f\ ^™ F^ A
> C PA
      United States
          Air Quality Criteria for
          Ozone and Related
          Photochemical Oxidants
          (Second External Review
          Draft)
          Volume II of

-------
                                                 EPA/600/R-05/004bB
                                                      August 2005
Air Quality Criteria for Ozone and  Related
           Photochemical  Oxidants
                    Volume II
         National Center for Environmental Assessment-RTF Office
                Office of Research and Development
               U.S. Environmental Protection Agency
                  Research Triangle Park, NC

-------
                                   DISCLAIMER

     This document is a second 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.
                                     PREFACE

     National Ambient Air Quality Standards (NAAQS) are promulgated by the United States
Environmental Protection Agency (EPA) to meet requirements set forth in Sections 108 and 109
of the U.S. Clean Air Act (CAA). Sections 108 and 109 require the EPA Administrator (1) to
list widespread air pollutants that reasonably may be expected to endanger public health or
welfare; (2) to issue air quality criteria for them that assess the latest available scientific
information on nature and effects of ambient exposure to them; (3) to set "primary" NAAQS to
protect human health with adequate margin of safety and to set "secondary" NAAQS to protect
against welfare effects (e.g., effects on vegetation, ecosystems, visibility, climate, manmade
materials, etc); and (5) to periodically review and revise,  as appropriate, the criteria and NAAQS
for a given listed pollutant or class of pollutants.
     In 1971, the U.S. Environmental Protection Agency (EPA) promulgated National Ambient
Air Quality Standards (NAAQS) to protect the public  health and welfare from adverse effects of
photochemical oxidants. The EPA promulgates the NAAQS on the basis of scientific
information contained in air quality criteria issued under  Section 108 of the Clean Air Act.
Following the review of criteria as contained in the EPA  document, Air Quality Criteria for
Ozone and Other Photochemical Oxidants published in 1978, the chemical designation of the
standards was changed from photochemical oxidants to ozone (O3) in 1979 and a 1-hour O3
NAAQS was set. The 1978 document focused mainly on the air quality criteria for O3 and, to a
lesser extent, on those for other photochemical oxidants (e.g., hydrogen peroxide and the
peroxyacyl nitrates), as have subsequent revised versions of the ozone document.
                                          Il-ii

-------
     To meet Clean Air Act requirements noted above for periodic review of criteria and
NAAQS, the O3 criteria document, Air Quality Criteria for Ozone and Other Photochemical
Oxidants, was next revised and then released in August 1986; and a supplement, Summary of
Selected New Information on Effects of Ozone on Health and Vegetation, was issued in January
1992. These documents were the basis for a March 1993 decision by EPA that revision of the
existing 1-h NAAQS for O3 was not appropriate at that time. That decision, however, did not
take into account some newer scientific data that became available after completion of the  1986
criteria document.  Such literature was assessed in the next periodic revision of the O3 air quality
criteria document (completed in 1996) and provided scientific bases supporting the setting by
EPA in  1997 of an  8-h O3 NAAQS that is currently in force together with the 1-h O3 standard.
     The purpose of this revised air quality criteria document for O3 and related photochemical
oxidants is to critically evaluate and assess the latest scientific information published since that
assessed in the above 1996 Ozone Air Quality Criteria Document (O3 AQCD), with the main
focus being on pertinent new information useful in evaluating health and environmental effects
data associated with ambient air O3 exposures. However,  some other scientific data are also
presented and evaluated in order to provide a better understanding of the nature, sources,
distribution, measurement, and concentrations of O3 and related photochemical oxidants and
their precursors in the environment. The document mainly assesses pertinent literature published
or accepted for publication through 2004.
     The present Second Draft O3 AQCD (dated August 2005) is being released for public
comment and review by the Clean Air Scientific Advisory Committee (CASAC) to obtain
comments on the organization and structure of the document, the issues addressed, the
approaches employed in assessing and interpreting the newly available information on O3
exposures and effects, and the key findings and conclusions arrived at as a consequence of this
assessment.  Public comments and recommendations will be taken into account making any
appropriate further revisions to this document for incorporation into the final version of the
document to be completed and issued by February 28, 2006. Evaluations contained in the
present document will be drawn on to provide inputs to associated PM Staff Paper analyses
prepared by EPA's Office of Air Quality Planning and Standards (OAQPS)  to pose options for
consideration by the EPA Administrator with regard to proposal and, ultimately, promulgation of
decisions on potential retention or revision, as appropriate, of the current O3 NAAQS.

-------
     Preparation of this document was coordinated by staff of EPA's National Center for
Environmental Assessment in Research Triangle Park (NCEA-RTP). NCEA-RTP scientific
staff, together with experts from other EPA/ORD laboratories and academia, contributed to
writing of document chapters.  Earlier drafts of document materials were reviewed by non-EPA
experts in peer consultation workshops held by EPA. The document describes the nature,
sources, distribution, measurement, and concentrations of O3  in outdoor (ambient) and indoor
environments.  It also evaluates the latest data on human exposures to ambient O3 and
consequent health effects in exposed human populations, to support decision making regarding
the primary, health-related O3 NAAQS. The document also evaluates ambient O3 environmental
effects on vegetation and ecosystems, man-made materials, and surface level solar UV radiation
flux and global climate change, to support decision making on secondary O3 NAAQS.
     NCEA acknowledges the valuable contributions provided by authors, contributors, and
reviewers and the diligence of its staff and contractors in the preparation of this draft document.
                                         Il-iv

-------
           Air Quality Criteria for Ozone and Related
                   Photochemical Oxidants
                (Second External Review Draft)


