both during short-term exposure (less than
9 30 days) and long-term exposure (0-90 days). However, the authors comment that this difference
10 could be partially explained by different distributions of exposure to traffic pollution at work.
11 A few studies have examined the association between short-term O3 concentrations and
12 mortality stratified by sex and in contrast with studies of other endpoints, were more consistent in
13 reporting elevated risks among females. These studies, conducted in the U.S. (Medina-Ramon and
14 Schwartz, 2008, 193829). Italy (Stafoggia et al., 2010, 625034). and Asia (Kan et al., 2008, 156621).
15 reported higher effects in females. In the U.S. study, the elevated risk of mortality among females
16 was greater specifically among those 60 years of age and older (Medina-Ramon and Schwartz, 2008,
17 193829). One long-term O3 exposure study of respiratory mortality stratified their results by sex and
18 reported relative risks of 1.01 (95 % CI: 0.99, 1.04) for males and 1.04 (95% CIs 1.03, 1.07) for
19 females (Jerrett et al., 2009, 194160).
20 Experimental research provides a further understanding of the possible differential
21 susceptibility of males and females to O3 exposure. Several studies have suggested that physiological
22 differences between sexes may predispose females to a greater susceptibility to O3. Lower plasma
23 and nasal lavage fluid (NLF) levels of uric acid (most prevalent antioxidant) in females, the initial
24 defense mechanism of O3 neutralization, may be a contributing factor (Housley et al., 1996, 080811).
25 Consequently, reduced absorption of O3 in the upper airways of females may promote its deeper
26 penetration. Dosimetric measurements have shown that the absorption distribution of O3 is
27 independent of gender when absorption is normalized to anatomical dead space (Bush et al., 1996,
28 080763). Thus, a differential removal of O3 by uric acid seems to be minimal. In general, the
29 physiologic response of young healthy females to O3 exposure appears comparable to the response
30 of young males (Hazucha et al., 2003, 048168). During the follicular phase of the menstrual cycle,
31 lung function response to O3 is enhanced (Fox et al., 1993, 043906). Seal et al. (1996, 044251) later
32 reported no effect of menstrual cycle phase in their analysis of responses of 150 women, but
33 conceded that the methods used by Fox et al. (1993, 043906) more precisely defined the menstrual
34 cycle phase. In a toxicological study, small sex differences were seen in adult mice with respect to
35 pulmonary inflammation and injury after a 5-h exposure to 0.8 ppm O3, and although adult females
36 were generally more susceptible, these differences were strain-dependent, with some strains
37 exhibiting greater susceptibility in males (Vancza et al., 2009, 596419). The most obvious sex
38 difference was apparent in lactating females, which incurred the greatest lung injury or inflammation
39 among several of the strains.
March 2011 8-13 DRAFT - DO NOT CITE OR QUOTE
-------
1 Overall, results have varied, with evidence for increased susceptibility for O3-related health
2 effects present for females in some studies and males in other studies. Most studies examining O3
3 and mortality report females to be more susceptible than males. Little evidence is available regarding
4 a difference between the sexes for other outcomes. Mixed findings are reported on whether effect
5 measure modification exists by sex for respiratory and cardiovascular HA and ED visits, although
6 this inconsistency, at least in part, could be attributable to additional comparisons of different age
7 groups and different respiratory health endpoints by the various studies.
8.4. Genetics
8 Multiple studies that examined the effect of short- and long-term O3 exposure on respiratory
9 function have focused on whether various genes modify the effect of O3 on health. A study
10 examining the relationship between a mother's asthma and her infant's respiratory health illustrated
11 the potential for genetics to play a role in O3-related susceptibility to health effects. A study of
12 wheeze in infants reported larger associations between short-term O3 exposure and wheeze and
13 difficulty breathing in infants whose mothers have asthma compared to infants of mothers without
14 asthma (Triche et al., 2006, 093274).
15 Multiple genes, including glutathione S-transferase Mu 1 (GSTM1) and tumor necrosis
16 factor-a (TNF-a) were evaluated in the 2006 AQCD (U.S. EPA, 2006, 088089) and found to have a
17 "potential role... in the innate susceptibility to O3." Studies performed since the last AQCD have
18 continued to examine the roles of GSTM1 and TNFa on O3-related health effects and have also
19 examined other gene variants that may increase susceptibility to O3-related health effects.
20 Epidemiologic studies that examined the effects of short-term exposure to O3 on lung function
21 included analyses of potential gene-environment interactions. Romieu et al. (2006, 090969) reported
22 an association between O3 and respiratory symptoms that were larger among children with GSTM1
23 null or glutathione S-transferase P 1 (GSTP1) Val/Val genotypes. However, results suggested that O3-
24 associated decreases in lung function may be greater among children with GSTP1 lie/lie or Ile/Val
25 compared to GSTP1 Val/Val. Alexeef et al. (2008, 195864) reported greater decreases in lung
26 function among GSTP1 Val/Val adults than those with other genotypes. In addition, they detected
27 greater decreases for adults with long GT dinucleotide repeats in heme-oxygenase-1 (HMOX1)
28 promoters.
29 Several controlled human exposure studies have reported that genetic polymorphism of
30 antioxidant enzymes may modulate pulmonary function and inflammatory response to O3 challenge.
31 It appears that healthy carriers of NAD(P)H quinone oxidoreductase 1 (NQO1) wild type (wt) in
32 combination with GSTM1 null genotype had greater decreases in lung function parameters with
33 exposure to O3 (Bergamaschi et al., 2001, 052670). Adults with GSTM1 null only genotype did not
34 show the same response to O3. In contrast, asthmatic children with GSTM1 null genotype (Romieu
35 et al., 2004, 056796) were reported to have greater decreases in lung function in relation to O3
36 exposure. In a similar study, Vagaggini et al. (2010, 387127) exposed mild-to-moderate asthmatics to
March 2011 8-14 DRAFT - DO NOT CITE OR QUOTE
-------
1 O3 with moderate exercise. In subjects with NQO1 wt and GSTM1 null, there was no evidence of
2 changes in lung function or inflammatory responses to O3.
3 In a study of healthy volunteers with GSTM1 sufficient (n=19; 24±3) and GSTM1 null (n=16;
4 25 ± 5) exposed to 400 ppb O3 for 2 hours with exercise, Alexis et al. (2009, 628542) found
5 genotype effects on inflammatory responses but not lung function responses to O3. At 4 h post O3
6 exposure, individuals with both GSTM1 genotypes had significant increases in sputum neutrophils
7 with a tendency for a greater increase in GSTM1 sufficient than nulls. At 24 hours postexposure,
8 neutrophils had returned to baseline levels in the GSTM1 sufficient individuals. In the GSTM1 null
9 subjects, neutrophil levels increased from 4 to 24 hours and were significantly greater than both
10 baseline levels and levels at 24 hours in the GSTM1 sufficient individuals. Since there was no FA
11 control in the Alexis et al. (2009, 628542) study, effects of the exposure other than O3 can not be
12 ruled out. In general, the findings between studies are inconsistent and additional, better-controlled
13 studies are needed to clarify the influence of genetic polymorphisms on O3 responsiveness in
14 humans.
15 Several epidemiologic studies of long-term O3 exposure examined interactions with different
16 gene variants, including GSTP1, HMOX1, and TNF-a. A study among children reported a three-way
17 interaction effect between He 105 homozygotes of GSTP1, O3, and playing more than two team
18 sports, and new onset of asthma (Islam et al., 2009, 196715). Additionally, Islam et al. found that
19 non-Hispanic white children with less than 23 repeats in the HMOX-1 gene had decreased risk of
20 new-onset asthma (Islam et al., 2008, 097348). ARG1 and ARG2 (encoded by arginases)
21 modification were examined for the association between genotypes and new-onset asthma (Salam et
22 al., 2009, 596644). Reduced asthma risk was observed among atopic children living in high O3 areas
23 and having the ARG1 haplotypes. There was no difference in risk for children with ARG2
24 haplotypes. A decreased risk of bronchitic symptoms was observed among asthmatic children in low
25 O3 areas with TNF-a variant G-308A (TNF-308GG genotype), a variant that may alter gene
26 expression. There was no decrease in risk for children with this TNF-a variant but living in areas
27 with high O3 concentrations. Additionally, this modification for high and low levels of O3 was not
28 present among non-asthmatic children (Lee et al., 2009, 199915). Wenten et al. (2009, 597084)
29 observed increased risk of respiratory-related school absences among children with variants of
30 catalase (CAT) and myeloperoxidase (MPO) genes, especially when the children were living in high
31 O3 areas.
32 Toxicological studies have reported differences in cardiac and respiratory effects after O3
33 exposure among different mouse strains, which alludes to susceptibility among individuals due to
34 genetic variability (Chuang et al., 2009, 197202: Hamade and Tankersley, 2009, 596386: Hamade et
35 al., 2008, 156515: Tankersley et al., 2010, 628062). Altered O3 responses between two strains could
36 be due to genetic variability in nuclear factor erythroid 2-related factor 2 (Nrf-2), suggesting a role
37 for genetic differences in altering the formation of ROS. Another difference among strains is
38 attributed to differences in O3-induced lung hyperpermeability (Kleeberger et al., 2000, 014895:
39 Kleeberger et al., 2001, 016163). Additionally, some studies have reported O3-related effects to vary
March 2011 8-15 DRAFT - DO NOT CITE OR QUOTE
-------
1 by Inf-1 and Inf-2 genes (Tankersley and Kleeberger, 1994, 021420) and a gene coding for Clara cell
2 secretory protein (CCSP) (Broeckaert et al, 2003, 055490: Wattiez et al, 2003, 043783).
3 Voynow et al. (2009, 194311) have shown that NQO1 deficient mice, like their human
4 counterparts, are resistant to O3-induced airway hyperresponsiveness and inflammation. Reduced
5 production of inflammatory mediators and cells and blunted airway hyperresponsiveness were
6 observed in NQOl-null mice after exposure to 1 ppm O3 for 3 hours. These results correlated with
7 those from in vitro experiments in which human bronchial epithelial cells treated with an NQO1
8 inhibitor exhibited reduced inflammatory responses to exposure to 0.4 ppm O3 for 5 hours.
9 The role of TNF-a signaling in O3-induced responses has been previously established through
10 depletion experiments, but a more recent toxicological study investigated the effects of combined O3
11 and PM exposure in transgenic TNF overexpressing mice. Kumarathasan et al. (2005, 596398) found
12 that subtle effects of these pollutants were difficult to identify in the midst of the severe pathological
13 changes caused by constitutive TNF-a overexpression. However, there was evidence that TNF
14 transgenic mice were more susceptible to O3/PM-induced oxidative stress, and they exhibited
15 elevation of a serum creatine kinase after pollutant exposure, which may suggest potential systemic
16 or cardiac related effects. Differential susceptibility to O3 among inbred strains of animals does not
17 seem to be dose dependent since absorption of 18O in various strains of mice did not correlate with
18 resistance or sensitivity (Vancza et al., 2009, 596419).
19 Defects in DNA repair mechanisms may also confer susceptibility to O3-related health effects.
20 Cockayne syndrome, a rare autosomal recessive disorder in humans, is characterized by UV
21 sensitivity abnormalities, neurological abnormalities, and premature aging. The same genetic defect
22 in mice (Csb~'~) makes them sensitive to oxidative stressors, including O3. Kooter et al. (2007,
23 596397) demonstrated that Csb"7" mice produced significantly more TNF-a after exposure to 0.8 ppm
24 O3 than their wild-type counterparts. However, there were no significant differences in other markers
25 of inflammation or lung injury between the two strains of mice.
8.5. Diet
26 Diet was not examined as a susceptibility factor in previous AQCDs, but recent studies have
27 examined modification of the association between O3 and health effects by dietary factors. Because
28 O3 mediates its toxic effects through oxidative stress, the antioxidant status of an individual is an
29 important factor that may contribute to increased susceptibility to O3-related health effects.
30 Supplementation with vitamin E has been investigated in a number of studies as a means of
31 inhibiting O3-mediated damage.
32 Epidemiologic studies have examined effect measure modification by diet and found evidence
33 that certain dietary components are related to the effect of O3 has on respiratory outcomes. The most
34 recent study examined fruit/vegetable intake and Mediterranean diet (Romieu et al., 2009, 548788).
35 Increases in these food patterns, which have been noted for their high vitamins C and E and omega-3
36 fatty acid content, protected against O3-related decreases in lung function among children living in
March 2011 8-16 DRAFT - DO NOT CITE OR QUOTE
-------
1 Mexico City. Another study examined supplementation of the diets of asthmatic children in Mexico
2 with Vitamins C and E (Sienra-Monge et al., 2004, 196422). Associations were detected between
3 short-term O3 and nasal airway inflammation among children in the placebo group but not in those
4 receiving the supplementation. The authors concluded that "vitamin C and E supplementation above
5 the minimum dietary requirement in asthmatic children with a low intake of vitamin E might provide
6 some protection against the nasal acute inflammatory response to ozone."
7 The epidemiologic evidence is supported by the controlled human exposure studies, which
8 have shown that the first line of defense against oxidative stress is antioxidants-rich extracellular
9 lining fluid (ELF) which scavenge free radicals and limit lipid peroxidation. Exposure to O3 depletes
10 the antioxidant level in nasal ELF probably due to scrubbing of O3 (Mudway et al., 1999, 001270):
11 however, the concentration and the activity of antioxidant enzymes either in ELF or plasma do not
12 appear to be related to O3 responsiveness (Avissar et al., 2000, 012528; Blomberg et al., 1999,
13 001267: Samet et al., 2001, 019034). Carefully controlled studies of dietary antioxidant
14 supplementation have demonstrated some protective effects of alpha-tocopherol (a form of vitamin
15 E) and ascorbate (vitamin C) on spirometric lung function from O3 but not on the intensity of
16 subjective symptoms and inflammatory response including cell recruitment, activation and a release
17 of mediators (Samet et al., 2001, 019034: Trenga et al., 2001, 019845). Dietary antioxidants have
18 also afforded partial protection to asthmatics by attenuating postexposure bronchial
19 hyperresponsiveness (Trenga et al., 2001, 019845).
20 Toxicological studies provide evidence of biological plausibility to the epidemiologic and
21 controlled human exposure studies. Wagner et al. (2007, 596420: 2009, 201574) have shown
22 reductions in O3-exacerbated nasal allergy responses in rats with gamma-tocopherol treatment (a
23 form of vitamin E). Ozone-induced inflammation and mucus production were also inhibited by
24 gamma-tocopherol. Inconsistent results are observed in toxicological studies of ascorbate deficiency
25 and O3 responses. Guinea pigs deficient in ascorbate displayed only minimal injury and
26 inflammation after exposure to O3 (Kodavanti et al., 1995, 077440). A recent study in mice
27 demonstrated a protective effect of beta-carotene in the skin, where it limited the production of
28 proinflammatory markers and indicators of oxidative stress induced by O3 exposure (Valacchi et al.,
29 2009, 201554). In addition to the studies of antioxidants, one toxicological study examined protein
30 deficiency. Protein deficiency alters the levels of enzymes and chemicals in the brain involved with
31 redox status; exposure to 0.75 ppm O3 has been shown to differentially affect Na+/K+ ATPase,
32 glutathione, and lipid peroxidation, depending on the nutritional status of the animal, but the
33 significance of these changes is unclear (Calderon Guzman et al., 2006, 596371).
8.6. Body Mass Index
34 Obesity, defined as a BMI of 30 kg/m2 or greater, is an issue of increasing importance in the
35 U.S., with self-reported rates of 26.7% in 2009, up from 19.8% in 2000 (Sherry et al., 2010,
36 667866). A few studies have been performed examining the association between BMI and lung
March 2011 8-17 DRAFT - DO NOT CITE OR QUOTE
-------
1 function. An epidemiologic study reported decreased lung function with increased short-term O3
2 exposure for both obese and non-obese subjects; however, the magnitude of the reduction in lung
3 function was greater for those subjects who were obese (Alexeeff et al, 2007, 195862). Further
4 decrements in lung function were noted for obese individuals with airway hyperresponsiveness.
5 Controlled human exposure studies have also detected differential effects of O3 on lung function for
6 individuals with varying BMIs. In a retrospective analysis of data from 541 healthy, nonsmoking,
7 white males between the ages of 18-35 years from 15 studies conducted at the U.S. EPA Human
8 Studies Facility in Chapel Hill, North Carolina, McDonnell et al. (2010, 383972) found that
9 increased body mass index (BMI) was found to be associated with enhanced FEVi responses. The
10 BMI effect was of the same order of magnitude but in the opposite direction of the age effect
11 whereby FEVi responses diminish with increasing age. In a similar analysis, Bennett et al. (2007,
12 418827) found enhanced FEVi decrements following O3 exposure with increasing BMI in a group of
13 healthy, nonsmoking, women (BMI range 15.7 to 33.4), but not among healthy, nonsmoking men
14 (BMI range 19.1 to 32.9). In the women, greater O3-induced FEVi decrements were seen in
15 overweight/obese (BMI >25) than in normal weight (BMI from 18.5 to 25), and in normal weight
16 than in underweight (BMI <18.5). Even disregarding the five underweight women, a greater O3
17 response in the overweight/obese category (BMI >25), compared with the normal weight group
18 (BMI from 18.5 to 24.9).
19 Studies in genetically and dietarily obese mice have shown enhanced pulmonary inflammation
20 and injury with acute O3exposure, but responses to longer exposures at more relevant doses appear
21 to differ. A recent study found that obese mice are actually resistant to O3-induced pulmonary injury
22 and inflammation and reduced lung compliance following exposure to 0.3 ppm O3 for 72 hours,
23 regardless of whether obesity was genetic or diet-induced (Shore et al., 2009, 201551).
24 Multiple epidemiologic and human clinical studies have reported increased O3-related
25 respiratory health effects among obese individuals. Future research of the effect modification of the
26 relationship between O3 and other health-related outcomes besides respiratory health effects by BMI
27 will advance understanding of obesity as a potential susceptibility factor.
8.7. Socioeconomic Status
28 SES is often represented by personal or neighborhood SES, educational attainment, health
29 insurance status, and other such factors. Based on the 2000 Census data, 12.4% of Americans live in
30 poverty (poverty threshold for family of four was $17,463) (Social Science Data Analysis Network;
31 CensusScope and CensusScope, 2010, 647297).
32 Multiple epidemiologic studies have reported individuals of low SES to be more susceptible to
33 the effects of short-term O3 exposure on respiratory HAs and ED visits. A study performed in Korea
34 examined the association between O3 concentrations and asthma HA and reported larger effect
35 estimates in areas of moderate and low SES compared with areas of high SES (SES was based on
36 average regional insurance rates) (Lee et al., 2006, 098248). Another Canadian study reported
March 2011 8-18 DRAFT - DO NOT CITE OR QUOTE
-------
1 inverse effects of O3 on respiratory HA and ED visits regardless of SES, measured by average census
2 tract household income (Burra et al., 2009, 195868). In addition, a study conducted across 10 cities
3 in Canada found the largest association between O3 and respiratory HA was among those with an
4 educational level less than grade 9, but no consistent trend in the effect was seen across quartiles of
5 income (Cakmak et al., 2006, 093272). One study performed among children in New York State
6 reported greater associations between long-term O3 exposure and asthma HA among children of
7 mothers who did not graduate from high school, whose births were covered by Medicaid/self-paid,
8 and who were living in poor neighborhoods compared to children whose mothers graduated from
9 high school, whose births were covered by other insurance, and who were not living in poor
10 neighborhoods, respectively (Lin et al., 2008, 196680).
11 One study reported the association between short-term O3 and ED visits for cardiac disease by
12 quartiles of neighborhood-level education and income. No effect measure modification was apparent
13 for either measure of SES (Cakmak et al., 2006, 099068).
14 Several studies were conducted examining the modification by SES of the relationship
15 between short-term O3 concentrations and mortality. A U.S. multicity study reported that
16 communities with a higher proportion of the population unemployed had higher mortality effect
17 estimates (Bell and Dominici, 2008, 193828). A study examining effect measure modification of the
18 association between O3 and mortality by percentage unemployed reported a higher percent change in
19 mortality with increased percent unemployed but this varied across the regions included in the study
20 (U.S., Canada, Europe) (Katsouyanni et al., 2009, 199899). A Chinese study reported that the
21 greatest effects between O3 concentrations and mortality at lag day 0 were among those living in
22 areas of high social deprivation (i.e. low SES), but this association was not consistent across lag days
23 (at other lag times, the middle social deprivation index category had the greatest association) (Wong
24 et al., 2008, 157151). However, another study in Asia comparing low to high educational attainment
25 populations reported no evidence of greater mortality effects (total, CVD, or respiratory) (Kan et al.,
26 2008, 156621). Additionally, a study in Italy reported no difference in risk of mortality among
27 census-block level derived income levels (Stafoggia et al., 2010, 625034). A study of infant mortality
28 in Mexico reported no association between O3 concentrations and infant mortality among any of the
29 three levels of SES determined using a socioeconomic index based on residential areas (Romieu et
30 al., 2004, 093074). Another study in Mexico reported a positive association between O3 levels at lag
31 0 and respiratory-related infant mortality in only the low SES group (determined based on education,
32 income, and household conditions across residential areas), but no association was observed in any
33 of the SES groups with other lags (Carbajal-Arroyo et al., In Press, 667773).
34 Evidence from a controlled human exposure study that examined O3 effects on lung function
35 does not provide support for greater O3-related health effects in individuals of lower SES. In a
36 follow-up study (Seal et al., 1993, 039357) on modification by race, Seal et al. (1996, 044251)
37 reported that, of three SES categories, individuals in the middle SES category showed greater
38 concentration-dependent decline in percent-predicted FEVi (4-5% at 400 ppb O3) than in low and
39 high SES groups. The authors did not have an "immediately clear" explanation for this finding.
March 2011 8-19 DRAFT - DO NOT CITE OR QUOTE
-------
1 Overall, most studies of individuals and those living in neighborhoods with low SES have
2 reported that individuals with low SES are more susceptible to O3-related health effects, resulting in
3 higher odds of respiratory HAs and ED visits. This was not supported by a single controlled human
4 exposure study conducted to examine O3-related effects on lung function for individuals from
5 varying SES groups. Inconsistent results have been observed in the few studies examining effect
6 modification of associations with mortality.
8.8. Air Conditioning Use
7 Air conditioning use is an important component of exposure, as use of central air conditioning
8 will limit exposure to O3 by blocking the penetration of O3 into the indoor environment (further
9 information can be found in Section 4.4). Air conditioning use is a difficult effect measure modifier
10 to examine, as it represents multiple components. More generally, air conditioning prevalence is
11 associated with temperature of a region; those areas with higher temperatures have a greater
12 prevalence of households with air conditioning. Second, it is a marker of SES, with individuals of
13 low SES less likely to have an air conditioner. Finally, air conditioning use is often measured based
14 on area prevalence and may not reflect individual-level use. Despite these limitations, a few studies
15 have examined effect measure modification by prevalence of air conditioning use in an area.
16 Studies examining multiple cities across the U.S. have assessed whether associations between
17 O3 concentrations and HA and mortality varied among areas with high and low prevalence of air
18 conditioning. Medina-Ramon et al. (2006, 087721) conducted a study during the warm season and
19 observed a greater association between O3 levels and pneumonia HAs among areas with a lower
20 proportion of households having central air conditioning compared to areas with a larger proportion
21 of households without air conditioning. The same trend of increased association for areas with a
22 lower prevalence of central air conditioning was noted in a study of O3 concentrations and mortality
23 (Bell and Dominici, 2008, 193828V Conversely, Medina-Ramon and Schwartz (2008, 193829) found
24 that among individuals with atrial fibrillation, a lower risk of mortality was observed for areas with a
25 lower prevalence of central air conditioning.
8.9. Involvement in Outdoor Activities
26 Studies included in the 2006 O3 AQCD reported individuals who participate in outdoor
27 activities or work to be a susceptible population based on consistently reported associations between
28 O3 exposure and respiratory health outcomes in these groups (U.S. EPA, 2006, 088089). Outdoor
29 workers are exposed to ambient O3 concentrations outside for a greater period of time than
30 individuals who spend their days indoors. Additionally, an increase in dose to the lower airways is
31 possible with exercise due to both increases in amount of air breathed (i.e., minute ventilation) and a
32 shift from nasal to oronasal breathing (Hu et al., 1994, 041323; Nodelman and Ultman, 1999,
33 015112; Sawyer et al., 2007, 195142). For further discussion of the association between FEVi
March 2011 8-20 DRAFT - DO NOT CITE OR QUOTE
-------
1 responses to O3 and minute ventilation, refer to Section 6.2.3.1 of the 2006 O3 AQCD (U.S. EPA,
2 2006, 088089).
3 A recent study has explored the potential effect measure modification of O3 exposure and
4 DNA damage by indoor/outdoor workplace (Tovalin et al, 2006, 091322). In a study of indoor and
5 outdoor workers in Mexico, individuals who worked outdoors in Mexico City had a slight
6 association between O3 exposure and DNA damage (measured by comet tail length assay), whereas
7 no association was observed for indoor workers in Mexico City. Workers in another Mexican city,
8 Puebla, demonstrated no association between O3 levels and DNA damage, regardless of whether they
9 worked indoors or outdoors.
10 Although there is no evidence of modification by outdoor activity in this recent study, previous
11 work has shown that increased dose of O3 concentrations from outdoor work leads to increased
12 susceptibility to O3-related health effects among individuals who participate in outdoor activities or
13 work.
8.10. Race/Ethnicity
14 Based on the 2000 Census, 69.1% of the U.S. population comprises Non-Hispanic Whites.
15 Approximately 12.1% of people reported their race/ethnicity as Non-Hispanic Black and 12.6%
16 reported being Hispanic (Social Science Data Analysis Network; CensusScope and CensusScope,
17 2010.647301).
18 A couple of studies examined the associations between short-term O3 concentrations and
19 mortality and reported higher effect estimates among Blacks (Medina-Ramon and Schwartz, 2008,
20 193829) and among communities with larger proportions of Blacks (Bell and Dominici, 2008,
21 193828). Another study examined short-term exposure to O3 concentrations and asthma HAs among
22 children in New York State. These authors reported no statistically significant difference in the odds
23 of asthma HA for Blacks compared to other races but did detect higher odds for Hispanics compared
24 to non-Hispanics (Lin et al., 2008, 196680).
25 Support for the epidemiologic studies is provided by a controlled human exposure study (Seal
26 et al., 1993, 039357). which has compared lung function responses of Whites and Blacks exposed to
27 a range of O3 concentrations. The independent effects of gender-race group and O3 concentration on
28 lung function were positive, but the interaction between gender-race group and O3 concentration was
29 not statistically significant. The findings indicate some overall difference between the gender-race
30 groups that is independent of O3 concentration (the concentration-response curves for the four
31 gender-race groups are parallel). In a multiple comparison procedure on data collapsed across all O3
32 concentrations for each sex-race group, both Black men and Black women had larger decrements in
33 FEVi than did White men. The authors noted that the O3 dose per unit of lung tissue would be
34 greater in blacks and females than whites and males, respectively. That this difference in tissue dose
35 might have affected responses to O3 cannot be ruled out. The college students recruited for the Seal
March 2011 8-21 DRAFT-DO NOT CITE OR QUOTE
-------
1 et al. (1993, 039357) study are probably from better educated and more SES advantaged families,
2 thus reducing potential for these variables to be confounding factors.
3 Overall, the results of recent studies suggest that there may be race-related susceptibility for
4 some outcomes, although the overall understanding of potential effect measure modification by race
5 is limited by the small number of studies. Additionally, these results may be confounded by other
6 factors, such as socioeconomic status.
8.11. Physical Conditioning
7 The 2008 Summary of Health Statistics for U.S. Adults from the CDC reported the prevalence
8 of regular leisure-time physical activity as slightly above 30% for adults 18 years of age and older in
9 the U.S. Forty-nine percent of individuals 65 and older reported no leisure-time physical activity
10 (Pleis et al., 2009, 629608). Physical activity is of interest as a susceptibility factor because studies
11 have demonstrated that exercise affects both the amount of air breathed (flow rate and breathing
12 frequency increase) and type of breathing (switch from nasal to oronasal) (Hu et al., 1994, 041323;
13 Nodelman and Ultman, 1999, 015112: Sawyer et al., 2007, 195142). A study of effect measure
14 modification by exercise habits ten years prior to death observed excess risk of mortality with
15 increasing O3 concentrations among individuals that never exercised compared to individuals that
16 exercised at least once a month for both adults 30 years of age and older and adults 65 years of age
17 and older (Wong et al., 2007, 093278). No recent studies examining modification of O3-related
18 health effects by current physical activity were identified.
8.12. Smoking
19 Previous O3 AQCDs have concluded that smoking does not increase susceptibility to
20 O3-related health effects; in fact, in controlled human exposure studies, smokers have been found to
21 be less susceptible to O3-related health effects than non-smokers. Data from recent interviews
22 conducted as part of the 2008 National Health Interview Survey (NHIS) (Pleis et al., 2009, 629608)
23 have shown the rate of smoking among adults 18 year and older to be approximately 20% in the U.S.
24 Approximately 21% of individuals surveyed were identified as former smokers.
25 Baccarelli et al. (2007, 091310) performed a study of O3 concentrations and plasma
26 homocysteine levels (a risk factor for vascular disease). They found no interaction of smoking
27 (smokers versus non-smokers) for the associations between O3 concentrations and plasma
28 homocysteine levels. Another study examined the association between O3 and resting heart rate and
29 also reported no interaction with smoking status (current smokers versus current non-smokers)
30 (Ruidavets et al., 2005, 089443).
31 A study examining correlations between O3 levels and oxidative DNA damage examined
32 results stratified by current versus never and former smokers (Palli et al., 2009, 196688). Ozone was
33 positively associated with DNA damage for short-term and long-term exposures among never/former
March 2011 8-22 DRAFT - DO NOT CITE OR QUOTE
-------
1 smokers. For current smokers, short-term O3 concentrations were inversely associated with DNA
2 damage; however, the number of current smokers was small (n=12).
3 The findings of Palli et al. (2009, 196688) are consistent with those from controlled human
4 exposure studies that have confirmed that smokers are less responsive to O3 than non-smokers.
5 Spirometric and plethysmographic pulmonary function decline, nonspecific airway hyperreactivity,
6 and inflammatory response of smokers to O3 were all weaker than the ones reported for non-
7 smokers. Similarly, the time course of development and recovery of these effects as well their
8 reproducibility was not different from non-smokers. Chronic airway inflammation with
9 desensitization of bronchial nerve endings and an increased production of mucus may plausibly
10 explain the pseudo-protective effect of smoking (Frampton et al., 1997, 082692; Torres et al., 1997,
11 084265V
12 These findings for smoking are consistent with previous AQCD conclusions. An
13 epidemiologic study of O3-associated DNA damage reported smokers to be less susceptible to O3-
14 related health effects. However, both epidemiologic studies of short-term exposure and CVD
15 outcomes found no effect measure modification by smoking.
8.13. Hyperthyroidism
16 A potential susceptibility factor has been identified in toxicological studies but has not yet
17 been explored in epidemiologic or controlled human exposure studies. Lung damage and
18 inflammation due to oxidative stress may be modulated by thyroid hormones. Compared to controls,
19 hyperthyroid rats exhibited elevated levels of BAL neutrophils and albumin after a 4-h exposure to
20 O3, indicating inflammation and damage. Hyperthyroidism did not affect production of reactive
21 oxygen or nitrogen species, but BAL phospholipids were increased, indicating greater activation of
22 Type II cells and surfactant protein production compared to normal rats (Huffman et al., 2006,
23 596388). Thus, this study provides some underlying evidence which suggests that individuals with
24 hyperthyroidism may represent a susceptible population. Future studies in humans have the potential
25 to identify this as additional susceptibility factors.
8.14. Summary
26 In this section, epidemiologic, controlled human exposure, and toxicological studies have been
27 evaluated that contribute information on potential susceptibility factors. Overall, this review provides
28 evidence that various factors may lead to increased susceptibility to O3-related health effects.
29 The populations identified in this section that are most susceptible to O3-related health effects
30 are individuals with influenza/infection, individuals with asthma, and older age groups. There were a
31 small number of studies on influenza/infection but both reported influenza/infection to modify the
32 association between O3 exposure and respiratory effects, with individuals having influenza or an
33 infection being at increased susceptibility. Asthma as a susceptibility factor was supported by
34 controlled human exposure and toxicological studies, as well as some evidence from epidemiologic
March 2011 8-23 DRAFT - DO NOT CITE OR QUOTE
-------
1 studies. Most studies comparing age groups reported greater effects of short-term O3 exposure on
2 mortality among older adults. Diet and obesity are also both likely susceptibility factors. Multiple
3 epidemiologic, controlled human exposure, and toxicological studies reported that diets deficient in
4 Vitamins E and C are associated with susceptibility to O3-related health effects. Similarly, studies of
5 effect measure modification by BMI observed greater O3-related respiratory decrements for
6 individuals who were obese.
7 Other potential factors [pre-existing conditions (such as COPD and CVD) young age, sex, and
8 multiple genes (such as GSTM1, GSTP1, HMOX-1, NQO1, and TNF-a)] provided some evidence
9 of susceptibility, but further evidence is needed. In addition, examination of modification of the
10 associations between O3 exposure and health effects by SES and race were available in a limited
11 number of studies, and demonstrated possible increased odds of health effects related to O3 exposure
12 among those with low SES and Blacks.
13 Individuals involved in outdoor activities were examined in a recent study but no effect
14 modification was observed. However, previous evidence along with biological plausibility from
15 toxicological and controlled human studies has shown this population to be susceptible to O3-related
16 health effects. The only studies examining effect measure modification by diabetes examined O3
17 exposure and cardiovascular outcomes, but none of the studies reported any change in the
18 association by diabetes.
19 Studies of air conditioning use, physical conditioning, and smoking were conducted but not
20 much evidence was available to determine whether susceptibility to O3-related health effects is
21 present for these factors. Toxicological studies also identified hyperthyroidism and the lifestage of
22 gestation to be factors warranting further examination. Future research on these will provide
23 additional insight into whether these factors affect susceptibility to O3-related health effects.
March 2011 8-24 DRAFT - DO NOT CITE OR QUOTE
-------
References
Alexeeff, S. E.; Litonjua, A. A.; Suh, H.; Sparrow, D.; Vokonas, P. S.; Schwartz, J. (2007). Ozone exposure and lung
function: effect modified by obesity and airways hyperresponsiveness in the VA normative aging study. Chest, 132:
1890-1897. http://dx.doi.org/10.1378/chest.07-1126 195862
Alexeeff, S. E.; Litonjua, A. A.; Wright, R. O.; Baccarelli, A.; Suh, H.; Sparrow, D.; Vokonas, P. S.; Schwartz, J. (2008).
Ozone exposure, antioxidant genes, and lung function in an elderly cohort: VA normative aging study. Occup
Environ Med, 65: 736-742. http://dx.doi.org/10.1136/oem.2007.035253 195864
Alexis, N. E.; Zhou, H.; Lay, J. C.; Harris, B.; Hernandez, M. L.; Lu, T. S.; Bromberg, P. A.; Diaz-Sanchez, D.; Devlin, R.
B.; Kleeberger, S. R.; Peden, D. B. (2009). The glutathione-S-transferase Mu 1 null genotype modulates ozone-
induced airway inflammation in human subjects. J Allergy Clin Immunol, 124: 1222-1228.
http://dx.doi.0rg/10.1016/i.iaci.2009.07.036628542
Arbex, A. M.; de Souza Conceicao, G. M.; Perez Cendon, S.; Arbex, F. F.; Lopes, A. C.; Providello Moyses, E.; Santiago,
S. L.; Saldiva, P. H. N.; Pereira, L. A. A.; Ferreira Braga, A. L. (2009). Urban air pollution and COPD-related
emergency room visits. J Epidemiol Community Health, 966: 777-783. http://dx.doi.org/10.1136/jech.2008.078360
184334
Auten, R. L.; Potts, E. N.; Mason, S. N.; Fischer, B.; Huang, Y; Foster, W. M. (2009). Maternal exposure to particulate
matter increases postnatal ozone-induced airway hyperreactivity in juvenile mice. Am J Respir Crit Care Med, 180:
1218-1226. http://dx.doi.org/10.1164/rccm.200901-01160C 200760
Avissar, N. E.; Reed, C. K.; Cox, C.; Frampton, M. W.; Finkelstein, J. N. (2000). Ozone, but not nitrogen dioxide, exposure
decreases glutathione peroxidases in epithelial lining fluid of human lung. Am J Respir Crit Care Med, 162: 1342-
1347. 012528
Baccarelli, A.; Zanobetti, A.; Martinelli, I.; Grillo, P.; Hou, L.; Lanzani, G; Mannucci, P. M.; Bertazzi, P. A.; Schwartz, J.
(2007). Air pollution, smoking, and plasma homocysteine. Environ Health Perspect, 115: 176-181. 091310
Barraza-Villarreal, A.; Sunyer, J.; Hernandez-Cadena, L.; Escamilla-Nunez, M. C.; Sienra-Monge, J. J.; Ramirez-Aguilar,
M.; Cortez-Lugo, M.; Holguin, F.; Diaz-Sanchez, D.; Olin, A. C.; Romieu, I. (2008). Air pollution, airway
inflammation, and lung function in a cohort study of Mexico City schoolchildren. Environ Health Perspect, 116:
832-838. http://dx.doi.org/10.1289/ehp.10926 156254
Basha, M. A.; Gross, K. B.; Gwizdala, C. J.; Haidar, A. H.; Popovich, J. Jr (1994). Bronchoalveolar lavage neutrophilia in
asthmatic and healthy volunteers after controlled exposure to ozone and filtered purified air. Chest, 106: 1757-1765.
075950
Bell, M. L.; Dominici, F. (2008). Effect modification by community characteristics on the short-term effects of ozone
exposure and mortality in 98 US communities. Am J Epidemiol, 167: 986-997.
http://dx.doi.org/10.1093/aje/kwm396 193828
Bennett, W. D.; Hazucha, M. J.; Folinsbee, L. J.; Bromberg, P. A.; Kissling, G. E.; London, S. J. (2007). Acute pulmonary
function response to ozone in young adults as a function of body mass index. Inhal Toxicol, 19:1147-1154.
http://dx.doi.org/10.1080/08958370701665475418827
Bergamaschi, E.; De Palma, G; Mozzoni, P.; Vanni, S.; Vettori, M. V; Broeckaert, F.; Bernard, A.; Mutti, A. (2001).
Polymorphism of quinone-metabolizing enzymes and susceptibility to ozone-induced acute effects. Am J Respir
Crit Care Med, 163: 1426-1431. 052670
Blomberg, A.; Mudway, I. S.; Nordenhall, C.; Hedenstrom, H.; Kelly, F. J.; Frew, A. J.; Holgate, S. T.; Sandstrom, T.
(1999). Ozone-induced lung function decrements do not correlate with early airway inflammatory or antioxidant
responses. Eur Respir J, 13: 1418-1428. 001267
Broeckaert, F.; Clippe, A.; Wattiez, R.; Falmagne, P.; Bernard, A. (2003). Lung hyperpermeability, Clara-cell secretory
potein (CC16), and susceptibility to ozone of five inbred strains of mice. Inhal Toxicol, 15: 1209-1230. 055490
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments such as the Integrated Science Assessments (ISAs) and the Integrated Risk Information System (IRIS).
March 2011 8-25 DRAFT - DO NOT CITE OR QUOTE
-------
Buadong, D.; Jinsart, W.; Funatagawa, I.; Karita, K.; Yano, E. (2009). Association between PM10 and O-3 levels and
hospital visits for cardiovascular diseases in Bangkok, Thailand. J Epidemiol, 19: 182-188.
http://dx.doi.org/10.2188/iea.JE20080047602060
Burra, T. A.; Moineddin, R.; Agha, M. M.; Glazier, R. H. (2009). Social disadvantage, air pollution, and asthma physician
visits in Toronto, Canada. Environ Res, 109: 567-574. http://dx.doi.Org/10.1016/j.envres.2009.03.004 195868
Bush, M. L.; Asplund, P. T; Miles, K. A.; Ben-Jebria, A.; Ultman, J. S. (1996). Longitudinal distribution of O3 absorption
in the lung: gender differences and intersubject variability. J Appl Physiol, 81: 1651-1657. 080763
Cakmak, S.; Dales, R. E.; Judek, S. (2006). Do gender, education, and income modify the effect of air pollution gases on
cardiac disease? J Occup Environ Med, 48: 89-94. http://dx.doi.org/10.1097/01.jom.0000184878.11956.4b 099068
Cakmak, S.; Dales, R. E.; Judek, S. (2006). Respiratory health effects of air pollution gases: Modification by education and
income. Arch Environ Occup Health, 61:5-10. 093272
Cakmak, S.; Dales, R. E.; Vidal, C. B. (2007). Air pollution and mortality in Chile: Susceptibility among the elderly.
Environ Health Perspect, 115: 524-7. 091170
Calderon Guzman, D.; Barragan Mejia, G; Hernandez Garcia, E.; Juarez Olguin, H. (2006). Effect of nutritional status and
ozone exposure on some biomarkers of oxidative stress in rat brain regions. Nutr Cancer, 55: 195-200.
http://dx.doi.org/10.1207/sl5327914nc5502_ll 596371
Carbajal-Arroyo, L.; Miranda-Soberanis, V; Medina-Ramon, M.; Rojas-Bracho, L.; Tzintzun, G; Solis-Gutierrez, P.;
Mendez-Ramirez, L; Hurtado-Diaz, M.; Schwartz, J.; Romieu, I. (In Press). Effect of PM10 and O3 on infant
mortality among residents in the Mexico City Metropolitan Area: A case-crossover analysis, 1997-2005. J
Epidemiol Community Health, http://dx.doi.org/10.1136/jech.2009.101212 667773
Carey, S. A.; Minard, K. R.; Trease, L. L.; Wagner, J. G; Garcia, G. J.; Ballinger, C. A.; Kimbell, J. S.; Plopper, C. G;
Corley, R. A.; Postlethwait, E. M.; Harkema, J. R.; Einstein, D. R. (2007). Three-dimensional mapping of ozone-
induced injury in the nasal airways of monkeys using magnetic resonance imaging and morphometric techniques.
Toxicol Pathol, 35: 27-40. http://dx.doi.org/10.1080/01926230601072343 195752
Chiu, H. F.; Yang, C. Y (2009). Air Pollution and Emergency Room Visits for Arrhythmias: Are There Potentially Sensitive
Groups? J Toxicol Environ Health A, 72: 817-823. http://dx.doi.org/10.1080/15287390902800405 603236
Chuang, G. C.; Yang, Z.; Westbrook, D. G; Pompilius, M.; Ballinger, C. A.; White, R. C.; Krzywanski, D. M.; Postlethwait,
E. M.; Ballinger, S. W. (2009). Pulmonary ozone exposure induces vascular dysfunction, mitochondrial damage,
and atherogenesis. Am J Physiol Lung Cell Mol Physiol, 297: L209-L216.
http://dx.doi.org/10.1152/aiplung.00102.2009 197202
Delfmo, R. J.; Staimer, N.; Tjoa, T; Arhami, M.; Polidori, A.; Gillen, D. L.; George, S. C.; Shafer, M. M.; Schauer, J. J.;
Sioutas, C. (2010). Associations of primary and secondary organic aerosols with airway and aystemic inflammation
in an elderly panel cohort. Epidemiology, 21: 892-902. http://dx.doi.org/10.1097/EDE.Ob013e3181f20e6c 647222
Escamilla-Nunez, M. C.; Barraza-Villarreal, A.; Hernandez-Cadena, L.; Moreno-Macias, H.; Ramirez-Aguilar, M.; Sienra-
Monge, J. J.; Cortez-Lugo, M.; Texcalac, J. L.; del Rio-Navarro, B.; Romieu, I. (2008). Traffic-related air pollution
and respiratory symptoms among asthmatic children, resident in Mexico City: The EVA cohort study. Respir Res, 9:
74. http://dx.doi.org/10.1186/1465-9921-9-74 594284
Fanucchi, M. V; Plopper, C. G; Evans, M. J.; Hyde, D. M.; Van Winkle, L. S.; Gershwin, L. J.; Schelegle, E. S. (2006).
Cyclic exposure to ozone alters distal airway development in infant rhesus monkeys. Am J Physiol Lung Cell Mol
Physiol, 291: L644-L650. http://dx.doi.org/10.1152/ajplung.00027.2006 096491
Fortino, V; Maioli, E.; Torricelli, C.; Davis, P.; Valacchi, G. (2007). Cutaneous MMPs are differently modulated by
environmental stressors in old and young mice. Toxicol Lett, 173: 73-79.
http://dx.doi.0rg/10.1016/i.toxlet.2007.06.004596382
Fox, S. D.; Adams, W. C.; Brookes, K. A.; Lasley, B. L. (1993). Enhanced response to ozone exposure during the follicular
phase of the menstrual cycle. Environ Health Perspect, 101: 242-244. 043906
Frampton, M. W.; Morrow, P. E.; Torres, A.; Cox, C.; Voter, K. Z.; Utell, M. J.; Gibb, F. R.; Speers, D. M. (1997). Ozone
responsiveness in smokers and nonsmokers. Am J Respir Crit Care Med, 155: 116-121. 082692
March 2011 8-26 DRAFT - DO NOT CITE OR QUOTE
-------
Funabashi, H.; Shima, M.; Kuwaki, T.; Hiroshima, K.; Kuriyama, T. (2004). Effects of repeated ozone exposure on
pulmonary function and bronchial responsiveness in mice sensitized with ovalbumin. Toxicology, 204: 75-83.
http://dx.doi.0rg/10.1016/i.tox.2004.06.047596384
Goldberg, M. S.; Burnett, R. T.; Yale, J. F.; Valois, M. F.; Brook, J. R. (2006). Associations between ambient air pollution
and daily mortality among persons with diabetes and cardiovascular disease. Environ Res, 100: 255-267. 088641
Halonen, J. I.; Lanki, T.; Tiittanen, R; Niemi, J. V.; Loh, M.; Pekkanen J (2009). Ozone and cause-specific
cardiorespiratory morbidity and mortality. J Epidemiol Community Health, TBD: TBD.
http://dx.doi.org/10.1136/iech.2009.087106625764
Hamade, A. K.; Misra, V.; Rabold, R.; Tankersley, C. G. (2010). Age-related changes in cardiac and respiratory adaptation
to acute ozone and carbon black exposures: Interstrain variation in mice. Inhal Toxicol, 22: 84-94.
http://dx.doi.org/10.3109/08958378.2010.503974666324
Hamade, A. K.; Rabold, R.; Tankersley, C. G. (2008). Adverse cardiovascular effects with acute particulate matter and
ozone exposures: Interstrain variation in mice. Environ Health Perspect, 116: 1033-1039. 156515
Hamade, A. K.; Tankersley, C. G. (2009). Interstrain variation in cardiac and respiratory adaptation to repeated ozone and
particulate matter exposures. Am J Physiol Regul Integr Comp Physiol, 296: R1202-R1215.
http://dx.doi.org/10.1152/aipregu.90808.2008596386
Harkema, J. R.; Plopper, C. G; Hyde, D. M.; StGeorge, J. A.; Dungworth, D. L. (1987). Effects of an ambient level of
ozone on primate nasal epithelial mucosubstances: quantitative histochemistry. Am J Pathol, 127: 90-96.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMCl899606/. 040816
Hazucha, M. J.; Folinsbee, L. J.; Bromberg, P. A. (2003). Distribution and reproducibility of spirometric response to ozone
by gender and age. J Appl Physiol, 95: 1917-1925. 048168
Henrotin, J. B.; Besancenot, J. P.; Bejot, Y; Giroud, M. (2007). Short-term effects of ozone air pollution on ischaemic
stroke occurrence: A case-crossover analysis from a 10-year population-based study in Dijon, France. Occup
Environ Med, 64: 439-445. 093270
Hernandez-Cadena, L.; Holguin, F.; Barraza-Villarreal, A.; Del Rio-Navarro, B. E.; Sienra-Monge, J. J.; Romieu, I. (2009).
Increased levels of outdoor air pollutants are associated with reduced bronchodilation in children with asthma.
Chest, 136: 1529-1536. http://dx.doi.org/10.1378/chest.08-1463 594283
Horstman, D. H.; Ball, B. A.; Brown, J.; Gerrity, T; Folinsbee, L. J. (1995). Comparison of pulmonary responses of
asthmatic and nonasthmatic subjects performing light exercise while exposed to a low level of ozone. Toxicol Ind
Health, 11: 369-385. 075834
Housley, D. G.; Eccles, R.; Richards, R. J. (1996). Gender difference in the concentration of the antioxidant uric acid in
human nasal lavage. Acta Otolaryngol, 116: 751-754. 080811
Hu, S.-C.; Ben-Jebria, A.; Ultman, J. S. (1994). Longitudinal distribution of ozone absorption in the lung: effects of
respiratory flow. J Appl Physiol, 77: 574-583. 041323
Huffman, L. J.; Beighley, C. M.; Frazer, D. G; McKinney, W. G; Porter, D. W. (2006). Increased susceptibility of the lungs
of hyperthyroid rats to oxidant injury: Specificity of effects. Toxicology, 225: 119-127.
http://dx.doi.0rg/10.1016/i.tox.2006.05.008596388
Islam, T; Berhane, K.; McConnell, R.; Gauderman, W. I; Avol, E.; Peters, J. M.; Gilliland, F. D. (2009). Glutathione-S-
transferase (GST) PI, GSTM1, exercise, ozone and asthma incidence in school children. Thorax, 64: 197-202.
http://dx.doi.org/10.1136/thx.2008.099366 196715
Islam, T; McConnell, R.; Gauderman, W. J.; Avol, E.; Peters, J. M.; Gilliland, F. D. (2008). Ozone, oxidant defense genes
and risk of asthma during adolescence. Am J Respir Crit Care Med, 177: 388-395.
http://dx.doi.org/10.1164/rccm.200706-863OC097348
Jerrett, M.; Burnett, R. T; Pope, C. A.; Ito, K.; Thurston, G; Krewski, D.; Shi, Y; Calle, E.; Thun, M. (2009). Long-term
ozone exposure and mortality. N Engl J Med, 360: 1085-1095. http://dx.doi.org/10.1056/NEJMoa0803894 194160
Joad, J. P.; Kott, K. S.; Brie, J. M.; Peake, J. L.; Plopper, C. G; Schelegle, E. S.; Gershwin, L. J.; Pinkerton, K. E. (2006).
Structural and functional localization of airway effects from episodic exposure of infant monkeys to allergen and/or
ozone. Toxicol Appl Pharmacol, 214: 237-243. http://dx.doi.Org/10.1016/j.taap.2005.12.012 596390
March 2011 8-27 DRAFT - DO NOT CITE OR QUOTE
-------
Kan, H.; London, S. I; Chen, G; Zhang, Y; Song, G; Zhao, N.; Jiang, L.; Chen, B. (2008). Season, sex, age, and education
as modifiers of the effects of outdoor air pollution on daily mortality in Shanghai, China: The Public Health and Air
Pollution in Asia (PAPA) Study. Environ Health Perspect, 116: 1183-1188. 156621
Katsouyanni, K.; Samet, J. M.; Anderson, H. R.; Atkinson, R.; Le Tertre, A.; Medina, S.; Samoli, E.; Touloumi, G; Burnett,
R. T.; Krewski, D.; Ramsay, T.; Dominici, F.; Peng, R. D.; Schwartz, J.; Zanobetti, A. (2009). Air pollution and
health: A European and North American approach (APHENA) (Report No. Research Report 142). Boston, MA:
Health Effects Institute. http://pubs.healtheffects.org/view.php?id=327. 199899
Khatri, S. B.; Holguin, F. C.; Ryan, P. B.; Mannino, D.; Erzurum, S. C.; league, W. G. (2009). Association of ambient
ozone exposure with airway inflammation and allergy in adults with asthma. J Asthma, 46: 777-785.
http://dx.doi.org/10.1080/02770900902779284594282
Kleeberger, S. R.; Reddy, S. P.; Zhang, L.-Y.; Cho, H.-Y; Jedlicka, A. E. (2001). Toll-like receptor 4 mediates ozone-
induced murine lung hyperpermeability via inducible nitric oxide synthase. Am J Physiol, 280: L326-L333. 016163
Kleeberger, S. R.; Reddy, S.; Zhang, L.-Y; Jedlicka, A. E. (2000). Genetic susceptibility to ozone-induced lung
hyperpermeability: role of toll-like receptor 4. Am J Respir Cell Mol Biol, 22: 620-627. 014895
Ko, F. W. S.; Tarn, W.; Wong, T. W.; Lai, C. K. W.; (2007). Effects of air pollution on asthma hospitalization rates in
different age groups in Hong Kong. Clin Exp Allergy, 37: 1312-1319. 092844
Kodavanti, U. P.; Costa, D. L.; Dreher, K. L.; Crissman, K.; Hatch, G. E. (1995). Ozone-induced tissue injury and changes
in antioxidant homeostasis in normal and ascorbate-deficient guinea pigs. Biochem Pharmacol, 50: 243-251.
http://dx.doi.org/10.1016/0006-2952(95)00122-0 077440
Kooter, I. M.; Pennings, J. L.; Fokkens, P. H.; Leseman, D. L.; Boere, A. J.; Gerlofs-Nijland, M. E.; Cassee, F. R.; Schalk,
J. A.; Orzechowski, T. J.; Schaap, M. M.; Breit, T. M.; Dormans, J. A.; van Oostrom, C. T; de Vries, A.; van Steeg,
H. (2007). Ozone induces clear cellular and molecular responses in the mouse lung independently of the
transcription-coupled repair status. J Appl Physiol, 102: 1185-1192.
http://dx.doi.org/10.1152/iapplphvsiol.00796.2006596397
Kumarathasan, P.; Blais, E.; Goegan, P.; Yagminas, A.; Guenette, J.; Adamson, I. Y; Crapo, J. D.; Mason, R. J.; Vincent, R.
(2005). 90-day repeated inhalation exposure of surfactant Protein-C/tumor necrosis factor-alpha, (SP-C/TNF-alpha)
transgenic mice to air pollutants. Int J Toxicol, 24: 59-67.
http://www.ncbi. nlm.nih.gov/entrez/query. fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids= 15981741.
596398
Lagorio, S.; Forastiere, F.; Pistelli, R.; lavarone, I.; Michelozzi, P.; Fano, V; Marconi, A.; Ziemacki, G; Ostro, B. D.
(2006). Air pollution and lung function among susceptible adult subjects: A panel study. Environ Health, 5:11.
http://dx.doi.org/10.1186/1476-069X-5-11089800
Lee, I. M.; Tsai, S. S.; Ho, C. K.; Chiu, H. F.; Wu, T. N.; Yang, C. Y. (2008). Air pollution and hospital admissions for
congestive heart failure: are there potentially sensitive groups? Environ Res, 108: 348-353.
http://dx.doi.0rg/10.1016/i.envres.2008.07.024 192076
Lee, J. T; Son, J. Y; Kim, H.; Kim, S. Y. (2006). Effect of air pollution on asthma-related hospital admissions for children
by socioeconomic status associated with area of residence. Arch Environ Occup Health, 61: 123-130. 098248
Lee, Y. L.; McConnell, R.; Berhane, K.; Gilliland, F. D. (2009). Ambient ozone modifies the effect of tumor necrosis factor
G-308A on bronchitic symptoms among children with asthma. Allergy, 64: 1342-1348.
http://dx.doi.0rg/10.llll/i.1398-9995.2009.02014.xl99915
Lewis, T. C.; Robins, T. G; Dvonch, J. T; Keeler, G. J.; Yip, F. Y; Mentz, G. B.; Lin, X.; Parker, E. A.; Israel, B. A.;
Gonzalez, L.; Hill, Y. (2005). Air pollution-associated changes in lung function among asthmatic children in
Detroit. Environ Health Perspect, 113: 1068-1075. 081079
Liao, D.; Heiss, G; Chinchilli, V. M.; Duan, Y; Folsom, A. R.; Lin, H. M.; Salomaa, V. (2005). Association of criteria
pollutants with plasma hemostatic/inflammatory markers: a population-based study. J Expo Sci Environ Epidemiol,
15: 319-328. 088677
Lim, Y; Phung, A. D.; Corbacho, A. M.; Aung, H. H.; Maioli, E.; Reznick, A. Z.; Cross, C. E.; Davis, P. A.; Valacchi, G.
(2006). Modulation of cutaneous wound healing by ozone: Differences between young and aged mice. Toxicol Lett,
160: 127-134. http://dx.doi.Org/10.1016/j.toxlet.2005.06.013 670834
March 2011 8-28 DRAFT - DO NOT CITE OR QUOTE
-------
Lin, M.; Stieb, D. M.; Chen, Y. (2005). Coarse particulate matter and hospitalization for respiratory infections in children
younger than 15 years in Toronto: A case-crossover analysis. Pediatrics, 116: 235-240. 087828
Lin, S.; Liu, X.; Le, L. H.; Hwang, S. A. (2008). Chronic exposure to ambient ozone and asthma hospital admissions
among children. Environ Health Perspect, 116: 1725-1730. http://dx.doi.org/10.1289/ehp.11184 196680
Liu, L.; Poon, R.; Chen, L.; Frescura, A. M.; Montuschi, P.; Ciabattoni, G; Wheeler, A.; Dales, R. (2009). Acute effects of
air pollution on pulmonary function, airway inflammation, and oxidative stress in asthmatic children. Environ
Health Perspect, 117: 668-674. http://dx.doi.org/10.1289/ehpll813 192003
Lopez, L; Sanchez, L; Bizarro, P.; Acevedo, S.; Ustarroz, M.; Fortoul, T. (2008). Ultrastructural alterations during
embryonic rats' lung development caused by ozone. J Electron Microsc (Tokyo), 57: 19-23.
http://dx.doi.org/10.1093/jmicro/dfm033 197786
Mar, T. F.; Koenig, J. Q. (2009). Relationship between visits to emergency departments for asthma and ozone exposure in
greater Seattle, Washington. Ann Allergy Asthma Immunol, 103: 474-479.
http://www.ingentaconnect.com/content/acaai/aaai/2009/00000103/00000006/art00005. 594410
McDonnell, W. F.; Stewart, P. W.; Smith, M. V. (2010). Prediction of ozone-induced lung function responses in humans.
Inhal Toxicol, 22: 160-168. http://dx.doi.org/10.1080/08958370903089557 383972
McDonnell, W. F.; Stewart, P. W.; Smith, M. V; Pan, W. K.; Pan, J. (1999). Ozone-induced respiratory symptoms:
Exposure-response models and association with lung function. Eur Respir J, 14: 845-853. 010939
Medina-Ramon, M.; Schwartz, J. (2008). Who is more vulnerable to die from ozone air pollution? Epidemiology, 19: 672-
679. 193829
Medina-Ramon, M.; Zanobetti, A.; Schwartz, J. (2006). The effect of ozone and PM10 on hospital admissions for
pneumonia and chronic obstructive pulmonary disease: A national multicity study. Am J Epidemiol, 163: 579-588.
http://dx.doi.org/10.1093/aje/kwj078 087721
Middleton, N.; Yiallouros, P.; Kleanthous, S.; Kolokotroni, O.; Schwartz, J.; Dockery, D. W.; Demokritou, P.; Koutrakis, P.
(2008). A 10-year time-series analysis of respiratory and cardiovascular morbidity in Nicosia, Cyprus: The effect of
short-term changes in air pollution and dust storms. Environ Health, 7: 39. 156760
Mudway, I. S.; Blomberg, A.; Frew, A. J.; Holgate, S. T; Sandstrom, T; Kelly, F. J. (1999). Antioxidant consumption and
repletion kinetics in nasal lavage fluid following exposure of healthy human volunteers to ozone. Eur Respir J, 13:
1429-1438. 001270
Mudway, I. S.; Stenfors, N.; Blomberg, A.; Helleday, R.; Dunster, C.; Marklund, S. L.; Frew, A. J.; Sandstrom, T; Kelly, F.
J. (2001). Differences in basal airway antioxidant concentrations are not predictive of individual responsiveness to
ozone: A comparison of healthy and mild asthmatic subjects. Free Radic Biol Med, 31: 962-974. 025327
Neidell, M.; Kinney, P. L. (2010). Estimates of the association between ozone and asthma hospitalizations that account for
behavioral responses to air quality information. Environ Sci Pol, 13: 97-103.
http://dx.doi.0rg/10.1016/i.envsci.2009.12.006384492
Nodelman, V.; Ultman, J. S. (1999). Longitudinal distribution of chlorine absorption in human airways: A comparison to
ozone absorption. J Appl Physiol, 87: 2073-2080. http://jap.physiology.Org/cgi/content/full/87/6/2073. 015112
Oyarzun, M.; Dussaubat, N.; Gonzalez, S. (2005). Effect of 0.25 ppm ozone exposure on pulmonary damage induced by
bleomycin. Biol Res, 38: 353-358. 596407
Palli, D.; Sera, F.; Giovannelli, L.; Masala, G; Grechi, D.; Bendinelli, B.; Caini, S.; Dolara, P.; Saieva, C. (2009).
Environmental ozone exposure and oxidative DNA damage in adult residents of Florence, Italy. Environ Pollut,
157: 152171525. http://dx.doi.0rg/10.1016/j.envpol.2008.09.Oil 196688
Paulu, C.; Smith, A. E. (2008). Tracking associations between ambient ozone and asthma-related emergency department
visits using case-crossover analysis. J Public Health Manag Pract, 14: 581-591. 180168
Peden, D. B.; Boehlecke, B.; Horstman, D.; Devlin, R. (1997). Prolonged acute exposure to 0.16 ppm ozone induces
eosinophilic airway inflammation in asthmatic subjects with allergies. J Allergy Clin Immunol, 100: 802-808.
085842
Peel, J. L.; Metzger, K. B.; Klein, M.; Flanders, W. D.; Mulholland, J. A.; Tolbert, P. E. (2007). Ambient air pollution and
cardiovascular emergency department visits in potentially sensitive groups. Am J Epidemiol, 165: 625-633. 090442
March 2011 8-29 DRAFT - DO NOT CITE OR QUOTE
-------
Pleis, J. R.; Lucas, J. W.; Ward, B. W. (2009). Summary health statistics for U.S. adults: National health interview survey,
2008 (Report No. DHHS Publication No. (PHS) 2010-1570). Atlanta, GA: Centers for Disease Control and
Prevention, National Center for Health Statistics. 629608
Plopper, C. G; Smiley-Jewell, S. M.; Miller, L. A.; Fanucchi, M. V; Evans, M. J.; Buckpitt, A. R.; Avdalovic, M.;
Gershwin, L. J.; Joad, J. P.; Kajekar, R.; Larson, S.; Pinkerton, K. E.; Van Winkle, L. S.; Schelegle, E. S.; Pieczarka,
E. M.; Wu, R.; Hyde, D. M. (2007). Asthma/allergic airways disease: Does postnatal exposure to environmental
toxicants promote airway pathobiology? Toxicol Pathol, 35: 97-110. http://dx.doi.org/10.1080/01926230601132030
596412
Qian, Z.; Lin, H. M.; Chinchilli, V. M.; Lehman, E. B.; Duan, Y; Craig, T. J.; Wilson, W. E.; Liao, D.; Lazarus, S. C.;
Bascom, R. (2009). Interaction of ambient air pollution with asthma medication on exhaled nitric oxide among
asthmatics. Arch Environ Occup Health, 64: 168-176. http://dx.doi.org/10.1080/19338240903240616 548793
Romieu, I.; Barraza-Villarreal, A.; Escamilla-Nunez, C.; Texcalac-Sangrador, J. L.; Hernandez-Cadena, L.; Diaz-Sanchez,
D.; De Batlle, J.; Del Rio-Navarro, B. E. (2009). Dietary intake, lung function and airway inflammation in Mexico
City school children exposed to air pollutants. Respir Res, 10: 122. 548788
Romieu, I.; Ramirez-Aguilar, M.; Moreno-Macias, H.; Barraza-Villarreal, A.; Miller, P.; Hernandez-Cadena, L.; Carbajal-
Arroyo, L. A.; Hernandez-Avila, M. (2004). Infant mortality and air pollution: Modifying effect by social class. J
Occup Environ Hyg, 46: 1210-1216. 093074
Romieu, I.; Ramirez-Aguilar, M.; Sienra-Monge, J. J.; Moreno-Macias, H.; Del Rio-Navarro, B. E.; David, G; Marzec, J.;
Hernandez-Avila, M.; London, S. (2006). GSTM1 and GSTP1 and respiratory health in asthmatic children exposed
to ozone. Eur Respir J, 28: 953-959. http://dx.doi.org/10.1183/09031936.06.00114905 090969
Romieu, I.; Sienra-Monge, J. J.; Ramirez-Aguilar, M.; Moreno-Macias, H.; Reyes-Ruiz, N. L; Estela del Rio-Navarro, B.;
Hernandez-Avila, M.; London, S. J. (2004). Genetic polymorphism of GSTM1 and antioxidant supplementation
influence lung function in relation to ozone exposure in asthmatic children in Mexico City. Thorax, 59: 8-10.
056796
Ruidavets, J.-B.; Cassadou, S.; Cournot, M.; Bataille, V; Meybeck, M.; Ferrieres, J. (2005). Increased resting heart rate
with pollutants in a population based study. J Epidemiol Community Health, 59: 685-693. 089443
Sacks, J. D.; Stanek, L. W.; Luben, T. J.; Johns, D. O.; Buckley, B. J.; Brown, J. S.; Ross, M. (In Press). Particulate matter
induced health effects: Who's susceptible? Environ Health Perspect. http://dx.doi.org/10.1289/ehp.1002255 664486
Salam, M. T.; Islam, T.; Gauderman, W. J.; Gilliland, F. D. (2009). Roles of arginase variants, atopy, and ozone in
childhood asthma. J Allergy Clin Immunol, 123: 596-602. http://dx.doi.org/10.1016/jjaci.2008.12.020 596644
Samet, J. M.; Hatch, G. E.; Horstman, D.; Steck-Scott, S.; Arab, L.; Bromberg, P. A.; Levine, M.; McDonnell, W. F.;
Devlin, R. B. (2001). Effect of antioxidant supplementation on ozone-induced lung injury in human subjects. Am J
Respir Crit Care Med, 164: 819-825. 019034
Sarangapani, R.; Gentry, P. R.; Covington, T. R.; Teeguarden, J. G; Clewell HJ 3rd (2003). Evaluation of the potential
impact of age- and gender-specific lung morphology and ventilation rate on the dosimetry of vapors. Inhal Toxicol,
15: 987-1016. http://dx.doi.org/10.1080/713857276 054581
Sawyer, K.; Brown, J.; HazuchaM; Bennett, W. D. (2007). The effect of exercise on nasal uptake of ozone in healthy
human adults. J Appl Physiol, 102: 1380-1386. http://dx.doi.org/10.1152/iapplphvsiol.00269.2006 195142
Scannell, C.; Chen, L.; Aris, R. M.; Tager, L; Christian, D.; Ferrando, R.; Welch, B.; Kelly, T; Balmes, J. R. (1996).
Greater ozone-induced inflammatory responses in subjects with asthma. Am J Respir Crit Care Med, 154: 24-29.
080755
Schelegle, E. S.; Miller, L. A.; Gershwin, L. J.; Fanucchi, M. V; Van Winkle, L. S.; Gerriets, J. E.; Walby, W. F.; Mitchell,
V; Tarkington, B. K.; Wong, V J.; Baker, G. L.; Pantle, L. M.; Joad, J. P.; Pinkerton, K. E.; Wu, R.; Evans, M. J.;
Hyde, D. M.; Plopper, C. G. (2003). Repeated episodes of ozone inhalation amplifies the effects of allergen
sensitization and inhalation on airway immune and structural development in Rhesus monkeys. Toxicol Appl
Pharmacol, 191: 74-85. 053778
Seal, E. Jr; McDonnell, W. F.; House, D. E. (1996). Effects of age, socioeconomic status, and menstrual cycle on
pulmonary response to ozone. Arch Environ Occup Health, 51:132-137. 044251
March 2011 8-30 DRAFT - DO NOT CITE OR QUOTE
-------
Seal, E. Jr; McDonnell, W. R; House, D. E.; Salaam, S. A.; Dewitt, P. I; Butler, S. O.; Green, I; Raggio, L. (1993). The
pulmonary response of white and black adults to six concentrations of ozone. Am J Respir Crit Care Med, 147: 804-
810.039357
Servais, S.; Boussouar, A.; Molnar, A.; Douki, T.; Pequignot, J. M.; Favier, R. (2005). Age-related sensitivity to lung
oxidative stress during ozone exposure. Free Radic Res, 39: 305-316.
http://dx.doi.org/10.1080/10715760400011098 195667
Sherry, B.; Blanck, H. M.; Galuska, D. A.; Pan, L.; Dietz, W. H.; Balluz, L. (2010). Vital signs: State-specific obesity
prevalence among adults — United States, 2009. MMWR Recomm Rep, 59: 951-955.
http://www.cdc.gov/mmwr/preview/mmwrhtml/mm59e0803al.htm. 667866
Shore, S. A.; Lang, J. E.; Kasahara, D. L; Lu, F. L.; Verbout, N. G; Si, H.; Williams, E. S.; Terry, R. D.; Lee, A.; Johnston,
R. A. (2009). Pulmonary responses to subacute ozone exposure in obese vs. lean mice. J Appl Physiol, 107: 1445-
1452. http://dx.doi.org/10.1152/japplphysiol.00456.2009 201551
Sienra-Monge, J. J.; Ramirez-Aguilar, M.; Moreno-Macias, H.; Reyes-Ruiz, N. I.; Del Rio-Navarro, B. E.; Ruiz-Navarro,
M. X.; Hatch, G; Crissman, K.; Slade, R.; Devlin, R. B.; Romieu, I. (2004). Antioxidant supplementation and nasal
inflammatory responses among young asthmatics exposed to high levels of ozone. Clin Exp Immunol, 138: 317-
322. http://dx.doi.0rg/10.llll/i.1365-2249.2004.02606.x 196422
Silverman, R. A.; Ito, K. (2010). Age-related association of fine particles and ozone with severe acute asthma in New York
City. J Allergy Clin Immunol, 125: 367-373.e5. http://dx.doi.Org/10.1016/j.jaci.2009.10.061 386252
Social Science Data Analysis Network; CensusScope (2010). United States: Age Distribution. Retrieved October 07, 2010
from http://www.censusscope.org/us/chart_age.html. 647298
Social Science Data Analysis Network; CensusScope (2010). United States: Population by race. Retrieved October 07,
2010 from http://www.censusscope.org/us/chart_race.html. 647301
Social Science Data Analysis Network; CensusScope (2010). United States: Poverty by age. Retrieved October 07, 2010
from http://www.censusscope.org/us/chart_poverty.html. 647297
Stafoggia, M.; Forastiere, F.; Faustini, A.; Biggeri, A.; Bisanti, L.; Cadum, E.; Cernigliaro, A.; Mallone, S.; Pandolfi, P.;
Serinelli, M.; Tessari, R.; Vigotti, M. A.; Perucci, C. A. (2010). Susceptibility factors to ozone-related mortality: A
population-based case-crossover analysis. Am J Respir Crit Care Med, 182: 376-384.
http://dx.doi.org/10.1164/rccm.200908-1269OC625034
Steinvil, A.; Kordova-Biezuner, L.; Shapira, L; Berliner, S.; Rogowski, O. (2008). Short-term exposure to air pollution and
inflammation-sensitive biomarkers. Environ Res, 106: 51-61. 188893
Tankersley, C. G; Kleeberger, S. R. (1994). Ozone-induced inflammation and altered ventilation in genetically susceptible
mice: a comparison of acute and subacute exposures. Toxicol Lett, 72: 279-289. 021420
Tankersley, C. G; Peng, R. D.; Bedga, D.; Gabrielson, K.; Champion, H. C. (2010). Variation in echocardiographic and
cardiac hemodynamic effects of PM and ozone inhalation exposure in strains related to Nppa and Nprl gene knock-
out mice. Inhal Toxicol, 22: 695-707. http://dx.doi.org/10.3109/08958378.2010.487549 628062
Thaller, E. I.; Petronella, S. A.; Hochman, D.; Howard, S.; Chhikara, R. S.; Brooks, E. G. (2008). Moderate increases in
ambient PM2.5 and ozone are associated with lung function decreases in beach lifeguards. J Occup Environ Med,
50: 202-211. http://dx.doi.org/10.1097/JOM.Ob013e31816386b4 195869
Torres, A.; Utell, M. J.; Morow, P. E.; Voter, K. Z.; Whitin, J. C.; Cox, C.; Looney, R. J.; Speers, D. M.; Tsai, Y; Frampton,
M. W. (1997). Airway inflammation in smokers and nonsmokers with varying responsiveness to ozone. Am J
Respir Crit Care Med, 156: 728-736. 084265
Tovalin, H.; Valverde, M.; Morandi, M. T; Blanco, S.; Whitehead, L.; Rojas, E. (2006). DNA damage in outdoor workers
occupationally exposed to environmental air pollutants. Occup Environ Med, 63: 230-236. 091322
Trenga, C. A.; Koenig, J. Q.; Williams, P. V (2001). Dietary antioxidants and ozone-induced bronchial
hyperresponsiveness in adults with asthma. Arch Environ Occup Health, 56: 242-249. 019845
Triche, E. W; Gent, J. F.; Holford, T. R.; Belanger, K.; Bracken, M. B.; Beckett, W. S.; Naeher, L.; McSharry, J.-E.;
Leaderer, B. P. (2006). Low-level ozone exposure and respiratory symptoms in infants. Environ Health Perspect,
114: 911-916. http://dx.doi.org/10.1289/ehp.8559 093274
March 2011 8-31 DRAFT - DO NOT CITE OR QUOTE
-------
U.S. Census Bureau (2010). U.S. population projections. Retrieved October 07, 2010 from
http://www.census.gov/population/www/projections/projectionsagesex.html. 647312
U.S. EPA (1996). Air quality criteria for ozone and related photochemical oxidants (Report No. EPA/600/P-93/004AF).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development. (NTIS
No. PB94173127) 017831
U.S. EPA (1996). Air quality criteria for ozone and related photochemical oxidants Volume III of III (Report No.
EPA/600/P-93/004cF). Research Triangle Park, NC: U.S. Environmental Protection Agency. (NTIS No. PB96-
185608)080828
U.S. EPA (2006). Aging and Toxic Response: Issues Relevant to Risk Assessment (Final). Washington, DC: U.S.
Environmental Protection Agency. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=156648. 192082
U.S. EPA (2006). Air quality criteria for ozone and related photochemical oxidants (Report No. EPA/600/R-05/004AF).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid= 149923. 088089
U.S. EPA (2009). Integrated science assessment for particulate matter (Report No. EPA/600/R-08/139F). Research Triangle
Park, NC: U.S. Environmental Protection Agency, Office of Research and Development.
http://cfpub.epa.gov/ncea/cfm/recordisplay. cfm?deid=216546. 179916
U.S. EPA (2010). Integrated science assessment for carbon monoxide (Report No. EPA/600/R-09/019F). Research Triangle
Park, NC: U.S. Environmental Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=218686. 626035
Vagaggini, B.; Bartoli, M. L.; Cianchetti, S.; Costa, R; Bacci, E.; Dente, F. L.; Di Franco, A.; Malagrino, L.; Paggiaro, P.
(2010). Increase in markers of airway inflammation after ozone exposure can be observed also in stable treated
asthmatics with minimal functional response to ozone. Respir Res, 11: 5. http://dx.doi.org/10.1186/1465-9921-ll-5
387127
Valacchi, G; Pecorelli, A.; Mencarelli, M.; Maioli, E.; Davis, P. A. (2009). Beta-carotene prevents ozone-induced
proinflammatory markers in murine skin. Toxicol Ind Health, 25: 241-247.
http://dx.doi.org/10.1177/0748233709103030201554
Valacchi, G; Vasu, V T; Yokohama, W.; Corbacho, A. M.; Phung, A.; Lim, Y; Aung, H. H.; Cross, C. E.; Davis, P. A.
(2007). Lung vitamin E transport processes are affected by both age and environmental oxidants in mice. Toxicol
Appl Pharmacol, 222: 227-234. http://dx.doi.Org/10.1016/j.taap.2007.04.010 596418
Vancza, E. M.; Galdanes, K.; Gunnison, A.; Hatch, G; Gordon, T. (2009). Age, strain, and gender as factors for increased
sensitivity of the mouse lung to inhaled ozone. Toxicol Sci, 107: 535-543. http://dx.doi.org/10.1093/toxsci/kfn253
596419
Villeneuve, P. J.; Chen, L.; Rowe, B. H.; Coates, F. (2007). Outdoor air pollution and emergency department visits for
asthma among children and adults: a case-crossover study in northern Alberta, Canada. Environ Health, 6: 40.
http://dx.doi.org/10.1186/1476-069X-6-40 195859
Villeneuve, P. J.; Chen, L.; Stieb, D.; Rowe, B. H. (2006). Associations between outdoor air pollution and emergency
department visits for stroke in Edmonton, Canada. Eur J Epidemiol, 21: 689-700. 090191
Voynow, J. A.; Fischer, B. M.; Zheng, S.; Potts, E. N.; Grover, A. R.; Jaiswal, A. K.; Ohio, A. J.; Foster, W. M. (2009).
NAD(P)H quinone oxidoreductase 1 is essential for ozone-induced oxidative stress in mice and humans. Am J
Respir Cell Mol Biol, 41: 107-113. 194311
Wagner, J. G; Harkema, J. R.; Jiang, Q.; Illek, B.; Ames, B. N.; Peden, D. B. (2009). Gamma-tocopherol attenuates ozone-
induced exacerbation of allergic rhinosinusitis in rats. Toxicol Pathol, 37: 481-491.
http://dx.doi.org/10.1177/0192623309335630201574
Wagner, J. G; Jiang, Q.; Harkema, J. R.; Illek, B.; Patel, D. D.; Ames, B. N.; Peden, D. B. (2007). Ozone enhancement of
lower airway allergic inflammation is prevented by gamma-tocopherol. Free Radic Biol Med, 43:1176-1188.
http://dx.doi.0rg/10.1016/i.freeradbiomed.2007.07.013596420
Wattiez, R.; Noel-Georis, L; Cruyt, C.; Broeckaert, F.; Bernard, A.; Falmagne, P. (2003). Susceptibility to oxidative stress:
proteomic analysis of bronchoalveolar lavage from ozone-sensitive and ozone-resistant strains of mice. Proteomics,
3: 658-665. http://dx.doi.org/10.1002/pmic.200300417043783
March 2011 8-32 DRAFT - DO NOT CITE OR QUOTE
-------
Wenten, M.; Gauderman, W. I; Berhane, K.; Lin, P. C.; Peters, J.; Gilliland, F. D. (2009). Functional variants in the
catalase and myeloperoxidase genes, ambient air pollution, and respiratory-related school absences: An example of
epistasis in gene-environment interactions. Am J Epidemiol, 170: 1494-1501. http://dx.doi.org/10.1093/aje/kwp310
597084
Wong, C. M.; Ou, C. Q.; Chan, K. P.; Chau, Y. K.; Thach, T. Q.; Yang, L.; Chung, R. Y; Thomas, G. N.; Peiris, J. S.; Wong,
T. W.; Hedley, A. J.; Lam, T. H. (2008). The effects of air pollution on mortality in socially deprived urban areas in
Hong Kong, China. Environ Health Perspect, 116: 1189-1194. 157151
Wong, C. M.; Yang, L.; Thach, T. Q.; Chau, P. Y; Chan, K. P.; Thomas, G. N.; Lam, T. H.; Wong, T. W.; Hedley, A. J.;
Peiris, J. S. (2009). Modification by influenza on health effects of air pollution in Hong Kong. Environ Health
Perspect, 117: 248-253. http://dx.doi.org/10.1289/ehp.11605 196722
Wong, C.-M.; Ou, C.-Q.; Thach, T.-Q.; Chau, Y.-K.; Chan, K.-P; Ho, S.-Y; Chung, R. Y; Lam, T.-H.; Hedley, A. J. (2007).
Does regular exercise protect against air pollution-associated mortality? Prev Med, 44: 386-392. 093278
Zanobetti, A.; Schwartz, J. (2008). Is there adaptation in the ozone mortality relationship: A multi-city case-crossover
analysis. Environ Health, 7: 22. http://dx.doi.org/10.1186/1476-069X-7-22 195755
March 2011 8-33 DRAFT - DO NOT CITE OR QUOTE
-------
Chapter 9. Environmental Effects: Ozone
Effects on Vegetation and Ecosystems
9.1. Introduction
1 This chapter synthesizes and evaluates the most policy-relevant science to help form the
2 scientific foundation for the review of a vegetation- and ecologically-based secondary NAAQS for
3 O3. The secondary NAAQS are based on welfare effects. The Clean Air Act (CAA) definition of
4 welfare effects includes, but is not limited to, effects on soils, water, wildlife, vegetation, visibility,
5 weather, and climate, as well as effects on materials, economic values, and personal comfort and
6 well-being. The effects of O3 as a greenhouse gas and its direct effects on climate are discussed in
7 Chapter 10 of this document.
8 The intent of the ISA, according to the CAA, is to "accurately reflect the latest scientific
9 knowledge expected from the presence of [a] pollutant in ambient air" (42U.S.C.7408 (1990,
10 080701) 42U.S.C.7409 (1990, 037658)). This chapter of the ISA includes scientific research from
11 biogeochemistry, soil science, plant physiology, and ecology conducted at multiple scales (e.g.,
12 organ, individual, population, community, ecosystem). Key information and judgments formerly
13 found in the AQCDs regarding O3 effects on vegetation and ecosystems are found in this chapter.
14 This chapter of the O3 ISA serves to update and revise Chapter 9 and AX9 of the 2006 O3 AQCD
15 (U.S. EPA, 2006,
16 Numerous studies of the effects of O3 on vegetation and ecosystems were reviewed in the
17 2006 O3 AQCD. That document concluded that the effects of ambient O3 on vegetation and
18 ecosystems appear to be widespread across the U.S., and experimental studies demonstrated
19 plausible mechanisms for these effects. Ozone effect studies published from 2005 to September 2010
20 are reviewed in this document in the context of the previous O3 AQCDs (U.S. EPA, 2006,
21 088089)(U.S. EPA, 1996, 080827)01.5. EPA, 1984, 029711)(U.S. EPA, 1978, 040586). From 2005
22 to 2010, some areas have had very little new research published and the reader is referred back to
23 sections of the 2006 O3 AQCD for a more comprehensive treatment of those subjects. This chapter is
24 focused on studies of vegetation and ecosystems that occur in the U.S. and that report endpoints or
25 processes most relevant to the review of the secondary standard. Many studies have been published
26 about vegetation and ecosystems outside of the U.S. and North America, largely in Europe and Asia.
27 This document includes discussion of studies of vegetation and ecosystems outside of North America
28 only if those studies contribute to the general understanding of O3 effects across species and
29 ecosystems. For example, studies outside North America are discussed that consider physiological
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments such as the Integrated Science Assessments (ISAs) and the Integrated Risk Information System (IRIS).
March 2011 9-1 DRAFT - DO NOT CITE OR QUOTE
-------
1 and biochemical processes that contribute to the understanding of effects of O3 across species. Also,
2 ecosystem studies outside of North America that contribute to the understanding of O3 effects on
3 general ecosystem processes are discussed in the chapter.
4 This chapter is organized in sections that discuss exposure methods, followed by effects on
5 vegetation and ecosystems at various scales and ends with policy-relevant discussions of exposure
6 indices and exposure-response. First, Section 9.2 presents summary information and conclusions
7 based on the literature assessed in this chapter and the assessments and conclusions of the previous
8 O3 AQCDs. A brief overview of various methodologies that have been, and continue to be, central to
9 quantifying O3 effects on vegetation is provided in Section 9.3 (AX9.1 of the 2006 O3 AQCD for
10 more detailed discussion) (U.S. EPA, 2006, 088089). Sections 9.4 through 9.6 begin with a
11 discussion of effects at the cellular and subcellular level followed by consideration of the whole
12 plant and finally, O3 impacts on ecosystem-level processes (Figure 9-1). In Section 9.4, research is
13 reviewed from the molecular to the biochemical and physiological levels in impacted plants, offering
14 insight into the mode of action of O3. Section 9.5 provides a review of the effects of O3 exposure on
15 major endpoints at the whole plant scale including growth, reproduction, visible foliar injury and leaf
16 gas exchange in woody and herbaceous plants in the U.S., as well as a brief discussion of O3 effects
17 on agricultural crop yield and quality. The response of plants to O3 as influenced by numerous
18 environmental biotic and abiotic factors is also discussed Section 9.5. In Section 9.6, available
19 research for assessing the effect of O3 on ecosystems is reviewed, along with data potentially
20 available for estimating the loss of various ecosystem services. The development of indices of O3
21 exposure and dose modeling is discussed in Section 9.7. Finally, exposure-response relationships for
22 a number of tree species, native vegetation, and crop species and cultivars are reviewed, tabulated,
23 and compared in Section 9.8 to form the basis for an assessment of the potential risk to vegetation
24 from current ambient levels of O3.
9.2. Summary and Integration
9.2.1. Introduction
25 The subsequent sections of this chapter will present the most policy-relevant information
26 related to this review of the NAAQS for the effects of O3 on vegetation and ecosystems. This section
27 integrates the key findings from the disciplines evaluated in this current assessment of the O3
28 scientific literature, which includes plant physiology, biochemistry, whole plant biology, ecosystems
29 and exposure-response.
30 Ozone effects at small scales, such as the leaf of an individual plant, can result in effects at a
31 continuum of larger scales. Figure 9-1 is a simplified diagram of the major pathway through which
32 O3 enters plants and the major endpoints O3 may affect from small to large scales. The subsequent
33 sections in this chapter are organized around this paradigm of effects at the cellular and subcellular
34 level followed by consideration of the whole plant and finally, O3 impacts on ecosystem-level
March 2011 9-2 DRAFT - DO NOT CITE OR QUOTE
-------
1 processes. Ozone enters leaves through stomata, and can alter stomatal conductance and disrupt CO2
2 fixation (Section 9.4). These effects can change rates of leaf gas exchange, growth and reproduction
3 at the individual plant level (Section 9.5). Those O3-induced effects can translate from the individual
4 plant level to the ecosystem level, and cause changes in ecosystem services, such as C storage, water
5 production, nutrient cycling, and community composition (Section 9.6). The EPA framework for
6 causal determinations described in Chapter 1 has been applied to the body of scientific evidence to
7 collectively examine effects attributed to O3 exposure (See Table 9-1).
March 2011 9-3 DRAFT - DO NOT CITE OR QUOTE
-------
Leaf
\7
Plant
\7
Ecosystem
Aboveground process
Belowground process
Water cycling
Light
Cuticle
Epidermis
Vascular
System
Epidermis
Cuticle
Photosynthesis
transpiration
Aboveground
growth
Reproduction;
Foliar Injury
Crop yield
and quality
^-^S^
respiration
Root
growth
Microbial activity
decomposition
Ecosystem productivity ^,
*^s&
r
Insects and wildlife
Competition and biodiversity
Water production
Soil C storage
Nutrient cycling
Figure 9-1. The effects of ozone at leaf, plant and ecosystem scales.
March 2011
9-4
DRAFT - DO NOT CITE OR QUOTE
-------
Table 9-1 Summary of ozone causal determinations for vegetation and ecosystem effects
Vegetation and Ecosystem Effects
Reduced Vegetation Growth
Alteration of Vegetation Reproduction
Visible Foliar Injury Effects on Vegetation
Alteration of Leaf Gas Exchange in Vegetation
Reduced Yield and Quality of Agricultural Crops
Reduced Productivity in Terrestrial Ecosystems
Reduced Carbon (C) Sequestration in Terrestrial Ecosystems
Alteration of Terrestrial Ecosystem Water Cycling
Alteration of Below-ground Biogeochemical Cycles
Alteration of Terrestrial Community Composition
Causality Determination
Causal
Causal
Causal
Causal
Causal
Causal
Likely Causal
Likely Causal
Causal
Likely Causal
9.2.2. Mechanisms Governing Response
1 Section 9.4 focuses on the effects of O3 stress on plants and their responses to that stress on the
2 molecular, biochemical and physiological levels. Many of the studies focus on the molecular
3 mechanisms that underlie the observed biochemical and physiological changes observed in many
4 plant species in response to O3 exposure. The results support and strengthen those reported in the
5 2006 O3 AQCD. The most significant change in this section from the 2006 O3 AQCD is the emphasis
6 on molecular mechanisms as new techniques, such as those used in evaluating transcriptomes (total
7 set of RNA transcripts in a particular cell at a particular time) and proteomes (total set of proteins
8 expressed in a particular cell at a particular time), have been utilized to perform very comprehensive
9 analyses of changes in gene transcription and protein expression in plants exposed to O3. These
10 newer molecular studies not only provide very important and wide-ranging information regarding
11 the many mechanisms of plant responses to O3, they also allow for the analysis of interactions
12 between various biochemical pathways which are induced in response to O3. However, many of
13 these studies are conducted in artificial conditions with model plants which are typically exposed to
14 very high, short doses of O3. Therefore, additional work remains to elucidate whether these plant
15 responses are transferable to other plant species exposed to more realistic ambient conditions.
16 Ozone is taken up into leaves through open stomata. Once inside the substomatal cavity, O3 is
17 thought to rapidly react with the aqueous layer surrounding the cell (apoplast) to form breakdown
18 products such as hydrogen peroxide (H2O2), superoxide (O2), hydroxyl radicals (HO) and peroxy
19 radicals (HO2). These radicals may play a role in signaling processes and may also interact with
20 sensitive molecules both outside and inside the cell to cause damage. This process was very
March 2011 9-5 DRAFT - DO NOT CITE OR QUOTE
-------
1 comprehensively described in the 2006 O3 AQCD and is only summarized in this document in
2 Section 9.4.2.
3 As plants have been shown to respond to O3 exposure very rapidly, this response must result
4 from a plant's ability to sense the presence of O3 and/or its breakdown products and then
5 communicate that information to the nucleus to initiate changes in gene expression. While it remains
6 unclear what the exact mechanism is by which the plant senses the presence of O3, whether there are
7 multiple simultaneous mechanisms by which O3 can be sensed, and how much variation exists in O3
8 sensing between species and exposure conditions, some progress has been made in the understanding
9 of this process since the 2006 O3 AQCD. Experimental evidence described in Section 9.4.3.1
10 suggests that O3 and/or its breakdown products may be directly sensed by apoplastic receptor
11 proteins (although they have not yet been identified). Additionally, a change in cellular redox state
12 due to plant exposure to O3 could be the manner in which plants sense the presence of the pollutant.
13 Once the plant has sensed the presence of the pollutant, there is much evidence to suggest that
14 mitogen-activated protein kinases (MAPK) play an important role in communicating signals to the
15 nucleus that result in gene expression changes in response to O3. Calcium has also been implicated to
16 play a role in the signal transduction processes. To summarize, the evidence to date suggests there
17 may be several mechanisms by which plants sense the presence of O3 and then communicate this
18 signal to the nucleus to induce changes in gene expression.
19 New technologies have allowed for the evaluation of changes in the entire transcriptome and
20 proteome, rather than analyzing the modification of the expression of individual genes and proteins;
21 the results of these studies are presented in Section 9.4.3.2. While transcriptome and proteome
22 analyses per se were not previously addressed, the 2006 O3 AQCD did provide much information
23 regarding changes in gene expression and protein quantity of individual genes and proteins in
24 O3-treated plants. In the transcriptome and proteome studies described here, O3 exposure conditions
25 (concentration, duration of exposure), plant species and sampling times vary significantly; however,
26 functional classification of the genes and proteins that are either up- or down-regulated by plant
27 exposure to O3 exhibit common trends. In summary, genes involved in plant defense, signaling, and
28 those associated with the synthesis of plant hormones and secondary metabolism are generally up-
29 regulated in plants exposed to O3, while those related to photosynthesis and general metabolism are
30 typically down-regulated. Proteome studies support these results by demonstrating concomitant
31 increases or decreases in the proteins encoded by these genes. The transcriptome and proteome
32 results support and enhance the findings of the 2006 O3 AQCD.
33 The 2006 O3 AQCD included a discussion on the role of phytohormones, including salicylic
34 acid, ethylene and jasmonic acid, in plant response to O3. Many additional studies using microarray
35 technology (used to determine changes in the transcriptome) and a variety of Arabidopsis mutants
36 are described in Section 9.4.3.3 and support the conclusions from the 2006 O3 AQCD. Transcriptome
37 analysis has also illuminated the complex interactions that exist between these hormones to better
38 define plant response to O3. To summarize, the results indicate that while ethylene and salicylic acid
39 are needed to develop O3-induced leaf lesions, jasmonic acid acts antagonistically to ethylene and
March 2011 9-6 DRAFT - DO NOT CITE OR QUOTE
-------
1 salicylic acid to limit the spread of the lesions. Abscisic acid, in addition to its role in regulating
2 stomatal aperture, may also act antagonistically to the jasmonic acid signaling pathway. Nitric oxide
3 has also been proposed to play a role in regulating O3-induced changes in gene expression; however,
4 its role is not yet well defined. Changes in phytohormones and the interactions between them reveal
5 some of the complexity of plant responses to an oxidative stressor such as O3.
6 Antioxidant metabolites, such as ascorbate and glutathione, and the enzymes that regenerate
7 them are a critical part of plant defense responses to oxidative stress. The role of ascorbate, which is
8 located in several cellular compartments and also in the apoplast, was comprehensively evaluated in
9 the 2006 O3 AQCD as a first line of defense (due to its location in the apoplast) against oxidative
10 stress. Ascorbate has also been the focus of studies investigating differences in O3 tolerance between
11 plant species or cultivars/genotypes within one species. While the studies evaluated for the current
12 document support the important role of ascorbate, several studies suggest that ascorbate quantity,
13 especially in the apoplast, is not the primary factor in determining plant tolerance to O3. In summary,
14 antioxidant metabolites and enzymes increase in quantity in plants exposed to O3. In most cases,
15 there is a correlation between the degree to which these defensive systems are induced and the
16 ability of the plant to tolerate exposure to O3. This up-regulation of antioxidant defenses and the
17 need to keep antioxidant metabolites in a reduced state requires a significant shift in C metabolism
18 away from growth and reproduction to sustain the energy needs of the plant for defense.
19 While declines in C fixation as a result of plant exposure to O3 were extensively described in
20 the 2006 O3 AQCD, some recent studies (described in Section 9.4.5.1) of O3-induced declines in
21 photosynthesis have focused also on O3 effects on the light reactions. Declines in the Fv/Fm ratio (a
22 measure of the maximum efficiency of the light reactions of photosynthesis) were observed in
23 several studies using a variety of plant species and exposure conditions. Additionally, O3 increased
24 the coefficient of non-photochemical quenching in several species, an indication that defense and
25 repair mechanisms of a non-photochemical nature are activated in these plants while less absorbed
26 light is being used to drive photosynthesis. This indicates a shift away from photosynthesis to
27 defense, resulting in negative impacts on growth and reproduction.
28 Section 9.4.5.2 evaluates the effects of O3 on respiration. While C assimilation declines in O3
29 exposed plants, respiration is generally up-regulated. These increases in respiration are thought to
30 result from a plant's greater energy needs for defense (maintaining its antioxidant metabolites in a
31 reduced state) and repair. The increased energy needs will negatively impact plant growth and
32 reproduction.