                         VOLUME I


Executive Summary	E-l

1.   INTRODUCTION  	1-1

2.   PHYSICS AND CHEMISTRY OF OZONE IN THE ATMOSPHERE  	2-1

3.   ENVIRONMENTAL CONCENTRATIONS, PATTERNS, AND
    EXPOSURE ESTIMATES	3-1

4.   DOSIMETRY, SPECIES HOMOLOGY, SENSITIVITY, AND
    ANIMAL-TO-HUMAN EXTRAPOLATION	4-1

5.   TOXICOLOGICAL EFFECTS OF OZONE AND RELATED
    PHOTOCHEMICAL OXIDANTS IN LABORATORY ANIMALS
    AND IN VITRO TEST SYSTEMS  	5-1

6.   CONTROLLED HUMAN EXPOSURE STUDIES OF OZONE AND
    RELATED PHOTOCHEMICAL OXIDANTS  	6-1

7.   EPIDEMIOLOGICAL STUDIES OF HUMAN HEALTH EFFECTS
    ASSOCIATED WITH AMBIENT OZONE EXPOSURE	7-1

8.   INTEGRATIVE SYNTHESIS: EXPOSURE AND HEALTH EFFECTS	8-1

9.   ENVIRONMENTAL EFFECTS: OZONE EFFECTS ON
    VEGETATION AND ECOSYSTEMS  	9-1

10.  TROPOSPHERIC OZONE EFFECTS ON UV-B FLUX AND
    CLIMATE CHANGE PROCESSES 	10-1

11.  EFFECT OF OZONE ON MAN-MADE MATERIALS	11-1
                             II-v

-------
           Air Quality Criteria for Ozone and Related
                   Photochemical Oxidants
                (Second External Review Draft)
                           (cont'd)


                         VOLUME II


CHAPTER 2 ANNEX (ATMOSPHERIC PHYSICS/CHEMISTRY)  	AX2-1

CHAPTER 3 ANNEX (AIR QUALITY AND EXPOSURE)	AX3-1

CHAPTER 4 ANNEX (DOSIMETRY)  	AX4-1

CHAPTER 5 ANNEX (ANIMAL TOXICOLOGY) 	AX5-1

CHAPTER 6 ANNEX (CONTROLLED HUMAN EXPOSURE)	AX6-1

CHAPTER 7 ANNEX (EPIDEMIOLOGY)	AX7-1
                         VOLUME III


CHAPTER 9 ANNEX (ENVIRONMENTAL EFFECTS)	AX9-1
                            Il-vi

-------
                                  Table of Contents

                                                                                    Page

List of Tables	II-xiii
List of Figures  	II-xviii
Authors, Contributors, and Reviewers	II-xxix
U.S. Environmental Protection Agency Project Team for Development of Air Quality
      Criteria for Ozone and Related Photochemical Oxidants  	  II-xxxv
U.S. Environmental Protection Agency Science Advisory Board (SAB) Staff Office
      Clean Air Scientific Advisory Committee (CASAC) Ozone Review Panel  	II-xxxviii
Abbreviations and Acronyms	 II-xli

AX2.    PHYSICS AND CHEMISTRY OF OZONE IN THE ATMOSPHERE  	AX2-1
        AX2.1    INTRODUCTION	AX2-1
        AX2.2    TROPOSPHERIC OZONE CHEMISTRY	AX2-2
                 AX2.2.1     Atmospheric Structure	AX2-2
                 AX2.2.2     Overview of Ozone Chemistry  	AX2-3
                 AX2.2.3     Initiation of the Oxidation of VOCs 	AX2-6
                 AX2.2.4     Chemistry of Nitrogen Oxides in the Troposphere 	AX2-11
                 AX2.2.5     The Methane Oxidation Cycle	AX2-14
                 AX2.2.6     The Atmospheric Chemistry of Alkanes 	AX2-20
                 AX2.2.7     The Atmospheric Chemistry of Alkenes 	AX2-23
                 AX2.2.8     The Atmospheric Chemistry of Aromatic Hydrocarbons . . . .  AX2-31
                             AX2.2.8.1     Chemical Kinetics and Atmospheric
                                          Lifetimes of Aromatic Hydrocarbons	AX2-32
                             AX2.2.8.2     Reaction Products and Mechanisms
                                          of Aromatic Hydrocarbon Oxidation  	AX2-36
                             AX2.2.8.3     The Formation of Secondary Organic
                                          Aerosol as a Sink for Ozone Precursors	AX2-44
                 AX2.2.9     Importance of Oxygenated VOCs 	AX2-44
                 AX2.2.10    Influence of Multiphase Chemical Processes  	AX2-45
                             AX2.2.10.1    HOX and Aerosols	AX2-47
                             AX2.2.10.2    NOX Chemistry 	AX2-50
                             AX2.2.10.3    Halogen Radical Chemistry  	AX2-52
                             AX2.2.10.4    Reactions on the Surfaces of Crustal
                                          Particles	AX2-56
                             AX2.2.10.5    Reactions on the Surfaces of Aqueous
                                          H2SO4 Solutions 	AX2-57
                             AX2.2.10.6    Oxidant Formation in Particles	AX2-58
        AX2.3    PHYSICAL PROCESSES INFLUENCING THE ABUNDANCE
                 OF OZONE	AX2-59
                 AX2.3.1     Stratospheric-Tropospheric Ozone Exchange (STE)	AX2-61
                 AX2.3.2     Deep Convection in the Troposphere 	AX2-70
                             AX2.3.2.1     Observations of the Effects of
                                          Convective Transport	AX2-72
                             AX2.3.2.2     Modeling the Effects of Convection	AX2-75
                 AX2.3.3     Nocturnal Low-Level Jets 	AX2-78

-------
                          Table of Contents
                                (cont'd)
                                                                            Page
         AX2.3.4    Intercontinental Transport of Ozone and Other Pollutants  ... AX2-82
                    AX2.3.4.1     The Atmosphere/Ocean Chemistry
                                  Experiment, AEROCE 	AX2-82
                    AX2.3.4.2     The North Atlantic Regional Experiment,
                                  NARE	AX2-84
         AX2.3.5    The Relation of Ozone to Solar Ultraviolet Radiation,
                    Aerosols, and Air Temperature 	AX2-88
                    AX2.3.5.1     Solar Ultraviolet Radiation and Ozone	AX2-88
                    AX2.3.5.2     Impact of Aerosols on Radiation and
                                  Photolysis Rates and Atmospheric
                                  Stability	AX2-89
                    AX2.3.5.3     Temperature and Ozone  	AX2-89
AX2.4   THE RELATION OF OZONE TO ITS PRECURSORS AND
         OTHER OXIDANTS	AX2-93
         AX2.4.1    Summary of Results for the Relations Among Ozone,
                    its Precursors and Other Oxidants from Recent
                    Field Experiments 	AX2-96
                    AX2.4.1.1     Results from the Southern Oxidant Study
                                  and Related Experiments 	AX2-96
                    AX2.4.1.2     Results from Studies on Biogenic and
                                  Anthropogenic Hydrocarbons and
                                  Ozone Production	AX2-100
                    AX2.4.1.3     Results of Studies on Ozone Production
                                  in Mississippi and Alabama  	AX2-101
                    AX2.4.1.4     The Nocturnal Urban Plume Over
                                  Portland, Oregon	AX2-102
                    AX2.4.1.5     Effects of VOCs in Houston on
                                  Ozone Production	AX2-102
                    AX2.4.1.6     Chemical and Meteorological Influences
                                  on the Phoenix Urban Ozone Plume	AX2-103
                    AX2.4.1.7     Transport of Ozone and Precursors
                                  on the Regional Scale	AX2-103
                    AX2.4.1.8     Model Calculations and Aircraft
                                  Observations of Ozone Over
                                  Philadelphia	AX2-104
                    AX2.4.1.9     The Two-Reservoir System  	AX2-105
AX2.5   METHODS USED TO CALCULATE RELATIONS BETWEEN
         OZONE AND ITS PRECURSORS	AX2-105
         AX2.5.1    Chemistry-Transport Models	AX2-107
         AX2.5.2    Emissions of Ozone Precursors	AX2-121
         AX2.5.3    Observationally-Based Models  	AX2-128
         AX2.5.4    Chemistry-Transport Model Evaluation  	AX2-129

-------
                                Table of Contents
                                      (cont'd)
                                                                              Page
                           AXA.5.4.1    Evaluation of Emissions Inventories	AX2-141
                           AX2.5.4.2    Availability and Accuracy of Ambient
                                       Measurements	AX2-143
       AX2.6    TECHNIQUES FOR MEASURING OZONE AND ITS
                PRECURSORS 	AX2-144
                AX2.6.1     Sampling and Analysis of Ozone	AX2-144
                AX2.6.2     Sampling and Analysis of Nitrogen Oxides	AX2-147
                           AX2.6.2.1    Calibration Standards 	AX2-148
                           AX2.6.2.2    Measurement of Nitric Oxide	AX2-149
                           AX2.6.2.3    Measurements of Nitrogen Dioxide  	AX2-150
                           AX2.6.2.4    Monitoring for NO2 Compliance Versus
                                       Monitoring for Ozone Formation	AX2-151
                AX2.6.3     Measurements of Nitric Acid Vapor, HNO3  	AX2-151
                AX2.6.4     Sampling and Analysis of Volatile Organic
                           Compounds 	AX2-153
                           AX2.6.4.1    Polar Volatile Organic Compounds  	AX2-154
       REFERENCES	AX2-156

AX3.   ENVIRONMENTAL CONCENTRATIONS, PATTERNS, AND
       EXPOSURE ESTIMATES	AX3-1
       AX3.1    INTRODUCTION	AX3-1
       AX3.2    SURFACE OZONE CONCENTRATIONS  	AX3-4
                AX3.2.1     Nationwide Distribution of Metrics for Characterizing
                           Exposures of Vegetation to Ozone  	AX3-27
       AX3.3    SPATIAL VARIABILITY IN OZONE CONCENTRATIONS	AX3-39
                AX3.3.1     Spatial Variability of Ozone Concentrations in
                           Urban Areas  	AX3-40
                AX3.3.2     Small-scale Horizontal and Spatial Variability in
                           Ozone Concentrations 	AX3-55
                AX3.3.3     Ozone Concentrations at High Elevations	AX3-57
       AX3.4    DIURNAL PATTERNS IN OZONE CONCENTRATION	AX3-67
                AX3.4.1     Introduction	AX3-67
                AX3.3.2     Diurnal Patterns in Urban Areas  	AX3-69
                AX3.3.3     Diurnal Patterns in Nonurban Areas 	AX3-83
       AX3.5    SEASONAL VARIATIONS IN OZONE CONCENTRATIONS  	AX3-96
                AX3.5.1     Seasonal Variations in Urban Areas 	AX3-96
                AX3.5.2     Seasonal Variations in Nonurban Areas 	AX3-102
       AX3.6    TRENDS IN OZONE CONCENTRATIONS	AX3-103
       AX3.7    RELATIONS BETWEEN OZONE, OTHER OXIDANTS, AND
                OXIDATION PRODUCTS	AX3-114
       AX3.8    RELATIONSHIP BETWEEN SURFACE OZONE AND
                OTHER POLLUTANTS	AX3-121
                AX3.8.1     Introduction	AX3-121
                AX3.8.2     Co-Occurrence of Ozone with Nitrogen Oxides  	AX3-123