33 Secondary metabolism is most often up-regulated in a variety of species exposed to either
34 acute or chronic O3 exposures as a part of a generalized plant defense mechanism. Changes in gene
35 expression, quantity and activity of enzymes associated with secondary metabolism and alterations
36 in secondary metabolite quantity have been documented in plants exposed to O3. Some secondary
37 metabolites, such as flavonoids and polyamines, are of particular interest as they are known to have
38 antioxidant properties. Investigations on the importance of isoprenes in plant response to O3 have
39 revealed conflicting results; however, there is some evidence to suggest that they may play a
March 2011 9-7 DRAFT - DO NOT CITE OR QUOTE
-------
1 protective role. In summary, secondary metabolites increase in quantity in O3-treated plants as part of
2 a generalized plant defense response. Some secondary metabolites are of particular importance in
3 Os-treated plants as they may have antioxidant functions. Increased synthesis of secondary
4 metabolites represents a large energy investment of the plant into defense responses and away from
5 growth and reproduction.
6 Section 9.4.6 focuses on O3-induced changes in stomatal function. Stomata play a critical role
7 in limiting O3 uptake into the plant by reducing stomatal aperture. Declines in stomatal conductance
8 in response to O3 have been documented for many plant species, and much evidence suggests that
9 this results from increases in intercellular CO2 concentration due to reductions in C fixation.
10 Additionally, sensitivity of some plants to O3 has been related to a sluggish stomatal response, in
11 which plants are unable to close their stomata rapidly in response to O3. To summarize, stomatal
12 response to O3 can help to determine plant sensitivity to the pollutant, and the decreases in stomatal
13 conductance are thought to be related to declines in C fixation rates. Reduced stomatal conductance
14 will decrease rates of C assimilation and lead to diminished growth and reproduction in plants.
9.2.3. Nature of Effects on Vegetation
15 Ambient O3 concentrations have long been known to cause visible foliar injury, decreases in
16 photosynthetic rate, decreases in growth, and decreases in the quality and yield of some plant species
17 (U.S. EPA, 2006, 088089KU.S. EPA, 1996, 080827)01.5. EPA, 1984, 029711)OJ.S. EPA, 1978,
18 040586). Numerous studies have related O3 exposure to plant responses, with most research effort
19 focused on the growth of tree seedlings and the yield of crops as endpoints. The response of a plant
20 species to O3 exposure depends upon many factors, including genetic characteristics, biochemical
21 and physiological status, and previous and current exposure to other stressors. The associated
22 sections in Section 9.5 focus mainly on studies published since the release of the 2006 O3 AQCD
23 (U.S. EPA, 2006, 088089). However, because much O3 research was conducted prior to the 2006 O3
24 AQCD, the conclusions presented below are collectively based on this ISA as well as the 1978,
25 1986, 1996, and 2006 AQCDs (U.S. EPA, 2006, 088089)(U.S. EPA, 1996, 080827)01.5. EPA, 1984,
26 0297_li)(U.S. EPA, 1978, 040586).
9.2.3.1. Effects on Woody and Herbaceous Vegetation
Growth and Biomass Allocation
27 The previous O3 AQCDs concluded that there is strong and consistent evidence that ambient
28 concentrations of O3 decrease growth in numerous plant species across the U.S. Studies published
29 since the last review continue to support that conclusion (Section 9.5.2.1).
30 In a recently published meta-analysis of 263 studies, Wittig et al. (2009, 191631) reported that
31 current ambient O3 concentrations (~40 ppb) significantly decreased annual total biomass growth of
32 forest species by an average of 7%, with potentially greater decreases (11-17%) in areas that have
March 2011 9-8 DRAFT - DO NOT CITE OR QUOTE
-------
1 higher O3 concentrations and as background O3 increases in the future. This meta-analysis
2 demonstrates the coherence of O3 effects across numerous studies and species using a variety of
3 experimental techniques. In a study conducted on mature forest trees, McLaughlin et al. (2007,
4 090348) reported that the cumulative effects of ambient levels of O3 decreased seasonal stem growth
5 by 30-50% for most of the species in a high O3 year in comparison to a low O3 year.
6 Since the 2006 O3 AQCD, several studies were published based on the Aspen free-air carbon-
7 dioxide/ozone enrichment (FACE) experiment using "free air", O3, and CO2 exposures in a forest in
8 Wisconsin. It was found that O3 caused reductions in total biomass relative to the control in aspen,
9 paper birch, and sugar maple communities during the first seven years of stand development.
10 Overall, the studies at the Aspen FACE experiment were consistent with many of the open-top
11 chamber (OTC) studies that were the foundation of previous O3 NAAQS reviews. These results
12 strengthen our understanding of O3 effects on forests and demonstrate the relevance of the
13 knowledge gained from trees grown in open-top chamber studies.
14 In recent studies, O3 was shown to have either negative, non-significant, or positive effects on
15 root biomass and root:shoot ratio. While the findings of individual studies were mixed, recent meta-
16 analyses have generally indicated that O3 reduced C allocated to roots (Grantz et al., 2006,
17 191545)(Wittig et al., 2009, 191631).
18 For some annual species, particularly crops, the endpoint for an assessment of the risk of O3
19 exposure can be defined as yield or growth, e.g., production of grain. For plants grown in mixtures
20 such as hayfields, and natural or semi-natural grasslands (including native nonagricultural species),
21 endpoints other than production of biomass may be important. Such endpoints include biodiversity
22 or species composition, and measures of plant quality. Effects may also result from competitive
23 interactions among plants in mixed-species communities. Most of the available data on non-crop
24 herbaceous species are for grasslands with many of the recent studies conducted in Europe.
25 Evidence is sufficient to conclude that there is a causal relationship between O3 exposure
26 and reduced growth of woody and herbaceous vegetation.
Reproduction
27 Studies during recent decades have demonstrated O3 effects on different stages of plant
28 reproduction (Section 9.5.2.2). Several recent studies published since the 2006 O3 AQCD further
29 demonstrate the effects of O3 on reproductive processes in herbaceous and woody plant species.
30 The impacts of O3 on reproductive development can occur by influencing (1) age at time of
31 initial flowering, particularly in long-lived trees that often have long juvenile periods of early growth
32 without flower and seed production; (2) flower bud initiation and development; (3) pollen
33 germination and pollen tube growth; and (4) seed, fruit, or cone yields and seed quality.
34 Evidence is sufficient to conclude that there is a causal relationship between O3 exposure
35 and changes in reproduction of vegetation.
March 2011 9-9 DRAFT - DO NOT CITE OR QUOTE
-------
Visible Foliar Injury
1 Visible foliar injury resulting from exposure to O3 has been well characterized and
2 documented over several decades of research on many tree, shrub, herbaceous, and crop species
3 (U.S. EPA, 2006, 088089KU.S. EPA, 1996, 080827KU.S. EPA, 1984, 029711KU.S. EPA, 1978,
4 040586)(Section 9.5.2.3). Ozone-induced visible foliar injury symptoms on certain bioindicator plant
5 species are considered diagnostic as they have been verified experimentally in exposure-response
6 studies, using exposure methodologies such as continuous stirred tank reactors (CSTRs), OTCs, and
7 free-air fumigation. Experimental evidence has clearly established a consistent association of visible
8 injury with O3 exposure, with greater exposure often resulting in greater and more prevalent injury.
9 Since the 2006 O3 AQCD, several multiple-year field surveys of O3-induced visible foliar injury
10 have been conducted at National Wildlife Refuges in Maine, Michigan, New Jersey, and South
11 Carolina. New sensitive species showing visible foliar injury continue to be identified from field
12 surveys and verified in controlled exposure studies.
13 The use of biological indicators in field surveys to detect phytotoxic levels of O3 is a
14 longstanding and effective methodology. The USD A Forest Service through the Forest Health
15 Monitoring (FHM) Program (1990-2001) and currently the Forest Inventory and Analysis (FIA)
16 Program has been collecting data regarding the incidence and severity of visible foliar injury on a
17 variety of O3 sensitive plant species throughout the U.S. The network has provided evidence that O3
18 concentrations were high enough to induce visible symptoms on sensitive vegetation. From repeated
19 observations and measurements made over a number of years, specific patterns of areas experiencing
20 visible O3 injury symptoms can be identified.
21 In addition, a study by Kohut (2007, 093289) assessed the risk of O3-induced visible foliar
22 injury on bioindicator plants (NPS, 2006, 677536) in 244 national parks in support of the National
23 Park Service's Vital Signs Monitoring Network (NPS, 2007, 677537). Kohut (2007, 093289)
24 concluded that the risk of visible foliar injury was high in 65 parks (27%), moderate in 46 parks
25 (19%), and low in 131 parks (54%). Some of the well-known parks with a high risk of O3-induced
26 visible foliar injury include Gettysburg, Valley Forge, Delaware Water Gap, Cape Cod, Fire Island,
27 Antietam, Harpers Ferry, Manassas, Wolf Trap Farm Park, Mammoth Cave, Shiloh, Sleeping Bear
28 Dunes, Great Smoky Mountains, Joshua Tree, Sequoia and Kings Canyon, and Yosemite.
29 Evidence is sufficient to conclude that there is a causal relationship between ambient O3
30 exposure and the occurrence of O3-induced visible foliar injury on sensitive vegetation across
31 the U.S.
Gas Exchange
32 There is strong experimental evidence over several decades of research that exposure to O3
33 reduces photosynthesis and alters stomatal conductance in a wide variety of plant species. The mode
34 of action, as characterized in Section 9.4 and in previous reviews, provides biological plausibility for
35 O3 effects on leaf gas exchange.
March 2011 9-10 DRAFT - DO NOT CITE OR QUOTE
-------
1 In compiling more than 55 studies, Wittig et al. (2007, 191695) reported that current O3
2 concentrations in the northern hemisphere are decreasing photosynthesis (11%) and stomatal
3 conductance (13%) across tree species. It was also found that younger trees (<4 year) were affected
4 less by O3 than older trees. Further, the authors also found that decreases in photosynthesis are
5 consistent with the cumulative uptake of O3 into the leaf. In contrast, several studies reported that O3
6 exposure may result in loss of stomatal control, incomplete stomatal closure at night and a
7 decoupling of photosynthesis and stomatal conductance, which may have implications for whole-
8 plant water use (Section 9.6.3).
9 Evidence is sufficient to conclude that there is a causal relationship between O3 exposure
10 and the alteration of leaf gas exchange in vegetation.
9.2.3.2. Agricultural Crops
Yield and Crop Quality
11 The detrimental effect of O3 on crop production has been recognized since the 1960's and a
12 large body of research has subsequently stemmed from those initial findings. Previous O3 AQCDs
13 have extensively reviewed this body of literature (U.S. EPA, 2006, 088089). Recent experimental
14 studies of O3 effects on crops are discussed in Section 9.5.3 and summarized in Table 9-3 and 9-16
15 Current O3 concentrations across the U.S. are high enough to cause yield loss for a variety of
16 agricultural crops including, but not limited to, soybean, wheat, cotton, potato, watermelon, beans,
17 turnip, onion, lettuce, and tomato. Continued increases in O3 concentration may further decrease
18 yield in these sensitive crops while also initiating yield losses in less sensitive crops. Despite the
19 well-documented yield losses due to increasing O3 concentration, there is still a knowledge gap
20 pertaining to the exact mechanism of O3-induced yield loss. Research has linked increasing O3
21 concentration to decreased photo synthetic rates and accelerated senescence, which are related to
22 yield.
23 Recent modeling research has correlated satellite air-column observations with direct air-
24 sampling O3 data and modeled the yield-loss due to O3 over the continuous tri-state area of Illinois,
25 Iowa and Wisconsin. This modeling data correlates well with the previous results from FACE-type
26 experiments and OTC experiments.
27 New research is beginning to consider the mechanism of damage caused by long, lower O3
28 concentration (so-called chronic exposure) compared to short, very high O3 concentration (so-called
29 acute exposure). Both types of O3 exposure cause damage to agricultural crops, but through very
30 different mechanisms. Until recently, most research on the mechanism of O3 damage has used acute
31 exposure studies. It has become clear that the same cellular and biochemical processes involved in
32 the response to acute O3 exposure are not involved in response to chronic O3 exposure, yet both
33 cause yield-loss in agriculturally important crops.
34 In addition, new research has highlighted the effects of O3 on crop quality. Increasing O3
35 concentration decreases nutritive quality of grasses, decreases macro- and micro-nutrient
March 2011 9-11 DRAFT-DO NOT CITE OR QUOTE
-------
1 concentrations in fruits and vegetable crops, and decreases cotton fiber quality. These areas of
2 research require further investigation to determine the mechanism and dose-responses.
3 Evidence is sufficient to conclude that there is a causal relationship between O3 exposure
4 and reduced yield and quality of agricultural crops.
9.2.3.3. Factors That Modify Functional and Growth Response
5 Many biotic and abiotic factors, including insects, pathogens, root microbes and fungi,
6 temperature, water and nutrient availability, and other air pollutants, as well as elevated CO2,
7 influence or alter plant response to O3. These modifying factors were comprehensively reviewed in
8 the 2006 O3 AQCD. A limited number of studies published since 2006 provide further support for
9 our understanding of the role of these interactions in modifying O3-induced plant responses and are
10 discussed in Section 9.5.4.
9.2.4. Ecosystems and Services
11 Ozone has been found to alter plant physiological processes such as growth, biomass
12 allocation, reproduction and gas exchange (Section 9.5). Those O3-induced effects at the individual
13 plant scale have the potential to translate to effects at the ecosystem level, and cause changes in
14 biogeochemical cycling and community composition. Information presented in the associated section
15 (Section 9.6) was collected at multiple scales, ranging from responses at the population level to the
16 ecosystem level. The effects of O3 on ecosystem productivity, C sequestration, water cycling,
17 nutrient cycling, and community composition are reviewed.
9.2.4.1. Productivity and Carbon Sequestration
18 During the previous NAAQS reviews, there were very few studies that investigated the effect
19 of O3 exposure on ecosystem productivity and C sequestration. Recent studies from long-term FACE
20 experiments provided evidence of the association of O3 exposure and reduced productivity at the
21 ecosystem level. Elevated O3 reduced stand-level biomass by 13-23% at Aspen FACE after 7 years
22 of O3 exposure (King et al., 2005, 191701). and annual volume growth by 9.5 m3/ha at the Kranzberg
23 Forest (Germany) FACE (Pretzsch et al., 2010, 580435). Studies at the leaf and plant scales showed
24 that O3 reduced photosynthesis and plant growth, which provided coherence and biological
25 plausibility for the decrease in ecosystem productivity. Results across different ecosystem models
26 were consistent with the FACE experimental evidence, which showed that O3 reduced ecosystem
27 productivity.
28 Although O3 generally causes negative effects on plant growth, the magnitude of the response
29 varies among plant communities. For example, O3 had little impact on white fir, but greatly reduced
30 growth of ponderosa pine in southern California (Weinstein et al., 2005, 179965). Ozone decreased
31 net primary production (NPP) of most forest types by 7-8% in Mid-Atlantic region, but had small
32 impacts on spruce-fir forest, which was decreased by only 1% (Pan et al., 2009, 596032). Among
March 2011 9-12 DRAFT - DO NOT CITE OR QUOTE
-------
1 crop species, the estimated yield loss for wheat (7-12%) and soybean (6-16%) were higher than rice
2 (3-4%) and maize (3-5%) (Van Dingenen et al, 2009, 199765V
3 In addition to plant growth, other indicators that are typically estimated by model studies
4 include net ecosystem CO2 exchange (NEE), C sequestration, and crop yield. Model simulations
5 consistently found that O3 exposure caused negative impacts on those indicators (Section 9.6.2,
6 Table 9-5), but the severity of these impacts was influenced by multiple interactions of biological
7 and environmental factors. For example, the largest O3-induced crop yield losses occurred in high-
8 production areas exposed to high O3 concentrations, such the Midwest and the Mississippi Valley
9 regions of the U.S. (Van Dingenen et al., 2009, 199765V
10 The suppression of ecosystem C sinks results in more CO2 accumulation in the atmosphere.
11 Globally, the indirect radiative forcing, reported in Watts/square meter (W/m2), caused by O3
12 exposure through lowering ecosystem C sink (0.62-1.09 W/m2) could have an even greater impact
13 on global warming than the direct radiative forcing of O3 (0.89 W/m2) (Sitch et al., 2007, 093294V
14 Ozone could also affect regional C budgets through interacting with multiple factors, such as N
15 deposition, elevated CO2 and land use history. Model simulations suggested that O3 partially offset
16 the growth stimulation caused by elevated CO2 and N deposition in both Northeast- and Mid-
17 Atlantic-region forest ecosystems of the U.S. (Ollinger et al., 2002, 180189)(Pan et al., 2009,
18 596032V
19 The evidence is sufficient to infer that there is a causal relationship between O3 exposure
20 and reduced productivity, and a likely causal relationship between O3 exposure and reduced
21 carbon sequestration in terrestrial ecosystems.
9.2.4.2. Water Cycling
22 Although the evidence was from a limited number of field and modeling studies, these
23 findings showed an association of O3 exposure and the alteration of water cycle at the ecosystem
24 level. Field studies conducted by McLaughlin et al. (2007, 090348K2007. 090347) suggested that
25 peak hourly O3 exposure increased the rate of water loss from several tree species, and led to a
26 reduction in the late-season modeled stream flow in three forested watersheds in eastern Tennessee.
27 Evidence of sluggish stomatal responses during O3 exposure was found in their study and several
28 other studies (Section 9.6.3), which provided biological plausibility for the observed higher water
29 loss at the ecosystem level. However, many experiments, mostly based on short-term O3 exposure,
30 found that O3 generally reduced stomatal conductance. The O3-induced reduction in stomatal
31 aperture is the biological assumption for most process-based models. Therefore, results of those
32 models normally found that O3 reduced water loss. For example, Felzer (2009, 191460) found that
33 O3 damage and N limitation together reduced evapotranspiration and increase runoff.
34 Although the direction of the response differed among studies, the evidence is sufficient to
35 conclude that there is likely to be a causal relationship between O3 exposure and the alteration
36 of ecosystem water production.
March 2011 9-13 DRAFT - DO NOT CITE OR QUOTE
-------
9.2.4.3. Below-Ground Processes
1 Since the 2006 O3 AQCD, more evidence has shown that although the responses are often
2 species specific, O3 altered the quality and quantity of C input to soil, microbial community
3 composition, and C and nutrient cycling. Biogeochemical cycling of below-ground processes is
4 driven by C input from plants. Studies at the leaf and plant level have provided biologically plausible
5 mechanisms, such as reduced photosynthetic rates, increased metabolic cost, and reduced root C
6 allocation (Section 9.6.4) for the association of O3 exposure and the alteration of below-ground
7 processes.
8 Results from Aspen FACE and other experimental studies consistently found that O3 reduced
9 litter production and altered C chemistry, such as soluble sugars, soluble phenolics, condensed
10 tannins, lignin, and macro/micro nutrient concentration in litter (Liu et al., 2005, 187005)(Parsons et
11 al., 2008, 191853)(Kasurinen et al., 2006, 191269). The changes in substrate quality and quantity
12 could alter microbial metabolism under elevated O3, and therefore soil C and nutrient cycling.
13 Several studies indicated that O3 generally suppressed soil enzyme activities (Chung et al., 2006,
14 191729)(Pritsch et al., 2009, 626808). However, the impact of O3 on litter decomposition was
15 inconsistent and varied among species, sites and exposure length. Ozone had small impact on
16 dynamics of micro and macro nutrients, except for N. Ozone was found to reduce N release from
17 leaf litter and decrease gross N mineralization, which could potentially decrease N availability to
18 plants (Holmes et al., 2006, 191372¥Liu et al., 2007, 093286V
19 Studies from the Aspen FACE experiment suggested that the response of below-ground
20 C cycle to O3 exposure, such as litter decomposition, soil respiration and soil C content, changed
21 over time. For example, in the early part of the experiment (1998-2003), O3 had no impact on soil
22 respiration but reduced the formation rates of total soil C under elevated CO2. However, after 10-
23 11 yr of exposure, O3 was found to increase soil respiration but have no significant impact on soil
24 C formation under elevated CO2 (Section 9.6.4.3).
25 The evidence is sufficient to infer that there is a causal relationship between O3 exposure
26 and the alteration of below-ground biogeochemical cycles.
9.2.4.4. Community Composition
27 In the 2006 O3 AQCD, the impact of O3 exposure on species competition and community
28 composition was assessed. Ozone was found to cause a significant decline in ponderosa and Jeffrey
29 pine in the San Bernardino Mountains in southern California. Ozone exposure also tended to shift
30 the grass-legume mixtures in favor of grass species (U.S. EPA, 2006, 088089). Since the 2006 O3
31 AQCD, more evidence has shown that O3 exposure changed the competitive interactions and led to
32 loss of O3 sensitive species or genotypes. Studies at plant level found that the severity of O3 damage
33 on growth, reproduction and foliar injury varied among species (Section 9.6.5), which provided the
34 biological plausibility for the alteration of community composition. Additionally, research since the
March 2011 9-14 DRAFT - DO NOT CITE OR QUOTE
-------
1 last review has shown that O3 can alter community composition and diversity of soil microbial
2 communities.
3 The decline of conifer forests under O3 exposure was continually observed in several regions.
4 Ozone damage was believed to be an important causal factor in the dramatic decline of sacred fir in
5 the valley of Mexico (de Lourdes de Bauer and Hernandez-Tejeda, 2007, 196891). as well as
6 cembran pine in southern France and Carpathian Mountains (Wieser et al., 2006, 191391). Results
7 from the Aspen FACE site indicated that O3 could alter community composition of broadleaf forests
8 as well. At the Aspen FACE site, O3 reduced growth and increased mortality of a sensitive aspen
9 clone, while the O3 tolerant clone emerged as the dominant clone in the pure aspen community. In
10 the mixed aspen-birch and aspen-maple communities, O3 reduced the competitive capacity of aspen
11 compared to birch and maple (Kubiske et al., 2007, 191336).
12 The tendency for O3-exposure to shift the biomass of grass-legume mixtures in favor of grass
13 species, was reported in the 2006 O3 AQCD (U.S. EPA, 2006, 088089) and has been generally
14 confirmed by recent studies. However, in a high elevation mature/species-rich grass-legume pasture,
15 O3 fumigation showed no significant impact on community composition (Bassin et al., 2007,
16 191534).
17 Ozone exposure not only altered community composition of plant species, but also
18 microorganisms. The shift in community composition of bacteria and fungi has been observed in
19 both natural and agricultural ecosystems, although no general patterns could be identified (Kanerva
20 et al., 2008, 191264)(Morskv et al., 2008, 191507)(Kasurinen et al., 2005, 191245).
21 The evidence is sufficient to conclude that there is likely a causal relationship between O3
22 exposure and the alteration of community composition.
9.2.5. Air Quality Indices
23 Exposure indices are metrics that quantify exposure as it relates to measured plant damage
24 (i.e., reduced growth). They are summary measures of monitored ambient O3 concentrations over
25 time intended to provide a consistent metric for reviewing and comparing exposure-response effects
26 obtained from various studies. No new information is available since 2006 that alters the basic
27 conclusions put forth in the 2006 and 1996 O3 AQCDs (U.S. EPA, 1996, 080827)(U.S. EPA, 2006,
28 088089). These AQCDs focused on the research used to develop various exposure indices to help
29 quantify effects on growth and yield in crops, perennials, and trees (primarily seedlings). The
30 performance of indices was compared through regression analyses of earlier studies designed to
31 support the estimation of predictive O3 exposure-response models for growth and/or yield of crops
32 and tree (seedling) species.
33 The main conclusions from the 1996 and 2006 O3 AQCDs (U.S. EPA, 1996,
34 080827)(U.S. EPA, 2006, 088089) regarding an index based on ambient exposure are still valid.
35 These key conclusions can be restated as follows:
36 • O3 effects in plants are cumulative;
March 2011 9-15 DRAFT - DO NOT CITE OR QUOTE
-------
1 • higher O3 concentrations appear to be more important than lower concentrations in
2 eliciting a response;
3 • plant sensitivity to O3 varies with time of day and plant development stage; and
4 • exposure indices that cumulate hourly O3 concentrations and preferentially weight the
5 higher concentrations have better statistical fits to growth/yield response data than do the
6 mean and peak indices.
7 Various weighting functions have been used, including threshold-weighted (e.g., SUM06) and
8 continuous sigmoid-weighted (e.g., W126) functions. Based on statistical goodness-of-fit tests, these
9 cumulative, concentration-weighted indices could not be differentiated from one another using data
10 from previous exposure studies. Additional statistical forms for O3 exposure indices have been
11 discussed in Lee et al. (1988, 042136) and in Section 9.7 of this ISA. The majority of studies
12 published since the 2006 O3 AQCD (2006, 088089) do not change earlier conclusions, including the
13 importance of peak concentrations, and the duration and occurrence of O3 exposures in altering plant
14 growth and yield.
15 Given the current state of knowledge and the best available data, exposure indices that
16 cumulate and differentially weight the higher hourly average concentrations and also include the
17 mid-level values continue to offer the most defensible approach for use in developing response
18 functions and comparing studies, as well as for defining future indices for vegetation protection.
9.2.5.1. Modeled Ozone Deposition or "Flux"
19 Another approach for improving risk assessment of vegetation response to ambient O3 is based
20 on determining the O3 concentration from the atmosphere that enters the leaf (i.e., flux or
21 deposition). Interest has been increasing in recent years, particularly in Europe, in using
22 mathematically tractable flux models for O3 assessments at the regional, national, and European
23 scale (Matyssek et al., 2008, 191262)(Paoletti and Manning, 2007, 180174)(Emberson et al., 2000,
24 040350)(Emberson et al., 2000, 042537)(ICP M&M, 2004, 677471). While some efforts have been
25 made in the U.S. to calculate O3 flux into leaves and canopies (Grantz et al., 1997, 026664)(Grantz
26 et al., 1995, 026659)(Grulke et al., 2004, 042646)(Turnipseed et al., 2009, 5887_52)(Uddling et al.,
27 2009, 5962_19)(Bergweiler et al., 2008, 191656)(Hogg et al., 2007, 199349). little information has
28 been published relating these fluxes to effects on vegetation. There is also concern that not all O3
29 stomatal uptake results in a yield reduction, which depends to some degree on the amount of internal
30 detoxification occurring with each particular species. Those species having high amounts of
31 detoxification potential may, in fact, show little relationship between O3 stomatal uptake and plant
32 response (Musselman and Massman, 1999, 040706). The lack of data in the U.S. and the lack of
33 understanding of detoxification processes have made this technique less viable for vulnerability and
34 risk assessments in the U.S.
March 2011 9-16 DRAFT - DO NOT CITE OR QUOTE
-------
9.2.5.2. Night-Time Exposures
1 A 12-h (8:00 a.m. to 8:00 p.m.) daylight period for a W126 cumulating exposure was proposed
2 in 2007 and 2009 following the release of the 2006 O3 AQCD and was based primarily on evidence
3 that the conditions for uptake of O3 into the plant occur mainly during the daytime hours. Plants have
4 the highest stomatal conductance during the daytime and atmospheric turbulent mixing is greatest
5 then as well (U.S. EPA, 2006, 088089)(Uddling et al., 2010, 387073). Recent reviews of the
6 literature reported that a large number of species had varying degrees of nocturnal stomatal
7 conductance (Caird et al., 2007, 199337)(Dawson et al., 2007, 67038 l)(Musselman and Minnick,
8 2000, 011612). In general, stomatal conductance at night is at a much lower rate compared to
9 daytime conductance (Caird et al., 2007, 199337). For significant nocturnal stomatal flux and O3
10 effects to occur, specific conditions must exist. A susceptible plant with nocturnal stomatal
11 conductance and low defense must be growing in an area with relatively high night-time O3 and
12 appreciable nocturnal turbulence. It is unclear how many areas there are in the U.S. where these
13 conditions occur. More information is needed in these locations in order to assess the local O3
14 patterns, micrometeorology and responses of potentially vulnerable plant species.
9.2.6. Exposure-Response
15 None of the information on effects of O3 on vegetation published since the 2006 O3 AQCD has
16 modified the assessment of quantitative exposure-response relationships that was presented in that
17 document (U.S. EPA, 2006, 088089). This assessment updates the 2006 exposure-response models
18 by computing them using the W126 metric, cumulated over 90 days. Almost all of the experimental
19 research on the effects of O3 on growth or yield of plants published since 2006 used only two levels
20 of exposure. In addition, hourly O3 concentration data that would allow calculations of exposure
21 using the W126 scale are generally unavailable. However, two long-term experiments, one with a
22 crop species (soybean), one with a tree species (aspen), have produced data that can be used to
23 validate the exposure-response models presented in the 2006 O3 AQCD, and methodology used to
24 derive them.
25 Quantitative characterization of exposure-response in the 2006 O3 AQCD was based on
26 experimental data generated for that purpose by the National Crop Loss Assessment Network
27 (NCLAN) and EPA National Health and Environmental Effects Research Laboratory, Western
28 Ecology Division (NHEERL-WED) projects, using OTCs to expose crops and trees seedling to O3.
29 In recent years, yield and growth results for two of the species that had provided extensive exposure-
30 response information in those projects have become available from studies that used FACE
31 technology, which is intended to provide conditions much closer to natural environments (Dickson et
32 al., 2000, 628220: Morgan et al., 2004, 072764: Morgan et al., 2006, 079186: Pregitzer et al., 2008,
33 191677). The robust methods that were used previously with exposure measured as SUM06 were
34 applied to the NCLAN and NHEERL-WED data with exposure measured as W126, in order to
35 derive single-species median models for soybean and aspen from studies involving different
March 2011 9-17 DRAFT - DO NOT CITE OR QUOTE
-------
1 genotypes, years, and locations. The resulting models were used to predict the change in yield of
2 soybean and biomass of aspen between the two levels of exposure reported in current FACE
3 experiments. Results from these new experiments were exceptionally close to predictions from the
4 models. The accuracy of model predictions for two widely different plant species provides support
5 for the validity of the corresponding multiple-species models for crops and trees in the NCLAN and
6 NHEERL-WED projects. However, variability among species in those projects indicates that the
7 range of sensitivity is likely quite wide. This was confirmed by a recent experiment with cottonwood
8 in a naturally occurring gradient of exposure (Gregg et al., 2006, 186961). which established the
9 occurrence of species with responses substantially more severe under currently existing conditions
10 than are predicted by the median model for multiple species.
11 Results from several meta-analyses have provided approximate values for responses of yield
12 of soybean, wheat, rice and other crops under broad categories of exposure, relative to charcoal -
13 filtered air (Ainsworth, 2008, 191646: Feng et al., 2008, 191453: Morgan et al., 2003, 055527).
14 Likewise, Feng and Kobayashi (2009, 199223) have summarized yield data for six crop species
15 under various broad comparative exposure categories, while Wittig et al. (2009, 191631) reviewed
16 263 studies that reported effects on tree biomass. However, these analyses have proved difficult to
17 compare with exposure-response models, especially given that exposure was not expressed on the
18 same W126 scale.
9.3. Experimental Exposure Methodologies
9.3.1. Introduction
19 A variety of methods for studying plant response to O3 exposures have been developed over
20 the last several decades. Methodological advancements since 2006 have not fundamentally altered
21 our understanding of O3 effects on plants or ecosystems. The majority of methodologies currently
22 used have been discussed in detail in the 1996 O3 AQCD (Section 5.2; U.S. EPA, 1996, 080828) and
23 2006 O3 AQCD (Section AX9.1; U.S. EPA, 2006, 088089). This section will serve as a short
24 overview of the methodologies and the reader is referred to the previous O3 AQCDs for more in-
25 depth discussion.
9.3.2. "Indoor," Controlled Environment, and Greenhouse Chambers
26 The earliest experimental investigations of the effects of O3 on plants utilized simple glass or
27 plastic-covered chambers, often located within greenhouses, into which a flow of O3-enriched air or
28 oxygen could be passed to provide the exposure. The types, shapes, styles, materials of construction,
29 and locations of these chambers have been numerous. Hogsett et al. (1987, 043465) have
30 summarized the construction and performance of more elaborate and better instrumented chambers
31 since the 1960s, including those installed in greenhouses (with or without some control of
32 temperature and light intensity).
March 2011 9-18 DRAFT - DO NOT CITE OR QUOTE
-------
1 One greenhouse chamber approach that continues to yield useful information on the
2 relationships of O3 uptake to both physiological and growth effects employs continuous stirred tank
3 reactors (CSTRs) first described by Heck et al. (1978, 037673). Although originally developed to
4 permit mass-balance studies of O3 flux to plants, their use has more recently widened to include
5 short-term physiological and growth studies of O3 * CO2 interactions (Heagle et al., 1994,
6 Q26699)(Loats and Rebbeck, 1999, Q29709)(Rao et al., 1995, 030221)(Reinert and Ho, 1995,
7 030247)(Reinert et al., 1997, 030252). and validation of visible foliar injury on a variety of plant
8 species (Kline et al., 2009, 196918)(Orendovici et al., 2003, 049080). In many cases, supplementary
9 lighting and temperature control of the surrounding structure have been used to control or modify the
10 environmental conditions (Heagle et al., 1994, 026699).
11 Many investigations have utilized commercially available controlled environment chambers
12 and walk-in rooms adapted to permit the introduction of a flow of O3 into the controlled air-volume.
13 Such chambers continue to find use in genetic screening and in physiological and biochemical
14 studies aimed primarily at improving our understanding of modes of action. For example, some of
15 the studies of the O3 responses of common plantain (Plantago major) populations have been
16 conducted in controlled environment chambers (Reiling and Davison, 1994, 035373)(Whitfield et
17 al.. 1996. 055544).
18 More recently, some researchers have been interested in direct O3 effects on reproductive
19 processes, separate from the effects on vegetative processes (Black et al., 2010, 625575). For this
20 purpose, controlled exposure systems have been employed to expose the reproductive structures of
21 annual plants to gaseous pollutants independently of the vegetative component (Black et al., 2010,
22 625575)(Stewart et al., 1996, 036384).
9.3.3. Field Chambers
23 In general, field chamber studies are dominated by the use of various versions of the open top
24 chamber (OTC) design, first described by Heagle et al. (1973, 038348) and Mandl et al. (1973,
25 039967). The OTC method continues to be a widely used technique in the U.S. and Europe for
26 exposing plants to varying levels of O3. Most of the new information confirms earlier conclusions
27 and provides additional support for OTC use in assessing plant species and in developing exposure-
28 response relationships. Chambers are generally ~3 m in diameter with 2.5-m-high walls. Hogsett et
29 al. (1987, 043780) described in detail many of the various modifications to the original OTC designs
30 that appeared subsequently, e.g., the use of larger chambers for exposing small trees (Kats et al.,
31 1985, 055511) or grapevines (Mandl et al., 1989, 043987). the addition of a conical baffle at the top
32 to improve ventilation (Kats et al., 1976, 039799). a frustum at the top to reduce ambient air
33 incursions, and a plastic rain-cap to exclude precipitation (Hogsett et al., 1985, 039383). All versions
34 of OTCs included the discharge of air via ports in annular ducting or interiorly perforated double-
35 layered walls at the base of the chambers to provide turbulent mixing and the upward mass flow of
36 air.
March 2011 9-19 DRAFT - DO NOT CITE OR QUOTE
-------
1 Chambered systems, including OTCs, have several advantages. For instance, they can provide
2 a range of treatment levels including charcoal-filtered (CF), clean-air control, and several above
3 ambient concentrations for O3 experiments. Depending on experimental intent, a replicated, clean-air
4 control treatment is an essential component in many experimental designs. The OTC can provide a
5 consistent, definable exposure because of the constant wind speed and delivery systems. Statistically
6 robust concentration-response (C-R) functions can be developed using such systems for evaluating
7 the implications of various alternative air quality scenarios on vegetation response. Nonetheless,
8 there are several characteristics of the OTC design and operation that can lead to exposures that
9 might differ from those experienced by plants in the field. First, the OTC plants are subjected to
10 constant air flow turbulence, which, by lowering the boundary layer resistance to diffusion, may
11 result in increased uptake. This may lead to an overestimation of effects relative to areas with less
12 turbulence (Krupa et al., 1995, 038895)(Legge et al, 1995, 040689). Conversely, however, other
13 research has found that OTC's may slightly change vapor pressure deficit (VPD) in a way that may
14 decrease the uptake of O3 into leaves (Piikki et al., 2008, 199265). As with all methods that expose
15 vegetation to modified O3 concentrations in chambers, OTCs create internal environments that differ
16 from ambient air. This so-called "chamber effect" refers to the modification of microclimatic
17 variables, including reduced and uneven light intensity, uneven rainfall, constant wind speed,
18 reduced dew formation, and increased air temperatures (Fuhrer, 1994, 055549)(Manning and Krupa,
19 1992, 044155). However, in at least one case where canopy resistance was quantified in OTCs and in
20 the field, it was determined that gaseous pollutant exposure to crops in OTCs was similar to that
21 which would have occurred at the same concentration in the field (Unsworth et al., 1984,
22 041239)(Unsworth et al., 1984, 040024). Because of the standardized methodology and protocols
23 used in National Crop Loss Assessment Network (NCLAN) and other programs, the database can be
24 assumed to be internally consistent.
25 While it is clear that OTCs can alter some aspects of the microenvironment and plant growth,
26 it is important to establish whether or not these differences affect the relative response of a plant to
27 O3. As noted in the 1996 O3 AQCD (U.S. EPA, 1996, 080828). evidence from a number of
28 comparative studies of OTCs and other exposure systems suggested that responses were essentially
29 the same regardless of exposure system used and chamber effects did not significantly affect
30 response. For example, a study of chamber effects examined the responses of tolerant and sensitive
31 white clover clones (Trifolium repens) to ambient O3 in greenhouse, open top, and ambient plots
32 (Heagle et al., 1996, 042660). The response found in OTCs was the same as in ambient plots.
33 Another type of field chamber called a "terracosm" has been developed and used in recent
34 studies (Lee et al., 2009, 595904). Concern over the need to establish realistic plant-litter-soil
35 relationships as a prerequisite to studies of the effects of O3 and CO2 enrichment on ponderosa pine
36 (Pinus ponderosa) seedlings led Tingey et al. (1996, 055536) to develop closed, partially
37 environmentally controlled, sun-lit chambers ("terracosms") incorporating 1-m-deep lysimeters
38 containing forest soil in which the appropriate horizon structure was retained.
March 2011 9-20 DRAFT - DO NOT CITE OR QUOTE
-------
1 Other researchers have recently published studies using another type of out-door chamber
2 called recirculating Outdoor Plant Environment Chambers (OPECs) (Flowers et al., 2007, 191852).
3 These closed chambers are approximately 2.44 m* 1.52 m with a growth volume of approximately
4 3.7 m3 in each chamber. These chambers admit 90% of full sunlight and control temperature,
5 humidity and vapor pressure (Fiscus et al., 1999, 672964).
9.3.4. Plume and FACE-Type Systems
6 Plume systems are chamberless exposure facilities in which the atmosphere surrounding plants
7 in the field is modified by the injection of pollutant gas into the air above or around them from
8 multiple orifices spaced to permit diffusion and turbulence, so as to establish relatively homogeneous
9 conditions as the individual plumes disperse and mix with the ambient air. They can only be used to
10 increase the O3 levels in the ambient air.
11 The most common plume system used in the U.S. is a modification of the free-air carbon-
12 dioxide/ozone enrichment (FACE) system (Hendrey and Kimball, 1994, 040397)(Hendrey et al.,
13 1999, 042641). Although originally designed to provide chamberless field facilities for studying the
14 CO2 effects of climate change, FACE systems have been adapted to include the dispensing of O3
15 (Karnosky et al., 1999, 035307). This method has been employed in Illinois (SoyFACE) to study
16 soybeans (Morgan et al., 2004, 072764)(Rogers et al., 2004, 079201) and in Wisconsin (Aspen
17 FACE) to study trembling aspen (Populus tremuloides), birch (Betula papyrifera) and maple (Acer
18 saccharum)(Kamosky et al., 1999, 035307). Volk et al. (2003, 055568) also described a similar
19 system for exposing grasslands that uses 7-m diameter plots. FACE systems discharge the pollutant
20 gas (O3 and/or CO2) through orifices spaced along an annular ring (or torus) or at different heights on
21 a ring of vertical pipes. Computer-controlled feedback from the monitoring of gas concentration
22 regulates the feed rate of enriched air to the dispersion pipes. Feedback of wind speed and direction
23 information ensures that the discharges only occur upwind of the treatment plots, and that discharge
24 is restricted or closed down during periods of low wind speed or calm conditions. The diameter of
25 the arrays and their height (25-30 m) in some FACE systems requires large throughputs of enriched
26 air per plot, particularly in forest tree systems. The cost of the throughputs tends to limit the number
27 of enrichment treatments, although Hendrey et al. (1999, 042641) argued that the cost on an enriched
28 volume basis is comparable to that of chamber systems.
29 Although plume systems make virtually none of the modifications to the physical environment
30 that are inevitable with chambers, their successful use depends on selecting the appropriate numbers,
31 sizes, and orientations of the discharge orifices to avoid "hot-spots" resulting from the direct
32 impingement of jets of pollutant-enriched air on plant foliage (Werner and Fabian, 2002, 053040).
33 Because mixing is unassisted and completely dependent on wind turbulence and diffusion, local
34 gradients are inevitable especially in large-scale systems. FACE systems have provisions for shutting
35 down under low wind speed or calm conditions and for an experimental area that is usually defined
36 within a generous border in order to strive for homogeneity of the exposure concentrations within the
37 treatment area. They are also dependent upon continuous computer-controlled feedback of the O3
March 2011 9-21 DRAFT-DO NOT CITE OR QUOTE
-------
1 concentrations in the mixed treated air and of the meteorological conditions. Plume and FACE
2 systems also are unable to reduce O3 levels below ambient in areas where O3 concentrations are
3 phytotoxic.
9.3.5. Ambient Gradients
4 Ambient O3 gradients that occur in the U.S. hold potential for the examination of plant
5 responses over multiple levels of exposure that are occurring. However, few such gradients can be
6 found that meet the rigorous statistical requirements for comparable site characteristics such as soil
7 type, temperature, rainfall, radiation, and aspect (Manning and Krupa, 1992, 044155); although with
8 small plants, soil variability can be avoided by the use of plants in large pots. The use of soil
9 monoliths transported to various locations along natural O3 gradients is another possible approach to
10 overcome differences in soils; however, this approach is also limited to small plants.
11 Studies in the 1970s used the natural gradients occurring in southern California to assess yield
12 losses of alfalfa and tomato) (Oshima et al, 1976, 038475)(Oshima et al, 1977, 038938). Atransect
13 study of the impact of O3 on the growth of white clover and barley in the U.K. was confounded by
14 differences in the concurrent gradients of SO2 and NO2 pollution (Ashmore et al., 1988, 037038).
15 Studies of forest tree species in national parks in the eastern U.S. (Winner et al., 1989, 043403)
16 revealed increasing gradients of O3 and visible foliar injury with increased elevation.
17 Several studies have used the San Bernardino Mountains Gradient Study in southern
18 California to study the effects of O3 and N deposition on forests dominated by ponderosa and Jeffrey
19 pine (Arbaugh et al., 2003, 052925)(Miller and Elderman, 1977, 038488)(Grulke. 1999,
20 Q52983.)(Jones and Paine, 2006, 191301). However, it is difficult to separate the effects of N and O3
21 in some instances in these studies (Arbaugh et al., 2003, 052925). An O3 gradient in Wisconsin has
22 been used to study foliar injury in a series of trembling aspen clones (Populus tremuloides) differing
23 in O3 sensitivity (Karnosky et al., 1999, 035307)(lVIankovska et al., 2005, 672965).
24 More recently, some studies have been published that have used natural gradients to study a
25 variety of endpoints and species. For example, Gregg et al. (2003, 046996) studied cottonwood
26 saplings grown in an urban to rural gradient of O3 in the New York City area. The secondary nature
27 of the reactions of O3 formation and NOX titration reactions within the city center resulted in
28 significantly higher cumulative O3 exposures in the rural sites. The results of this gradient study were
29 similar to those of a parallel OTC study. Also, the U.S. forest service Forest Inventory and Analysis
30 (FIA) program uses large-scale O3 exposure patterns across the continental U.S. to study occurrences
31 of foliar injury due to O3 exposure (Smith et al., 2003, 044183)( Section 9.5.2.3). Finally,
32 McLaughlin et al. (2007, 090348)(2007. 090347) used spatial and temporal O3 gradients to study
33 forest growth and water use in the southern Appalachians. These studies found varying O3 exposures
34 between years and between sites.
March 2011 9-22 DRAFT - DO NOT CITE OR QUOTE
-------
9.3.6. Comparative Studies
1 All experimental approaches used to expose plants to O3 have shortcomings. The use of
2 laboratory, greenhouse, or field chambers raises concerns for the roles of chamber effects on
3 micrometeorology. In contrast, plume, FACE and gradient systems suffer from limited exposure
4 levels, few replicates and an inability to reduce O3 levels below ambient in areas where O3
5 concentrations are phytotoxic.
6 While it is clear that chambers can alter some aspects of plant growth, it is important to
7 establish whether or not these differences affect plant response to O3. As noted in the 1996 O3 AQCD
8 (U.S. EPA, 1996, 080828). evidence from the comparative studies of OTCs and from closed
9 chamber and O3-exclusion exposure systems on the growth of alfalfa (Medicago sativa) by Olszyk et
10 al. (1986, 055530) suggested that, since significant differences were found for fewer than 10% of the
11 growth parameters measured, the responses were, in general, essentially the same regardless of
12 exposure system used; and chamber effects did not significantly affect response. In 1988, Heagle et
13 al. (1988, 043559) concluded: "Although chamber effects on yield are common, there are no results
14 showing that this will result in a changed yield response to O3." A study of chamber effects examined
15 the responses of tolerant and sensitive white clover clones (Trifolium repens) to ambient O3 in
16 greenhouse, open-top, and ambient plots (Heagle et al., 1996, 042660). For individual harvests,
17 greenhouse O3 exposure reduced the forage weight of the sensitive clone 7 to 23% more than in
18 OTCs. However, the response in OTCs was the same as in ambient plots. Several studies have
19 shown very similar response of yield to O3 for plants grown in pots or in the ground, suggesting that
20 even such a significant change in environment does not alter the proportional response to O3, at least
21 as long as the plants are well watered (Heagle, 1979, Q39329)(Heagle et al., 1983, 039372).
22 A few recent studies have compared results of O3 experiments between OTCs, FACE
23 experiments, and gradient studies. For example, a series of studies undertaken at Aspen FACE
24 (Isebrands et al., 2000, 044174)(Isebrands et al., 2001, 036345) showed that O3-symptom expression
25 was generally similar in OTCs, FACE, and ambient-O3 gradient sites, and supported the previously
26 observed variation among trembling aspen clones using OTCs (Karnosky et al., 1999,
27 035307)(Mafikovska et al., 2005, 672965). In the SoyFACE experiment in Illinois, soybean (Pioneer
28 93B15 cultivar) yield loss data from a two-year study was published (Morgan et al., 2006, 079186).
29 This cultivar is a recent selection and, like most modern cultivars, has been selected under an already
30 high current O3 exposure. It was found to have average sensitivity to O3 compared to 22 other
31 cultivars tested at SoyFACE. In this experiment, ambient hourly O3 concentrations were increased by
32 approximately 20% and measured yields were decreased by 15% in 2002 as a result of the increased
33 O3 exposure (Morgan et al., 2006, 079186). To compare these results to chamber studies, Morgan et
34 al. (2006, 079186) calculated the expected yield loss from a linear relationship constructed from
35 chamber data using 7-h seasonal averages (Ashmore, 2002, 672967). They calculated an 8%
36 expected yield loss from the 2002 O3 exposure using that linear relationship. In another study, Gregg
37 et al. (2003, 046996)(2006. 186961) found similar O3 effects on cottonwood sapling biomass growth
March 2011 9-23 DRAFT - DO NOT CITE OR QUOTE
-------
1 and physiology along an ambient O3 gradient in the New York City area and a parallel OTC study.
2 Additionally, Section 9.8.3 of this document presents comparisons of exposure-response in from
3 OTC studies in trees and crops with results from more recent FACE experiments.
9.4. Mechanisms Governing Vegetation Response to
Ozone
9.4.1. Introduction
4 This section focuses on the effects of O3 stress on plants and their responses to that stress on
5 the molecular, biochemical and physiological levels. First, the pathway of O3 uptake into the leaf and
6 the initial chemical reactions occurring in the substomatal cavity and apoplast will be described
7 (Section 9.4.2). Once O3 has entered the substomatal cavity and apoplast, it is thought that the cell
8 must be able to sense the presence of O3 or its breakdown products in order to initiate the rapid
9 changes in gene expression that have been measured in O3-treated plants. While an "O3 sensor" still
10 remains elusive, much progress has been made in examining several different mechanisms that may
11 contribute both to sensing the presence of O3 and its breakdown products, and also transducing a
12 signal to the nucleus to initiate changes in gene transcription, which will be described in Section
13 9.4.3.1. The next section focuses on changes in gene expression in response to O3 exposure, with
14 particular emphasis on results from transcriptome and proteome analyses (Section 9.4.3.2).
15 Subsequently, the role of phytohormones such as salicylic acid (SA), ethylene (ET), jasmonic acid
16 (JA), and abscisic acid (ABA) and their interactions in both signal transduction and determining
17 plant response to O3 is discussed in Section 9.4.3.3. After O3 uptake and sensing, plants can respond
18 to the oxidative stress to minimize damage. These mechanisms of detoxification, with particular
19 emphasis on antioxidant enzymes and metabolites, are reviewed in Section 9.4.4. The next section
20 focuses on the effects of O3 on primary and secondary metabolism in plants, looking at
21 photosynthesis, respiration and several secondary metabolites, some of which may also act as
22 antioxidants and protect the plant from oxidative stress (Section 9.4.5). The last section focuses on
23 the mechanisms underlying changes in stomatal function (Section 9.4.6). For many of these topics,
24 information from the 2006 O3 AQCD (U.S. EPA, 2006, 088089) has been summarized, as this
25 information is still valid and supported by more recent findings. For other topics, such as genomics
26 and proteomics, which have arisen due to the availability of new technologies, the information is
27 based solely on new publications with no reference to the 2006 O3 AQCD.
28 As Section 9.4 focuses on mechanisms underlying vegetation response to O3, the conditions
29 that are used to study these mechanisms are often artificial and do not necessarily reflect conditions
30 that a plant may be exposed to in an agricultural setting or natural ecosystem. The goal of many of
31 these studies is to elicit a plant response to O3 in a relatively short period of time and not always to
32 simulate ambient O3 exposures. Therefore, plants are often exposed to unrealistically high O3
33 concentrations for several hours or days (acute exposure), and only in a few cases to ambient or
March 2011 9-24 DRAFT - DO NOT CITE OR QUOTE
-------
1 slightly elevated O3 concentrations for longer time periods (chronic exposure). Additionally, the
2 plant species utilized in these studies are often not agriculturally important or commonly found as
3 part of natural ecosystems. Model organisms such as Arabidopsis thaliana are frequently used as
4 they are easy to work with, and mutants or transgenic plants are easy to develop or have already been
5 developed. Furthermore, the Arabidopsis genome has been sequenced, and much is known about the
6 molecular basis of many biochemical and cellular processes.
7 Many of the studies described in this section focus on changes in the expression of genes in
8 O3-treated plants. However, changes in gene expression (i.e., either up- or down-regulation of gene
9 expression) do not always translate into changes in protein quantity and/or activity, as there are many
10 levels of post-transcriptional and post-translational modifications which impact protein quantity and
11 activity. Frequently, these studies do not evaluate whether the observed changes in gene expression
12 lead to changes at the protein level and, therefore, it is not always clear how relevant the changes in
13 gene expression are in determining plant response to O3.
14 The advent of new technologies, such as those employed in genomics and proteomics, has
15 allowed for a more comprehensive analysis of the many molecular and biochemical mechanisms of
16 plant response to O3 and how all these responses interact with or affect each other to determine the
17 ultimate response of plants to a stressor such as O3. While the studies of transcriptome changes are
18 very valuable, further work needs to be done to evaluate whether the transcriptome changes result in
19 concomitant changes in the proteome. A few recent studies have evaluated proteome changes in
20 response to O3, and those results are also discussed in Section 9.4.3.2.
21 The most significant change in this section in relation to the 2006 O3 AQCD (U.S. EPA, 2006,
22 088089) is the increased emphasis on the understanding of molecular mechanisms underlying plant
23 responses to O3, as a significant number of the newer studies that were evaluated for this section
24 focus on changes in gene expression in plants exposed to elevated O3. Conclusions from the 2006 O3
25 AQCD have been supported by these new studies, and the advent of new technologies has allowed
26 for a more comprehensive understanding of the mechanisms governing plant response to O3.
27 In summary, the goal of many of these new studies reported on in this section was to increase
28 knowledge of the mechanisms of plant response to O3 by using artificial exposure conditions and
29 model organisms. This information adds to the understanding of the basic biology of plant response
30 to oxidative stress in the absence of any other potential stressors. The results of these studies are
31 important and valid, even though they may not always directly translate into effects observed in
32 other plants under more realistic exposure conditions. They represent one step in the process of
33 comprehensively understanding plant responses to oxidative stress, which then need to be followed
34 up with additional experiments using other plant species exposed to O3 under more natural
35 conditions.
March 2011 9-25 DRAFT - DO NOT CITE OR QUOTE
-------
9.4.2. Ozone Uptake into the Leaf
1 AX9.2.3 of the 2006 O3 AQCD clearly described the process by which O3 enters plant leaves
2 through open stomata (U.S. EPA, 2006, 088089). This information continues to be valid and is only
3 summarized here.
4 Ozone moves into the leaf interior by diffusing through open stomata, and environmental
5 conditions which promote high rates of gas exchange will favor the uptake of the pollutant by the
6 leaf. Factors that may limit uptake include boundary layer resistance and the size of the stomatal
7 aperture (Figure 9-2) (U.S. EPA, 2006, 088089). Once inside the substomatal cavity, O3 is thought to
8 rapidly react with the aqueous apoplast to form breakdown products known as reactive oxygen
9 species (ROS), such as hydrogen peroxide (H2O2), superoxide (O2 ), hydroxyl radicals (HO) and
10 peroxy radicals (HO2) (Figure 9-3). Hydrogen peroxide is not only a toxic breakdown product of O3,
11 but has been shown to function as a signaling molecule, which is activated in response to both biotic
12 and abiotic stressors. The role of H2O2 in signaling was described in detail in the 2006 O3 AQCD
13 (U.S. EPA, 2006, 088089). Additional organic molecules present in the apoplast or cell wall, such as
14 those containing double bonds or sulfhydryls that are sensitive to oxidation, could also be converted
15 to oxygenated molecules after interacting with O3 (Figure 9-4). These reactions are not only pH
16 dependent but are also influenced by the presence of other molecules in the apoplast (U.S. EPA,
17 2006, 088089). The 2006 O3 AQCD provided a comprehensive summary of what is known about the
18 possible interactions of O3 with other biomolecules (U.S. EPA, 2006, 088089). It is in the apoplast
19 that initial detoxification reactions by antioxidant metabolites and enzymes take place, and these
20 initial reactions are critical to reduce concentrations of the oxidative breakdown products of O3;
21 these reactions are described in more detail in Section 9.4.4 of this document.
March 2011 9-26 DRAFT - DO NOT CITE OR QUOTE
-------
Light
Cuticlec
Epidermis [
Pallisade
Mesophyll
Spongy
Mesophyll
Epidermis
Cuticle
mrnrn
Vascular
System
C0 = [C02]-
Figure 9-2. The microarchitecture of a dicot leaf. [While details among species vary, the general
overview remains the same. Light that drives photosynthesis generally falls upon
the upper (adaxial) leaf surface. Carbon dioxide and ozone enter through the
stomata on the lower (abaxial) leaf surface, while water vapor exits through the
stomata (transpiration).
a.
HO'
Hydroxyl
Radical
b.
Su peroxide
2
f
Ho2-
HO-
HO2-
Peroxyl
Radical
Figure 9-3. Possible reactions of ozone within water. [Ja) Ozone reacts at the double bonds to
form carbonyl groups, (b) Under certain circumstances, peroxides are generated.
March 2011
9-27
DRAFT - DO NOT CITE OR QUOTE
-------
a.
0
NO
H2C=CH2
2. OH- + H2C=CH2
Crigee n
Mechanism / \
^ O 0
\ i
H2C — CH2
OH
\
H2C=CH2
H2C - CH2
ON02
H2C — QH2
H
HC=0
0
n
HC-OH
b.
CH(OH)CH 02H
CH(OH)CH 02H
H2O2
HO OH
0=C CH(OH)CH02H
CHO , CHO
Further Oxidation
Source: Adapted from Mudd (1996, 0527951
Figure 9-4. The Crigee mechanism of ozone attack of a double bond. [Ja) The typical Crigee
mechanism is shown in which several reactions paths from the initial product is
shown, (b) Typical reaction of ascorbic acid with ozone.
March 2011
9-28
DRAFT - DO NOT CITE OR QUOTE
-------
9.4.3. Cellular to Systemic Responses
9.4.3.1. Ozone Sensing and Signal Transduction
1 New technologies allowing for large-scale analysis of oxidative stress-induced changes in
2 gene expression have facilitated the study of signal transduction processes associated with the
3 perception and integration of responses to the stress. Many of these studies have been conducted
4 using Arabidopsis or tobacco plants, for which a variety of mutants are available and/or which can be
5 easily genetically modified to generate either loss-of-function or over-expressing genotypes. Several
6 comprehensive review articles provide an overview of what is known of O3-induced signal
7 transduction processes and how they may help to explain differential sensitivity of plants to the
8 pollutant (Kangasjarvi et al, 2005, 180341)(Ludwikow and Sadowski, 2008, 191426)(Baier et al.,
9 2005, 186866). Additionally, analysis of several studies of transcriptome changes has also allowed
10 for the compilation of these data to determine an initial time-course for O3-induced activation of
11 various signaling compounds (Kangasjarvi et al., 2005, 180341).
12 A number of different mechanisms for plant sensing of O3 have been proposed; however, there
13 is still much that is not known about this process. Ozone and/or the ROS breakdown products could
14 be sensed by an apoplastic receptor protein, which would either be directly modified by O3 or ROS
15 or which might sense O3 or ROS modifications in other apoplastic components (Kangasjarvi et al.,
16 2005, 180341)(Baier et al., 2005, 186866). Some of the earliest events that occur in plant response to
17 O3 have been described in the guard cells of stomata. Reactive oxygen species were observed in the
18 chloroplasts of guard cells in the O3 tolerant Col-0 Arabidopsis thaliana ecotype plants within
19 5 minutes of plant exposure to 350 ppb O3 (Joo et al., 2005, 191307). Reactive oxygen species from
20 the breakdown of O3 in the apoplast are believed to activate GTPases (G-proteins), which, in turn,
21 activate several intracellular sources of ROS, including ROS derived from the chloroplasts.
22 G-proteins are also believed to play a role in activating membrane-bound NADPH oxidases to
23 produce ROS and, as a result, propagate the oxidative burst to neighboring cells (Joo et al., 2005,
24 191307). Therefore, G-proteins are recognized as important molecules involved in plant responses to
25 O3 and may play a role in perceiving ROS from the breakdown of O3 in the apoplast (Booker et al.,
26 2004, 02058l)(Kangasiarvi et al., 2005, 180341).
27 A change in the redox state of the plant and the oxidation of sensitive molecules in itself may
28 represent a means of perception and signaling of oxidative stress in plants. Disulfide-thiol
29 conversions in proteins and the redox state of the glutathione pool are likely to be important
30 components of redox sensing and signal transduction (Foyer and Noctor, 2005, 191555)(Foyer and
31 Noctor. 2005. 631198).
32 Calcium (Ca2+) has also been implicated in the transduction of signals to the nucleus in
33 response to oxidative stress. The influx of Ca2+ from the apoplast into the cell occurs early in plant
34 response to O3, and it is thought to play a role in regulating the activity of protein kinases, which are
35 discussed below (Hamel et al., 2005, 191214)(Baier et al., 2005, 186866). Calcium channel blockers
March 2011 9-29 DRAFT - DO NOT CITE OR QUOTE
-------
1 inhibited O3-induced activation of protein kinases in tobacco suspension cells exposed to 500 ppb O3
2 for 10 minutes, indicating that the opening of Ca2+ channels is an important upstream signaling event
3 or that the as yet unknown upstream process has a requirement for Ca2+ (Samuel et al, 2000,
4 625706V
5 Once ROS are generated by NADPH oxidase, signals are further transmitted to the nucleus to
6 initiate changes in gene expression associated with plant defense responses. Integral to these signal
7 cascades are mitogen-activated protein kinases (MAPK), which phosphorylate proteins and activate
8 various cellular responses (Hamel et al., 2005, 191214). Mitogen-activated protein kinases are
9 induced in several different plant species in response to O3 exposure, including tobacco (Samuel et
10 al., 2005, 199316). Arabidopsis (Ludwikow et al., 2004, 595939). the shrub Phillyrea latifolia
11 (Paolacci et al., 2007, 191422) and poplar (Hamel et al., 2005, 191214). In tobacco, the MAPK that
12 is induced by plant exposure to 500 ppb O3 is a salicylic acid (SA) induced protein kinase (SIPK),
13 which was found to positively regulate O3-induced ethylene (ET) production and negatively regulate
14 SA accumulation (Samuel et al., 2005, 199316). In poplar suspension cells exposed to 500 ppb O3,
15 O3-induced activation of two distinct MAP kinases was dependent on ROS formation and the
16 activity of Ca2+ channels (Hamel et al., 2005, 191214). Arabidopsis mutants with suppressed activity
17 of MAPK3 and MAPK6 (an ortholog of SIPK in tobacco) were more susceptible to exposure to
18 500 ppb O3, as evidenced by tissue death resembling the hypersensitive response (Miles et al., 2005,
19 191648). Similar results of increased susceptibility to 500 ppb O3 were found in tobacco lines either
20 over-expressing SIPK or with suppressed SIPK function. The authors concluded that these tobacco
21 lines were unable to cope with increased oxidative stress due to the alteration of the normal
22 O3-induced MAPK signal transduction process (Samuel and Ellis, 2002, 625703). The Arabidopsis
23 MAPK3 and MAPK6 are also differentially induced in Col-0 and the O3-sensitive Arabidopsis
24 mutant radical induced cell death (rcdl) exposed to 250-300 ppb O3 for 6 hours, indicating a role for
25 these kinases in plant response to oxidative stress (Overmyer et al., 2005, 191596). Similarly,
26 components of the MAPK cascades were down-regulated in the sensitive Wassilewskija (Ws)
27 Arabidopsis ecotype after exposure to 300 ppb O3 for 6 h, suggesting that the sensitive ecotypes may
28 not activate these defense responses (Mahalingam et al., 2006, 191221). Expression of MAPK5 was
29 also reduced in the Ws ecotype in response to chronic O3 exposures; Ws Arabidopsis plants were
30 exposed to O3 concentrations 20-25% above ambient for 8-12 days in the SoyFACE site (Li et al.,
31 2006. 191332).
32 The cysteine-rich RLKs (CRKs), which are part of the receptor-like/Pelle kinase (RLKs)
33 group, are thought to be involved in the regulation of defense responses and cell death in Arabidopsis
34 (Wrzaczek et al., 2010, 644190). CRKs were up-regulated by treatments that resulted in apoplastic
35 ROS production, such as O3 (250 ppb O3 for 6 hours) and pathogen treatments, but either remained
36 unchanged or were down-regulated by treatments resulting in ROS production in other cellular
37 compartments, such as the mitochondria or chloroplasts. Although their function remains unclear, it
38 has been postulated that a conserved cysteine motif in the CRKs could serve as a sensor for redox
39 modifications in the cell resulting from ROS production (Wrzaczek et al., 2010, 644190).
March 2011 9-30 DRAFT - DO NOT CITE OR QUOTE
-------
1 In conclusion, experimental evidence suggests that there may be several different mechanisms
2 involved in sensing the presence of O3 or its breakdown products. These mechanisms may vary by
3 species or developmental stage of the plant or may co-exist and be activated by different exposure
4 conditions. Calcium and protein kinases are likely involved in the transduction of the initial signal to
5 the nucleus and other cellular compartments to initiate the changes in gene transcription discussed in
6 Section 9.4.3.2.
9.4.3.2. Gene Expression Changes in Response to Ozone
7 The advent of DNA microarray technology has allowed for the study of gene expression in
8 cells on a large scale. Rather than assessing changes in gene expression of individual genes, DNA
9 microarrays facilitate the evaluation of entire transcriptomes, providing a comprehensive picture of
10 alterations in gene expression. In addition, these studies have provided more insight into the complex
11 interactions between molecules and signal pathways, which result in the regulation of plant
12 responses to stresses such as O3 (Ludwikow and Sadowski, 2008, 191426). Transcriptome analysis
13 of O3-treated plants has been performed in several species, including Arabidopsis thaliana (Tosti et
14 al, 2006, 191425)(Heidenreich et al, 2005, 191260KLJ et al, 2006, 191332)(Mahalingam et al.,
15 2005, 191693)(Tamaoki et al., 2003, 080053). pepper (Capsicum annuum) (Lee and Yun, 2006,
16 191592). clover (Medicago truncatula) (Puckette et al., 2008, 191698). Phillyrea latifolia (Paolacci
17 et al., 2007, 191422). and European beech (Fagus sylvatica) (Olbrich et al., 2005, 191697)(Olbrich
18 et al., 2009, 596020)(Olbrich et al., 2010, 625424). In some cases, researchers compared
19 transcriptomes of two or more cultivars, ecotypes or mutants that differed in their sensitivity to O3
20 (Lee and Yun, 2006, 191592)Oamaoki et al., 2003, 080053)qi et al., 2006,19L332)(Puckette et al.,
21 2008, 191698)(Rizzo et al., 2007, 191447). Species, O3 exposure conditions (concentration, duration
22 of exposure) and sampling times varied significantly in these studies. However, functional
23 classification of the genes that were either up- or down-regulated by plant exposure to O3 exhibited
24 common trends. Genes involved in plant defense, signaling and those associated with the synthesis
25 of plant hormones and secondary metabolism were generally up-regulated, while those related to
26 photosynthesis and general metabolism were generally down-regulated in O3-treated plants (Tosti et
27 al., 2006, 191425)(Olbrich et al., 2005, 191697)(Tamaoki et al., 2003, 08005!)(Lee and Yun5 2006,
28 191592)(Puckette et al., 2008, 191698)(Li et al., 2006, 191332).
29 Analysis of the transcriptome has been used to evaluate differences in gene expression
30 between O3 sensitive and tolerant plants. In pepper, 67% of the 180 genes studied that were affected
31 by O3 were differentially regulated in the sensitive and tolerant cultivars. At both 0 hours and
32 48 hours after a 3-day exposure at 150 ppb, O3 responsive genes were either up- or down-regulated
33 more markedly in the sensitive than in the tolerant cultivar (Lee and Yun, 2006, 191592).
34 Transcriptome analysis also revealed differences in timing and magnitude of changes in gene
35 expression between sensitive and tolerant clovers. Acute exposure (300 ppb O3 for 6 hours) led to the
36 production of an oxidative burst in both clovers (Puckette et al., 2008, 191698). However, the
37 sensitive Jemalong cultivar exhibited a sustained ROS burst and a concomitant down-regulation of
March 2011
9-31
DRAFT - DO NOT CITE OR QUOTE
-------
1 defense response genes at 12 hours after the onset of exposure, while the tolerant JE 154 accession
2 showed much more rapid and large-scale transcriptome changes than the Jemalong cultivar (Puckette
3 et al, 2008, 191698V
4 Arabidopsis ecotypes WS and Col-0 were exposed to 1.2 x ambient O3 concentrations for
5 8-12 days at the SoyFACE site (Li et al., 2006, 191332). The sensitive WS ecotype showed a far
6 greater number of changes in gene expression in response to this low-level O3 exposure than the
7 tolerant Col-0 ecotype. Exposure of the WS ecotype to 300 ppb O3 for 6 hours showed a rapid
8 induction of genes leading to cell death, such as proteases, and down-regulation or inactivation of
9 cell signaling genes, demonstrating an ineffective defense response in this O3 sensitive ecotype
10 (Mahalingam et al., 2006,191221).
11 The temporal response of plants to O3 exposure was evaluated in the Arabidopsis Col-0
12 ecotype during a 6-h exposure at 350 ppb O3 and for 6 hours after the exposure was completed.
13 Results of this study, shown in Figure 9-5, indicate that genes associated with signal transduction and
14 regulation of transcription were in the class of early up-regulated genes, while genes associated with
15 redox homeostasis and defense/stress response were in the class of late up-regulated genes
16 (Mahalingam et al.. 2005. 191693).
17 A few studies have been conducted to evaluate transcriptome changes in response to longer
18 term chronic O3 exposures in woody plant species. Longer term exposures resulted in the up-
19 regulation of genes associated with secondary metabolites, including isoprenoids, polyamines and
20 phenylpropanoids in 2-year-old seedlings of the Mediterranean shrub Phillyrea latifolia exposed to
21 110 ppb O3 for 90 days (Paolacci et al., 2007, 191422). In 3-year-old European beech saplings
22 exposed to O3 for 20 months, with monthly average twice ambient O3 concentrations ranging from
23 11 to 80 ppb, O3-induced changes in gene transcription were similar to those observed for
24 herbaceous species, including genes associated with plant stress response, primary metabolism,
25 hormone synthesis, cell structure and premature senescence (Olbrich et al., 2009, 596020). In
26 another study, the magnitude of these transcriptional changes was far greater in beech saplings than
27 in adult trees exposed to the same O3 concentrations for the same time period, indicating that adult
28 trees may be less responsive to this stressor than younger plants (Olbrich et al., 2010, 625424).
29 These results have been substantiated by results from proteome analysis in rice, poplar, wheat,
30 and soybean. Exposure of soybean to 120 ppb O3 for 12 h/day for 3 days in growth chambers
31 resulted in decreases in the quantity of proteins associated with photosynthesis, while proteins
32 involved with antioxidant defense and C metabolism increased (Ahsan et al., 2010, 644189). Young
33 poplar plants exposed to 120 ppb O3 in a growth chamber for 35 days also showed significant
34 changes in proteins involved in C metabolism (Bohler et al., 2007, 199408). Declines in enzymes
35 associated with C fixation, the Calvin cycle and photosystem II were measured, while ascorbate
36 peroxidase and enzymes associated with glucose catabolism increased in abundance. Two-week-old
37 rice seedlings exposed to varying levels of O3 (4, 40, 80, 120 ppb) in a growth chamber for 9 days
38 showed reductions in expression of proteins associated with photosynthesis and energy metabolism,
39 and increases in some antioxidant and defense related proteins (Feng et al., 2008, 191626). A
March 2011 9-32 DRAFT - DO NOT CITE OR QUOTE
-------
1 subsequent study of O3-treated rice seedlings (exposed to 200 ppb O3 for 24 h) focusing on the
2 integration of transcriptomics and proteomics, supported and further enhanced these results (Cho et
3 al., 2008, 603254). The authors found that of the 22,000 genes analyzed from the rice genome, 1,535
4 were differentially regulated by O3. Those genes were functionally categorized as transcription
5 factors, MAPK cascades, those encoding for enzymes involved in the synthesis of JA, ET, shikimate,
6 tryptophan and lignin, and those involved in glycolysis, citric acid cycle, oxidative respiration and
7 photosynthesis. The authors determined that the proteome and metabolome analysis supported the
8 results of the transcriptome changes described above (Cho et al., 2008, 603254). This type of study,
9 which ties together results from changes in gene expression, protein quantity and activity, and
10 metabolite levels, provides the most complete picture of the molecular and biochemical changes
11 occurring in plants exposed to a stressor such as O3. Sarkar et al. (2010, 657214) compared two
12 cultivars of wheat grown in OTCs at several O3 concentrations, including filtered air, ambient O3
13 (mean concentration 47 ppb), ambient +10 ppb and ambient + 20 ppb for 5 h/day for 50 days.
14 Declines in the rate of photosynthesis and stomatal conductance were related to decreases in proteins
15 involved in C fixation and electron transport, and evidence of increased proteolysis of photosynthetic
16 proteins such as the large subunit of ribulose-l,6-bisphosphate carboxylase/oxygenase (Rubisco).
17 Enzymes that take part in energy metabolism, such as ATP synthesis, were also down-regulated,
18 while defense/stress related proteins were induced with O3 treatment. In comparing the two wheat
19 cultivars, Sarkar et al. (2010, 657214) found that while the qualitative changes in protein expression
20 between the two cultivars was similar, the magnitude of these changes differed between the sensitive
21 and tolerant wheat cultivars.
22 All of these studies describe common trends for changes in gene and protein expression which
23 occur in a variety of plant species in response to O3. While genes associated with C assimilation and
24 general metabolism are down-regulated, genes associated with signaling, catabolism, and defense are
25 up-regulated. The magnitude of these changes in gene and protein expression appears to be related to
26 plant sensitivity or tolerance to O3.
March 2011 9-33 DRAFT - DO NOT CITE OR QUOTE
-------
(A) Oj-induccd oxidalive bursl
Signaling Redox homeoslasis
Transcription Defense/stress response
PR proteins
12 hr
ROSt
Afar
9hr
12 far
activation^,
SA-intluecd
SH
l.JIC PR )>CML'%
Photosynthesis
Source: Used with permission from Springer, Mahalingam et al. (2005,1916931.
Figure 9-5 Composite diagram of major themes in the temporal evolution of the genetic
response to ozone stress. I|A) Temporal profile of the oxidative stress response to
ozone. The biphasic ozone-induced oxidative burst is represented in black, with the
ROS control measurements shown as a broken line. Average transcript profiles are
shown for early up-regulated genes (yellow, peaks at 0.5-1 hours), and the 3 hours
(blue), 4.5 hours (red) and 9-12 hours (green) late up-regulated genes and forthe
down-regulated genes coding for photosynthesis proteins (brown). (B)
Diagrammatic representation of redox regulation of the oxidative stress response.
9.4.3.3. Role of Phytohormones in Plant Response to Ozone
1 Many studies of O3 effects on plants have analyzed the importance of plant hormones such as
2 SA, ET and JA in determining plant response to O3; some of the roles of these hormones were
3 described in the 2006 O3 AQCD (U.S. EPA, 2006, 088089). Transcriptome analysis and the use of a
4 variety of mutants have allowed for further elucidation of the complex interactions between SA, ET,
5 JA and the role of abscisic acid (ABA) in mediating plant response to O3 (Ludwikow and Sadowski,
6 2008, 191426). In addition to their roles in signaling pathways, phytohormones also appear to
March 2011
9-34
DRAFT - DO NOT CITE OR QUOTE
-------
1 regulate, and be regulated by, the MAPK signaling cascades described previously. Most evidence
2 suggests that while ET and S A are needed to develop O3-induced leaf lesions, JA acts
3 antagonistically to SA and ET to limit the lesions (Figure 9-6) (Kangasjarvi et al., 2005, 180341).
4 The rapid production of ET in response to O3 has been described in many plant species and
5 has been further characterized through the use of a variety of mutants that either over-produce or are
6 insensitive to ET. Production of stress ET in O3-treated plants, which is thought to be a wounding
7 response, was found to be correlated to the degree of injury development in leaves (U.S. EPA, 2006,
8 088089). More recent studies have supported these conclusions and have also focused on the
9 interactions occurring between several oxidative-stress induced phytohormones. Yoshida et al.
10 (2009, 191385) determined that ET likely amplifies the oxidative signal generated by ROS, thereby
11 promoting lesion formation. By analyzing the O3-induced transcriptome of several Arabidopsis
12 mutants of the Col-0 ecotype, Tamaoki et al. (2003, 080053) determined that at 12 hours after
13 initiating the O3 exposure (200 ppb for 12 hours), the ET and JA signaling pathways were the main
14 pathways used to activate plant defense responses, with a lesser role for SA. The authors also
15 demonstrated that low levels of ET production could stimulate the expression of defense genes,
16 rather than promoting cell death when ET production is high. Tosti et al. (2006, 191425) supported
17 these findings by showing that O3 not only activates the biosynthetic pathways of ET, JA and SA, but
18 also increases the expression of genes related to the signal transduction pathways of these
19 phytohormones in O3-treated Arabidopsis plants (300 ppb O3 for 6 hours). Conversely, in the O3
20 sensitive Ws ecotype, its sensitivity may, in part, be due to intrinsically high ET levels leading to SA
21 accumulation, and the high ET and SA may act to repress JA-associated genes, which would serve to
22 inhibit the spread of lesions (Mahalingam et al., 2006, 191221). Ogawa et al. (2005, 191653) found
23 that an O3-induced increase in S A leads to the formation of leaf lesions in tobacco plants exposed to
24 200 ppb O3 for 6 hours. Furthermore, several genes encoding for enzymes in the biosynthetic
25 pathway of SA were suppressed in transgenic tobacco plants with reduced levels of O3-induced ET
26 production, suggesting that SA levels are controlled by ET in the presence of O3.
27 Exposure of the Arabidopsis mutant rcdl to acute doses of O3 (250 ppb O3 for 8 h/day for
28 3 days) resulted in programmed cell death (PCD) and the formation of leaf lesions. Salicylic acid
29 was required for the initiation of PCD in these mutants, and increased ET synthesis resulted in
30 amplification of cell death, while JA was thought to contain the spreading of lesions (Overmyer et
31 al., 2000, 036361). In addition, the O3-treated rcdl mutants showed higher SA and JA accumulation
32 as compared to the Col-0 ecotype. High SA levels could be involved in amplifying cell death in these
33 mutants. Jasmonic acid, which is thought to accumulate as a direct result of cell death, may represent
34 an autocatalytic mechanism for limiting cell death, by which the increased cell death observed in the
35 rcdl mutant would determine the high amount of JA needed to contain lesion formation (Overmyer
36 et al., 2005, 191596). In cotton plants exposed to a range of O3 concentrations (0-120 ppb) and
37 methyl jasmonate (MeJA), Grantz et al. (2010, 625419) determined that endogenous applications of
38 MeJA did not protect plants from chronic O3 exposure.
March 2011 9-35 DRAFT - DO NOT CITE OR QUOTE
-------
1 There are distinct patterns of gene expression when evaluating early and later O3 responsive
2 genes (Mahalingam et al., 2005, 191693). While ET and JA pathways were stimulated rapidly after
3 Arabidopsis Col-0 exposure to O3, at 48 hours postexposure, JA and ET synthesis in O3-treated
4 Arabidopsis were reduced, while the synthesis of SA was stimulated (D'Haese et al., 2006, 191448).
5 Abscisic acid has been investigated for its role in regulating stomatal aperture and also for its
6 contribution to signaling pathways in the plant. The role of ABA and the interaction between ABA
7 and H2O2 in O3-induced stomatal closure was described in the 2006 O3 AQCD (U.S. EPA, 2006,
8 088089). More recently, it was determined that synthesis of ABA was induced in O3-treated
9 Arabidopsis plants (250-350 ppb O3 for 6 hours), with a more pronounced induction in several O3
10 sensitive red mutants as compared to the wildtype Col-0 (Overmyer et al., 2008, 191283). Ludwikow
11 et al. (2009, 199440) used Arabidopsis ABIltd mutants, in which a key negative regulator of ABA
12 action (abscisic acid insensitive 1 protein phosphatase 2C) has been knocked out, to examine O3
13 responsive genes in this mutant compared to the Arabidopsis Col-0. Results of this study indicate a
14 role for ABU in negatively regulating the synthesis of both ABA and ET in O3-treated plants
15 (350 ppb O3 for 9 hours). Additionally, ABU may stimulate JA-related gene expression, providing
16 evidence for an antagonistic interaction between ABA and JA signaling pathways (Ludwikow et al.,
17 2009. 199440).
18 Nitric oxide (NO) has also been shown to play a role in regulating O3-induced gene expression
19 in plants. However, little is known to date about NO and its role in the complex interactions of
20 molecules in response to O3. Exposure of tobacco to O3 (150 ppb for 5 hours) stimulated NO and
21 NO-dependent ET production, while NO production itself did not depend on the presence of ET
22 (Ederli et al., 2006, 191479). Analysis of O3-treated Arabidopsis indicated the possibility of a dual
23 role for NO in the initiation of cell death and later lesion containment (Ahlfors et al., 2009, 191533).
24 While much work remains to be done to better elucidate how plants sense O3 and how signals
25 are communicated to the nucleus to generate plant responses to oxidative stress, it is clear that the
26 mechanism for O3 sensing and signal transduction is very complex. Many of the phytohormones and
27 other signaling molecules thought to be involved in these processes are interactive and depend upon
28 a variety of other factors, which could be either internal or external to the plant. This results in a
29 highly dynamic and complex system, capable of generating a variety of plant responses to oxidative
30 stress.
March 2011 9-36 DRAFT - DO NOT CITE OR QUOTE
-------
ozone
Cell
death
Source: Used with permission from Blackwell Publishing Ltd., Kangasjarvi et al. (2005, 1803411.
Figure 9-6. The oxidative cell death cycle. (Ozone-derived radicals induce endogenous ROS
production (1) which results in salicylic acid (SA) accumulation and programmed
cell death; (2) Cell death triggers ethylene (ET) production, which is required for the
continuing ROS production responsible for the propagation of cell death; (3)
Jasmonates counteract the progression of the cycle by antagonizing the cell death
promoting function of SAand ET; (4) Abscisic acid (ABA) antagonizes ET function in
many situations and might also have this role in ozone-induced cell death; (5)
Mutually antagonistic interactions between ET, SA and jasmonic acid (JA) are
indicated with red bars.
9.4.4. Detoxification
9.4.4.1. Overview of Ozone-Induced Defense Mechanisms
1 Plants are exposed to an oxidizing environment on a continual basis, and many reactions that
2 are part of the basic metabolic processes, such as photosynthesis and respiration, generate ROS. As a
3 result, there is an extensive and complex mechanism in place to detoxify these oxidizing radicals,
4 including both enzymes and metabolites, which are located in several locations in the cell and also in
5 the apoplast of the cell. As O3 enters the leaf through open stomata, the first point of contact of O3
6 with the plant is thought to be in the apoplast, where it breaks down to form oxidizing radicals such
7 as H2O2, O2, HO- and HO2. Another source of oxidizing radicals is an oxidative burst, generated by
8 a membrane-bound NADPH oxidase enzyme, which is thought to be part of the plant's defense
9 system against pathogens (Schraudner et al., 1998, 056358). Antioxidant metabolites and enzymes
10 located in the apoplast are thought to form a first line of defense by detoxifying O3 and/or the ROS
11 that are formed as breakdown products of O3 (Section 9.4.2.). However, even with the presence of
12 several antioxidants, including ascorbate, the redox buffering capacity of the apoplast is far less than
March 2011
9-37
DRAFT - DO NOT CITE OR QUOTE
-------
1 that of the cytoplasm, as it lacks the regeneration systems necessary to retain a reduced pool of
2 antioxidants (Foyer and Noctor, 2005, 631198).
3 Redox homeostasis is regulated by the presence of a pool of antioxidants, which are typically
4 found in a reduced state and detoxify ROS produced by oxidases or electron transport components.
5 As ROS increase due to environmental stress such as O3, the antioxidant pool will no longer be able
6 to maintain its reduced state (Foyer and Noctor, 2005, 631198). As such, not only the quantity and
7 types of antioxidant enzymes and metabolites present, but also the cellular ability to regenerate those
8 antioxidants are important considerations in mechanisms of plant tolerance to oxidative stress
9 (Dizengremel et al, 2008, 191587). Molecules such as glutathione (GSH), thioredoxins and NADPH
10 play very important roles in this regeneration process; additionally, alterations in C metabolism will
11 be necessary to supply the needed reducing power for antioxidant regeneration (Dizengremel et al.,
12 2008, 191587). Increases in the activity of catabolic pathways allow the cell to generate more
13 NADPH (Section 9.4.5).
9.4.4.2. Role of Antioxidants in Plant Defense Responses
14 Ascorbate has been the focus of many different studies as an antioxidant metabolite that
15 protects plants from exposure to O3. It is found in several cellular locations, including the
16 chloroplast, the cytosol and the apoplast (Noctor and Foyer, 1998, 657213). Ascorbate is synthesized
17 in the cell and transported to the apoplast. Apoplastic ascorbate can be oxidized to dehydroascorbate
18 (DHA) with exposure to O3 and is then transported back to the cytoplasm. Here, DHA is reduced to
19 ascorbate by the enzyme dehydroascorbate reductase (DHAR) and reduced GSH, which is part of the
20 ascorbate-glutathione cycle (Noctor and Foyer, 1998, 657213). Many studies have focused on
21 evaluating whether ascorbate is the determining factor in differential sensitivity of plants to O3.
22 Cheng et al. (2007, 191499) exposed two soybean cultivars to elevated O3 (77 ppb) and filtered air
23 for 7 h/day for 6 days. The differences in sensitivity between the two cultivars could not be
24 explained by differential O3 uptake or by the fraction of reduced ascorbate present in the apoplast.
25 However, total antioxidant capacity of the apoplast was twofold higher in the tolerant Essex cultivar
26 as compared to the sensitive Forrest cultivar, indicating that there may be other compounds in the
27 leaf apoplast that scavenge ROS. D'Haese et al. (2005, 191551) exposed the NC-S (sensitive) and
28 NC-R (resistant) clones of white clover (Trifolium repens) to 60 ppb O3 for 7 h/day for 5 days in
29 environmental chambers. Surprisingly, the NC-S clone had a higher constitutive concentration of
30 apoplastic ascorbate with a higher redox status than the NC-R clone. However, the redox status of
31 symplastic GSH was higher in NC-R, even though the concentration of GSH was not higher than in
32 NC-S. In addition, total symplastic antioxidative capacity was not a determining factor in differential
33 sensitivity between these two clones. Severino et al. (2007, 199293) also examined the role of
34 antioxidants in the differential sensitivity of the two white clover clones by growing them in the field
35 for a growing season and then exposing them to elevated O3 (100 ppb for 8 h/day for 10 days) in
36 OTC at the end of the field season. The NC-R clone had greater quantities of total ascorbate and total
37 antioxidants than the NC-S clone at the end of the experiment. While the second study indicates a
March 2011 9-38 DRAFT - DO NOT CITE OR QUOTE
-------
1 possible relationship between O3 tolerance and ascorbate levels, the first study indicates that there
2 are other factors besides ascorbate that determine plant tolerance to O3. In snap bean, plants of the O3
3 tolerant Provider cultivar had greater total ascorbate and more ascorbate in the apoplast than the
4 sensitive S156 cultivar after exposure to 71 ppb O3 for 10 days in OTC (Burkey et al, 2003,
5 630251). While most of the apoplastic ascorbate was in the oxidized form, the ratio of reduced
6 ascorbate to total ascorbate was higher in Provider than S156, indicating that Provider is better able
7 to maintain this ratio to maximize plant protection from oxidative stress.
8 While the quantities of antioxidant metabolites such as ascorbate are an important indicator of
9 plant tolerance to O3, the ability of the plant to recycle oxidized ascorbate efficiently plays a large
10 role in determining the plant's ability to deal with a sustained exposure to oxidative stress. Tobacco
11 plants over-expressing DHAR were better protected from exposure to either chronic (100 ppb O3
12 4 h/day for 30 days) or acute (200 ppb O3 for 2 hours) conditions than control plants and those with
13 reduced expression of DHAR. The DHAR over-expressing plants exhibited an increase in guard cell
14 ascorbic acid, leading to a decrease in stomatal responsiveness to O3 and an increase in stomatal
15 conductance and O3 uptake. Despite this, the presence of higher levels of ascorbic acid led to a lower
16 oxidative load and a higher level of photosynthetic activity in the DHAR over-expressing plants
17 (Chen and Gallie, 2005, 191465). A subsequent study with tobacco plants over-expressing DHAR
18 confirmed some of these results. Levels of ascorbic acid were higher in the transgenic tobacco
19 plants, and they exhibited greater tolerance to O3 exposure (200 ppb O3) as demonstrated by higher
20 photosynthetic rates in the transgenic plants as compared to the control plants (Eltayeb et al., 2006,
21 191377). Over-expression of monodehydroascorbate reductase (MDAR) in tobacco plants also
22 showed enhanced stress tolerance in response to O3 exposure (200 ppb O3), with higher rates of
23 photosynthesis and higher levels of reduced ascorbic acid as compared to controls (Eltayeb et al.,
24 2007, 191553). Results of these studies show the importance of ascorbic acid as a detoxification
25 mechanism, but more importantly emphasize that the recycling of oxidized ascorbate and
26 maintenance of a reduced pool of ascorbate is critical in determining plant tolerance to oxidative
27 stress.
28 The roles of other antioxidant metabolites and enzymes, including GSH, catalase (CAT), and
29 superoxide dismutase (SOD), were comprehensively reviewed in the 2006 O3 AQCD (U.S. EPA,
30 2006, 088089). Additional studies have supported the findings reported in that document. Superoxide
31 dismutase (SOD) and peroxidase (POD) activities were measured in both the tolerant Bel B and
32 sensitive Bel W3 tobacco cultivars exposed to ambient O3 concentrations for 2 weeks 3 times
33 throughout a growing season (Borowiak et al., 2009, 191247). Tulip poplar (Liriodendron tulipifera)
34 trees exposed to increasing O3 concentrations (from 100 to 300 ppb O3 during a 2-week period)
35 showed increases in activities of SOD, ascorbate peroxidase (APX), glutathione reductase (GR),
36 MDAR, DHAR, CAT and POD in the 2-week period, although individual enzyme activities
37 increased at different times during the 2-week period (Ryang et al., 2009, 191267).
38 Longer, chronic O3 exposures in trees revealed some distinctive patterns of increases in SOD
39 and APX activity that were measured in Quercus mongolica after 45 days of plant exposure to
March 2011 9-39 DRAFT - DO NOT CITE OR QUOTE
-------
1 80 ppb O3, which were followed by declines in the activities and quantities of these enzymes after
2 75 days of exposure (Yan et al, 2010, 628514). Similarly, activities of SOD, APX, DHAR, MDAR,
3 and GR increased in Gingko biloba trees during the first 50 days of exposure to 80 ppb O3, followed
4 by decreases in activity below control values after 50 days of exposure (He et al., 2006, 199430).
5 Soybean plants exposed to 70 or 100 ppb O3 for 4 h/day over the course of a growing season showed
6 elevated POD activity and a decrease in CAT activity at 40 and 60 days after germination (Singh et
7 al., 2010, 386275).
8 Antioxidant enzymes appear to increase in quantity in O3-treated plants as a defense
9 mechanism against oxidative stress. However, it appears that plants cannot maintain these elevated
10 levels of antioxidant enzymes for an extended period of time, likely due to the high metabolic costs
11 involved. Therefore, plants exposed to chronic O3 conditions may show more symptoms of exposure
12 as defensive mechanisms are down-regulated over time.
9.4.5. Effects on Primary and Secondary Metabolism
9.4.5.1. Light and Dark Reactions of Photosynthesis
13 Declines in the rate of photosynthesis and stomatal conductance in O3-treated plants have been
14 documented for many different plant species (U.S. EPA, 2006, 088089)(Booker et al., 2009, 191569)
15 (Wittig et al., 2007, 191695). Much of the literature regarding O3 effects on photosynthesis has
16 focused on C assimilation; however, more recently, attention has also been focused on deleterious
17 effects on the light reactions. Chlorophyll fluorescence provides a useful measure of changes to the
18 photosynthetic process from exposure to oxidative stress. Decreases in the Fv/Fm ratio (a measure of
19 the maximum efficiency of Photosystem II) in dark adapted leaves indicate a decline in the
20 efficiency of the PSII photosystems and a concomitant increase in non-photochemical quenching
21 (Guidi and Degl'lnnocenti, 2008, 191571)(Scebba et al., 2006, 191219). Changes in these parameters
22 have been correlated to differential sensitivity of plants to the pollutant. In a study to evaluate the
23 response of 4 maple species to O3 (exposed to an 8-h avg of 51 ppb for ambient and 79 ppb for
24 elevated treatment in OTC), the 2 species which were most sensitive based on visible injury and
25 declines in CO2 assimilation also showed the greatest decreases in Fv/Fm in symptomatic leaves. In
26 asymptomatic leaves, CO2 assimilation decreased significantly but there was no significant decline
27 in Fv/Fm (Calatayud et al., 2007, 191411). Degl 'Innocenti et al. (2007, 191350) measured
28 significant decreases in Fv/Fm in young and symptomatic leaves of a resistant tomato genotype (line
29 93.1033/1) in response to O3 exposure (150 ppb O3 for 3 hours in a growth chamber), but only minor
30 decreases in asymptomatic leaves with no associated changes in net photosynthetic rate. In the O3
31 sensitive tomato cultivar Cuor Di Bue, the Fv/Fm ratio did not change, while the photosynthetic rate
32 declined significantly in asymptomatic leaves (Degl'lnnocenti et al., 2007, 191350). In two soybean
33 cultivars, Fv/Fm also declined significantly with plant exposure to O3 (Singh et al., 2009, 199427). It
March 2011 9-40 DRAFT - DO NOT CITE OR QUOTE
-------
1 appears that in asymptomatic leaves, photoinhibition, as indicated by a decrease in Fv/Fm, is not the
2 main reason for a decline in photosynthesis.
3 An evaluation of photosynthetic parameters of two white clover (Trifolium repens cv. Regal)
4 clones that differ in their O3 sensitivity revealed that O3 (40-110 ppb O3 for 7 h/day for 5 days)
5 increased the coefficient of non-photochemical quenching (QNP) in both the resistant (NC-R) and
6 sensitive (NC-S) clones, however qNP was significantly lower for the sensitive clone (Crous et al,
7 2006, 199321). Sensitive Acer clones had a lower coefficient of non-photochemical quenching,
8 while exposure to O3 increased qNp in both sensitive and tolerant clones (Calatayud et al., 2007,
9 191411). While exposure to O3 also increased qNp in tomato, there were no differences in the
10 coefficient of photochemical quenching between cultivars thought to be differentially sensitive to O3.
11 (Degl'Innocenti et al., 2007, 191350). Higher qNP as a result of exposure to O3 indicates a reduction
12 in the proportion of absorbed light energy being used to drive photochemistry. A lower coefficient of
13 non-photochemical quenching in O3 sensitive plants could indicate increased vulnerability to ROS
14 generated during exposure to oxidative stress (Crous et al., 2006, 199321).
15 Several measures of the light reactions of photosynthesis are sensitive to exposure to O3,
16 however, photosynthetic C assimilation is generally considered to be more affected by pollutant
17 exposure, resulting in an overall decline in photosynthesis (Heath, 2008, 195632)(Fiscus et al., 2005,
18 079155)(Guidi and Degl'Innocenti, 2008, 191571). Loss of C assimilation capacity has been shown
19 to result primarily from declines in the quantity of Rubisco (Calatayud et al., 2007, 191411)(Singh et
20 al., 2009, 199427). Experimental evidence suggests that both decreases in Rubisco synthesis and
21 enhanced degradation of the protein contribute to the measured reduction in its quantity (U.S. EPA,
22 2006, 088089). Reduced C assimilation has been linked to reductions in biomass and yield (Keutgen
23 et al., 2005, 191295¥He et al., 2007,199789)(Novak et al., 2007,194.630)(Gregg et al., 2006,
24 186961), (Wang et al., 2009, 199303).