                                       Il-ix

-------
                          Table of Contents
                                (cont'd)
                                                                            Page
         AX3.8.3    Co-Occurrence of Ozone with Sulfur Dioxide  	AX3-125
         AX3.8.4    Co-Occurrence of Ozone and Daily PM25	AX3-126
         AX3.8.5    Co-Occurrence of Ozone with Acid Precipitation	AX3-128
         AX3.8.6    Co-Occurrence of Ozone with Acid Cloudwater	AX3-130
AX3.9    THE METHODOLOGY FOR DETERMINING POLICY
         RELEVANT BACKGROUND OZONE CONCENTRATIONS 	AX3-130
         AX3.9.1    Introduction	AX3-130
         AX3.9.2    Capability of Global Models to Simulate
                    Tropospheric Ozone	AX3-146
         AX3.9.3    Mean Background Concentrations:  Spatial and
                    Seasonal Variation	AX3-150
         AX3.9.4    Frequency of High-Ozone Occurrences at
                    Remote Sites  	AX3-152
AX3.10  OZONE EXPOSURE IN VARIOUS MICROENVIRONMENTS	AX3-160
         AX3.10.1   Introduction	AX3-160
         AX3.10.2   Summary of the Information Presented in the
                    Exposure Discussion in the 1996 Ozone
                    Criteria Document	AX3-161
         AX3.10.3   Concepts of Human Exposure	AX3-161
         AX3.10.4   Quantification of Exposure  	AX3-162
         AX3.10.5   Methods to Estimate Personal Exposure	AX3-162
                    AX3.10.5.1    Direct Measurement Method  	AX3-163
                    AX3.10.5.2    Indirect Measurement Method 	AX3-164
         AX3.10.6   Ozone Exposure Models 	AX3-166
                    AX3.10.6.1    Population Exposure Models  	AX3-172
                    AX3.10.6.2    Ambient Concentrations Models  	AX3-180
                    AX3.10.6.3    Microenvironmental Concentration
                                  Models 	AX3-181
         AX3.10.7   Measured Exposures and Monitored Concentrations 	AX3-185
                    AX3.10.7.1    Personal Exposure Measurements  	AX3-185
                    AX3.10.7.2    Monitored Ambient Concentrations	AX3-189
                    AX3.10.7.3    Ozone  Concentrations in
                                  Microenvironments	AX3-190
         AX3.7.4    Factors Affecting Ozone Concentrations Indoors	AX3-204
         AX3.10.8   Trends in Concentrations Within Microenvironments	AX3-216
         AX3.10.9   Characterization of Exposure	AX3-216
                    AX3.10.9.1    Use of Ambient Ozone Concentrations 	AX3-216
                    AX3.10.9.2    Exposure Selection in Controlled
                                  Exposure Studies	AX3-218
                    AX3.10.9.3    Exposure to Related Photochemical
                                  Agents	AX3-219
                                  II-x

-------
                                Table of Contents
                                     (cont'd)
                                                                              Page
AX4.   DOSIMETRY OF OZONE IN THE RESPIRATORY TRACT	AX4-1
       AX4.1    INTRODUCTION	AX4-1
       AX4.2    EXPERIMENTAL OZONE DOSIMETRY INVESTIGATIONS 	AX4-3
                AX4.2.1    Bolus-Response Studies	AX4-3
                AX4.2.2    General Uptake Studies	AX4-9
       AX4.3    DOSIMETRY MODELING	AX4-10
       AX4.4    SPECIES HOMOLOGY, SENSITIVITY AND ANIMAL-TO-
                HUMAN EXTRAPOLATION  	AX4-16
       REFERENCES	AX4-18

AX5.   ANNEX TO CHAPTER 5 OF OZONE AQCD	AX5-1
       REFERENCES	AX5-64

AX6.   CONTROLLED HUMAN EXPOSURE STUDIES OF OZONE AND
       RELATED PHOTOCHEMICAL OXIDANTS 	AX6-1
       AX6.1    INTRODUCTION	AX6-1
       AX6.2    PULMONARY FUNCTION EFFECTS OF OZONE EXPOSURE
                IN HEALTHY SUBJECTS	AX6-2
                AX6.2.1    Introduction	AX6-2
                AX6.2.2    Acute Ozone Exposures for Up to 2 Hours	AX6-3
                AX6.2.3    Prolonged Ozone Exposures  	AX6-10
                          AX6.2.3.1     Effect of Exercise Ventilation Rate
                                       on FEV] Response to 6.6 h Ozone
                                       Exposure	AX6-13
                          AX6.2.3.2     Exercise Ventilation Rate as a Function
                                       of Body/Lung Size on FEV] Response
                                       to 6.6 h Ozone Exposure	AX6-14
                          AX6.2.3.3     Comparison of 6.6 h Ozone Exposure
                                       Pulmonary Responses to Those
                                       Observed in 2 h Intermittent Exercise
                                       Ozone Exposures 	AX6-16
                AX6.2.4    Triangular Ozone Exposures	AX6-17
                AX6.2.5    Mechanisms of Pulmonary Function Responses	AX6-20
                          AX6.2.5.1     Pathophysiologic Mechanisms	AX6-22
                          AX6.2.5.2     Mechanisms at a Cellular and
                                       Molecular Level 	AX6-27
       AX6.3    PULMONARY FUNCTION EFFECTS OF OZONE EXPOSURE
                IN SUBJECTS WITH PREEXISTING DISEASE	AX6-28
                AX6.3.1    Subjects with Chronic Obstructive Pulmonary
                          Disease	AX6-33
                AX6.3.2    Subjects with Asthma	AX6-33
                AX6.3.3    Subjects with Allergic Rhinitis 	AX6-38
                AX6.3.4    Subjects with Cardiovascular Disease	AX6-39
                                       Il-xi

-------
                                Table of Contents
                                      (cont'd)
                                                                               Page
       AX6.4   INTERSUBJECT VARIABILITY AND REPRODUCIBILITY
                OF RESPONSE	AX6-40
       AX6.5   INFLUENCE OF AGE, GENDER, ETHNIC, ENVIRONMENTAL
                AND OTHER FACTORS	AX6-44
                AX6.5.1    Influence of Age  	AX6-44
                AX6.5.2    Gender and Hormonal Influences  	AX6-51
                AX6.5.3    Racial, Ethnic, and Socioeconomic Status Factors  	AX6-58
                AX6.5.4    Influence of Physical Activity	AX6-59
                AX6.5.5    Environmental Factors	AX6-60
                AX6.5.6    Oxidant-Antioxidant Balance  	AX6-65
                AX6.5.7    Genetic Factors  	AX6-68
       AX6.6   REPEATED EXPOSURES TO OZONE  	AX6-71
       AX6.7   EFFECTS ON EXERCISE PERFORMANCE	AX6-82
                AX6.7.1    Introduction	AX6-82
                AX6.7.2    Effect on Maximal Oxygen Uptake	AX6-82
                AX6.7.3    Effect on Endurance Exercise Performance	AX6-85
       AX6.8   EFFECTS ON AIRWAY RESPONSIVENESS  	AX6-86
       AX6.9   EFFECTS ON INFLAMMATION AND HOST DEFENSE  	AX6-98
                AX6.9.1    Introduction	AX6-98
                AX6.9.2    Inflammatory Responses in the Upper
                           Respiratory Tract	AX6-113
                AX6.9.3    Inflammatory Responses in the Lower
                           Respiratory Tract	AX6-115
                AX6.9.4    Adaptation of Inflammatory Responses	AX6-121
                AX6.9.5    Effect of Anti -Infl ammatory and Other
                           Mitigating Agents  	AX6-122
                AX6.9.6    Changes in Host Defense Capability Following
                           Ozone Exposure	AX6-124
       AX6.10  EXTRAPULMONARY EFFECTS  OF OZONE	AX6-127
       AX6.11  OZONE MIXED WITH OTHER POLLUTANTS	AX6-130
                AX6.11.1    Ozone and Sulfur Oxides	AX6-130
                AX6.11.2    Ozone and Nitrogen-Containing Pollutants	AX6-135
                AX6.11.3    Ozone and Other Pollutant Mixtures Including
                           Particulate Matter	AX6-138
       AX6.12  CONTROLLED STUDIES OF AMBIENT AIR EXPOSURES	AX6-139
                AX6.12.1    Mobile Laboratory Studies  	AX6-139
                AX6.12.2    Aircraft Cabin Studies	AX6-141
       REFERENCES	AX6-143

AX7.   EPIDEMIOLOGIC STUDIES OF HUMAN HEALTH EFFECTS
       ASSOCIATED WITH AMBIENT OZONE EXPOSURE	AX7-1
       AX7-1   Tables of Epidemiologic Studies of Human Health Effects
                Associated with Ambient Ozone Exposure	AX7-2
       AX7-2   Description of Summary Density Curves  	AX7-115

-------
                                     List of Tables

Number                                                                               Page

AX2-1       Comparison of the Atmospheric Lifetimes (T) of Low Molecular Weight
             Hydrocarbons Due to Reaction with OH, NO3, Cl, Br and O3  	AX2-7

AX2-2       Calculated Atmospheric Lifetimes of Biogenic Volatile Organic
             Compounds (adapted from Atkinson and Arey, 2003)  	AX2-25

AX2-3       Hydroxyl Rate Constants and Atmospheric Lifetimes of Mono- and
             Di-cyclic Aromatic Hydrocarbons (adapted from Atkinson 2000)	AX2-34

AX2-4       Chemistry-Transport Models (CTM) Contributing to the Oxcomp
             Evaluation of Predicting Tropospheric O3 and OH (Prather and
             Ehhalt, 2001)	AX2-120

AX2-5       Emissions of Nitrogen Oxides by Various Sources in the United States
             in 1999	AX2-121

AX2-6       Emissions of Volatile Organic Compounds by Various Sources in the
             United States in 1999	AX2-122

AX2-7       Emissions of Ammonia by Various Sources in the United States in 1999 	AX2-123

AX2-8       Emissions of Carbon Monoxide by Various Sources in the United States
             in 1999	AX2-124

AX3-1       Ozone Monitoring Seasons by State	AX3-3

AX3-2       Summary of Percentiles of Pooled Data Across Monitoring Sites for
             May to September 2000-2004  Concentrations are in ppb	AX3-11

AX3-3       Seasonal (April to October) Percentile Distribution of Hourly Ozone
             Concentrations (ppm), Number of Hourly Mean Ozone Occurrences
             >0.08 and >0.10, Seasonal 7-h Average Concentrations, SUM06, and
             W126 Values for Sites Experiencing Low Maximum Hourly Average
             Concentrations with Data Capture of >75%	AX3-18

AX3-4       The Top 10 Daily Maximum 8-h Average Concentrations (ppm) for Sites
             Experiencing Low Maximum Hourly Average Concentrations with Data
             Capture of >75%	AX3-23

AX3-5       Summary Statistics for Ozone  (in ppm) Spatial Variability in Selected
             U.S.  Urban Areas	AX3-42

AX3-6       Description of Mountain Cloud Chemistry Program Sites 	AX3-60

-------
                                     List of Tables
                                         (cont'd)

Number                                                                              Page

AX3-7       Seasonal (April-October) Percentiles, SUM06, SUM08, and W126 Values
             for the MCCP Sites  	AX3-61

AX3-8       Summary Statistics for 11 Integrated Forest Study Sites	AX3-65

AX3-9       Trends in Warm Season (May to September) Daily Maximum 8-h O3
             Concentrations at National Parks in the United States (1990 to 2004)	AX3-112

AX3-10      Range of Annual (January-December) Hourly Ozone Concentrations
             (ppb) at Background Sites Around the World (CMDL, 2004)  	AX3-136

AX3-11      Range of annual (January-December) Hourly Median and Maximum
             Ozone Concentrations (ppb) at Background Stations in Protected Areas
             of the United States (CASTNet, 2004)  	AX3-136