25 Most of the research on O3 effects on photosynthesis has focused on C3 (Calvin cycle) plants
26 because C4 (Hatch-Slack) plants have lower stomatal conductance and therefore assumed to be less
27 sensitive to O3 stress. However, a few studies have been conducted to evaluate the effects of O3 on
28 C4 photosynthesis. In older maize leaves, Leitao et al. (2007, 191456)(2007. 191263) found that the
29 activity, quantity and transcript levels of both Rubisco and phosphoenolpyruvate carboxylase (PEPc)
30 decreased as a function of rising O3 concentration. In younger maize leaves, the quantity, activity,
31 and transcript levels of the carboxylases were either increased or unaffected in plants exposed to
32 40 ppb O3 for 7 h/day for 28-33 days, but decreased at 80 ppb (Leitao et al., 2007, 191263¥Leitao et
33 al., 2007, 199379V
9.4.5.2. Respiration and Dark Respiration
34 While much research emphasis regarding O3 effects on plants has focused on the negative
35 impacts on C assimilation, other studies have measured impacts on catabolic pathways such as
36 respiration and photorespiration. Generally, respiration has been found to increase in plants exposed
37 to O3. Bean plants exposed to ambient (average 12-h mean 43 ppb) and twice ambient (average 12-h
March 2011 9-41 DRAFT - DO NOT CITE OR QUOTE
-------
1 mean 80 ppb) O3 showed increases in respiration. When mathematically partitioned, the maintenance
2 coefficient of respiration was significantly increased by O3 exposure, while the growth coefficient of
3 respiration was not affected (Amthor, 1988, 041870). Loblolly pines were exposed to ambient (12-h
4 daily mean was 45 ppb) and twice ambient (12 hours daily mean was 86 ppb) O3 for 12 h/day for
5 approximately seven months per year for 3 and 4 years. While photosynthetic activity declined with
6 the age of the needles and increasing O3 concentration, enzymes associated with respiration showed
7 higher levels of activity with increasing O3 concentration (Dizengremel et al., 1994, 187217). In
8 their review on the role of metabolic changes in plant redox status after O3 exposure, Dizengremel et
9 al. (2009, 199424) summarized multiple studies in which several different tree species were exposed
10 to O3 concentrations ranging from ambient to 200 ppb O3 for at least several weeks. In all cases, the
11 activity of enzymes, including phosphofructokinase, pyruvate kinase and fumarase, which are part of
12 several catabolic pathways, were increased in response to O3 exposure.
13 Photorespiration is a light-stimulated process which consumes O2 and releases CO2. While it
14 has been regarded as a wasteful process, more recent evidence suggests that it may play a role in
15 photoprotection during photosynthesis (Bagard et al., 2008, 191593). The few studies that have been
16 conducted on O3 effects on photorespiration suggest that rates of photorespiration decline
17 concomitantly with rates of photosynthesis. Soybean plants were exposed to ambient (daily averages
18 43-58 ppb) and 1.5 ambient O3 (daily averages 63-83 ppb) O3 in OTCs for 12 h/day for 4 months.
19 Rates of photosynthesis and photorespiration and photorespiratory enzyme activity declined only at
20 the end of the growing season and did not appear to be very sensitive to O3 exposure (Booker et al.,
21 1997, 026425). Young hybrid poplars exposed to 120 ppb O3 for 13 h/day for 35 days in phytotron
22 chambers showed that effects on photorespiration and photosynthesis were dependent upon the
23 developmental stage of the leaf. While young leaves were not impacted, reductions in photosynthesis
24 and photorespiration were measured in fully expanded leaves (Bagard et al., 2008, 191593).
9.4.5.3. Secondary Metabolism
25 Transcriptome analysis of Arabidopsis plants has revealed modulation of several genes
26 involved in plant secondary metabolism (Ludwikow and Sadowski, 2008, 191426). Phenylalanine
27 ammonia lyase (PAL) has been the focus of many studies involving plant responses to O3 due to its
28 importance in linking the phenylpropanoid pathway of plant secondary metabolism to primary
29 metabolism in the form of the shikimate pathway. Genes encoding several enzymes of the
30 phenylpropanoid pathway and lignin biosynthesis were up-regulated in transcriptome analysis of
31 Arabidopsis plants (Col-0) exposed to 350 ppb O3 for 6 hours, while 2 genes involved in flavonoid
32 biosynthesis were down-regulated (Ludwikow et al., 2004, 595939). Exposure of Arabidopsis
33 (Col-0) to lower O3 concentrations (150 ppb for 8 h/day for 2 days) resulted in the induction of 11
34 transcripts involved in flavonoid synthesis. In their exposure of 2-year-old Mediterranean shrub
35 Phillyrea latifolia to 110 ppb O3 for 90 days, Paolacci et al. (2007, 191422) identified four clones
36 that were up-regulated and corresponded to genes involved in the synthesis of secondary
37 metabolites, such as isoprenoids, polyamines and phenylpropanoids. Up-regulation of genes
March 2011 9-42 DRAFT - DO NOT CITE OR QUOTE
-------
1 involved in isoprene synthesis was also observed mMedicago trunculata exposed to 300 ppb O3 for
2 6 hours, while genes encoding enzymes of the flavonoid synthesis pathway were either up- or down-
3 regulated (Puckette et al, 2008, 191698). Exposure of red clover to 1.5 x ambient O3 (average
4 concentrations of 32.4 ppb) for up to 9 weeks in an open field exposure system resulted in increases
5 in leaf total phenolic content. However, the types of phenolics that were increased in response to O3
6 exposure differed depending upon the developmental stage of the plant. While almost all of the 31
7 different phenolic compounds measured increased in quantity initially during the exposure, after
8 3 weeks the quantity of isoflavones decreased while other phenolics increased (Saviranta et al., 2010,
9 102177). Exposure of beech saplings to ambient and 2 x ambient O3 concentrations over 2 growing
10 seasons resulted in the induction of several enzymes which contribute to lignin formation, while
11 enzymes involved in flavonoid biosynthesis were down-regulated (Olbrich et al., 2009, 596020).
12 Exposure of tobacco Bel W3 to 160 ppb O3 for 5 hours showed up-regulation of almost all genes
13 encoding for enzymes which are part of the prechorismate pathway (Janzik et al., 2005, 191581).
14 Isoprenoids can serve as antioxidant compounds in plants exposed to oxidative stress (Paolacci et al.,
15 2007. 191422).
16 The prechorismate pathway is the pathway leading to the formation of chorismate, a precursor
17 to the formation of the aromatic amino acids tryptophan, tyrosine and phenylalanine. These amino
18 acids are precursors for the formation of many secondary aromatic compounds, and, therefore, the
19 prechorismate pathway represents a branch-point in the regulation of metabolites into either primary
20 or secondary metabolism (Janzik et al., 2005, 191581). Exposure of the O3 sensitive Bel W3 tobacco
21 cultivar at 160 ppb for 5 hours showed an increase in transcript levels of most of the genes encoding
22 enzymes of the prechorismate pathway. However, shikimate kinase (SK) did not show any change in
23 transcript levels and only one of three isoforms of DAHPS (3-deoxy-D-arabino-heptulosonat-7-
24 phosphate synthase), the first enzyme in this pathway, was induced by O3 exposure (Janzik et al.,
25 2005, 191581). Differential induction of DAHPS isoforms was also observed in European beech
26 after 40 days of exposure to 150-190 ppb O3. At this time point in the beech experiment, transcript
27 levels of shikimate pathway enzymes, including SK, were generally strongly induced after an only
28 weak initial induction after the first 40 days of exposure. Both soluble and cell-wall bound phenolic
29 metabolites showed only minimal increases in response to O3 for the duration of the exposure period
30 (Alonso et al., 2007, 199289). Total leaf phenolics decreased with leafage in Populus nigra exposed
31 to 80 ppb O3 for 12 h/day for 14 days. Ozone increased the concentration of total leaf phenolics in
32 newly expanded leaves, with the most significant increases occurring in compounds such as
33 quercitin glycoside, which has a high antioxidant capacity (Fares et al., 2010, 628522). While O3
34 exposure induced the activity of several phenylpropanoid pathway enzymes, the degree of induction
35 differed in the two poplar clones exposed to 60 ppb for 5 h/day for 15 days. In the tolerant 1-214
36 clone, PAL activity increased ninefold in O3-treated plants as compared to controls, while there was
37 no significant difference in PAL activity in the sensitive Eridano clone (Di Baccio et al., 2008,
38 199850).
March 2011 9-43 DRAFT - DO NOT CITE OR QUOTE
-------
1 Polyamines such as putrescine, spermidine and spermine play a variety of roles in plants and
2 have been implicated in plant defense responses to both abiotic and biotic stresses. They exist in both
3 a free form and conjugated to hydroxycinnamic acids. Investigations on the role of polyamines have
4 found that levels of putrescine increase in response to oxidative stress. This increase stems largely
5 from the increase in the activity of arginine decarboxylase (ADC), a key enzyme in the synthesis of
6 putrescine (Groppa and Benavides, 2008, 191616). Langebartels et al. (1991, 043506) described
7 differences in putrescine accumulation in O3-treated tobacco plants exposed to several O3
8 concentrations, ranging from 0-400 ppb for 5-7 hours. A large and rapid increase in putrescine
9 occurred in the tolerant Bel B cultivar and only a small increase in the Bel W3 cultivar, which
10 occurred only after the formation of necrotic leaf lesions. Van Buuren et al. (2002, 631199) further
11 examined the role of polyamines in these two tobacco cultivars during an acute (130 ppb O3 for 7 h
12 in a growth chamber) exposure. They found that while free putrescine accumulated in undamaged
13 tissue of both cultivars, conjugated putrescine predominantly accumulated in tissues undergoing cell
14 death after plant exposure to O3 (van Buuren et al., 2002, 631199). The authors suggest that while
15 free putrescine may not play a role in conferring tolerance in the Bel B cultivar, conjugated
16 putrescine may play a role in O3-induced programmed cell death in Bel W3 plants.
17 Isoprene is emitted by some plant species and represents the predominant biogenic source of
18 hydrocarbon emissions in the atmosphere (Guenther et al., 2006, 607080). In the atmosphere, the
19 oxidation of isoprene by hydroxyl radicals can enhance O3 formation in the presence of NOX,
20 thereby impacting the O3 concentration that plants are exposed to. While isoprene emission varies
21 widely between species, and it has been proposed to stabilize membranes and provide those plant
22 species that produce it with a mechanism of thermotolerance (Sharkey et al., 2008, 191492). It has
23 also been suggested that isoprene may act as an antioxidant compound to scavenge O3 (Loreto and
24 Velikova, 2001, 657212). Recent studies using a variety of plant species have shown conflicting
25 results in trying to understand the effects of O3 on isoprene emission. Acute doses of O3 (300 ppb for
26 3 h) stimulated isoprene emissions in detached leaves of Phragmites australis (Velikova et al., 2005,
27 199410). Similarly, isoprene emissions were stimulated in Populus nigra after exposure to 100 ppb
28 O3 for 5 days continuously (Fares et al., 2008, 191683). Isoprene emission in attached leaves of
29 Populus alba, which were exposed to 150 ppb O3 for 11 h/day for 30 days inside cuvettes, was
30 inhibited, while isoprene emission and transcript levels of isoprene synthase mRNA were increased
31 in the leaves exposed to ambient O3 (40 ppb), which were located above the leaves enclosed in the
32 exposure cuvettes (Fares et al., 2006, 191455). Exposure of 2 genotypes of hybrid poplar to 120 ppb
33 O3 for 6 h/day for 8 days resulted in a significant reduction in isoprene emission in the O3-sensitive
34 but not the tolerant genotype (Ryan et al., 2009, 191299). Similarly, O3 treatment (80 ppb 12 h/day
35 for 14 days) of Populus nigra showed that isoprene emission was reduced in the treated plants
36 relative to the control plants (Fares et al., 2010, 628522). Based on results of this and other studies,
37 Fares et al. (2010, 628522) concluded that the isoprenoid pathway may be induced in plants exposed
38 to acute O3 doses, while at lower doses isoprene emission may be inhibited. Vickers et al. (2009,
39 191497) developed transgenic tobacco plants with the isoprene synthase gene from Populus alba and
March 2011 9-44 DRAFT - DO NOT CITE OR QUOTE
-------
1 exposed them to 120 ppb O3 for 6 h/day for 2 days. They determined that the wildtype plants showed
2 significantly more O3 damage, including the development of leaf lesions and a decline in
3 photo synthetic rates, than the transgenic, isoprene-emitting plants. Transgenic plants also
4 accumulated less H2O2 and had lower levels of lipid peroxidation following exposure to O3 than the
5 wildtype plants (Vickers et al., 2009, 191497). These results indicate that isoprene may have a
6 protective role for plants exposed to oxidative stress.
9.4.6. Changes in Stomatal Function
7 There has been some debate as to whether O3-induced reductions in photosynthesis result from
8 changes in stomatal conductance or direct effects on C assimilation. A review of the literature
9 suggests there may be species-specific differences in how O3 affects photosynthesis through either
10 direct or indirect mechanisms. Recent studies utilizing a new simultaneous O3 exposure/gas
11 exchange device have demonstrated that exposure of Arabidopsis plants to 150 ppb O3 resulted in a
12 60-70% decline in stomatal conductance within 9-12 minutes of beginning the exposure. Twenty to
13 thirty minutes later, stomatal conductance had returned to its initial value, even with continuing
14 exposure to O3, indicating a rapid direct effect of O3 on stomatal function (Kollist et al., 2007,
15 191539). The contributions of stomatal versus non-stomatal factors in reducing net photosynthesis
16 have been investigated in a number of tree species. Ginkgo biloba trees exposed to ambient O3
17 (40 ppb) and elevated O3 (80 ppb) in OTCs for 90 days were investigated over the course of seven
18 months. Initially, stomatal limitation restricted C assimilation and, along with higher rates dark
19 respiration, served to protect against oxidative stress. However, as the season progressed, declines in
20 C assimilation resulted from direct oxidative damage to the photosynthetic apparatus (He et al.,
21 2007, 199789). In other studies focusing on O3 effects on adult trees, including Quercus mongolica
22 (exposed to 80 ppb O3 for 9 h/day for 4 months in OTCs) and Fagus sylvatica (exposed to the
23 2 x ambient concentrations, with a mean of 57 ppb O3 for 6 months in a free air fumigation system)
24 O3-induced reductions in stomatal conductance were accompanied by a decrease in intercellular CO2
25 concentration, indicating that stomatal closure did not result from high internal CO2 but rather from a
26 direct effect of O3 on the guard cells (Wang et al., 2009, 199303)(Kitao et al., 2009, 191331). In a
27 related study, Populus deltoides grown in rural areas showed a greater O3-induced decline in biomass
28 allocation than those grown in urban areas (with means in O3 concentrations ranging from 20-36 ppb
29 in urban areas and 36-47 ppb in rural areas). This decline in biomass was determined to result from
30 significantly higher rates of stomatal conductance in the rural trees, indicating a loss of stomatal
31 control (Gregg et al., 2006, 186961). An investigation of the differences between O3-sensitive and
32 insensitive cutleaf coneflower (Rudbeckia laciniata var. digitata) revealed differences in stomatal
33 behavior (Grulke et al., 2007, 186963). Cutleaf coneflower plants growing in Great Smoky Mountain
34 National Park (exposed to mean O3 concentration of 42 ppb from mid-May through mid-September)
35 showed variable stomatal responses to changes in environmental conditions, such as light and vapor
36 pressure deficit. It is likely that the O3-sensitive coneflower plants have a set of traits, such as a
March 2011 9-45 DRAFT - DO NOT CITE OR QUOTE
-------
1 sluggish stomatal response to changes in light intensity, which predispose them to being more
2 sensitive to O3 exposure (Grulke et al., 2007, 186963)(Paoletti and Grulke, 2010, 628561).
3 Feng et al. (2007, 191641) determined that reductions in net photosynthesis in O3-exposed
4 winter wheat (Triticum aestivum), exposed to a mean concentration of either 52 or 105 ppb O3 for
5 6 weeks in OTCs, likely occurred as a result of both stomatal and non-stomatal factors. Other studies
6 suggest that O3 impacts net photosynthesis only through non-stomatal factors. A direct effect of O3
7 on C fixation as described above would lead to a buildup of internal CO2, resulting in a reductions in
8 stomatal conductance through stomatal closure. In Acer opalus, a significant reduction in light-
9 saturated photosynthesis was accompanied by significant increases in internal CO2 concentration and
10 decreases in water use efficiency without appreciable changes in stomatal conductance (Calatayud et
11 al., 2007, 191411). A similar decrease in photosynthesis, accompanied by an increase in internal CO2
12 concentration, was measured in soybean plants exposed to O3. Singh et al. (2009, 199427) attributed
13 the declines in photosynthesis to direct damage to the photosynthetic apparatus rather than to
14 stomatal limitations.
9.5. Nature of Effects on Vegetation
9.5.1. Introduction
15 Ambient O3 concentrations have long been known to cause visible symptoms, decreases in
16 photosynthetic rates, decreases in growth, and decreases in the yield of plants (U.S. EPA, 1978,
17 040586)(U.S. EPA, 1986, 017607)01.5. EPA, 1996, 080828)01.5. EPA, 2006, 088089). Numerous
18 studies have related O3 exposure to plant responses, with most effort focused on the yield of crops
19 and the growth of tree seedlings. Most experiments exposed individual plants grown in pots or soil
20 under controlled conditions to known concentrations of O3 for a segment of daylight hours for some
21 portion of the plant's life span (Section 9.3). This section focuses on the responses of plants to
22 seasonal or multi-year exposures to known amounts of O3. Quantitative responses include changes in
23 growth and biomass allocation, changes in reproduction, onset of visible foliar injury, and changes in
24 leaf gas exchange. The response of a plant species or variety to O3 exposure depends upon many
25 factors, including biochemical and physiological status (Section 9.4), genetic characteristics, and
26 previous and current exposure to other stressors (Section 9.5.4.). Because of the available
27 information, most of this section focuses on the response of individual plants, especially tree
28 seedlings and crops, with limited discussion of mixtures of herbaceous species. Responses at the
29 ecosystem scale are discussed in Section 9.6.
30 This section will focus mainly on studies published since the release of the 2006 O3 AQCD
31 (U.S. EPA, 2006, 088089). However, because much O3 research was conducted prior to the 2006 O3
32 AQCD, the present discussion of vegetation response to O3 exposure is largely based on the
33 conclusions of the 1978, 1986, 1996, and 2006 O3 AQCDs (U.S. EPA, 1978, 040586)OJ.S. EPA,
34 1986, 017607)OJ.S. EPA, 1996, 080828)OJ.S. EPA, 2006, 088089).
March 2011 9-46 DRAFT - DO NOT CITE OR QUOTE
-------
9.5.2. Effects on Woody and Herbaceous Vegetation
9.5.2.1. Growth and Biomass Allocation
1 The previous O3 AQCDs concluded that there is strong evidence that exposure to O3 decreases
2 growth in numerous plant species. Studies published since the last review support those conclusions
3 and are summarized below.
4 In a recently published meta-analysis, Wittig et al. (2009, 191631) quantitatively compiled
5 peer reviewed studies from the past 40 years on the effect of current and future O3 exposures on the
6 physiology and growth of forest species. Wittig et al. (2009, 191631) reported that current ambient
7 O3 concentrations (-40 ppb) significantly decreased annual total biomass growth (7%) across 263
8 studies. However, this effect could be greater (11 to 17%) in areas that have higher O3 concentrations
9 and as background O3 increases in the future (Wittig et al., 2009, 191631). This meta-analysis
10 demonstrates the coherence of O3 effects across numerous studies and species using a variety of
11 experimental techniques.
12 In two companion papers, McLaughlin et al. (2007, 090348)(2007. 090347) investigated the
13 effects of ambient O3 on tree growth and hydrology at forest sites in the southern Appalachian
14 Mountains. The authors reported the cumulative effects of ambient levels of O3 decreased seasonal
15 stem growth by 30-50% for most trees species in a high O3 year in comparison to a low O3 year
16 (McLaughlin et al., 2007, 090348). The authors also report that high ambient O3 concentrations can
17 disrupt whole-tree water use and in turn reduce late-season streamflow (McLaughlin et al., 2007,
18 090347); see Section 9.6.3 for more on water cycling.
19 Since the 2006 O3 AQCD, several new studies based on the Aspen FACE "free air" O3 and
20 CO2 exposure experiment in a forest in Wisconsin were published (Darbah et al., 2007,
21 093288)(Darbah et al., 2008, 196890)(King et al., 2005, 191701)(Kubiske et al., 2006,
22 093284)(Kubiske et al., 2007, 191336)(Riikonen et al., 2008, 191258). King et al. (2005, 191701)
23 found that O3 fumigation over the first seven years of stand development reduced total biomass
24 relative to the control by 23, 13, and 14% in the aspen, aspen-birch, and aspen-maple communities,
25 respectively. Over the same time period, Kubiske et al. (2006, 093284) observed that elevated O3
26 decreased tree heights, diameters, and main stem volumes in the aspen community by 11, 16, and
27 20%, respectively. In addition, Kubiske et al. (2007, 191336) reported that elevated O3 may change
28 the intra- and inter-species competition. For example, O3 treatments increased the rate of conversion
29 from a mixed aspen-birch community to a birch dominated community. In another study at this site,
30 Percy et al. (2007, 093287) suggested that negative growth effects were seen below the previous 8-h
31 O3 standard level of 0.084 ppm, but the informativeness of the study was diminished by severe
32 methodological problems.
33 Several studies at the Aspen FACE site also considered other growth-related effects of elevated
34 O3. Darbah et al. (2007, Q93288)(2008, 196890) reported that O3 treatments decreased paper birch
3 5 seed weight and seed germination and that this would likely lead to a negative impact of
March 2011
9-47
DRAFT - DO NOT CITE OR QUOTE
-------
1 regeneration for that species. Riikonen et al. (2008, 191258) found that elevated O3 decreased the
2 amount of starch in birch buds by 16%, and reduced aspen bud size, which may have been related to
3 the observed delay in spring leaf development. The results suggest that elevated O3 concentrations
4 have the potential to alter C metabolism of overwintering buds, which may have carry-over effects in
5 the subsequent growing season (Riikonen et al., 2008, 191258).
6 Effects on growth of understory vegetation were also investigated at Aspen FACE. Bandeff et
7 al. (2006, 191733) found that the effects of elevated CO2 and O3 on understory species composition,
8 total and individual species biomass, N content, and 15N recovery were a result of overstory
9 community responses to those treatments; however, there were no apparent direct treatment effects
10 due to high variability of the data. Total understory biomass increased with increasing light and was
11 greatest under the open canopy of the aspen/maple community, as well as the more open canopy of
12 the elevated O3 treatments (Bandeff et al., 2006, 191733). Similarly, data from a study by Awmack et
13 al. (2007, 191415) suggest that elevated CO2 and O3 may have indirect growth effects on red
14 (Trifolium pratense) and white (Trifolium repens) clover in the understory via overstory community
15 effects; however, no direct effects of elevated O3 were observed.
16 Overall, the studies at the Aspen FACE experiment are consistent with many of the OTC
17 studies that were the foundation of previous O3 NAAQS reviews. These results strengthen our
18 understanding of O3 effects on forests and demonstrate the relevance of the knowledge gained from
19 trees grown in open-top chamber studies.
20 For some annual species, particularly crops, the endpoint for an assessment of the risk of O3
21 exposure can be defined as yield or growth, e.g., production of grain. For plants grown in mixtures
22 such as hayfields, and natural or semi-natural grasslands (including native nonagricultural species),
23 endpoints other than production of biomass may be important. Such endpoints include biodiversity
24 or species composition, and effects may result from competitive interactions among plants in mixed-
25 species communities. Most of the available data on non-crop herbaceous species are for grasslands
26 with many of the recent studies conducted in Europe. See Section 9.6.5 for a review of the recent
27 literature on O3 effects on competition and biodiversity in grasslands.
Root Growth
28 Although O3 does not penetrate soil, it could alter root development by decreasing
29 C assimilation via photosynthesis (Andersen, 2003, 041673). The response of root development to
30 O3 exposure depends on available photosynthate and could vary over time. Many biotic and abiotic
31 factors, such as community dynamics and drought stress, have been found to affect root production
32 under elevated O3. An earlier study at the AspenFACE experiment found that elevated O3 reduced
33 coarse root and fine roots biomass in young stands of paper birch and trembling aspen (King et al.,
34 2001, 041751). However, this reduction disappeared several years later. Ozone significantly
35 increased fine-root (<1.0 mm) in the aspen community (Pregitzer et al., 2008, 191677). This increase
36 in fine root production was due to changes in community composition, such as better survival of the
37 O3-tolerant aspen genotype, birch, and maple, rather than changes in C allocation at the individual
March 2011 9-48 DRAFT - DO NOT CITE OR QUOTE
-------
1 tree level (Pregitzer et al., 2008, 191677)(Zak et al., 2007, 191239). In an adult European
2 beech/Norway spruce forest in Germany, drought was found to nullify the O3-driven stimulation of
3 fine root growth. Ozone stimulated fine-root production of beech during the humid year, but had no
4 significant impact on fine root production in the dry year (Nikolova et al., 2010, 626810)(Matyssek
5 etal.. 2010. 628553).
6 Using a non-destructive method, Vollsnes et al. (2010, 625576) studied the in vivo root
7 development of subterranean clover (Trifolium subterraneum) before, during and after short-term O3
8 exposure. It was found that O3 reduced root tip formation, root elongation, the total root length, and
9 the ratios between below- and above-ground growth within one week after exposure. Those effects
10 persisted for up to three weeks; however, biomass and biomass ratios were not significantly altered
11 at the harvest five weeks after exposure.
12 Ozone has been shown to have negative (Jones et al., 2010, 567354). non-significant (Phillips
13 et al., 2009,199802)(Andersen et al., 2010, 628559) and positive effects (Pregitzer et al., 2008,
14 191677)(Grebenc and Kraigher, 2007, 191265) on root biomass and root: shoot ratio. While the
15 findings of individual studies were mixed, several recent meta-analyses have generally indicated that
16 O3 reduced C allocated to roots. In one meta-analysis, Grantz et al. (2006, 191545) estimated the
17 effect of O3 on the root:shoot allometric coefficient (k), the ratio between the relative growth rate of
18 the root and shoot. The results showed that O3 reduced k by 5.6%, and the largest decline was
19 observed in slow-growing plants. In another meta-analysis including 263 publications, Wittig et al.
20 (2009, 191631) found that current O3 exposure had no significant impacts on root biomass and
21 root:shoot ratio when compared to pre-industrial O3 exposure. However, if O3 concentrations rose to
22 81-101 ppb (projected O3 levels in 2100), both root biomass and root:shoot ratio were found to
23 significantly decrease. Gymnosperms and angiosperms differed in their responses, with
24 gymnosperms being less sensitive to elevated O3. In two other meta-analyses, Wang et al. (2010,
25 387478) found elevated O3 reduced biomass allocation to roots by 8.3% at ambient CO2 and 6.0% at
26 elevated CO2, and Morgan et al. (2003, 055527) found O3 reduced root dry weight of soybean.
9.5.2.2. Reproduction
27 Studies during recent decades have demonstrated O3 effects on different stages of plant
28 reproduction. The impacts of O3 on reproductive development, as reviewed by Black et al. (2000,
29 036322). can occur by influencing (1) age at which flowering occurs, particularly in long-lived trees
30 that often have long juvenile periods of early growth without flower and seed production; (2) flower
31 bud initiation and development; (3) pollen germination and pollen tube growth; and (4) seed, fruit, or
32 cone yields and seed quality (Table 9-2) (U.S. EPA, 2006,
33 Several recent studies since the 2006 O3 AQCD further demonstrate the effects of O3 on
34 reproductive processes in herbaceous and woody plant species. Ramo et al. (2007, 191441) exposed
35 several meadow species to elevated O3 (40-50 ppb) and CO2 (+100 ppm), both individually and
36 combined, over three growing seasons in ground-planted mesocosms, using OTCs. Elevated O3
March 2011 9-49 DRAFT - DO NOT CITE OR QUOTE
-------
1 delayed flowering of Campanula rotundifolia and Vicia cracca. Ozone also reduced the overall
2 number of produced flowers and decreased fresh weight of individual Fragaria vesca berries.
3 Black et al. (2007, 191558) exposed Brassica campestris to 70 ppb for two days during late
4 vegetative growth or ten days during most of the vegetative phase. The two-day exposure had no
5 effect on growth or reproductive characteristics, while the 10 day exposure reduced vegetative
6 growth and reproductive site number on the terminal raceme, emphasizing the importance of
7 exposure duration and timing. Mature seed number and weight per pod were unaffected due to
8 reduced seed abortion, suggesting that, although O3 affected reproductive processes, indeterminate
9 species such as B. campestris possess enough compensatory flexibility to avoid reduced seed
10 production (Black et al.. 2007. 191558).
11 In the determinate species, Plantago major, Black et al. (2010, 625575) found that O3 can
12 have direct effects on reproductive development in populations of differing sensitivity. Only the first
13 flowering spike was exposed to 120 ppb O3 for 7 hours per day on 9 successive days (corresponding
14 to flower development) while the leaves and second spike were exposed to charcoal-filtered air.
15 Exposure of the first spike to O3 affected seed number per capsule on both spikes even though spike
16 two was not exposed. The combined seed weight of spikes one and two was increased by 19% in the
17 two resistant populations, suggesting an overcompensation for injury; whereas, a decrease of 21%
18 was observed in the most sensitive population (Black et al., 2010, 625575).
19 A study by Darbah et al. (2007, 093288)(2008. 196890) of paper birch (Betula papyrifera)
20 trees at the Aspen FACE site in Rhinelander, WI investigated the effects of elevated O3 and/or CO2
21 on reproductive fitness. Elevated O3 increased flowering, but decreased seed weight and germination
22 success rate of seeds from the exposed trees. These results suggest that O3 can dramatically affect
23 flowering, seed production, and seed quality of paper birch, ultimately affecting its reproductive
24 fitness (Darbah et al., 2007, 093288)(Darbah et al., 2008, 196890).
March 2011 9-50 DRAFT - DO NOT CITE OR QUOTE
-------
Table 9-2. Ozone effects on plant reproductive processes (derived from Table AX9-22 of the 2006
ozone AQCD)
Species
Apocynun androsaemifolium
Buddleia davidii
Rubuscuneifolius
Plantago major
Fragaria * ananassa
Plantago major
Understory herbs
Condition Measures
Flowering time
Flowering time
Pollen germination
Pollen tube elongation
Fruit yield
Seed yield
Seed yield
References
Bergweilerand Manning (1999, 036321)
Findleyetal. (1997,036337)
Chappelka (2002, 0529461
Stewart (1998, 094233)
Drogoudi and Ashmore (2001, 052959); Drogoudi and Ashmore
(2000, 042517)
Lyons and Barnes (1 998, 036354); Pearson et al. (1
Reiling and Davison (1992, 0435911: Whitfield et al.
Harward and Treshow (1975, 038366)
996, 053074);
(1997.0363971
Source: Adapted from 200603 AQCD (U.S. EPA, 2006, 0880891
9.5.2.3. Visible Foliar Injury
1 Visible foliar injury resulting from exposure to O3 has been well characterized and
2 documented over several decades on many tree, shrub, herbaceous, and crop species (U.S. EPA,
3 2006, 088089KU.S. EPA, 1996, 080827)(U.S. EPA, 1984, 029711)01.5. EPA, 1978, 040586).
4 Visible foliar injury symptoms are considered diagnostic as they have been verified experimentally
5 in exposure-response studies, using exposure methodologies such as CSTRs, OTCs, and free-air
6 fumigation (see Section 9.3 for more detail on exposure methodologies). Several pictorial atlases and
7 guides have been published, providing details on diagnosis and identification of O3-induced visible
8 foliar injury on many plant species throughout North America (Penn State, 1987, 62697l)(Flagler.
9 1998, 025525) and Europe (Sanchez et al., 2001, 626980)(Sanchez et al., 2001, 626980)(Innes et al.,
10 2001, 048954). Typical visible injury symptoms on broad-leaved plants include: stippling, flecking,
11 surface bleaching, bifacial necrosis, pigmentation (e.g., bronzing), chlorosis, and/or premature
12 senescence. Typical visible injury symptoms for conifers include: chlorotic banding, tip burn,
13 flecking, chlorotic mottling, and/or premature senescence of needles. Although common patterns of
14 injury develop within a species, these foliar lesions can vary considerably between and within
15 taxonomic groups. Furthermore, the degree and extent of visible foliar injury development varies
16 from year to year and site to site (Chappelka et al., 2007, 093290)(Orendovici-Best et al., 2008,
17 196940)(Smith et al., 2003, 044183). even among co-members of a population exposed to similar O3
18 levels, due to the influence of co-occurring environmental and genetic factors. Nevertheless,
19 Chappelka et al. (2007, 093290) reported that the average incidence of O3-induced foliar injury was
20 73% on milkweed in the Great Smokey Mountain National Park in the years 1992-1996.
21 Although the majority of O3-induced visible foliar injury occurrence has been observed on
22 seedlings and small plants, many studies have reported visible injury of mature coniferous trees,
23 primarily in the western U.S. (Arbaugh et al., 1998, 040297) and to mature deciduous trees in eastern
24 North America (Chappelka et al., 1999, 041860: Chappelka et al., 1999, 052952: Hildebrand et al.,
March 2011
9-51
DRAFT - DO NOT CITE OR QUOTE
-------
1 1996, 042670: Schaub et al, 2005, 191608: Somers et al, 1998, 044010: Vollenweider et al., 2003,
2 051157).
3 It is important to note that visible foliar injury occurs only when sensitive plants are exposed
4 to elevated O3 concentrations in a predisposing environment. A major confounding factor for O3-
5 induced visible foliar injury is the amount of soil moisture available to a plant during the year that
6 the visible foliar injury is being assessed. This is because lack of soil moisture generally decreases
7 stomatal conductance of plants and, therefore, limits the amount of O3 entering the leaf that can
8 cause injury (Grulke et al., 2003, 042637)(Matyssek et al., 2006, 191481)(Panek. 2004,
9 Q79202)(Panek and Goldstein, 2001, 030190¥Temple et al., 1992, 043349)(Temple et al., 1988,
10 043237). Consequently, many studies have shown that dry periods in local areas tend to decrease the
11 incidence and severity of O3-induced visible foliar injury; therefore, the incidence of visible foliar
12 injury is not always higher in years and areas with higher O3, especially with co-occurring drought
13 (Smith et al., 2003, 044183). Other factors such as leafage influence the severity of symptom
14 expression with older leaves showing greater injury severity (Zhang et al., 2010, 628555).
15 Although visible injury is a valuable indicator of the presence of phytotoxic concentrations of
16 O3 in ambient air, it is not always a reliable indicator of other negative effects on vegetation. The
17 significance of O3 injury at the leaf and whole plant levels depends on how much of the total leaf
18 area of the plant has been affected, as well as the plant's age, size, developmental stage, and degree
19 of functional redundancy among the existing leaf area. Previous O3 AQCDs have noted the difficulty
20 in relating visible foliar injury symptoms to other vegetation effects such as individual plant growth,
21 stand growth, or ecosystem characteristics (U.S. EPA, 2006, 088089)01.5. EPA, 1996,
22 080827)(U.S. EPA, 1996, 080827). As a result, it is not presently possible to determine, with
23 consistency across species and environments, what degree of injury at the leaf level has significance
24 to the vigor of the whole plant. However, in some cases, visible foliar symptoms have been
25 correlated with decreased vegetative growth (Benoit et al., 1982, 039778: Karnosky et al., 1996,
26 036347: Peterson et al., 1987, 042148: Somers et al., 1998, 044010) and with impaired reproductive
27 function (Black et al., 2000, 036322: Chappelka, 2002, 052946). Conversely, the lack of visible
28 injury does not always indicate a lack of phytotoxic concentrations of O3 or a lack of non-visible O3
29 effects (Gregg et al., 2003, 046996: Gregg et al., 2006, 186961).
Biomonitoring
30 The use of biological indicators to detect phytotoxic levels of O3 is a longstanding and
31 effective methodology (Chappelka and Samuelson, 1998, 093687: Manning and Krupa, 1992,
32 044155). A plant bioindicator can be defined as a vascular or nonvascular plant exhibiting a typical
33 and verifiable response when exposed to a plant stress such as an air pollutant (Manning, 2003,
34 053773). To be considered a good indicator species, plants must (1) exhibit a distinct, verified
35 response; (2) have few or no confounding disease or pest problems; and (3) exhibit genetic stability
36 (U.S. EPA, 2006, 088089). Such sensitive plants can be used to detect the presence of a specific air
37 pollutant such as O3 in the ambient air at a specific location or region and, as a result of the
March 2011 9-52 DRAFT - DO NOT CITE OR QUOTE
-------
1 magnitude of their response, provide unique information regarding specific ambient air quality.
2 Bioindicators can be either introduced sentinels, such as the widely used tobacco (Nicotiana
3 tabacum) variety Bel W3 (Calatayud et al, 2007, 191568: Gombert et al, 2006, 089309: Heggestad,
4 1991, 042533: Kostka-Rick and Hahn, 2005, 199428: Laffray et al., 2007, 191384: Nali et al., 2007,
5 191478) or detectors, which are sensitive native plant species (e.g., tall milkweed [Asclepias
6 exaltata]; Chappelka et al., 2007, 093290: Souza et al., 2006, 191658V The approach is especially
7 useful in areas where O3 monitors are not operated (Manning, 2003, 053773). For example, in
8 remote wilderness areas where instrument monitoring is generally not available, the use of
9 bioindicator surveys in conjunction with the use of passive samplers (Krupa et al., 2001, 040430)
10 may be a useful methodology (Manning, 2003, 053773). However, the method requires expertise or
11 training in recognizing those signs and symptoms uniquely attributable to exposure to O3 as well as
12 in their quantitative assessment.
13 Since the 2006 O3 AQCD (U.S. EPA, 2006, 088089). new sensitive plant species have been
14 identified from field surveys and verified in controlled exposure studies (Kline et al., 2008, 191591:
15 Kline et al., 2009, 196918). Several multiple-year field surveys have also been conducted at National
16 Wildlife Refuges in Maine, Michigan, New Jersey, and South Carolina (Davis, 2007, 093291: Davis,
17 2007, 093292: Davis, 2009, 199287: Davis and Orendovici, 2006, 093293).
18 The USDA Forest Service through the Forest Health Monitoring Program (FHM) (1990 -
19 2001) and currently the Forest Inventory and Analysis (FIA) Program has been collecting data
20 regarding the incidence and severity of visible foliar injury on a variety of O3 sensitive plant species
21 throughout the U.S. (Coulston et al., 2003, 041871)(Smith et al., 2003, 044183). The plots where
22 these data are taken are known as biosites. These biosites are located throughout the country and
23 analysis of visible foliar injury within these sites follows a set of established protocols. For more
24 details, see http://www.nrs.fs.fed.us/fia/topics/ozone/ (USDA, 2011, 677550). The network has
25 provided evidence of O3 concentrations high enough to induce visible symptoms on sensitive
26 vegetation. From repeated observations and measurements made over a number of years, specific
27 patterns of areas experiencing visible O3 injury symptoms can be identified. Coulston et al. (2003,
28 041871) used information gathered over a 6-year period (1994-1999) from the network to identify
29 several species that were sensitive to O3 over a regional scale including sweetgum (Liquidambar
30 styraciflua), loblolly pine (Pinus taeda), and black cherry (Prunus serotina). In a study of the west
31 coast of the U.S, Campbell et al. (2007, 602360) reported O3 injury in 25-37% of biosites in
32 California forested ecosystems from 2000-2005.
33 A study by Kohut (2007, 093289) assessed the risk of O3-induced visible foliar injury on
34 bioindicator plants (NPS, 2006, 677536) in 244 national parks in support of the National Park
35 Service's Vital Signs Monitoring Network (NPS, 2007, 677537). The risk assessment was based on a
36 simple model relating response to the interaction of the plant, the level of O3 exposure, and the
37 exposure environment. Kohut (2007, 093289) concluded that the risk of visible foliar injury was
38 high in 65 parks (27%), moderate in 46 parks (19%), and low in 131 parks (54%). Some of the well-
39 known parks with a high risk of O3-induced visible foliar injury include Gettysburg, Valley Forge,
March 2011 9-53 DRAFT - DO NOT CITE OR QUOTE
-------
1 Delaware Water Gap, Cape Cod, Fire Island, Antietam, Harpers Ferry, Manassas, Wolf Trap Farm
2 Park, Mammoth Cave, Shiloh, Sleeping Bear Dunes, Great Smoky Mountains, Joshua Tree, Sequoia
3 and Kings Canyon, and Yosemite.
9.5.2.4. Leaf Gas Exchange
4 In general, there is strong experimental evidence over several decades of research that
5 exposure to O3 reduces photosynthesis and alters stomatal conductance in a wide variety of plant
6 species (Wittig et al, 2007, 191695) and these effects have been discussed in detail in previous O3
7 AQCDs (U.S. EPA, 2006, 088089)01.5. EPA, 1996, 080827). Recent studies related to these effects
8 are discussed in several sections within this document; therefore, this section refers to those sections
9 for a more in depth discussion.
10 Ozone effects on photosynthesis were presented Section 9.4.5.1, with a focus on recent studies
11 on the effects of O3 on light reactions. Changes in stomatal function in response to O3 exposure were
12 discussed in Section 9.4.6. In addition, the implications of changes in stomatal control for water
13 cycling are discussed in Section 9.6.3. Leaf gas exchange as it relates to effects based air quality
14 exposure indices and dose modeling is discussed throughout Section 9.7.
9.5.3. Agricultural Crops
15 The detrimental effect of O3 on crop production has been recognized since the 1960's and a
16 large body of research has stemmed from that recognition. Previous O3 AQCDs have extensively
17 reviewed this body of literature (U.S. EPA, 2006, 088089). Table 9-3 summarizes recent
18 experimental studies of O3 effects on agricultural crops, exclusive of growth and yield. Growth and
19 yield results are summarized in Table 9-16.
20 Ozone diffuses into the leaf apoplast via the stomata where it is rapidly converted into other
21 ROS that signal a diverse metabolic response (Kangasjarvi et al., 2005, 180341)(Long and Naidu,
22 2002, 038179). The mechanism of O3 sensing and cellular response is detailed in Section 9.4.3.1.
23 Ozone stress has been characterized as either acute or chronic, depending on the O3 concentration
24 and the exposure duration (Fiscus et al., 2005, 079155). While the actual concentration and duration
25 threshold for O3 damage varies from species to species and sometimes even among genotypes of the
26 same species (Ariyaphanphitak et al., 2005, 191349)(Biswas et al., 2008, 191428)(Dalstein and Vas,
27 2005, 191570)(Guidi et al., 2009,199825)(Keutgen et al., 2005, 191295)(Sawada and Kohno, 2009,
28 199426). it is commonly accepted that acute damage results from a very high concentration of O3
29 (>150 ppb) over a short period of time, and chronic O3 damage results from a lower concentration of
30 exposure over a longer period of time. In general, acute O3 damage has been well characterized and
31 mimics the biochemical defense response of plants to pathogen attack (Kangasjarvi et al., 2005,
32 18034l)(Overmyer et al., 2003, 053537). In contrast, the mechanism leading to chronic O3 damage
33 is less well-characterized but hallmark physiological symptoms include: decreased photosynthetic
34 productivity, decreased Rubisco activity and chlorophyll content, lower stomatal conductance, leaf
March 2011 9-54 DRAFT - DO NOT CITE OR QUOTE
-------
1 chlorosis, accelerated senescence and a general decrease in green leaf area and plant productivity
2 (Ashmore et al., 2006, 191557). Despite the knowledge gap pertaining to the mechanism of O3
3 damage, a number of comprehensive reviews and meta-analyses have recently been published
4 discussing both the current understanding of the quantitative effects of O3 concentration on a variety
5 of crop species and the potential focus areas for biotechnological improvement to a future growing
6 environment that will include higher O3 concentrations (Ainsworth, 2008, 191646)(Booker et al.,
7 2009,191569)(Feng et al., 2008,191451)(Grantz et al., 2006, 191545)(Haves et al., 2007,
8 196911)(Mills et al., 2007, 180221)(Morgan et al., 2003, 055527)(Van Dingenen et al., 2009,
9 199765). Since the 2006 O3 AQCD, exposure-response indices for a variety of crops have been
10 suggested (Mills et al., 2007, 180221) and many reports have investigated the effects of O3
11 concentration on seed or fruit quality to extend the knowledge base beyond yield quantity. This
12 section will outline the key findings from these papers as well as highlight some of the recent
13 research addressing the endpoints such as yields and crop quality.
14 Genetic variability is not the only factor that determines the crop response to O3-damage.
15 Ozone concentrations throughout a growing-season is not homogeneous and other environmental
16 conditions, such as elevated CO2 concentrations, drought, cold or nutrient availability may alleviate
17 or exacerbate the oxidative stress response to a given O3 concentration. This section will also
18 highlight recent literature that focuses on O3 damage to crops as influenced by other environmental
19 factors.
9.5.3.1. Yield
20 It is well known that yield is negatively impacted in many crop species in response to high O3
21 concentrations. However the threshold for damage varies from species to species. Reproductive
22 organs such as seeds may be particularly sensitive to injury or biomass reductions due to O3, as
23 reviewed by Black et al. (2000, 036322). Numerous analyses of experiments conducted in OTCs and
24 with naturally occurring gradients demonstrate that the effects of O3 exposure vary depending on the
25 growth stage of the plant. Plants grown for seed or grain are often most sensitive to exposure during
26 the seed or grain-filling period (Lee et al., 1988, 594572)(Pleijel et al., 1998, 053021)(Soia et al.,
27 2000, Q30388)(Younglove et al., 1994, 044162). AX9.5.4.1 of the 2006 O3 AQCD summarized many
28 previous studies on crop yield (U.S. EPA, 2006, 088089).
29 The effect of O3 exposure on U.S. crops remains an important area of research and several
30 studies have been published on this topic since the 2006 O3 AQCD (Tables 9-3 and 9-16). For
31 example, one study with cotton in a crop-weed interaction study (Grantz and Shrestha, 2006,
32 191702) utilizing OTCs suggests that ambient O3 concentrations (12-h avg: 79.9 ppb) decreased
33 cotton biomass by 25% and 1.5 x ambient O3 concentration (12-h avg: 122.7 ppb) decreased cotton
34 biomass by 75% compared to charcoal filtered control (12-h avg: 12.8 ppb). Further, this study
35 suggests that the weed, yellow nutsedge, was less sensitive to increasing O3 concentration which
36 would increase weed competition (Grantz and Shrestha, 2006, 191702). In a study of peanuts in
37 North Carolina, near ambient and elevated exposures of O3 reduced photosynthesis and yield
March 2011
9-55
DRAFT - DO NOT CITE OR QUOTE
-------
1 compared to very low O3 conditions (Booker et al., 2007, 191370)(Burkey et al., 2007, 191371). In
2 another study, Grantz and Vu (2009, 195237) reported that sugarcane biomass growth significantly
3 declined under O3 exposure.
4 The average yield loss reported across a number of meta-analytic studies have been published
5 recently for soybean (Morgan et al., 2003, 055527). wheat (Feng et al., 2008, 191453). rice
6 (Ainsworth, 2008, 191646). semi-natural vegetation (Hayes et al., 2007, 196911). potato, bean and
7 barley (Feng and Kobayashi, 2009, 199223). The meta-analytic technique allows for the objective
8 development of a quantitative consensus of the effects of a treatment across a wide body of literature
9 and therefore provides an average response ratio compiled from these sources. Further, this
10 technique allows for a compilation of data across a range of O3-fumigation techniques, durations and
11 concentrations in order to assemble the existing literature in a meaningful manner.
12 Morgan et al. (2003, 055527) reported an average seed yield loss for soybean of 24%
13 compared to charcoal filtered air across all O3 concentrations used in the 53 compiled studies. The
14 decrease in seed yield appeared to be the product of nearly equal decreases (7-12%) in seed weight,
15 seed number and pod number. As would be expected, the lowest O3 concentration (30-59 ppb)
16 resulted in the smallest yield losses, approximately 8%, while the highest O3 concentration
17 (80-120 ppb ) resulted in the largest yield losses, approximately 35% (Morgan et al., 2003, 055527).
18 Further, the oil/protein ratio within the soybean seed was altered due to growth at elevated O3
19 concentrations, with a decrease in oil content. The studies included in this meta-analysis all used
20 enclosed fumigation systems or growth chambers which may have altered the coupling of the
21 atmosphere to the lower plant canopy (McLeod and Long, 1999, 688834). Utilizing the Soybean
22 Free Air gas Concentration Enrichment Facility (SoyFACE; www.soyface.illinois.edu). Morgan et al.
23 (2006, 079186) report a 20% seed yield loss due to a 23% increase in average daytime O3
24 concentration (56-69 ppb) within a single soybean cultivar across two growing seasons in Illinois,
25 supporting the results from the meta-analysis. A further breakdown of the effects of current O3
26 concentrations (AOT40 of 4.7 ppm-h) on bean seed quality (Phaseolus vulgaris) has identified that
27 growth at current O3 concentrations compared to charcoal-filtered air raised total lipids, total crude
28 protein and dietary fiber content (Iriti et al., 2009, 195635). An increase in total phenolics was also
29 observed, however the individual phenolics compounds responded differently, with significant
30 decreases in anthocyanin content. The seeds from ambient O3 exposed plants also displayed
31 increased total antioxidant capacity compared to charcoal-filtered air controls (Iriti et al., 2009,
32 195635). Betzelberger et al. (2010, 644183) has recently utilized the SoyFACE facility to compare
33 the impact of elevated O3 concentrations across 10 soybean cultivars to investigate intraspecific
34 variability of the O3 response to find physiological or biochemical markers for eventual O3 tolerance
35 breeding efforts (Betzelberger et al., 2010, 644183). They report an average 17% decrease in yield
36 across all 10 cultivars across two growing seasons due to a doubling of ambient O3 concentrations,
37 with a the individual cultivar responses ranging from -7% to -36%. The dose-response functions
38 derived for these 10 modern cultivars were similar to the response functions derived from the
March 2011 9-56 DRAFT - DO NOT CITE OR QUOTE
-------
1 NCLAN studies conducted in the 1980's (Heagle, 1989, 093985) suggesting there has not been any
2 inadvertent selection for more O3-tolerant cultivars in recent history.
3 A meta-analysis has also been performed on studies investigating the effects of O3
4 concentrations on wheat (Feng et al, 2008, 191453). Across 23 studies included, elevated O3
5 concentrations (ranging from a 7-h daily average of 31-200 ppb) decreased grain yield by 29%.
6 Winter wheat and spring wheat did not differ in their responses; however the response in both
7 varieties to increasing O3 concentrations resulted in successively larger decreases in yield, from a
8 20% decrease in 42 ppb to 60% in 153 ppb O3. These yield losses were mainly caused by a
9 combination of decreases in individual grain weight (-18%), ear number per plant (-16%), and grain
10 number per ear (-11%). Further, the grain starch concentration decreased by 8% and the grain protein
11 yield decreased by 18% due to growth at elevated O3 concentrations as well. However, increases in
12 grain calcium and potassium levels were reported (Feng et al., 2008, 191453).
13 A recent meta-analysis found that growth at elevated O3 concentrations negatively impacts
14 nearly every aspect of rice performance as well (Ainsworth, 2008, 191646). While rice is not a major
15 crop in the U.S., it provides a staple food for over half of the global population (IRRI, 2002, 688833)
16 and the effects of rising O3 concentrations on rice yields merits consideration. On average, rice
17 yields decreased 14% in 62 ppb O3 compared to charcoal-filtered air. This yield loss was largely
18 driven by a 20% decrease in grain number (Ainsworth, 2008, 191646).
19 Feng and Kobayashi (2009, 199223) have recently compiled yield data for six major crop
20 species, potato, barley, wheat, rice, bean and soybean and grouped the O3 treatments used in those
21 studies into three categories: baseline O3 concentrations (<26 ppb), current ambient 7- or 12-h daily
22 O3 concentrations (31-50 ppb), and future ambient 7- or 12-h daily O3 concentrations (51-75 ppb).
23 Using these categories, they have effectively characterized the effects of current O3 concentrations
24 and the effects of future O3 concentrations compared to the baseline O3 concentrations. At current O3
25 concentrations, which ranged from 41-49 ppb in the studies included, soybean (-7.7%), bean
26 (-19.0%), barley (-8.9%), wheat (-9.7%), rice (-17.5%) and potato (-5.3%) all reported yield losses
27 compared to the baseline O3 concentrations (<26 ppb). At future O3 concentrations, averaging
28 63 ppb, soybean (-21.6%), bean (-41.4%), barley (-14%), wheat (-28%), rice (-17.5%) and potato
29 (-11.9%) all reported significantly larger yield losses compared to the losses at current O3
30 concentrations (<26 ppb) (Feng and Kobayashi, 2009, 199223).
31 An extensive review of OTC literature has determined the AOT40 critical level that causes a
32 5% yield reduction across a variety of agricultural and horticultural species (Mills et al., 2007,
33 180221). They classify the species into three groups: sensitive, moderate and tolerant. The sensitive
34 crops, including watermelon, beans, cotton, wheat, turnip, onion, soybean, lettuce, and tomato,
35 respond with a 5% reduction in yield under a 3-month AOT40 of 6 ppm-h. Watermelon was the most
36 sensitive with a critical level of 1.6 ppm-h. The moderately sensitive crops, including sugar beet,
37 oilseed rape, potato, tobacco, rice, maize, grape and broccoli, responded with a 5% reduction in yield
38 between 8.6 and 20 ppm-h. The crops classified as tolerant, including strawberry, plum and barley,
39 responded with a 5% yield reduction between 62-83.3 ppm-h (Mills et al., 2007, 180221).
March 2011 9-57 DRAFT - DO NOT CITE OR QUOTE
-------
1 Feng and Kobayashi (2009, 199223) compared their response-exposure results to those
2 published by Mills et al. (2007, 180221) and found the ranges of yield loss to be similar for soybean,
3 rice and bean. However, Feng and Kobayasi reported smaller yield losses for potato and wheat and
4 larger yield losses for barley compared to the dose-response functions published by Mills et al.
5 (2007, 180221). which they attributed to their more lenient criteria for literature inclusion.
6 While the studies investigating the impact of various O3 concentrations on yield are important
7 and aid in determining the vulnerability of various crops to a variety of O3 concentrations, there is
8 still uncertainty as to how these crops will respond under field conditions with interacting
9 environmental factors such as temperature, soil moisture, CO2 concentration, and soil fertility
10 (Booker et al., 2009, 191569). Further, there appears to be a distinct developmental and genotype
11 dependant influence on plant sensitivity to O3 that has yet to be fully investigated across O3
12 concentrations in a field setting.
13 Because O3 is heterogeneous in both time and space and O3 monitoring stations are
14 predominantly near urban areas, the O3 impacts on current crop yields are difficult to estimate.
15 Fishman et al. (2010, 644259) have used satellite observations to estimate O3 concentrations in the
16 contiguous tri-state area of Iowa, Illinois and Indiana and have combined that information with other
17 measured environmental variables to model the historical impact of O3 concentrations on soybean
18 yield across the 2002-2006 growing seasons. When soybean yield across Iowa, Indiana and Illinois
19 was modeled as a function of seasonal temperature, soil moisture and O3 concentrations, O3 had the
20 largest contribution to the variability in yield for the southern-most latitudes included in the dataset.
21 Fishman et al. (2010, 644259) determined that O3 concentrations significantly reduced soybean yield
22 by -0.38 to -1.63% ppb/v across the 5 years. This value is consistent with previous chamber studies
23 (NCLAN; Heagle, 1989, 093985) and results from SoyFACE (Morgan et al., 2006, 079186).
24 Satellite estimates of tropospheric O3 concentrations exist globally (Fishman et al., 2008, 193927).
25 therefore utilizing this historical modeling approach is feasible across a wider geographical area,
26 longer time-span and perhaps for more crop species.
9.5.3.2. Crop Quality
27 In general, it appears that increasing O3 concentrations above current ambient concentrations
28 can cause species dependant biomass losses, decreases in root biomass and nutritive quality,
29 accelerated senescence and shifts in biodiversity. A study conducted with highbush blackberry has
30 demonstrated decreased nutritive quality with increasing O3 concentration despite no change in
31 biomass among a charcoal-filtered control, an ambient O3 and a 2 x ambient O3 treatment (Ditchkoff
32 et al., 2009, 192230). A study conducted with sedge using control (30 ppb), low (55 ppb), medium
33 (80 ppb) and high (105 ppb) O3 treatments has demonstrated decreased root biomass and accelerated
34 senescence in the medium and high O3 treatments (Jones et al., 2010, 567354). Alfalfa showed no
35 biomass changes across two years of double ambient O3 concentrations (AOT40 of 13.9 ppm-h)
36 using FACE fumigation (Maggio et al., 2009, 191645). However a modeling study has demonstrated
37 that 84% of the relative feed value in high-yielding alfalfa was due to the variability in mean O3
March 2011 9-58 DRAFT - DO NOT CITE OR QUOTE
-------
1 concentration from 1998-2002 (Lin et al., 2007, 196925). Further, in a managed grassland FACE
2 system, the reduction in total biomass harvest over five years decreased twice as fast in the elevated
3 treatment (AOT40 of 13-59 ppm-h) compared to ambient (AOT40 of 1-20.7 ppm-h). Compared with
4 the ambient control, loss in annual dry matter yield was 23% after 5 year. Further, there was
5 significant changes in the functional categories growing in each plot with legumes showing the
6 strongest negative response (Volk et al., 2006, 191434). This study suggests a shift in biodiversity
7 away from nitrogen-fixers in managed grasslands. An OTC study conducted with Trifolium
8 subterraneum exposed to filtered ( <15 ppb), ambient, and 40 ppb above ambient O3 demonstrates
9 decreases in biomass in the highest O3 treatment as well as 10-20% decreased nutritive quality which
10 was mainly attributed to accelerated senescence (Sanz et al., 2005, 196963). A study conducted with
11 Eastern gamagrass and big bluestem in OTCs suggests that big bluestem is not sensitive to O3, but
12 gamagrass displayed decreased nutritive quality, due to higher lignin content and decreased N, in the
13 2 x ambient O3 treatment (Lewis et al., 2006, 191542).
March 2011 9-59 DRAFT - DO NOT CITE OR QUOTE
-------
Table 9-3. Summary of recent studies of ozone effects on crops (exclusive of growth and yield)
Species
Facility
Alfalfa (Medicago
sativa cv. Beaver)
Growth chambers
Bean
(Phaseolus vulgarisl.
cv Borlotto)
OTC, ground-planted
Curno, Italy
Big Blue Stem
(Andropogon gerardii)
OTC
Alabama, U.S.
Brassica napus
Growth chambers
Belgium
Brassica napus
cv. Westar
Growth chambers
Finland
Eastern Gamagrass
(Tripsacum
dactybides)
OTC
Alabama, U.S.
Lettuce
(Lactuca sativa)
OTC
Carcaixent
Experimental Station,
Spain
(Arachis hypogaea)
OTC
Raleigh, NC; U.S.
Poa pratensis
OTC
Braunschweig,
Germany
Potato
(Solanum tuberosum
OTC
Sweden & Finland
Exposure
Duration
1,2or
4 days
4 months
4 months
4 days
17-26
days
4 months
30 days
Syr
3yr;
4-5 wk
in the
spring
9 \/r
L yi
Ozone Exposure3
(Additional treatment)
3, 5or7h/day
85 ppb
(Exposure duration)
Seasonal AOT40:
CF = 0.5ppm-h;
Ambient = 4.6 ppm-h
(N/A)
12-h avg:
CF=14ppb;
Ambient = 29 ppb;
Elevated = 71 ppb
(N/A)
CF&176ppb
for 4 h/day
(N/A)
CF&IOOppb
(Bt/non-Bt;
herbivory)
12-h avg:
CF=14ppb;
Ambient = 29 ppb;
Elevated = 71 ppb
(N/A)
12-h mean:
CF= 10.2 ppb;
NF = 30.1 ppb;
NF+03= 62.7 ppb
(4 cultivars)
12-h avg:
CF = 22 ppb;
Ambient = 46 ppb;
Elevated = 75 ppb
(C02: 375 ppm; 548 ppm;
730 ppm)
CF+25 = 21.7ppb;
NF+50 = 73.1 ppb
(Competition)
CF=10ppb;
Ambient = 25 ppb);
Ambient(+) = (36 ppb);
Ambient(++) = (47 ppb)
(N/A)
Variable(s) measured
Relative feed value
Seed lipid,
Protein content
Fiber content
Relative feed value
Glucosinolates
VOC emissions
Relative feed value
Lipid peroxidation;
Root length
Harvest biomass
Relative feed value
[K], [Ca], [Mg], [P], [N] per dry weight
of tubers "dose-response regression,
report significant positive or negative
slope with increasing [03]
percent change from
CFb
(percent change from
ambient)
"high variability among
treatment groups (N/A)
+28.5 (N/A)
+7.88 (N/A)
+ 14.54 (N/A)
n.s. (n.s.)
-41 (N/A)
-30.7 (N/A);
-34 (N/A)
-17 (-12)
+77 (+38)
-22 (-14)
-40 (-10)
N/A (n.s.; -8)
[N] [P] [Ca] n.s.;
[K] & [Mg] sig +
(N/A)
Reference
Muntiferingetal.
(2006, 1912701
Iritietal. (2009,
1956351
Lewis et al. (2006,
1915421
Gielenetal. (2006,
1912711
Himanen etal.
(2009, 1913381
Lewis et al. (2006,
1915421
Calatayud et. al.
(2002, 6842221
Booker et al. (2007,
1913701
Bender etal. (2006,
1914371
Piikki et al. (2007,
191451)
March 2011
9-60
DRAFT - DO NOT CITE OR QUOTE
-------
Species
Facility
Location
Potato
(Solarium tuberosum
cv. Indira)
Climate chambers
Germany
Soybean
OTC
Italy
Soybean
(Glycine max
cv. 93B15)
SoyFACE
Urbana, IL; U.S.
Soybean
(Glycine max
cv. 93B15)
SoyFACE
Urbana, IL; U.S.
Soybean
(Glycine max
cv. Essex)
OTC, ground-planted
Raleigh, NC; U.S.
Soybean
(Glycine max
cv. Essex)
OTCs, 21 L pots
Raleigh, NC; U.S.
Soybean
(Glycine max)
lOcultivars)
SoyFACE
Urbana, IL; U.S.
Spring Wheat
(Triticum aestivum
cv. Minaret; Satu;
Drabant; Dragon)
OTCs
Belgium, Finland,
& Sweden
Strawberry
(Fragaria x ananassa
Duch. Cv. Korona
& Elsanta)
Growth chambers
Bonn, Germany
Sweet Potato
Growth Chambers
Bonn, Germany
Exposure
Duration
8wk
Syr
Syr
May-Oct
4 months
2yr
2x3
months
2yr
7yr
2 months
4wk
Ozone Exposure3
(Additional treatment)
CF=10ppb;
Ambient = 50 ppb;
2*Ambient= 100 ppb
(C02: 400 ppm &
700 ppm)
AOT40:
CF = 0 ppm-h;
Ambient = 3.4 ppm-h;
Elevated = 9.0 ppm-h
(Well-watered &
water-stressed)
AOT40:
Ambient = 5-22 ppm-h;
Elevated = 20-43 ppm-h
(C02: 550 ppm;
environmental variability)
8-h avg:
Ambient = 38.5 ppb;
Elevated = 52 ppb
(Herbivory)
12-havg:
CF = 21 ppb;
1.5*Ambient= 74 ppb
(C02: 370 ppm &
714 ppm)
12-havg:
CF=18ppb);
Elevated = 72 ppb)
(C02:367&718)
8-h avg (ppb):
Ambient = 46.3 & 37.9;
Elevated = 82.5 & 61. 3
(Cultivar comparisons)
Seasonal AOT40s
ranged from
Oto16ppm-h
(N/A)
8-h avg:
CF = 0 ppb;
Elevated = 78 ppb
(N/A)
8-h avg:
CF = 0 ppb;
Ambient < 40 ppb;
Elevated = 255 ppb
(N/A)
Variable(s) measured
Pathogen infestation using %
necrosis
Daily
evapotranspiration
Photosynthesis in new leaves,
Herbivory
defense-related
genes
Post-harvest residue
Water-use efficiency
Total antioxidant capacity
Seed protein content;
1 ,000-seed weight regressed across
all experiments
Total leaf area
Tuberweight
percent change from
CFb
(percent change from
ambient)
+52(n.s.)
-28 (-14)
N/A(n.s.)
N/A (N/A)
N/A (-15.46)
n.s. (N/A)
N/A (+19)
N/A (Significant negative
correlation)
N/A (Significant negative
correlation)
-16 (N/A)
-14 (-11. 5)
Reference
Plessl et al. (2007,
1969521
Jaude et al. (2008,
1912221
Bernacchietal.
(2006, 1580011
Casteel et al.
(2008, 1916961
Booker et al. (2005,
0791511
Booker et al.
(Booker etal.,
2004, 0791381
Betzelbergeretal.
(2010,6441831
Piikki et al. (2008,
1998121
Keutgen et al.
(2005, 1912951
Keutgen et al.
(2008, 1916901
March 2011
9-61
DRAFT - DO NOT CITE OR QUOTE
-------
Species
Facility
Tomato
(Lycopersicon
esculentum)
OTC
Valencia, Spain
Tritolium repens &
Trifolium pretense
Aspen FACE
Rhinelander, Wl; U.S.
Exposure Ozone Exposure3
Duration (Additional treatment)
8- mean:
CF=16.3ppb;
NF = 30.1 ppb;
133 days NF(+) = 83.2 ppb
(Various cultivars;
early & late harvest)
3-mo daylight avg:
3 months Ambient = 34.8 ppb;
1.2xAmbient= 42.23 ppb
(C02; 560 ppm)
Variable(s) measured
Brix degree
Lignin;
Dry-matter
digestibility
percent change from
CFb
(percent change from
ambient)
-7.2 (-3.6)
N/A(+19.3)
N/A(-4.2)
Reference
Calvo, et al. (2005,
1915701
Muntiferingetal.
(2006, 1912701
aOzone exposure in ppb unless otherwise noted.
CF = Carbon-filtered air.
NF = Non-filtered air.
1
2
3
4
5
6
1
9
10
11
12
13
14
15
16
17
18
19
9.5.4. Factors that Modify Functional and Growth Response
Many biotic and abiotic factors, including insects, pathogens, root microbes and fungi,
temperature, water and nutrient availability, and other air pollutants, as well as elevated CO2,
influence or alter plant response to O3. These modifying factors were comprehensively reviewed in
AX9.3 of the 2006 O3 AQCD (U.S. EPA, 2006, 088089) and thus, this section serves mainly as a
brief summary of the previous findings. A limited number of new studies published since the 2006
O3 AQCD add to our understanding of the role of these interactions in modifying O3-induced plant
responses. Many of these modifying factors and interactions are integrated into discussions
elsewhere in this chapter and the reader is directed to those sections.
9.5.4.1. Genetics
It is well known that species vary greatly in their responsiveness to O3. Even within a given
species, individual genotypes or populations can also vary significantly with respect to O3 sensitivity
(see section AX 9.3.2 of the 2006 AQCD; U.S. EPA, 2006, 088089). Therefore, caution should be
taken when considering a species' degree of sensitivity to O3. Plant response to O3 is determined by
genes that are directly related to oxidant stress and to an unknown number of genes that are not
specifically related to oxidants, but instead control leaf and cell wall thickness, stomatal
conductance, and the internal architecture of the air spaces. It is rarely the case that single genes are
responsible for O3 tolerance. Studies using molecular biological tools and transgenic plants have
positively verified the role of various genes and gene products in O3 tolerance and are continuing to
increase the understanding of O3 toxicity and differences in O3 sensitivity. See Section 9.4.3.2 of this
document for a discussion of recent studies related to gene expression changes in response to O3.
March 2011
9-62
DRAFT - DO NOT CITE OR QUOTE
-------
9.5.4.2. Environmental Biological Factors
1 As stated in the 2006 O3 AQCD, the biological factors within the plant's environment that may
2 directly or indirectly influence its response to O3 in a positive or negative manner encompass insects
3 and other animal pests, diseases, weeds, and other competing plant species. Ozone may influence the
4 severity of a disease or infestation by a pest or weed, either by direct effects on the causal species, or
5 indirectly by affecting the host, or both. In addition, the interaction between O3, a plant, and a pest,
6 pathogen, or weed may influence the response of the target host species to O3 (U.S. EPA, 2006,
7 088089). Several recent studies on the effects of O3 on insects via their interactions with plants are
8 discussed in Section 9.6.6.1. In addition, O3 has also been shown to alter soil fauna communities
9 (Section 9.6.6.2).
10 In contrast to detrimental biological interactions, there are mutually beneficial relationships or
11 symbioses involving higher plants and bacteria or fungi. These include (1) the nitrogen-fixing
12 species Rhizobium and Frankia that nodulate the roots of legumes and alder and (2) the mycorrhizae
13 that infect the roots of many crop and tree species, all of which may be affected by exposure of the
14 host plants to O3. Some discussion of mycorrhizae can be found in Section 9.6.4.
15 In addition to the interactions involving animal pests, O3 also has indirect effects on higher
16 herbivorous animals, e.g., livestock, due to O3-induced changes in feed quality. Recent studies on the
17 effects of O3 on nutritive quality of plants are discussed in Sections 9.5.3 and 9.6.6.3.
18 Intra- and interspecific competition are also important factors in determining vegetation
19 response to O3. Plant competition involves the ability of individual plants to acquire the
20 environmental resources needed for growth and development: light, water, nutrients, and space.
21 Intraspecific competition involves individuals of the same species, typically in monoculture crop
22 situations, while interspecific competition refers to the interference exerted by individuals of
23 different species on each other when they are in a mixed culture. This topic was previously reviewed
24 in AX9.3.3.4 of the 2006 O3 AQCD (U.S. EPA, 2006, 088089). Recent studies on competition and its
25 implications for community composition are discussed in Section 9.6.5.
9.5.4.3. Physical Factors
26 Physical or abiotic factors play a large role in modifying plant response to O3, and have been
27 extensively discussed in previous O3 AQCDs (U.S. EPA, 1996, 080828)01.5. EPA, 2006,
28 This section summarizes those findings as well as recent studies published since the last review.
29 Although some studies have indicated that O3 impact significantly increases with increased
30 ambient temperature (Ball et al, 2000, 026354)(Mills et al, 2000, 030098). other studies have
31 indicated that temperature has little effect (Balls et al., 1996, 026370)(Fredericksen et al., 1996,
32 026653). A recent study by Riikonen et al. (2009, 195664) at the Ruohoniemi open air exposure field
33 in Kuopio, Finland found that the effects of temperature and O3 on total leaf area and photosynthesis
34 of Betulapendula were counteractive. Elevated O3 reduced the saplings' ability to utilize the warmer
35 growth environment by increasing the stomatal limitation for photosynthesis and by reducing the
March 2011 9-63 DRAFT - DO NOT CITE OR QUOTE
-------
1 redox state of ascorbate in the apoplast in the combination treatment as compared to temperature
2 alone (Riikonen et al, 2009, 195664).
3 Temperature affects the rates of all physiological processes based on enzyme catalysis and
4 diffusion; each process and overall growth (the integral of all processes) has a distinct optimal
5 temperature range. It is important to note that a plant's response to changes in temperature will
6 depend on whether it is growing near its optimum temperature for growth or near its maximum
7 temperature (Rowland-Bamford, 2000, 030257). However, temperature is very likely an important
8 variable affecting plant O3 response in the presence of the elevated CO2 levels contributing to global
9 climate change. In contrast, some evidence suggests that O3 exposure sensitizes plants to low
10 temperature stress (Colls and Unsworth, 1992, 026469) and, also, that O3 decreases below-ground
11 carbohydrate reserves, which may lead to responses in perennial species ranging from rapid demise
12 to impaired growth in subsequent seasons (i.e., carry-over effects) (Andersen et al., 1997, 052923).
13 Light, a component of the plant's physical environment, is an essential "resource" of energy
14 content that drives photosynthesis and C assimilation. It has been suggested that increased light
15 intensity may increase the O3 sensitivity of light-tolerant species while decreasing that of shade-
16 tolerant species, but this appears to be an oversimplification with many exceptions. Several studies
17 suggest that the interaction between O3 sensitivity and light environment is complicated by the
18 developmental stage as well as the light environment of individual leaves in the canopy (Chappelka
19 and Samuelson, 1998, 093687_)(Topa et al., 2001, 035398)(Kitao et al., 2009, 191331).
20 Although the relative humidity of the ambient air has generally been found to increase the
21 adverse effects of O3 by increasing stomatal conductance (thereby increasing O3 flux into the leaves),
22 abundant evidence also indicates that the ready availability of soil moisture results in greater O3
23 sensitivity (Mills, 2002, 035322). The partial "protection" against the adverse effects of O3 afforded
24 by drought has been observed in field experiments (Low et al., 2006, 191396) and modeled in
25 computer simulations (Broadmeadow and Jackson, 2000, 021325). Conversely, O3 may enhance the
26 negative effects of O3 on plants (Grulke et al., 2003, 052984)(Tollastrini et al., 2010, 644392). There
27 is also some evidence that O3 can predispose plants to drought stress (Maier-Maercker, 1998,
28 029961). Hence, the nature of the response is largely species-specific and will depend to some extent
29 upon the sequence in which the stressors occur.
9.5.4.4. Interactions with other Pollutants
Ozone-Nitrogen Interactions
30 Elevated O3 exposure and N deposition often co-occur as major pollutant types. However, the
31 interactions of O3 exposure and N deposition on vegetation are complex and less well understood
32 compared to their independent effects. Consistent with the conclusion of the 2006 O3 AQCD
33 (U.S. EPA, 2006, 088089). studies published since the last review indicated that the interactive
34 effects of N and O3 varied among species and ecosystems (Table 9-4). This section will focus on O3
35 and N interactions at the plant scale. Responses at the ecosystem scale are discussed in Section 9.6.
March 2011 9-64 DRAFT - DO NOT CITE OR QUOTE
-------
1 Nitrogen deposition could stimulate relative growth rate (RGR), and lead to increased stomatal
2 conductance. Therefore, plants might become more susceptible to O3 exposure. Alternatively, N
3 deposition may increase the availability of photosynthates for use in detoxification and plants could
4 become more tolerant to O3 (Bassin et al., 2007, 196879). Only a few recent studies have
5 investigated the interactive effects of O3 and N in the U.S. Grulke et al. (2005, 199433) measured
6 stomatal conductance of California black oak (Quercus kelloggii) at a long-term N-enrichment site
7 located in the San Bernardino Mountains, which is accompanied by high O3 exposure (80 ppb,
8 24-h avg. over a six month growing season). The authors found that N amendment led to poor
9 stomatal control in full sun in midsummer of the average precipitation years, but enhanced stomatal
10 control in shade leaves of California black oak. In an OTC study, Handley and Grulke (2008,
11 191485) found that O3 lowered photosynthetic ability and water-use efficiency, and increased leaf
12 chlorosis and necrosis of California black oak. Nitrogen fertilization tended to reduce plant
13 sensitivity to O3 exposure; however, the interaction was not statistically significant.
14 Studies conducted outside the U.S. are also summarized in Table 9-4. Generally, the responses
15 were species specific. The O3-induced reduction in photosynthetic rate and biomass loss were greater
16 in the relatively high N treatment for watermelon (Citrillus lanants) (Calatayud et al., 2006, 191482)
17 and Japanese beech (Fagus crenata) seedlings (Yamaguchi et al., 2007, 191438). However, there
18 was no significant interactive effect of O3 and N on biomass production for Quercus serrata
19 seedlings (Watanabe et al., 2007, 191474). young Norway spruce (Picea abies) trees (Thomas et al.,
20 2005, 075930). and young European beech (Fagus sylvatica) trees (Thomas et al., 2006, 191362).
March 2011 9-65 DRAFT - DO NOT CITE OR QUOTE
-------
Table 94. Response of plants to the interactive effects of elevated ozone exposure and N enrichment
Site
San
Bernardino
Mountains,
U.S.
San
Bernardino
Mountains,
U.S.
Switzerland
Switzerland
Switzerland
Switzerland
Switzerland
Spain
Spain
Japan
Japan
Species
California black
oak (Quercus
kelloggii)
California black
oak (Quercus
kelloggii)
spruce trees
(Pices abies)
beech trees
(Fagus sylvatica)
Alpine pasture
Alpine pasture
Alpine pasture
watermelon
(Citrillus tenants)
Trifolium striatum
Japanese beech
seedlings (Fagus
crenata)
Quercus serrate
seedlings
Ozone exposure
80 ppb
0,75, and 150 ppb
Filtered (19.4-28.1 ppb);
ambient (37.6-47.4 ppb)
Filtered (19.4-28.1 ppb);
ambient (37.6-47.4 ppb)
ambient (AOT40 of 11.1-
12.6ppm-h); 1.2 ambient
(AOT40 of 1 5.2-29.5 ppm-h)
and 1.6 ambient (28.4-64.9
ppm-h)
ambient (AOT40 of 11.1-
12.6 ppm-h); 1.2 ambient
(AOT40 of 15.2-29.5 ppm-h)
and 1.6 ambient (28.4-64.9
ppm-h)
ambient (AOT40 of 11.1-
12. 6 ppm-h); 1.2 ambient
(AOT40 of 15.2-29.5 ppm-h)
and 1.6 ambient (28.4-64.9
ppm-h)
03 free (AOT40 of 0 ppm-h),
ambient (AOT40 of 5. 1-6.3
ppm-h) and elevated 03
(AOT40 of 32.5-35.6 ppm-h)
Filtered (24-havg. of 8-22
ppb); ambient (29-34 ppb),
elevated 03 (35-56 ppb)
Filtered (24-h avg. of 10.3-
13.2 ppb); ambient (42.0-
43.3 ppb), 1.5 ambient (62.6-
63. 9 ppb) and 2.0 ambient
(82.7-84.7 ppb)
Filtered (24-h avg. of 10.3-
13. 2 ppb; ambient (42.0-
43.3 ppb, 1.5 ambient (62.6-
63.9 ppb) and 2.0 ambient
(82.7-84.7 ppb)
N addition
0, and 50 kg N/
ha/yr
0, and 50 kg N/
ha/yr
0,20, 40 and 80
kg N/ ha/yr
0,20, 40 and 80
kg N/ ha/yr
0,5, 10' 25, 50
kg N/ ha/yr
0,5, 10,25,50
kg N/ha/yr
0,5, 10' 25, 50
kg N/ ha/yr
140, 280, and
436 kg N/ ha/yr
10, 30, and 60
kg N/ ha/yr
0, 20 and 50 kg
N/ ha/yr
0, 20 and 50 kg
N/ ha/yr
Responses
N-amended trees had lower late summer
C gain and greater foliar chlorosis in the
drought year, and poor stomatal control
and lower leaf water use efficiency and in
midsummer of the average precipitation
year.
N fertilization tended to reduce plant
sensitivity to 03 exposure; however the
interaction was not statistically significant.
Higher nitrogen levels alleviated the
negative impact of 03 on root starch
concentrations
Nitrogen addition amplified the negative
effects of 03 on leaf area and shoot
elongation.
The positive effects of N addition on
canopy greenness were counteracted by
accelerated leaf senescence in the highest
03 treatment.
Only a small number of species showed
significant 03 and N interactive effects on
leaf chlorophyll concentration, leaf weight
and change in 180, and the patterns were
not consistent.
The positive effects of N addition on
canopy greenness were counteracted by
accelerated leaf senescence in the highest
03 treatment.
High N concentration enhanced the
detrimental effects of 03 on Chi a
fluorescence parameters, lipid
peroxidation, and the total yield.
Ozone reduced total aerial biomass. N
fertilization counterbalanced 03-induced
effects only when plants were exposed to
moderate 03 levels (ambient) but not
under elevated 03 concentrations.
The Oa-induced reduction in net
photosynthesis and whole-plant dry mass
were greater in the relatively high N
treatment than that in the low N treatment.
No significant interactive effects of 03 and
N load on the growth and net
photosynthetic rate were detected.
References
Grulkeetal. (2005,
1994331
Handley andGruIke
(2008, 1914851
Thomas etal. (2005,
0759301
Thomas etal. (2006,
1913621
Bassin etal. (2007,
1915341
Bassin etal. (2009,
1913331
Bassin etal. (2007,
1915341
Calatayud etal. (2006,
1914821
Sanz et al. (2007,
1992451
Yamaguchi etal. (2007,
1914381
Watanabe etal. (2007,
1914741
Ozone-Carbon Dioxide Interactions
1 Several decades of research has shown that exposure to elevated CO2 increases photosynthetic
2 rates (Tissue et al., 1999, 029011)(Bernacchi et al., 2006, ISSOOlKWill and Ceulemans, 1997,
3 679719)(Tissue et al., 1997, 679718)(Bernacchi et al., 2005, 679713). decreases stomatal
4 conductance (Bernacchi et al., 2006, 158001)(Ainsworth and Rogers, 2007, 092940)(Paoletti et al.,
5 2007, 199365)(Medlvn et al., 2001, 679716)(Leakev et al., 2006, 679714) and generally increases
6 the growth of plants(McCarthy et al., 2010, 679715)(Norby et al., 2005, 679717). This is in contrast
7 to the decrease on photosynthesis and growth in many plants that are exposed to elevated O3. The
March 2011
9-66
DRAFT - DO NOT CITE OR QUOTE
-------
1 interactive effects on vegetation have been the subject of research in the past two decades because
2 the implications on productivity and water use of ecosystems. This area of research was covered
3 thoroughly in AX9.3.8.1 of the 2006 O3 AQCD and much of the conclusions made then are still
4 relevant (U.S. EPA, 2006,
5 The bulk of the available evidence shows that, under the various experimental conditions used
6 (which almost exclusively employed abrupt or "step" increases in CO2 concentration, as discussed
7 below), increased CO2 levels (ambient + 200 to 400 ppm) may protect plants from the adverse
8 effects of O3 on growth. This protection may be afforded in part by CO2 acting together with O3 in
9 inducing stomatal closure, thereby reducing O3 uptake, and in part by CO2 reducing the negative
10 effects of O3 on Rubisco and its activity in CO2-fixation. Although both CO2-induced and
11 O3-induced decreases in stomatal conductance have been observed primarily in short-term studies,
12 recent data show a long-term and sustained reduction in stomatal conductance under elevated CO2
13 for a number of species (Ainsworth and Long, 2005, 042647)(Ellsworth et al., 2004,
14 080092)(Gunderson et al., 2002, 080097). Instances of increased stomatal conductance have also
15 been observed in response to O3 exposure, suggesting partial stomatal dysfunction after extended
16 periods of exposure (Maier-Maercker, 1998, 029961)(Grulke et al., 2007, 186963¥Paoletti and
17 Grulke, 2010, 628561).
18 Important caveats must be raised with regard to the findings presented in published research.
19 The first caveat concerns the distinctly different natures of the exposures to O3 and CO2 experienced
20 by plants in the field. Changes in the ambient concentrations of these gases have very different
21 dynamics. In the context of climate change, CO2 levels increase relatively slowly (globally
22 2 ppm/year) and may change little over several seasons of growth. On the other hand, O3 presents a
23 fluctuating stressor with considerable hour-to-hour, day-to-day and regional variability (Polle and
24 Pell, 1999, 093689). Almost all of the evidence presented comes from experimentation involving
25 plants subjected to an abrupt step increase to a higher, steady CO2 concentration. In contrast, the O3
26 exposure concentrations usually varied from day to day. Luo and Reynolds (1999, 035319). Hui et
27 al. (2002, 035288). and Luo (2001, 035318) noted the difficulties in predicting the likely effects of a
28 gradual CO2 increase from experiments involving a step increase or those using a range of CO2
29 concentrations. It is also important to note that the levels of elevated CO2 in many of the studies will
30 not be experienced in the field for 30 or 40 years, but elevated levels of O3 can occur in several areas
31 of the U.S. Therefore, the CO2 * O3 interaction studies may be less policy relevant for current
32 ambient conditions.
33 Another caveat concerns the interactions of O3 and CO2 with other climatic variables, such as
34 temperature and precipitation. In light of the key role played by temperature in regulating
35 physiological processes and modifying plant response to increased CO2 levels (Long, 1991,
36 029710)(Morison and Lawlor, 1999, 094194) and the knowledge that relatively modest increases in
37 temperature may lead to dramatic consequences in terms of plant development (Lawlor, 1998,
38 029015). it is important to consider that studying CO2 and O3 interactions alone may not create a
39 complete understanding of effects on plants under future climate change.
March 2011 9-67 DRAFT - DO NOT CITE OR QUOTE
-------
9.6. Effects of Ozone on Ecosystems and Services
9.6.1. Ecosystem Scale, Function, and Structure
1 Information presented in this section was collected at multiple scales, ranging from the
2 physiology of a given species to population, community, and ecosystem-level investigations. For this
3 assessment, "ecosystem" is defined as a functional entity consisting of interacting groups of living
4 organisms and their abiotic (chemical and physical) environment. Ecosystems cover a hierarchy of
5 spatial scales and can comprise the entire globe, biomes at the continental scale, or small, well-
6 circumscribed systems such as a small pond.
7 Ecosystems have both structure and function. Structure may refer to a variety of measurements
8 including the species richness, abundance, community composition and biodiversity as well as
9 landscape attributes. Competition among and within species and their tolerance to environmental
10 stressors are key elements of survivorship. When environmental conditions are shifted, for example,
11 by the presence of anthropogenic air pollution, these competitive relationships may change and
12 tolerance to stress may be exceeded. "Function" refers to the suite of processes and interactions
13 among the ecosystem components and their environment that involve nutrient and energy flow as
14 well as other attributes including water dynamics and the flux of trace gases. Plant processes
15 including photosynthesis, respiration, C allocation, nutrient uptake and evaporation, are directly
16 related to functions of energy flow and C, nutrient and water cycling. The energy accumulated and
17 stored by vegetation (via photo synthetic C capture) is available to other organisms. Energy moves
18 from one organism to another through food webs, until it is ultimately released as heat. Nutrients and
19 water can be recycled. Air pollution alters the function of ecosystems when elemental cycles or the
20 energy flow are altered. This alteration can also be manifested in changes in the biotic composition
21 of ecosystems.
22 There are at least three levels of ecosystem response to pollutants: (1) the individual organism
23 and its environment; (2) the population and its environment; and (3) the biological community
24 composed of many species and their environment (Billings, 1978, 034165). Individual organisms
25 within a population vary in their ability to withstand the stress of environmental change. The
26 response of individual organisms within a population is based on their genetic constitution, stage of
27 growth at time of exposure to stress, and the microhabitat in which they are growing (Levine and
28 Pinto, 1998, 029599). The stress range within which organisms can exist and function determines the
29 ability of the population to survive. Those best able to cope with environmental stressors survive and
30 reproduce. Competition among different species results in succession (community change over time)
31 and, ultimately, sensitive species may be progressively replaced and communities shift to favor those
32 species that may have the capability to tolerate stressors such as O3 (Guderian, 1985,
33 019325)(Rapport and Whitford, 1999, 004595). In the sections that follow, available information on
34 individual, population and community response to O3 will be discussed. Effects of O3 on productivity
35 and C sequestration, water cycling, below-ground processes, competition and biodiversity, and
March 2011 9-68 DRAFT - DO NOT CITE OR QUOTE
-------
1 insects and wildlife are considered below and in the context of ecosystem services where
2 appropriate.
9.6.1.1. Ecosystem Services
3 Ecosystem structure and function may be translated into ecosystem services. Ecosystem
4 services identify the varied and numerous ways that ecosystems are important to human welfare.
5 Ecosystems provide many goods and services that are of vital importance for the functioning of the
6 biosphere and provide the basis for the delivery of tangible benefits to human society. Hassan et al.
7 (2005, 092759) define these benefits to include supporting, provisioning, regulating, and cultural
8 services:
9 • Supporting services are necessary for the production of all other ecosystem services.
10 Some examples include biomass production, production of atmospheric O2, soil
11 formation and retention, nutrient cycling, water cycling, and provisioning of habitat.
12 Biodiversity is a supporting service that is increasingly recognized to sustain many of the
13 goods and services that humans enjoy from ecosystems. These provide a basis for three
14 higher-level categories of services.
15 • Provisioning services, such as products (Gitay et al., 2001, 092761). i.e., food (including
16 game, roots, seeds, nuts and other fruit, spices, fodder), fiber (including wood, textiles),
17 and medicinal and cosmetic products (such as aromatic plants, pigments).
18 • Regulating services that are of paramount importance for human society such as
19 (1) C sequestration, (2) climate and water regulation, (3) protection from natural hazards
20 such as floods, avalanches, or rock-fall, (4) water and air purification, and (5) disease and
21 pest regulation.
22 • Cultural services that satisfy human spiritual and aesthetic appreciation of ecosystems
23 and their components.
9.6.1.2. Assessing Ozone Effects at Larger Spatial Scales
24 Ozone effects at large spatial scales start as effects on plants at smaller spatial scales. Ozone enters
25 leaves through stomata, and has been shown to alter stomatal conductance and reduce the activity
26 and concentration of Rubisco (Section 9.4.6). Those changes alter the rates of CO2 uptake and water
27 loss from leaves, and therefore the rates of photosynthesis and transpiration at the plant level
28 (Section 9.5.2). Those O3-induced effects could translate from the plant level to the ecosystem level,
29 and cause changes in ecosystem services, such as C storage, water production, nutrient cycling, and
March 2011 9-69 DRAFT - DO NOT CITE OR QUOTE
-------
1 community composition (Figure 9-1). Changes at the ecosystem level are difficult to evaluate
2 directly due to the complexity and the large spatial and temporal scale of ecosystems. These
3 assessments generally involve the extrapolation of laboratory or field results by ecological models or
4 long-term field experiments as discussed below.
9.6.2. Productivity and Carbon Sequestration
5 During the previous NAAQS review, there were very few studies that investigated the effect of
6 O3 exposure on ecosystem productivity and C sequestration. Recent studies from long-term FACE
7 experiments provide new evidence of the association of O3 exposure and changes in productivity at
8 the ecosystem level. In addition to experimental studies, several model studies also assessed the
9 impact of O3 exposure on productivity and C sequestration from stand to global scales. Three types
10 of models are most often used to study the ecological consequences of O3 exposure: (1) regression
11 models such as dose-response function derived from the Nation Crop Loss Assessment Network
12 (NCLAN) (Wang and Mauzerall, 2004, 179978)(Tong and Mauzerall, 2008, 621169): (2) tree growth
13 models such as TREGRO and ECOPHYS (Hogsett et al, 2008, 19L229) (Martin et al, 2001,
14 043678); and (3) process-based ecosystem models such as PnET, Dynamic Land Ecosystem Model
15 (DLEM) and Terrestrial Ecosystem Model (TEM) (Ollinger et al., 2002, 180189)(Ren et al., 2007,
16 191366)(Felzer et al., 2009, 191460). The experimental and model studies on ecosystem productivity
17 and C sequestration, at the stand scale as well as regionally and globally, are reviewed in the
18 following section.
9.6.2.1. Stand Scale
19 The above- and below-ground biomass and net primary production (NPP) were measured at
20 the Aspen FACE site after 7-year O3 exposure. Elevated O3 caused 23, 13 and 14% reductions in
21 total biomass relative to the control in the aspen, aspen-birch and aspen-maple communities,
22 respectively (King et al., 2005, 191701). At the Kranzberg Forest FACE experiment, O3 reduced
23 annual volume growth by 9.5 m3/ha in a mixed mature stand of Norway spruce and European beech
24 (Pretzsch et al., 2010, 580435). Ozone also altered C accumulation and turnover in soil, and the
25 details of these studies are discussed in Section 9.6.4.
26 Changes in stand productivity under elevated O3 were assessed by several model studies.
27 TREGRO is a process-based, single tree growth model and has been widely used to simulate the
28 effects of O3 on the growth of several species in different regions in the U.S. Ozone acts within the
29 model by reducing the maximum potential photo synthetic rate as a function of the cumulative uptake
30 of O3 (Weinstein et al., 1991, 043993)(Tingev et al., 2004, 042385). Hogsett et al. (2008, 191229)
31 used TREGRO to evaluate the effectiveness of various forms and levels of air quality standards for
32 protecting tree growth in the San Bernardino Mountains of California. They found that O3 exposures
33 at the Crestline site resulted in a mean 20.9% biomass reduction from 1980 to 1985 and 10.3%
34 biomass reduction from 1995 to 2000, compared to the "background" O3 concentrations (O3
March 2011 9-70 DRAFT - DO NOT CITE OR QUOTE
-------
1 concentration in Crook County, Oregon). The level of vegetation protection projected was different
2 depending on the air quality scenarios under consideration. Specifically, when air quality was
3 simulated to just meet the California 8 h average maximum of 70 ppb and the maximum 3 months
4 12-h SUM06 of 25 ppm-h, annual growth reductions were limited to 1% or less, while air quality
5 that just met a previous NAAQS (the second highest 1-h max [125 ppb]) resulted in 6-7% annual
6 reduction in growth, resulting in the least protection relative to background O3 (Hogsett et al, 2008,
7 191229V
8 Combining TREGRO with ZELIG, Weinstein et al. (2005, 179965) simulated the effects of
9 different O3 levels ( 0.5, 1.5, 1.75, and 2 times ambient) on the growth and competitive interactions
10 of white fir and ponderosa pine at three sites in California: Lassen National Park, Yosemite National
11 Park, and Crestline. Their results suggested that O3 had little impact on white fir, but greatly reduced
12 the growth of ponderosa pine. If current O3 concentrations continue over the next century, ambient
13 O3 exposure (SUM06 of 110 ppm-h) at Crestline was predicted to decrease individual tree C budget
14 by 10% and decrease ponderosa pine abundance by 16%. Effects at Lassen National Park and
15 Yosemite National Park sites were found to be smaller because of lower O3 exposure levels
16 (Weinstein et al., 2005, 179965V
17 The effects of O3 on stand productivity and dynamics were also studied by other tree growth or
18 stand models, such as ECOPHYS, INTRAST and LINKAGES. ECOPHYS is a functional-structural
19 tree growth model. The model used the linear relationship between the maximum capacity of
20 carboxylation and O3 dose to predict the relative effect of O3 on leaf photosynthesis (Martin et al.,
21 2001, 043678). Simulations with ECOPHYS found that O3 decreased stem dry matter production,
22 stem diameter and leaf dry matter production, induced earlier leaf abscission, and inhibited root
23 growth (Martin et al., 2001, 043678V Simulation with INTRAST and LINKAGES showed similar
24 adverse effects on stand growth. Linking INTRAST with LINKAGES, Hanson et al. (2005, 191461)
25 found that a simulated increase O3 concentration in 2100 (a mean 20-ppb increase over the current
26 O3 concentration) yields a 35% loss of net ecosystem C exchange (NEE) with respect to the current
27 conditions (174 g C/m2/year).
9.6.2.2. Regional and Global Scales
28 Since the publication of the 2006 O3 AQCD, there is additional evidence suggesting that O3
29 exposure alters ecosystem productivity and biogeochemical cycling at the regional and continental-
30 scale. Most of those studies were conducted by using process-based ecosystem models (Table 9-5)
31 and are briefly reviewed in the following sections.
Carbon Dynamics in Natural Ecosystems
32 Results of Reich (1987, 019314) and Tjoelker et al. (1995, 035394) indicated that O3 effects
33 on photosynthesis can be determined as a function of O3 uptake to internal leaf surface. Pooling data
34 from Reich (1987, 019314) and Tjoelker et al. (1995, 035394). Ollinger et al. (1997, 040707)
35 derived an equation to simulate the O3 effects on photosynthesis by cumulative O3 dose (AOT40)
March 2011 9-71 DRAFT - DO NOT CITE OR QUOTE
-------
1 and stomatal conductance. They incorporated this equation into the PnET-II model and studied the
2 effect of O3 on hardwood forest productivity of 64 hardwood sites in northeastern U.S. Their model
3 indicated that O3 caused a 3-16% reduction in NPP from 1987 to 1992 (Table 9-5). Ollinger et al.
4 (2002, 180189) assessed the interactive effects of O3, N deposition, elevated CO2 and land use
5 history on C dynamics by PnET-CN. Their results indicated that O3 offset the increase in net
6 C exchange caused by elevated CO2 and N deposition by 13% (25.0 g C/m2/year) under agriculture
7 site history, and 23% (33.6 g C/m2/year) under timber harvest site history. PnET-CN was also used to
8 assess changes in C sequestration of U.S. Mid-Atlantic temperate forest. Pan et al. (2009, 596032)
9 designed a factorial modeling experiment to separate the effects of changes in atmospheric
10 composition, historical climatic variability and land-disturbances on the C cycle. They also found O3
11 acted as a negative factor, partially offsetting the growth stimulation caused by elevated CO2 and N
12 deposition of U.S. Mid-Atlantic temperate forest. Ozone decreased NPP of most forest types by 7-
13 8%. Among all the forest types, spruce-fir forest was most resistant to O3 damage, and NPP
14 decreased by only 1% (Pan et al., 2009, 596032V
15 Felzer et al. (2004, 186927) developed TEM 4.3 to simulate the effects of O3 on plant growth.
16 The effects of O3 on NPP and C sequestration of deciduous trees, conifers and crops in the
17 conterminous U.S. were estimated by TEM. The results indicated that O3 reduced NPP and
18 C sequestration in the U.S. (Table 9-5) and the largest decreases (over 13% in some locations) in
19 NPP occurred in the Midwest agricultural lands during the mid-summer. TEM was also used to
20 evaluate the magnitude of O3 damage at a global scale (Table 9-5) (Felzer et al., 2005, 186928).
21 Simulations for the historical period (1860-1995) show that the largest reductions in NPP and net
22 C exchange occurred in the mid western U.S., eastern Europe, and eastern China (Felzer et al., 2005,
23 186928). DLEM was developed to simulate the detrimental effect of O3 on ecosystems, and has been
24 used to examine the O3 damage on NPP and C sequestration in Great Smoky Mountains National
25 Park (Zhang et al., 2007, 196983). grassland ecosystems and terrestrial ecosystems in China (Ren et
26 al., 2007, 581541¥Ren et al., 2007, 191366). Results of those simulations are listed in Table 9-5.
27 Instead of using AOT40 as their O3 exposure metric as PnET, TEM and DLEM did, Sitch et al.
28 (2007, 093294) incorporated a different O3 metric named CUOt (cumulative stomatal uptake of O3),
29 derived from Pleijel et al. (2004, 056608). into the MOSES-TRIFFID coupled model. In the CUOt
30 metric, the fractional reduction of plant production is dependent on O3 uptake by stomata over a
31 critical threshold for damage and this threshold level is different for different plant functional types.
32 Consistent with previous studies, their model simulation indicated that O3 reduced global gross
33 primary production (GPP), C exchange rate and C sequestration (Table 9-5). The largest reductions
34 in GPP and land-C storage were projected over North America, Europe, China and India. In the
35 model, reduced ecosystem C uptake due to O3 damage, results in additional CO2 accumulation in the
36 atmosphere and an indirect radiative forcing of climate change. Their simulations indicated that the
37 indirect radiative forcing caused by O3 (0.62-1.09 W/m2) could have even greater impact on global
38 warming that the direct radiative forcing of O3 (0.89 W/m2) (Sitch et al., 2007, 093294).
March 2011 9-72 DRAFT - DO NOT CITE OR QUOTE
-------
Crop Yield Loss
1 Two large scale field studies were conducted in the U.S. (NCLAN) and in Europe (European
2 Open Top Chamber Programme, EOTCP) to assess the impact of O3 on crop production. Ozone
3 exposure-response regression models derived from the two programs have been widely used to
4 estimate crop yield loss (Wang and Mauzerall, 2004, 179978)(Tong and Mauzerall, 2008,
5 621169)(Van Dingenen et al., 2009, 199765). Those studies found that O3 generally reduced crop
6 yield and different crops showed different sensitivity to O3 pollution (Table 9-5). Ozone was
7 calculated to induce a possible 45-82 million metric tons loss for wheat globally. Production losses
8 for rice, maize and soybean were on the order of 17-23 million metric tons globally (Van Dingenen
9 et al., 2009, 199765). The largest yield losses occur in high-production areas exposed to high O3
10 concentrations, such the Midwest and the Mississippi Valley regions in the U.S., Europe, China and
11 India (Van Dingenen et al., 2009, 199765)Obng et al., 2007, 107431).
March 2011 9-73 DRAFT - DO NOT CITE OR QUOTE
-------
Table 9-5. The effects of ozone on
Scale
GPP Global
NPP Global
U.S.
U.S.
northeastern
U.S.
U.S. Mid-
Atlantic
China
C exchange Global
Global
C sequestration Global
U.S.
GSM National
Park
China
Crop yield loss Global
U.S.
U.S.
East Asia
Model
MOSES-
TRIFFID
TEM
TEM
TEM
PnET
PnET
DLEM
TEM
MOSES-
TRIFFID
MOSES-
TRIFFID
TEM
DLEM
DLEM
Dose-
response
function
Dose-
response
function
Dose-
response
function
Dose-
response
function
primary production, C exchange, C sequestration and
Index
CUOta
AOT40
AOT40
AOT40
AOT40
AOT40
AOT40
AOT40
cuot
cuot
AOT40
AOT40
AOT40
M7b;
M12c;
AOT40
M7; M12;
AOT40
SUM06
M7; M12
Ozone Impacts
Decreased by 14-23% over the period 1901-2100
Decreased by 0.8% without agricultural management and a
decrease of 2.9% with optimal agricultural management
Reduced by 2.3% without optimal N fertilization and 7.2%
with optimal N fertilization from 1983-1993
Reduced by 2.6-6.8% during the late 1980s-early 1990s.
A reduction of 3-1 6% from 1 987-1 992
Decreased NPP of most forest types by 7-8%
Reduced NPP of grassland in China by 8.5 Tg C from
1960s to 1990s
Reduced netC exchange (1950-1995) by 0.1 Pg C/yr
without agricultural management and 0.3 Pg C/yr with
optimal agricultural management
Decreased global mean land-atmosphere C fluxes by 1.3
Pg C/yr and 1.7 Pg C/yr for the 'high' and 'low' plant 03
sensitivity models, respectively
Reduced land-C storage accumulation by between 143 Pg C/yr
and 263 Pg C/yr from 1900-2100
Reduced C sequestration by 1 8-38 Tg C/yr from 1 950 to
1995
Decreased the ecosystem C storage of deciduous forests
by 2.5% and pine forest by 1.4% from 1971 to 2001
Reduced total C storage by 0.06% in 1960s and 1.6% in
1 990s in China's terrestrial ecosystems
Reduced by 7.3% to 12.3% for wheat, 5.4% to 15.6% for
soybean, 2.8% to 3.7% for rice, and 2.4% to 4.1% for maize
in year 2000.
Reduced by 4.1% to 4.4% for wheat, 7.1% to 17.7% for
soybean, 2.6% to 3.2% for rice, and 2.2% to 3.6% for maize
in year 2000.
Caused a loss of 53.8 million to 438 million bushels in
soybean production, which account for 1.7-14.2% of total
U.S. soybean production in 2005
Reduced the yield of wheat, rice and corn by 1-9% and
soybean by 23-27% in China, Japan and South Korea in
1990
yield loss
Reference
Sitchetal. (2007,
0932941
Felzeretal. (2005,
186928)
Felzeretal. (2005,
1869281
Felzeretal. (2004,
186927)
Ollingeretal.
(1997.0407071
Panetal. (2009,
5960321
Renetal. (2007,
5815411
Felzeretal. (2005,
1869281
Sitchetal. (2007,
0932941
Sitchetal. (2007,
0932941
Felzeretal. (2004,
1869271
Zhang etal. (2007,
1969831
Renetal. (2007,
1913661
Van Dingenen etal.
(2009, 1997651
Van Dingenen et al.
(2009, 1997651
Tong et al. (2007,
1074311
Wang and
Mauzerall (2004,
1799781
aCUOt is defined as the cumulative stomatal uptake of 03, using a constant 03-uptake rate threshold oft nmol/m2/s.
bM7 is defined as 7-h mean 03 concentration (ppb).
CM12 is defined as 12-h mean 03 concentration (ppb).
dPg equals 1 * 1015 grams.
9.6.3. Water Cycling
1 Ozone has been shown to alter stomatal performance, which affects plant and stand
2 transpiration and therefore hydrological cycling. However, there is not a clear consensus on stomatal
3 response to O3 exposure. A meta-analysis found that O3 reduced stomatal conductance by 11%
4 (Wittig et al., 2007, 191695). Stomatal closure could help protect the plant from water loss. On the
5 other hand, a number of studies suggested that O3 exposure could impair stomatal function, which
6 leads to greater stomatal apertures, delays stomatal closure at night and results in higher transpiration
7 (Grulke et al., 2004, 042646¥McLaughlin et al., 2007, Q9034_8)(McLaughlin et al., 2007,
March 2011
9-74
DRAFT - DO NOT CITE OR QUOTE
-------
1 Q90347)(Mills et al, 2009, 191272)(Wilkinson and Davies, 2009, 199758)(Wilkinson and Davies,
2 2010, 598245). This O3-induced impairment of stomatal control may be more pronounced for plants
3 growing under drought stress (McLaughlin et al., 2007, 090348)(McLaughlin et al., 2007,
4 09034_Z)(Wilkinson and Davies, 2010, 598245). The discrepancy regarding stomatal function has
5 lead to further debate on the effects of O3 on ecosystem water production.
6 Felzer et al. (2009, 191460) used TEM-Hydro to assess the interactions of O3, climate,
7 elevated CO2 and N limitation on the hydrological cycle in the eastern U.S. They found that elevated
8 CO2 decreased evapotranspiration by 2-4% and increased runoff by 3-7%, as compared to the effects
9 of climate alone. When O3 damage and N limitation were included, evapotranspiration was reduced
10 by an additional 4-7% and runoff was increased by an additional 6-11% (Felzer et al., 2009,
11 191460). Based upon simulation with INTRAST and LINKAGES, Hanson et al. (2005, 191461)
12 found that increasing O3 concentration by 20 ppb above the current ambient level yields a modest
13 3% reduction in water use. Those ecological models were generally built on the assumption that O3
14 induces stomatal closure. Therefore, results of those models normally found that O3 reduced water
15 use.
16 In contrast to the model simulations (Hanson et al., 2005, 191461)(Felzer et al., 2009,
17 191460). field studies conducted by McLaughlin et al. (2007, 090348)(2007. 090347) indicated that
18 O3 increases water use in a mixed deciduous forest in eastern Tennessee. McLaughlin et al. (2007,
19 090348)(2007. 090347) found that O3, with daily maximum levels ranging from 69.2 to 82.9 ppb,
20 reduced stem growth by 30-50% in the high-O3 year 2002. The decrease in growth rate was caused
21 in part by amplification of diurnal cycles of water loss and recovery. Peak hourly O3 exposure
22 increased the rate of water loss through transpiration as indicated by the increased stem sap flow.
23 The increased canopy conductance resulted in higher water uptake as reflected in the reduced soil
24 moisture in the rooting zone. The change in tree water use led to further impacts on the hydrological
25 cycle at the landscape level. Increased water use under high O3 exposure was reported to reduce late-
26 season modeled streamflow in three forested watersheds in eastern Tennessee (McLaughlin et al.,
27 2007, 090347).
28 In addition to the impacts on stomatal performance, O3-induced physiological changes, such as
29 reduced leaf area index and accelerated leaf senescence, could alter water use efficiency. At the
30 Aspen FACE experiment, stand-level water use, as indicated by sap flux per unit ground area, was
31 not significantly affected by elevated O3 despite a 22% decrease in leaf area index and 20% decrease
32 in basal area (Uddling et al., 2008, 191655). Several factors could contribute to the lack of negative
33 effect of elevated O3 on stand water use. The maximum sap flux per unit total leaf area was
34 substantially increased by elevated O3, suggesting that whole-plant hydraulic conductance per unit
35 leaf area was increased (Uddling et al., 2009, 596219). Other potential contributing factors included
36 the higher proportion of sun leaves, and similar or even increased fine root biomass under elevated
37 O3 (Uddling et al., 2008, 191655). Elevated O3 could also affect evapotranspiration by altering tree
38 crown interception of precipitation. Ozone has been shown to change branch architectural
39 parameters, and the effects were species specific at the Aspen FACE experiment (Rhea et al., 2010,
March 2011
9-75
DRAFT - DO NOT CITE OR QUOTE
-------
1 647084). The authors found that there was a significant correlation between canopy architecture
2 parameters and stem flow for birch but not aspen.
9.6.4. Below-Ground Processes
3 Above-ground and below-ground processes are tightly interconnected. Because roots and soil
4 organisms are not exposed directly to O3, below-ground processes are affected by O3 more through
5 altering the quality and quantity of C supply from photosynthates and litterfall (Andersen, 2003,
6 041673). Ozone can decrease leaf C uptake by altering stomatal function, reducing the activity and
7 concentration of Rubisco, and accelerating leaf senescence (Section 9.4). Ozone can also increase
8 the metabolic costs by stimulating the production of chemical compounds for defense and repair
9 processes, and increasing the synthesis of antioxidants to neutralize free radicals (see Section 9.4),
10 which increase the consumption of carbon for above-ground processes. Therefore, O3 could
11 significantly reduced the amount of C available for allocation to below-ground by decreasing
12 C uptake but increasing C consumption of above-ground processes (Andersen, 2003, 041673).
13 Since the 2006 O3 AQCD, there is additional evidence for O3 effects on below-ground
14 processes. Ozone has been found to alter root growth, soil food web structure, decomposer activities,
15 C turnover and nutrient flow (Figure 9-7). Ozone effects on root development and root biomass
16 production (Section 9.5.2.1) and soil food web structure (Section 9.6.5.3) are reviewed in other
17 sections. The focus in this section is on the response of litter input, decomposer activities, soil
18 respiration, soil C formation and nutrient cycling.
March 2011 9-76 DRAFT - DO NOT CITE OR QUOTE
-------
Carbon movement in plant and soil
co
Al location otC
co,
Litter inputs
CO. release
Soil physical *
Ctwmieil properttos
• Fungi
-Micro &
macro Inverteorales
Source: Used with permission from Andersen (2003, 0416731
Figure 9-7. Conceptual diagram showing where ozone disrupts C flow in a tree-soil system,
including transfer between biotic and abiotic components below ground that
influence soil physical and chemical properties. Arrows denote C flux pathways that
are affected by ozone. Dashed lines indicate where the impact of ozone is suspected
but unknown.
9.6.4.1. Litter Carbon Chemistry, Litter Nutrient and Their Ecosystem Budgets
Consistent with previous findings, recent studies show that, although the responses are often
species specific, O3 tends to alter litter chemistry (U.S. EPA, 2006, 08 8089). Alterations in chemical
2
3 parameters, such as changes in C chemistry and nutrient concentrations, were observed in both leaf
4 and root litter (Table 9-6).
5 At the Aspen FACE site, several studies investigated litter chemistry changes (Chapman et al.,
6 2005, 191345)(Liu et al., 2005, 187005)(Johnson and Pregitzer, 2007, 191287)(Parsons et al., 2008,
7 191853). In both aspen and birch leaf litter, elevated O3 increased the concentrations of soluble
8 sugars, soluble phenolics and condensed tannins (Liu et al., 2005, 187005)(Parsons et al., 2008,
9 191853). Compared to other treatments, aspen litter under elevated O3 had the highest fiber
10 concentration, with the lowest concentration associated with the birch litter under the same
1 1 conditions (Parsons et al., 2008, 191853). Chapman et al. (2005, 191345) measured chemical
12 changes in fine root litter and found that elevated O3 decreased lignin concentration. The O3-induced
March 201 1
9-77
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
chemistry changes were also reported from other experimental sites. Results from an OTC study in
Finland suggested that elevated O3 increased the concentration of acid-soluble lignin, but had no
significant impact on other chemicals such as total sugars, hemicelluloses, cellulose or total lignin in
the litter of silver birch (Kasurinen et al., 2006, 191269). Results from the free air canopy O3
exposure experiment at Kranzberg Forest, showed that O3 increased starch concentrations but had no
impact on cellulose and lignin in beech and spruce leaf litter (Aneja et al., 2007, 191472). The effect
of O3 on three antioxidants (ascorbate, glutathione and oc-tocopherol) in fine roots of beech was also
assessed at Kranzberg Forest. The results indicated that O3 had no significant effect on oc-tocopherol
and ascorbate concentrations, but decreased glutathione concentrations in fine roots (Haberer et al.,
2008, 191334). In addition to changing C chemistry, O3 also altered nutrient concentrations in green
leaves and litter (Table 9-6).
The combined effects of O3 on biomass productivity and chemistry changes could alter
C chemicals and nutrient contents at the canopy or ecosystem level. For example, although O3 had
different impacts on their concentrations, annual fluxes of C chemicals (soluble sugar, soluble
phenolics, condensed tannins, lipid and hemicelluloses), macro nutrients (N, P, K and S) and micro
nutrients (Mg, B, Cu and Zn) to soil were all reduced due to lower litter biomass productivity at
Aspen FACE (Liu et al., 2005, 187005)(Liu et al., 2007, 093286). At the Kranzberg Forest, N
content of spruce canopy in a mixed culture and Ca2+ content of beech canopy in a monoculture
increased due to elevated O3 increased leaf concentrations of those nutrients although leaf production
was not significantly altered by O3 (Rodenkirchen et al., 2009, 191540).
Table 9-6. The effect of elevated ozone on leaf/litter nutrient concentrations
Study Site
Suonenjoki Research
Station, Finland
Aspen FACE
Aspen FACE
Kranzberg Forest, Germany
Kranzberg Forest, Germany
Species
Silver birch
Aspen and birch
Birch
Beech and spruce
Beech and spruce
Ozone Concentration
Ambient: 10-60 ppb
Elevated: 2*ambient
Ambient: 50-60 ppb
Elevated: 1.5*ambient
Ambient: 50-60 ppb
Elevated: 1.5*ambient
Ambient: 9-41 ppb
Elevated: 2*ambient
Ambient: 9-41 ppb
Elevated: 2*ambient
Response
Decreased the concentration of P, Mn, Zn
and B in leaf litter
Decreased the concentrations of P, S, Ca
and Zn, but had no impact on the
concentrations of N, K, Mg, Mn, B and Cu in
leaf litter.
Increase N concentration in birch litter
Increased N concentration in beach leaf, but
not in spruce needle
1) Had no significant effects on spruce
needle chemistry; 2) increased Ca
concentration in beech leaves in
monoculture, but had no impacts on other
nutrients
Reference
Kasurinen et al.
(2006, 1912691
Liuetal. (2007,
0932861
Parsons etal. (2008,
1918531
Kozovits et al.
(2005, 191282)
Rodenkirchen etal.
(2009, 1915401
9.6.4.2. Decomposer Metabolism and Litter Decomposition
The above- and below-ground physiological changes caused by O3 exposure cascade through
the ecosystem and affect soil food webs. In the 2006 O3 AQCD, there were very few studies on the
effect of O3 on the structure and function of soil food webs, except two studies conducted by Larson
et al. (2002, 053015) and Phillips et al. (2002, 041768). Since the last O3 AQCD (U.S. EPA, 2006,
March 2011
9-78
DRAFT - DO NOT CITE OR QUOTE
-------
1 088089). new studies have provided more information on how O3 affects the metabolism of soil
2 microbes and soil fauna.
3 Chung et al.(2006, 191729) found that the activity of the cellulose-degrading enzyme 1,4-p-
4 glucosidase was reduced by 25% under elevated O3 at AspenFACE. The decrease in cellulose-
5 degrading enzymatic activity was associated with the lower cellulose availability under elevated O3
6 (Chung et al., 2006, 191729). The suppression of soil enzyme activities was also found in the
7 rhizosphere of beech trees (Fagus sylvatica) in a lysimeter study in Germany (Esperschutz et al.,
8 2009, 595669)(Pritsch et al., 2009, 626808). Except for xylosidase, enzyme activities involved in
9 plant cell wall degradation (cellobiohydrolase, beta-glucosidase and glucuronidase) were decreased
10 in rhizosphere soil samples under elevated O3 (2 x ambient level) (Pritsch et al., 2009, 626808).
11 Similarly, Chen et al. (2009, 191452) found O3 exposure, with a 3-month AOT40 of
12 21.4-44.1 ppm-h, decreased the microbial metabolic capability in the rhizosphere and bulk soil of
13 wheat, although the observed reduction in bulk soil was not significant.
14 Ozone-induced change in soil organisms' activities could affect litter decomposition rates.
15 However, no general decomposition pattern has been identified. The responses varied among
16 species, sites and exposure length. Parsons et al. (2008, 191853) collected litter from aspen and birch
17 seedlings at AspenFACE site, and conducted a 23-month field litter incubation starting in 1999. They
18 found that elevated O3 had different impacts on the decomposition of aspen and birch litter. Elevated
19 O3 was found to reduce aspen litter decomposition. However, O3 accelerated birch litter
20 decomposition under ambient CO2, but reduced it under elevated CO2 (Parsons et al., 2008, 191853).
21 Liu et al. (2009, 191470) conducted another litter decomposition study at Aspen FACE from 2003 to
22 2006, when stand leaf area index (LAI) reached its maximum. During the 935-day field incubation,
23 elevated O3 was shown to reduce litter mass loss in the first year, but not in the second year. They
24 suggested that higher initial tannin and phenolic concentrations under elevated O3 reduced microbial
25 activity in the first year (Liu et al., 2009, 191470). In an OTC experiment, Kasurinen et al. (2006,
26 191269) collected silver birch leaf litter from three consecutive growing seasons and conducted three
27 separate litter-bag incubation experiments. Litter decomposition was not affected by O3 exposure in
28 the first two incubations, but a slower decomposition rate was found in the third incubation. Their
29 principle component analysis indicated that the litter chemistry changes caused by O3 (decreased
30 Mn, P, B and increased C:N) might be partially responsible for the decreased mass loss of their third
31 incubation.
9.6.4.3. Soil respiration and carbon formation
32 Ozone could reduce the availability of photosynthates for export to roots, and increase root
33 mortality and turnover rates. Ozone has also been shown to reduce above-ground litter productivity
34 and alter litter chemistry, which would affect the quality and quantity of the C supply to soil
35 organisms (Section 9.6.6.1). The complex interactions among those changes make it difficult to
36 predict the response of soil C cycling under elevated O3. The 2006 O3 AQCD concluded that O3 had
37 no consistent impact on soil respiration (U.S. EPA, 2006, 088089). Ozone could increase or decrease
March 2011 9-79 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
soil respiration, depending on the approach and timing of the measurements. Ozone may also alter
soil C formation. However, very few experiments directly measured changes in soil organic matter
content under O3 fumigation (U.S. EPA, 2006, 088089). Recent studies on soil respiration and soil
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
C content also found mixed responses. Most importantly, new publications derived from long-term
fumigation experiments, such as the Aspen FACE experiment, suggest that ecosystem response to O3
exposure can change over time. Observations made during the late exposure years can be
inconsistent with those during the early years, highlighting the need for caution to assess O3 effects
based on short-term studies (Table 9-7).
Table 9-7. The temporal variation of ecosystem responses to ozone exposure at AspenFACE site
Endpoint
Litter decomposition
Fine root production
Soil respiration
Soil C formation
The Time of the
Measurement
1999-2001
2003-2006
1999
2002, 2005
1998-1999
2003-2007
1998-2001
2004-2008
Response
03 reduced aspen litter decomposition. However, 03 accelerated birch
litter decomposition under ambient C02, but reduced it under elevated
C02
03 reduced litter mass loss in the first year, but not in the second year.
03 had no significant impact on fine root biomass
03 increased fine root biomass
Soil respiration under +C02+03 treatment was lower than that under
+C02 treatment
Soil respiration under +C02+03 treatment was 5-25% higher than
under elevated C02 treatment.
03 reduced the formation rates of total soil C by 51 % and acid-
insoluble soil C by 48%
No significant effect of 03 on the new C formed under elevated C02
Reference
Parsons et al. (2008, 1918531
Liu et al. (2009, 1914701
Kingetal. (2001, 0417511
Pregitzer et al. (2008, 191677)
Kingetal. (2001, 0417511
Pregitzer et al. (2006,
191676), Pregitzer et al.
(2008, 1916771
Loya et al. (2003, 0743801
Talhelm et al. (2009, 5961891
Soil Respiration
Ozone has shown inconsistent impacts on soil respiration. A sun-lit controlled-environment
chamber study found that O3 had no significant effects on soil respiration, fine root biomass or any
of the soil organisms in a reconstructed ponderosa pine/soil-litter system (Tingey et al., 2006,
191341). In an adult European beech/Norway spruce forest at Kranzberg Forest, the free air O3
fumigation (AOT40 of 10.2-117 ppm-h) increased soil respiration under both beech and spruce
during a humid year (Nikolova et al., 2010, 626810) . The increased soil respiration under beech has
been accompanied by the increase in fine root biomass and ectomycorrhizal fungi diversity and
turnover (Grebenc and Kraigher, 2007, 191265). The stimulating effect on soil respiration
disappeared under spruce in a dry year, which was associated with a decrease in fine root production
in spruce under drought. This finding suggested that drought was a more dominant stress than O3 for
spruce (Nikolova et al., 2010, 626810). Andersen et al. (2010, 628559) labeled the canopies of
European beech and Norway spruce with CO2 depleted in 13C at the same site. They did not observe
any significant changes in soil respiration for either species.
The nearly 10 year long studies at AspenFACE indicated that the response of soil respiration to
O3 interacted with CO2 exposure and varied temporally (Table 9-7) (King et al., 2001,
04_175i)(Pregitzer et al., 2006, 191676)(Pregitzer et al., 2008, 191677). Ozone treatment alone
March 2011
9-80
DRAFT - DO NOT CITE OR QUOTE
-------
1 generally had the lowest mean soil respiration rates, although those differences between control and
2 elevated O3 were usually not significant. However, soil respiration rates were different with O3 alone
3 and when acting in combination with elevated CO2. In the first five years (1998-2002), soil
4 respiration under +CO2+O3 treatment was similar to that under control and lower than that under
5 +CO2 treatment (King et al., 2001, 041751)(Pregitzer et al., 2006, 191676V Since 2003, +CO2+O3
6 treatment started to show the greatest impact on soil respiration. Compared to elevated CO2, soil
7 respiration rate under +CO2+O3 treatment was 15-25% higher from 2003-2004, and 5-10% higher
8 from 2005-2007 (Pregitzer et al., 2006, 191676)(Pregitzer et al., 2008, 191677). Soil respiration was
9 highly correlated with the biomass of roots with diameters of <2 mm and <1 mm, across plant
10 community and atmospheric treatments. The authors suggested that the increase in soil respiration
11 rate may be due to +CO2+O3 increased fine root (<1.0 mm) biomass production (Pregitzer et al.,
12 2008. 191677).
Soil Carbon Formation
13 Ozone-induced reductions in plant growth can result in reduced C input to soil and therefore
14 soil C content (Andersen, 2003, 041673). The simulations of most ecosystem models support this
15 prediction (Felzer et al., 2004, 186927)(Zhang et al., 2007,196983)(Ren et al., 2007, 191366V
16 However, very few studies have directly measured soil C dynamics under elevated O3. After the first
17 four years of fumigation (from 1998 to 2001) at the Aspen FACE site, Loya et al. (2003, 074380)
18 found that forest stands exposed to both elevated O3 and CO2 accumulated 51% less total soil C, and
19 48% less acid-insoluble soil C compared to stands exposed only to elevated CO2. Soil organic carbon
20 (SOC) was continuously monitored at the Aspen FACE site, and the later data showed that the initial
21 reduction in new C formation (soil C derived from plant litter since the start of the experiment) by
22 O3 under elevated CO2 is only a temporary effect (Table 9-7) (Talhelm et al., 2009, 596189). The
23 amount of new soil C in the elevated CO2 and the combined elevated CO2 and O3 treatments has
24 converged since 2002. There was no significant effect of O3 on the new C formed under elevated
25 CO2 over the last four years of the study (2004-2008). Talhelm et al. (2009, 596189) suggested the
26 observed reduction in the early years of the experiment might be driven by a suppression of
27 C allocated to fine root biomass. During the early exposure years, O3 had no significant impact on
28 fine root production (King et al., 2001, 041751). However, the effect of O3 on fine root biomass was
29 observed later in the experiment. Ozone increased fine root production and the highest fine root
30 biomass was observed under the combined elevated CO2 and O3 treatment in the late exposure years
31 (Table 9-7) (Pregitzer et al., 2006, 191676). This increase in fine root production was due to changes
32 in community composition, such as better survival of O3-tolerant aspen genotype, birch and maple,
33 rather than changes in C allocation at the individual tree level (Pregitzer et al., 2008, 191677)(Zak et
34 al., 2007, 191239).
March 2011 9-81 DRAFT - DO NOT CITE OR QUOTE
-------
9.6.4.4. Nutrient cycling
1 Nutrient cycling is important for sustaining ecosystem production. Ozone can affect nutrient
2 cycling by changing nutrient release from litter and uptake by plants. Nitrogen is the limiting
3 nutrient for most ecosystems, and several studies examined N dynamics under elevated O3. Holmes
4 et al. (2006, 191372) found that elevated O3 decreased gross N mineralization at the Aspen FACE
5 site, indicating that O3 may reduce N availability. However, other N cycling processes, such as NH4+
6 immobilization, gross nitrification, microbial biomass N and soil organic N, were not affected by
7 elevated O3 (Holmes et al., 2006, 191372). Similarly, Kanerva et al. (2006, 191747) found total N,
8 NO3-, microbial biomass N, potential nitrification and denitrification in their meadow mesocosms
9 were not affected by elevated O3 (40-50 ppb). Ozone also showed small impact on other micro and
10 macro nutrients. Liu et al. (2007, 093286) assessed N, P, K, S, Ca, Mg, Mn, B, Zn and Cu release
11 dynamics at Aspen FACE, and they found that O3 had no effects on most nutrients, except to
12 decrease N and Ca release from litter.
13 Using the Simple Nitrogen Cycle model (SINIC), Hong et al. (2006, 186989) evaluated the
14 impacts of O3 exposure on soil N dynamics and streamflow nitrate flux. The detrimental effect of O3
15 on plant growth was found to reduce plant uptake of N and therefore increase nitrate leaching. Their
16 model simulation indicated that ambient O3 exposure increased the mean annual stream flow nitrate
17 export by 12% (0.042 g N/m2/year) at the Hubbard Brook Experimental Watershed from 1964-1994
18 (Hong et al., 2006, 186989).
9.6.4.5. Dissolved Organic Carbon and Biogenic Trace Gases Emission
19 The O3-induced changes in plant growth, C and N fluxes to soil and microbial metabolism can
20 alter other biogeochemical cycling processes, such as soil dissolved organic carbon (DOC) turnover
21 and trace gases emission.
22 Jones et al. (2009, 199881) collected fen cores from two peatlands in North Wales, UK and
23 exposed them to one of four levels of O3 (AOT40 of 0, 3.69, 5.87 and 13.80 ppm-h for 41 days).
24 They found the concentration of porewater DOC in fen cores was significantly decreased by
25 increased O3 exposure. A reduction of the low molecular weight fraction of DOC was concurrent
26 with the observed decrease in DOC concentration. Their results suggested that O3 damage to
27 overlying vegetation may decrease utilizable C flux to soil. Microbes, therefore, have to use labile C
28 in the soil to maintain their metabolism, which, the authors hypothesized, leads to a decreased DOC
29 concentration with a shift of the DOC composition to more aromatic, higher molecular weight
30 organic compounds.
31 Several studies since the 2006 O3 AQCD have examined the impacts of O3 on nitrous oxide
32 (N2O) and methane (CH4) emission. Kanerva et al. (2007, 191405) measured the fluxes of N2O and
33 CH4 in meadow mesocosms, which were exposed to elevated CO2 and O3 in OTCs in south-western
34 Finland. They found that the daily N2O fluxes were decreased in the NF+O3 (non-filtered air +
35 elevated O3, 40-50 ppb) after three seasons of exposure. Elevated O3 alone or combined with CO2
March 2011 9-82 DRAFT - DO NOT CITE OR QUOTE
-------
1 did not have any significant effect on the daily fluxes of CH4 (Kanerva et al., 2007, 191405). In
2 another study conducted in central Finland, the 4 year open air O3 fumigation (AOT40 of 20.8-
3 35.5 ppm-h for growing season) also did not affect the rate of potential CH4 production, but it
4 slightly increased potential CH4 oxidation by 15% in the peatland microcosms (Morsky et al., 2008,
5 191507). However, O3 has no overall effects on CH4 emissions, which is the net result of the
6 potential CH4 production and oxidation (Morsky et al., 2008, 191507).
9.6.5. Competition and biodiversity
7 The effects of O3 on species competition (AX9.3.3.4) and community composition (AX9.6.4)
8 were summarized in the 2006 O3 AQCD (U.S. EPA, 2006, 088089). Plant species differ in their
9 sensitivity to O3. Fast growing plants with high stomatal conductance and high specific leaf area
10 (SLA) were more likely to be sensitive to O3 exposure. Further, different genotypes of a given
11 species also vary in their sensitivity. This differential sensitivity could change the competitive
12 interactions that lead to loss in O3 sensitive species or genotypes. A shift in community composition
13 in forest and grassland ecosystems noted in the 2006 O3 AQCD (U.S. EPA, 2006, 088089) has
14 continued to be observed from experimental and gradient studies. Additionally, research since the
15 last review has shown that O3 can alter community composition and diversity of soil microbial
16 communities.
9.6.5.1. Forest
17 In the San Bernardino Mountains in southern California, O3 pollution caused a significant
18 decline in ponderosa pine (Pinus ponderosa ) and Jeffrey pine (Pinus jeffreyi} (U.S. EPA, 2006,
19 088089). Pine trees in the young mature age class group exhibited higher mortality rates compared
20 with mature trees at a site with severe O3 visible foliar injury. The vulnerability of young mature
21 pines was most likely caused by the fact that trees in this age class were emerging into the canopy,
22 where higher O3 concentrations were encountered (McBride and Laven, 1999, 053050). Because of
23 the loss of O3-sensitive pines, mixed forests of ponderosa pine, Jeffery Pine and white fir (Abies
24 concolor) shifted to predominantly white fir (Miller, 1973, 039165). Ozone may have indirectly
25 caused the decline in understory diversity in coniferous forests in the San Bernardino Mountains
26 through an increase in pine litterfall. This increase in litterfall from O3 exposure results in an
27 understory layer that may prohibit the establishment of native herbs, but not exotic annual Galium
28 aparine (Allen et al., 2007, 196876).
29 Ozone damage to conifer forests has also been observed in several other regions. In the Valley
30 of Mexico, a widespread mortality of sacred fir (Abies religiosa) was observed in the heavily
31 polluted area of the Desierto de los Leones National Park in the early 1980s (de Lourdes de Bauer
32 and Hernandez-Tejeda, 2007, 196891)(Fenn et al., 2002, 626806). Ozone damage was widely
33 believed to be an important causal factor in the dramatic decline of sacred fir. In alpine regions of
34 southern France and the Carpathians Mountains, O3 was also considered as the major cause of the
March 2011 9-83 DRAFT - DO NOT CITE OR QUOTE
-------
1 observed decline in cembran pine (Pinus cembm)(Wieser et al., 2006, 191391). However, for those
2 pollution gradient studies, several possible factors, such as drought, insect outbreak and forest
3 management, may also contribute to the mortality of trees (de Lourdes de Bauer and Hernandez-
4 Tejeda, 2007, 196891)(Wieser et al., 2006, 191391).
5 New evidence from long-term free O3 fumigation experiments provided additional support for
6 the potential impacts of O3 on species competition and community composition changes in forest
7 ecosystems. At the Aspen FACE site, community composition at both the genetic and species levels
8 was altered after seven years of fumigation with O3 (Kubiske et al., 2007, 191336). In the pure aspen
9 community, O3 fumigation reduced growth and increased mortality of sensitive clone 259, while the
10 O3 tolerant clone 8L emerged as the dominant clone. Growth of clone 8L was even greater under
11 elevated O3 compared to controls, probably due to O3 alleviated competitive pressure on clone 8L by
12 reducing growth of other clones. In the mixed aspen-birch and aspen-maple communities, O3
13 reduced the competitive capacity of aspen compared to birch and maple (Kubiske et al., 2007,
14 191336). In a phytotron study, O3 fumigation reduced growth of beech but not spruce in mixed
15 culture, suggesting a higher susceptibility of beech to O3 under interspecific competition (Kozovits
16 et al., 2005,191282).
9.6.5.2. Grassland and Agricultural Land
17 The response of managed pasture, often cultivated as a mixture of grasses and clover, to O3
18 pollution has been studied for many years. The tendency for O3-exposure to shift the biomass of
19 grass-legume mixtures in favor of grass species, reported in the previous O3 AQCD (U.S. EPA, 2006,
20 088089) has been generally confirmed by recent studies. In a mesocosm study, Trifolium repens and
21 Lolium perenne mixtures were exposed to an episodic rural O3 regime within solardomes for
22 12 weeks. T. repens showed significant changes in biomass but notZ. perenne, and the proportion of
23 T. repens decreased in O3-exposed mixtures compared to the control (Hayes et al., 2009, 191360).
24 After 5-year O3 fumigation (AOT40 of 13.3-59.5 ppm-h) at the Le Mouret FACE experiment,
25 Switzerland, legumes in fumigated plots declined from their initial over-representation (128%) to a
26 mere 59% in control plots(Volk et al., 2006, 191434). However, Stampfli and Fuhrer (2010, 102180)
27 re-analyzed the species and soil data and suggested that Volk et al. (2006, 191434) overestimated the
28 O3 effect. Stampfli and Fuhrer (2010, 102180) found that the difference in the species dynamics
29 between control and O3 treatment was more caused by heterogeneous initial conditions than O3
30 exposure. Several studies also suggested the mature/species-rich ecosystems were more resilient to
31 O3 exposure. At another FACE experiment, located at Alp Flix, Switzerland, O3 fumigation (AOT40
32 of 15.2-64.9 ppm-h) showed no significant impact on community composition of this species-rich
33 pasture (Bassin et al., 2007, 191534). Pfleeger et al. (2010, 644281) collected seed bank soil from an
34 agricultural field and examined how the plant community responded over several generations to
35 elevated O3 exposures. Sixty plant species from 22 families emerged in the chambers over their
36 four year study. Overall, they found that O3 appeared to have small impacts on seed germination and
37 only a minor effect on species richness of pioneer plant communities.
March 2011 9-84 DRAFT - DO NOT CITE OR QUOTE
-------
1 Several review papers have discussed the physiological and ecological characteristics of O3-
2 sensitive herbaceous plants. Hayes et al. (2007, 196911) assessed species traits associated with O3
3 sensitivity by the changes in biomass caused by O3 exposure. Plants of the therophyte (e.g., annual)
4 life form were particularly sensitive to O3. Species with higher mature leaf N concentration tended to
5 be more sensitive than those with lower leaf N concentration. Plants growing under high oxidative
6 stress environments, such as high light or high saline, were more sensitive to O3. Using the same
7 dataset from Hayes et al. (2007, 196911). Mills et al. (2007, 196934) identified the O3 sensitive
8 communities. They found that the largest number of these O3 sensitive communities were associated
9 with grassland ecosystems. Among grassland ecosystems, alpine grassland, sub-alpine grassland,
10 woodland fringe, and dry grassland were identified as the most sensitive communities.
9.6.5.3. Microbes
11 Several methods have been used to study microbial composition changes associated with
12 elevated O3. Phospholipid fatty acid (PLFA) analysis is widely used to determine whether O3 elicits
13 an overall effect on microbial community composition. However, since PLFA markers cover a broad
14 range of different fungi, resolution of this method may be not fine enough to detect small changes in
15 the composition of fungal communities. Methods, such as microscopic analyses and polymerase
16 chain reaction-denaturing gradient gel electrophoresis (PCR-DGGE), have better resolution to
17 specifically analyze the fungal community composition. The resolution differences among those
18 methods needs to be considered when assessing the O3 impact on microbial community composition.
19 Kanerva et al. (2008, 191264) found that elevated O3 (40-50 ppb) decreased total, bacterial,
20 actinobacterial and fungal PLFA biomass values as well as fungal:bacterial PLFA biomass ratio in
21 their meadow mesocosms in south-western Finland. The relative proportions of individual PLFAs
22 between the control and elevated O3 treatments were significantly different, suggesting that O3
23 modified the structure of the microbial community. Morsky et al. (2008, 191507) exposed boreal
24 peatland microcosms to elevated O3, with growing season AOT40 of 20.8-35.3 ppm-h, in an open-air
25 O3 exposure field in Central Finland. They also found that microbial composition was altered after
26 three growing seasons with O3 fumigation, as measured by PLFA. Ozone tended to increase the
27 presence of Gram-positive bacteria and the biomass of fungi in the peatland microcosms. Ozone also
28 resulted in higher microbial biomass, which co-occurred with the increases in concentrations of
29 organic acids and leaf density of sedges (Morsky et al., 2008, 191507). In a lysimeter study in
30 Germany, O3 was found to alter the PLFA profiles in the upper 0-20 cm rhizosphere soil of European
31 beech. Elevated O3 reduced bacterial abundance but had no detectable effect on fungal abundance
32 (Pritsch et al., 2009, 626808). Using microscopic analyses, Kasurinen et al. (2005, 191245) found
33 that elevated O3, with 5 or 6 months of AOT40 of 20.6-30.9 ppb-h, decreased the proportions of
34 black and liver-brown mycorrhizas and increased that of light brown/orange mycorrhizas. In an
35 herbaceous plant study, SSCP (single-strand conformation polymorphism) profiles indicated that O3
36 stress (about 75 ppb) had a very small effect on the structural diversity of the bacterial community in
37 rhizospheres (Dohrmann and Tebbe, 2005, 191320). At the Aspen FACE site, O3 had no significant
March 2011 9-85 DRAFT - DO NOT CITE OR QUOTE
-------
1 effect on fungal relative abundance, as indicated by PLFA profile. However, elevated O3 altered
2 fungal community composition, according to the identification of 39 fungal taxonomic units from
3 soil using polymerase chain reaction-denaturing gradient gel electrophoresis (PCR-DGGE) (Chung
4 et al., 2006, 191729). Ozone was found to change microbial community composition in an
5 agricultural system. Chen et al. (2010, 644257) found elevated O3 (100-150 ppb) had significant
6 effects on soil microbial composition expressed as PLFA percentage in a rice paddy in China.
9.6.6. Insects and Wildlife
9.6.6.1. Insects
7 Insects may respond indirectly to changes to plants (i.e., increased reactive oxygen species,
8 altered phytochemistry, altered nutrient content) that occur under elevated O3 conditions, or O3 can
9 have a direct effect on insect performance (Menendez et al., 2009, 191430). Effects of O3 on insects
10 occur at the species level (i.e., growth, survival, reproduction, development, feeding behavior) and at
11 the population and community-level (i.e., population growth rate, community composition). In
12 general, effects of O3 on insects are highly context- and species-specific (Bidart-Bouzat and Imeh-
13 Nathaniel, 2008, 191431; Lindroth, 2010, 596479). Furthermore, plant responses to O3 exposure and
14 herbivore attack have been demonstrated to share signaling pathways, complicating characterization
15 of these stressors (Lindroth, 2010, 596479: Menendez et al., 2009, 191430: Menendez et al., 2010,
16 384046)
Species-Level Responses
17 In considering insect growth, survival and reproduction in elevated O3 conditions, several
18 studies have indicated an effect while others have found no correlation. The performance of five
19 herbivore species (three moths and two weevils) was assessed in an OTC experiment at 2 x ambient
20 concentration (Peltonen et al., 2010, 596482). Growth of larvae of the Autumnal moth, Epirrita
21 autumna, was significantly decreased in the O3 treatment while no effects were observed in the other
22 species. In an aphid oviposition preference study using birch buds grown in a three year OTC
23 experiment, O3 had neither a stimulatory or deterring effect on egg-laying (Peltonen et al., 2006,
24 196946). Furthermore, changes in birch bud phenolic compounds associated with the doubled
25 ambient concentrations of O3 did not correlate with changes in aphid oviposition (Peltonen et al.,
26 2006, 196946). Reproduction in Popilliajaponica, that were fed soybeans and grown under elevated
27 O3,appeared to be unaffected (O'Neill et al., 2008, 195653). In a meta-analysis of effects of elevated
28 O3 on 22 species of trees and 10 species of insects, the rates of survival, reproduction and food
29 consumption were typically unaffected while development times were reduced and pupal masses
30 were increased (Valkama et al., 2007, 191348).
31 At the Aspen FACE site insect performance under elevated (50-60 ppb) O3 conditions
32 (approximately 1.5 x background ambient levels of 30-40 ppb O3) have been considered for several
March 2011 9-86 DRAFT - DO NOT CITE OR QUOTE
-------
1 species. Cumulative fecundity of aphids (Cepegillettea betulaefoliae), that were reared on O3-
2 exposed paper birch (Betula papyrifera) trees, was lower than aphids from control plots (Awmack et
3 al, 2004, 052926). No effects on growth, development, adult weight, embryo number and birth
4 weight of newborn nymphs were observed. In a study conducted using three aspen genotypes,
5 performance of the aspen beetle (Chrysomela crochi) decreased across all parameters measured
6 (development time, adult mass and survivorship) under elevated O3 (Vigue and Lindroth, 2010,
7 644185). There was an increase in the development time of male and female aspen beetle larvae
8 although the percentages varied across genotypes. Decreased beetle adult mass and survivorship was
9 observed across all genotypes under elevated O3 conditions. Another study from the Aspen FACE
10 site, did not find any significant effects of elevated O3 on performance (longevity, fecundity,
11 abundance) of the invasive weevil (Polydrusus sericeus) (Hillstrom et al., 2010, 644679).
12 Since the 2006 O3 AQCD, several studies have considered the effect of elevated O3 on feeding
13 behavior of insects. In a feeding preference study, the common leaf weevil (Phyllobius pyri)
14 consumed significantly more leaf discs from one aspen clone when compared to a second clone
15 under ambient air conditions (Freiwald et al., 2008, 196903). In a moderately elevated O3
16 environment (1.5 x ambient), this preference for a certain aspen clone was less evident, however,
17 leaves from O3-exposed trees were significantly preferred to leaves grown under ambient conditions.
18 Other plant-herbivore interactions have shown no effects of elevated O3 on feeding. Feeding
19 behavior of Japanese beetles (P. japonica) appeared to be unchanged when beetles were fed soybean
20 leaves grown under elevated O3 conditions (O'Neill et al., 2008, 195653). At the Aspen FACE site,
21 feeding by the invasive weevil (Polydrusus sericeus), as measured by leaf area consumption, was not
22 significantly different between foliage that was grown under elevated O3 versus ambient conditions
23 (Hillstrom et al.. 2010. 644679).
Population-Level and Community-Level Responses
24 Recent data on insects provide evidence of population-level and community-level responses to
25 O3. Elevated levels of O3 can affect plant phytochemistry and nutrient content which in turn can alter
26 population density and structure of the associated herbivorous insect communities and impact
27 ecosystem processes (Lindroth, 2010, 596479). In a long-term study of elevated O3 on herbivore
28 performance at the Aspen FACE site, individual performance and population-level effects of the
29 aphid C. betulaefoliae were assessed. Elevated O3 levels had a strong positive effect on the
30 population growth rates of the aphids; although effects were not detected by measuring growth,
31 development, adult weight, embryo number or birth weight of newborn nymphs (Awmack et al.,
32 2004, 052926). Conversely, a lower rate of population growth was observed in aphids previously
33 exposed to O3 in an OTC (Menendez et al., 2010, 384046). No direct effects of O3 were observed;
34 however, nymphs born from adults exposed to and feeding on O3 exposed plants were less capable of
35 infesting new plants when compared to nymphs in the control plots (Menendez et al., 2010, 384046).
36 Elevated O3 reduced arthropod abundance by 17% at Aspen FACE, largely as a result of the negative
37 effects on parasitoids, although phloem-feeding insects may benefit (Hillstrom and Lindroth, 2008,
March 2011 9-87 DRAFT - DO NOT CITE OR QUOTE
-------
1 191367). Herbivore communities effected by O3 and N were sampled along an air pollution gradient
2 in the Los Angeles basin (Jones and Paine, 2006, 191301). Abundance, diversity, and richness of
3 herbivores were not affected. However, a shift in community structure, from phloem-feeding to
4 chewing dominated communities, was observed along the gradient. No consistent effect of elevated
5 O3 on herbivory or insect population size was detected at SoyFACE (Dermody et al., 2008, 191850).
6 Evidence of modification of insect populations and communities in response to elevated O3
7 includes genotypic and phenotypic changes. In a study conducted at the Aspen FACE site, elevated
8 O3 altered the genotype frequencies of the pea aphid (Acyrthosiphon pisum) grown on red clover
9 (Trifolium pratense) over multiple generations (Mondor et al., 2005, 191217). Aphid color was used
10 to distinguish between the two genotypes. Ozone increased the genotypic frequencies of
11 pink-morph:green-morph aphids from 2:1 to 9:1, and depressed wing-induction responses more
12 strongly in the pink than the green genotype (Mondor et al., 2005, 191217). Growth and
13 development of individual green and pink aphids reared as a single genotype or mixed genotypes
14 were unaffected by elevated O3 (Mondor et al., 2010, 644271). Furthermore, growth of pea aphid
15 populations is not readily predictable using individual growth rates.
9.6.6.2. Wildlife
Herpatofauna
16 Since the 2006 O3 AQCD, direct effects of O3 exposure including physiological changes and
17 alterations of ecologically important behaviors such as feeding and thermoregulation have been
18 observed in wildlife. These studies have been conducted in limited laboratory exposures, and the
19 levels of O3 treatment (e.g. 0.2-0.8 ppm) were often unrealistically higher than the ambient levels.
20 Amphibians may be especially vulnerable to airborne oxidants due to the significant gas exchange
21 that occurs across the skin (Andrews et al., 2008, 645771)(Dohm et al., 2008, 604584). Exposure to
22 0.2 ppm to 0.8 ppm O3 for 4 h resulted in a decrease of oxygen consumption and depressed lung
23 ventilation in the California tree frog Pseudacris cadaverina (Mautz and Dohm, 2004, 644188).
24 Following a single 4-h exposure to O3, reduced pulmonary macrophage phagocytosis was observed
25 at 1 and 24 hours postexposure in the marine toad (Bufo marinus) indicating an effect on immune
26 system function (Dohm et al., 2005, 180452). There was no difference in macrophage function at
27 48 hours postexposure in exposed and control individuals.
28 Behavioral effects of O3 observed in amphibians include responses to minimize the surface
29 area of the body exposed to the air and a decrease in feeding rates (Dohm et al., 2008, 604584;
30 Mautz and Dohm, 2004, 644188). The adoption of a low-profile "water conservation posture" during
31 O3 exposure was observed in experiments with the California tree frog (Mautz and Dohm, 2004,
32 644188). Toads, Bufo marinus, exposed to 0.06 (iL/L O3 for 4 hours ate significantly fewer
33 mealworms at 1 hour and 48 hours postexposure than control toads (Dohm et al., 2008, 604584). In
34 the same study, escape/exploratory behavior as measured by total distance moved was not adversely
35 affected in the O3-exposed individuals as compared to the controls (Dohm et al., 2008, 604584).
March 2011 9-88 DRAFT - DO NOT CITE OR QUOTE
-------
1 Water balance and thermal preference in herpatofauna are altered with elevated O3. Toads
2 exposed to 0.8 ppm O3 for 4 hours exhibited behavioral hypothermia when preferred body
3 temperatures were measured at 1, 24 and 48 hours postexposure (Dohm et al., 2001, 017086).
4 Ozone-exposed individuals lost almost 5g more body mass on average than controls due to
5 evaporative water loss. At 24 hours after exposure, the individuals that had lost significant body
6 mass had lower preferred body temperatures(Dohm et al., 2001, 017086). Behavioral hypothermia
7 was also observed in reptiles following 4-h exposures to 0.6 ppm O3. Exposure of the Western Fence
8 Lizard (Sceloporus occidentalis) at 25°C induced behavioral hypothermia that recovered to control
9 temperatures by 24 hours (Mautz and Dohm, 2004, 644188). The behavioral hypothermic response
10 persisted in lizards exposed to O3 at 35°C at 24 hours postexposure resulting in a mean body
11 temperature 3.3°C over controls.
Soil Fauna Communities
12 Ozone has also been shown to alter soil fauna communities (Kasurinen et al., 2007, 199826;
13 Loranger et al., 2004, 072712; Meehan et al., 2010, 644184). Abundance of Acari (mites and ticks)
14 decreased by 47% under elevated O3 at AspenFACE site, probably due to the higher secondary
15 metabolites and lower N concentrations in litter and foliage under elevated O3 (Loranger et al., 2004,
16 072712). In another study from the AspenFACE site, leaf litter collected from aspen grown under
17 elevated O3 conditions were higher in fiber and lignin concentrations than trees grown under ambient
18 conditions. These chemical characteristics of the leaves were associated with an increased springtail
19 population growth following 10 weeks in a laboratory microcosm (Meehan et al., 2010, 644184).
20 Consumption rates of earthworms fed on leaf litter for 6 weeks from trees grown under elevated O3
21 conditions and ambient air did not vary significantly between treatments (Meehan et al., 2010,
22 644184). In another study on juvenile earthworms Lumbricus terrestris, individual growth was
23 reduced when worms were fed high-O3 birch litter from trees exposed for three years to elevated O3
24 in an OTC system (Kasurinen et al., 2007, 199826). In the same study no significant growth or
25 mortality effects were observed in isopods.
9.6.6.3. Indirect Effects on Wildlife
26 In addition to the direct effects of O3 exposure on physiological and behavioral endpoints
27 observed in the laboratory, there are indirect effects to wildlife. These effects include changes in
28 biomass and nutritive quality of O3-exposed plants (reviewed in Section 9.5) that are consumed by
29 wildlife. Reduced digestibility of O3-exposed plants may alter dietary intake and foraging strategies
30 in herbivores. In a study using native highbush blackberry (Rubus argutus) relative feed value of the
31 plants decreased in bushes exposed to double ambient concentrations of O3 (Ditchkoff et al., 2009,
32 192230). Indirect effects of elevated O3 on wildlife include changes in chemical signaling important
33 in ecological interactions reviewed below.
March 2011 9-89 DRAFT - DO NOT CITE OR QUOTE
-------
Chemical Signaling in Ecological Interactions
1 Ozone has been shown to degrade or alter biogenic VOC signals important to ecological
2 interactions including; (1) attraction of pollinators and seed dispersers; (2) defense against herbivory;
3 and (3) predator-prey interactions (McFrederick et al., 2009, 191329; Pinto et al, 2007, 196949;
4 Pinto et al., 2007, 196950; Pinto et al., 2010, 596490; Yuan et al., 2009, 199779). Each signal
5 released by emitters has an atmospheric lifetime and a unique chemical signature comprised of
6 different ratios of individual hydrocarbons that is susceptible to atmospheric oxidants such as O3
7 (Wright et al., 2005, 626463; Yuan et al., 2009, 199779). Under elevated O3 conditions, these
8 olfactory cues may travel shorter distances before losing their specificity (McFrederick et al., 2008,
9 196931; McFrederick et al., 2009, 191329). Additional non-phytogenic VOC-mediated
10 interrelationships with the potential to be modified by O3 include territorial marking, pheromones for
11 attraction of mates and various social interactions including scent trails, nestmate recognition and
12 signals involved in aggregation behaviors (McFrederick et al., 2009, 191329). In general, effects of
13 O3 on scent-mediated ecological interactions are highly context- and species-specific (Bidart-Bouzat
14 and Imeh-Nathaniel, 2008, 191431; Lindroth, 2010, 596479).
Pollination and Seed Dispersal
15 Phytogenic VOC's attract pollinators and seed dispersers to flowers and fruits (Dudareva et al.,
16 2006, 626458; Theis and Raguso, 2005, 626461). These floral scent trails in plant-insect interactions
17 may be destroyed or transformed by O3 (McFrederick et al., 2008, 196931). Using a Lagrangian
18 model, the rate of destruction of phytogenic VOC's was estimated in air parcels at increasing
19 distance from a source in response to increased regional levels of O3, hydroxyl and nitrate radicals
20 (McFrederick et al., 2008, 196931). Based on the model, the ability of pollinators to locate highly
21 reactive VOCs from emitting flowers may have decreased from kilometers during pre-industrial
22 times to <200 m at current ambient conditions (McFrederick et al., 2008, 196931). Scents that travel
23 shorter distances (0-10 m) are less susceptible to air pollutants, while highly reactive scents that
24 travel longer distances (10 to 100's of meters), are at a higher risk for degradation (McFrederick et
25 al., 2009, 191329). For example, male euglossine bees can detect bait stations from a distance of at
26 least one kilometer (Dobson, 1994, 626466). The alcohols, ketones and aldehydes comprising sex
27 pheromones in moths could be especially vulnerable to degradation by O3, since some males travel
28 >100 m to find mates (Carde and Haynes, 2004, 626467).
Defense Against Herbivory
29 Ozone can alter the chemical signature of VOCs emitted by plants and these VOCs are
30 subsequently detected by herbivores (Blande et al., 2010, 643928; Cannon, 1990, 626460; Iriti and
31 Faoro, 2009, 199313; Jackson et al., 1999, 026877; Pinto et al., 2007, 196949; Vuorinen et al., 2004,
32 626462). These modifications can make the plant either more attractive or repellant to phytophagous
33 insects (Pinto et al., 2010, 596490). For example, under elevated O3, the host plant preference by
34 forest tent caterpillars increased for birch compared to aspen (Agrell et al., 2005, 074324). Ozone-
March 2011 9-90 DRAFT - DO NOT CITE OR QUOTE
-------
1 induced emissions from red spruce needles were found to repel spruce budworm larvae (Cannon,
2 1990, 626460). Transcriptional profiles of field grown soybean (Glycine max) grown in elevated O3
3 conditions were altered due to herbivory by Japanese beetles. The herbivory resulted in a higher
4 number of transcripts in the leaves of O3-exposed plants and up-regulation of antioxidant metabolism
5 associated with plant defense (Casteel et al., 2008, 191696).
6 Ozone may modify signals involved in plant-to-plant interactions and plant defense against
7 pathogens (Blande et al., 2010, 643928: McFrederick et al., 2009, 191329: Pinto et al., 2010,
8 596490: Yuan et al., 2009, 199779). In a recent study with lima beans, 80 ppb O3 degraded several
9 herbivore-induced VOC's, reducing the distance over which plant-to-plant signaling occurred
10 (Blande et al., 2010, 643928).
Predator-Prey Interactions
11 Elevated O3 conditions are associated with disruption of pheromone-mediated interactions at
12 higher trophic levels (e.g., predators and parasitoids of herbivores). In a study from the Aspen FACE
13 site, predator escape behaviors of the aphid (Chatophorus stevensis) were enhanced on O3-fumigated
14 aspen trees although the mechanism of this response remains unknown (Mondor et al., 2004,
15 074334). The predatory mite Phytoseiulus persimilis can distinguish between the VOC signature of
16 ozonated lima bean plants and ozonated plants simultaneously damaged by T. urticae (Vuorinen et
17 al., 2004, 626462) however, other tritrophic interactions have shown no effect (Pinto et al., 2007,
18 196950).
19 There are few studies that consider host location behaviors of parasites under elevated O3. In
20 closed chambers fumigated with O3, the searching efficiency and proportion of the host larval fruit
21 flies parasitized by Asobara tabida, declined when compared to filtered air controls (Gate et al.,
22 1995, 026655). The host location behavior and rate of parasitism of the wasp (Coesia plutellae) on
23 Plutella xylostella-infested potted cabbage plants was tested under ambient and doubled O3
24 conditions in an open-air fumigation system (Pinto et al., 2008, 616554). The number of wasps
25 found in the field and the percentages of parasitized larvae were not significantly different from
26 controls under elevated O3.
27 Elevated O3 has the potential to perturb specialized food-web communication in transgenic
28 crops. In insect-resistant oilseed rape Brassica napus grown under 100 ppb O3 in a growth chamber,
29 reduced feeding damage by Putella xylostella led to deceased attraction of the endoparasitoid
30 (Costesia vestalis) (Himanen et al., 2009, 191338). Under chronic O3-exposure, the insect resistance
31 trait BT cryl Ac in transgenic B. napus was higher than the control (Himanen et al., 2009, 191369).
32 There was a negative relative growth rate of the Bt target herbivore, P. xylostella, in all O3
33 treatments.
March 2011 9-91 DRAFT - DO NOT CITE OR QUOTE
-------
9.7. Effects-Based Air Quality Exposure Indices and Dose
Modeling
9.7.1. Introduction
1 Exposure indices are metrics that quantify exposure as it relates to measured plant damage
2 (i.e., reduced growth). They are summary measures of monitored ambient O3 concentrations over
3 time, intended to provide a consistent metric for reviewing and comparing exposure-response effects
4 obtained from various studies. Such indices may also provide a basis for developing a biologically-
5 relevant air quality standard for protecting vegetation and ecosystems. Effects on plant growth and/or
6 yield have been a major focus of the characterization of O3 impacts on plants for purposes of the air
7 quality standard setting process (U.S. EPA, 1986, 017607KU.S. EPA, 1996, Q39046)(U.S. EPA,
8 2007, 090207). The quantitative characterization of the relationship of O3 and plant responses has
9 been referred to as "dose-response" and "exposure-response" alternatively. The distinction is in how
10 the pollutant concentration is expressed: "dose" is the pollutant concentration absorbed by the leaf
11 over some time period, and is very difficult to measure directly, whereas "exposure" is the ambient
12 air concentration measured near the plant over some time period, and summarized for that period
13 using an index. Exposure indices have been most useful in considering the form of secondary O3
14 NAAQS standard, in large part because they only require ambient air quality data rather than more
15 complex indirect calculations of dose to the plant. The attributes of exposure indices that most
16 require consideration are the weighting of O3 concentrations, and the daily and seasonal time-periods
17 that are most relevant to plant damage. Several different types of exposure indices are discussed in
18 Section 9.7.2.
19 Theoretically, a measure of plant O3 uptake or dose from ambient air (either rate of uptake or
20 cumulative seasonal uptake) might be a more ideal predictor of O3 damage to plants than an
21 exposure index and may be more useful in improving risk assessment. An uptake measure would
22 have to integrate all those environmental factors that influence stomatal conductance, including but
23 not limited to temperature, humidity, and soil water status (Section 9.7.4). Even when integrating
24 those environmental factors, a direct measure of the internal leaf concentration of O3, however, is
25 technically difficult. Therefore, uptake values are generally obtained with simulation models that
26 require knowledge of species- and site-specific values for the variables mentioned. In addition, it has
27 also been recognized that O3 detoxification processes and the temporal dynamics of detoxification
28 must be taken into account in dose modeling (Heath et al, 2009, 196783) (Section 9.7.4). Because of
29 this, research has focused historically on predictors of O3 damage to plants based only on exposure
30 as a summary measure of monitored ambient pollutant concentration over some integral of time,
31 rather than dose (Lee et al., 1988, 042136)(Lefohn and Benedict, 1982, 039395)(O'Gara. 1922,
32 015018KU.S. EPA, 1986, 017607KU.S. EPA, 1992, 042599KU.S. EPA, 1996, 080828V
March 2011 9-92 DRAFT - DO NOT CITE OR QUOTE
-------
9.7.2. Description of Exposure Indices Available in the Literature
1 Mathematical approaches for summarizing ambient air quality information in biologically
2 meaningful forms for O3 vegetation effects assessment purposes have been explored for more than
3 80 years (O'Gara, 1922, 015018)01.5. EPA, 1996, 080827). In the context of broad-scale national
4 NAAQS standards that protect for "known or anticipated" effects on many plant species in a variety
5 of habitats, exposure indices provide a numerical summary of very large numbers of ambient
6 observations of concentration over extended periods. Like any summary statistic, exposure indices
7 retain information on some characteristics of the original observations, at the exclusion of others. It
8 is expected that indices that incorporate the characteristics that are most relevant to plant response
9 will lead to more reliable protection. Several indices have attempted to incorporate some of the
10 biological, environmental, and exposure factors (directly or indirectly) that influence the magnitude
11 of the biological response and contribute to observed variability (Hogsett et al, 1988, 042128). In
12 the 1996 O3 AQCD (U.S. EPA, 1996, 080828). the exposure indices were arranged into five
13 categories; (1) One event, (2) Mean, (3) Cumulative, (4) Concentration weighted, and (5)
14 Multicomponent, and were discussed in detail (Lee et al., 1989, 042137). Figure 9-8 illustrates how
15 several of the indices weight concentration and accumulate exposure.
16
March 2011 9-93 DRAFT - DO NOT CITE OR QUOTE
-------
0.15
2
3
4
5
6
7
8
9
10
0246
Day
0.10
0.05
0.00
8 10 02468 10 02468 10
Day Day
Source: Used with permission from Air and Waste Management Association, Tingeyetal. (1991, 0426231
Figure 9-8. Diagrammatic representation of several exposure indices, illustrating howthey
weight concentration and accumulate exposure. [Ja) SUM06: the upper graphic
illustrates an episodic exposure profile; the shaded area under some of the peaks
illustrates the concentrations greater than or equal to 0.06 ppm that are
accumulated in the index. The insert shows the concentration weighting (0 to 1)
function. The lower portion of the graphic illustrates how concentration is
accumulated over the exposure period, (b) SIGMOID: the upper graphic illustrates
an episodic exposure profile; the variable shaded area under the peaks illustrates
the concentration-dependent weights that are accumulated in the index. The insert
shows the sigmoid concentration weighting function. This is similar to the W126
function. The lower portion of the graphic illustrates how concentration is
accumulated over the exposure period, (c) second HDM and M-7: the upper graphic
illustrates an episodic exposure profile. The lower portion of the graphic illustrates
that the second HDM considers only a single exposure peak, while the M-7 (average
of 7-h daily means) applies a constant exposure value over the exposure period.
Various components of the exposure-response relationship, including concentration, time of
day, respite time, frequency of peak occurrence, plant phenology, predisposition, etc., were weighted
with various functions and evaluated on their ability in ordering the regression of exposure versus
growth or yield response (Lee et al., 1989, 042137). The statistical evaluations for each of these
indices were completed using growth or yield response data from many earlier exposure studies
(e.g., NCLAN). This retrospective approach was necessary because there were no studies
specifically designed to test the goodness of fit of the various indices. The goodness of fit of a set of
linear and nonlinear models for exposure-response was ranked as various proposed indices were
used in turn to quantify exposure. This approach provided evidence for the best indices.
March 2011
9-94
DRAFT - DO NOT CITE OR QUOTE
-------
1 Most of the early retrospective studies reporting regression approaches used data from the
2 NCLAN program or data from Corvallis, Oregon or California (Lee et al., 1987, 042135)(Lee et al.,
3 1988, Q41136)(Lefohn et al., 1988, 042138¥Musselman et al., 1988, 042144)OJ.S. EPA, 1992,
4 Q42599)(U.S. EPA, 1986, 017607). These studies were previously reviewed by the EPA (U.S. EPA,
5 1992, Q42599)(U.S. EPA, 1996, 080828) and were in general agreement that the best fit of the data
6 were cumulative concentration-weighted exposure indices. Lee et al. (1987, 042135) suggested that
7 exposure indices that included all the 24-h data performed better than those that used only 7 hours of
8 data; this was consistent with the conclusions of Heagle et al. (1987, 042124) that plants receiving
9 exposures for an additional 5 h/day showed 10% greater yield loss than those exposed for 7 h/day. In
10 an earlier analysis using the National Crop Loss Assessment Network (NCLAN) data, Lee et al.
11 (1988, 042138) found the "best" exposure index was a phenologically weighted cumulative index,
12 with sigmoid weighting on concentration and a gamma weighting function as a surrogate for plant
13 growth stage. This index provided the best statistical fit when used in the models under
14 consideration, but it required data on species and site conditions, making specification of weighting
15 functions difficult for general use. The next best fits were the several indices which only cumulated
16 and weighted higher concentrations (e.g., W126, SUM06, SUM08, and AOT40). Amongst this group
17 no index lead to consistently better fits across all studies and species (Heagle et al., 1994,
18 04265 6)(Lefohn et al., 1988, 042138)(Musselman et al., 1988, 042144).
19 Other factors, including predisposition time (Hogsett et al., 1988, 042128)(McCool et al.,
20 1988, 041877) and crop development stage (Heagle et al., 1991, 042530)(Tingev et al., 2002,
21 040896) contributed to variation in the biological response and suggested the need for weighting O3
22 concentrations to account for predisposition time and phenology. However, the roles of
23 predisposition and phenology in plant response vary considerably with species and environmental
24 conditions; therefore, specification of a weighting function for general use in characterizing plant
25 exposure was not possible.
26 European scientists took a similar approach in developing indices describing growth and yield
27 loss in crops and tree seedlings, using OTCs with modified ambient exposures, but many fewer
28 species and study locations were employed in the European studies. There is evidence from some
29 European studies that a lower (Pleijel et al., 1997, 043747) or higher (Finnan et al., 1996,
30 042545)(Finnan et al., 1997, 083315) cutoff value in indices with a threshold may provide a better
31 statistical fit to the experimental data. Finnan et al. (1997, 083315) used seven exposure studies of
32 spring wheat to confirm that cumulative exposure indices emphasizing higher O3 concentrations
33 were best related to plant response and that cumulative exposure indices using weighting functions,
34 including cutoff concentrations, allometric and sigmoidal, provided a better fit and that the ranking
35 of these indices differed depending on the exposure-response model used. Weighting those
36 concentrations associated with sunshine hours in an attempt to incorporate an element of plant
37 uptake did not improve the index performance (Finnan et al., 1997, 083315). A more recent study
38 using data from several European studies of Norway spruce, analyzed the relationship between
39 relative biomass accumulation and several cumulative, weighted indices, including the AOT40 and
March 2011
9-95
DRAFT - DO NOT CITE OR QUOTE
-------
1 the SUM06 (Skarby et al., 2004, 080375). All the indices performed relatively well in regressing
2 biomass and exposure index, with the AOT20 and AOT30 doing slightly better than others (r2 =
3 0.46-0.47). In another comparative study of four independent data sets of potato yield and different
4 cumulative uptake indices with different cutoff values, a similarly narrow range of r2 was observed
5 (r2 = 0.3-0.4) (Pleijel et al., 2004, 036662).
6 In both the U.S. and Europe, the adequacy of these numerical summaries of exposure in
7 relating biomass and yield changes have, for the most part, all been evaluated using data from
8 studies not necessarily designed to compare one index to another (Lefohn et al., 1988, 042138)(Lee
9 et al., 1989, 042137)(Skarbv et al., 2004, 080375). Very few studies in the U.S. have addressed this
10 issue since the 2006 O3 AQCD. McLaughlin et al. (2007, 090348) reported that the cumulative
11 exposure index of AOT60 related well to reductions in growth rates at forest sites in the southern
12 Appalachian Mountains. However, the authors did not report an analysis to compare multiple
13 indices. Overall, given the available data from previous O3 AQCDs and the few recent studies, the
14 cumulative, concentration-weighted indices perform better than the peak or mean indices. It is still
15 not possible, however, to distinguish the differences in performance among the cumulative,
16 concentration-weighted indices.
17 The main conclusions from the 1996 and 2006 O3 AQCDs (U.S. EPA, 1996,
18 080828)(U.S. EPA, 2006, 088089) regarding an index based on ambient exposure are still valid. No
19 information has come forth since the 2006 O3 AQCD to alter those conclusions significantly. These
20 key conclusions can be restated as follows:
21 • O3 effects in plants are cumulative;
22 • higher O3 concentrations appear to be more important than lower concentrations in
23 eliciting a response;
24 • plant sensitivity to O3 varies with time of day and plant development stage; and
25 • exposure indices that accumulate the O3 hourly concentrations and preferentially weight
26 the higher concentrations have better statistical fits to growth/yield response than do the
27 mean and peak indices.
28 Following the 2006 criteria review process (U.S. EPA, 2006, 088089). the EPA proposed an
29 alternative form of the secondary NAAQS for O3 using a cumulative, concentration-weighted
30 exposure index to protect vegetation from damage (72 FR 37818 (2007, 684055). 75 FR 2938 (2010,
31 684211). p. 3003). The EPA considered two specific concentration-weighted indices: the cutoff
32 concentration weighted SUM06 and the sigmoid-weighted W126 exposure index (U.S. EPA, 2007,
33 090207). These two indices performed equally well in predicting the exposure-response relationships
34 observed in the crop and tree seedlings studies (Lee et al., 1989, 042137). At a workshop convened
March 2011 9-96 DRAFT - DO NOT CITE OR QUOTE
-------
1 to consider the science supporting these indices (Heck and Cowling, 1997, 084484) the participants
2 agreed that these cumulative concentration-weighted indices being considered were equally capable
3 of predicting plant response. A short-term daily 8-h avg was also proposed in the review that ended
4 in 2008 (73 FR 16436, (2008, 684051)). but at the time there were no scientific studies supporting
5 the utility of this index for vegetation. Below are the definitions of the three index forms considered
6 in the previous staff paper review (U.S. EPA, 2007, 090207):
7 • 8-h average form: 4th-highest daily max 8-h avg over the O3 season.
8 • SUM06: Sum of all hourly O3 concentrations greater than or equal to 0.06 ppm observed
9 during a specified daily and seasonal time window (Figure 9-8a).
10 • W126: Sigmoidally weighted sum of all hourly O3 concentrations observed during a
11 specified daily and seasonal time window (Similar to Figure 9-8b). The sigmoidal
12 weighting of hourly O3 concentration is given in the equation below, where C is the
13 hourly O3 concentration in ppm:
1
c 1 + 4403e-126C
Equation 9-1
14 The SUM06 and W126 indices have a variety of relevant time windows that may be applied and are
15 discussed in Section 9.7.3.
16 Other exposure indices are discussed in the literature and are currently used outside the U.S. In
17 Europe, the cutoff concentration-weighted index AOT40 was selected in developing exposure-
18 response relationships based on OTC studies of a limited number of crops and trees (Grunhage and
19 Jager, 2003, 052972). The United Nations Economic Commission for Europe (United Nations
20 Economic Commission for Europe UNECE, 1988, 055354) adopted the critical levels approach for
21 assessment of O3 risk to vegetation across Europe. As used by the UNECE, the critical levels are not
22 like the air quality regulatory standards used in the U.S., but rather planning targets for reductions in
23 pollutant emissions to protect ecological resources. Critical levels for O3 are intended to prevent
24 long-term deleterious effects on the most sensitive plant species under the most sensitive
25 environmental conditions, but not intended to quantify O3 effects. A critical level was defined as "the
26 concentration of pollutant in the atmosphere above which direct adverse effects on receptors, such as
27 plants, ecosystems, or materials may occur according to present knowledge" (United Nations
28 Economic Commission for Europe UNECE, 1988, 055354). The nature of the "adverse effects" was
29 not specified in the original definition, which provided for different levels for different types of
30 harmful effect (e.g., visible injury or loss of crop yield). There are also different critical levels for
31 crops, forests, and semi-natural vegetation. The caveat, "according to present knowledge" is
March 2011 9-97 DRAFT - DO NOT CITE OR QUOTE
-------
1 important because critical levels are not rigid; they are revised periodically as new scientific
2 information becomes available. For example, the original critical level for O3 specified
3 concentrations for three averaging times, but further research and debate led to the current critical
4 level being stated as the cumulative exposure (concentration x hours) over a cutoff concentration of
5 40 ppb (AOT40) (Fuhrer et al, 1997, 030380V
6 In Europe, a decision was made to work towards a flux-based approach for the critical levels
7 ("Level II"), with the goal of modeling O3 flux-effect relationships for three vegetation types: crops,
8 forests, and semi-natural vegetation (Grunhage and Jager, 2003, 052972). Progress has been made in
9 modeling flux (see section AX9.4.5; U.S. EPA, 2006, 088089) and the Mapping Manual is being
10 revised (Ashmore et al., 2004, 056621)(Ashmore MEmberson et al., 2004, 056624)(Grennfelt, 2004,
11 056625)(Karlsson et al., 2003, 055331). The revisions may include a flux-based approach for three
12 crops: wheat, potatoes, and cotton. However, because of a lack of flux-response data, a cumulative,
13 cutoff concentration-based (AOTx) exposure index will remain in use for the near future for most
14 crops and for forests and semi-natural herbaceous vegetation (Ashmore et al., 2004, 056623).
9.7.3. Important Components of Exposure Indices
15 The efficacy of exposure indices in predicting biological responses requires that researchers
16 identify a relationship between measured growth and/or yield effects and important components of
17 exposure indices. In the previous O3 AQCDs it was established that higher hourly concentrations
18 have greater effects on vegetation than lower concentrations (U.S. EPA, 1996, 080828)(U.S. EPA,
19 2006, 088089). Further, it was determined that the diurnal and seasonal duration of exposure is
20 important for plant response. Weighting of hourly concentrations and the diurnal and seasonal time
21 window of exposure are the most important variables in a cumulative exposure index and will be
22 discussed below. However, these variables must be taken in the context of plant phenology, diurnal
23 conductance rates, plant canopy structure, and detoxification mechanisms of vegetation as well as
24 the climate and meteorology, all of which are determinants of plant response. These more specific
25 factors will be discussed in the uptake and dose modeling section (Section 9.7.4).
9.7.3.1. Role of Concentration
26 The significant role of higher O3 concentrations was established based on several experimental
27 studies (U.S. EPA, 1996, 080828). Several studies (Nussbaum et al., 1995, 030141¥Oksanen and
28 Holopainen, 2001, 019538)(Yun and Laurence, 1999, 044165) have added support for the important
29 role that peak concentrations, as well as the pattern of occurrence, plays in plant response to O3.
30 Oksanen and Holopainen (2001, 019538) found that the peak concentrations and the shape of the O3
31 exposure (i.e., duration of the event) were important determinants of foliar injury in European white
32 birch saplings, but growth reductions were found to be more related to total cumulative exposure.
33 Based on air quality data from 10 U.S. cities, three 4-week exposure treatments having the same
34 SUM06 value were constructed by Yun and Laurence (1999, 044165). The authors used different
March 2011 9-98 DRAFT - DO NOT CITE OR QUOTE
-------
1 exposure regimes to explore effects of treatments with variable versus uniform peak occurrence
2 during the exposure period. The authors reported that the variable peak exposures were important in
3 causing injury, and that the different exposure treatments, although having the same SUM06,
4 resulted in very different patterns of foliar injury. Nussbaum et al. (1995, 030141) also found peak
5 concentrations and the pattern of occurrence to be critical in determining the measured response. The
6 authors recommended that to describe the effect on total forage yield, peak concentrations
7 >0.11 ppm must be emphasized by using an AOT with higher threshold concentrations.
8 A greater role for higher concentrations affecting plant growth might be inferred based on air
9 quality analyses for the southern California area (Lee et al., 2003, 05303 l)(Tingey et al., 2004,
10 042385). In the late 1960s and 1970s, extremely high O3 concentrations had impacted the San
11 Bernardino National Forest. However, over the past 20+ years, significant reductions in O3 exposure
12 have occurred (Davidson, 1993, 043366)(Lee et al., 2003, 053031)(Lefohn and Shadwick, 2000,
13 040483)(Bytnerowicz et al., 2008, 196881). An illustration of this improvement in air quality is
14 shown by the 37-year history of O3 air quality at the Crestline site in the San Bernardino Mountains
15 (Figure 9-9) (Lee et al., 2003, 053031). Ozone exposure increased from 1963 to 1979 concurrent
16 with increased population and vehicular miles, followed by a decline to the present mirroring
17 decreases in precursor emissions. The pattern in exposure was evident in various exposure indices
18 including the cumulative concentration weighted (SUM06), as well as maximum peak event (1 h
19 peak), and the number of days having hourly averaged O3 concentrations greater than or equal to
20 95 ppb. The number of days having hourly averaged O3 concentrations greater than or equal to
21 95 ppb declined significantly from 163 days in 1978 to 103 days in 1997. The changes in ambient O3
22 air quality for the Crestline site were reflected in the changes in frequency and magnitude of the peak
23 hourly concentration and the duration of exposure (Figure 9-9). Considering the role of exposure
24 patterns in determining response, the seasonal and diurnal patterns in hourly O3 concentration did not
25 vary appreciably from year to year over the 37-year period (Lee et al., 2003, 053031).
26 The inference for a role of higher concentrations comes both from results of measures of tree
27 conditions on established plots and from results of model simulations. Across a broad area of the San
28 Bernardino National Forest, the Forest Pest Management (FPM) method of injury assessment
29 indicated an improvement in crown condition from 1974 to 1988; and the area of improvement in
30 injury assessment is coincident with an improvement in O3 air quality (Miller and Rechel, 1999,
31 040702). A more recent analysis of forest changes in the San Bernardino National Forest using an
32 expanded network of monitoring sites has verified significant changes in growth, mortality rates,
33 basal area, and species composition throughout the area since 1974 (Arbaugh et al., 2003, 052925).
34 A model simulation of ponderosa pine growth over the 40-year period in the San Bernardino
35 National Forest showed a significant impact of O3 exposure on tree growth and indicates improved
36 growth with improving O3 air quality. This area has also experienced elevated N deposition and
37 based on a number of environmental indicators, it appears that this area is experiencing N saturation
38 (Fenn et al., 1996, 083540). To account for this potential interaction, the model simulations were
39 conducted under conditions of unlimited soil N. The actual interactions are not known. The
March 2011 9-99 DRAFT - DO NOT CITE OR QUOTE
-------
1 improvement in growth over the years was assigned to improved O3 air quality, but no distinction
2 was made regarding the relative role of mid-range and higher hourly concentrations, only that
3 improved growth tracked decreasing SUM06, maximum peak concentration, and number of days of
4 hourly O3 >95 ppb (Tingey et al, 2004, 042385). A summary of air quality data from 1980 to 2000
5 for the San Bernardino National Forest area of the number of "mid-range" hourly concentrations
6 indicated no dramatic changes over this 20-year period, ranging from about 1,500 to 2,000 hours per
7 year (Figure 9-10). There was a slow increase in the number of mid-range concentrations from 1980
8 to 1986, which corresponds to the period after implementation of the air quality standard. Another
9 sharper increase was observed in the late 1990s. This pattern of occurrence of mid-range hourly
10 concentrations suggests a lesser role for these concentration ranges compared to the higher values in
11 either of the ground-level tree injury observations of the model simulation of growth over the
12 40-year period.
March 2011 9-100 DRAFT - DO NOT CITE OR QUOTE
-------
c >.
.2-g
*J O
•—• o
^
EO
.0) °
o
o
CM
S E
3 S-
w S
(Q ^
0) .Q
0- 0.
o
o
CM
O 0> £
Z Al
OO
i O \ / \ o
V oo
,00 o oo^ /°.
• ° \/ O 00 X«*» •
1965 1970 1975 1980 1985 1990 1995 2000
Year
Source: Used with permission from Elsevier Science Ltd., (Lee et al., 2003, 0530311.
Figure 9-9. Trends in May to September 12-h SUM06, peak 1-h ozone concentration and number
of daily exceedances of 95 ppbforthe Crestline site in 1963 to 1999 in relation to
trends in mean daily maximum temperature for Crestline and daily reactive organic
gases (ROG) and oxides of nitrogen (NOX) for San Bernardino County. Annual ROG
and NOX emissions data for San Bernardino County were obtained from Alexis et al.
(2001, 079886) and the California Air Resource Board's emission inventory available
at http://www.arb.ca.gov/aqd/aqdpage.htm (Cal EPA, 2009, 677487).
March 2011
9-101
DRAFT - DO NOT CITE OR QUOTE
-------
Crestline, San Bernardino, CA
Number of Hours 50 - 89 ppb
060710005
(A
3
O
2500
2000
1500
1000
500
I "" I '" I t I I I
I ™~ I ~" I ™~ [
I • I • r ™ I
o
00
en
(N
CO
05
s
C>
tO
00
O
CO
CO
en
CM
(O
CO
O
O
CN
Year
Figure 9-10.The number of hourly average concentrations between 50 and 89 ppb for the period
1980-2000 for the Crestline, San Bernardino County, CA, monitoring site.
9.7.3.2. Diurnal and Seasonal Exposure
Diurnal Exposure
1 The diurnal patterns of maximal leaf/needle conductance and occurrence of higher ambient
2 concentrations can help determine which hours during the day over a season should be cumulated.
3 Stomatal conductance is species and phenology dependent and is linked to both diurnal and seasonal
4 meteorological activity as well as to soil/site conditions (e.g., soil moisture). Daily patterns of
5 leaf/needle conductance are often highest in midmorning, whereas higher ambient O3 concentrations
6 generally occurred in early to late afternoon when stomata were often partially closed and
7 conductances were lower. Total O3 flux depends on atmospheric and boundary layer resistances, both
8 of which exhibit variability throughout the day. Experimental studies with tree species demonstrated
9 the decoupling of ambient O3 exposure, peak occurrence, and gas exchange, particularly in areas of
10 drought (Panek, 2004, 079202). Several studies have suggested that ponderosa pine trees in the
11 southern and northern Sierra Nevada Mountains may not be as susceptible to high O3 concentrations
12 as to lower concentrations, due to reduced needle conductance and O3 uptake during the period when
13 the highest concentrations occur (Arbaugh et al, 1998, 040297)(Bauer et al., 2000, 040315)(Panek et
14 al., 2002, 0407_12)(Panek and Goldstein, 2001, 030190). Panek et al. (2002, 040712) compared direct
15 O3 flux measurements into a canopy of ponderosa pine and demonstrated a lack of correlation of
16 daily patterns of conductance and O3 occurrence, especially in the late season drought period; the
March 2011
9-102
DRAFT - DO NOT CITE OR QUOTE
-------
1 authors concluded that a consideration of climate or season was essential, especially considering the
2 role of soil moisture and conductance/uptake. In contrast, Grulke et al. (2002, 035283) reported high
3 conductance when O3 concentrations were high in the same species, but under different growing site
4 conditions. The decoupling of conductance and higher ambient O3 concentration would hold true for
5 more mesic environments as well as xeric landscapes. The longer-term biological responses reported
6 by Miller and Rechel (1999, 040702) for ponderosa pine in the same region, and the general
7 reduction in recent years in ambient O3 concentrations, suggest that stomatal conductance alone may
8 not be a sufficient indicator of potential vegetation injury or damage. Another consideration for the
9 effect of O3 uptake is the diurnal pattern of detoxification capacity of the plant. The detoxification
10 capacity may not follow the same pattern as stomatal conductance (Heath et al., 2009, 196783).
11 A 12-h (8:00 a.m. to 8:00 p.m.) daylight period for a W126 cumulating exposure was proposed
12 in 2007 (72 FR 37818, (2007, 684055)) and 2010 (75 FR 2938, (2010, 684211). p. 3003) following
13 the release of the 2006 O3 AQCD, based primarily on evidence that the conditions for uptake of O3
14 into the plant occur mainly during the daytime hours. In general, plants have the highest stomatal
15 conductance during the daytime and in many areas atmospheric turbulent mixing is greatest during
16 the day as well (U.S. EPA, 2006, 088089)(Uddling et al., 2010, 387073). However, notable
17 exceptions to maximum daytime conductance are cacti and other plants with crassulacean acid
18 metabolism (CAM photosynthesis) which only open their stomata at night. This section will focus on
19 plants with C3 and C4 photosynthesis, which generally have maximum stomatal conductance during
20 the daytime.
21 Recent reviews of the literature reported that a large number of species had varying degrees of
22 nocturnal stomatal conductance (Caird et al., 2007, 199337)(Dawson et al., 2007,
23 67038l)(Musselman and Minnick, 2000, 011612). This night-time conductance can also be
24 enhanced by O3 damage during the day that could result in loss of stomatal control, and less
25 complete closure of stomata, than under low O3 conditions (Grulke et al., 2007, 199376). In general,
26 the rate of stomatal conductance at night is much lower than during the day (Caird et al., 2007,
27 199337). Atmospheric turbulence at night is also often low, which results in stable boundary layers
28 and unfavorable conditions for O3 uptake into vegetation (Finkelstein et al., 2000, 024029).
29 Nevertheless, nocturnal turbulence does intermittently occur and may result in nonnegligible O3 flux
30 into the plants. In addition, plants might be more susceptible to O3 exposure at night than during the
31 daytime, because of potentially lower plant defenses (Musselman and Minnick, 2000,
32 011612)(Musselman et al., 2006, lH678)(Loreto and Fares, 2007, 180259)(Heath et al., 2009,
33 196783). For significant nocturnal stomatal flux and O3 effects to occur, specific conditions must
34 exist. A susceptible plant with nocturnal stomatal conductance and low defense must be growing in
35 an area with relatively high night-time O3 concentrations and appreciable nocturnal atmospheric
36 turbulence. It is unclear how many areas there are in the U.S. where these conditions occur. It may
37 be possible that these conditions exist in mountainous areas of southern California, front-range of
38 Colorado (Turnipseed et al., 2009, 588752) and the Great Smoky Mountains of North Carolina and
March 2011 9-103 DRAFT - DO NOT CITE OR QUOTE
-------
1 Tennessee. More information is needed in these locations to assess the local O3 patterns,
2 micrometeorology and responses of potentially vulnerable plant species.
3 Several field studies have attempted to quantify night-time O3 uptake with a variety of
4 methods. However, many of these studies have not linked the night-time flux to measured effects on
5 plants. Grulke et al. (2004, 042646) showed that the stomatal conductance at night for ponderosa
6 pine in the San Bernardino National Forest (CA) ranged from one tenth to one fourth that of
7 maximum daytime stomatal conductance. In June, at a high-elevation site, it was calculated that 11%
8 of the total daily O3 uptake of pole-sized trees occurred at night. In late summer, however, O3 uptake
9 at night was negligible. However, this study did not consider the turbulent conditions at night.
10 Finklestein et al. (2000, 024029) investigated O3 deposition velocity to forest canopies at three
11 different sites. The authors found the total flux (stomatal and non-stomatal) to the canopy to be very
12 low during night-time hours as compared to day-time hours. However, the authors did note that
13 higher nocturnal deposition velocities at conifer sites may be due to some degree of stomatal opening
14 at night (Finkelstein et al., 2000, 024029). Work by Mereu et al. (2009, 102168) in Italy on
15 mediterranean species indicated that nocturnal uptake was from 10 to 18% of total daily uptake
16 during a weak drought and up to 24% as the drought became more pronounced. The proportion of
17 night-time uptake was greater during the drought due to decreases in daytime stomatal conductance
18 (Mereu et al., 2009, 102168). In a recent study at the AspenFACE site in Wisconsin, calculated leaf-
19 level stomatal O3 flux was near zero from the night-time hours of 8:00 p.m. to 5:00 a.m. (Uddling et
20 al., 2010, 387073). This was likely due to low horizontal wind speed (>1 m/s) and low O3
21 concentrations (<25 ppb) during those same night-time hours (Figure 9-11).
March 2011 9-104 DRAFT - DO NOT CITE OR QUOTE
-------
O.I 5
•N 0.10
E
o"
~o
IL 0.05
0.00
5 10 15 20
Time of day
§o
o
E 40
"q
£ 30
c
O
c
8
o
10
S 10 15 20
Time of day
E 5
o
! 4
s
«5
IJL
en
U_
Fst.,,,,,2004
• Fstm<1,2004
10 15
Time of day
E 5
o
! 4
e
on
u.
Fst,.u,.20Q5
Fst,nn,2005
0 5 10 15 ZO
Time of day
Source: Used with permission from Elsevier Ltd., Uddling et al. (2010, 3870731.
Figure 9-11. Mean diurnal.
(a) conductance through boundary layer and stomata (gbs), (b) Ozone concentration, and leaf-
level stomatal ozone flux without flux cut-off threshold (FstOi) in control plots from
mid-June through August in (c) 2004 and (d) 2005 in the Aspen FACE experiment.
Subscripts "max" and "min" refer to stomatal fluxes calculated neglecting and
accounting for potential non-stomatal ozone flux, respectively.
1 A few studies have tested the biological effects of night-time O3 exposure on vegetation in
2 controlled chambers. Biomass of ponderosa pine seedlings was significantly reduced when seedlings
3 were exposed to either daytime or nighttime episodic profiles (Lee and Hogsett, 1999, 040451).
4 However, the biomass reductions were much greater with daytime peak concentrations than with
5 nighttime peak concentrations. Similarly, birch cuttings grown in field chambers that were exposed
6 to O3 at night only, daytime only, and 24 hours showed similar reductions in biomass in night only
7 and day only treatments. Birch seedling showed greater reductions in growth in 24-h exposures than
8 those exposed to O3 at night or day only (Matyssek et al., 1995, 040700). Field mustard (Brassica
9 rapa) plants exposed to O3 during the day or night showed little significant difference in the amounts
10 of injury or reduced growth response to O3 treatment, although the stomatal conductance was
11 70-80% lower at night (Winner et al., 1989, 043403). These studies show that effects can be seen
March 2011
9-105
DRAFT - DO NOT CITE OR QUOTE
-------
1 with night-time exposures to O3 but if atmospheric conditions are stable at night, it is uncertain how
2 these exposures may affect plants and trees with complex canopies in the field.
Seasonal Exposure
3 Vegetation across the U.S. has widely varying periods of physiological activity during the year
4 due to variability in climate and phenology. In order for a particular plant to be vulnerable to O3
5 pollution, it must have foliage and be physiologically active. Annual crops are typically grown for
6 periods of two to three months. In contrast, perennial species may be photosynthetically active
7 longer (up to 12 months each year for some species) depending on the species and where it is grown.
8 In general, the period of maximum physiological activity and thus, potential O3 uptake for vegetation
9 coincides with some or all of the intra-annual period defined as the O3 season, which varies on a
10 state-by-state basis (Figure 3-18). This is because the high temperature and high light conditions that
11 typically promote the formation of tropospheric O3 also promote physiological activity in vegetation.
12 There are very limited exceptions to this pattern where O3 can form in the winter in areas in the
13 western U.S. with intense natural gas exploration (Pinto, 2009, 187038). but this is typically when
14 plants are dormant and there is little chance of O3 uptake. The selection of any single window of
15 time for a national standard to consider hourly O3 concentrations represents a compromise, given the
16 significant variability in growth patterns and lengths of growing season among the wide range of
17 vegetation species that may experience adverse effects associated with O3 exposure.
18 Various intra-annual averaging and accumulation time periods have been considered for the
19 protection of vegetation. The 2010 proposal for secondary O3 standard (75 FR 2938, (2010, 684211).
20 p. 3003) proposed to use the maximum consecutive 3-month period within the O3 season. The U.S.
21 Forest Service and federal land managers have used a 24-h W126 accumulated for 6 months from
22 April through September (see FLAG report; Federal land managers' air quality related values
23 workgroup (FLAG) phase I report, 2000, 088923). However, some monitors in the U.S. are
24 operational for as little as four months and would not have enough data for a 6-month seasonal
25 window. The exposure period in the vast majority of O3 exposure studies conducted in the U.S. has
26 been much shorter than 6 months. Most of the crop studies done through NCLAN had exposures less
27 than three months with an average of 77 days. Open-top chamber studies of tree seedlings, compiled
28 by the EPA, had an average exposure of just over three months or 99 days. In more recent FACE
29 experiments, SoyFACE exposed soybeans for an average of approximately 120 days per year and the
30 Aspen FACE experiment exposed trees to an average of approximately 145 days per year of elevated
31 O3, which included the entire growing season at those particular sites. Despite the possibility that
32 plants may be exposed to ambient O3 longer than 3 months in some locations, there is a lack of
33 exposure experiments conducted for longer than 3 months.