AX3-12      Range of annual (January-December) Hourly Median and Maximum Ozone
             Concentrations (ppb) at Canadian Background Stations (CAPMoN3, 2003) .  . . AX3-136

AX3-13      Number of Hours >0.05 ppm for Selected Rural O3 Monitoring in the
             United States by Month for the Period 1988 to 2001  	AX3-138

AX3-14      Number of Hours >0.06 ppm for Selected Rural O3 Monitoring Sites
             in the United States by Month for the Period of 1988 to 2001  	AX3-140

AX3-15      Global Budgets of Tropospheric Ozone (Tg year:) for the
             Present-day Atmosphere  	AX3-145

AX3 -16      Description of Simulations Used for Source Attribution
             (Fiore et al., 2003a)	AX3-149

AX3-17      Number of Hours with Ozone Above 50 or 60 ppbv at U.S. CASTNet
             Sites in 2001  	AX3-154

AX3-18      Activity Pattern Studies Included in the Consolidated Human Activity
             Database (CHAD)	AX3-174

AX3-19      Personal and Population Exposure Models for Ozone  	AX3-179

AX3-20      Personal Exposure Measurements	AX3-186

AX3-21      Indoor/Outdoor Ozone Ratios	AX3-191

AX3-22      Indoor and Outdoor O3 Concentrations in Boston, MA  	AX3-196

-------
                                     List of Tables
                                         (cont'd)

Number                                                                              Page

AX3-23      Indoor and Outdoor O3 Concentrations in Hong Kong	AX3-198

AX3-24      Indoor and Outdoor Ozone Concentrations	AX3-201

AX3-25      Rate Constants (Ir1) for the Removal of Ozone by Surfaces in Different
             Indoor Environments	AX3-210

AX4-1       New Experimental Human Studies on Ozone Dosimetry	AX4-4

AX4-2       New Ozone Dosimetry Model Investigations	AX4-12

AX5-1       Cellular Targets of Ozone Interaction	AX5-2

AX5-2       Effects of Ozone on Lung Monooxygenases	AX5-3

AX5-3       Antioxidants, Antioxidant Metabolism, and Mitochondrial
             Oxygen Consumption 	AX5-4

AX5-4       Lipid Metabolism and Content of the Lung  	AX5-5

AX5-5       Effects of Ozone on Protein Synthesis  	AX5-7

AX5-6       Effects of Ozone on Differential Gene Expression	AX5-8

AX5-7       Effects of Ozone on Lung Host Defenses	AX5-9

AX5-8       Effects of Ozone on Lung Permeability and Inflammation	AX5-18

AX5-9       Effects of Ozone on Lung Structure: Acute and Subchronic Exposures  	AX5-29

AX5-10      Effects of Ozone on Lung Structure: Subchronic and Chronic Exposures	AX5-33

AX5-11      Effects of Ozone on Pulmonary Function	AX5-36

AX5-12      Effects of Ozone on Airway Responsiveness 	AX5-38

AX5-13      Effects of Ozone on Genotoxicity/Carcinogenicity	AX5-45

AX5-14      Systemic  Effects of Ozone	AX5-46

AX5-15      Interactions of Ozone With Nitrogen Dioxide	AX5-53

AX5-16      Interactions of Ozone with Formaldehyde  	AX5-55
                                           II-xv

-------
                                      List of Tables
                                          (cont'd)

Number                                                                               Page

AX5-17      Interactions of Ozone with Tobacco Smoke	AX5-56

AX5-18      Interactions Of Ozone With Particles 	AX5-57

AX5-19      Effects of Other Photochemical Oxidants	AX5-63

AX6-1       Controlled Exposure of Healthy Humans to Ozone for 1 to 2 Hours
             During Exercise	AX6-5

AX6-2       Pulmonary Function Effects after Prolonged Exposures to Ozone	AX6-11

AX6-3       Ozone Exposure in Subjects with Preexisting Disease	AX6-29

AX6-4       Classification of Asthma Severity	AX6-37

AX6-5       Age Differences in Pulmonary Function Responses to Ozone 	AX6-45

AX6-6       Gender and Hormonal Differences in Pulmonary Function Responses
             to Ozone	AX6-52

AX6-7       Influence of Ethnic, Environmental, and Other Factors  	AX6-61

AX6-8       Changes in Forced Expiratory Volume in One Second After Repeated
             Daily Exposure to Ozone	AX6-72

AX6-9       Pulmonary Function Effects with Repeated Exposures to Ozone	AX6-73

AX6-10      Ozone Effects on Exercise Performance	AX6-83

AX6-11      Airway Responsiveness Following Ozone Exposures	AX6-88

AX6-12      Studies of Respiratory Tract Inflammatory Effects from Controlled
             Human Exposure to Ozone 	AX6-99

AX6-13      Studies of Effects on Host Defense, on Drug Effects and Supportive
             In Vitro Studies Relating to Controlled Human Exposure to Ozone	AX6-106

AX6-14      Ozone Mixed with Other Pollutants 	AX6-131

AX6-15      Acute Effects of Ozone in Ambient Air in Field Studies with a
             Mobile Laboratory	AX6-140

AX7-1       Effects of Acute O3 Exposure on Lung Function and Respiratory
             Symptoms in Field Studies  	AX7-3

-------
                                      List of Tables
                                          (cont'd)

Number                                                                               Page

AX7-2       Effects of Acute O3 Exposure on Cardiovascular Outcomes in
             Field Studies  	AX7-28

AX7-3       Effects of O3 on Daily Emergency Department Visits 	AX7-35

AX7-4       Effects of O3 on Daily Hospital Admissions	AX7-47

AX7-5       Effects of Acute O3 Exposure on Mortality	AX7-64

AX7-6       Effects of Chronic O3 Exposure on Respiratory Health  	AX7-95

AX7-7       Effects of Chronic O3 Exposure on Mortality and Incidence of Cancer	AX7-112