34 In an analysis of the 3- and 6-month maximum W126 values calculated for over 1,200 AQS
35 (Air Quality System) and CASTNET (Clean Air Status and Trend Network) EPA monitoring sites for
36 the years 2008-2009, it was found that these 2 accumulation periods resulted in highly correlated
37 metrics (Figure 9-12). The two cumulation periods were centered on the yearly maximum for each
March 2011 9-106 DRAFT - DO NOT CITE OR QUOTE
-------
1 monitoring site, and it is possible that this correlation would be weaker if the two periods were not
2 temporally aligned. In the U.S., W126 cumulated over 3 months, and W126 cumulated over 6
3 months are proxies of one another, as long as the period in which daily W126 is accumulated
4 corresponds to the seasonal maximum. Therefore, it is expected that either statistic will predict
5 vegetation response equally well. In other words, the strength of the correlation between maximum
6 3-month W126 and maximum 6-month W126 is such that there is no material difference in their
7 predictive value for vegetation response.
March 2011 9-107 DRAFT - DO NOT CITE OR QUOTE
-------
Highest 3 month W126
A
Highest 3 month W126
B
Figure 9-12.Maximum 3-month, 12-h W126 plotted against maximum 6-month, 12-h W126. Data
are from the AQS and CASTNET monitors for the years 2008 and 2009. (A) W126,
3 month versus 6 month, 2008 (Pearson correlation = 0.99); (B) W126,3 month
versus 6 month, 2009 (Pearson correlation = 0.99).
9.7.4. Ozone Uptake/Dose Modeling for Vegetation
1 Another approach for improving risk assessment of vegetation response to ambient O3 is based
2 on estimating the O3 concentration from the atmosphere that enters the leaf (i.e., flux or deposition).
March 2011
9-108
DRAFT - DO NOT CITE OR QUOTE
-------
1 Interest has been increasing in recent years, particularly in Europe, in using mathematically tractable
2 flux models for O3 assessments at the regional, national, and European scale (Matyssek et al., 2008,
3 191262)(Paoletti and Manning, 2007, 180174)(Emberson et al., 2000, 04_0350)(Emberson et al.,
4 2000, 042537)(ICP M&M, 2004, 677471). Some researchers have claimed that using flux models
5 can be used to better predict vegetation responses to O3 than exposure-based approaches (Matyssek
6 et al., 2008, 191262). However, other research has suggested that flux models do not predict
7 vegetation responses to O3 better than exposure-based models, such as AOT40 (Gonzalez-Fernandez
8 et al., 2010, 381357). While some efforts have been made in the U.S. to calculate O3 flux into leaves
9 and canopies (Grantz et al., 1997, 026664)(Grantz et al., 1995, 026659)(Grulke et al., 2004,
10 042646)(Turnipseed et al., 2009, 588752)(Uddling et al., 2009, 596219)(Bergweiler et al., 2008,
11 191656)(Hogg et al., 2007, 199349). little information has been published relating these fluxes to
12 effects on vegetation. The lack of flux data in the U.S. and the lack of understanding of
13 detoxification processes have made this technique less viable for vulnerability and risk assessments
14 in the U.S.
15 Flux calculations are data intensive and must be carefully implemented. Reducing
16 uncertainties in flux estimates for areas with diverse surface or terrain conditions to within ±50%
17 requires "very careful application of dry deposition models, some model development, and support
18 by experimental observations" (Wesely and Hicks, 2000, 025018). As an example, the annual
19 average deposition velocity of O3 among three nearby sites in similar vegetation was found to vary
20 by ±10%, presumably due to terrain (Brook et al., 1997, 041857). Moreover, the authors stated that
21 the actual variation was even greater, because stomatal uptake was unrealistically assumed to be the
22 same among all sites, and flux is strongly influenced by stomatal conductance (Brook et al., 1997,
23 041857). This uptake-based approach to quantify the vegetation impact of O3 requires inclusion of
24 those factors that control the diurnal and seasonal O3 flux to vegetation (e.g., climate patterns,
25 species and/or vegetation-type factors and site-specific factors). The models have to distinguish
26 between stomatal and non-stomatal components of O3 deposition to adequately estimate actual
27 concentration reaching the target tissue of a plant to elicit a response (Uddling et al., 2009, 596219).
28 Determining this O3 uptake via canopy and stomatal conductance by necessity relies on models to
29 predict flux and ultimately the "effective" flux (Grunhage et al., 2004, 05662l)(Massman et al.,
30 2000, 011616)(Massman. 2004, 055350). "Effective flux" has been defined as the balance between
31 O3 flux and detoxification processes (Dammgen et al., 1993, 055312)(Grunhage and Haenel, 1997,
32 Q4_0392)(Musselman and Massman, 1999, 040706)(Heath et al., 2009, 196783). The time-integrated
33 "effective flux" is termed "effective dose." The uptake mechanisms and the resistances in this
34 process, including stomatal conductance and biochemical defense mechanisms, are discussed below.
35 The flux-based index is the goal for the "Level II" critical level for assessment of O3 risk to
36 vegetation and ecosystems across Europe (Ashmore MEmberson et al., 2004, 056624).
March 2011
9-109
DRAFT - DO NOT CITE OR QUOTE
-------
9.7.4.1. Canopy Structure
1 A factor important in both O3 exposure and uptake is how canopy structure affects O3
2 concentration in and under forest canopies. There have been several investigations of O3
3 concentrations under tree canopies (Enders, 1992, 040355)(Fontan et al, 1992,
4 040370)(Fredericksen et al., 1995, 038898)(Joss and Graber, 1996, 040408)(Kolb et al., 1997,
5 Q52597)(Lorenzini and Nali, 1995, 04069 DfNeufeld et al., 1992, 038971)(Samuelson and Kelly,
6 1997, 040832). In general, they indicated a reduction in O3 of ~20 to 40% in the area below the
7 canopy but above the shrub/herb layers. An essential component in the determination of the AOT40
8 as a critical level was the height at which the O3 concentration was measured. The measurement
9 heights are related to the O3 concentration measured at the top of the canopy, i.e., upper surface
10 boundary of the (quasi-) laminar layer (Grunhage and Jager, 2003, 052972). This location is
11 presumably more closely related to stomatal uptake. Weighting the O3 concentration at this location
12 takes into account stomatal opening and, if weighted with the Jarvis-Steward factors for radiation,
13 temperature, and soil moisture, the "toxicologically" effective AOT40 is obtained (Grunhage and
14 Jager, 2003, 052972). A question exists however as to whether this "canopy" O3 concentration is
15 clearly connected to stomatal O3 uptake. During site conditions that limit stomatal conductance (e.g.,
16 low soil moisture, high VPD), high concentrations of O3 can occur at the top of the canopy with
17 minimal risk.
9.7.4.2. Site and Climate Factors
18 Soil moisture is a critical factor in controlling O3 uptake through its effect on plant water status
19 and stomatal conductance. In an attempt to relate uptake, soil moisture, and ambient air quality to
20 identify areas of potential risk, available O3 monitoring data for 1983 to 1990 were used along with
21 literature-based seedling exposure-response data from regions within the southern Appalachian
22 Mountains that might have experienced O3 exposures sufficient to inhibit growth (Lefohn et al.,
23 1997, 082871). In a small number of areas within the region, O3 exposures and soil moisture
24 availability were sufficient to possibly cause growth reductions in some O3 sensitive species (e.g.,
25 black cherry). The conclusions were limited, however, because of the uncertainty in interpolating O3
26 exposures in many of the areas and because the hydrologic index used might not reflect actual water
27 stress.
9.7.4.3. Plant Defens e Mechanis m - Detoxification
28 The non-stomatal component of plant defenses are the most difficult to quantify, but some
29 studies are available (Barnes et al., 2002, 040313)(Chen et al., 1998, 040317)(Massman and Grantz,
30 1995, 040698)(Plochl et al., 2000, 040802)(Heath et al., 2009, 196783). Massman et al. (2000,
31 011616) developed a conceptual model of a dose-based index to determine how plant injury response
32 to O3 relates to the traditional exposure-based parameters. The index used time-varying-weighted
33 fluxes to account for the fact that flux was not necessarily correlated with plant injury or damage.
March 2011 9-110 DRAFT-DO NOT CITE OR QUOTE
-------
1 The model applied only to plant foliar injury and suggested that application of flux-based models for
2 determining plant damage (yield or biomass) would require a better understanding and quantification
3 of the relationship between injury and damage.
9.8. Ozone Exposure-Plant Response Relationships
9.8.1. Introduction
4 The adequate characterization of the effects of O3 on plants for the purpose of setting air
5 quality standards is contingent not only on the choice of the index used (i.e. SUM06, W126) to
6 summarize O3 concentrations (Section 9.7), but also on quantifying the response of the plant
7 variables of interest at specific values of the selected index. The many factors that determine the
8 response of plants to O3 exposure have been discussed in previous sections. They include species,
9 genotype and other genetic characteristics (Section 9.4), biochemical and physiological status
10 (Section 9.4), previous and current exposure to other stressors (Section 9.5), and characteristics of
11 the exposure itself (Section 9.7). Establishing a secondary air quality standard requires the capability
12 to generalize those observations, in order to obtain predictions that are reliable enough under a broad
13 variety of scenarios, taking into account these factors. This section reviews results that have related
14 specific quantitative observations of O3 exposure with quantitative observations of plant responses,
15 and the predictions of responses that have been derived from those observations through empirical
16 models.
17 For four decades, exposure to O3 at ambient concentrations found in many areas of the U.S.
18 has been known to cause detrimental effects in plants (U.S. EPA, 2006, 088089)(U.S. EPA, 1996,
19 080827KU.S. EPA, 1984, 029711KU.S. EPA, 1978, 040586V Results published after the 2006 O3
20 AQCD continue to support this finding, and the following sections deal with the quantitative
21 characterizations of the relationship, and what new insights may have appeared since 2006.
22 Detrimental effects on plants include visible injury, decreases in the rate of photosynthesis, reduced
23 growth, and reduced yield of marketable plant parts. Most published exposure-response data have
24 been reported O3 effects on the yield of crops and the growth of tree seedlings, and those two
25 variables have been the focus of the characterization of ecological impacts of O3 for the purpose of
26 setting secondary air quality standards. In order to support quantitative modeling of exposure-
27 response relationships, data should preferably include more than three levels of exposure, and some
28 control of potential confounding or interacting factors should be present in order to model the
29 relationship with sufficient accuracy. Letting potential confounders, such as other stressors, vary
30 freely when generating O3 exposure-response data might improve the 'realism' of the data, but it also
31 greatly increases the amount of data necessary to extract a clear quantitative description of the
32 relationship. Conversely however, experimental settings should not be so exhaustively restrictive as
33 to make generalization outside of them problematic. During the last four decades, many of the
34 studies of the effects of O3 on growth and yield of plants have not included enough levels of O3 to
March 2011 9-111 DRAFT-DO NOT CITE OR QUOTE
-------
1 parameterize more than the simplest linear model. The majority of these studies have only contrasted
2 two levels, ambient and elevated, or sometimes three by adding carbon filtration in OTC studies,
3 with little or no consideration of quantitatively relating specific values of exposure to specific values
4 of growth or yield. This is not to say that studies that did not include more than two or three levels of
5 O3 exposure, or studies that were conducted in uncontrolled environments, do not provide exposure-
6 response information that is highly relevant to reviewing air quality standards. In fact, they can be
7 essential in verifying the agreement between predictions obtained through the empirical models
8 derived from experiments such as NCLAN, and observations. The consensus of model predictions
9 and observations from a variety of studies conducted in other locations, at other times, and using
10 different exposure methods, greatly increases confidence in the reliability of both. Furthermore, if
11 they are considered in the aggregate, studies with few levels of exposure or high unaccounted
12 variability can provide additional independent estimates of decrements in plant growth and yield, at
13 least within a few broad categories of exposure.
14 Extensive exposure-response information on a wide variety of plant species has been produced
15 by two long-term projects that were designed with the explicit aim of obtaining quantitative
16 characterizations of the response of such an assortment of crop plants and tree seedlings to O3 under
17 North American conditions: the NCLAN project for crops, and the EPA National Health and
18 Environmental Effects Research Laboratory, Western Ecology Division tree seedling project
19 (NHEERL/WED). The NCLAN project was initiated by the EPA in 1980 primarily to improve
20 estimates of yield loss under field conditions and to estimate the magnitude of crop losses caused by
21 O3 throughout the U.S. (Heck et al., 1982, Q39525)(Heck et al, 1991, 042621). The cultural
22 conditions used in the NCLAN studies approximated typical agronomic practices, and the primary
23 objectives were: (1) to define relationships between yields of major agricultural crops and O3
24 exposure as required to provide data necessary for economic assessments and development of O3
25 NAAQS; (2) to assess the national economic consequences resulting from O3 exposure of major
26 agricultural crops; and (3) to advance understanding of cause-and-effect relationships that determine
27 crop responses to pollutant exposures.
28 NCLAN experiments yielded 54 exposure-response curves for 12 crop species, some of which
29 were represented by multiple cultivars at several of 6 locations throughout the U.S. The
30 NHEERL/WED project was initiated by EPA in 1988 with the same objectives for tree species, and
31 yielded 49 exposure-responses curves for multiple genotypes of 11 tree species grown for up to three
32 years in Oregon, Michigan, and the Great Smoky Mountain National Park. Both projects used OTCs
33 to expose plants to three to five levels of O3. Eight of the 54 crop datasets were from plants grown
34 under a combination of O3 exposure and experimental drought conditions. Figure 9-13 through 9-16
35 summarize some of the NCLAN and NHEERL/WED results.
36 It should be noted that data from FACE experiments might also be used for modeling
37 exposure-response. They only use two levels of O3 (ambient concentration at the site and a multiple
38 of it), but given that the value of both levels of exposure changes every year, and that they are
39 typically run for many consecutive years, aggregating data over time produces twice as many levels
March 2011 9-112 DRAFT-DO NOT CITE OR QUOTE
-------
1 of O3 as there are years. As described in Section 9.3.4, FACE experiments seek to impose fewer
2 constraints on the growth environment then OTCs. As a consequence, FACE studies have to contend
3 with larger variability, especially year-to-year variability, but the difference in experimental
4 conditions between the two methodologies makes comparisons between their results especially
5 useful.
6 Growth and yield of at least one crop (soybean) has been investigated in yearly experiments
7 since 2001 at a FACE facility in Illinois (Morgan et al, 2006, 079186)(University of Illinois, 2010,
8 670286). however almost all analyses of SoyFACE published so far have been based on subsets of
9 one or two years, and have only contrasted ambient versus elevated O3 as categorical variables. They
10 have not modeled the response of growth and yield to O3 exposure continuously over the range of
11 exposure values that have occurred over time. The only exception is a study by Betzelberger et al.
12 (2010, 644183). who used a linear regression model on data pooled over 2 years. Likewise, trees of
13 three species (trembling aspen, paper birch, and sugar maple) were grown between 1998 and 2009 in
14 a FACE experiment located in Rhinelander, Wisconsin (Dickson et al., 2000, 628220)(Pregitzer et
15 al., 2008, 191677). The Aspen FACE experiment has provided extensive data on responses of trees
16 beyond the seedling stage under long-term exposure, and also on ecosystem-level responses (Section
17 9.6), but the only attempt to use those data in a continuous model of the response of tree growth to
18 O3 exposure (Percy et al., 2007, 093287) suffered severe methodological problems, some of which
19 are discussed in Section 9.8.3. Finally, one experiment was able to exploit a naturally occurring
20 gradient of O3 concentrations to fit a linear regression model to the growth of cottonwood (Gregg et
21 al., 2003, 046996; Gregg et al., 2006, 186961). Factors such as genotype, soil type and soil moisture
22 were under experimental control, and the authors were able to partition out the effects of potential
23 confounders such as temperature, atmospheric N deposition, and ambient CO2.
24 A serious difficulty in assessing results of exposure-response research is the multiplicity of O3
25 metrics that have been used in reporting. As described in Section 9.7, metrics that entail either
26 weighting or thresholding of hourly values cannot be converted into one another, or into unweighted
27 metrics such as hourly average. When using weighted or thresholded metrics, which include W126,
28 AOTx or SUMx metrics; O3 exposure at every exposure-response data point must be computed
29 separately for each metric, starting with the hourly data. Comparisons of exposure-response models
30 can only be made between studies that used the same metric, and the value of exposure at which a
31 given plant response is expected on one scale of exposure cannot be exactly converted to another
32 scale. Determining the exposure value at which an effect would be observed in a different metric can
33 only be accomplished by first computing the experimental exposures in this metric from the hourly
34 data, then estimating (fitting) model coefficients again. This problem is irremediable, although useful
35 comparisons might be made using categorical exposures such as 'current ambient exposure' or '2050
36 projected exposure', which can serve as a common reference for quantitative values expressed in
37 various metrics. Studies that contained growth or yield exposure-response data at few levels of
38 exposure, and/or using metrics other than W126 are summarized in Tables 9-16 and 9-17.
39
March 2011 9-113 DRAFT-DO NOT CITE OR QUOTE
-------
9.8.2. 1996 and 2006 Ozone AQCDs Estimates Of Crop Yield Loss And
Tree Seedling Biomass Loss
1 The 1996 and 2006 O3 AQCDs relied extensively on analyses of NCLAN and NHEERL/WED
2 by Lee et al. (1987, 042135: 1988, 042136: 1989, 042137: 1994, 043268). Hogsett et al. (1997,
3 040402). Lee and Hogsett (1999, 040451). Heck et al. (1984, 039380). Rawlings and Cure (1985,
4 039419). Lesser et al. (1990, 043015). and Gumpertz and Rawlings (1992, 043259). Those analyses
5 concluded that a three-parameter Weibull model -
Y -ae n '
6 J l>t t^ Equation 9-2D
6 is the most appropriate model for the response of absolute yield and growth to O3 exposure, because
7 of the interpretability of its parameters, its flexibility (given the small number of parameters), and its
8 tractability for estimation. In addition, removing the intercept a results in a model of relative yield
9 (yield relative to [yield at exposure=0]) without any further reparameterization. Formulating the
10 model in terms of relative yield or relative yield loss (yield loss=[l - relative yield]) is essential in
1 1 comparing exposure-response across species, genotypes, or experiments for which absolute values of
12 the response may vary greatly. In the 1996 and 2006 O3 AQCDs, the two-parameter model of relative
13 yield was used in deriving common models for multiple species, multiple genotypes within species,
14 and multiple locations.
15 Given the disparate species, genotypes, and locations that were included in the NCLAN and
16 NHEERL/WED projects, and in the absence of plausible distributional assumptions with respect to
17 those variables, a three step process using robust methods was used to obtain parameter estimates
18 that could be generalized. The models that were derived for each species or group of species were
19 referred to as median composite functions. In the first step, the three parameters of the Weibull
20 model were estimated (fitted) for absolute yield or biomass data from each NCLAN and
21 NHEERL/WED experiment (54 crop datasets and 49 tree seedling datasets), using nonlinear
22 regression. When data were only available for three levels of exposure because of experimental
23 problems, the shape parameter (3 was constrained to 1, reducing the model to an exponential decay
24 model. In the second step, a was dropped, and predicted values of relative yield or biomass were
25 then computed for 12-hr W126 exposures between 0 and 60 ppm-h. At each of these W126 exposure
26 values, the 25th, 50th, and 75th percentiles of the response were identified among the predicted
27 curves of relative response. For example, for the 34 NCLAN studies of 12 crop species grown under
28 non-droughted conditions for a complete cropping cycle (Figure 9-13), the 3 quartiles of the
29 response were identified at every integer value of W126 between 0 and 60. The third step fitted a
30 two-parameter Weibull model to those percentiles, yielding the median composite function for the
31 relative yield or biomass response to O3 exposure for each grouping of interest (e.g., all crops, all
32 trees, all datasets for one species), as well as composite functions for the other quartiles. In the 1996
33 and 2006 O3 AQCDs, this modeling of crop yield loss and tree seedling biomass loss was conducted
March 2011 9-114 DRAFT-DO NOT CITE OR QUOTE
-------
1 using the SUM06 metric for exposure. This section updates those results by using the 12-hr W126 as
2 proposed in 2007 (72 FR 37818 (2007, 684055)) and 2010 (75 FR 2938 (2010, 684211). p. 3003).
3 Figures 9-13 through 9-16 present quantiles of predicted relative yield or biomass loss at seven
4 values of the 12-h W126 for some representative groupings of NCLAN and NHEERL/WED results.
5 Tables 9-8 through 9-10 give the 90-day 12-h W126 O3 exposure values at which 10 and 20% yield
6 or biomass losses are predicted in 50 and 75% of crop or tree species using the composite functions.
-------
100 -
90 -
80 -
3? 70 -
| 60 -
I 50 -
1 40 -
£ 30 -
20 -
10 -
0 -
C
100 •
90 •
80 •
3? 70 •
| 60 -
1 50 •
1 40 -
1 30 -
20 •
10 •
n
11 Soybean datasets
?75'»Pctile
SCPPctile
25-.p=,i,e
* ~ ' '
10 20 30 40 50 60
5 Cotton datasets
T r-L| 75thPctile
I — SCPPctile
T r~| LJ 25»Pctile
I U T
4, i T 1 1 !
100 -
90 -
80 -
70 -
60 -
50 -
40 -
30 -
20 -
10 -
7 Wheat datasets rn 75lhP=tile
I
I r
f
r^i
D
rn
T
1
i-U — 50thPctile
— LJ 25thPctile
^
- T
r
1
100 -
90 -
80 -
70 -
60 -
50 -
40 -
30 -
20 -
10 -
0 -
2 Com datasets
20 30 40 50
12hrW126 (ppm-hr
10 20 30 40 50
12hrW126 (ppm-hr)
Source of Weibull parameters: Lee and Hogsett (1996, 6702781.
Figure 9-14. Quantiles of predicted relative yield loss for 4 crop species in NCLAN experiments.D
Quantiles of the predicted relative yield loss at 7 values of 12-h W126 for Weibull
curves estimated using nonlinear regression for 4 species grown under well-
watered conditions for the full duration of 1 cropping cycle. The number of studies
available for each species is indicated on each plot.
March 2011
9-116
DRAFT - DO NOT CITE OR QUOTE
-------
100 -
90 •
80 -
ra" 70 "
in
° 60 -
in
in
1 50 -
o
2 40 -
c
-------
100
90
80
-^
~ 70
1 60
1 50
o
IS 40
s
y 30
°-
20
10
100 •
14Aspendatasets 90 •
i80 •
r-L] 75th Pctile
70 •
T 60 •
— 50th Pctile 50 .
!-
40 •
— L-7-1 25th Pctile
J T 30'
T
20-
1 10-
1
11 Ponderosapine datasets
p^l 75th Pctile
pL, — 50" Pctile
r^l |— | LJ 25" Pctile
E-3 \ T-
a T
0 10 20 30 40 50 60 0 10 20 30 40 50 60
100 -
90 -
80 -
f 70 "
5 60-
(fl
1 50 -
| 40 -
1 30 -
Q-
20 -
10 -
100 •
7Douglasf rdatasets 90 •
80 •
70 •
60 •
50 •
40 •
30 •
T 20-
I
r^-i rb B 50thpctile
3 Tulip poplardatasets
50* Pctile
0 10 20 30 40 50 60 0 10 20 30 40 50 60
90 day 12 hr W126 (ppm-hr)
90 day 12 hr W126 (ppm-hr)
Source of Weibull parameters: Lee and Hogsett (1996, 6702781.
Figure 9-16. Quantiles of predicted relative biomass loss for 4 tree species in NHEERL/WED
experiments. [Quantiles of the predicted relative above-ground biomass loss at 7
exposure values of 12-h W126 for Weibull curves estimated using nonlinear
regression on data for 4 tree species grown under well-watered conditions for 1 or 2
year. Curves were standardized to 90-day W126. The number of studies available for
each species is indicated on each plot.
March 2011
9-118
DRAFT - DO NOT CITE OR QUOTE
-------
Table 9-8. Ozone exposures at which 10 and 20% yield loss is predicted for 50 and 75% of crop
species, based on composite functions for the 50th and 75th percentiles of 34 Weibull
curves for relative yield loss data from 34 non-droughted NCLAN studies of 12 crop species;
curves were standardized to 90-day W126
90-day 12-h W126 for 10% yield loss (ppm-h) 90-day 12-h W126 for 20% yield loss (ppm-h)
Model for the 50th Percentile of 34 curves
Relative yield=exp(-(W126/104.82)"1.424) 22 37
Model for the 75th Percentile of 34 curves
Relative yield=exp(-(W126/78.12)"1.415) 16 27
Source of parameters for the 34 curves: Lee and Hogsett (1996, 6702781
Table 9-9. Ozone exposures at which 10 and 20% yield loss is predicted for 50 and 75% of crop
species under drought conditions and adequate moisture, based on composite functions
for the 50th and 75th percentiles of 16 Weibull curves for relative yield loss data from 8
NCLAN studies that paired draughted and watered conditions forthe same genotype;
curves were standardized to 90-day W126
90 day 12-h W12S for 10% yield loss 90 day 12-h W126 for 20% yield loss
(ppm-h) (ppm-h)
Model for the 60th Percentile of 2*8 curves
Watered
Draughted
Relative yield=exp(-(W1 26/1 32.86)**1 . 1 70)
Relative yield=exp(-(W126/179.84)**1. 713)
19
48
37
75
Model for the 76th Percentile of 2*8 curves
Watered
Draughted
Relative yield=exp(-(W1 26/90.43)**!. 310)
Relative yield=exp(-(W126/105.16)**1. 833)
16
31
29
46
Source of parameters forthe 16 curves: Lee and Hogsett (1996, 670278)
March 2011 9-119 DRAFT-DO NOT CITE OR QUOTE
-------
Table 9-10. Ozone exposures at which 10 and 20% biomass loss is predicted for 50 and 75 % of tree
species, based on composite functions for the 50th and 75th percentiles of 49 Weibull
curves for relative above-ground biomass loss data from 49 studies of 11 tree species
grown under well-watered conditions for 1 or 2 year; curves were standardized to 90-day
W126
90 day 12 h W126 for 10% yield loss90 day 12 h W126 for 20% yield loss
(ppm-h) (ppm-h)
Model for the 60th Percentile of 49 curves
Relative yield=exp(-(W126/131.57)**1.242) 21 39
Model for the 76th Percentile of 49 curves
Relative yield=exp(-(W126/65.49)**1.500) 15 24
Source of parameters for the 49 curves: Lee and Hogsett (1996, 6702781
9.8.3. Validation of 1996 and 2006 Ozone AQCD Models and Methodology
Using the 90 day 12-h W126 and Current FACE Data.
1 Since the completion of the NCLAN and NHEERL/WED projects, almost no studies have
2 been published that could provide a basis for estimates of exposure-response that can be compared to
3 those of the 1996 and 2006 O3 AQCDs. Most experiments, regardless of exposure methodology,
4 include only two levels of exposure. In addition, very few studies have included measurements of
5 exposure using the W126 metric, or the hourly O3 concentration data that would allow computing
6 exposure using the W126. Two FACE projects, however, were conducted over multiple years, and by
7 adding to the number of exposure levels over time, may support independent model estimation and
8 prediction using the same model and the same robust process as summarized in Section 9.8.2.
9 Hourly O3 data were available from both FACE projects.
10 The SoyFACE project is situated near Champaign, IL, and comprises 32 octagonal rings (20m-
11 diameter), 4 of which in a given year are exposed to ambient conditions, and 4 of which are exposed
12 to elevated O3 as a fixed proportion of the instantaneous ambient concentration (Betzelberger et al,
13 2010, 644183: Morgan et al., 2004, 072764: Morgan et al., 2006, 079186'XUniversitv of Illinois,
14 2010, 670286). Since 2002, yield data have been collected for up to 8 genotypes of soybean grown
15 in subplots within each ring. The Aspen FACE project is situated in Rhinelander, WI, and comprises
16 12 rings (30m-diameter), 3 of which are exposed to ambient conditions, and 3 of which are exposed
17 to O3 as a fixed proportion of the instantaneous ambient concentration (Dickson et al., 2000, 628220:
18 Karnosky et al., 2005, 095556: Pregitzer et al., 2008, 191677). In the summer of 1997, half the area
19 of each ring was planted with small (five to seven leaf sized) clonally propagated plants of five
20 genotypes of trembling aspen, which were left to grow in those environments until 2009. Biomass
21 data are currently available for the years 1997-2005 (King et al., 2005, 191701). Ozone exposure in
22 these two FACE projects can be viewed as a categorical variable with two levels: ambient, and
23 elevated. However, this overlooks the facts that yearly ambient and elevated exposure both vary with
24 every year, and that the proportionality between them also changes. This change has two sources:
March 2011 9-120 DRAFT - DO NOT CITE OR QUOTE
-------
1 first, the dispensing of O3 into the elevated exposure rings varies from the proportionality set point to
2 some extent, and for SoyFACE, the set point changed between years. Second, the proportionality
3 does not propagate predictably from the hourly data to the yearly value when using thresholded or
4 concentration-weighted cumulative metrics (such as AOT40, SUM06 or W126). Hourly average
5 elevated exposures that are, for example, 1.5 times ambient do not result in AOT40, SUM06 or
6 W126 values that are some constant multiple of the ambient values of those indices. The greater the
7 fraction of elevated hourly values that are above the threshold or heavily weighted, compared to the
8 fraction of hourly ambient values that are, the greater the difference between ambient and elevated
9 yearly exposure, as measured using weighted cumulative indices. When elevated exposure is a
10 multiple of ambient hourly intervals, the number of hours for which elevated exposure meets the
11 threshold for inclusion can vary widely, even though the hourly mean for the year retains the
12 proportionality. As a consequence, the number of exposure levels in multi-year experiments is twice
13 the number of years. In the case of SoyFACE for the period between 2002 and 2008, ambient
14 exposure in the highest year was approximately equal to elevated exposure in the lowest year, with
15 14 levels of O3 exposure evenly distributed from lowest to highest. The particular conditions of the
16 Aspen FACE experiment resulted in 12 exposure levels between 1998 and 2003, but they were not as
17 evenly distributed between minimum and maximum over the 6-year period.
18 There are necessary differences in the modeling of exposure-response in annual plants such as
19 soybean, and in perennial plants such as aspen trees, when exposure takes place over multiple years.
20 In annual plants, responses recorded at the end of the life cycle, i.e., yearly, are analyzed in
21 relationship to that year's exposure. Yield of soybeans is affected by exposure during the year the
22 crop was growing, and a new crop is planted every year. Thus an exposure-response relationship can
23 be modeled from yearly responses matched to yearly exposures, with those exposure-response data
24 points having been generated in separate years. For perennial organisms, which are not harvested
25 yearly and continue to grow from year to year, such pairing of exposure and response cannot be done
26 without accounting for time. Not only does the size of the organism at the beginning of each year of
27 exposure increase, but size is also dependent on the exposure from previous years. Therefore the
28 relationship of response and exposure must be analyzed either one year at a time, or by standardizing
29 the response as a yearly increment relative to size at the beginning of each year. Furthermore, the
30 relevant measurement of exposure is cumulative, or cumulative yearly average exposure, starting in
31 the year exposure was initiated, up to the end of the year of interest. When analyzing the growth of
32 trees over several years, it would be evidently incorrect to pair the exposure level in every discrete
33 year with absolute size of the trees that year, and posit a direct relationship between them. In the
34 Aspen FACE experiment, for example, one could not establish an exposure-response relationship by
35 matching 12 yearly exposures and 12 yearly tree sizes, as if size did not also depend on time. This is
36 the basis of the 2007 study of Aspen FACE data by Percy et al. (2007, 093287), and that study was
37 therefore not informative.
March 2011 9-121 DRAFT - DO NOT CITE OR QUOTE
-------
9.8.3.1. Comparison ofNCLAN-Based Prediction and SoyFACE Data.
1 For this ISA, EPA conducted a comparison between yield as predicted by the composite
2 function three-step process (Section 9.8.2) using NCLAN data for soybean yield as observed in
3 SoyFACE. The median composite function for relative yield was derived for the 11 NCLAN soybean
4 Weibull functions for non-droughted studies, and several comparisons between prediction and
5 SoyFACE observations were conducted as follows.
6 For the years 2007 and 2008, SoyFACE yield data were available for 7 and 6 genotypes,
7 respectively. The EPA used those data to compare the change in relative yield observed in a given
8 year between ambient O3 and elevated O3 in SoyFACE, versus the change in relative yield predicted
9 by the NCLAN-based median composite function between those same two values of O3 exposure.
10 The two parameter median composite function for relative yield of soybean was used to predict yield
11 response at the two observed values of exposure in each year, and the change between yield under
12 ambient and elevated was compared to the change observed in SoyFACE for the relevant year (Table
13 9-11). This approach results in a direct comparison of predicted versus observed change in yield.
14 Because the value of relative response between any two values of O3 exposure is independent of the
15 intercept a, this comparison does not require prediction of the absolute values of the responses.
16 Since comparisons of absolute values might be of interest, the predictive functions were also
17 scaled to the observed data, using two distinct methods. In the first method, the intercept a was
18 calculated algebraically by entering the observed W126 value at ambient exposure and the
19 corresponding value of the response into the three parameter model with the shape and scale
20 parameters (|3 and r\) set to their value for the NCLAN predictive model. This method provides a
21 comparison between the response observed under elevated exposure, and the response that would
22 have been predicted with only the knowledge of what the response was under ambient exposure
23 (Table 9-12; Method 1). In the second method, the intercept for the NCLAN predictive model was
24 estimated by regression using both ambient and elevated data. This method gives a comparison of
25 prediction and observation that takes all the observed information into account to provide the best
26 possible estimate of the intercept, and thus the best possible scaling (Table 9-12, Method 2 and
27 Figure 9-17). It should be noted that the similarity to each other of the predictions obtained by these
28 two scaling methods is a reflection of the accuracy of the predictions: the distance between the
29 predictions from the two methods increases as the distance between prediction and observation
30 increases.
31 For the comparison of NCLAN and SoyFACE, this validation was possible for 2007 and 2008,
32 where data for 7 and 6 soybean genotypes, respectively, were available. The median composite
33 function for relative yield was derived for the 11 NCLAN soybean Weibull functions for
34 nondroughted studies, and the values of median yield under ambient exposure at SoyFACE in 2007
35 and 2008 were used to obtain an estimate of the intercept a for the NCLAN median function in each
36 of the two years
37 Table 9-11 presents the results of ambient/elevated relative yield comparisons between the
38 NCLAN-derived predictions and SoyFACE observations. Table 9-12 presents the results of
March 2011 9-122 DRAFT - DO NOT CITE OR QUOTE
-------
1 comparisons between NCLAN-derived predictions and SoyFACE observations of yield, using two
2 methods for scaling the predictive function. Figure 9-17 presents yield observed in two years in
3 SoyFACE, and predicted by the median composite function derived from NCLAN data using the
4 second scaling method as described, with the intercept estimated using 14 observations in 2007, and
5 12 in 2008.
6 Finally, data were also available for one additional genotype from 2003 to 2007. By
7 aggregating data for each genotype over the years (5 years for one genotype, 2 years for the 6
8 others), a composite function for the 25th, 50th, and 75th percentiles was then developed for
9 SoyFACE, and compared to the corresponding NCLAN-based function. NCLAN functions were
10 obtained using 12-hr W126 standardized to 90 days. SoyFACE 12-hr W126 was cumulated over the
11 highest 90 days of the growing season.
12
Table 9-11. Comparison between relative yield observed in the SoyFACE experiment, and relative yield
predicted at the same values of ozone by the median composite function for NCLAN (two-
parameter relative yield model)
Year
2007
2008
90-day 12-hWI 26 (ppm-h)
Ambient
4.39
3.23
Elevated
46.23
28.79
Elevated
Observed in SoyFACE
0.76
0.88
Yield,
Relative to Am blent
Predicted by NCLAN
0.75
0.85
Table 9-12. Comparison between yield observed in the SoyFACE experiment and yield predicted at the
same values of ozone by the median composite function for NCLAN (three-parameter
absolute yield model), using two scaling methods to calculate the intercept
Year 90-day 12-h W126 (ppm-h)
Yield observed in Yield observed in Yield predicted by Yield predicted by Yield predicted by
Ambient Elevated SoyFACE ambient SoyFACE elevated NCLAN in elevated, NCLAN in ambient, NCLAN in elevated,
(g/m2) (g/m2) Method 1 (g/m2) Method 2 (g/m2) Method 2 (g/m2)
2007 4.39 46.23 305.2 230.6 227.6 309.2 230.6
~2008 3~232879 3448 3044 293~53503 298^2
March 2011 9-123 DRAFT - DO NOT CITE OR QUOTE
-------
2007, 7 genotypes
2008, 6 genotypes
10 20 30 40 50
90day12hrW126 (ppm-hr)
20 30 40 50
90 day 12 hr W126 (ppm-hr)
Note: Black
composite
Source of data: Betzelberger et al. (2010, 644183);Morgan et al. (2006, 0791861: Lee and Hoasett (1996, 670278V
dots are median of 7 or 6 soybean genotypes in SoyFACE (2007, 2008); bars are IQR for genotypes; dashed line is median
model for 11 studies in NCLAN.
Figure 9-17. Comparison of yield observed in SoyFACE experiment in a given year with yield
predicted by the median composite function based on NCLAN.
1 Finally, a composite function for the 25th, 50th, and 75th percentiles was developed from
2 SoyFACE annual yield data by pooling one genotype from 2003 to 2007, and six genotypes in 2007
3 and 2008. NCLAN functions were obtained using 12-h W126 standardized to 90 days. SoyFACE
4 12-h W126 was cumulated over the highest 90 days of the growing season. The correlation between
5 W126 cumulated for the entire season in SoyFACE and W126 for the highest 90 days was greater
6 than 0.99 in all years. The same process was used for SoyFACE: first, the three parameter Weibull
7 model described in Section 9.8.2 was estimated using nonlinear regression on exposure-yield data
8 for each genotype separately, over the years for which data were available, totaling seven curves.
9 The 25th, 50th, and 75th percentiles of the predicted values for the two parameter relative yield
10 curves were then identified at every integer of W126 between 0 and 60, and a two-parameter Weibull
11 model estimated by regression for the three quartiles. The comparison between these composite
12 functions for the quartiles of relative yield loss in SoyFACE and the corresponding composite
13 functions for NCLAN is presented in Figure 9-18.
March 2011
9-124
DRAFT - DO NOT CITE OR QUOTE
-------
100
90 -
80 -
g 70 -
| 60 H
1 50 -
-------
1 four clonally propagated genotypes. All plants were grown in OTCs for one growing season before
2 being destructively harvested. Aspen FACE data were from clonally propagated trees of five
3 genotypes grown from 1998 to 2003, with above-ground biomass calculated using allometric
4 equations derived from data for trees harvested destructively in 2000 and 2002 (King et al, 2005,
5 191701V
6 The two parameter median composite function for relative biomass was used to predict
7 biomass response under the observed elevated exposure, relative to its value under observed ambient
8 exposure. EPA first tested the accuracy of the prediction of biomass at elevated exposure relative to
9 biomass at ambient exposure, for each separate year of Aspen FACE. Comparisons between
10 observed and predicted biomass values were then conducted for each year by scaling the predictive
11 function to yearly Aspen FACE data using the two scaling methods described in Section 9.8.3.1.
12 Yearly 90 day 12-hour W126 values for Aspen FACE were computed as the cumulative average
13 from the year of planting up to the year of interest. A comparison of composite functions between
14 NHEERL/WED and Aspen FACE, similar to the one performed for NCLAN and SoyFACE, was not
15 possible: as discussed in the introduction to Section 9.8, the pairing of 12 exposure values from
16 separate years and 12 values of biomass cannot be the basis for a model of exposure-response,
17 because the trees continued growing for the six-year period of exposure. Table 9-13 presents the
18 results of ambient/elevated relative biomass comparisons between the NHEERL/WED-derived
19 predictions and Aspen FACE observations. Table 9-14 presents the results of comparisons between
20 NHEERL/WED-derived predictions and Aspen FACE observations of biomass, using two methods
21 for scaling the predictive function. Figure 9-19 presents biomass observed in six years at Aspen
22 FACE, and predicted by the median composite function derived from NHEERL/WED data using the
23 second scaling method as described, with the intercept estimated using 2 observations in each year.
Table 9-13. Comparison between above-ground biomass observed under elevated ozone in Aspen
FACE experiment in 6 year, relative to above-ground biomass observed under ambient
ozone and relative above-ground biomass above-ground biomass at the same values of
ozone predicted by the median composite function for NHEERL/WED (two-parameter
relative biomass model)
Year
1998
1999
2000
2001
2002
2003
90-day 12-hWI 26 (ppm-h)
Cumulative Average
Ambient
3.19
2.61
2.43
2.55
2.51
2.86
Elevated
30.08
33.85
30.16
31.00
30.27
29.12
Above-Ground Biomass,
Elevated Relative To Ambient
Observed in Aspen FACE
0.75
0.70
0.71
0.71
0.69
0.71
Predicted by NHEERL/WED
0.74
0.70
0.74
0.73
0.74
0.75
March 2011
9-126
DRAFT - DO NOT CITE OR QUOTE
-------
Table 9-14. Comparison between above-ground biomass observed in Aspen FACE experiment in
6 year and biomass predicted by the median composite function based on NHEERL/WED
(three-parameter absolute biomass model), using 2 scaling methods to calculate the
intercept
Year
90day12-hW126(ppm-h)
Cumulative Average
Biomass Observed in Aspen
FACE (g/m2)
5i?-m-'pSls(S2Si8tod by Biomass Predicted by NHEERL/WED,
™ 2
--pl(
Ko™
Method 2 (g/m2)
Ambient
Elevated
Ambient
Elevated
Elevated
Ambient
Elevated
1998
3.19
30.08
274.7
204.9
202.3
276.0
203.2
1999
2.61
33.85
955.3
673.3
665.9
958.7
668.3
2000
2.43
30.16
1400.3
998.6
1036.0
1382.4
1022.8
2001
2.55
31.00
1620.7
1154.9
1183.7
1607.0
1173.7
2002
2.51
30.27
2125.9
1468.41
1566.7
2079.0
1532.1
2003
2.86
29.12
2695.2
1907.8
2022.5
2640.1
1981.2
3UUU -
2500 •
_ 2000 •
(M
s
§ 1500 •
ro
E
0
m 1000 -
500 •
n .
^
^
_ ^
""-- "^1
" -. ^ • 2003
_ ** ^ ^
_ """"--^^ | 2002
^ ^ •» ^ ^
""""""--**$ 2001
_ ^ "" "J 2000
""""""-* 1999
-f 1998
10 20 30 40 50 60
90 day 12 hr W126 (yearly cumulative average, ppm-hr)
70
Source of data: Kingetal. (2005, 191701V Lee and Hogsett (1996, 670278).
Note: Black dots are aspen biomass/m for 3 FACE rings filled with an assemblage of 5 clonal genotypes of aspen at Aspen FACE; bars
are SE for 3 rings; dashed line is median composite model for 4 clonal genotypes and wild-type seedlings in 11 NHEERL/WED 1-year
OTC studies.
Figure 9-19. Comparison between above-ground biomass observed in Aspen FACE experiment
in 6 year and biomass predicted by the median composite function based on
NHEERL/WED.
March 2011
9-127
DRAFT - DO NOT CITE OR QUOTE
-------
1 As in the comparisons between NCLAN and SoyFACE, the agreement between predictions
2 based on NHEERL/WED data and Aspen FACE observations was exceptionally close. The results of
3 the two projects strongly reinforce each other with respect to the response of aspen biomass to O3
4 exposure. The methodology used for obtaining the median composite function is shown to be
5 capable of deriving a predictive model despite potential confounders, and despite the added
6 measurement error that is expected from calculating biomass using allometric equations. In addition,
7 the function based on one year of growth was shown to be applicable to subsequent years.
8 The results of experiments that used different exposure methodologies, different genotypes,
9 locations, and durations converged to the same values of response to O3 exposure for each of two
10 very dissimilar plant species, and predictions based on the earlier experiments were validated by the
11 data from current ones. However, in these comparisons, the process used in establishing predictive
12 functions involved aggregating data over variables such as time, locations, and genotypes, and the
13 use of a robust statistic (quartiles) for that aggregation. The validating data, from SoyFACE and
14 Aspen FACE, were in turn aggregated over the same variables. The accuracy of predictions is not
15 expected to be conserved for individual values of those variables over which aggregation occurred.
16 For example, the predicted values for soybean, based on data for five genotypes, are not expected to
17 be valid for each genotype separately. As shown in the validation, however, aggregation that
18 occurred over different values of the same variable did not affect accuracy: composite functions
19 based on one set of genotypes were predictive for another set, as long as medians were used for both
20 sets. A study of cottonwood (Populus deltoides) conducted using a naturally occurring gradient of O3
21 exposure (Gregg et al., 2003, 046996; Gregg et al., 2006, 186961) may provide an illustration of the
22 response of an individual species whose response is far from the median response for an aggregation
23 of species.
9.8.3.3. Exposure-Response in a Gradient Study
24 Gregg et al. (2003, 046996) grew saplings of one clonally propagated genotype of cottonwood
25 (Populus deltoides} in seven locations within New York City and in the surrounding region between
26 July and September in 1992, 1993 and 1994, and harvested them 72 days after planting. Owing to
27 regional gradients of atmospheric O3 concentration, the experiment yielded eight levels of exposure
28 (Figure 9-20), and the authors were able to rule out environmental variables other than O3 to account
29 for the large differences in biomass observed after one season of growth. The deficit in growth
30 increased substantially faster with increasing O3 exposure than has been observed in aspen, another
31 species of the same genus (Populus tremuloides, Section 9.8.3.2). Using a three parameter Weibull
32 model (Figure 9-20), the biomass of cottonwood at a W126 exposure of 15 ppm-h, relative to
33 biomass at 5 ppm-h, is estimated to be 0.18 (18% of growth at 5 ppm-h). The relative biomass of
34 trembling aspen within the same 5-15 ppm-h range of exposure is estimated to be 0.92, using the
35 median composite model for aspen whose very close agreement with Aspen FACE data was shown
36 in Section 9.8.3.2. Using a median composite function for all deciduous trees in the NHEERL/WED
37 project (6 species in 21 studies) also gives predictions that are very distant from the cottonwood
March 2011 9-128 DRAFT - DO NOT CITE OR QUOTE
-------
1 response observed in this experiment. For all deciduous tree species in NHEERL/WED, biomass at a
2 W126 exposure of 15 ppm-h, relative to biomass at 5 ppm-h, was estimated to be 0.87.
100 -
90 -
80
70 -
g 60 -
(A
| 50 -
o
in 40 -
30 -
20 -
10 -
0
10 20 30 40 50
72 day 12 hr W126 (ppm-hr)
60
70
Source: Modified with permission from Nature Publishing Group, Gregg et al. (2003, 0469961.
Figure 9-20. Above-ground biomass for one genotype of cottonwood grown in seven locations
for one season in 3 years. mine represents the three-parameter Weibull model.
3 These cottonwood data confirms that, as should be expected, some individual tree species are
4 substantially more sensitive than the median of NHEERL/WED (Figure 9-15). As shown in
5 Section 9.8.2, the median models available for trembling aspen and soybean have verifiable
6 predictive ability for those particular species. This suggests that the corresponding NCLAN- and
7 NHEERL/WED-based models for multiple crop and tree species can provide reliable estimates of
8 losses for similar assortments of species. However, their predictive ability would likely be poor for
9 individual species not tested.
10 An alternative hypothesis for the difference between the response of cottonwood in this
11 experiment and deciduous tree species in NHEERL/WED, or the difference between the response of
12 cottonwood and aspen in NHEERL/WED and Aspen FACE, could be the presence of confounding
13 factors in the environments where the experiment was conducted. However, variability in
14 temperature, moisture, soil fertility, and atmospheric deposition of N were all ruled out by Gregg
15 et al. (2003, 046996) as contributing to the observed response to O3. In addition, this hypothesis
16 would imply that the unrecognized confounder(s) were either absent from both OTC and FACE
17 studies, or had the same value in both. This is not impossible, but the hypothesis that cottonwood is
18 very sensitive to O3 exposure is more parsimonious, and sufficient.
March 2011
9-129
DRAFT - DO NOT CITE OR QUOTE
-------
9.8.3.4. Meta-analyses of growth and yield studies
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Since the 2006 O3 AQCD, five studies have used meta-analytic methods to integrate results
from experimental studies of crops or tree species relevant to the U.S. It is possible to obtain
exposure-response data for growth and yield from those meta-analyses, but because all of them
provided summary measurements of O3 exposure as hourly averages of various lengths of exposures,
comparisons with exposure-response results where exposure is expressed as W126 are problematic.
Table 9-15 summarizes the characteristics of the five meta-analyses. They all included studies
conducted in the U.S. and other locations worldwide, and all of them expressed responses as
comparative change between levels of exposure to O3, with carbon filtered air (CF) among those
levels. Using hourly average concentration to summarize exposure, CF rarely equates absence of O3,
although it almost always near zero when exposure is summarized as W126, SUM06, or AOT40.
Table 9-15. Meta-analyses of growth or yield studies published since 2005
Study
Ainsworth (2008,
1916461
Feng et al. (2008,
1914531
Feng and Kobayashi
(2009, 199223)
Grantzetal. (2006,
1915451
Wittig et al. (2009,
1916311
Number of articles
included
12
53
All crops together : 81
16
All responses:263
Articles that included
biomass:unreported
Years of
publication
surveyed
1980-2007
1980-2007
1980-2007
1992-2004
1970-2006
Crop, species or genera
rice
wheat
Potato, barley, wheat, rice,
bean, soybean
34 herbaceous dicots
21 herbaceous monocots
5 tree species
4 gymnosperm tree genera
11 angiosperm tree genera
Response
Yield
Yield
Yield
Relative
Growth Rate
Total biomass
Number of
03 levels
2
5
3
2
4
Duration of
exposure
unreported
> 10 days
> 1 0 days
2-24 weeks
> 7 days
The only effect of O3 exposure on yield of rice reported in Ainsworth (2008, 191646) was a
decrease of 14% with exposure increasing from CF to 62 ppb average concentration. Feng et al.
(2008, 191453) were able to separate exposure of wheat into four classes with average
concentrations of 42, 69, 97, and 153 ppb, in data where O3 was the only treatment. Mean responses
relative to CF were yield decreases of 17, 25, 49, and 61% respectively. Feng et al. (2008, 191453)
observed that wheat yield losses were smaller under conditions of drought, and that Spring wheat
and Winter wheat appeared similarly affected. However, mean exposure in studies of Winter wheat
was substantially higher than in studies of Spring wheat (86 versus 64 ppb), which suggests that the
yield of Spring wheat was in fact more severely affected, since yield was approximately the same,
even though Spring wheat was exposed to lower concentrations. Exposures of the six crops
considered in Feng and Kobayashi (2009, 199223) were classified into two ranges, each compared to
CF air. In the lower range of exposure (41-49 ppb), potato studies had the highest average exposure
(45 ppb), and wheat and rice the lowest (41 ppb). In the higher range (51-75 ppb), wheat studies had
the highest average exposure (65 ppb), and potato, barley and rice the lowest (63 ppb). In other
words, across the studies included, all crops were exposed to very similar levels of O3. At
March 2011
9-130
DRAFT - DO NOT CITE OR QUOTE
-------
1 approximately 42 ppb, the yield of potato, barley, wheat, rice, bean, and soybean declined by 5.3,
2 8.9, 9.7, 17.5, 19, and 7.7% respectively, relative to CF air. At approximately 64 ppb O3, declines
3 were 11.9, 12.5, 21.1, 37.5, 41.4, and 21.6%. Grantz et al. (2006, 191545) reported Relative Growth
4 Rate (RGR) rather than growth, and did not report O3 exposure values in a way that would allow
5 calculation of mean exposure for each of the three categories of plants for which RGR changes are
6 reported. All studies used only two levels of exposure, with CF air as the lower one, and most used
7 elevated exposure in the range of 40 to 70 ppb. Decline in RGR was 8.2% for the 34 herbaceous
8 dicots, 4.5% for the 21 herbaceous monocots, and 17.9 for the 5 tree species. Finally, Wittig et al.
9 (2009, 191631) divided the studies analyzed into three classes of comparisons: CF versus ambient,
10 CF versus elevated, and ambient versus elevated, but reported comparisons between three average
11 levels of exposure besides CF: 40 ppb, 64 ppb, and 97 ppb. Corresponding decreases in total biomass
12 relative to CF were 7, 17, and 17%.
13 These meta-analyses provide very strong confirmation of EPA's conclusions from previous O3
14 AQCDs: compared to lower levels of ambient O3, current levels in many locations are having a
15 substantial detrimental effect on the growth and yield of a wide variety of crops and natural
16 vegetation. They also confirm strongly that decreases in growth and yield continue at exposure levels
17 higher than current ambient levels. However, direct comparisons with the predictions of exposure-
18 response models that use concentration-weighted cumulative metrics are difficult.
9.8.3.5. Additional expos ure-res pons e data
19 The studies summarized in Tables 9-16 and 9-17 contain growth or yield exposure-response
20 data at too few levels of exposure for exposure-response models, and/or used metrics other than
21 W126. These tables update Tables AX9-16 through AX9-19 of the 2006 O3 AQCD.
March 2011 9-131 DRAFT - DO NOT CITE OR QUOTE
-------
Table 9-1 6. Summary of studies of effects of ozone exposure on growth and yield of agricultural crops
Species
Facility
Location
Alfalfa (Medicago
sativa)
OTC; 0.27m3 pots
Federico, Italy
Bean (Phaseolus
vulgaris I. cv Borlotto)
OTC; ground-planted
Curno, Italy
Big Blue Stem
(Andropogon gerardii)
OTC
Alabama
Brassica napus cv.
Westar
Growth chambers
Finland
Corn (lea mays cv.
Chambord)
OTC
France
Cotton cv. Pima
OTC; 9-L pots
France
Eastern Gamagrass
(Tripsacum
dactyloides)
OTC
Alabama
Grapevine (Vitis
vinivera)
OTC
Austria
Mustard (Brassica
campestris)
Chambers;
7.5-cm pots
Oilseed Rape
(Brassica napus)
OTC
Yangtze Delta, China
Peanut (Arachis
hypogaea)
OTC
Raleigh, NC
Exposure
Duration
2 yr, 2005,
2006
3 months,
2006
4 months,
2003
17-26 days
33 days
8wk
4 months,
2003
3 yr, May-Oct
10 days
39 days
Syr
03 Exposure
(Additional Treatment)
AOT40: CF 0 ppm-h
1 3.9 ppm-h (2005), 10.1 ppm-h
(2006)
(NaCI: 0.29, 0.65, 0.83,
1.06deciSiemens/meter)
Seasonal AOT40:
CF (0.5 ppm-h);
ambient (4.6 ppm-h)
(N/A)
12-havg:
CF(14ppb),
Ambient (29 ppb),
Elevated (71 ppb)
(N/A)
8-h avg:
CF(Oppb), 100 ppb
(Bt/non-Bt; herbivory)
AOT40 ppm-h: 1.1; 1.3; 4.9; 7.2;
9.3; 12.8
(N/A)
12-havg: 12.8 + 0.6; 79.9 + 6.3;
122.7 + 9.7
(N/A)
12-havg:
CF(14ppb),
Ambient (29 ppb),
Elevated (71 ppb)
(N/A)
AOT40 ppm-h:
CF (0),
Ambient (7-20),
Elevated. 1 (20-30), Elevated. 2
(38-48)
CF&
67. 8 ppb for 7 h
(N/A)
Daily avg: 1 00 ppb, one with
diurnal variation and one with
constant concentration
(N/A)
12-havg:
CF (22 ppb),
Ambient (46 ppb),
Elevated (75ppb)
(C02: 375 ppm; 548 ppm;
730 ppm)
Response Measured
Total shoot yield
# Seeds per plant;
100-seed weight
Final harvest biomass;
RVF
Shoot biomass
Total above-ground
biomass
Above-ground biomass
Final harvest biomass;
RVF
Total fruit yield/
Sugar yield
Seeds/plant
Biomass and pods per
plant
Yield (seed weight, g/m)
percent change from
CF
(percent change from
ambient)
n.s. (N/A)
-33 (N/A)
n.s. (N/A)
n.s. (n.s.)
-7 (-7)
-30.70 (N/A)
N/A (Highest treatment
caused -26% change)
-76 (n.s.)
+68 (+42);
-17 (-12)
-20 to -80 in different yr
(-20 to -90 in different yr)
n.s. (N/A)
Diurnal variability
reduced both biomass
and pod number more
than constant fumigation
(N/A)
-33 (-8)
Reference
Maggio etal.
(2009, 1916451
Gerosa et al.
(2009, 1914031
Lewis et al.
(2006, 1915421
Himanen etal.
(2009, 1913381
Leitao et al.
(2007, 1914561
Grantzand
Shrestha (2006,
1917021
Lewis et al.
(2006, 1915421
Soja et al. (2004,
0943971
Black etal.
(2007, 1915581
Wang et al.
(2008, 1912761
Burkeyetal.
(2007, 1913711
March 2011
9-132
DRAFT - DO NOT CITE OR QUOTE
-------
Species
Facility
Location
Poa pratensis
OTC
Braunschweig,
Germany
Potato (Solanum
tuberosum)
OTC; CHIP
6 northern European
locations
Rice (Oryza sativa)
OTC
Raleigh, NC
Rice (Oryza sativa) 20
Asian cultivars
OTC
Gunma Prefecture,
Japan
Seminatural grass
FACE
Le Mouret, Switzerland
Soybean
OTC; CRA
Bari, Italy
Soybean (G/yc/ne max
cv. 93B15)
SoyFACE
Urbana, IL
Soybean (G/yc/ne max
cv. Essex)
Chambers; 21 L
Raleigh, NC
Soybean (G/yc/ne max
cv. Essex)
OTCs;21-Lpots
Raleigh, NC
Soybean (G/yc/ne max
cv. Tracaja)
Chambers; pots
Brazil
Soybean (G/yc/ne
max) 10 cultivars
SoyFACE
Urbana, IL
Exposure
Duration
2000-2002:
4-5 wk in the
Spring
1988,1999.
Emergence to
harvest
1997-1998,
June-
September
2008 growing
season
Syr
2003-2005
growing
seasons
2002, 2003
growing
seasons
2x3 months
2x3 months
20 days
2007 & 2008
03 Exposure
(Additional Treatment)
8-h avg:
CF+25(21.7),
NF+50J73.1)
(Competition)
AOT40:CF (0);
Ambient (0.27-5. 19); NF (0.002-
2.93)
NF+ (3.10-24.78
(N/A)
12-h mean ppb:
CF (27.5),
Elevated (74.8)
(C0a
Daily avg (ppb):
CF (2),
O.Sxambient (23);
1 xambient (28);
1.5xambient(42);
2xambient (57)
(Cultivar comparisons)
Seasonal AOT40: Ambient
(0.1-7.2ppm-h);
Elevated. (1.8-24.1 ppm-h)
(N/A)
Seasonal AOT40 ppm-h: CF (0),
Ambient (3.4), High (9.0)
(Drought)
8-h avg:
Ambient (62 & 50 ppb), Elevated
(75 & 63 ppb)
(N/A)
12-h avg: CF (28),
Elevated (79),
Elevated flux (11 2)
(C02: 365 & 700)
12-havg:CF(18);
Elevated (72)
(C02:367&718)
12-h avg: CF&30 ppb
(N/A)
8-h avg: Ambient (46.3 & 37.9),
Elevated (82.5 & 61. 3)
(Cultivar comparisons)
Response Measured
Total biomass (g
DW/pot)
Tuber yield averaged
across 5 field-sites;
Tuber starch content
regressed against [03]
report sig. + slope with
increasing [03]
Total biomass;
Seed yield
Yield
Relative annual yield
Yield
Yield
Seed mass per plant
Seed mass per plant
Biomass
Yield
percent change from
CF
(percent change from
ambient)
N/A(n.s.)
N/A (-27 % -+27%, most
comparisons n.s.) Linear
regression slope =
-0.0098)
-25(N/A)
-13 to 20 (N/A)
From n.s. to -30 across
all cultivars
N/A (2xfaster
decrease in yield/yr)
-46 (-9)
N/A
(-15 in 2002;
-25 in 2003)
-30 (N/A)
-34 (N/A)
-18 (N/A)
N/A (-17.20)
Reference
Bender etal.
(2006, 1914371
Vandermeiren
et al. (2005,
1799921
Reid, etal.
(2008, 1915611
Sawada and
Kohno (2009,
1994261
Volk et al. (2006,
1914341
Jaoudeetal.
(2008, 1912231
Morgan etal.
(2006, 0791861
Booker and
Fiscus (2005,
1916521
Booker et al.
(2004, 0791381
Bulbovas et al.
(2007, 1994111
Betzelberger
etal. (2010,
6441831
March 2011
9-133
DRAFT - DO NOT CITE OR QUOTE
-------
Species
Facility
Location
Spring Wheat (Triticum
aestivumcv. Minaret;
Satu; Drabant; Dragon)
OTCs
Belgium, Finland, &
Sweden
Strawberry (Fragaria x
ananassa Duch. Cv
Korona & Elsanta)
Growth chambers
Bonn, Germany
Sugarbeet (Beta
vulgaris cv. Patriot)
OTC
Belgium
Sugarcane
(Saccharum spp)
CSTR
San Joaquin Valley, CA
Sweet Potato
Growth chambers
Bonn, Germany
Tomato (Lycopersicon
esculentum)
OTC
Valencia, Spain
Trifolium
Subterraneum
OTC; 2.5-L pots
Madrid, Spain
Watermelon (Citrullus
lanatus)
OTC
Valencia, Spain
Yellow Nutsedge
OTC; 9-L pots
Exposure
Duration
1990-2006
2 months
2003, 2004;
5 months
2007;
11-13wk.
4wk
133 days in
1998
29 days
2000,2001.
90 days
8wk
03 Exposure
(Additional Treatment)
Seasonal AOT40s ranged from 0
to16ppm-h
(N/A)
8-h avg: CF (0 ppb) &
Elevated (78 ppb)
(N/A)
8-h avg: Ambient (36 ppb);
Elevated (62 ppb)
(N/A)
12-havg:CF(4ppb);
Ambient (58);
Elevated (147)
(N/A)
8-h avg: CF (0 ppb),
Ambient (<40 ppb) Elevated (255
ppb)
(N/A)
8-h mean ppb:
CF 16.3, NF 30.1,
NF+ 83.2
(Various cultivars; early & late
harvest)
12-h avg: CF (<7.9±6.3); Ambient
(34.4+10.8);
Elevated (56.4+22.3)
(N:5, 15 & 30 kg/ha)
AOT40: CF (0 ppm-h)
Ambient (5.7 ppm-h), Elevated
(34.1 ppm-h)
(N:0, 1 4.0 & 29.6 g/pot)
12-h avg: 12.8 + 0.6; 79.9 + 6.3;
122.7 + 9.7
(N/A)
Response Measured
Seed protein content;
1,000-seed weight
regressed across all
experiments
Fruit yield (weight/plant)
Sugar yield
Total biomass (g/plant)
Tuberweight
Yield
Above-ground biomass
total fruit yield (kg)
above-ground biomass
percent change from
CF
(percent change from
ambient)
N/A (significant negative
correlation)
N/A (sig negative
correlation)
-16 (N/A)
N/A (-9)
-40 (-30)
-14 (-11. 5)
n.sfn.s.)
-45 (-35)
n.s. (54)
n.s. (n.s.)
Reference
Piikkietal.
(2008, 1998121
Keutgen et al.
(2005, 1912951
DeTemmerman
et al. (2007,
1913611
Grantzand Vu
(2009, 1952371
Keutgen et al.
(2008, 1916901
Calvo et al.
(2005, 1915701
Sanz et al. (2005,
1969631
Calatayud et al.
(2006, 1914821
Grantzand
Shrestha (2006,
191702)
In studies where variables other than 03were included in the experimental design, response to 03 is only provided for the control level of those variables.
March 2011
9-134
DRAFT - DO NOT CITE OR QUOTE
-------
Table 9-17. Summary of studies of effects of ozone exposure on growth of natural vegetation
Species
Facility
Location
Yellow nutsedge (Cyperus
esculentus)
CSTR
Parlier, CA
35 herbaceous species
OTC
Corvallis, OR
Highbush blackberry (Rubus
argutus)
OTC
Auburn, AL
Horseweed (Conyza
canadensis)
CSTR
San Joaquin Valley, CA
Red Oak (Quercus rubrum)
Forest sites
Look Rock & Twin Creeks
Fnrpctc TM
rUlcbLb, I IN
Pine species
Forest sites
Look Rock Forest, TN
Hickory species
Forest sites
Look Rock Forest, TN
Chestnut Oak (Quercus prinus)
Forest sites
Look Rock Forest, TN
Exposure
Duration
UUI dUUM
53 days in 2008
1999-2002,
May-August
2004,
May-August
2005, 2 runs,
28 days each
(July-Aug, Sept)
2001-2003,
April-October
2001-2003,
April-October
2001-2003,
April-October
2001-2003,
April-October
03 Exposure
(Additional
Treatment)
12-h mean ppb: CF
(4); CF+ (60);
CF2+(115)
4-yravg; yearly
W126ppm-h:
CF (0),
CF+(21),
CF 2+ (49.5)
12-h mean ppb:
CF(21.7),
Ambient (32.3),
Elevated (73.3)
W126ppm-hr:
CF(0),
CF+(11),
CF 2+ (30)
(Glyphosate
resistance)
AOT60:
2001 (11.5),
2002 24.0),
2003 11.7)
(Observational
study with multiple
environmental
variables)
AOT60:
2001 (11.5),
2002 24.0),
2003 11.7)
(Observational
study with multiple
environmental
variables)
AOT60:
2001 (11.5),
2002 24.0),
9nrn 11 7^
ZUUJ I I . 1 1
(Observational
study with multiple
environmental
variables)
AOT60:
2001 (11.5),
2002 24.0),
9nrn 11 7^
ZUUJ I I . 1 1
(Observational
study with multiple
environmental
variables)
Response Measured
Above-ground biomass;
tubers (g/plant)
Total community above-
ground biomass
(35 species) after 4 years
Vegetative regrowth after
pruning
Total biomass (g/plant)
Annual circumference
increment (change
relative to 2001 in year
2002;2003)
Annual circumference
increment (change
relative to 2001 in year
2002;2003)
Annual circumference
increment (change
relative to 2001 in year
2002;2003)
Annual circumference
increment (change
relative to 2001 in year
2002;2003)
Response
ns;CF(4.1)CF+(3.9)
CF2+(2.7)
CF (459 g/m2), CF+
(457 g/m2), CF2+ (398
g/m2)
CF(75.1 g/plant),
Ambient (76.4 g/plant),
Elevated (73.1 g/plant)
Glyphosate sensitive:
CF (0.354)
CF+(0.197)
CF2+(0.106)
Glyphosate resistant:
CF(0.510)
CF+(0.313)
CF2+(0.143)
-42.8%; +1%
-62.5%; -2.9%
-14%; +30%
+44%; +55%
Reference
Grantzetal. (2010,
1021611
Pfleegeretal. (2010,
6442811
Ditchkoffetal. (2009,
1922301
Grantzetal. ((2008,
1913121
Mclaughlin etal. (2007,
0903481
Mclaughlin etal. (2007,
0903481
Mclaughlin etal. (2007,
0903481
Mclaughlin etal. (2007,
0903481
March 2011
9-135
DRAFT - DO NOT CITE OR QUOTE
-------
Species
Facility
Location
Black Cherry (Prunus rigida)
Forest sites
Twin Creeks Forest, TN
Shortleaf pine (Pinus echinata)
Forest sites
Twin Creeks Forest, TN
Hemlock (Tsuga canadensis)
horest sites
Twin Creeks Forest, TN
Red Maple (Acer rubrum)
Forest sites
Twin Creeks Forest, TN
Yellow Poplar (Liriodendron
tulipifera)
Forest sites
Look Rock, Oak Ridge, & Twin
Creeks Forest, TN
Sugar Maple (Acer saccharum)
Forest sites
Twin Creeks Forest, TN
Trembling aspen (Populus
tremuloides), 5 genotypes
Aspen FACE
Rhinelander, Wl
Hybrid Poplar (Populus
trichocarpa x Populus deltoides)
OTC
Seattle, WA
Exposure
Duration
UUI dUUM
2002-2003,
April-October
2002-2003,
April-October
2002-2003,
April-October
2002-2003,
April-October
2002-2003,
April-October
2002-2003,
April-October
1998-2004,
May-October
2003, 3 months
03 Exposure
(Additional
Treatment)
AOT60:
2002 (24.0),
2003(117)
(Observational
study with multiple
environmental
variables)
AOT60:
2002 (24.0),
2003(117)
(Observational
study with multiple
environmental
variables)
AOT60:
2002 (24.0),
2003(117)
(Observational
study with multiple
environmental
variables)
AOT60:
2002 (24.0),
2003(117)
(Observational
study with multiple
environmental
variables)
AOT60:
2002 24.0),
2003 11.7)
(Observational
study with multiple
environmental
variables)
AOT60:
2002 (24.0),
2003(117)
(Observational
study with multiple
environmental
variables)
Cumulative avg
90-day 12-hW1 26.
Ambient 3.1 ppm-h
Elevated: 27.2
ppm-h
(Competition with
birch, maple)
Daily mean (ug/g):
CF(<9),
Elevated (85-1 28)
Response Measured Response
Annual circumference
increment (change 7r0/
relative to 2003 in year ~'0/0
2002)
Annual circumference
increment (change 1Rfio/
relative to 2003 in year •1b'H/0
2002)
Annual circumference
increment (change -, qo/
relative to 2003 in year -^i.»/°
2002)
Annual circumference
increment (change rQ R0/
relative to 2003 in year "3a'D/0
2002)
Annual circumference
increment (change KWI- mow
relative to 2001 in years -
-------
References
A list of all references considered for inclusion in this chapter can be found at
http://hero.epa. gov/index.cfm?action=search.do&submit=Search&group_id=476
Agrell, I; Kopper, B. I; McDonald, E. P.; Lindroth, R. L. (2005). CO2 and O3 effects on host plant preferences of the
forest tent caterpillar (Malacosoma disstria). Global Change Biol, 11: 588-599. http://dx.doi.org/10.lll l/j.1365-
2486.2005.00924.x 074324
Ahlfors, R.; Brosche, M.; Kollist, H.; Kangasjarvi, J. (2009). Nitric oxide modulates ozone-induced cell death, hormone
biosynthesis and gene expression in Arabidopsis thaliana. Plant J, 58: 1-12. http://dx.doi.org/10.1111/j.1365-
313X.2008.03756.x 191533
Ahsan, N.; Nanjo, Y; Sawada, H.; Kohno, Y; Komatsu, S. (2010). Ozone stress-induced proteomic changes in leaf total
soluble and chloroplast proteins of soybean reveal that carbon allocation is involved in adaptation in the early
developmental stage. Proteomics, 10: 2605-2619. http://dx.doi.org/10.1002/pmic.201000180 644189
Ainsworth, E. A. (2008). Rice production in a changing climate: a meta-analysis of responses to elevated carbon dioxide
and elevated ozone concentration. Global Change Biol, 14: 1642-1650. http://dx.doi.org/10.1111/j.1365-
2486.2008.01594.x 191646
Ainsworth, E. A.; Long, S. P. (2005). What have we learned from 15 years of free-air CO2 enrichment (FACE)? A meta-
analytic review of the responses of photosynthesis, canopy properties and plant production to rising CO2. New
Phytol, 165: 351-371.042647
Ainsworth, E. A.; Rogers, A. (2007). The response of photosynthesis and stomatal conductance to rising [CO2]:
mechanisms and environmental interactions. Plant Cell Environ, 30: 258-270. 092940
Air quality criteria and control techniques, Section 108 of the Clean Air Act. 42 USC A§ 7408. (1990).
http://www.epa.gov/air/caa/titlel.html#ia. 080701
Alexis, A.; Garcia, A.; Nystrom, M.; Rosenkranz, K. (2001). The 2001 California almanac of emissions and air quality.
Retrieved August 10, 2005 from http://www.arb.ca.gov/aqd/almanac/almanac01/almanac01.htm. 079886
Allen, E. B.; Temple, P. J.; Bytnerowicz, A.; Arbaugh, M. J.; Sirulnik, A. G; Rao, L. E. (2007). Patterns of understory
diversity in mixed coniferous forests of southern California impacted by air pollution. ScientificWorldJournal, 7:
247-263. http://dx.doi.org/10.1100/tsw.2007.72 196876
Alonso, R.; Bermejo, V; Sanz, J.; Vails, B.; Elvira, S.; Gimeno, B. S. (2007). Stomatal conductance of semi-natural
Mediterranean grasslands: Implications for the development of ozone critical levels. Environ Pollut, 146: 692-698.
http://dx.doi.0rg/10.1016/i.envpol.2006.06.009 199289
Amthor, J. S. (1988). Growth and maintenance respiration in leaves of bean (Phaseolus vulgaris L) exposed to ozone in
open-top chambers in the field. New Phytol, 110: 319-325. 041870
Andersen, C. P. (2003). Source-sink balance and carbon allocation below ground in plants exposed to ozone. New Phytol,
157:213-228.041673
Andersen, C. P.; Ritter, W.; Gregg, J.; Matyssek, R.; Grams, T E. E. (2010). Below-ground carbon allocation in mature
beech and spruce trees following long-term, experimentally enhanced O3 exposure in Southern Germany. Environ
Pollut, 158: 2604-2609. http://dx.doi.Org/10.1016/j.envpol.2010.05.008 628559
Andersen, C. P.; Wilson, R.; Plocher, M.; Hogsett, W. E. (1997). Carry-over effects of ozone on root growth and
carbohydrate concentrations of ponderosa pine seedlings. Tree Physiol, 17: 805-811. 052923
Andrews, K. M.; Gibbons, J. W; Jochimsen, D. M. (2008). Ecological effects of roads on amphibians and reptiles: A
literature review. In JC Mitchell; RE Jung Brown; B Bartholomew (Eds.), Urban Herpetology (pp. 121-143). Salt
Lake City: Society for the Study of Amphibians and Reptiles. http://www.uga.edu/srel/Reprint/3091.htm 645771
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments such as the Integrated Science Assessments (ISAs) and the Integrated Risk Information System (IRIS).
March 2011 9-137 DRAFT - DO NOT CITE OR QUOTE
-------
Aneja, M. K.; Sharma, S.; Fleischmann, R; Stich, S.; Heller, W.; Bahnweg, G; Munch, J. C.; Schloter, M. (2007). Influence
of ozone on litter quality and its subsequent effects on the initial structure of colonizing microbial communities.
MicrobEcol, 54: 151-160. http://dx.doi.org/10.1007/s00248-006-9183-0 191472
Arbaugh, M. J.; Miller, P. R.; Carroll, J. J.; Takemoto, B. L.; Proctor, T. (1998). Relationships of ozone exposure to pine
injury in the Sierra Nevada and San Bernardino Mountains of California, USA. Environ Pollut, 101: 291-301.
http://dx.doi.org/10.1016/S0269-7491(98)00027-X 040297
Arbaugh, M.; Bytnerowicz, A.; Grulke, N.; Fenn, M.; Poth, M.; Temple, P.; Miller, P. (2003). Photochemical smog effects
in mixed conifer forests along a natural gradient of ozone and nitrogen deposition in the San Bernardino Mountains.
Environ Int, 29: 401-406. 052925
Ariyaphanphitak, W.; Chidthaisong, A.; Sarobol, E.; Bashkin, V. N.; Towprayoon, S. (2005). Effects of elevated ozone
concentrations on Thai Jasmine rice cultivars (Oryza sativa L.). Water Air Soil Pollut, 167: 179-200.
http://dx.doi.org/10.1007/sll270-005-8650-4 191349
Ashmore MEmberson, L.; Karlsson, P. E.; Pleijel, H. (2004). Introduction for ozone deposition special issue. Atmos
Environ, 38: 2211-2212. 056624
Ashmore, M. R. (2002). Effects of oxidants at the whole plant and community level. In Air Pollution and Plants (pp. 89-
118). London: Wiley. 672967
Ashmore, M. R.; Bell, J. N. B.; Mimmack, A. (1988). Crop growth along a gradient of ambient air pollution. Environ
Pollut, 53: 99-121. 037038
Ashmore, M.; Emberson, L.; Karlsson, P. E.; Pleijel, H. (2004). New directions: a new generation of ozone critical levels
for the protection of vegetation in Europe (correspondence). Atmos Environ, 38: 2213-2214. 056623
Ashmore, M.; Toet, S.; Emberson, L. (2006). Ozone - a significant threat to future world food production? New Phytol,
170: 201-204. http://dx.doi.Org/10.llll/i.1469-8137.2006.01709.x 191557
Awmack, C. S.; Harrington, R.; Lindroth, R. L. (2004). Aphid individual performance may not predict population
responses to elevated CO2 or O3. Global Change Biol, Biol.10: 1414-1423. 052926
Awmack, C. S.; Mondor, E. B.; Lindroth, R. L. (2007). Forest understory clover populations in enriched CO2 and O-3
atmospheres: Interspecific, intraspecific, and indirect effects. Environ Exp Bot, 59: 340-346.
http://dx.doi.0rg/10.1016/i.envexpbot.2006.04.003 191415
Bagard, M.; Le Thiec, D.; Delacote, E.; Hasenfratz-Sauder, M. P.; Banvoy, J.; Gerard, J.; Dizengremel, P.; Jolivet, Y.
(2008). Ozone-induced changes in photosynthesis and photorespiration of hybrid poplar in relation to the
developmental stage of the leaves. Physiol Plant, 134: 559-574. http://dx.doi.Org/10.llll/j.1399-3054.2008.01160.x
191593
Baier, M.; Kandlbinder, A.; Golldack, D.; Dietz, K. (2005). Oxidative stress and ozone: Perception; signalling and
response. Plant Cell Environ, 28: 1012-1020. http://dx.doi.Org/10.llll/i.1365-3040.2005.01326.x 186866
Ball, G. R.; Palmer-Brown, D.; Fuhrer, J.; Skarby, L.; Gimeno, B. S.; Mills, G. (2000). Identification of non-linear
influences on the seasonal ozone dose-response of sensitive and resistant clover clones using artificial neural
networks. Ecol Modell, 129: 153-168. 026354
Balls, G. R.; Palmer-Brown, D.; Sanders, G. E. (1996). Investigating microclimatic influences on ozone injury in clover
(Trifolium subterraneum) using artificial neural networks. New Phytol, 132: 271-280. 026370
Bandeff, J. M.; Pregitzer, K. S.; Loya, W. M.; Holmes, W. E.; Zak, D. R. (2006). Overstory community composition and
elevated atmospheric CO2 and O-3 modify understory biomass production and nitrogen acquisition. Plant Soil,
282: 251-259. http://dx.doi.org/10.1007/slll04-005-5930-0 191733
Barnes, J.; Zheng, Y; Lyons, T. (2002). Plant resistance to ozone: The role of ascorbate. In Omasa, K.; Saji, H.; Youssefian,
S.; Kondo, N. (Eds.), Air pollution and plant biotechnology - Prospects for phytomonitoring and phytoremediation
(p. 235-252). Tokyo: Springer-Verlag. 040313
Bassin, S.; Volk, M.; Fuhrer, J. (2007). Factors affecting the ozone sensitivity of temperate European grasslands: An
overview. Environ Pollut, 146: 678-691. http://dx.doi.Org/10.1016/j.envpol.2006.06.010 196879
Bassin, S.; Volk, M.; Suter, M.; Buchmann, N.; Fuhrer, J. (2007). Nitrogen deposition but not ozone affects productivity
and community composition of subalpine grassland after 3 yr of treatment. New Phytol, 175: 523-534.
http://dx.doi.0rg/10.llll/i.1469-8137.2007.02140.xl91534
March 2011 9-138 DRAFT - DO NOT CITE OR QUOTE
-------
Bassin, S.; Werner, R. A.; Sorgel, K.; Volk, M.; Buchmann, N.; Fuhrer, J. (2009). Effects of combined ozone and nitrogen
deposition on the in situ properties of eleven key plant species of a subalpine pasture. Oecologia, 158: 747-756.
http://dx.doi.org/10.1007/s00442-008-1191-vl91333
Bauer, M. R.; Hultman, N. E.; Panek, J. A.; Goldstein, A. H. (2000). Ozone deposition to a ponderosa pine plantation in the
Sierra Nevada Mountains (CA): a comparison of two different climatic years. J Geophys Res, 105: 22,123-22,136.
040315
Bender, J.; Muntifering, R. B.; Lin, J. C.; Weigel, H. J. (2006). Growth and nutritive quality of Poa pratensis as influenced
by ozone and competition. Environ Pollut, 142: 109-115. http://dx.doi.Org/10.1016/j.envpol.2005.09.012 191437
Benoit, L. R; Skelly, J. M.; Moore, L. D.; Dochinger, L. S. (1982). Radial growth reductions of Pinus strobus L correlated
with foliar ozone sensitivity as an indicator of ozone-induced losses in eastern forests. Can J For Res, 12: 673-678.
039778
Bergweiler, C. J.; Manning, W. J. (1999). Inhibition of flowering and reproductive success in spreading dogbane
(Apocynum androsaemifolium) by exposure to ambient ozone. Environ Pollut, 105: 333-339. 036321
Bergweiler, C.; Manning, W. J.; Chevone, B. I. (2008). Seasonal and diurnal gas exchange differences in ozone-sensitive
common milkweed (Asclepias syriaca L.) in relation to ozone uptake. Environ Pollut, 152: 403-415.
http://dx.doi.0rg/10.1016/i.envpol.2007.06.019 191656
Bernacchi, C. J.; Leaky, A. D. B.; Heady, L. E.; Morgan, P. B.; Dohleman, F. G; McGrath, J. M.; Gillespie, K. M.; Wittig,
V. E.; Rogers, A.; Long, S. P.; Ort, D. R. (2006). Hourly and seasonal variation in photosynthesis and stomatal
conductance of soybean grown at future CO2 and ozone concentrations for 3 years under fully open-air field
conditions. Plant Cell Environ, 29: 2077-2090. http://dx.doi.Org/10.llll/j.1365-3040.2006.01581.x 158001
Bernacchi, C. J.; Morgan, P. B.; Ort, D. R.; Long, S. P. (2005). The growth of soybean under free air CO2 enrichment
(FACE) stimulates photosynthesis while decreasing in vivo Rubisco capacity. Planta, 220: 434-446.
http://dx.doi.org/10.1007/s00425-004-1320-8679713
Betzelberger, A. M.; Gillespie, K. M.; McGrath, J. M.; Koester, R. P.; Nelson, R. L.; Ainsworth, E. A. (2010). Effects of
chronic elevated ozone concentration on antioxidant capacity, photosynthesis and seed yield of 10 soybean
cultivars. Plant Cell Environ, 33: 1569-1581. http://dx.doi.Org/10.llll/i.1365-3040.2010.02165.x644183
Bidart-Bouzat, M. G; Imeh-Nathaniel, A. (2008). Global change effects on plant chemical defenses against insect
herbivores. J Integr Plant Biol, 50: 1339-1354. http://dx.doi.Org/10.llll/i.1744-7909.2008.00751.xl91431
Billings, W. D. (1978). Plants and the ecosystem. Belmont, CA: Wadsworth Publishing Company, Inc. 034165
Biswas, D. K.; Xu, H.; Li, Y. G; Sun, J. Z.; Wang, X. Z.; Han, X. G; Jiang, G. M. (2008). Genotypic differences in leaf
biochemical, physiological and growth responses to ozone in 20 winter wheat cultivars released over the past 60
years. Global Change Biol, 14: 46-59. http://dx.doi.Org/10.llll/i.1365-2486.2007.01477.xl91428
Black, V. J.; Black, C. R.; Roberts, J. A.; Stewart, C. A. (2000). Impact of ozone on the reproductive development of plants.
NewPhytol, 147: 421-447. 036322
Black, V. J.; Stewart, C. A.; Roberts, J. A.; Black, C. R. (2007). Ozone affects gas exchange, growth and reproductive
development in Brassica campestris (Wisconsin Fast Plants). New Phytol, 176: 150-163.
http://dx.doi.0rg/10.llll/i.1469-8137.2007.02163.xl91558
Black, V. J.; Stewart, C. A.; Roberts, J. A.; Black, C. R. (2010). Direct effects of ozone on reproductive development in
Plantago major L. populations differing in sensitivity. Environ Exp Bot, 69: 121-128.
http://dx.doi.0rg/10.1016/i.envexpbot.2010.04.006625575
Blande, J. D.; Holopainen, J. K.; Li, T. (2010). Air pollution impedes plant-to-plant communication by volatiles. Ecol Lett,
13: 1172-1181. http://dx.doi.0rg/10.llll/i.1461-0248.2010.01510.x 643928
Bohler, S.; Bagard, M.; Oufir, M.; Planchon, S.; Hoffmann, L.; Jolivet, Y; Hausman, J. F.; Dizengremel, P.; Renaut, J.
(2007). A DIGE analysis of developing poplar leaves subjected to ozone reveals major changes in carbon
metabolism. Proteomics, 7: 1584-1599. http://dx.doi.org/10.1002/pmic.200600822 199408
Booker, F. L.; Burkey, K. O.; Overmyer, K.; Jones, A. M. (2004). Differential responses of G-protein Arabidopsis thaliana
mutants to ozone. NewPhytol, 162: 633-641. 020581
March 2011 9-139 DRAFT - DO NOT CITE OR QUOTE
-------
Booker, F. L.; Burkey, K. O.; Pursley, W. A.; Heagle, A. S. (2007). Elevated carbon dioxide and ozone effects on peanut: I.
Gas-exchange, biomass, and leaf chemistry. Crop Sci, 47: 1475-1487.
http://dx.doi.org/10.2135/cropsci2006.08.0537 191370
Booker, F. L.; Fiscus, E. L. (2005). The role of ozone flux and antioxidants in the suppression of ozone injury by elevated
CO2 in soybean. J Exp Bot, 56: 2139-2151. http://dx.doi.org/10.1093/jxb/eri214 191652
Booker, F. L.; Fiscus, E. L.; Miller, J. E. (2004). Combined effects of elevated atmospheric carbon dioxide and ozone on
soybean whole-plant water use. J Environ Manage, 1: S355-S362. 079138
Booker, F. L.; Prior, S. A.; Torbert, H. A.; Fiscus, E. L.; Pursley, W. A.; Hu, S. (2005). Decomposition of soybean grown
under elevated concentrations of CO2 and O3. Global Change Biol, 11: 685-698. http://dx.doi.Org/10.llll/j.1365-
2486.2005.00939.x 079151
Booker, F. L.; Reid, C. D.; Brunschon-Harti, S.; Fiscus, E. L.; Miller, J. E. (1997). Photosynthesis and photorespiration in
soybean [Glycine max (L) Merr] chronically exposed to elevated carbon dioxide and ozone. J Exp Bot, 48: 1843-
1852. 026425
Booker, F.; Muntifering, R.; McGrath, M.; Burkey, K.; Decoteau, D.; Fiscus, E.; Manning, W.; Krupa, S.; Chappelka, A.;
Grantz, D. (2009). The ozone component of global change: Potential effects on agricultural and horticultural plant
yield, product quality and interactions with invasive species. J Integr Plant Biol, 51: 337-351.
http://dx.doi.0rg/10.llll/i.1744-7909.2008.00805.xl91569
Borowiak, K.; Rucinska-Sobkowiak, R.; Rymer, K.; Gwozdz, E. A.; Zbierska, J. (2009). Biochemical markers of
tropospheric ozone: Experimentation with test-plants. Polish Journal of Ecology, 57: 3-14. 191247
Broadmeadow, M. S. J.; Jackson, S. B. (2000). Growth responses of Quercus petraea, Fraxinus excelsior and Pinus
sylvestris to elevated carbon dioxide, ozone and water supply. New Phytol, 146: 437-451. 021325
Brook, J. R.; DiGiovanni, F.; Cakmak, S.; Meyers, T P. (1997). Estimation of dry deposition velocity using inferential
models and site-specific meteorology—uncertainty due to siting of meteorological towers. Atmos Environ, 31: 3911-
3919.041857
Bulbovas, P.; de Souza, S. R.; de Moraes, R. M.; Luizao, F.; Artaxo, P. (2007). Soybean 'Tracaja' seedlings exposed to
ozone under controlled conditions. Pesquisa Agropecuaria Brasileira, 42: 641-646. http://dx.doi.org/10.1590/S0100-
204X2007000500005 199411
Burkey, K. O.; Booker, F. L.; Pursley, W. A.; Heagle, A. S. (2007). Elevated carbon dioxide and ozone effects on peanut: II.
Seed yield and quality. Crop Sci, 47: 1488-1497. http://dx.doi.org/10.2135/cropsci2006.08.0538 191371
Burkey, K. O.; Eason, G; Fiscus, E. L. (2003). Factors that affect leaf extracellular ascorbic acid content and redox status.
Physiol Plant, 117: 51-57. http://dx.doi.Org/10.1034/i.1399-3054.2003.1170106.x630251
Bytnerowicz, A.; Arbaugh, M.; Schilling, S.; Fraczek, W.; Alexander, D. (2008). Ozone distribution and phytotoxic
potential in mixed conifer forests of the San Bernardino Mountains, Southern California. Environ Pollut, 155: 398-
408. http://dx.doi.0rg/10.1016/j.envpol.2008.01.046 196881
Caird, M. A.; Richards, J. H.; Donovan, L. A. (2007). Nighttime stomatal conductance and transpiration in C-3 and C-4
plants. Plant Physiol, 143: 4-10. http://dx.doi.org/10.1104/pp.106.092940 199337
Calatayud, A.; Alvarado, J. W.; Barreno, E. (2002). Similar effects of ozone on four cultivars of lettuce in open top
chambers during winter. Photosynthetica, 40: 195-200. http://dx.doi.Org/10.1023/A:1021333305592 684222
Calatayud, A.; Pomares, F.; Barreno, E. (2006). Interactions between nitrogen fertilization and ozone in watermelon
cultivar Reina de Corazones in open-top chambers. Effects on chlorophyll alpha fluorescence, lipid peroxidation,
and yield. Photosynthetica, 44: 93-101. http://dx.doi.org/10.1007/sll099-005-0163-2 191482
Calatayud, V.; Cervero, J.; Sanz, M. J. (2007). Foliar, physiologial and growth responses of four maple species exposed to
ozone. Water Air Soil Pollut, 185: 239-254. http://dx.doi.org/10.1007/sll270-007-9446-5 191411
Calatayud, V; Sanz, M. J.; Calvo, E.; Cervero, J.; Ansel, W.; Klumpp, A. (2007). Ozone biomonitoring with Bel-W3
tobacco plants in the city of Valencia (Spain). Water Air Soil Pollut, 183: 283-291.
http://dx.doi.org/10.1007/sll270-007-9376-2 191568
Cal EPA (2009). Air quality data branch main page. Retrieved January 28, 2011 from
http://www.arb.ca.gov/aqd/aqdpage.htm. 677487
March 2011 9-140 DRAFT - DO NOT CITE OR QUOTE
-------
Campbell, S. I; Wanek, R.; Coulston, J. W. (2007). Ozone injury in west coast forests: 6 years of monitoring -
Introduction. Portland, OR: U.S. Department of Agriculture. 602360
Cannon, W. N. (1990). Olfactory response of eastern spruce budworm larvae to red spruce needles exposed to acid rain and
elevated levels of ozone. J Chem Ecol, 16: 3255-3261. 626460
Carde, R. T; Haynes, K. F. (2004). Stucture of the pheromone communication channel in moths. In Advances in insect
chemical ecology (pp. 283-332). Cambridge: Cambridge University Press. 626467
Casteel, C. L.; O'Neill, B. R; Zavala, J. A.; Bilgin, D. D.; Berenbaum, M. R.; DeLucia, E. H. (2008). Transcriptional
profiling reveals elevated CO2 and elevated O-3 alter resistance of soybean (Glycine max) to Japanese beetles
(Popilliajaponica). Plant Cell Environ, 31: 419-434. http://dx.doi.Org/10.llll/i.1365-3040.2008.01782.xl91696
Chapman, J. A.; King, J. S.; Pregitzer, K. S.; Zak, D. R. (2005). Effects of elevated concentrations of atmospheric CO2 and
tropospheric O-3 on decomposition of fine roots. Tree Physiol, 25: 1501-1510. 191345
Chappelka, A. H. (2002). Reproductive development of blackberry (Rubus cuneifolius) as influenced by ozone. New
Phytol, 155: 249-255. 052946
Chappelka, A. H.; Samuelson, L. J. (1998). Ambient ozone effects on forest trees of the eastern United States: a review.