AX7-8       Ozone-Associated Cardiovascular Mortality Risk Estimates (95% CI)
             per Standardized Increment	AX7-121

-------
                                      List of Figures

Number                                                                                Page

AX2-1       Schematic overview of O3 photochemistry in the stratosphere
             and troposphere	AX2-4

AX2-2       General chemical mechanism for the oxidative degradation of VOCs	AX2-21

AX2-3       Hydroxyl radical initiated oxidation of a) propane and b) propene	AX2-24

AX2-4       Structures of a selected number of terpene and sesquiterpene compounds	AX2-27

AX2-5       Products from the reaction of terpenes with ozone	AX2-30

AX2-6       Initial steps in the photooxidation mechanism of toluene initiated by its
             reaction with OH radicals 	AX2-39

AX2-7a      Cross section through a tropopause folding event on March 13, 1978
             at 0000 GMT	AX2-63

AX2-7b      Ozone mixing ratios pphm (parts per hundred million) corresponding to
             Figure AX2-7A	AX2-66

AX2-7c      Condensation nuclei concentrations (particles cm"3) corresponding to
             Figure AX2-7a	AX2-67

AX2-8       Schematic diagram of a meteorological mechanism involved in high
             concentrations of O3 found in spring in the lower troposphere off the
             American east coast	AX2-70

AX2-9a,b    (a) Contour plot of CO mixing ratios (ppbv) observed in and near the
             June 15,  1985, mesoscale convective complex in eastern Oklahoma	AX2-74

AX2-10      The diurnal evolution of the planetary boundary layer while high pressure
             prevails over land	AX2-79

AX2-11      Locations of low level j et occurrences in decreasing order of prevalence
             (most frequent, common, observed) 	AX2-79

AX2-12      Schematic diagram showing the diurnal behavior of O3 and the
             development of secondary O3 maxima resulting from downward
             transport from the residual layer when a low-level jet is present  	AX2-81

AX2-13      The nocturnal low-level j et occupies a thin slice of the atmosphere near
             the Earth's surface	AX2-81

AX2-14      A scatter plot of daily maximum 8-h O3 concentration versus daily
             maximum temperature in the Baltimore, MD Air Quality Forecast Area	AX2-90

-------
                                      List of Figures
                                          (cont'd)

Number                                                                                 Page

AX2-15      A scatter plot of daily maximum 1-h average O3 concentration versus daily
             maximum temperature in the Baltimore, MD Air Quality Forecast Area	AX2-91

AX2-16      A scatter plot of daily maximum 8-h average O3 concentrations versus daily
             maximum temperature downwind of Phoenix, AZ	AX2-91

AX2-17      Measured values of O3 and NOZ (NOy - NOX) during the afternoon at rural
             sites in the eastern United States (gray circles) and in urban areas and urban
             plumes associated with Nashville, TN (gray dashes), Paris, FR (black
             diamonds) and Los Angeles, CA (X's)  	AX2-96

AX2-18      Conceptual two-reservoir model showing conditions in the PEL and in the
             lower free troposphere during a multi-day O3 episode  	AX2-106

AX2-19      Seasonal variability in O3 concentrations observed at a number of pressure
             surfaces at six ozonesonde sites and the predictions of 13 global scale
             chemistry-transport models	AX2-118

AX2-20      Seasonal variability in O3 concentrations observed at a number of pressure
             surfaces at six ozonesonde sites and the predictions of 13 global scale
             chemistry-transport models	AX2-119

AX2-21a,b   Impact of model uncertainty on control strategy predictions for O3 for
             two days (August 10[a] and ll[b], 1992) in Atlanta, GA	AX2-130

AX2-22      Ozone isopleths (ppb) as a function of the average emission rate
             for NOX and VOC (1012 molec. cm"2 s"1) in zero dimensional box
             model calculations	AX2-132

AX2-23a     Time series for measured gas-phase species in comparison with results
             from a photochemical model	AX2-133

AX2-23b     Time series for measured gas-phase species in comparison with results
             from a photochemical model	AX2-135

AX2-24      Correlations for O3 versus NOZ (NOy - NOX) in ppb from chemical
             transport models for the northeast corridor, Lake Michigan, Nashville,
             the San Joaquin Valley and Los Angeles  	AX2-137

AX2-25a,b   Evaluation of model versus measured O3 versus NOy for two model
             scenarios for Atlanta	AX2-138

AX2-26a,b   Evaluation of model versus: (a) measured O3 versus NOZ and
             (b) O3 versus the sum 2H2O2 + NOZ for Nashville, TN	AX2-139

-------
                                      List of Figures
                                          (cont'd)

Number                                                                                Page

AX2-27      Time series of concentrations of RO2, HO2, and OH radicals, local O3
             photochemical production rate and concentrations of NOX from
             measurements made during BERLIOZ	AX2-140

AX3-1       County wide mean daily maximum 8-h O3 concentrations, May to
             September 2002 to 2004 	AX3-6

AX3-2       County wide 95th percentile value of daily maximum 8-h O3 concentrations,
             May to September 2002 to 2004	AX3-7

AX3-3       Locations of monitoring sites used for calculating county wide averages
             across  the United States	AX3-8

AX3-4       Distribution of nationwide daily maximum 1-h average O3 concentrations
             from May to September 2000 to 2004	AX3-9

AX3-5       Distribution of nationwide daily maximum 8-h average O3 concentrations
             from May to September 2000 to 2004	AX3-9

AX3-6       Distribution of nationwide 24-h average O3 concentrations from May
             to September 2000 to 2004 	AX3-10

AX3-7       Box plots showing O3 averaged by month from 1993 to 2002 in the five
             regions in the eastern United States derived by Lehman et al. (2004)  	AX3-12

AX3-8       Hourly average O3 concentrations observed at selected rural-agricultural sites from April
             to October 2001	AX3-14