New Phytol, 139: 91-108. 093687
Chappelka, A. H.; Somers, G. L.; Renfro, J. R. (2007). Temporal patterns of foliar ozone symptoms on tall milkweed
(Asclepias exaltata L) in Great Smokey Mountains National Park. Environ Pollut, 149: 358-365.
http://dx.doi.0rg/10.1016/i.envpol.2007.05.015093290
Chappelka, A.; Skelly, J.; Somers, G; Renfro, J.; Hildebrand, E. (1999). Mature black cherry used as a bioindicator of
ozone injury. Water Air Soil Pollut, 116: 261-266. 052952
Chappelka, A.; Somers, G; Renfro, J. (1999). Visible ozone injury on forest trees in Great Smoky Mountains National
Park, USA. Water Air Soil Pollut, 116: 255-260. 041860
Chen, C. W;Tsai,W. T; Lucier, A. A. (1998). A model of air-tree-soil system for ozone impact analysis. Ecol Modell, 111:
207-222. 040317
Chen, Z.; Gallie, D. R. (2005). Increasing tolerance to ozone by elevating foliar ascorbic acid confers greater protection
against ozone than increasing avoidance. Plant Physiol, 138: 1673-1689. http://dx.doi.org/10.1104/pp.105.062000
191465
Chen, Z.; Wang, X. K.; Feng, Z. Z.; Xiao, Q.; Duan, X. N. (2009). Impact of Elevated O-3 on Soil Microbial Community
Function Under Wheat Crop. Water Air Soil Pollut, 198: 189-198. http://dx.doi.org/10.1007/sll270-008-9838-l
191452
Chen, Z.; Wang, X. K.; Yao, F. F.; Zheng, F. X.; Feng, Z. Z. (2010). Elevated ozone changed soil microbial community in a
rice paddy. Soil Sci Soc Am J, 74: 829-837. http://dx.doi.org/10.2136/sssaj2009.0258 644257
Cheng, F. Y; Burkey, K. O.; Robinson, J. M.; Booker, F. L. (2007). Leaf extracellular ascorbate in relation to O-3 tolerance
of two soybean cultivars. Environ Pollut, 150: 355-362. http://dx.doi.Org/10.1016/j.envpol.2007.01.022 191499
Cho, K.; Shibato, J.; Agrawal, G. K.; Jung, Y. H.; Kubo, A.; Jwa, N. S.; Tamogami, S.; Satoh, K.; Kikuchi, S.; Higashi, T;
Kimura, S.; Saji, H.; Tanaka, Y; Iwahashi, H.; Masuo, Y; Rakwal, R. (2008). Integrated transcriptomics,
proteomics, and metabolomics analyses to survey ozone responses in the leaves of rice seedling. J Proteome Res, 7:
2980-2998. http://dx.doi.org/10.1021/pr800128q 603254
Chung, H. G; Zak, D. R.; Lilleskov, E. A. (2006). Fungal community composition and metabolism under elevated CO2 and
O-3. Oecologia, 147: 143-154. http://dx.doi.org/10.1007/s00442-005-0249-3 191729
Colls, J. J.; Unsworth, M. H. (1992). Air pollution interactions with natural stressors. In Barker, J. R.; Tingey, D. T. (Eds.),
Air pollution effects on biodiversity New York, NY: Van Nostrand Reinhold. 026469
Coulston, J. W.; Smith, G. C.; Smith, W. D. (2003). Regional assessment of ozone sensitive tree species using bioindicator
plants. Environ Monit Assess, 83: 113-127. 041871
Crous, K. Y; Vandermeiren, K.; Ceulemans, R. (2006). Physiological responses to cumulative ozone uptake in two white
clover (Trifolium repens L. cv. Regal) clones with different ozone sensitivity. Environ Exp Bot, 58: 169-179.
http://dx.doi.0rg/10.1016/i.envexpbot.2005.07.007 199321
March 2011 9-141 DRAFT - DO NOT CITE OR QUOTE
-------
D'Haese, D.; Horemans, N.; De Coen, W.; Guisez, Y. (2006). Identification of late O-3-responsive genes in Arabidopsis
thalianaby cDNAmicroarray analysis. Physiol Plant, 128: 70-79. http://dx.doi.Org/10.llll/j.1399-
3054.2006.00711.x 191448
D'Haese, D.; Vandermeiren, K.; Asard, H.; Horemans, N. (2005). Other factors than apoplastic ascorbate contribute to the
differential ozone tolerance of two clones of Trifolium repens L. Plant Cell Environ, 28: 623-632.
http://dx.doi.0rg/10.llll/i.1365-3040.2005.01308.xl91551
Dalstein, L.; Vas, N. (2005). Ozone concentrations and ozone-induced symptoms on coastal and alpine mediterranean pines
in southern France. Water Air Soil Pollut, 160: 181-195. 191570
Dammgen, U.; Grunhage, L.; Haenel, H.-D.; Jager, H.-J. (1993). Climate and stress in ecotoxicology A coherent system of
definitions and terms. J Appl Bot Food Qual, 67: 157-162. 055312
Darbah, J. N. T.; Kubiske, M. E.; Neilson, N.; Oksanen, E.; Vaapavuori, E.; Karnosky, D. F. (2007). Impacts of elevated
atmospheric CO2 and O3 on paper birch (Betula papyrifera): reproductive fitness. ScientificWorldJournal, 7: 240-
246. 093288
Darbah, J. N. T.; Kubiske, M. E.; Nelson, N.; Oksanen, E.; Vapaavuori, E.; Kamosky, D. F. (2008). Effects of decadal
exposure to interacting elevated CO2 and/or O-3 on paper birch (Betula papyrifera) reproduction. Environ Pollut,
155: 446-452. http://dx.doi.Org/10.1016/j.envpol.2008.01.033 196890
Davidson, A. (1993). Update of ozone trends in California's South Coast Air Basin. J Air Waste Manag Assoc, 43: 226-227.
043366
Davis, D. D. (2007). Ozone injury to plants within the Seney National Wildlife Refuge in northern Michigan. Northeast
Nat 14:415-424.093291
Davis, D. D. (2007). Ozone-induced symptoms on vegetation within the Moosehorn National Wildlife Refuge in Maine.
Northeast Nat, 14: 403-414. 093292
Davis, D. D. (2009). Ozone-Induced Stipple on Plants in the Cape Remain National Wildlife Refuge, South Carolina.
Southeastern Naturalist, 8: 471-478. 199287
Davis, D. D.; Orendovici, T. (2006). Incidence of ozone symptoms on vegetation within a National Wildlife Refuge in New
Jersey, USA. Environ Pollut, 143: 555-564. http://dx.doi.Org/10.1016/j.envpol.2005.10.051 093293
Dawson, T. E.; Burgess, S. S.; Tu, K. P.; Oliveira, R. S.; Santiago, L. S.; Fisher, J. B.; Simonin, K. A.; Ambrose, A. R.
(2007). Nighttime transpiration in woody plants from contrasting ecosystems. Tree Physiol, 27: 561-575.
http://dx.doi.org/10.1093/treephys/27.4.561 670381
De Temmerman, L.; Legrand, G; Vandermeiren, K. (2007). Effects of ozone on sugar beet grown in open-top chambers.
EurJAgron, 26: 1-9. http://dx.doi.Org/10.1016/j.eja.2006.08.001 191361
DeglTnnocenti, E.; Guidi, L.; Soldatini, G. F. (2007). Effects of elevated ozone on chlorophyll a fluorescence in
symptomatic and asymptomatic leaves of two tomato genotypes. Biol Plantarum, 51: 313-321.
http://dx.doi.org/10.1007/sl0535-007-0061-5 191350
Dermody, O.; O'Neill, B. F.; Zangerl, A. R.; Berenbaum, M. R.; DeLucia, E. H. (2008). Effects of elevated CO2 and O-3 on
leaf damage and insect abundance in a soybean agroecosystem. Arthropod-Plant Interactions, 2: 125-135. 191850
de Lourdes de Bauer, M.; Hernandez-Tejeda, T. (2007). A review of ozone-induced effects on the forests of central Mexico.
Environ Pollut, 147: 446-453. 196891
Dickson, R. E.; Lewin, K. F.; Isebrands, J. G; Coleman, M. D.; Heilman, W. E.; Riemenschneider, D. E.; Sober, J.; Host, G.
E.; Zak, D. R.; Hendrey, G. R.; Pregitzer, K. S.; Karnosky, D. F. (2000). Forest Atmosphere Carbon Transfer and
Storage (FACTS-II) the Aspen Free-Air CO2 and O3 Enrichment (FACE) project: An overview (Report No.
General Technical Report NC-214). St. Paul, MN: U.S. Dept. of Agriculture, Forest Service, North Central
Research Station, http://nrs.fs.fed.us/pubs/278. 628220
Ditchkoff, S. S.; Lewis, J. S.; Lin, J. C.; Muntifering, R. B.; Chappelka, A. H. (2009). Nutritive quality of highbush
blackberry (Rubus argutus) exposed to tropospheric ozone. R E M, 62: 364-370. 192230
Dizengremel, P.; Le Thiec, D.; Bagard, M.; Jolivet, Y. (2008). Ozone risk assessment for plants: Central role of
metabolism-dependent changes in reducing power. Environ Pollut, 156: 11-15.
http://dx.doi.0rg/10.1016/i.envpol.2007.12.024 191587
March 2011 9-142 DRAFT - DO NOT CITE OR QUOTE
-------
Dizengremel, P.; Le Thiec, D.; Hasenfratz-Sauder, M. P.; Vaultier, M. N.; Bagard, M.; Jolivet, Y. (2009). Metabolic-
dependent changes in plant cell redox power after ozone exposure. Plant Biol (Stuttg), 11: 35-42.
http://dx.doi.0rg/10.llll/i.1438-8677.2009.00261.xl99424
Dizengremel, P.; Sasek, T.; Brown, K.; Richardson, C. (1994). Ozone-induced changes in primary carbon metabolism
enzymes of loblolly pine needles. J Plant Physiol, 144: 300-306. 187217
Di Baccio, D.; Castagna, A.; Paoletti, E.; Sebastiani, L.; Ranieri, A. (2008). Could the differences in O(3) sensitivity
between two poplar clones be related to a difference in antioxidant defense and secondary metabolic response to
O(3) influx? Tree Physiol, 28: 1761-1772. 199850
Dobson, H. E. M. (1994). Floral volatiles in insect biology. In Insect-Plant Interactions (pp. 47-82). Boca Raton, FL:
Bernays, E. 626466
Dohm, M. R.; Mautz, W. J.; Andrade, J. A.; Gellert, K. S.; Salas-Ferguson, L. J.; Nicolaisen, N.; Fujie, N. (2005). Effects of
ozone exposure on nonspecific phagocytic capacity of pulmonary macrophages from an amphibian, Bufo marinus.
Environ Toxicol Chem, 24: 205-210. 180452
Dohm, M. R.; Mautz, W. J.; Doratt, R. E.; Stevens, J. R. (2008). Ozone exposure affects feeding and locomotor behavior of
adult bufo marinus. Environ Toxicol Chem, 27: 1209-1216. 604584
Dohm, M. R.; Mautz, W. J.; Looby, P. G; Gellert, K. S.; Andrade, J. A. (2001). Effects of ozone on evaporative water loss
and thermoregulatory behavior of marine toads (Bufo marinus). Environ Res, 86: 274-286. 017086
Dohrmann, A. B.; Tebbe, C. C. (2005). Effect of elevated tropospheric ozone on the structure of bacterial communities
inhabiting the rhizosphere of herbaceous plants native to Germany. Appl Environ Microbiol, 71: 7750-7758.
http://dx.doi.org/10.1128/AEM.71.12.7750-7758.2005 191320
Drogoudi, P. D.; Ashmore, M. (2001). 14C-allocation of flowering and deblossomed strawberry in response to elevated
ozone. NewPhytol, 152: 455-461. 052959
Drogoudi, P. D.; Ashmore, M. R. (2000). Does elevated ozone have differing effects in flowering and deblossomed
strawberry? New Phytol, 147: 561-569. 042517
Dudareva, N.; Negre, F.; Nagegowda, D. A.; Orlova, I. (2006). Plant Volatiles: recent advances and future perspectives.
Crit Rev Plant Sci, 25: 417-440. 626458
Ederli, L.; Morettini, R.; Borgogni, A.; Wasternack, C.; Miersch, O.; Reale, L.; Ferranti, F.; Tosti, N.; Pasqualini, S. (2006).
Interaction between nitric oxide and ethylene in the induction of alternative oxidase in ozone-treated tobacco plants.
Plant Physiol, 142: 595-608. http://dx.doi.org/10.1104/pp.106.085472 191479
Ellsworth, D. S.; Reich, P. B.; Naumburg, E. S.; Koch, G. W.; Kubiske, M. E.; Smith, S. D. (2004). Photosynthesis,
carboxylation and leaf nitrogen responses of 16 species to elevated pCO2 across four free-air CO2 enrichment
experiments in forest, grassland and desert. Global Change Biol, 10: 2121-2138. 080092
Eltayeb, A. E.; Kawano, N.; Badawi, G. H.; Kaminaka, H.; Sanekata, T; Morishima, L; Shibahara, T; Inanaga, S.; Tanaka,
K. (2006). Enhanced tolerance to ozone and drought stresses in transgenic tobacco overexpressing
dehydroascorbate reductase in cytosol. Physiol Plant, 127: 57-65. http://dx.doi.Org/10.llll/j.1399-
3054.2005.00624.x 191377
Eltayeb, A. E.; Kawano, N.; Badawi, G. H.; Kaminaka, H.; Sanekata, T; Shibahara, T; Inanaga, S.; Tanaka, K. (2007).
Overexpression of monodehydroascorbate reductase in transgenic tobacco confers enhanced tolerance to ozone, salt
and polyethylene glycol stresses. Planta, 225: 1255-1264. http://dx.doi.org/10.1007/s00425-006-0417-7 191553
Emberson, L. D.; Wieser, G; Ashmore, M. R. (2000). Modelling of stomatal conductance and ozone flux of Norway
spruce: comparison with field data. Environ Pollut, 109: 393-402. 042537
Emberson, L.; Ashmore, M. R.; Cambridge, H. M.; Simpson, D.; Tuovinen, J.-P (2000). Modelling stomatal ozone flux
across Europe. Environ Pollut, 109: 403-413. 040350
Enders, G. (1992). Deposition of ozone to a mature spruce forest: measurements and comparison to models. Environ Pollut,
75: 61-67. 040355
Esperschutz, J.; Pritsch, K.; Gattinger, A.; Welzl, G; Haesler, F.; Buegger, F.; Winkler, J. B.; Munch, J. C.; Schloter, M.
(2009). Influence of chronic ozone stress on carbon translocation pattern into rhizosphere microbial communities of
beech trees (Fagus sylvaticaL.) during a growing season. Plant Soil, 323: 85-95. http://dx.doi.org/10.1007/slll04-
009-0090-2 595669
March 2011 9-143 DRAFT - DO NOT CITE OR QUOTE
-------
Fares, S.; Barta, C.; Brilli, R; Centritto, M.; Ederli, L.; Ferranti, R; Pasqualini, S.; Reale, L.; Tricoli, D.; Loreto, R (2006).
Impact of high ozone on isoprene emission, photosynthesis and histology of developing Populus alba leaves
directly or indirectly exposed to the pollutant. Physiol Plant, 128: 456-465. http://dx.doi.Org/10.llll/j. 1399-
3054.2006.00750.x 191455
Fares, S.; Loreto, F.; Kleist, E.; Wildt, J. (2008). Stomatal uptake and stomatal deposition of ozone in isoprene and
monoterpene emitting plants. Plant Biol (Stuttg), 10: 44-54. http://dx.doi.org/10.1055/s-2007-965257 191683
Fares, S.; Oksanen, E.; Lannenpaa, M.; Julkunen-Tiitto, R.; Loreto F (2010). Volatile emissions and phenolic compound
concentrations along a vertical profile of Populus nigra leaves exposed to realistic ozone concentrations. Photosynth
Res, 104: 61-74. http://dx.doi.org/10.1007/slll20-010-9549-5 628522
Federal land managers' air quality related values workgroup (FLAG) phase I report. (2000). Lakewood, CO: U.S. Forest
Service. 088923
Felzer, B. S.; Cronin, T. W.; Melillo, J. M.; Kicklighter, D. W.; Schlosser, C. A. (2009). Importance of carbon-nitrogen
interactions and ozone on ecosystem hydrology during the 21st century. J Geophys Res, 114: G01020.
http://dx.doi.org/10.1029/2008ig000826 191460
Felzer, B.; Kicklighter, D.; Melillo, J.; Wang, C.; Xhuang, Q.; Prinn, R. (2004). Effects of ozone on net primary production
and carbon sequestration in the conterminous United States using a biogeochemistry model. Tellus B Chem Phys
Meteorol, 56B: 230-248. 186927
Felzer, B.; Reilly, J.; Melillo, J.; Kicklighter, D.; Sarofim, M.; Wang, C.; Prinn, R.; Zhuang, Q. (2005). Future effects of
ozone on carbon sequestration and climate change policy using a global biogeochemical model. Clim Change, 73:
345-373. http://dx.doi.org/10.1007/sl0584-005-6776-4 186928
Feng, Y. W.; Komatsu, S.; Furukawa, T.; Koshiba, T.; Kohno, Y. (2008). Proteome analysis of proteins responsive to
ambient and elevated ozone in rice seedlings. Agric Ecosyst Environ, 125: 255-265.
http://dx.doi.0rg/10.1016/i.agee.2008.01.018 191626
Feng, Z. Z.; Kobayashi, K. (2009). Assessing the impacts of current and future concentrations of surface ozone on crop
yield with meta-analysis. Atmos Environ, 43: 1510-1519. http://dx.doi.Org/10.1016/i.atmosenv.2008.ll.033 199223
Feng, Z. Z.; Kobayashi, K.; Ainsworth, E. A. (2008). Impact of elevated ozone concentration on growth, physiology, and
yield of wheat (Triticum aestivum L.): a meta-analysis. Global Change Biol, 14: 2696-2708.
http://dx.doi.0rg/10.llll/i.1365-2486.2008.01673.xl91453
Feng, Z. Z.; Yao, F. F.; Chen, Z.; Wang, X. K.; Meng, Q. W.; Feng, Z. W. (2007). Response of gas exchange and yield
components of field-grown Triticum aestivum L. to elevated ozone in China. Photosynthetica, 45: 441-446.
http://dx.doi.org/10.1007/sll099-007-0073-6 191641
Fenn, M. E.; de Bauer, L. L; Hernandez-Tejeda, T. (2002). Summary of air pollution impacts on forests in the Mexico City
air basin. In Urban Air Pollution and Forests (pp. 337-355). New York, NY: Springer-Verlag New York, Inc. 626806
Fenn, M. E.; Poth, M. A.; Johnson, D. W. (1996). Evidence for nitrogen saturation in the San Bernardino Mountains in
southern California. For Ecol Manage, 82: 211-230. 083540
Findley, D. A.; Keever, G J.; Chappelka, A. H.; Eakes, D. J.; Gillian, D. J. (1997). Differential responses of buddleia
(Buddleia davidii Franch) to ozone. Environ Pollut, 98: 105-111. 036337
Finkelstein, P. L.; Ellestad, T. G; Clarke, J. R; Meyers, T. P.; Schwede, D. B.; Hebert, E. O.; Neal, J. A. (2000). Ozone and
sulfur dioxide dry deposition to forests: observations and model evaluation. J Geophys Res, 105: 15,365-15,377.
024029
Finnan, J. M.; Burke, J. L.; Jones, M. B. (1997). An evaluation of indices that describe the impact of ozone on the yield of
spring wheat (Triticum aestivum L). Atmos Environ, 31: 2685-2693. 083315
Finnan, J. M.; Jones, M. B.; Burke, J. I. (1996). A time-concentration study on the effects of ozone on spring wheat
(Triticum aestivum L) 2 A comparison of indices. Agric Ecosyst Environ, 57: 169-177. 042545
Fiscus, E. L.; Booker, F. L.; Burkey, K. O. (2005). Crop responses to ozone: uptake, modes of action, carbon assimilation
and partitioning. Plant Cell Environ, 28: 997-1011. 079155
Fiscus, E. L.; Philbeck, R.; Britt, A. M.; Booker, F. L. (1999). Growth of Arabidopsis flavonoid mutants under solar
radiation and UV filters. Environ Exp Bot, 41: 231-245. http://dx.doi.org/10.1016/80098-8472(99)00011-8 672964
March 2011 9-144 DRAFT - DO NOT CITE OR QUOTE
-------
Fishman, I; Bowman, K. W.; Burrows, J. P.; Richter, A.; Chance, K. V; Edwards, D. P.; Martin, R. V; Morris, G. A.;
Pierce, R. B.; Ziemke, J. R.; Al-Saadi, J. A.; Creilson, J. K.; Schaack, T. K.; Thompson, A. M. (2008). Remote
sensing of tropospheric pollution from space. Bull Am Meteorol Soc, 89: 805-821.
http://dx.doi.org/10.1175/2008BAMS2526.! 193927
Fishman, J.; Creilson, J. K.; Parker, P. A.; Ainsworth, E. A.; Vining, G. G; Szarka, J.; Booker, F. L.; Xu, X. J. (2010). An
investigation of widespread ozone damage to the soybean crop in the upper Midwest determined from ground-
based and satellite measurements. Atmos Environ, 44: 2248-2256.
http://dx.doi.0rg/10.1016/i.atmosenv.2010.01.015644259
Flagler, R. B. (1998). Recognition of air pollution injury to vegetation: A pictorial atlas. Pittsburgh, PA: Air & Waste
Management Association. 025525
Flowers, M. D.; Fiscus, E. L.; Burkey, K. O.; Booker, F. L.; Dubois, J. J. B. (2007). Photosynthesis, chlorophyll
fluorescence, and yield of snap bean (Phaseolus vulgaris L.) genotypes differing in sensitivity to ozone. Environ
Exp Bot, 61: 190-198. http://dx.doi.Org/10.1016/j.envexpbot.2007.05.009 191852
Fontan, J. A.; Minga, A.; Lopez, A.; Druilhet, A. (1992). Vertical ozone profiles in a pine forest. Atmos Environ, 26A: 863-
869. 040370
Foyer, C. H.; Noctor, G. (2005). Oxidant and antioxidant signalling in plants: Are-evaluation of the concept of oxidative
stress in a physiological context. Plant Cell Environ, 28: 1056-1071. 191555
Foyer, C. H.; Noctor, G. (2005). Redox homeostasis and antioxidant signaling: A metabolic interface between stress
perception and physiological responses. Plant Cell, 17: 1866-1875. http://dx.doi.org/10.1105/tpc.105.033589
631198
Fredericksen, T. S.; Joyce, B. J.; Skelly, J. M.; Steiner, K. C.; Kolb, T. E.; Kouterick, K. B.; Savage, J. E.; Snyder, K. R.
(1995). Physiology, morphology, and ozone uptake of leaves of black cherry seedlings, saplings, and canopy trees.
Environ Pollut, 89: 273-283. 038898
Fredericksen, T. S.; Kolb, T. E.; Skelly, J. M.; Steiner, K. C.; Joyce, B. J.; Savage, J. E. (1996). Light environment alters
ozone uptake per net photosynthetic rate in black cherry trees. Tree Physiol, 16: 485-490. 026653
Freiwald, V.; Haikio, E.; Julkunen-Tiitto, R.; Holopainen, J. K.; Oksanen, E. (2008). Elevated ozone modifies the feeding
behaviour of the common leaf weevil on hybrid aspen through shifts in developmental, chemical, and structural
properties of leaves. Entomol Exp Appl, 128: 66-72. http://dx.doi.Org/10.llll/j.1570-7458.2008.00677.x 196901
Fuhrer, J. (1994). Effects of ozone on managed pasture 1 Effects of open-top chambers on microclimate, ozone flux, and
plant growth. Environ Pollut, 86: 297-305. 055549
Fuhrer, J.; Skarby, L.; Ashmore, M. R. (1997). Critical levels for ozone effects on vegetation in Europe. Environ Pollut, 97:
91-106.030380
Gate, I. M.; McNeill, S.; Ashmore, M. R. (1995). Effects of air pollution on the searching behaviour of an insect parasitoid.
Water Air Soil Pollut, 85: 1425-1430. 026655
Gerosa, G; Marzuoli, R.; Rossini, M.; Panigada, C.; Meroni, M.; Colombo, R.; Faoro, F.; Iriti, M. (2009). A flux-based
assessment of the effects of ozone on foliar injury, photosynthesis, and yield of bean (Phaseolus vulgaris L. cv.
Borlotto Nano Lingua di Fuoco) in open-top chambers. Environ Pollut, 157: 1727-1736.
http://dx.doi.0rg/10.1016/i.envpol.2008.06.028 191403
Gielen, B.; Vandermeiren, K.; Horemans, N.; D'Haese, D.; Serneels, R.; Valcke, R. (2006). Chlorophyll a fluorescence
imaging of ozone-stressed Brassica napus L. plants differing in glucosinolate concentrations. Plant Biol (Stuttg), 8:
698-705. http://dx.doi.org/10.1055/s-2006-924150 191271
Gitay, H.; Brown, S.; Easterling, W.; Jallow, B. (2001). Ecosystems and their goods and services. In Climate change 2001:
impacts, adaptation and vulnerability. Contribution of working group II to the third assessment report of the
Intergovernmental Panel on Climate Change (pp. 237-342). Cambridge, United Kingdom: Cambridge University
Press. 092761
Gombert, S.; Asta, J.; Seaward, M. R. D. (2006). Lichens and tobacco plants as complementary biomonitors of air pollution
in the Grenoble area (Isere, southeast France). Ecol Indicat, 6: 429-443. 089309
March 2011 9-145 DRAFT - DO NOT CITE OR QUOTE
-------
Gonzalez-Fernandez, I.; Kaminska, A.; Dodmani, M.; Goumenaki, E.; Quarrie, S.; Barnes, J. D. (2010). Establishing ozone
flux-response relationships for winter wheat: Analysis of uncertainties based on data for UK and Polish genotypes.
Atmos Environ, 44: 621-630. http://dx.doi.Org/10.1016/j.atmosenv.2009.ll.021 381357
Grantz, D. A.; Gunn, S.; Vu, H. B. (2006). O-3 impacts on plant development: a meta-analysis of root/shoot allocation and
growth. Plant Cell Environ, 29: 1193-1209. http://dx.doi.Org/10.llll/j.1365-3040.2006.01521.x 191545
Grantz, D. A.; Shrestha, A. (2006). Tropospheric ozone and interspecific competition between yellow nutsedge and pima
cotton. Crop Sci, 46: 1879-1889. http://dx.doi.org/10.2135/cropsci2005.06.0167191702
Grantz, D. A.; Shrestha, A.; Vu, H. B. (2008). Early vigor and ozone response in horseweed (Conyza canadensis) biotypes
differing in glyphosate resistance. Weed Sci, 56: 224-230. http://dx.doi.org/10.1614/ws-07-130.! 191312
Grantz, D. A.; Shrestha, A.; Vu, H. B. (2010). Ozone impacts on assimilation and allocation to reproductive sinks in the
vegetatively propagated C-4 weed, yellow nutsedge. Crop Sci, 50: 246-252.
http://dx.doi.org/10.2135/cropsci2009.03.0127 102161
Grantz, D. A.; Vu, H. B. (2009). O3 sensitivity in a potential C4 bioenergy crop: Sugarcane in California. Crop Sci, 49:
643-650. 195237
Grantz, D. A.; Vu, H. B.; Aguilar, C.; Rea MA (2010). No interaction between methyl jasmonate and ozone in Pima cotton:
Growth and allocation respond independently to both. Plant Cell Environ, 33: 717-728.
http://dx.doi.0rg/10.llll/i.1365-3040.2009.02096.x625419
Grantz, D. A.; Zhang, X. J.; Massman, W. J.; Den Hartog, G; Neumann, H. H.; Pederson, J. R. (1995). Effects of stomatal
conductance and surface wetness on ozone deposition in field-grown grape. Atmos Environ, 29: 3189-3198. 026659
Grantz, D. A.; Zhang, X. J.; Massman, W.; Delany, A.; Pederson, R. (1997). Ozone deposition to a cotton (Gossypium
hirsutum L) field: stomatal and surface wetness effects during the California Ozone Deposition experiment. Agr
Forest Meteorol, 85: 19-31. 026664
Grebenc, T; Kraigher, H. (2007). Changes in the community of ectomycorrhizal fungi and increased fine root number
under adult beech trees chronically fumigated with double ambient ozone concentration. Plant Biol (Stuttg), 9: 279-
287. http://dx.doi.org/10.1055/s-2006-924489 191265
Gregg, J. W.; Jones, C. G; Dawson, T. E. (2003). Urbanization effects on tree growth in the vicinity of New York City
[letter]. Nature, 424: 183-187. 046996
Gregg, J.; Jones, C.; Dawson, T. (2006). Physiological and developmental effects of O3 on cottonwood growth in urban
and rural sites. EcolAppl, 16: 2368-2381. http://dx.doi.org/10.1890/1051-0761(2006)016[2368:PADEOO]2.0.CO;2
186961
Grennfelt, P. (2004). New directions: recent research findings may change ozone control policies. Atmos Environ, 38:
2215-2216.056625
Groppa, M. D.; Benavides, M. P. (2008). Polyamines and abiotic stress: Recent advances. Amino Acids, 34: 35-45.
http://dx.doi.org/10.1007/s00726-007-0501-8 191616
Grulke, N. E. (1999). Physiological responses of ponderosa pine to gradients of environmental stressors. In Miller, P. R.;
McBride, J. R. (Eds.), Oxidant Air Pollution Impacts in the Montane Forests of Southern California New York, NY:
Springer-Verlag. 052983
Grulke, N. E.; Alonso, R.; Nguyen, T.; Cascio, C.; Dobrowolski, W. (2004). Stomata open at night in pole-sized and mature
ponderosa pine: implications for O3 exposure metrics. Tree Physiol, 24: 1001-1010. 042646
Grulke, N. E.; Dobrowolski, W.; Mingus, P.; Fenn, M. E. (2005). California black oak response to nitrogen amendment at a
high O3, nitrogen-saturated site. Environ Pollut, 137: 536-545. http://dx.doi.Org/10.1016/j.envpol.2005.01.039
199433
Grulke, N. E.; Johnson, R.; Esperanza, A.; Jones, D.; Nguyen, T.; Posch, S.; Tausz, M. (2003). Canopy transpiration of
Jeffrey pine in mesic and xeric microsites: O3 uptake and injury response. Trees Struct Funct, 17: 292-298. 042637
Grulke, N. E.; Johnson, R.; Monschein, S.; Nikolova, P.; Tausz, M. (2003). Variation in morphological and biochemical O3
injury attributes of mature Jeffrey pine within canopies and between microsites. Tree Physiol, 23: 923-929. 052984
Grulke, N. E.; Paoletti, E.; Heath, R. L. (2007). Chronic vs. short-term acute O3 exposure effects on nocturnal transpiration
in two Californian oaks. ScientificWorldJournal, 7: 134-140. http://dx.doi.org/10.1100/tsw.2007.33 199376
March 2011 9-146 DRAFT - DO NOT CITE OR QUOTE
-------
Grulke, N. E.; Preisler, H. K.; Rose, C.; Kirsch, J.; Balduman, L. (2002). O3 uptake and drought stress effects on carbon
acquisition of ponderosa pine in natural stands. New Phytol, 154: 621-631. 035283
Grulke, N.; Neufeld, H.; Davison, A.; Roberts, M.; Chappelka, A. (2007). Stomatal behavior of ozone-sensitive and -
insensitive coneflowers (Rudbeckia laciniata var. digitata) in Great Smoky Mountains National Park. New Phytol,
173: 100-109. http://dx.doi.0rg/10.llll/i.1469-8137.2006.01872.xl86963
Grunhage, L.; Haenel, H.-D. (1997). PLATIN (PLant-ATmosphere-INteraction) I: a model of plant-atmosphere interaction
for estimating absorbed doses of gaseous air pollutants. Environ Pollut, 98: 37-50. 040392
Grunhage, L.; Jager, H.-J. (2003). From critical levels to critical loads for ozone: a discussion of a new experimental and
modelling approach for establishing flux-response relationships for agricultural crops and native plant species.
Environ Pollut, 125: 99-110. 052972
Grunhage, L.; Krupa, S. V; Legge, A. H.; Jager, H.-J. (2004). Ambient flux-based critical values of ozone for protecting
vegetation: differing spatial scales and uncertainties in risk assessment. Atmos Environ, 38: 2433-2437. 056621
Guderian, R. (1985). Air pollution by photochemical oxidants: formation, transport, control, and effects on plants. New
York: Springer-Verlag. 019325
Guenther, A.; Karl, T; Harley, P.; Wiedinmyer, C.; Palmer, P. L; Geron, C. (2006). Estimates of global terrestrial isoprene
emissions using MEGAN (Model of Emissions of Gases and Aerosols from Nature). Atmos Chem Phys, 6: 3181-
3210. http://dx.doi.org/10.5194/acp-6-3181-2006 607080
Guidi, L.; Degl'Innocenti, E.; Martinelli, F.; Piras, M. (2009). Ozone effects on carbon metabolism in sensitive and
insensitive Phaseolus cultivars. Environ Exp Bot, 66: 117-125. http://dx.doi.Org/10.1016/j.envexpbot.2008.12.005
199825
Guidi, L.; Degl'Innocenti, E. (2008). Ozone effects on high light-induced photoinhibition in Phaseolus vulgaris. Plant Sci,
174: 590-596. http://dx.doi.Org/10.1016/j.plantsci.2008.03.003 191571
Gumpertz, M. L.; Rawlings, J. O. (1992). Nonlinear regression with variance components: modeling effects of ozone on
crop yield. Crop Sci, 32: 219-224. 043259
Gunderson, C. A.; Sholtis, J. D.; Wullschleger, S. D.; Tissue, D. T.; Hanson, P. J.; Norby, R. J. (2002). Environmental and
stomatal control of photosynthetic enhancement in the canopy of a sweetgum (Liquidambar styraciflua L)
plantation during 3 years of CO2 enrichment. Plant Cell Environ, 25: 379-393. 080097
Haberer, K.; Herbinger, K.; Alexou, M.; Rennenberg, H.; Tausz, M. (2008). Effects of drought and canopy ozone exposure
on antioxidants in fine roots of mature European beech (Fagus sylvatica). Tree Physiol, 28: 713-719.
http://dx.doi.0rg/10.1093/treephvs/28.5.713 191334
Hamel, L. P.; Miles, G P.; Samuel, M. A.; Ellis, B. E.; Seguin, A.; Beaudoin, N. (2005). Activation of stress-responsive
mitogen-activated protein kinase pathways in hybrid poplar (Populus trichocarpa x Populus deltoides). Tree
Physiol, 25: 277-288. http://dx.doi.Org/10.1093/treephvs/25.3.277191214
Handley, T; Grulke, N. E. (2008). Interactive effects of O-3 exposure on California black oak (Quercus kelloggii Newb.)
seedlings with and without N amendment. Environ Pollut, 156: 53-60.
http://dx.doi.0rg/10.1016/i.envpol.2008.01.002 191485
Hanson, P. J.; Wullschleger, S. D.; Norby, R. J.; Tschaplinski, T. J.; Gunderson, C. A. (2005). Importance of changing CO2,
temperature, precipitation, and ozone on carbon and water cycles of an upland-oak forest: incorporating
experimental results into model simulations. Global Change Biol, 11: 1402-1423. http://dx.doi.Org/10.llll/j.1365-
2486.2005.00991.x 191461
Harward, M.; Treshow, M. (1975). Impact of ozone on the growth and reproduction of understorey plants in the Aspen
zone of western USA. Environ Conserv, 2: 17-23. 038366
Hassan, R.; Scholes, R.; Ash, N. (2005). Ecosystems and human well-being: current state and trends, volume 1. United
Kingdom: Shearwater Books. 092759
Hayes, F.; Jones, M. L. M.; Mills, G; Ashmore, M. (2007). Meta-analysis of the relative sensitivity of semi-natural
vegetation species to ozone. Environ Pollut, 146: 754-762. http://dx.doi.Org/10.1016/j.envpol.2006.06.011 196911
Hayes, F.; Mills, G; Ashmore, M. (2009). Effects of ozone on inter- and intra-species competition and photosynthesis in
mesocosms of Lolium perenne and Trifolium repens. Environ Pollut, 157: 208-214.
http://dx.doi.0rg/10.1016/i.envpol.2008.07.002 191360
March 2011 9-147 DRAFT - DO NOT CITE OR QUOTE
-------
He, X. Y; Fu, S. L.; Chen, W.; Zhao, T. H.; Xu, S.; Tuba, Z. (2007). Changes in effects of ozone exposure on growth,
photosynthesis, and respiration of Ginkgo biloba in Shenyang urban area. Photosynthetica, 45: 555-561.
http://dx.doi.org/10.1007/sll099-007-0095-0 199789
He, X. Y; Ruan, Y. N.; Chen, W.; Lu, T. (2006). Responses of anti-oxidative system in leaves of Ginkgo biloba to elevated
ozone concentration in urban area. Botanical Studies, 47: 409-416. 199430
Heagle, A. S. (1979). Effects of growth media, fertiliser rate and hour and season of exposure on sensitivity of four soybean
cultivarsto ozone. Environ Pollut, 18: 313-322. 039329
Heagle, A. S. (1989). Ozone and crop yield*. Annu Rev Phytopathol, 27: 397-423.
http://dx.doi.org/10.1146/annurev.pv.27.090189.002145093985
Heagle, A. S.; Body, D. E.; Heck, W. W. (1973). An open-top field chamber to assess the impact of air pollution on plants. J
Environ Qual, 2: 365-368. 038348
Heagle, A. S.; Brandenburg, R. L.; Burns, J. C.; Miller, J. E. (1994). Ozone and carbon dioxide effects on spider mites in
white clover and peanut. J Environ Qual, 23: 1168-1176. 026699
Heagle, A. S.; Heck, W. W.; Lesser, V. M.; Rawlings, J. O. (1987). Effects of daily ozone exposure duration and
concentration fluctuation on yield of tobacco. Phytopathology, 77: 856-862. 042124
Heagle, A. S.; Kress, L. W; Temple, P. J.; Kohut, R. J.; Miller, J. E.; Heggestad, H. E. (1988). Factors influencing ozone
dose-yield response relationships in open-top field chamber studies. In Assessment of crop loss from air pollutants
(pp. 141-149). London: Elsevier Applied Science. 043559
Heagle, A. S.; Letchworth, M. B.; Mitchell, C. A. (1983). Effects of growth medium and fertilizer rate on the yield response
of soybeans exposed to chronic doses of ozone. Phytopathology, 73: 134-139. 039372
Heagle, A. S.; Miller, J. E.; Rawlings, J. O.; Vozzo, S. F. (1991). Effect of growth stage on soybean response to chronic
ozone exposure. J Environ Qual, 20: 562-570. 042530
Heagle, A. S.; Miller, J. E.; Sherrill, D. E. (1994). A white clover system to estimate effects of tropospheric ozone on
plants. J Environ Qual, 23: 613-621. 042656
Heagle, A. S.; Reinert, R. A.; Miller, J. E. (1996). Response of white clover to ozone in different environments. J Environ
Qual, 25: 273-278. 042660
Heath, R. L. (2008). Modification of the biochemical pathways of plants induced by ozone: What are the varied routes to
change? Environ Pollut, 155: 453-463. http://dx.doi.Org/10.1016/i.envpol.2008.03.010 195632
Heath, R. L.; Lefohn, A. S.; Musselman, R. C. (2009). Temporal processes that contribute to nonlinearity in vegetation
responses to ozone exposure and dose. Atmos Environ, 43: 2919-2928.
http://www.sciencedirect.eom/science/article/B6VH3-4VV2NR8-5/2/lf59clOd7c941154fd5cfa8d69f22dbc. 196783
Heck, W. W.; Cowling, E. B. (1997). The need for a long term cumulative secondary ozone standard - an ecological
perspective. EM, January: 23-33. 084484
Heck, W. W.; Cure, W. W.; Rawlings, J. O.; Zaragoza, L. J.; Heagle, A. S.; Heggestad, H. E.; Kohut, R. J.; Kress, L. W.;
Temple, P. J. (1984). Assessing impacts of ozone on agricultural crops: II. Crop yield functions and alternative
exposure statistics. JAir Pollut Control Assoc, 34: 810-817. 039380
Heck, W. W.; Heagle, A. S.; Miller, J. E.; Rawlings, J. O. (1991). A national program (NCLAN) to assess the impact of
ozone on agricultural resources. In Berglund, R. L.; Lawson, D. R.; McKee, D. J. (Eds.), Tropospheric ozone and
the environment: papers from an international conference; March 1990; Los Angeles, CA (pp. 225-254). Pittsburgh,
PA: Air & Waste Management Association. 042621
Heck, W. W.; Philbeck, R. B.; Dunning, J. A. (1978). A continuous stirred tank reactor (CSTR) system for exposing plants
to gaseous air contaminants: principles, specifications, construction, and operation. Washington, DC: U.S.
Government Printing Office. 037673
Heck, W. W.; Taylor, O. C.; Adams, R.; Bingham, G; Miller, J.; Preston, E.; Weinstein, L. (1982). Assessment of crop loss
from ozone. JAir Waste Manag Assoc, 32: 353-361. 039525
Heggestad, H. E. (1991). Origin of Bel-W3, Bel-C and Bel-B tobacco varieties and their use as indicators of ozone.
Environ Pollut, 74: 263-291.042533
March 2011 9-148 DRAFT - DO NOT CITE OR QUOTE
-------
Heidenreich, B.; Haberer, G; Mayer, K.; Sandermann, H.; Ernst, D. (2005). CDNA array-analysis of mercury- and ozone-
induced genes in Arabidopsis thaliana. Acta Physiologiae Plantarum, 27: 45-51. http://dx.doi.org/10.1007/sll738-
005-0035-1 191260
Hendrey, G. R.; Ellsworth, D. S.; Lewin, K. F.; Nagy, J. (1999). A free-air enrichment system for exposing tall forest
vegetation to elevated atmospheric CO2. Global Change Biol, 5: 293-309. 042641
Hendrey, G. R.; Kimball, B. A. (1994). The FACE program. Agr Forest Meteorol, 70: 3-14. 040397
Hildebrand, E.; Skelly, J. M.; Fredericksen, T. S. (1996). Foliar response of ozone-sensitive hardwood tree species from
1991 to 1993 in the Shenandoah National Park, Virginia. Can J For Res, 26: 658-669. 042670
Hillstrom, M. L.; Lindroth, R. L. (2008). Elevated atmospheric carbon dioxide and ozone alter forest insect abundance and
community composition. Insect Conservation and Diversity, 1: 233-241. http://dx.doi.org/10.1111/j.1752-
4598.2008.00031.x 191367
Hillstrom, M. L.; Vigue, L. M.; Coyle, D. R.; Raffa, K. F.; Lindroth, R. L. (2010). Performance of the invasive weevil
Polydrusus sericeus is influenced by atmospheric CO2 and host species. Agricultural and Forest Entomology, 12:
285-292. http://dx.doi.0rg/10.llll/i.1461-9563.2010.00474.x644679
Himanen, S. J.; Nerg, A. M.; Nissinen, A.; Pinto, D. M.; Stewart, C. N.; Poppy, G. M.; Holopainen, J. K. (2009). Effects of
elevated carbon dioxide and ozone on volatile terpenoid emissions and multitrophic communication of transgenic
insecticidal oilseed rape (Brassica napus). NewPhytol, 181: 174-186. http://dx.doi.org/!0.1111/j. 1469-
8137.2008.02646.x 191338
Himanen, S. J.; Nerg, A. M.; Nissinen, A.; Stewart, C. N.; Poppy, G. M.; Holopainen, J. K. (2009). Elevated atmospheric
ozone increases concentration of insecticidal Bacillus thuringiensis (Bt) CrylAc protein in Bt Brassica napus and
reduces feeding of a Bt target herbivore on the non-transgenic parent. Environ Pollut, 157: 181-185.
http://dx.doi.0rg/10.1016/i.envpol.2008.07.006 191369
Hogg, A.; Uddling, J.; Ellsworth, D.; Carroll, M. A.; Pressley, S.; Lamb, B.; Vogel, C. (2007). Stomatal and non-stomatal
fluxes of ozone to a northern mixed hardwood forest. Tellus B Chem Phys Meteorol, 59: 514-525.
http://dx.doi.0rg/10.llll/i.1600-0889.2007.00269.xl99349
Hogsett, W. E.; Olszyk, D.; Ormrod, D. P.; Taylor, G. E. Jr; Tingey, D. T. (1987). Air pollution exposure systems and
experimental protocols: volume I: A review and evaluation of performance. Washington, DC: U.S. Environmental
Protection Agency. 043780
Hogsett, W. E.; Olszyk, D.; Ormrod, D. P.; Taylor, G. E. Jr; Tingey, D. T. (1987). Air pollution exposure systems and
experimental protocols: volume II: Description of facilities. Washington, DC: U.S. Environmental Protection
Agency. 043465
Hogsett, W. E.; Tingey, D. T; Holman, S. R. (1985). A programmable exposure control system for determination of the
effects of pollutant exposure regimes on plant growth. Atmos Environ, 19: 1135-1145. 039383
Hogsett, W. E.; Tingey, D. T; Lee, E. H. (1988). Ozone exposure indices: Concepts for development and evaluation of their
use. In Assessment of crop loss from air pollutants: Proceedings of an international conference (pp. 107-138). New
York: Elsevier Applied Science. 042128
Hogsett, W. E.; Tingey, D. T; Lee, E. H.; Beedlow, P. A.; Andersen, C. P. (2008). An approach for evaluating the
effectiveness of various ozone Air Quality Standards for protecting trees. Environ Manage, 41: 937-948.
http://dx.doi.org/10.1007/s00267-007-9057-3 191229
Hogsett, W. E.; Weber, J. E.; Tingey, D.; Herstrom, A.; Lee, E. H.; Laurence, J. A. (1997). Environmental auditing: an
approach for characterizing tropospheric ozone risk to forests. J Environ Manage, 21: 105-120. 040402
Holmes, W. E.; Zak, D. R.; Pregitzer, K. S.; King, J. S. (2006). Elevated CO2 and O-3 alter soil nitrogen transformations
beneath trembling aspen, paper birch, and sugar maple. Ecosystems, 9: 1354-1363.
http://dx.doi.org/10.1007/sl0021-006-0163-5 191372
Hong, B.; Weinstein, D.; Swaney, D. (2006). Assessment of ozone effects on nitrate export from Hubbard Brook Watershed
6. Environ Pollut, 141: 8-21. http://dx.doi.Org/10.1016/i.envpol.2005.08.030 186989
Hui, D.; Sims, D. A.; Johnson, D. W.; Chang, W.; Luo, Y. (2002). Effects of gradual versus step increases in carbon dioxide
on Plantago photosynthesis and growth in a microcosm study. Environ Exp Bot, 47: 51-66. 035288
March 2011 9-149 DRAFT - DO NOT CITE OR QUOTE
-------
ICP M&M (2004). Mapping critical levels for vegetation. In Manual on Methodologies and Criteria for Modelling and
Mapping Critical Loads and Levels; and Air Pollution Effects, Risks and Trends International Cooperative
Programme on Modelling and Mapping. Retrieved January 28, 2011, from:
http://www.rivm.nl/en/themasites/icpmm/manual-and-downloads/manual-english/index.html. 677471
Innes, J. L.; Skelly, J. M.; Schaub, M. (2001). Ozone and broadleaved species: A guide to the identification of ozone-
induced foliar injury. Bern; Stuttgart; Wien: Haupt. 048954
Iriti, M.; Di Maro, A.; Bernasconi, S.; Burlini, N.; Simonetti, P.; Picchi, V.; Panigada, C.; Gerosa, G; Parente, A.; Faoro, F.
(2009). Nutritional traits of bean (Phaseolus vulgaris) seeds from plants chronically exposed to ozone pollution. J
Agric Food Chem, 57: 201-208. http://dx.doi.org/10.1021/jf802819m 195635
Iriti, M.; Faoro, F. (2009). Chemical diversity and defence metabolism: How plants cope with pathogens and ozone
pollution. International Journal of Molecular Sciences, 10: 3371-3399. http://dx.doi.org/10.3390/ijmsl0083371
199313
IRRI (2002). Annual Report. Los Bafios, Laguna in the Philippines: International Rice Research Institute, http://irri.org/.
688833
Isebrands, J. G; Dickson, R. E.; Rebbeck, J.; Karnosky, D. F. (2000). Interacting effects of multiple stresses on growth and
physiological processes in northern forest trees. In Mickler, R. A.; Birsdey, R. A.; Horn, J. (Eds.), Responses of
northern U.S. forests to environmental change New York, NY: Springer-Verlag. 044174
Isebrands, J. G; McDonald, E. P.; Kruger, E.; Hendrey, G; Percy, K.; Pregitzer, K.; Sober, J.; Karnosky, D. F. (2001).
Growth responses of Populus tremuloides clones to interacting carbon dioxide and tropospheric ozone. Environ
Pollut, 115: 359-371. 036345
Jackson, D. M.; Heagle, A. S.; Eckel, R. V. W. (1999). Ovipositional response of tobacco hornworm moths (Lepidoptera:
Sphyngidae) to tobacco plants grown under elevated levels of ozone. Environ Entomol, 28: 566-571. 026877
Janzik, I.; Preiskowski, S.; Kneifel, H. (2005). Ozone has dramatic effects on the regulation of the prechorismate pathway
in tobacco (Nicotiana tabacum L. cv. Bel W3). Planta, 223: 20-27. http://dx.doi.org/10.1007/s00425-005-0060-8
191581
Jaoude, M. B.; Katerji, N.; Mastrorilli, M.; Rana, G. (2008). Analysis of the ozone effect on soybean in the Mediterranean
region II. The consequences on growth, yield and water use efficiency. Eur J Agron, 28: 519-525.
http://dx.doi.0rg/10.1016/j.eja.2007.09.001 191223
Jaude, M. B.; Katerji, N.; Mastrorilli, M.; Rana, G. (2008). Analysis of the effect of ozone on soybean in the Mediterranean
region I: the consequences on crop-water status. Eur J Agron, 28: 508-518.
http://dx.doi.0rg/10.1016/i.eia.2007.09.002 191222
Johnson, R. M.; Pregitzer, K. S. (2007). Concentration of sugars, phenolic acids, and amino acids in forest soils exposed to
elevated atmospheric CO2 and O3. Soil Biol Biochem, 39: 3159-3166.
http://dx.doi.0rg/10.1016/i.soilbio.2007.07.010 191287
Jones, M. E.; Paine, T D. (2006). Detecting changes in insect herbivore communities along a pollution gradient. Environ
Pollut, 143: 377-387. http://dx.doi.Org/10.1016/j.envpol.2005.12.013 191301
Jones, M. L. M.; Hodges, G; Mills, G. (2010). Nitrogen mediates above-ground effects of ozone but not below-ground
effects in a rhizomatous sedge. Environ Pollut, 158: 559-565. http://dx.doi.Org/10.1016/j.envpol.2009.08.002
567354
Jones, T. G; Freeman, C.; Lloyd, A.; Mills, G. (2009). Impacts of elevated atmospheric ozone on peatland below-ground
doc characteristics. EcolEng, 35: 971-977. http://dx.doi.Org/10.1016/j.ecoleng.2008.08.009 199881
Joo, J. H.; Wang, S. Y; Chen, J. G; Jones, A. M.; Fedoroff, N. V. (2005). Different signaling and cell death roles of
heterotrimeric G protein alpha and beta subunits in the arabidopsis oxidative stress response to ozone. Plant Cell,
17: 957-970. http://dx.doi.org/10.1105/tpc.104.029603 191307
Joss, U.; Graber, W. K. (1996). Profiles and simulated exchange of H2O, O3, NO2 between the atmosphere and the HartX
Scots pine plantation. Theor Appl Climatol, 53: 157-172. 040408
Kanerva, T; Palojarvi, A.; Ramo, K.; Manninen, S. (2008). Changes in soil microbial community structure under elevated
tropospheric O3 and CO2. Soil Biol Biochem, 40: 2502-2510. http://dx.doi.Org/10.1016/j.soilbio.2008.06.007
191264
March 2011 9-150 DRAFT - DO NOT CITE OR QUOTE
-------
Kanerva, T.; Palojarvi, A.; Ramo, K.; Ojanpera, K.; Esala, M.; Manninen, S. (2006). A 3-year exposure to CO2 and O3
induced minor changes in soil N cycling in a meadow ecosystem. Plant Soil, 286: 61-73.
http://dx.doi.org/10.1007/slll04-006-9026-2 191747
Kanerva, T.; Regina, K.; Ramo, K.; Ojanpera, K.; Manninen, S. (2007). Fluxes of N2O, CH4 and CO2 in a meadow
ecosystem exposed to elevated ozone and carbon dioxide for three years. Environ Pollut, 145: 818-828.
http://dx.doi.0rg/10.1016/j.envpol.2006.03.055 191405
Kangasjarvi, J.; Jaspers, P.; Kollist, H. (2005). Signalling and cell death in ozone-exposed plants. Plant Cell Environ, 28:
1021-1036. 180341
Karlsson, P. E.; Sellden, G; Plaijel, H. (2003). Establishing ozone critical levels II UNECE workshop report. Gothenburg,
Sweden: IVL Swedish Environmental Institute. 055331
Karnosky, D. F.; Gagnon, Z. E.; Dickson, R. E.; Coleman, M. D.; Lee, E. H.; Isebrands, J. G. (1996). Changes in growth,
leaf abscission, biomass associated with seasonal tropospheric ozone exposures of Populus tremuloides clones and
seedlings. Can J For Res, 26: 23-37. 036347
Karnosky, D. R; Mankovska, B.; Percy, K.; Dickson, R. E.; Podila, G. K.; Sober, J.; Noormets, A.; Hendrey, G; Coleman,
M. D.; Kubiske, M.; Pregitzer, K. S.; Isebrands, J. G. (1999). Effects of tropospheric ozone on trembling aspen and
interaction with CO2: results from an OS-gradient and a FACE experiment. Water Air Soil Pollut, 116: 311-322.
035307
Karnosky, D. R; Pregitzer, K. S.; Zak, D. R.; Kubiske, M. E.; Hendrey, G. R.; Weinstein, D.; Nosal, M.; Percy, K. E.
(2005). Scaling ozone responses of forest trees to the ecosystem level in a changing climate. Plant Cell Environ, 28:
965-981. http://dx.doi.0rg/10.llll/i. 1365-3040.2005.01362.x 095556
Kasurinen, A.; Keinanen, M. M.; Kaipainen, S.; Nilsson, L. O.; Vapaavuori, E.; Kontro, M. H.; Holopainen, T. (2005).
Below-ground responses of silver birch trees exposed to elevated CO2 and O-3 levels during three growing
seasons. Global Change Biol, 11: 1167-1179. http://dx.doi.Org/10.llll/i.1365-2486.2005.00970.x 191245
Kasurinen, A.; Peltonen, P. A.; Julkunen-Tiitto, R.; Vapaavuori, E.; Nuutinen, V; Holopainen, T.; Holopainen, J. K. (2007).
Effects of elevated co2 and o-3 on leaf litter phenolics and subsequent performance of litter-feeding soil
macrofauna. Plant Soil, 292: 25-43. http://dx.doi.org/10.1007/slll04-007-9199-3 199826
Kasurinen, A.; Riikonen, J.; Oksanen, E.; Vapaavuori, E.; Holopainen, T. (2006). Chemical composition and decomposition
of silver birch leaf litter produced under elevated CO2 and O-3. Plant Soil, 282: 261-280.
http://dx.doi.org/10.1007/slll04-005-6026-6 191269
Kats, G; Olszyk, D. M.; Thompson, C. R. (1985). Open top experimental chambers for trees. J Air Waste Manag Assoc, 35:
1298-1301.055511
Kats, G; Thompson, C. R.; Kuby, W. C. (1976). Improved ventilation of open top greenhouses. J Air Waste Manag Assoc,
26: 1089-1090. 039799
Keutgen, A. J.; Noga, G; Pawelzik, E. (2005). Cultivar-specific impairment of strawberry growth, photosynthesis,
carbohydrate and nitrogen accumulation by ozone. Environ Exp Bot, 53: 271-280.
http://dx.doi.0rg/10.1016/i.envexpbot.2004.04.003 191295
Keutgen, N.; Keutgen, A. J.; Janssens, M. J. J. (2008). Sweet potato [Ipomoea batatas (L.) Lam.] cultivated as tuber or
leafy vegetable supplier as affected by elevated tropospheric ozone. J Agric Food Chem, 56: 6686-6690.
http://dx.doi.org/10.1021/if8006272191690
King, J. S.; Kubiske, M. E.; Pregitzer, K. S.; Hendrey, G. R.; McDonald, E. P.; Giardina, C. P.; Quinn, V S.; Karnosky, D.
F. (2005). Tropospheric O-3 compromises net primary production in young stands of trembling aspen, paper birch
and sugar maple in response to elevated atmospheric CO2. New Phytol, 168: 623-635.
http://dx.doi.0rg/10.llll/i.1469-8137.2005.01557.xl91701
King, J. S.; Pregitzer, K. S.; Zak, D. R.; Sober, J.; Isebrands, J. G; Dickson, R. E.; Hendrey, G. R.; Karnosky, D. F. (2001).
Fine-root biomass and fluxes of soil carbon in young stands of paper birch and trembling aspen as affected by
elevated atmospheric CO2 and tropospheric O3. Oecologia, 128: 237-250. 041751
Kitao, M.; Low, M.; Heerdt, C.; Grams, T. E. E.; Haberle, K. H.; Matyssek, R. (2009). Effects of chronic elevated ozone
exposure on gas exchange responses of adult beech trees (Fagus sylvatica) as related to the within-canopy light
gradient. Environ Pollut, 157: 537-544. http://dx.doi.Org/10.1016/j.envpol.2008.09.016 191331
March 2011 9-151 DRAFT - DO NOT CITE OR QUOTE
-------
Kline, L. I; Davis, D. D.; Skelly, J. M.; Decoteau, D. R. (2009). Variation in ozone sensitivity within Indian hemp and
common milkweed selections from the Midwest. Northeast Nat, 16: 307-313.
http://dx.doi.org/10.1656/045.016.0210 196918
Kline, L. J.; Davis, D. D.; Skelly, J. M.; Savage, J. E.; Ferdinand, J. (2008). Ozone sensitivity of 28 plant selections
exposed to ozone under controlled conditions. Northeast Nat, 15: 57-66. http://dx.doi.org/10.1656/1092-
6194(2008)15[57:OSOPSE]2.0.CO;2 191591
Kohut, R. (2007). Assessing the risk of foliar injury from ozone on vegetation in parks in the US National Park Service's
Vital Signs Network. Environ Pollut, 149: 348-357. http://dx.doi.Org/10.1016/j.envpol.2007.04.022 093289
Kolb, T. E.; Fredericksen, T. S.; Steiner, K. C.; Skelly, J. M. (1997). Issues in scaling tree size and age responses to ozone: a
review. Environ Pollut, 98: 195-208. 052597
Kollist, T.; Moldau, H.; Rasulov, B.; Oja, V; Ramma, H.; Huve, K.; Jaspers, P.; Kangasjarvi, J.; Kollist, H. (2007). A novel
device detects a rapid ozone-induced transient stomatal closure in intact Arabidopsis and its absence in abi2 mutant.
Physiol Plant, 129: 796-803. http://dx.doi.Org/10.llll/i.1399-3054.2006.00851.xl91539
Kostka-Rick, R.; Hahn, H. U. (2005). Biomonitoring using tobacco Bel W3 provides supplemental information for risk
assessment of vegetation injury due to ozone. Gefahrstoffe Reinhaltung Der Luft, 65: 485-491. 199428
Kozovits, A. R.; Matyssek, R.; Blaschke, H.; Gottlein, A.; Grams, T. E. E. (2005). Competition increasingly dominates the
responsiveness of juvenile beech and spruce to elevated CO2 and/or O-3 concentrations throughout two subsequent
growing seasons. Global Change Biol, 11: 1387-1401. http://dx.doi.Org/10.llll/i.1365-2486.2005.00993.x 191282
Krupa, S. V; Grunhage, L.; Jager, H.-I; Nosal, M.; Manning, W. J.; Legge, A. H.; Hanewald, K. (1995). Ambient ozone
(O3) and adverse crop response: a unified view of cause and effect. Environ Pollut, 87: 119-126. 038895
Krupa, S. V; Nosal, M.; Peterson, D. L. (2001). Use of passive ozone O3 samplers in vegetation effects assessment.
Environ Pollut, 112: 303-309. 040430
Kubiske, M. E.; Quinn, V S.; Heilman, W. E.; McDonald, E. P.; Marquardt, P. E.; Teclaw, R. M.; Friend, A. L.; Karnoskey,
D. F. (2006). Interannual climatic variation mediates elevated CO2 and O3 effects on forest growth. Global Change
Biol, 12: 1054-1068. http://dx.doi.Org/10.llll/i.1365-2486.2006.01152.x093284
Kubiske, M. E.; Quinn, V S.; Marquardt, P. E.; Karnosky, D. F. (2007). Effects of elevated atmospheric CO2 and/or O-3 on
intra- and interspecific competitive ability of aspen. Plant Biol (Stuttg), 9: 342-355. http://dx.doi.org/10.1055/s-
2006-924760 191336
Laffray, X.; Rose, C.; Garrec, J. P. (2007). Estimation of ozone concentration in a valley of the alps mountains based on
bel-w3 tobacco leaf injury. Water Air Soil Pollut, 186: 29-42. http://dx.doi.org/10.1007/sll270-007-9460-7 191384
Langebartels, C.; Kerner, K.; Leonardi, S.; Schraudner, M.; Trost, M.; Heller, W.; Sandermann, H. Jr (1991). Biochemical
plant responses to ozone: I. Differential induction of polyamine and ethylene biosynthesis in tobacco. J Plant
Physiol, 95: 882-889. http://www.plantphysiol.Org/cgi/content/abstract/95/3/882. 043506
Larson, J. L.; Zak, D. R.; Sinsabaugh, R. L. (2002). Extracellular enzyme activity beneath temperate trees growing under
elevated carbon dioxide and ozone. Soil Sci Soc Am J, 66: 1848-1856. 053015
Lawlor, D. W. (1998). Plant responses to global change: temperature and drought stress. In De Kok, L. J.; Stulen, I. (Eds.),
Responses of plant metabolism to air pollution and global change Leiden, The Netherlands: Backhuys Publishers.
029015
Leakey, A. D. B.; Bernacchi, C. J.; Ort, D. R.; Long, S. P. (2006). Long-term growth of soybean at elevated CO2 does not
cause acclimation of stomatal conductance under fully open-air conditions. Plant Cell Environ, 29: 1794-1800.
http://dx.doi.0rg/10.llll/i.1365-3040.2006.01556.x679714
Lee, E. H.; Hogsett, W. E. (1996). Methodology for calculating inputs for ozone secondary standard benefits anaylsis: Part
II. Research Triangle Park, NC: U.S. Environmental Protection Agency. 670278
Lee, E. H.; Hogsett, W. E. (1999). Role of concentrations and time of day in developing ozone exposure indices for a
secondary standard. J Air Waste Manag Assoc, 49: 669-681. 040451
Lee, E. H.; Hogsett, W. E.; Tingey, D. T. (1994). Attainment and effects issues regarding alternative secondary ozone air
quality standards. J Environ Qual, 23: 1129-1140. http://jeq.scijournals.Org/cgi/content/abstract/23/6/l 129. 043268
March 2011 9-152 DRAFT - DO NOT CITE OR QUOTE
-------
Lee, E. H.; Tingey, D. T.; Hogsett, W. E. (1987). Selection of the best exposure-response model using various 7-hour ozone
exposure statistics. Research Triangle Park, NC: U.S. Environmental Protection Agency. 042135
Lee, E. H.; Tingey, D. T; Hogsett, W. E. (1988). Evaluation of ozone exposure indices in exposure-response modeling.
Environ Pollut, 53: 43-62. 042136
Lee, E. H.; Tingey, D. T; Hogsett, W. E. (1988). Evaluation of ozone-exposure indices for relating exposure to plant
production and for estimating agricultural losses (Report No. EPA/600/3-88/039). Washington, DC: U.S.
Environmental Protection Agency. 594572
Lee, E. H.; Tingey, D. T; Hogsett, W. E. (1989). Interrelation of experimental exposure and ambient air quality data for
comparison of ozone exposure indices and estimating agricultural losses (Report No. EPA/600/3-89/047). Corvallis,
OR: U.S. Environmental Protection Agency. 042137
Lee, E. H.; Tingey, D. T; Hogsett, W. E.; Laurence, J. A. (2003). History of tropospheric ozone for the San Bernardino
Mountains of southern California, 1963-1999. Atmos Environ, 37: 2705-2717. 053031
Lee, E. H.; Tingey, D. T; Waschmann, R. S.; Phillips, D. L.; Olszyk, D. M.; Johnson, M. G; Hogsett, W. E. (2009).
Seasonal and long-term effects of CO2 and O-3 on water loss in ponderosa pine and their interaction with climate
and soil moisture. Tree Physiol, 29: 1381-1393. http://dx.doi.org/10.1093/treephys/tpp071 595904
Lee, S.; Yun, S. C. (2006). The ozone stress transcriptome of pepper (Capsicum annuum L.). Molecules and Cells, 21: 197-
205. 191592
Lefohn, A. S.; Benedict, H. M. (1982). Development of a mathematical index that describes ozone concentration,
frequency and duration. Atmos Environ, 16: 2529-2532. 039395
Lefohn, A. S.; Jackson, W.; Shadwick, D. S.; Knudsen, H. P. (1997). Effect of surface ozone exposures on vegetation
grown in the southern Appalachian Mountains: identification of possible areas of concern. Atmos Environ, 31:
1695-1708. 082871
Lefohn, A. S.; Laurence, J. A.; Kohut, R. J. (1988). A comparison of indices that describe the relationship between
exposure to ozone and reduction in the yield of agricultural crops. Atmos Environ, 22: 1229-1240.
http://dx.doi.org/10.1016/0004-6981(88)90353-8 042138
Lefohn, A. S.; Shadwick, D. S. (2000). Differences in trending estimates in the United States using several ozone metrics.
In Proceedings of the 93rd Air & Waste Management Association Annual Conference and Exhibition (pp. AS Id-
645). Pittsburgh, PA: Air & Waste Management Association. 040483
Legge, A. H.; Grunhage, L.; Nosal, M.; Jager, H.-J.; Krupa, S. V. (1995). Ambient ozone and adverse crop response: an
evaluation of North American and European data as they relate to exposure indices and critical levels. J Appl Bot
Food Qual, 69: 192-205. 040689
Leitao, L.; Bethenod, O.; Biolley, J. P. (2007). The impact of ozone on juvenile maize (Zea mays L.) plant photosynthesis:
Effects on vegetative biomass, pigmentation, and carboxylases (PEPc and Rubisco). Plant Biol (Stuttg), 9: 478-488.
http://dx.doi.org/10.1055/s-2007-964942 191456
Leitao, L.; Delacote, E.; Dizengremel, P.; Le Thiec, D.; Biolley, J. P. (2007). Assessment of the impact of increasing
concentrations of ozone on photosynthetic components of maize (Zea mays L.), a C-4 plant. Environ Pollut, 146: 5-
8. http://dx.doi.0rg/10.1016/i.envpol.2006.05.019 199379
Leitao, L.; Maoret, J. J.; Biolley, J. P. (2007). Changes in PEP carboxylase, rubisco and rubisco activase mRNA levels from
maize (Zea mays) exposed to a chronic ozone stress. Biol Res, 40: 137-153. http://dx.doi.org/10.4067/S0716-
97602007000200005 191263
Lesser, V. M.; Rawlings, J. O.; Spruill, S. E.; Somerville, M. C. (1990). Ozone effects on agricultural crops: statistical
methodologies and estimated dose-response relationships. Crop Sci, 30: 148-155. 043015
Levine, J. S.; Pinto, J. P. (1998). The production of CO by biomass burning. In MAK Khalil; JP Pinto; MJ Shearer (Eds.),
Atmospheric carbon monoxide and its environmental effects: Proceedings of the international conference;
December 1997; Portland, Oregon (pp. 251-256). Research Triangle Park, NC: U.S. Environmental Protection
Agency, Office of Research and Development. 029599
Lewis, J. S.; Ditchkoff, S. S.; Lin, J. C.; Muntifering, R. B.; Chappelka, A. H. (2006). Nutritive quality of big bluestem
(Andropogon gerardii) and eastern gamagrass (Tripsacum dactyloides) exposed to tropospheric ozone. R E M, 59:
267-274. 191542
March 2011 9-153 DRAFT - DO NOT CITE OR QUOTE
-------
Li, P. H.; Mane, S. P.; Sioson, A. A.; Robinet, C. V; Heath, L. S.; Bohnert, H. I; Grene, R. (2006). Effects of chronic ozone
exposure on gene expression in Arabidopsis thaliana ecotypes and in Thellungielia halophila. Plant Cell Environ,
29: 854-868. http://dx.doi.0rg/10.llll/i.1365-3040.2005.01465.xl91332
Lin, J. C.; Nosal, M.; Muntifering, R. B.; Krupa, S. V. (2007). Alfalfa nutritive quality for ruminant livestock as influenced
by ambient air quality in west-central Alberta. Environ Pollut, 149: 99-103.
http://dx.doi.0rg/10.1016/i.envpol.2006.12.009 196925
Lindroth, R. L. (2010). Impacts of elevated atmospheric CO2 and O3 on forests: Phytochemistry, trophic interactions, and
ecosystem dynamics. J Chem Ecol, 36: 21-Feb. http://dx.doi.org/10.1007/sl0886-009-9731-4 596479
Liu, L. L.; King, J. S.; Giardina, C. P.; Booker, F. L. (2009). The influence of chemistry, production and community
composition on leaf litter decomposition under elevated atmospheric CO2 and tropospheric O-3 in a northern
hardwood ecosystem. Ecosystems, 12: 401-416. http://dx.doi.org/10.1007/sl0021-009-9231-y 191470
Liu, L.; King, J. S.; Giardina, C. P. (2007). Effects of elevated atmospheric CO2 and tropospheric O3 on nutrient dynamics:
Decomposition of leaf litter in termbling aspen and paper birch communities. Plant Soil, 299: 65-82. 093286
Liu, L.; King, J.; Giardina, C. (2005). Effects of elevated concentrations of atmospheric CO2 and tropospheric O3 on leaf
litter production and chemistry in trembling aspen and paper birch communities. Tree Physiol, 25: 1511-1522.
187005
Loats, K. V.; Rebbeck, J. (1999). Interactive effects of ozone and elevated carbon dioxide on the growth and physiology of
black cherry, green ash, and yellow poplar seedlings. Environ Pollut, 106: 237-248. 029709
Long, S. P. (1991). Modification of the response of photosynthetic productivity to rising temperature by atmospheric CO2
concentrations: has its importance been underestimated? Plant Cell Environ, 14: 729-739. 029710
Long, S. P.; Naidu, S. L. (2002). Effects of oxidants at the biochemical, cell and physiological levels, with particular
reference to ozone. In Bell, J. N. B.; Treshow, M. (Eds.), Air Pollution and Plant Life West Sussex, United
Kingdom: John Wiley & Sons, Ltd. 038179
Loranger, G L; Pregitzer, K. S.; King, J. S. (2004). Elevated CO2 and O3t concentrations differentially affect selected
groups of the fauna in temperate forest soils. Soil Biol Biochem, 36: 1521-1524. 072712
Lorenzini, G; Nali, C. (1995). Analysis of vertical ozone and nitrogen oxides profiles in a Prunus cerasifera canopy. Int J
Biometeorol, 39: 1-4. 040691
Loreto, F.; Fares, S. (2007). Is ozone flux inside leaves only a damage indicator? Clues from volatile isoprenoid studies.
Plant Physiol, 143: 1096-1100. http://dx.doi.org/10.1104/pp.106.091892 180259
Loreto, F.; Velikova, V. (2001). Isoprene produced by leaves protects the photosynthetic apparatus against ozone damage,
quenches ozone products, and reduces lipid peroxidation of cellular membranes. Plant Physiol, 127: 1781-1787.
http://www.ncbi.nlm.nih.gOV/pubmed/l 1743121. 657212
Low, M.; Herbinger, K.; Nunn, A. J.; Haberle, K. H.; Leuchner, M.; Heerdt, C.; Werner, H.; Wipfler, P.; Pretzsch, H.; Tausz,
M.; Matyssek, R. (2006). Extraordinary drought of 2003 overrules ozone impact on adult beech trees (Fagus
sylvatica). Trees Struct Funct, 20: 539-548. http://dx.doi.org/10.1007/s00468-006-0069-z 191396
Loya, W. M.; Pregitzer, K. S.; Karberg, N. J.; King, J. S.; Giardina, C. P. (2003). Reduction of soil carbon formation by
tropospheric ozone under elevated carbon dioxide. Nature, 425: 705-707. 074380
Ludwikow, A.; Gallois, P.; Sadowski, J. (2004). Ozone-induced oxidative stress response in Arabidopsis: Transcription
profiling by microarray approach. Cell Mol Biol Lett, 9: 829-842. 595939
Ludwikow, A.; Kierzek, D.; Gallois, P.; Zeef, L.; Sadowski, J. (2009). Gene expression profiling of ozone-treated
Arabidopsis abiltd insertional mutant: Protein phosphatase 2C ABU modulates biosynthesis ratio of ABA and
ethylene. Planta, 230: 1003-1017. http://dx.doi.org/10.1007/s00425-009-1001-8 199440
Ludwikow, A.; Sadowski, J. (2008). Gene networks in plant ozone stress response and tolerance. J Integr Plant Biol, 50:
1256-1267. http://dx.doi.0rg/10.llll/i.1744-7909.2008.00738.xl91426
Luo, Y. (2001). Transient ecosystem response to free-air CO2 enrichment (FACE): experimental evidence and methods of
analysis. New Phytol, 152: 3-8. 035318
Luo, Y; Reynolds, J. F. (1999). Validity of extrapolating field CO2 experiments to predict carbon sequestration in natural
ecosystems. Ecology, 80: 1568-1583. 035319
March 2011 9-154 DRAFT - DO NOT CITE OR QUOTE
-------
Lyons, T. M.; Barnes, J. D. (1998). Influence of plant age on ozone resistance in Plantago major. New Phytol, 138: 83-89.
036354
Maggio, A.; Chiaranda, F. Q.; Cefariello, R.; Fagnano, M. (2009). Responses to ozone pollution of alfalfa exposed to
increasing salinity levels. Environ Pollut, 157: 1445-1452. http://dx.doi.Org/10.1016/j.envpol.2008.09.013 191645
Mahalingam, R.; Jambunathan, N.; Gunjan, S. K.; Faustin, E.; Weng, H.; Ayoubi, P. (2006). Analysis of oxidative
signalling induced by ozone in Arabidopsis thaliana. Plant Cell Environ, 29: 1357-1371.
http://dx.doi.0rg/10.llll/i.1365-3040.2006.01516.xl91221
Mahalingam, R.; Shah, N.; Scrymgeour, A.; Fedoroff, N. (2005). Temporal evolution of the Arabidopsis oxidative stress
response. Plant Mol Biol, 57: 709-730. http://dx.doi.org/10.1007/slll03-005-2860-4 191693
Maier-Maercker, U. (1998). Predisposition of trees to drought stress by ozone. Tree Physiol, 19: 71-78. 029961
Mandl, R. H.; Laurence, J. A.; Kohut, R. J. (1989). Development and testing of open-top chambers for exposing large,
perennial plants to air pollutants. J Environ Qual, 18: 534-540. 043987
Mandl, R. H.; Weinstein, L. H.; McCune, D. C.; Keveny, M. (1973). A cylindrical, open-top chamber for the exposure of
plants to air pollutants in the field. J Environ Qual, 2: 371-376. 039967
Manning, W. J. (2003). Detecting plant effects is necessary to give biological significance to ambient ozone monitoring
data and predictive ozone standards. Environ Pollut, 126: 375-379. 053773
Manning, W. J.; Krupa, S. V. (1992). Experimental methodology for studying the effects of ozone on crops and trees. In
Lefohn, A. S. (Ed.), Surface level ozone exposures and their effects on vegetation Chelsea, MI: Lewis Publishers,
Inc. 044155
Martin, M. J.; Host, G. E.; Lenz, K. E.; Isebrands, J. G (2001). Simulating the growth response of aspen to elevated ozone:
A mechanistic approach to scaling a leaf-level model of ozone effects on photosynthesis to a complex canopy
architecture. Environ Pollut, 115: 425-436. 043678
Massman, W. J. (2004). Toward an ozone standard to protect vegetation based on effective dose: A review of deposition
resistances and a possible metric. Atmos Environ, 38: 2323-2337. 055350
Massman, W. J.; Grantz, D. A. (1995). Estimating canopy conductance to ozone uptake from observations of
evapotranspiration at the canopy scale and at the leaf scale. Global Change Biol, 1: 183-198. 040698
Massman, W. J.; Musselman, R. C.; Lefohn, A. S. (2000). A conceptual ozone dose-response model to develop a standard
to protect vegetation. Atmos Environ, 34: 745-759. 011616
Matyssek, R.; Gunthardt-Goerg, M. S.; Maurer, S.; Keller, T. (1995). Nighttime exposure to ozone reduces whole-plant
production in Betula pendula. Tree Physiol, 15: 159-165. 040700
Matyssek, R.; Le Thiec, D.; Low, M.; Dizengremel, P.; Nunn, A. J.; Haberle, K. H. (2006). Interactions between drought
and O-3 stress in forest trees. Plant Biol (Stuttg), 8: 11-17. http://dx.doi.org/10.1055/S-2005-873025 191481
Matyssek, R.; Sandermann, H.; Wieser, G; Booker, F.; Cieslik, S.; Musselman, R.; Ernst, D. (2008). The challenge of
making ozone risk assessment for forest trees more mechanistic. Environ Pollut, 156: 567-582.
http://dx.doi.0rg/10.1016/i.envpol.2008.04.017 191262
Matyssek, R.; Wieser, G; Ceulemans, R.; Rennenberg, H.; Pretzsch, H.; Haberer, K.; Low, M.; Nunn, A. J.; Werner, H.;
Wipfler, P.; Obwald, W.; Nikolova, P.; Hanke, D. E.; Kraigher, H.; Tausz, M.; Bahnweg, G; Kitao, M.; Dieler, J.;
Sandermann, H.; Herbinger, K.; Grebenc, T; Blumenrother, M.; Deckmyn, G; Grams, T. E. E.; Heerdt, C.;
Leuchner, M.; Fabian, P.; Haberle, K. H. (2010). Enhanced ozone strongly reduces carbon sink strength of adult
beech (Fagus sylvatica)- Resume from the free-air fumigation study at Kranzberg Forest. Environ Pollut, 158:
2527-2532. http://dx.doi.Org/10.1016/j.envpol.2010.05.009 628553
Mautz, W. J.; Dohm, M. R. (2004). Respiratory and behavioral effects of ozone on a lizard and a frog. Comp Biochem
Physiol A Mol Integr Physiol, 139: 371-377. 644188
Mankovska, B.; Percy, K. E.; Karnosky, D. F. (2005). Impacts of greenhouse gases on epicuticular waxes of Populus
tremuloides Michx.: Results from an open-air exposure and a natural O3 gradient. Environ Pollut, 137: 580-586.
http://dx.doi.0rg/10.1016/i.envpol.2005.01.043672965
March 2011 9-155 DRAFT - DO NOT CITE OR QUOTE
-------
McBride, J. R.; Laven, R. D. (1999). Impact of oxidant air pollutants on forest succession in the mixed conifer forests of
the San Bernardino Mountains. In PR Miller; JR McBride (Eds.), Oxidant air pollution impacts in the Montane
forests of southern California: a case study of the San Bernardino Mountains (pp. 338-352). New York, NY:
Springer-Verlag New York, Inc. 053050
McCarthy, H. R.; Oren, R.; Johnsen, K. H.; Gallet-Budynek, A.; Pritchard, S. G; Cook, C. W.; LaDeau, S. L.; Jackson, R.
B.; Finzi, A. C. (2010). Re-assessment of plant carbon dynamics at the Duke free-air CO2 enrichment site:
Interactions of atmospheric CO2 with nitrogen and water availability over stand development. New Phytol, 185:
514-528. http://dx.doi.0rg/10.llll/i.1469-8137.2009.03078.x679715
McCool, P. M.; Musselman, R. C.; Younglove, T.; Teso, R. R. (1988). Response of kidney bean to sequential ozone
exposures. Environ Exp Bot, 28: 307-313. 041877
McFrederick, Q. S.; Fuentes, J. D.; Roulston, T.; Kathilankal, J. C.; Lerdau, M. (2009). Effects of air pollution on biogenic
volatiles and ecological interactions. Oecologia, 160: 411-420. http://dx.doi.org/10.1007/s00442-009-1318-9
191329
McFrederick, Q. S.; Kathilankal, J. C.; Fuentes, J. D. (2008). Air pollution modifies floral scent trails. Atmos Environ, 42:
2336-2348. http://dx.doi.Org/10.1016/j.atmosenv.2007.12.033 196931
McLaughlin, S. B.; Nosal, M.; Wullschleger, S. D.; Sun, G. (2007). Interactive effects of ozone and climate on tree growth
and water use in a southern Appalachian forest in the USA. New Phytol, 174: 109-124. 090348
McLaughlin, S. B.; Wullschleger, S. D.; Sun, G; Nosal, M. (2007). Interactive effects of ozone and climate on water use,
soil moisture content and streamflow in a southern Appalachian forest in the USA. New Phytol, 174: 125-136.
090347
McLeod, A. R.; Long, S. P. (1999). Free-air carbon dioxide enrichment (FACE) in global change research: A review. Adv
Ecol Res, 28: 1-56. http://dx.doi.org/10.1016/80065-2504(08)60028-8 688834
Medlyn, B. E.; Barton, C. V. M.; Broadmeadow, M. S. J.; Ceulemans, R.; De Angelis, P.; Forstreuter, M.; Freeman, M.;
Jackson, S. B.; Kellomaki, S.; Laitat, E.; Rey, A.; Roberntz, P.; Sigurdsson, B. D.; Strassemeyer, J.; Wang, K.;
Curtis, P. S.; Jarvis, P. G. (2001). Stomatal conductance of forest species after long-term exposure to elevated CO2
concentration: A synthesis. New Phytol, 149: 247-264. 679716
Meehan, T. D.; Crossley, M. S.; Lindroth, R. L. (2010). Impacts of elevated CO2 and O3 on aspen leaf litter chemistry and
earthworm and springtail productivity. Soil Biol Biochem, 42: 1132-1137.
http://dx.doi.0rg/10.1016/i.soilbio.2010.03.019644184
Menendez, A. L; Romero, A. M.; Folcia, A. M.; Martinez-Ghersa, M. A. (2009). Getting the interactions right: Will higher
O-3 levels interfere with induced defenses to aphid feeding? Basic Appl Ecol, 10: 255-264.
http://dx.doi.0rg/10.1016/i.baae.2008.03.010 191430
Menendez, A. L; Romero, A. M.; Folcia, A. M.; Martinez-Ghersa, M. A. (2010). Aphid and episodic O sub(3) injury in
arugula plants (Eruca sativa Mill) grown in open-top field chambers. Agric Ecosyst Environ, 135: 10-14.
http://dx.doi.0rg/10.1016/i.agee.2009.08.005 384046
Mereu, S.; Gerosa, G; Finco, A.; Fusaro, L.; Muys, B.; Manes, F. (2009). Improved sapflow methodology reveals
considerable night-time ozone uptake by Mediterranean species. Biogeosciences, 6: 3151-3162. 102168
Miles, G. P.; Samuel, M. A.; Zhang, Y L.; Ellis, B. E. (2005). RNA interference-based (RNAi) suppression of AtMPK6, an
Arabidopsis mitogen-activated protein kinase, results in hypersensitivity to ozone and misregulation of AtMPK3.
Environ Pollut, 138: 230-237. http://dx.doi.Org/10.1016/i.envpol.2005.04.017191648
Miller, P. L. (1973). Oxidant-induced community change in a mixed conifer forest. In JA Naegele (Ed.), Air pollution
damage to vegetation (pp. 101-117). Washington, DC: American Chemical Society. 039165
Miller, P. R.; Elderman, M. J. (1977). Photochemical oxidant air pollutant effects on a mixed conifer forest ecosystem: A
progress report, 1976. Corvallis, Oregon: U.S. Environmental Protection Agency. 038488
Miller, P. R.; Rechel, J. (1999). Temporal changes in crown condition indices, needle litterfall, and collateral needle injuries
of Ponderosa and Jeffrey pines. In Miller, P. R.; McBride, J. R. (Eds.), Oxidant air pollution impacts in the Montane
forests of southern California: a case study of the San Bernardino Mountains New York, NY: Springer. 040702
Mills, G. (2002). Modification of plant response by environmental conditions. In Bell, J. N. B.; Treshow, M. (Eds.), Air
pollution and plant life Chichester, United Kingdom: John Wiley & Sons Ltd. 035322
March 2011 9-156 DRAFT - DO NOT CITE OR QUOTE
-------
Mills, G; Ball, G; Hayes, R; Fuhrer, I; Skarby, L.; Gimeno, B.; De Temmerman, L. (2000). Development of a multi-factor
model for predicting the effects of ambient ozone on the biomass of white clover. Environ Pollut, 109: 533-542.
030098
Mills, G; Buse, A.; Gimeno, B.; Bermejo, V; Holland, M.; Emberson, L.; Pleijel, H. (2007). A synthesis of AOT40-based
response functions and critical levels of ozone for agricultural and horticultural crops. Atmos Environ, 41: 2630-
2643. http://dx.doi.0rg/10.1016/j.atmosenv.2006.ll.016 180221
Mills, G; Hayes, R; Jones, M. L. M.; Cinderby, S. (2007). Identifying ozone-sensitive communities of (semi-)natural
vegetation suitable for mapping exceedance of critical levels. Environ Pollut, 146: 736-743.
http://dx.doi.0rg/10.1016/i.envpol.2006.04.005 196934
Mills, G; Hayes, R; Wilkinson, S.; Davies, W. J. (2009). Chronic exposure to increasing background ozone impairs
stomatal functioning in grassland species. Global Change Biol, 15: 1522-1533. http://dx.doi.Org/10.llll/j.1365-
2486.2008.01798.x 191272
Mondor, E. B.; Awmack, C. S.; Lindroth, R. L. (2010). Individual growth rates do not predict aphid population densities
under altered atmospheric conditions. Agricultural and Forest Entomology, 12: 293-299.
http://dx.doi.0rg/10.llll/i.1461-9563.2010.00478.x644271
Mondor, E. B.; Tremblay, M. N.; Awmack, C. S.; Lindroth, R. L. (2004). Divergent pheromone-mediated insect behaviour
under global atmospheric change. Global Change Biol, 10: 1820-1824. 074334
Mondor, E. B.; Tremblay, M. N.; Awmack, C. S.; Lindroth, R. L. (2005). Altered genotypic and phenotypic frequencies of
aphid populations under enriched CO2 and O-3 atmospheres. Global Change Biol, 11: 1990-1996.
http://dx.doi.0rg/10.llll/i.1365-2486.2005.01054.xl91217
Morgan, P. B.; Ainsworth, E. A.; Long, S. P. (2003). How does elevated ozone impact soybean? A meta-analysis of
photosynthesis, growth and yield. Plant Cell Environ, 26: 1317-1328. 055527
Morgan, P. B.; Bernacchi, C. J.; Ort, D. R.; Long, S. P. (2004). An in vivo analysis of the effect of season-long open-air
elevation of ozone to anticipated 2050 levels on photosynthesis in soybean. J Plant Physiol, 135: 2348-2357.
072764
Morgan, P. B.; Mies, T. A.; Bollero, G. A.; Nelson, R. L.; Long, S. P. (2006). Season-long elevation of ozone concentration
to projected 2050 levels under fully open-air conditions substantially decreases the growth and production of
soybean. NewPhytol, 170: 333-343. 079186
Morison, J. I. L.; Lawlor, D. W. (1999). Interactions between increasing CO2 concentration and temperature on plant
growth. Plant Cell Environ, 22: 659-682. http://dx.doi.Org/10.1046/j.1365-3040.1999.00443.x 094194
Morsky, S. K.; Haapala, J. K.; Rinnan, R.; Tiiva, P.; Saarnio, S.; Silvola, J.; Holopainen, T.; Martikainen, P. J. (2008).
Long-term ozone effects on vegetation, microbial community and methane dynamics of boreal peatland
microcosms in open-field conditions. Global Change Biol, 14: 1891-1903. http://dx.doi.Org/10.llll/j.1365-
2486.2008.01615.x 191507
Mudd, J. B. (1996). Biochemical basis for the toxicity of ozone. In Yunus, M.; Iqbal, M. (Eds.), Plant response to air
pollution (pp. 267-283). New York, NY: John Wiley & Sons. 052795
Muntifering, R. B.; Chappelka, A. H.; Lin, J. C.; Karnosky, D. R; Somers, G. L. (2006). Chemical composition and
digestibility of Trifolium exposed to elevated ozone and carbon dioxide in a free-air (FACE) fumigation system.
FunctEcol 20: 269-275. http://dx.doi.0rg/10.llll/i.1365-2435.2006.01093.x 191270
Musselman, R. C.; Lefohn, A. S.; Massman, W. J.; Heath, R. L. (2006). A critical review and analysis of the use of
exposure- and flux-based ozone indices for predicting vegetation effects. Atmos Environ, 40: 1869-1888. 121678
Musselman, R. C.; Massman, W. J. (1999). Ozone flux to vegetation and its relationship to plant response and ambient air
quality standards. Atmos Environ, 33: 65-73. 040706
Musselman, R. C.; McCool, P. M.; Younglove, T. (1988). Selecting ozone exposure statistics for determining crop yield
loss from air pollutants. Environ Pollut, 53: 63-78. 042144
Musselman, R. C.; Minnick, T. J. (2000). Nocturnal stomatal conductance and ambient air quality standards for ozone.
Atmos Environ, 34: 719-733. 011612
March 2011 9-157 DRAFT - DO NOT CITE OR QUOTE
-------
Nali, C.; Balducci, E.; Frati, L.; Paoli, L.; Loppi, S.; Lorenzini, G. (2007). Integrated biomonitoring of air quality with
plants and lichens: A case study on ambient ozone from central Italy. Chemosphere, 67: 2169-2176.
http://dx.doi.0rg/10.1016/i.chemosphere.2006.12.036 191478
National ambient air quality standards for ozone (final rule). 73 Fed. Reg. 16436-16514 (2008). 684051
National Ambient Air Quality Standards for Ozone (Proposed Rule). 75 Fed. Reg. 2938-3052 (2010). 684211
National ambient air quality standards for ozone - Proposed rule. 72 Fed. Reg. 37818 (2007). 684055
National primary and secondary ambient air quality standards, Section 109 of the Clean Air Act. 42 USC A§ 7409. (1990).
http://www.epa.gov/air/caa/titlel .html#ia. 037658
Neufeld, H. S.; Renfro, J. R.; Hacker, W. D.; Silsbee, D. (1992). Ozone in Great Smoky Mountains National Park:
dynamics and effects on plants. Abstract from Tropospheric Ozone and the Environment II, Pittsburgh, PA. 038971
Nikolova, P. S.; Andersen, C. P.; Blaschke, H.; Matyssek, R.; Haberle, K. H. (2010). Belowground effects of enhanced
tropospheric ozone and drought in a beech/spruce forest (Fagus sylvatica L./Picea abies [L.] Karst). Environ Pollut,
158: 1071-1078. http://dx.doi.Org/10.1016/i.envpol.2009.07.036626810
Noctor, G; Foyer, C. H. (1998). Ascorbate and glutathione: Keeping active oxygen under control. Annu Rev Plant Physiol
Plant Mol Biol, 49: 249-279. 657213
Norby, R. J.; DeLucia, E. H.; Gielen, B.; Calfapietra, C.; Giardina, C. P.; King, J. S.; Ledford, J.; McCarthy, H. R.; Moore,
D. J. P.; Ceulemans, R.; De Angelis, P.; Finzi, A. C.; Karnosky, D. F.; Kubiske, M. E.; Lukac, M.; Pregitzer, K. S.;
Scarascia-Mugnozza, G. E.; Schlesinger, W. H.; Oren, R. (2005). Forest response to elevated CO2 is conserved
across a broad range of productivity. PNAS, 102: 18052-18056. http://dx.doi.org/10.1073/pnas.0509478102 679717
Novak, K.; Cherubini, P.; Saurer, M.; Fuhrer, J.; Skelly, J. M.; Krauchi, N.; Schaub, M. (2007). Ozone air pollution effects
on tree-ring growth, 13C, visible foliar injury and leaf gas exchange in three ozone-sensitive woody plant species.
Tree Physiol, 27: 941-949. 194630
NFS (2006). Ozone bioindicators. Retrieved January 28, 2011 from
http://www.nature.nps.gov/air/Pubs/bioindicators/index.cfm. 677536
NFS (2007). Ozone effects studies. Retrieved January 28, 2011 from http://www.nature.nps.gov/air/studies/ecoOzone.cfm.
677537
Nussbaum, S.; Geissmann, M.; Fuhrer, J. (1995). Ozone exposure-response relationships for mixtures of perennial ryegrass
and white clover depend on ozone exposure patterns. Atmos Environ, 29: 989-995. 030141
O'Gara, P. J. (1922). Sulfur dioxide and fume problems and their solution. Journal of Industrial and Engineering Chemistry,
14: 744-745. 015018
O'Neill, B. F.; Zangerl, A. R.; Delucia, E. H.; Berenbaum, M. R. (2008). Longevity and fecundity of Japanese beetle
(Popillia japonica) on foliage grown under elevated carbon dioxide. Environ Entomol, 37: 601-607.
www.life.illinois.edu/delucia/PUBLICATIONS/O'Neill%20Env%20Ent.pdf 195653
Ogawa, D.; Nakajima, N.; Sano, T; Tamaoki, M.; Aono, M.; Kubo, A.; Kanna, M.; loki, M.; Kamada, H.; Saji, H. (2005).
Salicylic acid accumulation under O-3 exposure is regulated by ethylene in tobacco plants. Plant Cell Physiol, 46:
1062-1072. http://dx.doi.org/10.1093/pcp/pcill8 191653
Oksanen, E.; Holopainen, T. (2001). Responses of two birch (Betula pendula Roth) clones to different ozone profiles with
similar AOT40 exposure. Atmos Environ, 35: 5245-5254. 019538
Olbrich, M.; Betz, G; Gerstner, E.; Langebartels, C.; Sandermann, H.; Ernst, D. (2005). Transcriptome analysis of ozone-
responsive genes in leaves of European beech (Fagus sylvatica L.). Plant Biol (Stuttg), 7: 670-676.
http://dx.doi.org/10.1055/s-2005-873001 191697
Olbrich, M.; Gerstner, E.; Bahnweg, G; Haberle, K. H.; Matyssek, R.; Welzl, G; Heller, W.; Ernst, D. (2010).
Transcriptional signatures in leaves of adult European beech trees (Fagus sylvatica L.) in an experimentally
enhanced free air ozone setting. Environ Pollut, 158: 977-982. http://dx.doi.Org/10.1016/j.envpol.2009.08.001
625424
Olbrich, M.; Gerstner, E.; Welzl, G; Winkler, J. B.; Ernst, D. (2009). Transcript responses in leaves of ozone-treated beech
saplings seasons at an outdoor free air model fumigation site over two growing seasons. Plant Soil, 323: 61-74.
http://dx.doi.org/10.1007/slll04-009-0129-4596020
March 2011 9-158 DRAFT - DO NOT CITE OR QUOTE
-------
Ollinger, S. V; Aber, J. D.; Reich, P. B. (1997). Simulating ozone effects on forest productivity: interactions among leaf-
and stand-level processes. Ecol Appl, 123: 351-358. 040707
Ollinger, S. V; Aber, J. D.; Reich, P. B.; Freuder, R. J. (2002). Interactive effects of nitrogen deposition, tropospheric
ozone, elevated CO2 and land use history on the carbon dynamics of northern hardwood forests. Global Change
Biol, 8: 545-562. 180189
Olszyk, D. M.; Kats, G; Dawson, P. J.; Bytnerowicz, A.; Wolf, J.; Thompson, C. R. (1986). Characteristics of air exclusion
systems vs chambers for field air pollution studies. J Environ Qual, 15: 326-334. 055530
Orendovici, T.; Skelly, J. M.; Ferdinand, J. A.; Savage, J. E.; Sanz, M.-I; Smith, G. C. (2003). Response of native plants of
northeastern United States and southern Spain to ozone exposures; determining exposure/response relationships.
Environ Pollut, 125: 31-40. 049080
Orendovici-Best, T.; Skelly, J. M.; Davis, D. D.; Ferdinand, J. A.; Savage, J. E.; Stevenson, R. E. (2008). Ozone uptake
(flux) as it relates to ozone-induced foliar symptoms of Prunus serotina and Populus maximowizii x trichocarpa.
Environ Pollut, 151: 79-92. http://dx.doi.Org/10.1016/j.envpol.2007.03.003 196940
Oshima, R. J.; Braegelmann, P. K.; Baldwin, D. W.; Van Way, V.; Taylor, O. C. (1977). Reduction of tomato fruit size and
yield by ozone. JAm Soc Hortic Sci, 102: 289-293. 038938
Oshima, R. J.; Poe, M. P.; Braegelmann, P. K.; Baldwin, D. W; Van Way, V (1976). Ozone dosage-crop loss function for
alfalfa: A standardized method for assessing crop losses from air pollutants. J Air Waste Manag Assoc, 26: 861-865.
038475
Overmyer, K.; Brosche, M.; Kangasjarvi, J. (2003). Reactive oxygen species and hormonal control of cell death. Trends
Plant Sci, 8: 335-342. 053537
Overmyer, K.; Brosche, M.; Pellinen, R.; Kuittinen, T; Tuominen, H.; Ahlfors, R.; Keinanen, M.; Saarma, M.; Scheel, D.;
Kangasjarvi, J. (2005). Ozone-induced programmed cell death in the Arabidopsis radical-induced cell deathl
mutant. Plant Physiol, 137: 1092-1104. http://dx.doi.org/10.1104/pp.104.055681 191596
Overmyer, K.; Kollist, H.; Tuominen, H.; Betz, C.; Langebartels, C.; Wingsle, G; Kangasjarvi, S.; Brader, G; Mullineaux,
P.; Kangasjarvi, J. (2008). Complex phenotypic profiles leading to ozone sensitivity in Arabidopsis thaliana
mutants. Plant Cell Environ, 31: 1237-1249. http://dx.doi.Org/10.llll/i.1365-3040.2008.01837.xl91283
Overmyer, K.; Tuominen, H.; Kettunen, R.; Betz, C.; Langebartels, C.; Sandermann, H. Jr; Kangasjarvi, J. (2000). Ozone-
sensitive Arabidopsis rcdl mutant reveals opposite roles for ethylene and jasmonate signaling pathways in
regulating superoxide-dependent cell death. Plant Cell, 12: 1849-1862. 036361
Pan, Y. D.; Birdsey, R.; Horn, J.; McCullough, K. (2009). Separating effects of changes in atmospheric composition,
climate and land-use on carbon sequestration of US Mid-Atlantic temperate forests. For Ecol Manage, 259: 151-
164. http://dx.doi.0rg/10.1016/j.foreco.2009.09.049 596032
Panek, J. A. (2004). Ozone uptake, water loss and carbon exchange dynamics in annually drought-stressed Pinus ponderosa
forests: Measured trends and parameters for uptake modeling. Tree Physiol, 24: 277-290. 079202
Panek, J. A.; Goldstein, A. H. (2001). Responses of stomatal conductance to drought in ponderosa pine: Implications for
carbon and ozone uptake. Tree Physiol, 21: 337-344. 030190
Panek, J.; Kurpius, M. R.; Goldstein, A. H. (2002). An evaluation of ozone exposure metrics for a seasonally drought-
stressed ponderosa pine ecosystem. Environ Pollut, 117: 93-100. 040712
Paolacci, A. R.; Miraldi, C.; Tanzarella, O. A.; Badiani, M.; Porceddu, E.; Nali, C.; Lorenzini, G; Ciaffi, M. (2007). Gene
expression induced by chronic ozone in the Mediterranean shrub Phillyrea latifolia: Analysis by cDNA-AFLP. Tree
Physiol, 27: 1541-1550. http://dx.doi.org/10.1093/treephys/27.ll.1541 191422
Paoletti, E.; Grulke, N. E. (2010). Ozone exposure and stomatal sluggishness in different plant physiognomic classes.
Environ Pollut, 158: 2664-2671. http://dx.doi.Org/10.1016/j.envpol.2010.04.024 628561
Paoletti, E.; Manning, W. J. (2007). Toward a biologically significant and usable standard for ozone that will also protect
plants. Environ Pollut, 150: 85-95. http://dx.doi.Org/10.1016/i.envpol.2007.06.037 180174
Paoletti, E.; Seufert, G; Delia Rocca, G; Thomsen, H. (2007). Photosynthetic responses to elevated CO2 and O-3 in
Quercus ilex leaves at a natural CO2 spring. Environ Pollut, 147: 516-524.
http://dx.doi.0rg/10.1016/i.envpol.2006.08.039 199365
March 2011 9-159 DRAFT - DO NOT CITE OR QUOTE
-------
Parsons, W. F. I; Bockheim, J. G; Lindroth, R. L. (2008). Independent, interactive, and species-specific responses of leaf
litter decomposition to elevated CO2 and O3 in a northern hardwood forest. Ecosystems, 11: 505-519. 191853
Pearson, S.; Davison, A. W.; Reiling, K.; Ashenden, T.; Ollerenshaw, J. H. (1996). The effects of different ozone exposures
on three contrasting populations of Plantago major. New Phytol, 132: 493-502. 053074
Peltonen, P. A.; Julkunen-Tiitto, R.; Vapaavuori, E.; Holopainen, J. K. (2006). Effects of elevated carbon dioxide and ozone
on aphid oviposition preference and birch bud exudate phenolics. Global Change Biol, 12: 1670-1679.
http://dx.doi.0rg/10.llll/i.1365-2486.2006.01226.xl96946
Peltonen, P. A.; Vapaavuori, E.; Heinonen, J.; Julkunen-Tiitto, R.; Holopainen, J. K. (2010). Do elevated atmospheric CO2
and O-3 affect food quality and performance of folivorous insects on silver birch? Global Change Biol, 16: 918-
935. http://dx.doi.0rg/10.llll/i.1365-2486.2009.02073.x 596482
Penn State (1987). Diagnosing injury to eastern forest trees. University Park, PA: Penn State College of Agricultural
sciences, http://pubs.cas.psu.edu/PubTitle.asp?varTitle=diagnosing 626971
Percy, K. E.; Nosal, M.; Heilman, W.; Dann, T; Sober, J.; Legge, A. H.; Karnosky, D. F. (2007). New exposure-based
metric approach for evaluating O3 risk to North American aspen forests. Environ Pollut, 147: 554-566.
http://dx.doi.0rg/10.1016/i.envpol.2006.10.009093287
Peterson, D. L.; Arbaugh, M. J.; Wakelield, V. A.; Miller, P. R. (1987). Evidence of growth reduction in ozone-injured
Jeffrey pine (Pinus jeffreyi Grev and Ball) in Sequoia and Kings Canyon National Parks. J Air Waste Manag Assoc,
37: 906-912. 042148
Pfleeger, T. G; Plocher, M.; Bichel, P. (2010). Response of pioneer plant communities to elevated ozone exposure. Agric
Ecosyst Environ, 138: 116-126. http://dx.doi.Org/10.1016/i.agee.2010.04.009644281
Phillips, D. L.; Johnson, M. G; Tingey, D. T; Storm, M. J. (2009). Elevated CO2 and O3 effects on fine-root survivorship
in ponderosa pine mesocosms. Oecologia, 160: 827-837. http://dx.doi.org/10.1007/s00442-009-1339-4 199802
Phillips, R. L.; Zak, D. R.; Holmes, W. E.; White, D. C. (2002). Microbial community composition and function beneath
temperate trees exposed to elevated atmospheric carbon dioxide and ozone. Oecologia, 131: 236-244. 041768
Piikki, K.; De Temmerman, L.; Hogy, P.; Pleijel, H. (2008). The open-top chamber impact on vapour pressure deficit and
its consequences for stomatal ozone uptake. Atmos Environ, 42: 6513-6522.
http://dx.doi.0rg/10.1016/i.atmosenv.2008.04.014 199265
Piikki, K.; De Temmerman, L.; Ojanpera, K.; Danielsson, H.; Pleijel, H. (2008). The grain quality of spring wheat
(Triticum aestivum L.) in relation to elevated ozone uptake and carbon dioxide exposure. Eur J Agron, 28: 245-254.
http://dx.doi.0rg/10.1016/i.eia.2007.07.004 199812
Piikki, K.; Vorne, V.; Ojanpera, K.; Pleijel, H. (2007). Impact of elevated O-3 and CO2 exposure on potato (Solanum
tuberosum L. cv. Bintje) tuber macronutrients (N, P, K, Mg, Ca). Agric Ecosyst Environ, 118: 55-64.
http://dx.doi.0rg/10.1016/i.agee.2006.04.012 191451
Pinto, D. M.; Blande, J. D.; Nykanen, R.; Dong, W. X.; Nerg, A. M.; Holopainen, J. K. (2007). Ozone degrades common
herbivore-induced plant volatiles: Does this affect herbivore prey location by predators and parasitoids? J Chem
Ecol, 33: 683-694. http://dx.doi.org/10.1007/sl0886-007-9255-8 196949
Pinto, D. M.; Blande, J. D.; Souza, S. R.; Nerg, A. M.; Holopainen, J. K. (2010). Plant volatile organic compounds (VOCs)
in ozone (O-3) polluted atmospheres: The ecological effects. J Chem Ecol, 36: 22-34.
http://dx.doi.org/10.1007/sl0886-009-9732-3 596490
Pinto, D. M.; Himanen, S. J.; Nissinen, A.; Nerg, A. M.; Holopainen, J. K. (2008). Host location behavior of Cotesia
plutellae Kurdjumov (Hymenoptera: Braconidae) in ambient and moderately elevated ozone in field conditions.
Environ Pollut, 156: 227-231. http://dx.doi.Org/10.1016/i.envpol.2007.12.009616554
Pinto, D. M.; Nerg, A. M.; Holopainen, J. K. (2007). The role of ozone-reactive compounds, terpenes, and green leaf
volatiles (GLVs), in the orientation of Cotesia plutellae. J Chem Ecol, 33: 2218-2228.
http://dx.doi.org/10.1007/sl0886-007-9376-0 196950
Pinto, J. (2009). Wyoming winter smog. Nat Geosci, 2: 88-90. http://dx.doi.org/10.1038/ngeo430 187038
Pleijel, H.; Danielsson h; Ojanpera, K.; De Temmerman, L.; Hogy, P. (2004). Relationships between ozone exposure and
yield loss in wheat and potato - Suggestions of critical levels for ozone effects on crops. Atmos Environ, 38: 2259-
2269. http://dx.doi.0rg/10.1016/j.atmosenv.2003.09.076 036662
March 2011 9-160 DRAFT - DO NOT CITE OR QUOTE
-------
Pleijel, H.; Danielsson, H.; Gelang, I; Sild, E.; Sellden, G. (1998). Growth stage dependence of the grain yield response to
ozone in spring wheat (Triticum aestivum L). Agric Ecosyst Environ, 70: 61-68. 053021
Pleijel, H.; Danielsson, H.; Ojanpera, K.; De Temmerman, L.; Hogy, P.; Badiani, M.; Karlsson, P. E. (2004). Relationships
between ozone exposure and yield loss in European wheat and potato—a comparison of concentration- and flux-
based exposure indices. Atmos Environ, 38: 2259-2269. 056608
Pleijel, H.; Ojanpera, K.; Mortensen, L. (1997). Effects of tropospheric ozone on the yield and grain protein content of
spring wheat (Triticum aestivum L) in the nordic countries. Acta Agric Scand B Soil Plant Sci, 47: 20-25. 043747
Plessl, M.; Elstner, E. F.; Rennenberg, H.; Habermeyer, I; Heiser, I. (2007). Influence of elevated CO2 and ozone
concentrations on late blight resistance and growth of potato plants. Environ Exp Bot, 60: 447-457.
http://dx.doi.0rg/10.1016/i.envexpbot.2007.01.003 196952
Plochl, M.; Lyons, T.; Ollerenshaw, I; Barnes, J. (2000). Simulating ozone detoxification in the leaf apoplast through the
direct reaction with ascorbate. Planta, 210: 454-467. 040802
Pollastrini, M.; Desotgiu, R.; Cascio, C.; Bussotti, F.; Cherubini, P.; Saurer, M.; Gerosa, G; Marzuoli, R. (2010). Growth
and physiological responses to ozone and mild drought stress of tree species with different ecological requirements.
Trees Struct Funct, 24: 695-704. http://dx.doi.org/10.1007/s00468-010-0439-4 644392
Polle, A.; Pell, E. J. (1999). Role of carbon dioxide in modifying the plant response to ozone. In Luo, Y; Mooney, H. A.
(Eds.), Carbon dioxide and environmental stress (pp. 193-213). San Diego, CA: Academic Press. 093689
Pregitzer, K. S.; Burton, A. J.; King, J. S.; Zak, D. R. (2008). Soil respiration, root biomass, and root turnover following
long-term exposure of northern forests to elevated atmospheric Co-2 and tropospheric O-3. New Phytol, 180: 153-
161. http://dx.doi.0rg/10.llll/i.1469-8137.2008.02564.x 191677
Pregitzer, K.; Loya, W.; Kubiske, M.; Zak, D. (2006). Soil respiration in northern forests exposed to elevated atmospheric
carbon dioxide and ozone. Oecologia, 148: 503-516. http://dx.doi.org/10.1007/s00442-006-0381-8 191676
Pretzsch, H.; Dieler, J.; Matyssek, R.; Wipfler, P. (2010). Tree and stand growth of mature Norway spruce and European
beech under long-term ozone fumigation. Environ Pollut, 158: 1061-1070.
http://dx.doi.0rg/10.1016/i.envpol.2009.07.035 580435
Pritsch, K.; Esperschuetz, J.; Haesler, F.; Raidl, S.; Winkler, B.; Schloter, M. (2009). Structure and activities of
ectomycorrhizal and microbial communities in the rhizosphere of Fagus sylvatica under ozone and pathogen stress
in a lysimeter study. Plant Soil, 323: 97-109. http://dx.doi.org/10.1007/sll 104-009-9972-6 626808
Puckette, M. C.; Tang, Y. H.; Mahalingam, R. (2008). Transcriptomic changes induced by acute ozone in resistant and
sensitive Medicago truncatula accessions. BMC Plant Biol, 8: 46. http://dx.doi.org/10.1186/1471-2229-8-46
191698
Ramo, K.; Kanerva, T; Ojanpera, K.; Manninen, S. (2007). Growth onset, senescence, and reproductive development of
meadow species in mesocosms exposed to elevated O3 and CO2. Environ Pollut, 145: 850-860.
http://dx.doi.0rg/10.1016/i.envpol.2006.03.054 191441
Rao, M. V.; Hale, B. A.; Ormrod, D. P. (1995). Amelioration of ozone-induced oxidative damage in wheat plants grown
under high carbon dioxide Role of antioxidant enzymes. J Plant Physiol, 109: 421-432. 030221
Rapport, D. J.; Whitford, W. G. (1999). How ecosystems respond to stress: Common properties of arid and aquatic systems.
Bioscience, 49: 193-203.004595
Rawlings, J. O.; Cure, W. W. (1985). The Weibull function as a dose-response model to describe ozone effects on crop
yields. Crop Sci, 25: 807-814. 039419
Reich, P. B. (1987). Quantifying plant response to ozone: Aunifying theory. Tree Physiol, 3: 63-91. 019314
Reid, C. D.; Fiscus, E. L. (2008). Ozone and density affect the response of biomass and seed yield to elevated CO2 in rice.
Global Change Biol, 14: 60-76. http://dx.doi.Org/10.llll/i.1365-2486.2007.01472.xl91561
Reiling, K.; Davison, A. W. (1992). Effects of a short ozone exposure given at different stages in the development of
Plantago major L. New Phytol, 121: 643-647. 043591
Reiling, K.; Davison, A. W. (1994). Effects of exposure to ozone at different stages in the development of Plantago major L
on chlorophyll fluorescence and gas exchange. New Phytol, 128: 509-514. 035373
March 2011 9-161 DRAFT - DO NOT CITE OR QUOTE
-------
Reinert, R. A.; Eason, G; Barton, J. (1997). Growth and fruiting of tomato as influenced by elevated carbon dioxide and
ozone. NewPhytol, 137: 411-420. 030252
Reinert, R. A.; Ho, M. C. (1995). Vegetative growth of soybean as affected by elevated carbon dioxide and ozone. Environ
Pollut, 89: 89-96. 030247
Ren, W.; Tian, H. Q.; Liu, M. L.; Zhang, C.; Chen, G S.; Pan, S. R; Felzer, B.; Xu, X. R (2007). Effects of tropospheric
ozone pollution on net primary productivity and carbon storage in terrestrial ecosystems of China. J Geophys Res,
112: D22S09. http://dx.doi.org/10.1029/2007jd008521 191366
Ren, W.; Tian, H.; Chen, G; Liu, M.; Zhang, C.; Chappelka, A. H.; Pan, S. (2007). Influence of ozone pollution and climate
variability on net primary productivity and carbon storage in China's grassland ecosystems from 1961 to 2000.
Environ Pollut, 149: 327-335. http://dx.doi.Org/10.1016/i.envpol.2007.05.029581541
Rhea, L.; King, J.; Kubiske, M.; Saliendra, N.; Teclaw, R. (2010). Effects of elevated atmospheric CO2 and tropospheric O-
3 on tree branch growth and implications for hydrologic budgeting. Environ Pollut, 158: 1079-1087.
http://dx.doi.0rg/10.1016/i.envpol.2009.08.038647084
Riikonen, J.; Kets, K.; Darbah, J.; Oksanen, E.; Sober, A.; Vapaavuori, E.; Kubiske, M. E.; Nelson, N.; Karnosky, D. R
(2008). Carbon gain and bud physiology in Populus tremuloides and Betula papyrifera grown under long-term
exposure to elevated concentrations of CO2 and O3. Tree Physiol, 28: 243-254.
http://dx.doi.org/10.1093/treephys/28.2.243 191258
Riikonen, J.; Maenpaa, M.; Alavillamo, M.; Silfver, T; Oksanen, E. (2009). Interactive effect of elevated temperature and
O3 on antioxidant capacity and gas exchange in Betula pendula saplings. Planta, 230: 419-427.
http://dx.doi.org/10.1007/s00425-009-0957-8 195664
Rizzo, M.; Bernardi, R.; Salvini, M.; Nali, C.; Lorenzini, G; Durante, M. (2007). Identification of differentially expressed
genes induced by ozone stress in sensitive and tolerant poplar hybrids. J Plant Physiol, 164: 945-949.
http://dx.doi.0rg/10.1016/i.iplph.2006.07.012 191447
Rodenkirchen, H.; Gottlein, A.; Kozovits, A. R.; Matyssek, R.; Grams, T. E. E. (2009). Nutrient contents and efficiencies of
beech and spruce saplings as influenced by competition and O3/CO2 regime. European Journal of Forest Research,
128: 117-128. http://dx.doi.org/10.1007/sl0342-008-0221-vl91540
Rogers, A.; Allen, D. J.; Davey, P. A.; Morgan, P. B.; Ainsworth, E. A.; Bernacchi, C. J.; Cornic, G; Dermody, O. C.;
Dohleman, R G; Heaton, E. A.; Mahoney, J.; Zhu, X.-G; Delucia, E. H.; Ort, D. R.; Long, S. P. (2004). Leaf
photosynthesis and carbohydrate dynamics of soybean grown throughout their life-cycle under free-air carbon
dioxide enrichment. Plant Cell Environ, 27: 449-458. 079201
Rowland-Bamford, A. J. (2000). Plant responses to changing carbon dioxide and temperature. In Singh, S. N. (Ed.), Trace
gas emissions and plants Dordecht, The Netherlands: Kluwer Academic Publishers. 030257
Ryan, A.; Cojocariu, C.; Possell, M.; Davies, W. J.; Hewitt, C. N. (2009). Defining hybrid poplar (Populus deltoides x
Populus trichocarpa) tolerance to ozone: identifying key parameters. Plant Cell Environ, 32: 31-45.
http://dx.doi.0rg/10.llll/i.1365-3040.2008.01897.xl91299
Ryang, S. Z.; Woo, S. Y; Kwon, S. Y; Kim, S. H.; Lee, S. H.; Kim, K. N.; Lee, D. K. (2009). Changes of net
photosynthesis, antioxidant enzyme activities, and antioxidant contents of Liriodendron tulipifera under elevated
ozone. Photosynthetica, 47: 19-25. http://dx.doi.org/10.1007/sll099-009-0005-8 191267
Samuel, M. A.; Ellis, B. E. (2002). Double jeopardy: Both overexpression and suppression of a redox-activated plant
mitogen-activated protein kinase render tobacco plants ozone sensitive. Plant Cell, 14: 2059-2069.
http://dx.doi.org/10.1105/tpc.002337625703
Samuel, M. A.; Miles, G. P.; Ellis, B. E. (2000). Ozone treatment rapidly activates MAP kinase signalling in plants. Plant J,
22: 367-376. http://dx.doi.0rg/10.1046/i.1365-313x.2000.00741.x625706
Samuel, M. A.; Walia, A.; Mansfield, S. D.; Ellis, B. E. (2005). Overexpression of SIPK in tobacco enhances ozone-
induced ethylene formation and blocks ozone-induced SA accumulation. J Exp Bot, 56: 2195-2201.
http://dx.doi.org/10.1093/ixb/eri219 199316
Samuelson, L. J.; Kelly, J. M. (1997). Ozone uptake in Prunus serotina, Acer rubrum and Quercus rubra forest trees of
different sizes. NewPhytol, 136: 255-264. 040832
March 2011 9-162 DRAFT - DO NOT CITE OR QUOTE
-------
Sanchez, M. J. S.; Pena, G. S.; Lorente, V. C.; Gallego, T. M.; Albert, J. C. (2001). La contamination atmosferica en los
bosques: Guia para la identificacion de danos visibles causados por Ozono. Madrid, Spain: Ministerio de Medio
Ambiente. http://www.ceam.es/VentaLibros/guia_O3/index.htm 626980
Sanz, J.; Bermejo, V; Gimeno, B. S.; Elvira, S.; Alonso, R. (2007). Ozone sensitivity of the Mediterranean terophyte
Trifolium striatum is modulated by soil nitrogen content. Atmos Environ, 41: 8952-8962.
http://dx.doi.0rg/10.1016/i.atmosenv.2007.08.016 199245
Sanz, J.; Muntifering, R. B.; Bermejo, V.; Gimeno, B. S.; Elvira, S. (2005). Ozone and increased nitrogen supply effects on
the yield and nutritive quality of Trifolium subterraneum. Atmos Environ, 39: 5899-5907.
http://dx.doi.0rg/10.1016/i.atmosenv.2005.06.022 196963
Sarkar, A.; Rakwal, R.; Agrawal, S. B.; Shibato, J.; Ogawa, Y; Yoshida, Y; Agrawal, G. K.; Agrawal, M. (2010).
Investigating the impact of elevated levels of ozone on tropical wheat using integrated phenotypical, physiological,
biochemical, and proteomics approaches. J Proteome Res, 9: 4565-4584. http://dx.doi.org/10.1021/Prl002824
657214
Saviranta, N. M. M.; Julkunen-Tiitto, R.; Oksanen, E.; Karjalainen, R. O. (2010). Leaf phenolic compounds in red clover
(Trifolium pratense L.) induced by exposure to moderately elevated ozone. Environ Pollut, 158: 440-446.
http://dx.doi.0rg/10.1016/i.envpol.2009.08.029 102177
Sawada, H.; Kohno, Y. (2009). Differential ozone sensitivity of rice cultivars as indicated by visible injury and grain yield.
Plant Biol (Stuttg), 11: 70-75. http://dx.doi.Org/10.llll/i.1438-8677.2009.00233.xl99426
Scebba, R; Giuntini, D.; Castagna, A.; Soldatini, G; Ranieri, A. (2006). Analysing the impact of ozone on biochemical and
physiological variables in plant species belonging to natural ecosystems. Environ Exp Bot, 57: 89-97.
http://dx.doi.0rg/10.1016/i.envexpbot.2005.04.005 191219
Schaub, M.; Skelly, J. M.; Zhang, J. W.; Ferdinand, J. A.; Savage, J. E.; Stevenson, R. E.; Davis, D. D.; Steiner, K. C.
(2005). Physiological and foliar symptom response in the crowns of Prunus serotina, Fraxinus americana and Acer
rubrum canopy trees to ambient ozone under forest conditions. Environ Pollut, 133: 553-567.
http://dx.doi.0rg/10.1016/i.envpol.2004.06.012 191608
Schraudner, M.; Moeder, W.; Wiese, C.; Van Camp, W.; Inze, D.; Langebartels, C.; Sandermann, H. Jr (1998). Ozone-
induced oxidative burst in the ozone biomonitor plant, tobacco Bel W3. Plant J, 16: 235-245. 056358
Severino, J. R; Stich, K.; Soja, G. (2007). Ozone stress and antioxidant substances in Trifolium repens and Centaureajacea
leaves. Environ Pollut, 146: 707-714. http://dx.doi.Org/10.1016/j.envpol.2006.04.006 199293
Sharkey, T. D.; Wiberley, A. E.; Donohue, A. R. (2008). Isoprene emission from plants: Why and how. Ann Bot, 101: 5-18.
http://dx.doi.org/10.1093/aob/mcm240 191492
Singh, E.; Tiwari, S.; Agrawal, M. (2009). Effects of elevated ozone on photosynthesis and stomatal conductance of two
soybean varieties: A case study to assess impacts of one component of predicted global climate change. Plant Biol
(Stuttg), 11: 101-108. http://dx.doi.0rg/10.llll/i.1438-8677.2009.00263.xl99427
Singh, E.; Tiwari, S.; Agrawal, M. (2010). Variability in antioxidant and metabolite levels, growth and yield of two soybean
varieties: An assessment of anticipated yield losses under projected elevation of ozone. Agric Ecosyst Environ, 135:
168-177. http://dx.doi.0rg/10.1016/j.agee.2009.09.004 386275
Sitch, S.; Cox, P. M.; Collins, W. J.; Huntingford, C. (2007). Indirect radiative forcing of climate change through ozone
effects on the land-carbon sink. Nature, 448: 791-794. http://dx.doi.org/10.1038/nature06059 093294
Skarby, L.; Ottosson, S.; Karlsson, P. E.; Wallina, G; Sellden, G; Medina, E. L.; Pleijel, H. (2004). Growth of Norway
spruce (Picea abies) in relation to different ozone exposure indices: a synthesis. Atmos Environ, 38: 2225-2236.
080375
Smith, G; Coulston, J.; Jepsen, E.; Prichard, T. (2003). A national ozone biomonitoring program: Results from field
surveys of ozone sensitive plants in northeastern forests (1994-2000). Environ Monit Assess, 87: 271-291. 044183
Soja, G; Barnes, J. D.; Posch, M.; Vandermeiren, K.; Pleijel, H.; Mills, G. (2000). Phenological weighting of ozone
exposures in the calculation of critical levels for wheat, bean and plantain. Environ Pollut, 109: 517-524. 030388
Soja, G; Reichenauer, T. G; Eid, M.; Soja, A.-M.; Schaber, R.; Gangl, H. (2004). Long-term ozone exposure and ozone
uptake of grapevines in open-top chambers. Atmos Environ, 38: 2313-2321. 094397
March 2011 9-163 DRAFT - DO NOT CITE OR QUOTE
-------
Somers, G. L.; Chappelka, A. H.; Rosseau, P.; Renfro, J. R. (1998). Empirical evidence of growth decline related to visible
ozone injury. For Ecol Manage, 104: 129-137. 044010
Souza, L.; Neufeld, H. S.; Chappelka, A. H.; Burkey, K. O.; Davison, A. W. (2006). Seasonal development of ozone-
induced foliar injury on tall milkweed (Asclepias exaltata) in Great Smoky Mountains National Park. Environ
Pollut, 141: 175-183. http://dx.doi.org/10.1016/ienvpol.2005.07.022191658
Stampfli, A.; Fuhrer, J. (2010). Spatial heterogeneity confounded ozone-exposure experiment in semi-natural grassland.
Oecologia, 162: 515-522. http://dx.doi.org/10.1007/s00442-009-1462-2 102180
Stewart, C. A.. (1998). Impact of ozone on the reproductive biology of Brassica campestris L and Plantago major L. [Ph. D]
(Doctoral dissertation). Loughborough University, England. 094233
Stewart, C. A.; Black, V. J.; Black, C. R.; Roberts, J. A. (1996). Direct effects of ozone on the reproductive development of
Brassica species. J Plant Physiol, 148: 172-178. 036384
Talhelm, A. F.; Pregitzer, K. S.; Zak, D. R. (2009). Species-specific responses to atmospheric carbon dioxide and
tropospheric ozone mediate changes in soil carbon. Ecol Lett, 12: 1219-1228. http://dx.doi.Org/10.llll/j.1461-
0248.2009.01380.x 596189
Tamaoki, M.; Nakajima, N.; Kubo, A.; Aono, M.; Matsuyama, T; Saji, H. (2003). Transcriptome analysis of OS-exposed
Arabidopsis reveals that multiple signal pathways act mutually antagonistically to induce gene expression. Plant
MolBiol, 53: 443-456. http://dx.doi.Org/10.1023/B:PLAN.0000019064.55734.52080053
Temple, P. J.; Kupper, R. S.; Lennox, R. W.; Rohr, K. (1988). Injury and yield responses of differentially irrigated cotton to
ozone. Agron Sustain Dev, 80: 751-755. 043237
Temple, P. J.; Riechers, G. H.; Miller, P. R. (1992). Foliar injury responses of ponderosa pine seedlings to ozone, wet and
dry acidic deposition, and drought. Environ Exp Bot, 32: 101-113. 043349
Theis, N.; Raguso, R. A. (2005). The effect of pollination on floral fragrance in thistles. J Chem Ecol, 31: 2581-2600.
626461
Thomas, V. F. D.; Braun, S.; Fluckiger, W. (2005). Effects of simultaneous ozone exposure and nitrogen loads on
carbohydrate concentrations, biomass, and growth of young spruce trees (Picea abies). Environ Pollut, 137: 507-
516. http://dx.doi.0rg/10.1016/j.envpol.2005.02.002 075930
Thomas, V. F. D.; Braun, S.; Fluckiger, W. (2006). Effects of simultaneous ozone exposure and nitrogen loads on
carbohydrate concentrations, biomass, growth, and nutrient concentrations of young beech trees (Fagus sylvatica).
Environ Pollut, 143: 341-354. http://dx.doi.Org/10.1016/i.envpol.2005.ll.036 191362
Tingey, D. T; Hogsett, W. E.; Lee, E. H.; Herstrom, A. A.; Azevedo, S. H. (1991). An evaluation of various alternative
ambient ozone standards based on crop yield loss data. In RL Berglund; DR Lawson; DJ McKee (Eds.),
Tropospheric Ozone and the Environment (pp. 272-288). Pittsburgh, PA: Air & Waste Management Association.
042623
Tingey, D. T; Hogsett, W. E.; Lee, E. H.; Laurence, J. A. (2004). Stricter ozone ambient air quality standard has beneficial
effect on ponderosa pine in California. J Environ Manage, 34: 397-405. 042385
Tingey, D. T; Johnson, M. G; Lee, E. H.; Wise, C.; Waschmann, R.; Olszyk, D. M.; Watrud, L. S.; Donegan, K. K. (2006).
Effects of elevated CO2 and O-3 on soil respiration under ponderosa pine. Soil Biol Biochem, 38: 1764-1778.
http://dx.doi.0rg/10.1016/i.soilbio.2005.12.003 191341
Tingey, D. T; McVeety, B. D.; Waschmann, R.; Johnson, M. G; Phillips, D. L.; Rygiewicz, P. T; Olszyk, D. M. (1996). A
versatile sun-lit controlled-environment facility for studying plant and soil processes. J Environ Qual, 25: 614-625.
055536
Tingey, D. T; Rodecap, K. D.; Lee, E. H.; Hogsett, W. E.; Gregg, J. W. (2002). Pod development increases the ozone
sensitivity of Phaseolus vulgaris. Water Air Soil Pollut, 139: 325-341. 040896
Tissue, D. T; Griffin, K. L.; Ball, T. (1999). Photosynthetic adjustment in field-grown ponderosa pine trees after six years
of exposure to elevated CO2. Tree Physiol, 19: 221-228. 029011
Tissue, D. T; Thomas, R. B.; Strain, B. R. (1997). Atmospheric CO2 enrichment increases growth and photosynthesis of
Pinus taeda: A 4 year experiment in the field. Plant Cell Environ, 20: 1123-1134. http://dx.doi.Org/10.1046/j. 1365-
3040.1997.d01-140.x 679718
March 2011 9-164 DRAFT - DO NOT CITE OR QUOTE
-------
Tjoelker, M. G; Volin, J. C.; Oleksyn, I; Reich, P. B. (1995). Interaction of ozone pollution and light effects on
photosynthesis in a forest canopy experiment. Plant Cell Environ, 18: 895-905. 035394
long, D. Q.; Mauzerall, D. L. (2008). Summertime State-Level Source-Receptor Relationships between Nitrogen Oxides
Emissions and Surface Ozone Concentrations over the Continental United States. Environ Sci Technol, 42: 7976-
7984. http://dx.doi.org/10.1021/es7027636 621169
Tong, D.; Mathur, R.; Schere, K.; Kang, D.; Yu, S. (2007). The use of air quality forecasts to assess impacts of air pollution
on crops: Methodology and case study. Atmos Environ, 41: 8772-8784.
http://dx.doi.0rg/10.1016/i.atmosenv.2007.07.060 107431
Topa, M. A.; Vanderklein, D. W.; Corbin, A. (2001). Effects of elevated ozone and low light on diurnal and seasonal carbon
gain in sugar maple. Plant Cell Environ, 24: 663-677. 035398
Tosti, N.; Pasqualini, S.; Borgogni, A.; Ederli, L.; Falistocco, E.; Crispi, S.; Paolocci, F. (2006). Gene expression profiles of
O-3-treated Arabidopsis plants. Plant Cell Environ, 29: 1686-1702. http://dx.doi.org/10.1111/j.1365-
3040.2006.01542.x 191425
Turnipseed, A. A.; Burns, S. P.; Moore, D. J. P.; Hu, J.; Guenther, A. B.; Monson, R. K. (2009). Controls over ozone
deposition to a high elevation subalpine forest. Agr Forest Meteorol, 149: 1447-1459.
http://dx.doi.0rg/10.1016/i.agrformet.2009.04.001 588752
U.S. EPA (1978). Air quality criteria for ozone and other photochemical oxidants (Report No. EPA/600/8-78/004).
Washington, DC: U.S. Environmental Protection Agency, Office of Research and Development. 040586
U.S. EPA (1984). Air quality criteria for ozone and other photochemical oxidants, v 3 (Report No. EPA/600/8-84/020 A).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development.
http://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=2000AVEV.txt. 029711
U.S. EPA (1986). Air quality criteria for ozone and other photochemical oxidants (Report No. EPA-600/8-84-020aF - EPA-
600/8-84-020eF). Research Triangle Park, NC: U.S. Environmental Protection Agency. 017607
U. S. EPA (1992). Summary of selected new information on effects of ozone on health and vegetation: Supplement to 1986
air quality criteria for ozone and other photochemical oxidants (Report No. EPA/600/8-88/105F). Research Triangle
Park, NC: U.S. Environmental Protection Agency, Office of Health and Environmental Assessment.
http://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=30001HVP.txt. 042599
U.S. EPA (1996). Air quality criteria for ozone and related photochemical oxidants Volume II of III (Report No.
EPA/600/P-93/004BF). Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research
and Development. 080827
U.S. EPA (1996). Air quality criteria for ozone and related photochemical oxidants Volume III of III (Report No.
EPA/600/P-93/004cF). Research Triangle Park, NC: U.S. Environmental Protection Agency. (NTIS No. PB96-
185608)080828
U.S. EPA (1996). Review of national ambient air quality standards for ozone: assessment of scientific and technical
information OAQPS staff paper (Report No. EPA/452/R-96/007). Research Triangle Park, NC: U.S. Environmental
Protection Agency. (NTIS No. PB96-203435) 039046
U.S. EPA (2006). Air quality criteria for ozone and related photochemical oxidants (Report No. EPA/600/R-05/004AF).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid= 149923. 088089
U.S. EPA (2007). Review of the national ambient air quality standards for ozone: policy assessment of scientific and
technical information OAQPS staff paper (Report No. EPA/452/R-07/003). Research Triangle Park, NC: U.S.
Environmental Protection Agency, Office of Air Quality Planning and Standards. 090207
Uddling, J.; Hogg, A. J.; Teclaw, R. M.; Carroll, M. A.; Ellsworth, D. S. (2010). Stomatal uptake of O3 in aspen and aspen-
birch forests under free-air CO2 and O3 enrichment. Environ Pollut, 158: 2023-2031.
http://dx.doi.0rg/10.1016/j.envpol.2009.12.001 387073
Uddling, J.; Teclaw, R. M.; Kubiske, M. E.; Pregitzer, K. S.; Ellsworth, D. S. (2008). Sap flux in pure aspen and mixed
aspen-birch forests exposed to elevated concentrations of carbon dioxide and ozone. Tree Physiol, 28: 1231-1243.
191655
March 2011 9-165 DRAFT - DO NOT CITE OR QUOTE
-------
Uddling, I; Teclaw, R. M.; Pregitzer, K. S.; Ellsworth, D. S. (2009). Leaf and canopy conductance in aspen and aspen-
birch forests under free-air enrichment of carbon dioxide and ozone. Tree Physiol, 29: 1367-1380.
http://dx.doi.org/10.1093/treephys/tpp070 596219
United Nations Economic Commission for Europe UNECE (1988). ECE critical levels workshop; March; Bad Harzburg,
Germany [final report]. In [Book title missing] [City missing]: [Publisher missing] 055354
University of Illinois (2010). SoyFACE. Retrieved December 08, 2010 from http://soyface.illinois.edu/. 670286
Unsworth, M. H.; Heagle, A. S.; Heck, W. W. (1984). Gas exchange in open-top field chambers -1 measurement and
analysis of atmospheric resistances to gas exchange. Atmos Environ, 18: 373-380. 041239
Unsworth, M. H.; Heagle, A. S.; Heck, W. W. (1984). Gas exchange in open-top field chambers - II resistances to ozone
uptake by soybeans. Atmos Environ, 18: 381-385. 040024
USDA(2011). Ozone Biomonitoring Program. Retrieved January 28, 2011 from http://www.nrs.fs.fed.us/fia/topics/ozone/.
677550
Valkama, E.; Koricheva, I; Oksanen, E. (2007). Effects of elevated O3, alone and in combination with elevated CO2, on
tree leaf chemistry and insect herbivore performance: A meta-analysis. Global Change Biol, 13: 184-201.
http://dx.doi.0rg/10.llll/i.1365-2486.01284.xl91348
Vandermeiren, K.; Black, C.; Pleijel, H.; de Temmerman, L. (2005). Impact of rising tropospheric ozone on potato: effects
on photosynthesis, growth, productivity and yield quality. Plant Cell Environ, 28: 982-996.
http://dx.doi.0rg/10.llll/i.1365-3040.2005.01316.xl79992
van Buuren, M. L.; Guidi, L.; Fornale, S.; Ghetti, F.; Franceschetti, M.; Soldatini, G F.; Bagni, N. (2002). Ozone-response
mechanisms in tobacco: Implications of polyamine metabolism. NewPhytol, 156: 389-398.
http://dx.doi.0rg/10.1046/i.1469-8137.2002.00539.x631199
Van Dingenen, R.; Dentener, F. J.; Raes, F.; Krol, M. C.; Emberson, L.; Cofala, J. (2009). The global impact of ozone on
agricultural crop yields under current and future air quality legislation. Atmos Environ, 43: 604-618.
http://dx.doi.0rg/10.1016/i.atmosenv.2008.10.033 199765
Velikova, V; Pinelli, P.; Pasqualini, S.; Reale, L.; Ferranti, F.; Loreto, F. (2005). Isoprene decreases the concentration of
nitric oxide in leaves exposed to elevated ozone. NewPhytol, 166: 419-426. http://dx.doi.org/! 0.111 l/j.1469-
8137.2005.01409.x 199410
Vickers, C. E.; Possell, M.; Cojocariu, C. L; Velikova, V B.; Laothawornkitkul, J.; Ryan, A.; Mullineaux, P. M.; Hewitt, C.
N. (2009). Isoprene synthesis protects transgenic tobacco plants from oxidative stress. Plant Cell Environ, 32: 520-
531. http://dx.doi.0rg/10.llll/i.1365-3040.2009.01946.x 191497
Vigue, L. M.; Lindroth, R. L. (2010). Effects of genotype, elevated CO2 and elevated O3 on aspen phytochemistry and
aspen leaf beetle Chrysomela crotchi performance. Agricultural and Forest Entomology, 12: 267-276.
http://dx.doi.0rg/10.llll/i.1461-9563.2010.00475.x644185
Volk, M.; Bungener, P.; Contat, F.; Montani, M.; Fuhrer, J. (2006). Grassland yield declined by a quarter in 5 years of free-
air ozone fumigation. Global Change Biol, 12: 74-83. http://dx.doi.Org/10.llll/j.1365-2486.2005.01083.x 191434
Volk, M.; Geissmann, M.; Blatter, A.; Contat, F.; Fuhrer, J. (2003). Design and performance of a free-air exposure system
to study long-term effects of ozone on grasslands. Atmos Environ, 37: 1341-1350. 055568
Vollenweider, P.; Woodcock, H.; Kelty, M. J.; Hofer, R.-M. (2003). Reduction of stem growth and site dependency of leaf
injury in Massachusetts black cherries exhibiting ozone symptoms. Environ Pollut, 125: 467-480. 051157
Vollsnes, A. V; Kruse, O. M. O.; Eriksen, A. B.; Oxaal, U.; Futsaether, C. M. (2010). In vivo root growth dynamics of
ozone exposed Trifolium subterraneum. Environ Exp Bot, 69: 183-188.
http://dx.doi.0rg/10.1016/i.envexpbot.2010.03.007625576
Vuorinen, T; Nerg, A. M.; Holopainen, J. K. (2004). Ozone exposure triggers the emission of herbivore-induced plant
volatiles, but does not disturb tritrophic signalling. Environ Pollut, 131: 305-311. 626462
Wang, L. L.; He, X. Y; Chen, W. (2009). Effects of elevated ozone on photosynthetic CO2 exchange and chlorophyll a
fluorescence in leaves of Quercus mongolica grown in urban area. Bull Environ Contam Toxicol, 82: 478-481.
http://dx.doi.org/10.1007/s00128-008-9606-3 199303
March 2011 9-166 DRAFT - DO NOT CITE OR QUOTE
-------
Wang, X. K.; Zheng, Q. W.; Feng, Z. Z.; Xie, J. Q.; Feng, Z. W.; Ouyang, Z.; Manning, W. J. (2008). Comparison of a
diurnal vs steady-state ozone exposure profile on growth and yield of oilseed rape (Brassica napus L.) in open-top
chambers in the Yangtze Delta, China. Environ Pollut, 156: 449-453.
http://dx.doi.0rg/10.1016/i.envpol.2008.01.027 191276
Wang, X.; Mauzerall, D. L. (2004). Characterizing distributions of surface ozone and its impact on grain production in
China, Japan and South Korea: 1990 and 2020. Atmos Environ, 38: 4383-4402.
http://dx.doi.0rg/10.1016/i.atmosenv.2004.03.067179978
Wang, X.; Taub, D. R. (2010). Interactive effects of elevated carbon dioxide and environmental stresses on root mass
fraction in plants: A meta-analytical synthesis using pairwise techniques. Oecologia, 163: 1-11.
http://dx.doi.org/10.1007/s00442-010-1572-x387478
Watanabe, M.; Yamaguchi, M.; Tabe, C.; Iwasaki, M.; Yamashita, R.; Funada, R.; Fukami, M.; Matsumura, H.; Kohno, Y;
Izuta, T. (2007). Influences of nitrogen load on the growth and photosynthetic responses of Quercus serrata
seedlings to O3. Trees Struct Funct, 21: 421-432. http://dx.doi.org/10.1007/s00468-007-0134-2 191474
Weinstein, D. A.; Beloin, R. M.; Yanai, R. D. (1991). Modeling changes in red spruce carbon balance and allocation in
response to interacting ozone and nutrient stresses. Tree Physiol, 9: 127-146. 043993
Weinstein, D. A.; Laurence, J. A.; Retzlaff, W. A.; Kern, J. S.; Lee, E. H.; Hogsett, W. E.; Weber, J. (2005). Predicting the
effects of tropospheric ozone on regional productivity of ponderosa pine and white fir. For Ecol Manage, 205: 73-
89. http://dx.doi.0rg/10.1016/i.foreco.2004.10.007 179965
Werner, H.; Fabian, P. (2002). Free-air fumigation of mature trees A novel system for controlled ozone enrichment in
grown-up beech and spruce canopies. Environ Sci Pollut Res Int, 9: 117-121. 053040
Wesely, M. L.; Hicks, B. B. (2000). Areview of the current status of knowledge on dry deposition. Atmos Environ, 34:
2261-2282. 025018
Whitfield, C. P.; Davison, A. W.; Ashenden, T. W. (1996). Interactive effects of ozone and soil volume on Plantago major.
NewPhytol, 134: 287-294. 055544
Whitfield, C. P.; Davison, A. W.; Ashenden, T. W. (1997). Artificial selection and heritability of ozone resistance in two
populations of Plantago major. NewPhytol, 137: 645-655. 036397
Wieser, G; Manning, W. J.; Tausz, M.; Bytnerowicz, A. (2006). Evidence for potential impacts of ozone on Pinus cembra
L. at mountain sites in Europe: An overview. Environ Pollut, 139: 53-58.
http://dx.doi.0rg/10.1016/i.envpol.2005.04.037 191391
Wilkinson, S.; Davies, W. J. (2009). Ozone suppresses soil drying- and abscisic acid (ABA)-induced stomatal closure via
an ethylene-dependent mechanism. Plant Cell Environ, 32: 949-959. http://dx.doi.org/10.1111/j.1365-
3040.2009.01970.x 199758
Wilkinson, S.; Davies, W. J. (2010). Drought, ozone, ABA and ethylene: New insights from cell to plant to community.
Plant Cell Environ, 33: 510-525. http://dx.doi.Org/10.llll/i.1365-3040.2009.02052.x598245
Will, R. E.; Ceulemans, R. (1997). Effects of elevated CO2 concentration on photosynthesis, respiration and carbohydrate
status of coppice Populus hybrids. Physiol Plant, 100: 933-939. http://dx.doi.org/10.1111/j.1399-
3054.1997.tb00020.x 679719
Winner, W. E.; Lefohn, A. S.; Cotter, I. S.; Greitner, C. S.; Nellessen, J.; McEvoy, L. R. Jr; Olson, R. L.; Atkinson, C. J.;
Moore, L. D. (1989). Plant responses to elevational gradients of O3 exposures in Virginia. PNAS, 86: 8828-8832.
043403
Wittig, V. E.; Ainsworth, E. A.; Long, S. P. (2007). To what extent do current and projected increases in surface ozone
affect photosynthesis and stomatal conductance of trees? A meta-analytic review of the last 3 decades of
experiments. Plant Cell Environ, 30: 1150-1162. 191695
Wittig, V. E.; Ainsworth, E. A.; Naidu, S. L.; Karnosky, D. F.; Long, S. P. (2009). Quantifying the impact of current and
future tropospheric ozone on tree biomass, growth, physiology and biochemistry: a quantitative meta-analysis.
Global Change Biol, 15: 396-424. http://dx.doi.Org/10.llll/i.1365-2486.2008.01774.xl91631
Woo, S. Y; Hinckley, T. M. (2005). The effects of ozone on growth and stomatal response in the F-2 generation of hybrid
poplar (Populus trichocarpa x Populus deltoides). Biol Plantarum, 49: 395-404. http://dx.doi.org/10.1007/sl0535-
005-0014-9 191359
March 2011 9-167 DRAFT - DO NOT CITE OR QUOTE
-------
Wright, G. A.; Lutmerding, A.; Dudareva, N.; Smith, B. H. (2005). Intensity and the ratios of compounds in the scent of
snapdragon flowers affect scent discrimination by honeybees (Apis mellifera). J Comp Physiol A Neuroethol Sens
Neural Behav Physiol, 191: 105-114.626463
Wrzaczek, M.; Brosche, M.; Salojarvi, J.; Kangasjarvi, S.; Idanheimo, N.; Mersmann, S.; Robatzek, S.; Karpinski, S.;
Karpinska, B.; Kangasjarvi, J. (2010). Transcriptional regulation of the CRK/DUF26 group of receptor-like protein
kinases by ozone and plant hormones in Arabidopsis. BMC Plant Biol, 10: 95. 644190
Yamaguchi, M.; Watanabe, M.; Iwasaki, M.; Tabe, C.; Matsumura, H.; Kohno, Y; Izuta, T. (2007). Growth and
photosynthetic responses of Fagus crenata seedlings to O-3 under different nitrogen loads. Trees Struct Funct, 21:
707-718. http://dx.doi.org/10.1007/s00468-007-0163-x 191438
Yan, K.; Chen, W.; He, X. Y; Zhang, G. Y; Xu, S.; Wang, L. L. (2010). Responses of photosynthesis, lipid peroxidation
and antioxidant system in leaves of Quercus mongolica to elevated O-3. Environ Exp Bot, 69: 198-204.
http://dx.doi.0rg/10.1016/i.envexpbot.2010.03.008628514
Yoshida, S.; Tamaoki, M.; loki, M.; Ogawa, D.; Sato, Y; Aono, M.; Kubo, A.; Saji, S.; Saji, H.; Satoh, S.; Nakajima, N.
(2009). Ethylene and salicylic acid control glutathione biosynthesis in ozone-exposed Arabidopsis thaliana. Physiol
Plant, 136: 284-298. http://dx.doi.Org/10.llll/i.1399-3054.2009.01220.xl91385
Younglove, T.; McCool, P. M.; Musselman, R. C.; Kahl, M. E. (1994). Growth-stage dependent crop yield response to
ozone exposure. Environ Pollut, 86: 287-295. 044162
Yuan, J. S.; Himanen, S. J.; Holopainen, J. K.; Chen, F.; Stewart, C. N. Jr (2009). Smelling global climate change:
Mitigation of function for plant volatile organic compounds. Trends Ecol Evol, 24: 323-331.
http://dx.doi.0rg/10.1016/i.tree.2009.01.012 199779
Yun, S.-C.; Laurence, J. A. (1999). The response of sensitive and tolerant clones of Populus tremuloides to dynamic ozone
exposure under controlled environmental conditions. New Phytol, 143: 305-313. 044165
Zak, D. R.; Holmes, W. E.; Pregitzer, K. S. (2007). Atmospheric CO2 and O-3 alter the flow of N-15 in developing forest
ecosystems. Ecology, 88: 2630-2639. 191239
Zhang, C.; Tian, H. Q.; Chappelka, A. H.; Ren, W.; Chen, H.; Pan, S. F.; Liu, M. L.; Styers, D. M.; Chen, G. S.; Wang, Y. H.
(2007). Impacts of climatic and atmospheric changes on carbon dynamics in the Great Smoky Mountains National
Park. Environ Pollut, 149: 336-347. http://dx.doi.Org/10.1016/j.envpol.2007.05.028 196983
Zhang, J.; Schaub, M.; Ferdinand, J. A.; Skelly, J. M.; Steiner, K. C.; Savage, J. E. (2010). Leafage affects the responses of
foliar injury and gas exchange to tropospheric ozone in Prunus serotina seedlings. Environ Pollut, 158: 2627-2634.
http://dx.doi.0rg/10.1016/i.envpol.2010.05.003628555
March 2011 9-168 DRAFT - DO NOT CITE OR QUOTE
-------
Chapter 10. The Role of Tropospheric
Ozone in Climate Change and
UV-B Effects
10.1. Introduction
1 Atmospheric O3 plays an important role in the Earth's energy budget by interacting with
2 incoming solar radiation and outgoing infrared radiation. Tropospheric O3 makes up only a small
3 portion of the total column of O3, but it has important incremental effects on the overall radiation
4 budget. This chapter assesses the specific role that tropospheric O3 plays in the earth's radiation
5 budget and how perturbations in tropospheric O3 might affect climate through its role as a
6 greenhouse gas (Section 10.2), and health, ecology and welfare through its role in shielding the
7 earth's surface from solar ultraviolet radiation (Section 10.3).
10.2. Effects of Tropospheric Ozone on Climate
10.2.1. Background
8 Tropospheric O3 is a major greenhouse gas, and increases in its abundance may contribute to
9 climate change (IPCC, 2007, 092980). Models calculate that the global burden of tropospheric O3
10 has doubled since the preindustrial era (Gauss et al, 2006, 630275). while observations indicate that
11 in some regions O3 may have increased by factors as great as 4 or 5 (Marenco et al., 1994, 047733;
12 Staehelin et al., 1994, 055369). These increases are tied to the rise in emissions of O3 precursors
13 from human activity, mainly fossil fuel consumption and agricultural processes. The impact on
14 climate of the O3 change since preindustrial times has been estimated to be about 25-40% of
15 anthropogenic CO2 impact and about 75% of anthropogenic CH4 impact (IPCC, 2007, 092980).
16 ranking it third in importance of the greenhouse gases. In the 21st century as the Earth's population
17 continues to grow and energy technology spreads to developing countries, a further rise in the global
18 burden of tropospheric O3 is possible, with consequences for future climate.
19 To examine the science of a changing climate and to provide balanced and rigorous
20 information to policy makers, the World Meteorological Organization (WMO) and the United
21 Nations Environment Programme (UNEP) formed the Intergovernmental Panel on Climate Change
22 (IPCC) in 1988. The IPCC supports the work of the Conference of Parties (COP) to the United
23 Nations Framework Convention on Climate Change (UNFCCC). The IPCC periodically brings
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments such as the Integrated Science Assessments (ISAs) and the Integrated Risk Information System (IRIS).
March 2011 10-1 DRAFT - DO NOT CITE OR QUOTE
-------
1 together climate scientists from member countries of WMO and the United Nations to review
2 knowledge of the physical climate system, past and future climate change, and evidence of human-
3 induced climate change. IPCC climate assessment reports are issued every 5 to 7 years.
4 This chapter draws in part on the fourth IPCC Assessment Report (AR4) (IPCC, 2007,
5 092980). as well as other peer-reviewed published research. Section 10.2.2 reviews the physics and
6 chemistry of climate change and radiative forcing, together with evidence of climate change in the
7 recent past and projections of future climate change. It also offers a brief description of tropospheric
8 O3 as compared to other greenhouse gases. Section 10.2.3 describes factors that influence the
9 magnitude of O3 effects of climate. Section 10.2.4 considers the competing effects of O3 precursors
10 on climate. Sections 10.2.5 and 10.2.6 describe the effects of changing tropospheric O3 on present-
11 day and future climate, respectively. Finally, Section 10.2.7 presents a summary of the effects of
12 tropospheric O3 on climate.
10.2.2. Physics and Chemistry of Climate Change and Radiative Forcing
10.2.2.1. Physics of Greenhouse Gases
13 The Earth's climate depends upon the flux of energy from the sun and its redistribution in the
14 earth-atmosphere-ocean system. Radiant energy from the sun enters the atmosphere in a range of
15 wavelengths, but peaks strongly in the shortwave (visible) part of the spectrum. Most solar energy at
16 very short wavelengths (e.g., ultraviolet) is absorbed at high altitudes by gases such as stratospheric
17 O3. About 30% of incoming solar radiation is reflected back to space, mainly by clouds or surfaces
18 with high albedo (reflectivity), such as snow, ice, and desert sand. In the troposphere, gases and
19 particles can interact with a fraction of the incoming solar radiation, but for the most part the
20 troposphere is transparent to shortwave radiation. Thus about 70% of shortwave solar radiation
21 penetrates to the Earth's surface and is absorbed. About one-third of the absorbed energy is then re-
22 emitted in the longwave (infrared) portion of the spectrum. The rest goes into evaporating water or
23 soil moisture or emerges as sensible heat.
24 The troposphere is opaque to the outgoing longwave radiation. Polyatomic gases such as CO2,
25 CH4, and O3 absorb and re-emit the radiation upwelling from the Earth's surface, reducing the
26 efficiency with which that energy returns to space. In effect, these gases act as a blanket warming the
27 Earth's surface. This phenomenon, known as the "Greenhouse Effect," was first quantified in the 19th
28 century (Arrhenius, 1896, 043125). and gives rise to the term "greenhouse gas".
10.2.2.2. Climate Change in the Recent Past
29 From the end of the Last Ice Age 12,000 years ago until the mid-1800s, observations from ice
30 cores show that concentrations of the long-lived greenhouse gases CO2, CH4, and N2O have been
31 relatively stable. Unlike these greenhouse gases, O3 is not preserved in ice, and no record of it before
32 the late 1800s exists. Models, however, suggest that it, too, has remained relatively constant during
March 2011 10-2 DRAFT - DO NOT CITE OR QUOTE
-------
1 this time period (Thompson, 1992, 043463; Thompson et al., 1993, 029510). The stable mix of
2 greenhouse gases in the atmosphere has kept the global mean temperature of the Earth close to 15°C.
3 Without the presence of greenhouse gases in the atmosphere, the Earth's temperature would be about
4 30°C cooler, or -15°C. Since the start of the Industrial Revolution, human activity has led to
5 significant increases of greenhouse gases in the atmosphere, mainly through fossil fuel combustion.
6 According to the IPCC AR4 (IPCC, 2007, 092980), we now have "very high confidence" that the net
7 effect of anthropogenic greenhouse gas emissions since 1750 has led to warming, and it is "very
8 likely" that human activity contributed to the 0.76°C rise in global mean temperature observed over
9 the last century. The increase of tropospheric O3 may have contributed 0.1-0.3°C warming to the
10 global climate during this time period (Hansen et al., 2005, 190596; Mickley et al., 2004, 057416).
11 Global cooling due to anthropogenic aerosols (IPCC, 2007, 092980) has likely masked the full
12 warming effect of the anthropogenic greenhouse gases.
10.2.2.3. Projections of Future Climate Change
13 The IPCC AR4 projects a warming of ~0.2°C per decade for the remainder of the 21st century
14 (IPCC, 2007, 092980). Even at constant concentrations of greenhouse gases in the atmosphere,
15 temperatures are expected to increase by about 0.1°C per decade, due to the slow response of oceans
16 to the warming applied so far. It is likely that the Earth will experience longer and more frequent
17 heat waves in the 21st century, together with more frequent droughts and/or heavy precipitation
18 events in some regions, due to perturbations in the hydrological cycle that result from changing
19 temperatures (IPCC, 2007, 092980). Sea levels could increase by 0.3-0.8 m by 2300 due to thermal
20 expansion of the oceans. The extent of Arctic sea ice is expected to decline, and contraction of the
21 Greenland ice sheet could further contribute to the sea level rise (IPCC, 2007, 092980).
22 Projections of future climate change are all associated with some degree of uncertainty. A
23 major uncertainty involves future trends in the anthropogenic emissions of greenhouse gases or their
24 precursors. For the IPCC AR4 climate projections, a set of distinct "storylines" or emission pathways
25 was developed (IPCC, 2000, 080704). Each storyline took into account factors such as population
26 growth, mix of energy technologies, and the sharing of technology between developed and
27 developing nations, and each resulted in a different scenario for anthropogenic emissions. When
28 these trends in emissions are applied to models, these scenarios yield a broad range of possible
29 climate trajectories for the 21st century.
30 A second factor bringing large uncertainty to model projections of future climate is the
31 representation of climate and, especially, climate feedbacks. Arise in surface temperatures would
32 perturb a suite of other processes in the earth-atmosphere-ocean system, which may in turn either
33 amplify the temperature increase (positive feedback) or diminish it (negative feedback). One
34 important feedback involves the increase of water vapor content of the atmosphere that would
35 accompany higher temperatures (Bony et al., 2006, 630272). Water vapor is a potent greenhouse gas;
36 accounting for the water vapor feedback may increase the climate sensitivity to a doubling of CO2 by
37 nearly a factor of two (Held and Soden, 2000, 630279). The ice-albedo feedback is also strongly
March 2011 10-3 DRAFT - DO NOT CITE OR QUOTE
-------
1 positive; a decline in snow cover and sea ice extent would diminish the Earth's albedo, allowing
2 more solar energy to be deposited to the surface (Holland and Bitz, 2003, 630280; Rind et al., 1995,
3 630285). A final example of a climate feedback involves the effects of changing cloud cover in a
4 warming atmosphere. Models disagree on the magnitude and even the sign of this feedback on
5 surface temperatures (Soden and Held, 2006, 631183).
10.2.2.4. Metrics of Potential Climate Change
6 Two different metrics are frequently used to estimate the potential climate impact of some
7 perturbation such as a change in greenhouse gas concentration: (1) global warming potential (GWP);
8 and (2) radiative forcing (RF).
9 GWP indicates the integrated radiative forcing over a specified period (usually 100 years)
10 from a unit mass pulse emission of a greenhouse gas or its precursor, and are reported as the
11 magnitude of this forcing relative to that of CO2. GWP is most useful for comparing the potential
12 climate impacts of long-lived gases, such as N2O or CH4. Since tropospheric O3 has a lifetime on the
13 order of weeks to months, GWP is not seen as a valuable metric for quantifying the importance of O3
14 on climate (Forster et al., 2007, 092936).
15 Radiative forcing is a change in the radiative balance at a particular level of the atmosphere or
16 at the surface when a perturbation is introduced in the earth-atmosphere-ocean system. In the global
17 mean, radiative forcing of greenhouse gases at the tropopause (top of the troposphere) is roughly
18 proportional to the surface temperature response
19 (Committee on Radiative Forcing Effects on Climate; Climate Research Committee; National
20 Research Council et al., 2005, 057409; Hansen et al., 2005, 190596). It thus provides a useful metric
21 for policymakers for assessing the response of the earth's surface temperature to a given change in
22 the concentration of greenhouse gas. Positive values of radiative forcing indicate warming in a test
23 case relative to the control; negative values indicate cooling. The units of radiative forcing are
24 energy flux per area, or W/m2.
25 Radiative forcing requires just a few model years to calculate, and it shows consistency from
26 model to model. However, radiative forcing does not take into account the climate feedbacks that
27 could amplify or dampen the actual surface temperature response, depending on region (Section
28 10.2.2.3). Quantifying the change in surface temperature requires a climate simulation in which all
29 important feedbacks are accounted for. As these processes are not well understood, the surface
30 temperature response to a given radiative forcing is highly uncertain and can vary greatly among
31 models and even from region to region within the same model.
10.2.2.5. Tropospheric Ozone as a Greenhouse Gas
32 Tropospheric O3 differs in important ways from other greenhouse gases. It is not emitted
33 directly, but is produced through photochemical oxidation of CO, CH4, and nonmethane volatile
34 organic compounds (VOCs) in the presence of nitrogen oxide radicals (NOX = NO + NO2; see
35 Section 3.2). It is also supplied by vertical transport from the stratosphere. The lifetime of O3 in the
March 2011 10-4 DRAFT - DO NOT CITE OR QUOTE
-------
1 troposphere is typically a few weeks, resulting in an inhomogeneous distribution that varies
2 seasonally; the distribution of the long-lived greenhouse gases like CO2 and CH4 are much more
3 uniform. The longwave radiative forcing by O3 is mainly due to absorption in the 9.6 um window,
4 where absorption by water vapor is weak. It is therefore less sensitive to local humidity than the
5 radiative forcing by CO2 or CH4, for which there is much more overlap with the water absorption
6 bands (Lenoble, 1993, 630969). And unlike other major greenhouse gases, O3 absorbs in the
7 shortwave as well as the longwave part of the spectrum.
8 Figure 10-1 shows the main steps involved in the influence of tropospheric O3 on climate. An
9 increase in the emissions of O3 precursors leads to an increase in the burden of tropospheric O3. The
10 added O3 then perturbs the radiative balance of the atmosphere, an effect quantified by the radiative
11 forcing metric. This forcing results in climate change, usually expressed as a change in surface
12 temperature. Climate change can also perturb tropospheric O3, as will be discussed in Section
13 10.2.6.3. As shown in Figure 10-2, the IPCC (IPCC, 2007, 092980) reports a radiative forcing of
14 0.35 W/m2 for the change in tropospheric O3 since the preindustrial era, ranking it third in
15 importance after the greenhouse gases CO2 (1.66 W/m2) and CH4 (0.48 W/m2). The error bars
16 encompassing the tropospheric O3 radiative forcing estimate range from 0.25 to 0.65 W/m2, making
17 it relatively more uncertain than the long-lived greenhouse gases.
Precursor emissions:
CO, VOCs, CH4, NOX
(Tgyear1)
Tropospheric ozone
abundance (Tg)
Radiative forcing due to
ozone change (W rrv2)
Climate response
Climate
effects
Figure 10-1. Flowchart for the effects of tropospheric ozone on climate. [Emissions of the ozone
precursors CO, VOCs, CH4, and NOx lead to production of tropospheric ozone. A
change in the burden of tropospheric ozone perturbs the radiative balance of the
atmosphere, leading to radiative forcing. The earth-atmosphere-ocean system
responds to the forcing with a change in climate. Climate change, in turn, can affect
the abundance of tropospheric ozone through multiple mechanisms. Units shown
are those typical for each quantity, with the climate response expressed as a change
in surface temperature.
March 2011
10-5
DRAFT - DO NOT CITE OR QUOTE
-------
RADIATIVE FORCING COMPONENTS
Rr Terns
Long-lived
greenhouse gases *
Ozone
Stratospheric water
vapour from CH,,
Surface albedo
(Direct effect
Cloud albedo
effect
Linear contrails
Solar irradiancc
Total not
anthropogenic
RF values (W rrT) Spatial scale LOSU
1.66[1.49to 1.83]
0.48 [0.43 to 0.53]
0.16 [0.14 to 0.18]
•0.05 [-0.15 to 0.05]
0.35 [0.25 to 0.65]
0.07 [0.02 to 0.12]
-0.2 [-0.4 to 0.0]
0.1 [0.0 to 0.2]
-0.5 [-0.9 to-0.1]
-0.7 [-1.8 to-0.3]
0.01 [0.003 to 0.03]
0.12 [0.06 to 0.30]
1.6 [0.6 to 2.4]
Global
Global
Continental
to global
Global
Local to
continental
Continental
to global
Continental
to global
Continental
Global
High
High
Med
Low
Med
-Low
Med
-Low
LOW
Low
-2
-1 0 1
Radiative Forcing (W m~2)
Source: Used with permission from Cambridge University Press, IPCC (IPCC, 2007, 0929801
Figure 10-2. Global average radiative forcing (RF) estimates and ranges in 2005 for
anthropogenic C02, CH4, ozone and other important agents and mechanisms,
together with the typical geographical extent (spatial scale) of the forcing and the
assessed level of scientific understanding (LOSU).
The net anthropogenic radiative forcing and its range are also shown. These require summing
asymmetric uncertainty estimates from the component terms, and cannot be
obtained by simple addition. Additional forcing factors not included here are
considered to have a very low LOSU.
10.2.3. Factors that Influence the Effect of Tropospheric Ozone on Climate
1 This section describes the main factors that influence the magnitude of the climate response to
2 changes in tropospheric O3. They include: (1) trends in the burden of tropospheric O3; (2) the effect
3 of surface albedo on O3 forcing; (3) the effect of vertical distribution on O3 forcing; (4) feedback
4 factors that can alter the climate response to O3 forcing; and (5) the indirect effects of tropospheric
5 O3 on the carbon cycle. Trends in stratospheric O3 may also affect temperatures at the Earth's
6 surface, but that topic is beyond the scope of this assessment.
March 2011
10-6
DRAFT - DO NOT CITE OR QUOTE
-------
10.2.3.1. Trends in the Burden of Tropospheric Ozone
1 To first order, the effect of tropospheric O3 on climate is proportional to the change in O3
2 burden. The earth's surface temperatures are most sensitive to O3 perturbations in the mid to upper
3 troposphere. This section therefore focuses mainly on observed O3 trends in the free troposphere or
4 in regions far from O3 sources, where a change in O3 concentrations may indicate change throughout
5 the troposphere. Data from ozonesondes, mountaintops, and remote surface sites are discussed, as
6 well as satellite data.
Observed Trends in Ozone Since the Preindustrial Era
7 Measurements of O3 at two European mountain sites dating from the late 1800s to early 1900s
8 show values at about 10 ppb, about one-fifth the values observed today at similar sites (Marenco et
9 al., 1994, 047733; Pavelin et al., 1999, 087296). The accuracy of these early measurements is
10 questionable however, in part because they exhibit O3 concentrations equivalent to or only a couple
11 of parts per billion greater than those observed at nearby low-altitude sites during the same time
12 period (Mickley et al., 2001, 080134: Volz and Kley, 1988, 041650). A larger vertical gradient in
13 tropospheric O3 would be expected because of its stratospheric source and its longer lifetime aloft. In
14 another study, Staehelin et al. (1994, 055369) revisited observations made in the Swiss mountains
15 during the 1950s and found a doubling in O3 concentrations from that era to 1989-1991.
16 Routine observations of O3 in the troposphere began in the 1970s with the use of balloon-
17 borne ozonesondes, but even this record is sparse. Trends from ozonesondes have been highly
18 variable and dependent on region (Logan et al., 1999, 631175). Over most sites in the U.S.,
19 ozonesondes reveal little trend. Over Canada, observations show a decline in O3 between 1980 and
20 1990, then a rebound in the following decade (Tarasick et al., 2005, 631184). Ozonesondes over
21 Europe give a mixed picture, with Hohenpeissenberg in Germany showing declines through the
22 troposphere in recent decades, while Zugspitze, also in Germany, exhibiting small increases
23 (Oltmans et al., 2006, 180188). Over Japan, O3 in the lower troposphere increased about 0.2-
24 0.4 ppb/y during the 1990s (Naja and Akimoto, 2004, 631178).
25 Ground-based measurements in remote regions provide a record of background tropospheric
26 O3 extending as far back as the 1980s or, for ship measurements, the late 1970s. Springtime O3
27 observations from several mountain sites in the western U.S. show a positive trend of about of 0.5-
28 0.7 ppb/y since the 1980s (Cooper et al., 2010, 380093: Jaffe et al., 2003, 052229). Ship-borne O3
29 measurements for the time period 1977 to 2002 indicate increases of 0.1-0.7 ppb/y over the tropical
30 and South Atlantic, but no significant change over the North Atlantic (Lelieveld et al., 2004,
31 630578). The lack of trend for the North Atlantic would seem at odds with O3 observations at Mace
32 Head on the west coast of Ireland, which show a significant positive trend of about 0.5 ppb/y from
33 1987 to 2003 (Simmonds et al., 2004, 631182). Over Japan, O3 at a remote mountain site has
34 increased 1 ppb/y from 1998 to 2003 (Tanimoto, 2009, 620751). a rate more than double that
35 recorded by ozonesondes in the lower troposphere over Japan during the 1990s (Naja and Akimoto,
36 2004, 631178).
March 2011 10-7 DRAFT - DO NOT CITE OR QUOTE
-------
1 The satellite record is now approaching a length that can be useful for diagnosing trends in
2 total column amounts of tropospheric O3. In contrast to the surface data from ships, tropospheric O3
3 columns from the Total Ozone Mapping Spectrometer (TOMS) show no trend over the tropical
4 Atlantic for the period 1980-1990 (Thompson and Hudson, 1999, 631185). Over the Pacific, a
5 longer, 25 year record of TOMS data again reveals no trend over the tropics, but shows increases in
6 tropospheric column O3 of about 2-3 Dobson Units (DU [1 DU = 2.69 x 1016 molecules of O3/cm2])
7 at midlatitudes in both hemispheres (Ziemke et al., 2005, 631193): for comparison, the tropospheric
8 O3 burden averages about 34 DU.
9 Interpreting these recent trends in tropospheric O3 is challenging. The first difficulty is
10 reconciling apparently contradictory trends in the observations, e.g., over tropical oceans. A second
11 difficulty is that the O3 trends depend on several factors, not all of which can be well characterized.
12 These factors include (1) trends in emissions of O3 precursors, (2) variation in the stratospheric
13 source of O3, (3) changes in solar radiation resulting from stratospheric O3 depletion, and (4) trends
14 in tropospheric temperatures (Fusco and Logan, 2003, 051229). The positive trends in the western
15 U.S. and over Japan are consistent with the rapid increase in emissions of O3 precursors from
16 mainland Asia and transport of pollution across the Pacific (Cooper et al., 2010, 380093; Tanimoto,
17 2009, 620751). The satellite trends over the northern mid-latitudes are consistent with this picture as
18 well (Ziemke et al., 2005, 631193). Increases in tropospheric O3 in the Southern Hemisphere are also
19 likely due to increased anthropogenic NOX emissions, especially from biomass burning. The declines
20 in O3 over Europe can be at least partly explained by decreases in O3 precursor emissions there
21 (Jonson et al., 2005, 630282). though recent O3 depletion in the lower stratosphere may also
22 contribute to the decreases by reducing stratospheric input to the troposphere (Fusco and Logan,
23 2003, 051229).
Calculation of Ozone Trends for the Recent Past
24 Attempts to simulate trends in tropospheric O3 allow us to test current knowledge of O3
25 processes and to predict with greater confidence trends in future O3 concentrations. Time-dependent
26 emission inventories of O3 precursors have also been developed (e.g., Lamarque et al., 2010, 630289
27 for 1850-2000; Van Aardenne et al., 2001, 055564 for 1890-1990). These inventories allow for the
28 calculation of changing O3 burden over time.
29 One recent multi-model study calculated an increase in the O3 burden since preindustrial times
30 of 8-14 DU, or about 30-70% (Gauss et al., 2006, 630275). The large spread in modeled estimates
31 reveals our limited knowledge of processes in the pristine atmosphere. Models typically overestimate
32 the late nineteenth and early twentieth century observations available in surface air and at mountain
33 sites by 50-100% (Kiehl et al., 1999, 047917) (Lamarque et al., 2005, 630287: Mickley et al., 2001,
34 080134; Shindell et al., 2003, 057417). Reconciling the differences between models and
35 measurements will require more accurate simulation of the natural sources of O3 (Mickley et al.,
36 2001, 080134) and/or implementation of novel sinks such as bromine radicals, which may reduce
37 background O3 in the pristine atmosphere by as much as 30% (Parrella et al., In Press, 664506).
March 2011 10-8 DRAFT - DO NOT CITE OR QUOTE
-------
1 For the more recent past (since 1970), application of time-dependent emissions reveals an
2 equatorward shift in the distribution of tropospheric O3 in the Northern Hemisphere due to the
3 industrialization of societies at low-latitudes (Berntsen et al, 2000, 047916; Lamarque et al., 2005,
4 630287). By constraining a model with historical (1950s-2000) observations, Shindell et al. (2002,
5 080130) calculated a large increase of 8.2 DU in tropospheric O3 over polluted continental regions
6 since 1950. Their result appears consistent with the large change in tropospheric O3 since
7 preindustrial times implied by the observations from the late 1800s (Marenco et al., 1994, 047733;
8 Pavelin et al., 1999, 087296).
10.2.3.2. The Effect of Surface Albedo on Ozone Forcing
9 The Earth's surface albedo plays a role in O3 forcing. Through most of the troposphere,
10 absorption of incoming shortwave solar radiation by O3 is small relative to its absorption of outgoing
11 longwave terrestrial radiation. However, over surfaces characterized by high albedo (e.g., over snow,
12 ice, or desert sand), incoming radiation is more likely to be reflected than over darker surfaces, and
13 the probability that O3 will absorb shortwave solar energy is therefore larger. In other words, energy
14 that would otherwise return to space may instead be deposited in the atmosphere. Several studies
15 have shown that transport of O3 to the Arctic from mid-latitudes leads to radiative forcing estimates
16 greater than 1.0 W/m2 in the region, especially in summer (Liao et al., 2004, 057414; Mickley et al.,
17 1999, 047918; Shindell et al., 2006, 631181). Because the Arctic is especially sensitive to radiative
18 forcing through the ice-albedo feedback, the large contribution in the shortwave to the total radiative
19 forcing in the region may be important.
10.2.3.3. The Effect of Vertical Distribution on Ozone Forcing
20 In the absence of feedbacks, O3 increments added near the tropopause produce the largest
21 increases in surface temperature (Lacis et al., 1990, 037834; Wang et al., 1980, 674821). This is a
22 result of the colder temperature of the tropopause relative to the rest of the troposphere and
23 stratosphere. Since radiation emitted by the atmosphere is approximately proportional to the fourth
24 power of its temperature1, the colder the added O3 is relative to the earth's surface, the weaker the
25 radiation emitted and the greater the "trapping" of longwave radiation in the troposphere.
10.2.3.4. Feedback Factors that Alter the Climate Response to Changes in Ozone
Forcing
26 Estimates of radiative forcing provide a first-order assessment of the effect of tropospheric O3
27 on climate. In the real atmosphere, climate feedbacks and transport of heat alter the sensitivity of
28 Earth's surface temperature to addition of tropospheric O3. Assessment of the full climate response to
29 increases in tropospheric O3 requires use of a climate model to simulate these interactions.
1 As described by the Stefan-Boltzmann law, an ideal blackbody—which the atmosphere approximates--absorbs at all wavelengths and re-
radiates proportional to the fourth power of its temperature.
March 2011 10-9 DRAFT - DO NOT CITE OR QUOTE
-------
1 Due to its short lifetime, O3 is heterogeneously distributed through the troposphere. Sharp
2 horizontal gradients exist in the radiative forcing of O3, with the greatest radiative forcing since
3 preindustrial times occurring over the northern mid-latitudes (more on this in Section 10.2.5). If
4 climate feedbacks are particularly powerful, they may obscure or even erase the correlation between
5 regional radiative forcing and climate response (Boer and Yu, 2003, 630271; Harvey, 2004, 190598).
6 For example, several model studies have reported that the horizontal pattern of surface temperature
7 response from 2000-2100 trends in short-lived species (including O3) closely matches the pattern
8 from the trends in the long-lived greenhouse gases over the same time period (Levy H et al., 2008,
9 631174: Shindell et al., 2007, 521350: Shindell et al., 2008, 190393). This correspondence occurs
10 even though the patterns of radiative forcing for the short-lived and long-lived species differ
11 significantly. In a separate paper, Shindell (2007, 521350) found that Arctic temperatures are
12 especially sensitive to the mid-latitude radiative forcing from tropospheric O3.
13 Other studies have found that the signature of warming due to tropospheric O3 does show
14 some consistency with the O3 forcing. For example, Mickley et al. (2004, 057416) examined the
15 change in O3 since preindustrial times and found greater warming in the Northern Hemisphere than
16 in the Southern Hemisphere (+0.4°C versus +0.2°C), as well as higher surface temperatures
17 downwind of Europe and Asia and over the North American interior in summer. For an array of
18 short-lived species including O3, Shindell and Faluvegi (2009, 631180) found that radiative forcing
19 applied over northern mid-latitudes yield more localized responses due to local cloud, water vapor,
20 and albedo feedbacks than radiative forcing applied over the tropics.
21 Climate feedbacks can also alter the sensitivity of surface temperature to the vertical
22 distribution of tropospheric O3. The previous section (Section 10.2.3.3) described the greater impact
23 of O3 added to the upper troposphere (near the tropopause) on radiative forcing, relative to additions
24 in the mid- to lower troposphere. However, warming induced by increased O3 in the upper
25 troposphere could stabilize the atmosphere to some extent, limiting the transport of heat to the
26 Earth's surface and mitigating the impact of the added O3 on surface temperature (Christiansen,
27 1999, 047920: Joshi et al., 2003, 193752). Hansen et al. (1997, 043104) determined that allowing
28 cloud feedbacks in a climate model meant that O3 enhancements in the mid-troposphere had the
29 greatest effect on surface temperature.
30 Finally, climate feedbacks can amplify or diminish the climate response of one greenhouse gas
31 relative to another. For example, Mickley et al. (2004, 057416) found a greater temperature response
32 to CO2 forcing than to an O3 forcing of similar global mean magnitude, due in part to the relatively
33 weak ice-albedo feedback for O3. Since CO2 absorbs in the same bands as water vapor, CO2 forcing
34 saturates in the middle troposphere and is also shifted toward the drier poles. A poleward shift in
35 radiative forcing amplifies the ice-albedo feedback in the case of CO2, and the greater mid-
36 troposphere radiative forcing allows for greater surface temperature response, relative to that for O3.
March 2011 10-10 DRAFT-DO NOT CITE OR QUOTE
-------
10.2.3.5. Indirect Effects of Tropospheric Ozone on the Carbon Cycle
1 A proposed indirect effect of tropospheric O3 on climate involves the carbon cycle. By directly
2 damaging plant life in ways discussed in Chapter 9, increases in tropospheric O3 may depress the
3 land-carbon sink of CO2, leading to accumulation of CO2 in the atmosphere and ultimately warming
4 of the Earth's surface. Sitch et al. (2007, 093294) calculated that this indirect warming effect of O3
5 on climate has about the same magnitude as the O3 direct effect. Their results suggest a doubled
6 sensitivity of surface temperatures to O3 forcing, compared to current model estimates.
10.2.4. Competing Effects of Ozone Precursors on Climate
7 Changes in O3 precursors affect not just O3 concentrations, but also other species that have
8 importance to the radiative balance of the earth's climate system. For example, an increase in CO or
9 VOCs would lead to a decrease in hydroxyl (OH) concentrations. Since OH is a major sink of the
10 greenhouse gas CH4, a decline in OH would lengthen the CH4 lifetime, enhance the CH4 burden, and
11 amplify surface warming. A rise in NOX emissions, on the other hand, could lead to an increase in
12 OH in certain locations, shortening the CH4 lifetime and leading to surface cooling (Fuglestvedt et
13 al.. 1999.047431).
14 Analyzing the net radiative forcing per unit emission for a suite of O3 precursors, Shindell and
15 Faluvegi (2009, 631180) calculated positive (+0.25 W/m2) radiative forcing from the increase in
16 anthropogenic emissions of CO and VOCs since preindustrial times, as well as for CH4 (+1 W/m2).
17 In contrast, they found negative (-0.29 W/m2) radiative forcing from anthropogenic emissions of
18 NOX due mainly to the link between NOX and CH4. Other studies have found a near cancellation of
19 the positive O3 forcing and the negative CH4 forcing that arise from an incremental change in
20 anthropogenic NOX emissions (Fiore et al., 2002, 051221: Fuglestvedt et al., 1999, 047431: Naik et
21 al., 2005, 193194). In addition, Wild et al. (2001, 193196) found that an increase in surface NO
22 emissions would lead to net cooling, while an increase in aircraft NO emissions would lead to net
23 warming.
24 These results point out the need for careful assessment of net radiative forcing involving
25 multiple pollutants in developing climate change policy (Unger et al., 2008, 631186). Naik et al.
26 (2005, 193194) has calculated that a carefully combined reduction of CO, VOCs, and NOX emissions
27 could lead to net cooling. In addition, several studies point to CH4 as an attractive target for
28 emissions control since CH4 is itself an important precursor of O3 (Fiore et al., 2002, 051221: West
29 et al., 2007, 622733). Fiore et al. (2002, 051221) found that reducing anthropogenic CH4 emissions
30 by 50% would lead to a global radiative cooling of -0.37 W/m2, mostly from CH4.
10.2.5. Calculating Radiative Forcing and Climate Response to Past Trends
in Tropospheric Ozone
31 The magnitude of the radiative forcing from the change in tropospheric O3 since the
32 preindustrial era is uncertain. This uncertainty derives in part from the scarcity of early
March 2011 10-11 DRAFT - DO NOT CITE OR QUOTE
-------
1 measurements and in part from our limited knowledge regarding processes in the natural
2 atmosphere. As noted previously, the IPCC AR4 reports a radiative forcing of 0.35 W/m2 from the
3 change in tropospheric O3 since 1750 (Forster et al., 2007, 092936). ranking it third in importance
4 among greenhouse gases after CO2 and CH4. The O3 forcing could, in fact, be as large as 0.7 W/m2,
5 if reconstructions of preindustrial and mid-20th century O3 based on the measurement record are
6 valid (Mickley et al., 2001, 080134: Shindell and Faluvegi, 2002, 080130). In any event, Unger et al.
7 (2010, 387104) showed that present-day O3 forcing can be attributed to emissions from across many
8 economic sectors, including on-road vehicles, household biofuel, power generation, and biomass
9 burning. As much as one-third of the radiative forcing from the 1890 to 1990 change in tropospheric
10 O3 could be due to increased biomass burning (Ito et al., 2007, 608803).
11 These calculated radiative forcing estimates can be compared to those obtained from satellite
12 data. Using data from TOMS, Worden et al. (2008, 631188) estimated a reduction in clear-sky
13 outgoing longwave radiation of 0.48 W/m2 by O3 in the upper troposphere over oceans in 2006. This
14 radiative forcing includes contributions from both anthropogenic and natural O3. Assuming that the
15 burden of O3 has roughly doubled since preindustrial times (Gauss et al., 2006, 630275). the total O3
16 forcing estimated with TOMS is consistent with that obtained from models estimating just the
17 anthropogenic contribution.
18 Calculation of the climate response to the O3 radiative forcing is challenging due to
19 complexity of feedbacks, as mentioned in Sections 10.2.2.3 and 10.2.3.4. In their model study,
20 Mickley et al. (2004, 057416) reported a global mean increase of 0.28°C since preindustrial times,
21 with values as large as 0.8°C in continental interiors. For the time period since 1870, Hansen et al.
22 (2005, 190596) estimated a much smaller increase in global mean surface temperature (0.11°C), but
23 they implemented 1880s anthropogenic emissions in their base simulation and also took into account
24 trends in both stratospheric and tropospheric O3; the modeled decline of lower stratospheric O3,
25 especially over polar regions, cooled surface temperatures in this study, counteracting the warming
26 effect of increasing tropospheric O3.
27 Figure 10-3 shows the Hansen et al. (2005, 190596) results as reported in Shindell et al. (2006,
28 631181). In that figure, summertime O3 has the largest radiative impact over the continental interiors
29 of the Northern Hemisphere. In winter, the impact of tropospheric O3 is greatest over the snow and
30 ice regions of the Arctic, where the probability of O3 absorption of shortwave radiation is high
31 (Section 10.2.3.2). Shindell et al. (2006, 631181) estimated that the change in tropospheric O3 over
32 the 20th century could have contributed about 0.3°C to annual mean Arctic warming and as much as
33 0.4-0.5°C during winter and spring. Over eastern China, Chang et al. (2009, 630273) calculated a
34 surface temperature increase of 0.4°C to the 1970-2000 change in tropospheric O3. It is not clear,
35 however, to what degree regional changes in O3 burden influenced this response, as opposed to more
36 global changes.
March 2011 10-12 DRAFT-DO NOT CITE OR QUOTE
-------
Annual surface air temperature
11
Annual radiative forcing
-1.1 -.9 -.7 -.5 -.3 -.1 .1 .3 .5 .7 .9 1.1
.3
Summer (JJA) surface air temperature .10 Winter (DJF) surface air temperature
-1.1 -.9 -.7 -.5 -.3 -.1 .1 .3 .5 .7 .9 1.5 -1.1 -£ -.7 -.5 -.3 -.1 .1 .3 .5 .7 .9 1.4
Source: Used with permission from American Geophysical Union, Shindell etal. (2006, 6311811
Figure 10-3. Ensemble average 1900-2000 surface temperature trends (°C per century) in
response to tropospheric ozone changes and the input radiative forcing (W/m2), as
computed by the NASA GISS chemistry-climate model.
Values are surface temperature trends for the annual average (top left), June-August (bottom
left), and December-February (bottom right) and annual average tropopause
instantaneous radiative forcing from 1880 to 1990 (top right). Temperature trends
greater than about 0.1°C are significant over the oceans, while values greater than
0.3°C are typically significant over land, except for northern middle and high
latitudes during winter where values in excess of about 0.5°C are significant. Values
in the top right corner give area-weighted global averages in the same units as the
plots.
10.2.6. Calculating the Radiative Forcing and Climate Response to Future
Trends in Tropospheric Ozone
1 Future trends in tropospheric O3 concentrations depend in large part on what pathways in
2 energy technology the world's societies will follow in coming decades. The trends in O3 will also
3 depend on the changes in a suite of climate-sensitive factors, such as the water vapor content of the
4 atmosphere. This section describes the following issues: (1) projected trends in the anthropogenic
5 emissions of O3 precursors; (2) the effects of these emissions on the tropospheric O3 burden; (3) the
March 2011
10-13
DRAFT - DO NOT CITE OR QUOTE
-------
1 effects of changing climate on tropospheric O3; and (4) radiative forcing and climate response to 21st
2 century trends in tropospheric O3.
10.2.6.1. Emissions of Anthropogenic Ozone Precursors across the 21st Century
3 The IPCC SRES effort devised scenarios for short-lived O3 precursors as well as the well-
4 mixed greenhouse gases including NOX, CO, and VOCs (IPCC, 2000, 080704). Using the IMAGE
5 socioeconomic model, Streets et al. (2004, 190423) provided speciation for NOX and VOCs and
6 allocated the trends in emissions over 17 regions and 8 economic sectors for the 2000-2050 time
7 period. The worst-case IPCC scenario, A2, features continued dependence on fossil fuels, rapid
8 population growth, and little sharing of technology between developed and developing nations. By
9 2100 in this scenario, global NOX emissions increase by a factor of 3.4 and CO emissions and CH4
10 by ~2.7, relative to 2000 (IPCC, 2000, 080704). Most of these increases in emissions occur over
11 developing countries. For example over Asia, NOX emissions in the A2 scenario increase by more
12 than a factor of 4 by 2100. The more moderate A1B scenario has global NOX and CO emissions
13 increasing by 25% and 90%, respectively by 2100, but global CH4 emissions decreasing by 10%. In
14 the Bl scenario, with its emphasis on clean and efficient technologies, global emissions of NOX, CO,
15 and CH4 all decrease by 2100 relative to the present day (-40%, -60%, and -30%, respectively).
16 Other emissions scenarios have been recently developed to describe trends in the short-term
17 (up to 2030). The Current Legislation (CLE) scenario provides trends consistent with existing air
18 quality regulations; the Maximum Feasible Reduction (MFR) scenario seeks to reduce emissions of
19 O3 precursors to the maximum extent possible. Emission source changes relative to the present day
20 for CLE, MFR, and A2 are given in Stevenson et al. (2006, 089222).
21 For the Fifth Assessment Report (IPCC AR5), a new set of scenarios has been developed: the
22 Representative Concentration Pathways (RCPs) (Moss et al., 2010, 664501). The RCPs will explore
23 for the first time approaches to climate change mitigation. The scenarios are designed to achieve
24 radiative forcing targets of 2.6, 4.5, 6.0 and 8.5 W/m2 by 2100, and have been designated RCP 2.6,
25 RCP 4.5, RCP 6.0, and RCP 8.5 (RCP 2.6 is also known as RCP3-PD.) In all scenarios, global
26 anthropogenic NOX emissions decline 30-50% during the 21st century, though RCP 8.5 shows a peak
27 during the 2020s at a value -15% greater than that of 2000. Global anthropogenic VOC and CO
28 emissions are relatively flat during the 2000-2050 time range, and then decline by 30-50% by the
29 end of the century. For CH4, global mean emission trends for the four RCP scenarios differ
30 significantly across the 21st century, with RCP 8.5 showing atripling of emissions by 2100, and
31 RCP 2.6 showing the emissions cut by half in this time range. All these global trends, however,
32 contain some regional variation. For example, Asian emissions of both NOX and VOCs show
33 significant increases in the near term (2030s to 2050s). Plots of the RCP trends can be found at
34 http://iiasa.ac.at/web-apps/tnt/RcpDb/dsd?Action=htmlpage&page=about (RCP, 2009, 677552).
March 2011 10-14 DRAFT-DO NOT CITE OR QUOTE
-------
10.2.6.2. Impact of 21st Century Trends in Emissions on Tropospheric Ozone
1 Due to its short lifetime, tropospheric O3 will respond readily to changes in anthropogenic
2 emissions of its precursors. As shown in Table 10-1, a recent multi-model study found increases in
3 the tropospheric O3 burden of 15% and 6% for the IPCC A2 and CLE scenarios respectively for the
4 2000-2030 time period, and a decrease for the MFR scenario of 5% (Stevenson et al, 2006, 089222).
5 These results indicate that the growth in tropospheric O3 between 2000 and 2030 could be reduced or
6 even reversed, depending on emission controls. For the relatively moderate A1B emissions scenario
7 over the 2000-2050 time period, Wu et al. (2008, 190039) calculated a change in O3 burden of about
8 20%. Looking further into the 21st century, Gauss et al. (2003, 094204) reported O3 burden changes
9 of 30-60% in response to application of the A2p anthropogenic emissions over the 2000-2100 time
10 period (the A2p scenario was a preliminary version of the A2 scenario). Using the A2 scenario for
11 the same 100 year time period, Pyle et al. (2007, 630284) projected a 50% increase in the O3 burden,
12 consistent with Gauss et al. (2003, 094204) and with Liao et al. (2006, 664500) who calculated an O3
13 change of 60% for the same conditions. Given the large (+40 ppb) monthly mean increases in
14 surface O3 that the A2 or A2p scenarios would yield over Asia and elsewhere by the end of the 21st
15 century (Prather et al., 2003, 047879). these 100-yr projections of the O3 burden would lead to
16 extremely unhealthy air quality.
17 As noted above, the RCP scenarios of AR5 show long-term declines in the global mean
18 emissions of O3 precursors, with some regional increases in the near-term. As of this writing, no
19 model study has reported the response of the tropospheric O3 burden to any of the IPCC AR5
20 scenarios.
Table 10-1.2000-2030 changes in anthropogenic emissions, and CH4 and tropospheric ozone burdens,
and the associated tropospheric ozone forcing for three scenarios; values are ensemble
means
Scenario
Percent change i
Percent change i
Percent change i
Percent change i
Radiative forcing
n NOX emissions
n CO emissions
n CH4 burden
n tropospheric 03 burden
due to 03 change3 (W/m2)
IPCCA2
+96%
+62%
+23%
+ 15%
0.3
Current Legislation
(CLE)
+ 18%
-16%
+ 19%
+6%
0.18
Maximum Feasible Reduction
(MFR)
-53%
-53%
0%
-5%
-0.05
'Includes radiative forcing due to corresponding CH4 change.
Source: Adapted from Stevenson et al. (2006, 089222V
March 2011 10-15 DRAFT-DO NOT CITE OR QUOTE
-------
10.2.6.3. Impact of 21st Century Climate on Tropospheric Ozone
1 For the time period from the 1800s to the present-day, most of the increase in the burden of
2 tropospheric O3 can be traced to changing emissions. Model studies show that climate change so far
3 has likely had little impact on the tropospheric O3 (e.g., Grenfell et al, 2001, 664496). In the future,
4 however, climate change is expected to bring large changes in a suite of variables that could affect
5 O3 production, loss, and transport. For example, increased water vapor in a warming atmosphere is
6 expected to enhance OH concentrations, which in remote, NOx-poor regions will accelerate O3 loss
7 rates (Johnson et al., 1999, 052390).
8 In the 2050s A1B climate, Wu et al. (2008, 629684) calculated a 5 ppb decrease in surface O3
9 over oceans. A rise in temperatures will also likely promote emissions of isoprene, an important
10 biogenic precursor of O3. Model studies have calculated 21st-century increases in isoprene emissions
11 ranging from 25-50%, depending on climate scenario and time horizon (Wu et al., 2008, 190039. and
12 references therein). These studies however did not take into account the effects of changing climate
13 and CO2 burden on vegetation extent, which could have large consequences for biogenic emissions
14 (Heald et al., 2008, 191617; Sanderson et al., 2003, 630286). In any event, enhanced isoprene
15 emissions will increase O3 concentrations in VOC-limited regions, but decrease O3 in NOx-limited
16 regions (Pyle et al., 2007, 630284: Sanderson et al., 2003, 630286: Wu et al., 2008, 190039).
17 Convection frequencies and lightning flash rates will also likely change in a changing climate, with
18 consequences for lightning NOX emissions and O3 concentrations in the upper troposphere (Price and
19 Rind, 1994, 630283: Sinha and Toumi, 1997, 047932). While Wu et al. (2008, 190039) calculated an
20 increase in lightning NOX by 2050 due to enhanced deep convection, Jacobson and Streets (2009,
21 630281) projected a decrease in lightning NOX due to a declining cloud ice in their future
22 atmosphere. Finally, changes in transport processes will almost certainly accompany global climate
23 change. For the 2050 A1B climate, Wu et al. (2008, 629684) showed that flattening of the meridional
24 temperature gradient in a warming world would lead to slower intercontinental transport of
25 tropospheric O3. For the A2 climate in 2100, Zeng and Pyle (2003, 047492) projected an 80%
26 increase in the flux of stratospheric O3 into the troposphere, relative to the present-day.
27 Taken together, these climate-driven processes could have significant effects on the burden
28 and distribution of tropospheric O3. As shown in Wu et al. (2008, 629684). model projections of the
29 change in O3 burden due solely to future climate change range from -12% to +3%, depending on the
30 model, scenario, and time horizon.
10.2.6.4. Radiative Forcing and Climate Response from 21st Century Trends in
Tropospheric Ozone
31 In the near term (2000-2030), Stevenson et al. (2006, 089222) estimated an O3 forcing of near
32 zero for MFR, 0.18 W/m2 for CLE, and +0.3 W/m2 for the A2 scenario (Table 10-1). Menon et al.
33 (2008, 613861). following the moderate A1B scenario, calculated a radiative forcing of 0.12 W/m2
34 from the 2000-2030 change in tropospheric O3, about the same as that derived by Stevenson et al.
March 2011 10-16 DRAFT-DO NOT CITE OR QUOTE
-------
1 (2006, 089222) for the CLE scenario. Over the longer term (2000 to 2100) for the A1B scenario,
2 Gauss et al. (2003, 094204) reported large positive radiative forcing (0.40 to 0.78 W/m2) due to the
3 change in tropospheric O3, as shown in Figure 10-4. Normalized radiative forcing for these model
4 calculations fell within a relatively narrow range, 0.032 to 0.040 W/m2/DU, indicating that the
5 largest uncertainty lies in the model-calculated changes in O3 burden. Applying the A2 scenario,
6 Chen et al. (2007, 630274) estimated a global mean radiative forcing of 0.65 W/m2 from
7 tropospheric O3 by 2100, consistent with the Gauss et al. (2003, 094204) results. These studies took
8 into account only the impact of changing emissions on tropospheric O3. In their calculations of the
9 2000-2100 radiative forcing from O3 in the A2 scenario, Liao et al. (2006, 664500) found that
10 inclusion of climate effects on tropospheric O3 reduced their radiative forcing estimate by 20%.
11
12
0.90
0.50
O./O
0.60
g1 0.50
'u
^
; 0.40
>
^ 0.30
0.20
0.10
0.00
sw t
LW
Net t
SVV H s LW t + sB iNet t + s
i\\
ULAQ UI01 UCI IASB KNMI UCAM MOZ1 MOZ2 HCIS LJKN/0 UI02
Source: Used with permission from American Geophysical Union, Gauss et al. (2003, 0942041
Figure 10-4. Global mean radiative forcing estimates calculated by a set of models for the 2000-
2100 change in tropospheric ozone. Shown are the components of radiative forcing
in W/m2. SW = shortwave component; LW = longwave component; Net = total
forcing; t = tropospheric ozone changes only; and t + s = both tropospheric and
stratospheric changes.
Several studies have included tropospheric O3 in their investigations of the response in the
future atmosphere to a suite of short-lived species (e.g., Levy H et al., 2008, 631174; Shindell et al.,
March 2011
10-17
DRAFT - DO NOT CITE OR QUOTE
-------
1 2007, 521350; Shindell et al., 2008, 190393). Few studies, however, have calculated the climate
2 response to changes in tropospheric O3 alone in the future atmosphere. For the A2 atmosphere, Chen
3 et al. (2007, 630274) estimated a global mean surface temperature increase of+0.34°C by 2100 in
4 response to the change in O3. The largest temperature increases in this study, as much as 5°C,
5 occurred over the populous regions of Asia and the Middle East and downwind of biomass burning
6 regions in South Africa and South America.
10.2.7. Summary of the Effects of Tropospheric Ozone on Climate
7 Tropospheric O3 is a major greenhouse gas, third in importance after CO2 and CH4. While the
8 developed world has successfully reduced emissions of O3 precursors in recent decades, many
9 developing countries have experienced large increases in precursor emissions and these trends are
10 expected to continue, at least in the near term. Projections of radiative forcing due to changing O3
11 over the 21st century show wide variation, due in large part to the uncertainty of future emissions of
12 source gases. In the near-term (2000-2030), projections of O3 radiative forcing range from near zero
13 to +0.3 W/m2, depending on the emissions scenario (Stevenson et al., 2006, 089222). Reduction of
14 tropospheric O3 concentrations could therefore provide an important means to slow climate change
15 in addition to the added benefit improving surface air quality.
16 It is clear that increases in tropospheric O3 lead to warming. However the precursors of O3 also
17 have competing effects on the greenhouse gas CH4, complicating emissions reduction strategies. A
18 decrease in CO or VOC emissions would enhance OH concentrations, shortening the lifetime of
19 CH4, while a decrease in NOX emissions could depress OH concentrations in certain regions and
20 lengthen the CH4 lifetime. Recent research, however, has shown that a carefully combined reduction
21 of CO, VOCs, and NOX emissions could lead to net cooling (Naik et al., 2005, 193194). In addition,
22 abatement of CH4 emissions would provide a straightforward means to address climate change since
23 CH4 is itself an important precursor of background O3 (Fiore et al., 2002, 051221; West et al., 2006,
24 196558: West et al., 2007, 622733).
25 Important uncertainties remain regarding the impact of O3 on future climate change. To
26 address these uncertainties, further research is needed to: (1) enhance our knowledge of the natural
27 atmosphere; (2) interpret observed trends of O3 in the free troposphere and remote regions; (3)
28 understand the relationship between regional O3 forcing and regional climate change; and (4)
29 determine the optimal mix of emissions reductions that would act to limit future climate change.
10.3. UV-B Related Effects and Tropospheric Ozone
10.3.1. Background
30 Ultraviolet (UV) radiation emitted from the Sun contains sufficient energy when it reaches the
31 Earth to break (photolyze) chemical bonds in molecules, thereby leading to damaging effects on
32 living organisms and materials. Atmospheric O3 plays a crucial role in reducing exposure to solar
March 2011 10-18 DRAFT-DO NOT CITE OR QUOTE
-------
1 UV radiation at the Earth's surface. Stratospheric O3 is responsible for the majority of this shielding
2 effect, as approximately 90% of total atmospheric O3 is located there over mid-latitudes (Crist et al.,
3 1994, 668881; Kar et al., 2010, 670423). Investigation of the supplemental shielding of UV radiation
4 provided by tropospheric O3 is important for quantifying UV exposure and the incidence of related
5 human health effects, ecosystem effects, and materials damage. The role of tropospheric O3 in
6 shielding of UV radiation is discussed in this section.
10.3.2. Physics of UV Radiation and Flux
7 Solar UV radiation is subdivided into classes based on wavelength: UV-A refers to wavelengths from
8 400-315 nm; UV-B from 315-280 nm; and UV-C from 280-100 nm. Since the energy possessed by a
9 photon is inversely proportional to its wavelength, UV-A radiation is the least energetic and UV-C is
10 the most energetic, with UV-B falling in-between. The wavelength determines how the photons
11 interact with the complex mixture of gases, clouds and particles present in the atmosphere (see
12 Figure 10-5). UV-A radiation can be scattered but is not absorbed to any meaningful degree by
13 atmospheric gases including O3. UV-B radiation is absorbed and scattered in part within the
14 atmosphere. UV-C is almost entirely blocked by the Earth's upper atmosphere, where it participates
15 in photoionization and photodissociation processes. Since UV-A is less energetic and does not
16 interact with O3 and UV-C is almost entirely blocked by stratospheric O3, UV-B is the most
17 important band of UV radiation to consider in relation to tropospheric O3 shielding.
March 2011 10-19 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Backscattered
Radiation
Incident Solar UV Radiation
Stratospheric O3
Source: 2006 03AQCD (2006,
Figure 10-5. Diagram of the factors that determine human exposure to ultraviolet radiation.
Solar flux has a temporal dependence, while radiative scattering and absorption have strong
wavelength, path length, and/or particle concentration dependencies. These combine to create
nonlinear effects on UV flux at the Earth's surface. Thus, careful quantification of atmospheric
absorbers and scatterers, along with a well-resolved description of the physics of these interactions,
is necessary for predicting the impact of ground-level O3 on UV flux. Chapter 10 of the 2006 O3
AQCD (U.S. EPA, 2006, 088089) describes in detail several key factors that influence the
spatiotemporal distribution of ground-level UV radiation flux, including: (1) long-term solar activity
including sunspot cycle; (2) solar rotation; (3) the position of the Earth in its orbit around the sun; (4)
atmospheric absorption and scattering of UV radiation by gas molecules and aerosol particles; (5)
absorption and scattering by stratospheric and tropospheric clouds; and (6) surface albedo. The
efficiencies of absorption and scattering are highly dependent on the concentration of the scattering
medium, particle size (for aerosols and clouds), and the altitude at which these processes are
occurring. These properties are sensitive to meteorology, which introduces additional elements of
temporal dependency in ground-level UV radiation flux. As seen in data collected by the Global
Ozone Monitoring by Occultation of Stars (GOMOS) instrument onboard the European Space
Agency's ENVISAT satellite (Figure 10-6), atmospheric O3 density undergoes wide natural variation
on relatively short timescales, particularly at mid-latitudes (Kyrola et al., 2010, 667819).
March 2011
10-20
DRAFT - DO NOT CITE OR QUOTE
-------
40
1
2
3
4
5
6
7
03
04
05 06
Time (years)
Source: Used with permission from Copernicus Publications, Kyrola et al., (2010, 6678191.
Figure 10-6. Monthly stratospheric ozone number density (scaled by 1 x 1012 molecules/cm3), in
3 latitude belts as a function of time (August 2002 - December 2008) and altitude
(15-40 km) from the Global Ozone Monitoring by Occultation of Stars (GOMOS)
instrument onboard the European Space Agency's ENVISAT satellite.
Latitude belts: 30°N-50°N (top), 10°S-10°N (middle), 30°S-50°S (bottom). White space in the
panels means that there are not enough data available.
The lower atmospheric pressure in the stratosphere means fewer gas molecules are present that
can absorb or scatter radiation. Stratospheric clouds and aerosols are also thinner and more dispersed
than those in the troposphere. In the language of the radiative transfer literature, these conditions
make the stratosphere a "single scattering" regime for UV radiation. The troposphere, due to its high
gas and particle concentrations is referred to as a "multiple scattering" regime. In practical terms,
UV radiation traverses the stratosphere with a substantially lower probability of encountering a gas
molecule, cloud, or aerosol particle than it would in the troposphere. The multiple scattering of UV
radiation in the troposphere accounts for the "disproportionate" role that tropospheric O3 is said to
play in absorbing UV radiation versus stratospheric O3 on a molecule per molecule basis (Balis et al.,
March 2011
10-21
DRAFT - DO NOT CITE OR QUOTE
-------
1 2002, 055023: Bruhl and Crutzen, 1989, 012518: Crist et al., 1994, 668881: Zerefos et al, 2002,
2 055169V
3 Latitude and altitude are primary variables in defining UV-B flux at the Earth's surface,
4 immediately followed in importance by clouds, surface albedo, PM concentration and composition,
5 and then by gas phase pollution. Of all these variables, only latitude and altitude can be defined with
6 small uncertainty in any effort to develop a UV climatology for use in a public health benefits
7 analysis relevant to the areas not presently attaining the NAAQS for O3. Cloud cover, and its effect
8 on surface UV flux, continues to be extremely difficult to define and predict. Particulate matter and
9 gas-phase tropospheric pollutants are subject to similarly high degrees of uncertainty in predicting
10 their relative concentration distributions, but recent advancements have been made (e.g., Bais et al.,
11 2005, 669135: Bergstrom et al., 2004, 669158: Goering et al., 2005, 669164). Land cover and,
12 consequently, surface albedo is highly variable at the geographic scales relevant to NAAQS
13 attainment.
14 The 2006 WMO assessment (WMO, 2006, 669178) reported that global average total column
15 O3 had declined by 3.5% from pre-1980 concentrations due to the presence of anthropogenic O3-
16 depleting substances in the atmosphere. In the period 2002-2005, no additional declines were found
17 in the global average due to bans on and reduced emissions of O3-depleting substances. The report
18 found that O3 depletion has a strong latitude and seasonal dependence: total column O3 declined by
19 ~3% in the Northern Hemisphere, declined by ~6% in the Southern Hemisphere, and remained
20 essentially unchanged over the tropics relative to pre-1980 total column O3 abundances. Polar
21 stratospheric O3 depletion is more complex and exhibits large interannual variations driven by
22 changes in meteorology.
10.3.3. Human Exposure and Susceptibility to Ultraviolet Radiation
23 The factors that potentially influence UV radiation exposure were discussed in detail in
24 Chapter 10 of the 2006 O3 AQCD (U.S. EPA, 2006, 088089) and are summarized here. These factors
25 included outdoor activity, occupation, age, gender, geography, and protective behavior. Outdoor
26 activity and occupation both influenced the amount of time people spend outdoors during daylight
27 hours, the predominant factor for exposure to solar UV radiation. Participation in outdoor sports
28 (e.g., basketball, soccer, golf, swimming, cycling) significantly increased UV radiation exposure
29 (Moehrle, 2001, 057502: Moehrle et al., 2000, 057503: Thieden et al., 2004, 057558: Thieden et al.,
30 2004, 057557). Occupations that substantially increased exposure to UV radiation included farming
31 (Airey et al., 1997, 057458: Schenker et al., 2002, 057522). fishing (Rosenthal et al., 1988, 057513).
32 landscaping (Rosenthal et al., 1988, 057513). construction (Gies and Wright, 2003, 057477).
33 physical education (Vishvakarman et al., 2001, 057614). mail delivery (Vishvakarman et al., 2001,
34 057614). and various other occupations that require workers to spend the majority of their day
35 outdoors during peak UV radiation hours.
36 Age and gender were found to be factors that influence human exposure to UV radiation,
37 particularly by influencing other factors of exposure such as outdoor activity and risk behavior.
March 2011 10-22 DRAFT - DO NOT CITE OR QUOTE
-------
1 Studies indicated that females generally spent less time outdoors and, consequently, had lower UV
2 radiation exposure compared to males (Gies et al., 1998, 057478; Godar et al., 2001, 057165;
3 Shoveller et al., 1998, 057524). The lowest exposure to UV radiation among Americans in the Godar
4 et al. (2001, 057165) study was received in females during their child raising years (age
5 22-40 years); the highest exposure was observed in males aged 41-59 years. A similar Canadian
6 survey found that younger adult males had the greatest exposures to UV radiation (Shoveller et al.,
7 1998.057524).
8 Geography influences the degree of solar UV flux to the surface, and hence exposure to UV
9 radiation. In the U.S. study by Godar et al. (2001, 057165). northerners and southerners were found
10 to spend an equal amount of time outdoors; however, the higher solar flux at lower latitudes
11 significantly increased the annual UV radiation dose for southerners. The annual UV radiation doses
12 in southerners were 25 and 40% higher in females and males, respectively, compared to northerners.
13 Other studies also have shown that altitude and latitude influence personal exposure to UV radiation
14 (e.g., Kimlin et al., 1998, 057491; Rigel et al., 1999, 057511).
15 Protective behaviors such as using sunscreen (e.g., Nole and Johnson, 2004, 057505). wearing
O V O ? " " /" O
16 protective clothing (e.g., Rosenthal et al., 1988, 057513). and spending time in shaded areas (e.g.,
17 Moise et al., 1999, 057504) were shown to reduce exposure to UV radiation. In one study, the use of
18 sunscreen was associated with extended intentional UV radiation exposure (Autier et al., 1999,
19 057459); however, a follow-up study indicated that sunscreen use increased duration of exposures to
20 doses of UV radiation that were below the threshold level for erythema (Autier et al., 2000, 057069).
21 Given these and other factors that potentially influence UV radiation exposure, the 2006 O3
22 AQCD (U.S. EPA, 2006, 088089) listed the following subpopulations potentially at risk for higher
23 exposures to UV radiation:
24 • Individuals who engage in high-risk behavior (e.g., sunbathing);
25 • Individuals who participate in outdoor sports and activities;
26 • Individuals who work outdoors with inadequate shade (e.g., farmers, construction
27 workers, etc.); and
28 • Individuals living in geographic areas with higher solar flux including lower latitudes
29 (e.g., Honolulu, HI) and higher altitudes (e.g., Denver, CO).
30 The risks associated with all these factors are, of course, highly dependent on season and region
31 (Sliney and Wengraitis, 2006, 651896).
March 2011 10-23 DRAFT - DO NOT CITE OR QUOTE
-------
10.3.4. Human Health Effects due to UV-B Radiation
1 Chapter 10 of the 2006 O3 AQCD (U.S. EPA, 2006, 088089) covered in detail the human
2 health effects associated with solar UV-B radiation exposure. These effects include erythema, skin
3 cancer, ocular damage, and immune system suppression. These adverse effects, along with protective
4 effects of UV radiation through increased production of vitamin D are summarized in this section.
5 For additional details, the reader is referred to Chapter 10 of the 2006 O3 AQCD (U.S. EPA, 2006,
6 088089) and references therein.
7 The most conspicuous and well-recognized acute response to UV radiation is erythema, or the
8 reddening of the skin. Erythema is likely caused by direct damage to DNA by UV radiation
9 (Matsumura and Ananthaswamy, 2004, 057184). Many studies discussed in the 2006 O3 AQCD
10 (U.S. EPA, 2006, 088089) found skin type to be a significant risk factor for erythema. Additional
11 risk factors include atopic dermatitis (ten Berge et al, 2009, 651897).
12 Skin cancer is another prevalent health effect associated with UV radiation. Exposure to UV
13 radiation is considered to be a major risk factor for all forms of skin cancer (Diepgen and Mahler,
14 2002, 093593; Gloster and Brodland, 1996, 057479). Ultraviolet radiation is especially effective in
15 inducing genetic mutations and acts as both a tumor initiator and promoter. Keratinocytes have
16 evolved DNA repair mechanisms to correct the damage induced by UV; however, mutations can
17 occur, leading to skin cancers that are appearing with increasing frequency (Hildesheim and Fornace,
18 2004, 057168). The relationship between skin cancer and chronic exposure to UV radiation is further
19 explored in Chapter 10 of the 2006 O3 AQCD (U.S. EPA, 2006, 088089).
20 Ocular damage from UV radiation exposure includes effects on the cornea, lens, iris, and
21 associated epithelial and conjunctival tissues. The region of the eye effected by exposure to UV
22 radiation depends on the wavelength of the incident UV radiation. Depending on wavelength,
23 common health effects associated with UV radiation include photokeratitis (snow blindness; short
24 wavelengths) and cataracts (opacity of the lens; long wavelengths).
25 Experimental studies have suggested that exposure to UV radiation may suppress local and
26 systemic immune responses to a variety of antigens (Clydesdale et al., 2001, 057105; Garssen and
27 Van Loveren, 2001, 057161; Selgrade et al., 1997, 036165). In rodent models, these effects have
28 been shown to worsen the course and outcome of some infectious diseases and cancers (Granstein
29 and Matsui, 2004, 057747; Norval et al., 1999, 036089). Results from human clinical studies suggest
30 that immune suppression induced by UV radiation may be a risk factor contributing to skin cancer
31 induction (Caforio et al., 2000, 080058; Lindelof et al., 2000, 080084; Ullrich, 2005, 095635). There
32 is also evidence that UV radiation has indirect involvement in viral oncogenesis through the human
33 papillomavirus (Pfister, 2003, 057515). dermatomyositis (Okada et al., 2003, 057769). human
34 immunodeficiency virus (Breuer-McHam et al., 2001, 057736) and other forms of
35 immunosuppression (Selgrade M-JK; Smith et al., 2001, 057454).
36 A potential health benefit of increased UV-B exposure relates to the production of vitamin D
37 in humans. Most humans depend on sun exposure to satisfy their requirements for vitamin D
March 2011 10-24 DRAFT - DO NOT CITE OR QUOTE
-------
1 (Holick, 2004, 057691). Vitamin D deficiency can cause metabolic bone disease among children and
2 adults, and also may increase the risk of many common chronic diseases, including type I diabetes
3 mellitus and rheumatoid arthritis (Holick, 2004, 057691). Substantial in vitro and toxicological
4 evidence also support a role for vitamin D activity against the incidence or progression of various
5 forms of cancer (Freedman et al., 2002, 035530: Garland et al, 1990, 025242: Giovannucci, 2005,
6 074099: Gorham et al., 1990, 000682: Grant, 2002, 034981: Grant, 2002, 025244: Grant and
7 Garland, 2004, 075093: Hanchette and Schwartz, 1992, 025257: Hughes et al., 2004, 074101: John
8 et al., 1999, 057489: John et al., 2005, 670422: Lefkowitz and Garland, 1994, 025253: Smedby et
9 al., 2005, 669175: Studzinski and Moore, 1995, 057554). In some studies, UV-B related production
10 of vitamin D had potential beneficial immunomodulatory effects on multiple sclerosis, insulin-
11 dependent diabetes mellitus, and rheumatoid arthritis (Ponsonby et al., (2002, 080061): Cantorna,
12 (2000, 080060)). More details on UV-B protective studies are provided in Chapter 10 of the 2006 O3
13 AQCD (U.S. EPA, 2006, 088089).
14 In establishing guidelines on limits of exposure to UV radiation, the International commission
15 on Non-Ionizing Radiation Protection (ICNIRP) agreed that some low-level exposure to UV
16 radiation has health benefits (ICNIRP, 2004, 057187). However, the adverse health effects of higher
17 UV exposures necessitated the development of exposure limits for UV radiation. The ICNIRP
18 recognized the challenge in establishing exposure limits that would achieve a realistic balance
19 between beneficial and adverse health effects. As concluded by ICNIRP (2004, 057187). "[t]he
20 present understanding of injury mechanisms and long-term effects of exposure to [UV radiation] is
21 incomplete, and awaits further research".
10.3.5. Ecosystem and Materials Damage Effects Due to UV-B Radiation
22 A 2009 progress report on the environmental effects of O3 depletion from the UNEP,
23 Environmental Effects Assessment Panel (UNEP, 2009, 669084) lists many ecosystem and materials
24 damage effects from UV-B radiation. An in-depth assessment of the global ecosystem and materials
25 damage effects from UV-B radiation per se is out of the scope of this assessment. However, a brief
26 summary of some mid-latitude effects is provided in this section to provide context for UV-B related
27 issues pertaining to tropospheric O3. The reader is referred to the UNEP report (UNEP, 2009,
28 669084) and references therein for further details. All of these UV-B related ecosystem and materials
29 effects can also be influenced by climate change through temperature and other meteorological
30 alterations, making quantifiable predictions of UV-B effects difficult.
31 Terrestrial ecosystem effects from increased UV-B radiation include reduced plant productivity
32 and plant cover, changes in biodiversity, susceptibility to infection, and increases in natural UV
33 protective responses. In general, however, these effects are small for moderate UV-B increases at
34 mid-latitudes. A field study on wheat in southern Chile found no substantial changes in crop yield
35 with moderate increases in UV-B radiation (Calderini et al., 2008, 668893). Similarly, field studies
36 on silver birch (Betula pendula) in Finland found no significant effects in photosynthetic function
37 with increases in UV-B radiation (Aphalo et al., 2009, 668923). Subtle, but important, changes in
March 2011 10-25 DRAFT - DO NOT CITE OR QUOTE
-------
1 habitat and biodiversity have also been linked to increases in UV-B radiation (Mazza et al, 2010,
2 668911: Obara et al., 2008, 668913: Wahl, 2008, 668921). Some plants have natural coping
3 mechanisms for dealing with changes in UV-B radiation (Brown and Jenkins, 2008, 668892: Favory
4 et al., 2009, 668897: loki et al., 2008, 668904: Jenkins, 2009, 668905). but these defenses may have
5 costs in terms of reduced growth (Clarke and Robinson, 2008, 668928: Phoenix et al., 2000, 668915:
6 Semerdjieva et al., 2003, 668919: Snell et al., 2009, 668920).
7 Aquatic ecosystem effects from increased UV-B radiation include sensitivity in growth,
8 immune response, and behavioral patterns of aquatic organisms. One study looking at
9 coccolithophores, an abundant phytoplankton group, found a 25% reduction in cellular growth with
10 UV-B exposure (Gao et al., 2009, 668899). Exposure to relevant levels of UV-B radiation has been
11 shown to modify immune response, blood chemistry, and behavior in certain species of fish (Holtby
12 and Bothwell, 2008, 668903: Jokinen et al., 2008, 668906: Markkula et al., 2009, 613291). Adverse
13 effects on growth and development from UV-B radiation have also been observed for amphibians,
14 sea urchins, mollusks, corals, and zooplankton (Croteau et al., 2008, 668894: Croteau et al., 2008,
15 603785: Garcia et al., 2009, 668935: Marquis and Miaud, 2008, 668908: Marquis et al., 2008,
16 668909: Oromi et al., 2008, 668914: Romansic et al., 2009, 515730).
17 Biogeochemical cycles, particularly the carbon cycle, can also be influenced by increased UV-
18 B radiation. A study on high latitude wetlands found UV-induced increases in CO2 uptake through
19 soil respiration (Haapala et al., 2009, 607260) while studies on arid terrestrial ecosystems found
20 evidence for UV-induced release of CO2 through photodegradation of above-ground plant litter
21 (Brandt et al., 2009, 668891: Caldwell et al., 2007, 668927: Henry et al., 2008, 668902: Zepp et al.,
22 2007, 668896). Changes in solar UV radiation may also have effects on carbon cycling and CO2
23 uptake in the oceans (Brewer and Peltzer, 2009, 669197: Fritz et al., 2008, 668898: Hader et al.,
24 2007, 668901: Meador et al., 2009, 668912: Zepp et al., 2008, 668922) as well as release of
25 dissolved organic matter from sediment and algae (Mayer et al., 2009, 668910: Riggsbee et al., 2008,
26 668917). Additional studies showing effects on these and additional biogeochemical cycles including
27 the water cycle and halocarbon cycle can be found in the UNEP report (UNEP, 2009, 669084) and
28 references therein.
29 Materials damage from increased UV-B radiation include UV-induced photodegradation of
30 wood (Kataoka et al., 2007, 670425) and plastics (Pickett et al., 2008, 668916). These studies and
31 others summarizing photo-resistant coatings and materials designed to reduce photodegradation of
32 materials are summarized in the UNEP report (UNEP, 2009, 669084) and references therein.
10.3.6. UV-B Related Effects Associated with Changes in Tropospheric
Ozone Concentrations
33 There are multiple complexities in attempting to quantify the relationship between changes in
34 tropospheric O3 concentrations and UV radiation exposure, as described above. Furthermore,
35 quantifying the relationship between UV radiation and health or welfare effects is complicated by the
36 uncertainties involved in the selection of an action spectrum and appropriate characterization of dose
March 2011 10-26 DRAFT - DO NOT CITE OR QUOTE
-------
1 (e.g., peak or cumulative levels of exposure, timing of exposures, etc.) The lack of published studies
2 that critically examine these issues together—that is the incremental health or welfare effects
3 attributable specifically to UV-B changes resulting from reductions in tropospheric O3
4 concentrations—reflects the significant challenges in this field.
5 As reported in the 2006 O3 AQCD (U.S. EPA, 2006, 088089). one analysis by Lutter and Wolz
6 (1997, 082672) attempted to estimate the effects of a nationwide 10 ppb reduction in seasonal
7 average tropospheric O3 on the incidence of nonmelanoma and melanoma skin cancers and cataracts
8 in humans. Their estimate, however, depended upon several simplifying assumptions, ranging from
9 an assumed generalized 10-ppb reduction in O3 column density, national annual average incidence
10 rates for the two types of skin cancer, and simple, linear biological amplification factors.
11 Specifically, the decrease of 10 ppbv in seasonally averaged O3 concentrations is likely an
12 overestimate since it doesn't account for the influence of background O3 coming from the global
13 accumulation or generation of regional chemistry (Adamowicz et al., 2004, 670421). Further, the
14 methodologies used in this analysis have ignored area-specific factors that are important in
15 estimating the extent to which small, variable changes in ground-level O3 mediate long-term
16 exposures to UV-B radiation.
17 A handful of studies have addressed the relationship between changes in tropospheric pollutant
18 concentrations and UV-B radiation exposure, providing some additional insight. A study by Palancar
19 and Toselli (2002, 057207) looked at changes in measured UV-B radiation in relation to ground-level
20 air pollutants during several air pollution episodes in Cordoba, Argentina. They found that changes
21 in aerosol concentrations explained the majority of UV-B radiation fluctuations, and that changes in
22 tropospheric O3 and SO2 had little effect. Repapis et al. (1998, 038015) performed a similar study on
23 UV-B exposures during high and low air pollution days in Athens, Greece. They found cloud cover
24 and aerosols to be the major factors in observed UV-B exposures reductions. Studies by Acosta and
25 Evans (2000, 670420) in Mexico City and Koronakis et al. (2002, 129938) in Athens, Greece both
26 found significant reductions in surface-level UV exposures during pollution episodes. Both these
27 studies include tropospheric O3 as a potential driver for the reductions, but neither study was able to
28 quantify the influence of individual atmospheric components involved in the observed attenuation in
29 UV-B radiation.
30 In the absence of studies specifically addressing UV-B related health effects from a reduction
31 in tropospheric O3, inferences were made in the 2006 O3 AQCD (U.S. EPA, 2006, 088089) on the
32 basis of studies focused on stratospheric O3 depletion. Studies included in that review examined the
33 potential effect of stratospheric O3 depletion on the risk of erythema (Longstreth et al., 1998,
34 001200). skin cancer (De Gruijl, 1995, 057471: Longstreth et al., 1995, 055174: Madronich and
35 De Gruijl, 1993, 055183: Slaper et al., 1996, 055128: Urbach, 1997, 086255). nonmelanoma skin
36 cancer (Longstreth et al., 1995, 055174: Slaper et al., 1996, 055128). and cataracts (Longstreth et al.,
37 1995, 055174). Note that several of the concerns expressed above in relation to the Lutter and Wolz
38 (1997, 082672) analysis are relevant to these analyses as well. Furthermore, these studies have a
39 high degree of uncertainty due to inadequate information on the action spectrum and dose-response
March 2011 10-27 DRAFT - DO NOT CITE OR QUOTE
-------
1 relationships. As a result, caution is advised when assessing and interpreting the quantitative results
2 of health risks due to stratospheric O3 depletion in the context of tropospheric O3 shielding.
3 Although the UV-B related health effects attributed to marginal reductions in tropospheric or
4 ground-level O3 that would result from attainment of the O3 NAAQS have not been directly
5 assessed, they would be expected to be small or nonexistent given the above findings and the fact
6 that tropospheric O3 makes up only -10% of the total atmospheric O3 column at mid-latitudes (Kar
7 et al., 2010, 670423). Furthermore, O3 present in the planetary boundary layer makes up only -10%
8 of tropospheric O3 (Thompson et al., 2007, 090796) and the NAAQS has only a fractional influence
9 on those ground-level O3 concentrations (i.e., it is not the intent of the NAAQS to entirely eliminate
10 ground-level O3). The net result is a very small influence on total column O3 through attainment of
11 the O3 standard. In addition, the health benefits of UV-B in the production of vitamin D suggests that
12 increased risks of human disease due to a slight excess in UV-B radiation exposure may be offset by
13 the benefits of enhanced vitamin D production. However, as with other impacts of UV-B on human
14 health, this beneficial effect of UV-B has not been studied in sufficient detail to allow for a credible
15 health benefits assessment. Hence, the above mentioned health and welfare effects associated with
16 UV-B exposures resulting from changes in ground-level O3 concentrations would likely be small or
17 nonexistent based on current information.
18 More reasonable estimates of the human health impacts of enhanced UV-B penetration
19 following reduced ground-level O3 concentrations require both (a) a solid understanding of the
20 multiple factors that define the extent of human exposure to UV-B, and (b) well-defined and
21 quantifiable links between human disease and UV-B exposure. Within the uncertain context of
22 presently available information on UV-B surface fluxes, a risk assessment of UV-B-related health
23 effects would need to factor in human habits (e.g., daily activities, recreation, dress, and skin care) in
24 order to adequately estimate UV-B exposure levels. Little is known about the impact of variability in
25 these human factors on individual exposure to UV radiation. Furthermore, detailed information does
26 not exist regarding the relevant type (e.g., peak or cumulative) and time period (e.g., childhood,
27 lifetime, or current) of exposure, wavelength dependency of biological responses, and inter-
28 individual variability in UV resistance. In conclusion, the effect of changes in surface-level O3
29 concentrations on UV-induced health outcomes cannot yet be critically assessed within reasonable
30 uncertainty. The reader is referred to the U.S. EPA 2002 Final Response to Court Remand (2003,
31 015702) for detailed discussions of the data and scientific issues associated with the determination of
32 public health benefits resulting from the attenuation of UV-B by surface-level O3.
March 2011 10-28 DRAFT - DO NOT CITE OR QUOTE
-------
References
A list of all references considered for inclusion in this chapter can be found at
http://hero.epa. gov/index.cfm?action=search.do&submit=Search&group_id=478
Acosta, L. R.; Evans, W. F. J. (2000). Design of the Mexico City UV monitoring network: UV-B measurements at ground
level in the urban environment. J Geophys Res, 105: 5017-5026. http://dx.doi.org/10.1029/1999JD900250 670420
Adamowicz, V; Dales, R.; Hale, B. A.; Hrudey, S. E.; Krupnick, A.; Lippman, M.; McConnell, J.; Renzi, P. (2004). Report
of an expert panel to review the socio-economic models and related components supporting the development of
Canada-Wide Standards (CWS) for particulate matter (PM) and ozone to the Royal Society of Canada. J Toxicol
Environ Health B Crit Rev, 7: 147-266. http://dx.doi.org/10.1080/10937400490253238 670421
Airey, D. K.; Wong, J. C. F.; Fleming, R. A.; Meldrum, L. R. (1997). An estimate of the UV-B exposure for outdoor
workers during a south-east Queensland summer. Health Phys, 72: 544-549. 057458
Aphalo, P. J.; Vapaavuori, E. M.; de la Rosa, T. M.; Lehto, T. (2009). Does supplemental UV-B radiation affect gas
exchange and RuBisCO activity of Betula pendula Roth, seedlings grown in forest soil under greenhouse
conditions? Plant Ecol Divers, 2: 37-43. http://dx.doi.org/10.1080/17550870902780299 668923
Arrhenius, S. (1896). On the influence of carbonic acid in the air upon the temperature of the ground. Philos Mag, 41: 237-
276. http://www.rsc.org/images/Arrheniusl896_tcml8-173546.pdf 043125
Autier, P.; Dore, J. F.; Negrier, S.; Lienard, D.; Panizzon, R.; Lejeune, F. J.; Guggisberg, D.; Eggermont, A. M. (1999).
Sunscreen use and duration of sun exposure: a double-blind, randomized trial. J Natl Cancer Inst, 91: 1304-1309.
057459
Autier, P.; Dore, J.-F.; Reis, A. C.; Grivegnee, A.; Ollivaud, L.; Truchetet, F.; Chamoun, E.; Rotmensz, N.; Severi, G;
Cesarini, J. P.; EORTC Melanoma Co-operative Group (2000). Sunscreen use and intentional exposure to
ultraviolet A and B radiation: a double blind randomized trial using personal dosimeters. Br J Cancer, 83: 1243-
1248. 057069
Bais, A. F.; Kazantzidis, A.; Kazadzis, S.; Balis, D. S.; Zerefos, C. S.; Meleti, C. (2005). Deriving an effective aerosol
single scattering albedo from spectral surface UV irradiance measurements. Atmos Environ, 36: 1093-1102.
http://dx.doi.0rg/10.1016/i.atmosenv.2004.09.080669135
Balis, D. S.; Zerefos, C. S.; Kourtidis, K.; Bais, A. F.; Hofzumahaus, A.; Kraus, A.; Schmitt, R.; Blumthaler, M.; Gobbi, G.
P. (2002). Measurements and modeling of photolysis rates during the photochemical activity and ultraviolet
radiation (PAUR) II campaign. J Geophys Res, 107: 8138. http://dx.doi.org/10.1029/2000JD000136 055023
Bergstrom, R. W.; Pilewskie, P.; Pommier, J.; Rabbette M.; Russell, P. B.; Schmid, B.; Redemann, J.; Higurashi, A.;
Nakajima, T.; Quinn, P. K. (2004). Spectral absorption of solar radiation by aerosols during ACE-Asia. J Geophys
Res, 109: D19S15. http://dx.doi.org/10.1029/2003JD004467669158
Berntsen, T. K.; Myhre, G; Stordal, F.; Isaksen, I. S. A. (2000). Time evolution of tropospheric ozone and its radiative
forcing. J Geophys Res, 105: 8915-8930. http://dx.doi.org/10.1029/1999JD901139 047916
Boer, G. J.; Yu, B. (2003). Climate sensitivity and response. Clim Dynam, 20: 415-429. http://dx.doi.org/10.1007/s00382-
002-0283-3 630271
Bony, S.; Colman, R.; Kattsov, V M.; Allan, R. P.; Bretherton, C. S.; Dufresne, J. L.; Hall, A.; Hallegatte, S.; Holland, M.
M.; Ingram, W.; Randall, D. A.; Soden, B. J.; Tselioudis, G; Webb, M. J. (2006). How well do we understand and
evaluate climate change feedback processes? J Clim, 19: 3445-3482.
http://journals.ametsoc.Org/doi/abs/10.1175/JCLI3819.l. 630272
Brandt, L. A.; Bohnet, C.; King, J. Y. (2009). Photochemically induced carbon dioxide production as a mechanism for
carbon loss from plant litter in arid ecosystems. J Geophys Res, 114: G02004.
http://dx.doi.org/10.1029/2008jg000772 668891
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and
Environmental Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of
developing science assessments such as the Integrated Science Assessments (ISAs) and the Integrated Risk Information System (IRIS).
March 2011 10-29 DRAFT - DO NOT CITE OR QUOTE
-------
Breuer-McHam, I; Simpson, E.; Dougherty, I.; Bonkobara, M.; Ariizumi, K.; Lewis, D. E.; Dawson, D. B.; Duvic, M.;
Cruz, P. D. Jr (2001). Activation of HIV in human skin by ultraviolet B radiation and its inhibition by NF"kappa"B
blocking agents. Photochem Photobiol, 74: 805-810. 057736
Brewer, P. G; Peltzer, E. T. (2009). Limits to marine life. Science, 324: 347-348. http://dx.doi.org/10.1126/science.1170756
669197
Brown, B. A.; Jenkins, G. I. (2008). UV-B signaling pathways with different fluence-rate response profiles are
distinguished in mature Arabidopsis leaf tissue by requirement for UVR8, HY5, and HYH. Plant Physiol, 146: 576-
588. http://dx.doi.org/10.1104/pp.107.108456 668892
Bruhl, C.; Crutzen, P. J. (1989). On the disproportionate role of tropospheric ozone as a filter against solar UV-B radiation.
Geophys Res Lett, 16: 703-706. http://dx.doi.org/10.1029/GL016i007p00703 012518
Caforio, A. L. P.; Fortina, A. B.; piaserico, S.; Alaibac, M.; Tona, F.; Feltrin, G; Pompei, E.; Testolin, L.; Gambino, A.;
Volta, S. D.; Thiene, G; Casarotto, D.; Peserico, A. (2000). Skin cancer in heart transplant recipients: risk factor
analysis and relevance of immunosuppressive therapy. Circulation, 102: 111-222 -111-227. 080058
Calderini, D. F.; Lizana, X. C.; Hess, S.; Jobet, C. R.; Zuniga, J. A. (2008). Grain yield and quality of wheat under
increased ultraviolet radiation (UV-B) at later stages of the crop cycle. Journal of Agricultural Science, 146: 57-64.
http://dx.doi.org/10.1017/S0021859607007447668893
Caldwell, M. M.; Bornman, J. F.; Ballare, C. L.; Flint, S. D.; Kulandaivelu, G. (2007). Terrestrial ecosystems, increased
solar ultraviolet radiation, and interactions with bother climate change factors. Photochem Photobiol Sci, 6: 252-
266. http://dx.doi.org/10.1039/B700019g 668927
Cantorna, M. T. (2000). Vitamin D and autoimmunity: is vitamin D status an environmental factor affecting autoimmune
disease prevalence? Exp Biol Med, 223: 230-233. 080060
Chang, W.; Liao, H.; Wang, H. (2009). Climate responses to direct radiative forcing of anthropogenic aerosols,
tropospheric ozone, and long-lived greenhouse gases in eastern China over 1951-2000. Adv Atmos Sci, 26: 748-
762. http://dx.doi.org/10.1007/s00376-009-9032-4 630273
Chen, W. T.; Liao, H.; Seinfeld, J. H. (2007). Future climate impacts of direct radiative forcing of anthropogenic aerosols,
tropospheric ozone, and long-lived greenhouse gases. J Geophys Res, 112: D14209.
http://dx.doi.org/10.1029/2006JD008051 630274
Christiansen, B. (1999). Radiative forcing and climate sensitivity: The ozone experience. Q J Roy Meteorol Soc, 125:
3011-3035. http://dx.doi.org/10.1002/qj.49712556011 047920
Clarke, L. J.; Robinson, S. A. (2008). Cell wall-bound ultraviolet-screening compounds explain the high ultraviolet
tolerance of the Antarctic moss, Ceratodon purpureus. NewPhytol, 179: 776-783. http://dx.doi.org/10.1111/j.1469-
8137.2008.02499.x 668928
Clydesdale, G. J.; Dandie, G. W.; Muller, H. K. (2001). Ultraviolet light induced injury: immunological and inflammatory
effects. Immunol Cell Biol, 79: 547-568. 057105
Committee on Radiative Forcing Effects on Climate; Climate Research Committee; National Research Council (2005).
Radiative forcing of climate change: expanding the concept and addressing uncertainties. Washington, DC: The
National Academies Press. http://books.nap.edu/openbook.php?record_id=11175&page=Rl 057409
Cooper, O. R.; Parrish, D. D.; Stohl, A.; Trainer, M.; Nedelec, P.; Thouret, V; Cammas, J. P.; Oltmans, S. J.; Johnson, B. J.;
Tarasick, D.; Leblanc, T; McDermid, I. S.; Jaffe, D.; Gao, R.; Stith, J.; Ryerson, T; Aikin, K.; Campos, T;
Weinheimer, A.; Avery, M. A. (2010). Increasing springtime ozone mixing ratios in the free troposphere over
western North America. Nature, 463: 344-348. http://dx.doi.org/10.1038/nature08708 380093
Crist, K. C.; Carmichael, G. R.; John, K. (1994). UV-B exposure and atmospheric ozone - Evaluation of radiative flux to
changes in ambient ozone levels. J Hazard Mater, 37: 527-538. http://dx.doi.org/10.1016/0304-3894(93)E0096-K
668881
Croteau, M. C.; Davidson, M. A.; Lean, D. R. S.; Trudeau, V L. (2008). Global increases in ultraviolet B radiation:
Potential impacts on amphibian development and metamorphosis. Physiol Biochem Zool, 81: 743-761.
http://dx.doi.org/10.1086/591949668894
Croteau, M. C.; Martyniuk, C. J.; Trudeau, V L.; Lean, D. R. S. (2008). Chronic Exposure of Rana pipiens Tadpoles to
UVB Radiation and the Estrogenic Chemical 4-tert-octylphenol. J Toxicol Environ Health A, 71: 134-144. 603785
March 2011 10-30 DRAFT - DO NOT CITE OR QUOTE
-------
De Gruijl, F. R. (1995). Action spectrum for photocarcinogenesis. Recent Results Cancer Res, 139: 21-30. 057471
Diepgen, T. L.; Mahler, V. (2002). The epidemiology of skin cancer. Br J Dermatol, 146: 1-6. 093593
Favory, J. J.; Stec, A.; Gruber, H.; Rizzini, L.; Oravecz, A.; Funk, M.; Albert, A.; Cloix, C.; Jenkins, G. L; Oakeley, E. J.;
Seidlitz, H. K.; Nagy, F.; Ulm, R. (2009). Interaction of COP1 and UVR8 regulates UV-B-induced
photomorphogenesis and stress acclimation in Arabidopsis. EMBO J, 28: 591-601.
http://dx.doi.org/10.1038/emboi.2009.4668897
Fiore, A. M.; Jacob, D. J.; Field, B. D.; Streets, D. G; Fernandes, S. D.; Jang, C. (2002). Linking ozone pollution and
climate change: The case for controlling methane. Geophys Res Lett, 29: 1919.
http://dx.doi.org/10.1029/2002GL015601 051221
Forster, R; Ramaswamy, V; Artaxo, R; Berntsen, T; Betts, R.; Fahey, D. W.; Haywood, J.; Lean, J.; Lowe, D. C.; Myhre,
G; Nganga, J.; Prinn, R.; Raga, G; Schultz, M.; Van Borland, R. (2007). Changes in atmospheric constituents and
in radiative forcing. In S Solomon; D Qin; M Manning; Z Chen; M Marquis; KB Averyt; M Tignor; HL Miller
(Eds.), Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
Assessment Report of the Intergovernmental Panel on Climate Change (pp. 129-234). Cambridge, U.K. and New
York, NY: Cambridge University Press, http://www.ipcc.ch/pdf/assessment-report/ar4/wgl/ar4-wgl-chapter2.pdf
092936
Freedman, D. M.; Dosemeci, M.; McGlynn, K. (2002). Sunlight and mortality from breast, ovarian, colon, prostate, and
non-melanoma skin cancer: a composite death certificate based case-control study. Occup Environ Med, 59: 257-
262. 035530
Fritz, J. J.; Neale, P. J.; Davis, R. F.; Peloquin, J. A. (2008). Response of Antarctic phytoplankton to solar UVR exposure:
inhibition and recovery of photosynthesis in coastal and pelagic assemblages. Mar Ecol Prog Ser, 365: 1-16.
http://dx.doi.org/10.3354/Meps07610668898
Fuglestvedt, J. S.; Berntsen, T. K.; Isaksen, I. S. A.; Mao, H.; Liang, X.-Z.; Wang, W.-C. (1999). Climatic forcing of
nitrogen oxides through changes in tropospheric ozone and methane; global 3D model studies. Atmos Environ, 33:
961-978. http://dx.doi.org/10.1016/81352-2310(98)00217-9 047431
Fusco, A. C.; Logan, J. A. (2003). Analysis of 1970-1995 trends in tropospheric ozone at Northern Hemisphere
midlatitudes with the GEOS-CHEM model. J Geophys Res, 108: 4449. http://dx.doi.org/10.1029/2002JD002742
051229
Gao, K. S.; Ruan, Z. X.; Villafane, V. E.; Gattuso, J. P.; Helbling, E. W. (2009). Ocean acidification exacerbates the effect
of UV radiation on the calcifying phytoplankter Emiliania huxleyi. Limnol Oceanogr, 54: 1855-1862. 668899
Garcia, T. S.; Paoletti, D. J.; Blaustein, A. R. (2009). Correlated trait responses to multiple selection pressures in larval
amphibians reveal conflict avoidance strategies. Freshw Biol, 54: 1066-1077. http://dx.doi.Org/10.llll/j.1365-
2427.2008.02154.x 668935
Garland, F. C.; Garland, C. F.; Gorham, E. D.; Young, J. F. (1990). Geographic variation in breast cancer mortality in the
United States: a hypothesis involving exposure to solar radiation. Prev Med, 19: 614-622. 025242
Garssen, J.; Van Loveren, H. (2001). Effects of ultraviolet exposure on the immune system. Crit Rev Immunol, 21: 359-
397.057161
Gauss, M.; Myhre, G; Isaksen, I. S. A.; Grewe, V; Pitari, G; Wild, O.; Collins, W. J.; Dentener, F. J.; Ellingsen, K.; Gohar,
L. K.; Hauglustaine, D. A.; lachetti, D.; Lamarque, J. F.; Mancini, E.; Mickley, L. J.; Prather, M. J.; Pyle, J. A.;
Sanderson, M. G; Shine, K. P.; Stevenson, D. S.; Sudo, K.; Szopa, S.; Zeng, G. (2006). Radiative forcing since
preindustrial times due to ozone change in the troposphere and the lower stratosphere. Atmos Chem Phys, 6: 575-
599. wwwatmos-chem-phys.net/6/575/2006/. 630275
Gauss, M.; Myhre, G; Pitari, G; Prather, M. J.; Isaksen, I. S. A.; Berntsen, T. K.; Brasseur, G. P.; Dentener, F. J.; Derwent,
R. G; Hauglustaine, D. A.; Horowitz, L. W; Jacob, D. J.; Johnson, M.; Law, K. S.; Mickley, L. J.; Miiller, J.-F.;
Plantevin, P. H.; Pyle, J. A.; Rogers, H. L.; Stevenson, D. S.; Sundet, J. K.; Van Weele, M.; Wild, O. (2003).
Radiative forcing in the 21st century due to ozone changes in the troposphere and the lower stratosphere. J Geophys
Res, 108: 4292. http://dx.doi.org/10.1029/2002JD002624 094204
Gies, P.; Roy, C.; Toomey, S.; MacLennan, R.; Watson, M. (1998). Solar UVR exposures of primary school children at
three locations in Queensland. Photochem Photobiol, 68: 78-83. 057478
March 2011 10-31 DRAFT - DO NOT CITE OR QUOTE
-------
Gies, P.; Wright, J. (2003). Measured solar ultraviolet radiation exposures of outdoor workers in Queensland in the building
and construction industry. Photochem Photobiol, 78: 342-348. 057477
Giovannucci, E. (2005). The epidemiology of vitamin D and cancer incidence and mortality: a review (United States).
Cancer Causes Control, 16: 83-95. 074099
Gloster, H. M. Jr; Brodland, D. G (1996). The epidemiology of skin cancer. Dermatol Surg, 22: 217-226. 057479
Godar, D. E.; Wengraitis, S. P.; Shreffler, J.; Sliney, D. H. (2001). UV doses of Americans. Photochem Photobiol, 73: 621-
629. 057165
Goering, C. D.; L'Ecuyer, T S.; Stephens, G. L.; Slusser, J. R.; Scott, G; Davis, J.; Barnard, J. C.; Madronich, S. (2005).
Simultaneous retrievals of column ozone and aerosol optical propertiesfrom direct and diffuse solar irradiance
measurements. J Geophys Res, 110: D05204. http://dx.doi.org/10.1029/2004JD005330 669164
Gorham, E. D.; Garland, F. C.; Garland, C. F. (1990). Sunlight and breast cancer incidence in the USSR. Int J Epidemiol,
19: 820-824. 000682
Granstein, R. D.; Matsui, M. S. (2004). UV radiation-induced immunosuppression and skin cancer. Cutis, 5: 4-9. 057747
Grant, W. B. (2002). An ecologic study of dietary and solar ultraviolet-B links to breast carcinoma mortality rates. Cancer,
94:272-281.034981
Grant, W. B. (2002). An estimate of premature cancer mortality in the US due to inadequate doses of solar ultraviolet-B
radiation. Cancer, 94: 1867-1875. 025244
Grant, W. B.; Garland, C. F. (2004). A critical review of studies on vitamin D in relation to colorectal cancer. Nutr Cancer,
48: 115-123.075093
Grenfell, J. L.; Shindell, D. T; Koch, D.; Rind, D. (2001). Chemistry-climate interactions in the Goddard Institute for
Space Studies general circulation model 2. New insights into modeling the preindustrial atmosphere. J Geophys
Res, 106: 33435-33451. 664496
Haapala, J. K.; Morsky, S. K.; Saarnio, S.; Rinnan, R.; Suokanerva, H.; Kyr, E.; Latola, K.; Martikanen, P. J.; Holopainen,
T; Silvola, J. (2009). Carbon dioxide balance of a fen ecosystem in northern Finland under elevated UV-B
radiation. Global Change Biol, 15: 943-954. http://dx.doi.Org/10.llll/i.1365-2486.2008.01785.x607260
Hader, D. P.; Kumar, H. D.; Smith, R. C.; Worrest, R. C. (2007). Effects of solar UV radiation on aquatic ecosystems and
interactions with climate change. Photochem Photobiol Sci, 6: 267-285. http://dx.doi.org/10.1039/B700020k
668901
Hanchette, C. L.; Schwartz, G. G. (1992). Geographic patterns of prostate cancer mortality. Cancer, 70: 2861-2869. 025257
Hansen, J. E.; Sato, M.; Ruedy, R. (1997). Radiative forcing and climate response. J Geophys Res, 102: 6831-6864.
http://dx.doi.org/10.1029/96JD03436043104
Hansen, J.; Sato, M.; Ruedy, R.; Nazarenko, L.; Lacis, A.; Schmidt, G. A.; Russell, G; Aleinov, L; Bauer, M.; Bauer, S.;
Bell, N.; Cairns, B.; Canute, V; Chandler, M.; Cheng, Y; Del Genio, A.; Faluvegi, G; Fleming, E.; Friend, A.; Hall,
T; Jackman, C.; Kelley, M.; Kiang, N.; Koch, D.; Lean, J.; Lerner, J.; Lo, K.; Menon, S.; Miller, R.; Minnis, P.;
Novakov, T; Oinas, V; Perlwitz, J.; Perlwitz, J.; Rind, D.; Romanou, A.; Shindell, D.; Stone, P.; Sun, S.; Tausnev,
N.; Thresher, D.; Wielicki, B.; Wong, T; Yao, M.; Zhang, S. (2005). Efficacy of climate forcings. J Geophys Res,
110: D18104. http://dx.doi.org/10.1029/2005JD005776 190596
Harvey, L. D. D. (2004). Characterizing the annual-mean climatic effect of anthropogenic CO2 and aerosol emissions in
eight coupled atmosphere-ocean GCMs. Clim Dynam, 23: 569-599. http://dx.doi.org/10.1007/s00382-004-0455-4
190598
Heald, C. L.; Henze, D. K.; Horowitz, L. W; Feddema, J.; Lamarque, J. F.; Guenther, A.; Hess, P. G; Vitt, F.; Seinfeld, J.
H.; Goldstein, A. H.; Fung, I. (2008). Predicted change in global secondary organic aerosol concentrations in
response to future climate, emissions, and land use change. J Geophys Res, 113: D05211.
http://dx.doi.org/10.1029/2007id009092 191617
Held, I. M.; Soden, B. J. (2000). Water vapor feedback and global warming. Annual Review of Energy and the
Environment, 25: 441-475. 630279
March 2011 10-32 DRAFT - DO NOT CITE OR QUOTE
-------
Henry, H. A. L.; Brizgys, K.; Field, C. B. (2008). Litter decomposition in a California annual grassland: Interactions
between photodegradation and litter layer thickness. Ecosystems, 11: 545-554. http://dx.doi.org/10.1007/sl0021-
008-9141-4668902
Hildesheim, I; Fornace, A. J. Jr (2004). The dark side of light: the damaging effects of UV rays and the protective efforts
of MAP kinase signaling in the epidermis. DNA Repair, 3: 567-580. 057168
Holick, M. F. (2004). Sunlight and vitamin D for bone health and prevention of autoimmune diseases, cancers, and
cardiovascular disease. Am J Clin Nutr, 80: 1678S-1688S. 057691
Holland, M. M.; Bitz, C. M. (2003). Polar amplification of climate change in coupled models. Clim Dynam, 21: 221-232.
http://dx.doi.org/10.1007/s00382-003-0332-6630280
Holtby, L. B.; Bothwell, M. L. (2008). Effects of solar ultraviolet radiation on the behaviour of juvenile coho salmon
(Oncorhynchus kisutch): avoidance, feeding, and agonistic interactions. Can J Fish Aquat Sci, 65: 701-711.
http://dx.doi.org/10.1139/F08-013668903
Hughes, A. M.; Armstrong, B. K.; Vajdic, C. M.; Turner, J.; Grulich, A. E.; Fritschi, L.; Milliken, S.; Kaldor, J.; Benke, G;
Kricker, A. (2004). Sun exposure may protect against non-Hodgkin lymphoma: a case-control study. Int J Cancer,
112: 865-871.074101
ICNIRP (2004). Guidelines on limits of exposure to ultraviolet radiation of wavelengths between 180 nm and 400 nm
(incoherent optical radiation). In ICNIRP Guidelines (pp. 171-186). Oberschleissheim, Germany: International
Commission on Non-Ionizing Radiation Protection. 057187
loki, M.; Takahashi, S.; Nakajima, N.; Fujikura, K.; Tamaoki, M.; Saji, H.; Kubo, A.; Aono, M.; Kanna, M.; Ogawa, D.;
Fukazawa, J.; Oda, Y; Yoshida, S.; Watanabe, M.; Hasezawa, S.; Kondo, N. (2008). An unidentified ultraviolet-B-
specific photoreceptor mediates transcriptional activation of the cyclobutane pyrimidine dimer photolyase gene in
plants. Planta, 229: 25-36. http://dx.doi.org/10.1007/s00425-008-0803-4 668904
IPCC (2000). Emissions scenarios: A special report of working group III of the Intergovernmental Panel on Climate
Change. Cambridge, UK: Intergovernmental Panel on Climate Change; Cambridge University Press.
http://www.grida.no/climate/ipcc/emission/. 080704
IPCC (2007). Summary for policymakers. In: The Physical Science Basis. Contribution of Working Group I to the Fourth
Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, United Kingdom and New
York, NY, USA: Cambridge University Press, http://www.ipcc.ch/ipccreports/ar4-wgl.htm. 092980
Ito, A.; Sudo, K.; Akimoto, H.; Sillman, S.; Penner, J. E. (2007). Global modeling analysis of tropospheric ozone and its
radiative forcing from biomass burning emissions in the twentieth century. J Geophys Res, 112: D24307.
http://dx.doi.org/10.1029/2007JD008745 608803
Jacobson, M. Z.; Streets, D. G. (2009). Influence of future anthropogenic emissions on climate, natural emissions, and air
quality. J Geophys Res, 114: D08118. http://dx.doi.org/10.1029/2008JD011476 630281
Jaffe, D.; Price, H.; Fairish, D.; Goldstein, A.; Harris, J. (2003). Increasing background ozone during spring on the west
coast of North America. Geophys Res Lett, 30: 1613. http://dx.doi.org/10.1029/2003GL017024 052229
Jenkins, G. I. (2009). Signal transduction in responses to UV-B radiation. Annu Rev Plant Biol, 60: 407-431.
http://dx.doi.org/10.1146/annurev.arplant.59.032607.092953 668905
John, E. M.; Schwartz, G. G; Dreon, D. M.; Koo, J. (1999). Vitamin D and breast cancer risk: the NHANES I
Epidemiologic Follow-up Study, 1971-1975 to 1992. Cancer Epidemiol Biomarkers Prev, 8: 399-406. 057489
John, E. M.; Schwartz, G. G; Koo, J.; Van Den Berg, D.; Ingles, S. A. (2005). Sun exposure, vitamin D receptor gene
polymorphisms, and risk of advanced prostate cancer. Cancer Res, 65: 5470-5479. http://dx.doi.org/10.1158/0008-
5472.CAN-04-3134 670422
Johnson, C. E.; Collins, W. J.; Stevenson, D. S.; Derwent, R. G. (1999). The relative roles of climate and emissions changes
on future tropospheric oxidant concentrations. J Geophys Res, 104: 18631-18645.
http://dx.doi.org/10.1029/1999JD900204052390
Jokinen, I. E.; Markkula, E. S.; Salo, H. M.; Kuhn, P.; Nikoskelainen, S.; Arts, M. T; Browman, H. I. (2008). Exposure to
increased ambient ultraviolet B radiation has negative effects on growth, condition and immune function of juvenile
Atlantic salmon (Salmo salar). Photochem Photobiol, 84: 1265-1271. http://dx.doi.Org/10.llll/j.1751-
1097.2008.00358.x 668906
March 2011 10-33 DRAFT - DO NOT CITE OR QUOTE
-------
Jonson, J. E.; Simpson, D.; Fagerli, H.; Solberg, S. (2005). Can we explain the trends in European ozone levels? Atmos
Chem Phys, 6: 51-66. http://dx.doi.org/10.5194/acp-6-51-2006 630282
Joshi, M.; Shine, K. P.; Ponater, M.; Stuber, N.; Sausen, R.; Li (2003). A comparison of climate response to different
radiative forcings in three general circulation models: Towards an improved metric of climate change. Clim Dynam,
20: 843-854. http://dx.doi.org/10.1007/s00382-003-0305-9 193752
Kar, J.; Fishman, J.; Creilson, J. K.; Richter, A.; Ziemke, J.; Chandra, S. (2010). Are there urban signatures in the
tropospheric ozone column products derived from satellite measurements? Atmos Chem Phys, 10: 5213-5222.
http://dx.doi.org/10.5194/acp-10-5213-2010670423
Kataoka, Y; Kiguchi, M.; Williams, R. S.; Evans, P. D. (2007). Violet light causes photodegradation of wood beyond the
zone affected by ultraviolet radiation. Holzforschung und Holzverwertung, 61: 23-27.
http://dx.doi.org/10.1515/HF.2007.005670425
Kiehl, J. T.; Schneider, T. L.; Portmann, R. W.; Solomon, S. (1999). Climate forcing due to tropospheric and stratospheric
ozone. J Geophys Res, 104: 31239-31254. http://dx.doi.org/10.1029/1999JD900991 047917
Kimlin, M. G; Wong, J. C. F.; Parisi, A. V. (1998). Simultaneous comparison of the personal UV exposure of two human
groups at different altitudes. Health Phys, 74: 429-434. 057491
Koronakis, P. S.; Sfantos, G. K.; Paliatsos, A. G; Kaldellis, J. K.; Garofalakis, J. E.; Koronaki, I. P. (2002). Interrelations of
UV-global/global/diffuse solar irradiance components and UV-global attenuation on air pollution episode days in
Athens, Greece. Atmos Environ, 36: 3173-3181. 129938
Kyrola, E.; Tamminen, J.; Sofieva, V; Bertaux, J. L.; Hauchecorne, A.; Dalaudier, F.; Fussen, D.; Vanhellemont, F.;
dAndon, O. F.; Barrot, G; Guirlet, M.; Fehr, T.; de Miguel, L. S. (2010). GOMOS O-3, NO2, and NO3
observations in 2002-2008. Atmos Chem Phys, 10: 7723-7738. 667819
Lacis, A. A.; Wuebbles, D. J.; Logan, J. A. (1990). Radiative forcing of climate by changes in the vertical distribution of
ozone. J Geophys Res, 95: 9971-9981. http://dx.doi.org/10.1029/JD095iD07p09971 037834
Lamarque, J. F.; Bond, T. C.; Eyring, V.; Granier, C.; Heil, A.; Klimont, Z.; Lee, D.; Liousse, C.; Mieville, A.; Owen, B.;
Schultz, M. G; Shindell, D.; Smith, S. J.; Stehfest, E.; Van Aardenne, J.; Cooper, O. R.; Kainuma, M.; Mahowald,
N.; McConnell, J. R.; Naik, V; Riahi, K.; van Vuuren, D. P. (2010). Historical (1850-2000) gridded anthropogenic
and biomass burning emissions of reactive gases and aerosols: Methodology and application. Atmos Chem Phys
Discuss, 10: 4963-5019. http://dx.doi.org/10.5194/acpd-10-4963-2010630289
Lamarque, J. F.; Hess, P.; Emmons, L.; Buja, L.; Washington, W.; Granier, C. (2005). Tropospheric ozone evolution
between 1890 and 1990. J Geophys Res, 110: D08304. http://dx.doi.org/10.1029/2004JD005537 630287
Lefkowitz, E. S.; Garland, C. F. (1994). Sunlight, vitamin D, and ovarian cancer mortality rates in US women. Int J
Epidemiol,23: 1133-1136. 025253
Lelieveld, J.; van Aardenne, J.; Fischer, H.; de Reus, M.; Williams, J.; Winkler, P. (2004). Increasing ozone over the
Atlantic Ocean. Science, 304: 1483-1487. http://dx.doi.org/10.1126/science.1096777630578
Lenoble, J. (1993). Atmospheric radiative transfer. Hampton, VA: A. Deepak Publishing.
http://www.worldcat.org/title/atmospheric-radiative-transfer/oclc/27769441 630969
Levy H, I. L; Schwaarzkopf, M. D.; Horowitz, L.; Ramaswamy, V; Findell, K. L. (2008). Strong sensitivity of late 21st
century climate to projected changes in short-lived air pollutants. J Geophys Res, 113: D06102.
http://dx.doi.org/10.1029/2007JD009176631174
Liao, H.; Chen, W. T; Seinfeld, J. H. (2006). Role of climate change in global predictions of future tropospheric ozone and
aerosols. J Geophys Res, 111: D12304. http://dx.doi.org/10.1029/2005jd006852 664500
Liao, H.; Seinfeld, J. H.; Adams, P. J.; Mickley, L. J. (2004). Global radiative forcing of coupled tropospheric ozone and
aerosols in a unified general circulation model. J Geophys Res, 109: D16207.
http://dx.doi.org/10.1029/2003JD004456057414
Lindelof, B.; Sigurgeirsson, B.; Gabel, H.; Stern, R. S. (2000). Incidence of skin cancer in 5356 patients following organ
transplantation. Br J Dermatol, 143: 513-519. 080084
March 2011 10-34 DRAFT - DO NOT CITE OR QUOTE
-------
Logan, J. A.; Megretskaia, I. A.; Miller, A. I; Tiao, G. C.; Choi, D.; Zhang, L.; Stolarski, R. S.; Labow, G. I;
Hollandsworth, S. M.; Bodeker, G. E.; Claude, H.; De Muer, D.; Kerr, J. B.; Tarasick, D. W.; Oltmans, S. J.;
Johnson, B.; Schmidlin, R; Staehelin, J.; Viatte, P.; Uchino, O. (1999). Trends in the vertical distribution of ozone:
A comparison of two analyses of ozonesonde data. J Geophys Res, 104: 26373-26399.
http://dx.doi.org/10.1029/1999JD900300631175
Longstreth, J. D.; De Gruijl; Kripke, M. L.; Takizawa, Y; van der Leun, J. C. (1995). Effects of increased solar ultraviolet
radiation on human health. Ambio, 24: 153-165. 055174
Longstreth, J.; de Gruijl, F. R.; Kripke, M. L.; Abseck, S.; Arnold, R; Slaper, H. L; Velders, G; Takizawa, Y; van der Leun,
J. C. (1998). Health risks. J Photochem Photobiol B, 46: 20-39. 001200
Lutter, R.; Wolz, C. (1997). UV-B screening by tropospheric ozone: Implications for the national ambient air quality
standard. Environ Sci Technol, 31: 142A-146A. 082672
Madronich, S.; De Gruijl, F. (1993). Skin cancer and UV radiation. Nature, 3666450: 23. 055183
Marenco, A.; Gouget, H.; Nedelec, P.; Pages, J.-P; Karcher, F. (1994). Evidence of a long-term increase in tropospheric
ozone from Pic du Midi data series: Consequences: Positive radiative forcing. J Geophys Res, 99: 16617-16632.
http://dx.doi.org/10.1029/94JD00021 047733
Markkula, E.; Salo, H. M.; Rikalainen, K.; Jokinen, I. E. (2009). Long-term UVB irradiation affects the immune functions
of carp (Cyprinus carpio) and rainbow trout (Oncorhynchus mykiss). Photochem Photobiol, 85: 347-352.
http://dx.doi.0rg/10.llll/i.1751-1097.2008.00446.x613291
Marquis, O.; Miaud, C. (2008). Variation in UV sensitivity among common frog Rana temporaria populations along an
altitudinal gradient. Zoology (Jena), 111: 309-317. http://dx.doi.Org/10.1016/j.zool.2007.09.003 668908
Marquis, O.; Miaud, C.; Lena, J. P. (2008). Developmental responses to UV-B radiation in common frog Rana temporaria
embryos from along an altitudinal gradient. Population Ecology, 50: 123-130. http://dx.doi.org/10.1007/sl0144-
007-0071-3 668909
Matsumura, Y; Ananthaswamy, H. N. (2004). Toxic effects of ultraviolet radiation on the skin. Toxicol Appl Pharmacol,
195:298-308.057184
Mayer, L. M.; Schick, L. L.; Hardy, K. R.; Estapa, M. L. (2009). Photodissolution and other photochemical changes upon
irradiation of algal detritus. Limnol Oceanogr, 54: 1688-1698. 668910
Mazza, C. A.; Izaguirre, M. M.; Curiale, J.; Ballare, C. L. (2010). A look into the invisible: Ultraviolet-B sensitivity in an
insect (Caliothrips phaseoli) revealed through a behavioural action spectrum. Proc Biol Sci, 277: 367-373.
http://dx.doi.org/10.1098/rspb.2009.1565668911
Meador, J. A.; Baldwin, A. J.; Catala, P.; Jeffrey, W. H.; Joux, F.; Moss, J. A.; Pakulski, J. D.; Stevens, R.; Mitchell, D. L.
(2009). Sunlight-induced DNA damage in marine micro-organisms collected along a latitudinal gradient from 70
degreesNto 68 degrees S. Photochem Photobiol, 85: 412-421. http://dx.doi.Org/10.llll/j.1751-1097.2008.00462.x
668912
Menon, S.; Unger, N.; Koch, D.; Francis, J.; Garrett, T; Sednev, L; Shindell, D.; Streets, D. (2008). Aerosol climate effects
and air quality impacts from 1980 to 2030. Environmental Research Letters, 3: 024004.
http://dx.doi.Org/10.1088/1748-9326/3/2/024004613861
Mickley, L. J.; Jacob, D. J.; Field, B. D.; Rind, D. (2004). Climate response to the increase in tropospheric ozone since
preindustrial times: A comparison between ozone and equivalent CO2 forcings. J Geophys Res, 109: D05106.
http://dx.doi.org/10.1029/2003JD003653057416
Mickley, L. J.; Jacob, D. J.; Rind, D. (2001). Uncertainty in preindustrial abundance of tropospheric ozone: Implications
for radiative forcing calculations. J Geophys Res, 106: 3389-3399. http://dx.doi.org/10.1029/2000JD900594
080134
Mickley, L. J.; Murti, P. P.; Jacob, D. J.; Logan, J. A.; Koch, D. M.; Rind, D. (1999). Radiative forcing from tropospheric
ozone calculated with a unified chemistry-climate model. J Geophys Res, 104: 30153-30172.
http://dx.doi.org/10.1029/1999JD900439047918
Moehrle, M. (2001). Ultraviolet exposure in the Ironman triathlon. Med Sci Sports Exerc, 33: 1385-1386. 057502
Moehrle, M.; Heinrich, L.; Schmid, A.; Garbe, C. (2000). Extreme UV exposure of professional cyclists. Dermatology,
201:44-45.057503
March 2011 10-35 DRAFT - DO NOT CITE OR QUOTE
-------
Moise, A. R; Buttner, P. G; Harrison, S. L. (1999). Sun exposure at school. Photochem Photobiol, 70: 269-274. 057504
Moss, R. H.; Edmonds, J. A.; Hibbard, K. A.; Manning, M. R.; Rose, S. K.; van Vuuren, D. P.; Carter, T. R.; Emori, S.;
Kainuma, M.; Kram, T.; Meehl, G. A.; Mitchell, J. R; Nakicenovic, N.; Riahi, K.; Smith, S. J.; Stouffer, R. J.;
Thomson, A. M.; Weyant, J. P.; Wilbanks, T. J. (2010). The next generation of scenarios for climate change research
and assessment. Nature, 463: 747-756. http://dx.doi.org/10.1038/nature08823 664501
Naik, V.; Mauzerall, D.; Horowitz, L.; Schwarzkopf, M. D.; Ramaswamy, V.; Oppenheimer, M. (2005). Net radiative
forcing due to changes in regional emissions of tropospheric ozone precursors. J Geophys Res, 110: D24306.
http://dx.doi.org/10.1029/2005JD005908 193194
Naja, M.; Akimoto, H. (2004). Contribution of regional pollution and long-range transport to the Asia-Pacific region:
Analysis of long-term ozonesonde data over Japan. J Geophys Res, 109: D21306.
http://dx.doi.org/10.1029/2004JD004687 631178
National ambient air quality standards for ozone: Final response to remand; final rule. 68 Fed. Reg. 614-645 (2003).
015702
Nole, G; Johnson, A. W. (2004). An analysis of cumulative lifetime solar ultraviolet radiation exposure and the benefits of
daily sun protection. Dermatol Ther, 17: 57-62. 057505
Norval, M.; Garssen, J.; Van Loveren, H.; El-Ghorr, A. A. (1999). UV-induced changes in the immune response to
microbial infections in human subjects and animal models. J Epidemiol, 6: S84-S92. 036089
Obara, Y; Koshitaka, H.; Arikawa, K. (2008). Better mate in the shade: Enhancement of male mating behaviour in the
cabbage butterfly, Pieris rapae crucivora, in a UV-rich environment. J Exp Biol, 211: 3698-3702.
http://dx.doi.org/10.1242/ieb.021980668913
Okada, S.; Weatherhead, E.; Targoff, I. N.; Wesley, R.; Miller, F. W.; International Myositis Collaborative Study Group
(2003). Global surface ultraviolet radiation intensity may modulate the clinical and immunologic expression of
autoimmune muscle disease. Arthritis Rheum, 48: 2285-2293. 057769
Oltmans, S. J.; Lefohn, A. S.; Harris, J. M.; Galbally, L; Scheel, H. E.; Bodeker, G; Brunke, E.; Claude, H.; Tarasick, D.;
Johnson, B. J.; Simmonds, P.; Shadwick, D.; Anlauf, K.; Hayden, K.; Schmidlin, F.; Fujimoto, T; Akagi, K.; Meyer,
C.; Nichol, S.; Davies, J.; Redondas, A.; Cuevas, E. (2006). Long-term changes in tropospheric ozone. Atmos
Environ, 40: 3156-3173. http://dx.doi.Org/10.1016/j.atmosenv.2006.01.029 180188
Oromi, N.; Marquis, O.; Miaud, C.; Sanuy, D. (2008). Influence of ambient ultraviolet radiation on Bufo calamita egg
development in a semiarid zone (Catalonia, Spain). J Environ Biol, 29: 135-137. 668914
Palancar, G. G; Toselli, B. M. (2002). Erythemal ultraviolet irradiance in Cordoba, Argentina. Atmos Environ, 36: 287-292.
057207
Parrella, J. P.; Evans, M. J.; Jacob, D. J.; Liang, Q.; Mickley, L. J.; Miller, B.; Pyle, J. A.; Yang, X. (In Press). Effect of
bromine chemistry on natural tropospheric ozone: Improved simulation of observations from the turn of the 20th
century. Nature, http://acmg.seas.harvard.edu/publications/parrella2010.pdf 664506
Pavelin, E. G; Johnson, C. E.; Rughooputh, S.; Toumi, R. (1999). Evaluation of pre-industrial surface ozone measurements
made using Schonbein's method. Atmos Environ, 33: 919-929. http://dx.doi.org/10.1016/S1352-2310(98)00257-X
087296
Pfister, H. (2003). Human papillomavirus and skin cancer. J Natl Cancer Inst, Monographs No. 31: 52-56. 057515
Phoenix, G. K.; Gwynn-Jones, D.; Lee, J. A.; Callaghan, T. V. (2000). The impacts of UV-B radiation on the regeneration of
a sub-arctic heath community. Plant Ecol, 146: 67-75. http://dx.doi.Org/10.1023/A:1009839506658 668915
Pickett, J. E.; Gibson, D. A.; Gardner, M. M. (2008). Effects of irradiation conditions on the weathering of engineering
thermoplastics. Polym Degrad Stabil, 93: 1597-1606. http://dx.doi.Org/10.1016/j.polymdegradstab.2008.02.009
668916
Ponsonby, A.-L.; McMichael, A.; Van der Mei, I. (2002). Ultraviolet radiation and autoimmune disease: insights from
epidemiological research. Toxicology, 181/182: 71-78. 080061
Prather, M.; Gauss, M.; Berntsen, T; Isaksen, I.; Sundet, J.; Bey, I.; Brasseur, G; Dentener, F.; Derwent, R.; Stevenson, D.;
Grenfell, L.; Hauglustaine, D.; Horowitz, L.; Jacob, D.; Mickley, L.; Lawrence, M.; Von Kuhlmann, R.; Muller, J.
R; Pitari, G; Rogers, H.; Johnson, M.; Pyle, J.; Law, K.; Van Weele, M.; Wild, O. (2003). Fresh air in the 21st
century? Geophys Res Lett, 30: 1100. http://dx.doi.org/10.1029/2002GL016285 047879
March 2011 10-36 DRAFT - DO NOT CITE OR QUOTE
-------
Price, C.; Rind, D. (1994). Possible implications of global climate change on global lightning distributions and frequencies.
J Geophys Res, 99: 10823-10831. http://dx.doi.org/10.1029/94JD00019630283
Pyle, J. A.; Warwick, N.; Yang, X.; Young, P. J.; Zeng, G. (2007). Climate/chemistry feedbacks and biogenic emissions.
Philos Transact A Math PhysEng Sci, 365: 1727-1740. http://dx.doi.org/10.1098/rsta.2007.2041 630284
RCP (2009). RCP Database (version 2.0). Retrieved January 28, 2011 from http://iiasa.ac.at/web-
apps/tnt/RcpDb/dsd?Action=htmlpage&page=about. 677552
Repapis, C. C.; Mantis, H. T.; Paliatsos, A. G; Philandras, C. M.; Bais, A. R; Meleti, C. (1998). Case study of UV-B
modification during episodes of urban air pollution. Atmos Environ, 38: 2203-2208. 038015
Rigel, D. S.;Rigel,E. G;Rigel,A. C. (1999). Effects of altitude and latitude on ambient UVB radiation. JAmAcad
Dermatol, 40: 114-116. 057511
Riggsbee, J. A.; Orr, C. H.; Leech, D. M.; Doyle, M. W.; Wetzel, R. G. (2008). Suspended sediments in river ecosystems:
Photochemical sources of dissolved organic carbon, dissolved organic nitrogen, and adsorptive removal of
dissolved iron. J Geophys Res, 113: G03019. http://dx.doi.org/10.1029/2007jg000654 668917
Rind, D.; Healy, R.; Parkinson, C.; Martinson, D. (1995). The role of sea ice in 2xCO2 climate model sensitivity. Part I:
The total influence of sea ice thickness and extent. J Clim, 8: 449-463. http://dx.doi.org/10.1175/1520-
0442(1995)008<0449:TROSII>2.0.CO;2 630285
Romansic, J. M.; Waggener, A. A.; Bancroft, B. A.; Blaustein, A. R. (2009). Influence of ultra violet-B radiation on growth,
prevalence of deformities, and susceptibility to predation in Cascades frog (Rana cascadae) larvae. Hydrobiologia,
624: 219-233. http://dx.doi.org/10.1007/sl0750-009-9703-2 515730
Rosenthal, F. S.; Phoon, C.; Bakalian, A. E.; Taylor, H. R. (1988). The ocular dose of ultraviolet radiation to outdoor
workers. Invest Ophthalmol Vis Sci, 29: 649-656. 057513
Sanderson, M. G; Jones, C. D.; Collins, W. J.; Johnson, C. E.; Derwent, R. G. (2003). Effect of climate change on isoprene
emissions and surface ozone levels. Geophys Res Lett, 30: 1936. http://dx.doi.org/10.1029/2003GL017642 630286
Schenker, M. B.; Orenstein, M. R.; Samuels, S. J. (2002). Use of protective equipment among California farmers. Am J Ind
Med, 42: 455-464. 057522
Selgrade M-JK; Smith, M. V; Oberhelman-Bragg, L. J.; LeVee, G. J.; Koren, H. S.; Cooper, K. D. (2001). Dose response
for UV-induced immune suppression in people of color: Differences based on erythemal reactivity rather than skin
pigmentation. Photochem Photobiol, 74: 88-95. 057454
Selgrade, M. K.; Repacholi, M. H.; Koren, H. S. (1997). Ultraviolet radiation-induced immune modulation: Potential
consequences for infectious, allergic, and autoimmune disease. Environ Health Perspect, 105: 332-334. 036165
Semerdjieva, S. I.; Phoenix, G. K.; Hares, D.; Gwynn-Jones, D.; Callaghan, T. V; Sheffield, E. (2003). Surface
morphology, leaf and cuticle thickness of four dwarf shrubs from a sub-Arctic heath following long-term exposure
to enhanced levels of UV-B. Physiol Plant, 117: 289-294. http://dx.doi.Org/10.1034/j.1399-3054.2003.00006.x
668919
Shindell, D. T; Faluvegi, G. (2002). An exploration of ozone changes and their radiative forcing prior to the
chlorofluorocarbon era. Atmos Chem Phys Discuss, 2: 363-374. http://dx.doi.org/10.5194/acp-2-363-2002 080130
Shindell, D. T; Faluvegi, G; Bauer, S. E.; Koch, D. M.; Unger, N.; Menon, S.; Miller, R. L.; Schmidt, G. A.; Streets, D. G.
(2007). Climate response to projected changes in short-lived species under an A1B scenario from 2000-2050 in the
GISS climate model. J Geophys Res, 112: D20103. http://dx.doi.org/10.1029/2007jd008753 521350
Shindell, D. T; Faluvegi, G; Bell, N. (2003). Preindustrial-to-present-day radiative forcing by tropospheric ozone from
improved simulations with the GISS chemistry-climate GCM. Atmos Chem Phys, 3: 1675-1702.
http://dx.doi.org/10.5194/acp-3-1675-2003057417
Shindell, D. T; Levy H, I. L; Schwarzkopf, M. D.; Horowitz, L. W.; Lamarque, J. F.; Faluvegi, G. (2008). Multimodel
projections of climate change from short-lived emissions due to human activities. J Geophys Res, 113: D11109.
http://dx.doi.org/10.1029/2007JD009152 190393
Shindell, D.; Faluvegi, G. (2009). Climate response to regional radiative forcing during the twentieth century. Nat Geosci,
2: 294-300. http://dx.doi.org/10.1038/ngeo473 631180
March 2011 10-37 DRAFT - DO NOT CITE OR QUOTE
-------
Shindell, D.; Faluvegi, G; Lacis, A.; Hansen, I; Ruedy, R.; Aguilar, E. (2006). Role of tropospheric ozone increases in
20th-century climate change. J Geophys Res, 111: D08302. http://dx.doi.org/10.1029/2005JD006348 631181
Shoveller, J. A.; Lovato, C. Y; Peters, L.; Rivers, J. K. (1998). Canadian National Survey on Sun Exposure & Protective
Behaviours: adults at leisure. Cancer Prev Control, 2: 111-116. 057524
Simmonds, P. G; Derwent, R. G; Manning, A. L.; Spain, G. (2004). Significant growth in surface ozone at Mace Head,
Ireland, 1987-2003. Atmos Environ, 38: 4769-4778. http://dx.doi.Org/10.1016/j.atmosenv.2004.04.036 631182
Sinha, A.; Toumi, R. (1997). Tropospheric ozone, lightning, and climate change. J Geophys Res, 102: 10667-10672.
http://dx.doi.org/10.1029/96JD03710047932
Sitch, S.; Cox, P. M.; Collins, W. J.; Huntingford, C. (2007). Indirect radiative forcing of climate change through ozone
effects on the land-carbon sink. Nature, 448: 791-794. http://dx.doi.org/10.1038/nature06059 093294
Slaper, H.; Velders, G. J. M.; Daniel, J. S.; de Gruijl, F. R.; Van der Leun, J. C. (1996). Estimates of ozone depletion and
skin cancer incidence to examine the Vienna Convention achievements. Nature, 384: 256-258. 055128
Sliney, D. H.; Wengraitis, S. (2006). Is a differentiated advice by season and region necessary? Prog Biophys Mol Biol, 92:
150-160. http://dx.doi.0rg/10.1016/i.pbiomolbio.2006.02.007651896
Smedby, K. E.; Hjalgrim, H.; Melbye, M.; Torrang, A.; Rostgaard, K.; Munksgaard, L.; Adami, J.; Hansen, M.; Porwit-
MacDonald, A.; Jensen, B. A.; Roos, G; pedersen, B. B.; Sundstrom, C.; Glimelius, B.; Adami, H.-O. (2005).
Ultraviolet radiation exposure and risk of malignant lymphomas. J Natl Cancer Inst, 97: 199-209. 669175
Snell, K. R. S.; Kokubun, T.; Griffiths, H.; Convey, P.; Hodgson, D. A.; Newsham, K. K. (2009). Quantifying the metabolic
cost to an Antarctic liverwort of responding to an abrupt increase in UVB radiation exposure. Global Change Biol,
15: 2563-2573. http://dx.doi.0rg/10.llll/i.1365-2486.2009.01929.x 668920
Soden, B. J.; Held, I. M. (2006). An assessment of climate feedbacks in coupled ocean-atmosphere models. J Clim, 19:
3354-3360. http://iournals.ametsoc.Org/doi/abs/10.1175/JCLI3799.l. 631183
Staehelin, J.; Thudium, J.; Buehler, R.; Volz-Thomas, A.; Graber, W. (1994). Trends in surface ozone concentrations at
Arosa (Switzerland). Atmos Environ, 28: 75-87. http://dx.doi.org/10.1016/1352-2310(94)90024-8 055369
Stevenson, D. S.; Dentener, F. J.; Schultz, M. G; Ellingsen, K.; Van Noije, T. P. C.; Wild, O.; Zeng, G; Amann, M.;
Atherton, C. S.; Bell, N.; Bergmann, D. J.; Bey, L; Butler, T; Cofala, J.; Collins, W. J.; Derwent, R. G; Doherty, R.
M.; Drevet, J.; Eskes, H. J.; Fiore, A. M.; Gauss, M.; Hauglustaine, D. A.; Horowitz, L. W.; Isaksen, I. S. A.; Krol,
M. C.; Lamarque, J. F.; Lawrence, M. G; Montanaro, V; Muller, J. F.; Pitari, G; Prather, M. J.; Pyle, J. A.; Rast, S.;
Rodriguez, J. M.; Sanderson, M. G; Savage, N. H.; Shindell, D. T; Strahan, S. E.; Sudo, K.; Szopa, S. (2006).
Multimodel ensemble simulations of present-day and near-future tropospheric ozone. J Geophys Res, 111: D08301.
http://dx.doi.org/10.1029/2005JD006338089222
Streets, D.; Bond, T; Lee, T; Jang, C. (2004). On the future of carbonaceous aerosol emissions. J Geophys Res, 109:
D24212. http://dx.doi.org/10.1029/2004JD004902 190423
Studzinski, G. P.; Moore, D. C. (1995). Sunlight—can it prevent as well as cause cancer? Cancer Res, 55: 4014-4022.
057554
Tanimoto, H. (2009). Increase in springtime tropospheric ozone at a mountainous site in Japan for the period 1998-2006.
Atmos Environ, 43: 1358-1363. http://dx.doi.Org/10.1016/j.atmosenv.2008.12.006 620751
Tarasick, D. W.; Fioletov, V E.; Wardle, D. L; Kerr, J. B.; Davies, J. (2005). Changes in the vertical distribution of ozone
over Canada from ozonesondes: 1980-2001. J Geophys Res, 110: D02304.
http://dx.doi.org/10.1029/2004JD004643 631184
ten Berge, O.; van Weelden, H.; Bruijnzeel-Koomen, C.; de Bruin-Weller, M. S.; Sigurdsson, V (2009). Throwing a light
on photosensitivity in atopic dermatitis: A retrospective study. Am J Clin Dermatol, 10: 119-123.
http://dx.doi.org/10.2165/00128071-200910020-00004651897
Thieden, E.; Philipsen, P. A.; Heydenreich, J.; Wulf, H. C. (2004). UV radiation exposure related to age, sex, occupation,
and sun behavior based on time-stamped personal dosimeter readings. Arch Dermatol, 140: 197-203. 057558
Thieden, E.; Philipsen, P. A.; Sandby-Moller, J.; Heydenreich, J.; Wulf, H. C. (2004). Proportion of lifetime UV dose
received by children, teenagers and adults based on time-stamped personal dosimetry. J Invest Dermatol, 123:
1147-1150.057557
March 2011 10-38 DRAFT - DO NOT CITE OR QUOTE
-------
Thompson, A. M. (1992). The oxidizing capacity of the Earth's atmosphere: Probable past and future changes. Science,
256: 1157-1165. http://dx.doi.org/10.1126/science.256.5060.1157043463
Thompson, A. M.; Chappellaz, J. A.; Fung, I. Y; Kucsera, T L. (1993). The atmospheric CH4 increase since the last glacial
maximum (2) Interactions with oxidants. Tellus B Chem Phys Meteorol, 45: 242-257.
http://dx.doi.Org/10.1034/i.1600-0889.1993.t01-2-00003.x029510
Thompson, A. M.; Hudson, R. D. (1999). Tropical tropospheric ozone (TTO) maps from Nimbus 7 and Earth Probe TOMS
by the modified-residual method: Evaluation with sondes, ENSO signals, and trends from Atlantic regional time
series. J Geophys Res, 104: 26961-26975. http://dx.doi.org/10.1029/1999JD900470 631185
Thompson, A. M.; Stone, J. B.; Witte, J. C.; Miller, S. K.; Oltmans, S. J.; Kucsera, T. L.; Ross, K. L.; Pickering, K. E.;
Merrill, J. T; Forbes, G; Tarasick, D. W.; Joseph, E.; Schmidlin, F. J.; McMillan, W. W.; Warner, J.; Hintsa, E. J.;
Johnson, J. E. (2007). Intercontinental Chemical Transport Experiment Ozonesonde Network study (IONS) 2004: 2
Tropospheric ozone budgets and variability over northeastern North America. J Geophys Res, 112: D12S13.
http://dx.doi.org/10.1029/2006JD007670 090796
U.S. EPA (2006). Air quality criteria for ozone and related photochemical oxidants (Report No. EPA/600/R-05/004AF).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid= 149923. 088089
Ullrich, S. E. (2005). Mechanisms underlying UV-induced immune suppression. Mutat Res, 571: 185-205.
http://dx.doi.0rg/10.1016/i.mrfmmm.2004.06.059095635
UNEP (2009). Environmental effects of ozone depletion and its interactions with climate change. Nairobi, Kenya: United
Nations Environment Programme. http://ozone.unep. org/Assessment_Panels/EEAP/EEAP-Progress-report-
2009.pdf 669084
Unger, N.; Bond, T. C.; Wang, J. S.; Koch, D. M.; Menon, S.; Shindell, D. T; Bauer, S. (2010). Attribution of climate
forcing to economic sectors. PNAS, 107: 3382-3387. http://dx.doi.org/10.1073/pnas.0906548107 387104
Unger, N.; Shindell, D. T; Koch, D. M.; Streets, D. G. (2008). Air pollution radiative forcing from specific emissions
sectors at 2030. J Geophys Res, 113: D02306. http://dx.doi.org/10.1029/2007JD008683 631186
Urbach, F. (1997). Ultraviolet radiation and skin cancer of humans. J Photochem Photobiol B, 40: 3-7. 086255
Van Aardenne, J. A.; Dentener, F. J.; Olivier, J. G. J.; Klein Goldewijk, C. G. M.; Lelieveld, J. (2001). A l°xl° resolution
data set of historical anthropogenic trace gas emissions for the period 1890-1990. Global Biogeochem Cycles, 15:
909-928. http://dx.doi.org/10.1029/2000GB001265 055564
Vishvakarman, D.; Wong, J. C. F.; Boreham, B. W. (2001). Annual occupational exposure to ultraviolet radiation in central
Queensland. Health Phys, 81: 536-544. 057614
Volz, A.; Kley, D. (1988). Evaluation of the Montsouris series of ozone measurements made in the nineteenth century.
Nature, 332: 240-242. http://dx.doi.org/10.1038/332240aO 041650
Wahl, M. (2008). Ecological modulation of environmental stress: interactions between ultraviolet radiation, epibiotic snail
embryos, plants and herbivores. JAnimEcol, 77: 549-557. http://dx.doi.Org/10.llll/j.1365-2656.2007.01352.x
668921
Wang, W.-C.; Pinto, J. P.; Yung, Y. L. (1980). Climatic effects due to halogenated compounds in the earth's atmosphere. J
Atmos Sci, 37: 333-338. http://dx.doi.org/10.1175/1520-0469(1980)037<0333:CEDTHO2.0.CO;2 674821
West, J. J.; Fiore, A. M.; Horowitz, L. W.; Mauzerall, D. L. (2006). Global health benefits of mitigating ozone pollution
with methane emission controls. PNAS, 103: 3988-93. http://dx.doi.org/10.1073/pnas.0600201103 196558
West, J. J.; Fiore, A. M.; Naik, V; Horowitz, L. W.; Schwarzkopf, M. D.; Mauzerall, D. L. (2007). Ozone air quality and
radiative forcing consequences of changes in ozone precursor emissions. Geophys Res Lett, 34: L06806.
http://dx.doi.org/10.1029/2006GL029173622733
Wild, O.; Prather, M. J.; Akimoto, H. (2001). Indirect long-term global radiative cooling from NOX emissions. Geophys
Res Lett, 28: 1719-1722. http://dx.doi.org/10.1029/2000GL012573 193196
WMO (2006). Scientific assessment of ozone depletion: 2006. Geneva, Switzerland: World Meterological Organization.
http://www.wmo.int/pages/prog/arep/gaw/ozone_2006/ozone_asst_report.html. 669178
March 2011 10-39 DRAFT - DO NOT CITE OR QUOTE
-------
Worden, H. M.; Bowman, K. W.; Worden, J. R.; Eldering, A.; Beer, R. (2008). Satellite measurements of the clear-sky
greenhouse effect from tropospheric ozone. Nat Geosci, 1: 305-308. http://dx.doi.org/10.1038/ngeol82 631188
Wu, S.; Mickley, L. J.; Jacob, D. J.; Rind, D.; Streets, D. G. (2008). Effects of 2000-2050 changes in climate and emissions
on global tropospheric ozone and the policy-relevant background surface ozone in the United States. J Geophys
Res, 113: D18312. http://dx.doi.org/10.1029/2007JD009639 190039
Wu, S.; Mickley, L. J.; Leibensperger, E. M.; Jacob, D. J.; Rind, D.; Streets, D. G. (2008). Effects of 2000-2050 global
change on ozone air quality in the United States. J Geophys Res, 113: D06302.
http://dx.doi.org/10.1029/2007JD008917629684
Zeng, G; Pyle, J. A. (2003). Changes in tropospheric ozone between 2000 and 2100 modeled in a chemistry-climate model.
Geophys Res Lett, 30: 1392. http://dx.doi.org/10.1029/2002GL016708 047492
Zepp, R. G; Erickson, D. J.; Paul, N. D.; Sulzberger, B. (2007). Interactive effects of solar UV radiation and climate
change on biogeochemical cycling. In The Environmental Effects of Ozone Depletion and its Interactions with
Climate Change: 2006 Assessment (pp. 135-164). Nairobi, Kenya: United Nations Environment Programme.
668896
Zepp, R. G; Shank, G. C.; Stabenau, E.; Patterson, K. W.; Cyterski, M.; Fisher, W.; Bartels, E.; Anderson, S. L. (2008).
Spatial and temporal variability of solar ultraviolet exposure of coral assemblages in the Florida Keys: Importance
of colored dissolved organic matter. Limnol Oceanogr, 53: 1909-1922. 668922
Zerefos, C. S.; Kourtidis, K. A.; Melas, D.; Balis, D.; Zanis, P.; Katsaros, L.; Mantis, H. T.; Repapis, C.; Isaksen, I.; Sundet,
J.; Herman, J.; Bhartia, P. K.; Calpini, B. (2002). Photochemical activity and solar ultraviolet radiation (PAUR)
modulation factors: An overview of the project. J Geophys Res, 107: 8134.
http://dx.doi.org/10.1029/2000JD000134055169
Ziemke, J. R.; Chandra, S.; Bhartia, P. K. (2005). A 25-year data record of atmospheric ozone in the Pacific from Total
Ozone Mapping Spectrometer (TOMS) cloud slicing: Implications for ozone trends in the stratosphere and
troposphere. J Geophys Res, 110: D15105. http://dx.doi.org/10.1029/2004JD005687631193
March 2011 10-40 DRAFT - DO NOT CITE OR QUOTE
-------
|