AX3-9       Hourly average O3 concentrations observed at selected rural-forest sites
             from April to October 2001	AX3-15

AX3-10      Hourly average O3 concentrations observed at selected rural-commercial
             or -residential sites from April to October 2001  	AX3-16

AX3-1 la-d   Daily 8-h maximum O3 concentrations observed at selected national
             park sites 	AX3-17

AX3-12      Seasonal SUM06 and W126 exposure indices for the Ouachita National
             Forest  for the period of 1991 to  2001  	AX3-27

AX3-13      Six-month (April to September) 24-h  cumulative W126 exposure index
             with the number of hourly average concentrations >0.10 ppm (N100)
             occurring during 2001 for the eastern United States	AX3-29
                                           II-xx

-------
                                     List of Figures
                                         (cont'd)

Number                                                                               Page

AX3-14      Six-month (April to September) 24-h cumulative SUM06 exposure index
             with the number of hourly average concentrations >0.10 ppm (N100)
             occurring during 2001 for the eastern United States	AX3-30

AX3-15      Six-month (April to September) 24-h cumulative W126 exposure index
             with the number of hourly average concentrations >0.10 ppm (N100)
             occurring during 2001 for the central United States	AX3-31

AX3-16      Six-month (April to September) 24-h cumulative SUM06 exposure index
             with the number of hourly average concentrations >0.10 ppm (N100)
             occurring during 2001 for the central United States	AX3-32

AX3-17      Six-month (April to September) 24-h cumulative W126 exposure index
             with the number of hourly average concentrations >0.10 ppm (N100)
             occurring during 2001 for the western United States 	AX3-33

AX3-18      Six-month (April to September) 24-h cumulative SUM06 exposure index
             with the number of hourly average concentrations >0.10 ppm (N100)
             occurring during 2001 for the western United States 	AX3-34

AX3-19      The 95% confidence interval for the 6-month (April to September) 24-h
             cumulative W126 exposure index for 2001 for the eastern United States	AX3-35

AX3-20      The 95% confidence interval for the 6-month (April to September) 24-h
             cumulative SUM06 exposure index for 2001 for the eastern United States	AX3-35

AX3-21      The 95% confidence interval for the 6-month (April to September) 24-h
             cumulative N100 exposure index for 2001 for the eastern United States	AX3-36

AX3-22      The 95% confidence interval for the 6-month (April to September) 24-h
             cumulative W126 exposure index for 2001 for the central United States	AX3-36

AX3-23      The 95% confidence interval for the 6-month (April to September) 24-h
             cumulative SUM06 exposure index for 2001 for the central United States  	AX3-37

AX3-24      The 95% confidence interval for the 6-month (April to September) 24-h
             cumulative N100 exposure index for 2001 for the central United States 	AX3-37

AX3-25      The 95% confidence interval for the 6-month (April to September) 24-h
             cumulative W126 exposure index for 2001 for the western United States	AX3-38

AX3-26      The 95% confidence interval for the 6-month (April to September) 24-h
             cumulative SUM06 exposure index for 2001 for the western United States	AX3-38

-------
                                      List of Figures
                                          (cont'd)

Number                                                                                 Page

AX3-27      The 95% confidence interval for the 6-month (April to September) 24-h
             cumulative N100 exposure index for 2001 for the western United States  	AX3-39

AX3-28      Locations of O3 sampling sites (a) by AQS ID# (b) and intersite correlation
             statistics (c) for the Charlotte, NC-Gastonia-Rock Hill, SC MSA  	AX3-43

AX3-29      Locations of O3 sampling sites (a) by AQS ID# (b) and intersite correlation
             statistics (c) for the Baton Rouge, LA MSA	AX3-44

AX3-30      Locations of O3 sampling sites (a) by AQS ID# (b) and intersite correlation
             statistics (c) for the Detroit-Ann Arbor-Flint, MI CMSA	AX3-45

AX3-31      Locations of O3 sampling sites (a) by AQS ID# (b) and intersite correlation
             statistics (c) for the St. Louis, MO-IL MSA  	AX3-46

AX3-32      Locations of O3 sampling sites (a) by AQS ID# (b) and intersite correlation
             statistics (c) for the Phoenix-Mesa, AZ MSA	AX3-48

AX3-33      Locations of O3 sampling sites (a) by AQS ID# (b) and intersite correlation
             statistics (c) for the Fresno, CA MSA	AX3-49

AX3-34      Locations of O3 sampling sites (a) by AQS ID# (b) and intersite correlation
             statistics (c) for the Bakersfield, CA MSA 	AX3-50

AX3-35      Locations of O3 sampling sites (a) by AQS ID# (b) and intersite correlation
             statistics (c) for the Los Angeles-Orange County, CA CMSA	AX3-51

AX3-36      Locations of O3 sampling sites (a) by AQS ID# (b) and intersite correlation
             statistics (c) for the Riverside-Orange County, CA CMSA	AX3-53

AX3-37      Vertical profile of O3 obtained over low vegetation  	AX3-57

AX3-38      Vertical profile of O3 obtained in a spruce forest 	AX3-58

AX3-39a-d   Seven- and 12-h seasonal means at (a) Whiteface Mountain and
             (b) Shenandoah National Park for May to September 1987, and integrated
             exposures at (c) Whiteface Mountain and (d) Shenandoah National Park
             for May to September 1987	AX3-63

AX3-40a-e   Integrated exposures for three non-Mountain Cloud Chemistry Program
             Shenandoah National Park sites, 1983 to 1987	AX3-64

AX3-41      Composite, nationwide diurnal variability in hourly averaged O3  in
             urban areas	AX3-69

-------
                                      List of Figures
                                          (cont'd)

Number                                                                                 Page

AX3-42       Composite, nationwide diurnal variability in 8 hour average O3 in
              urban areas	AX3-70

AX3-43a-f    Diurnal variability in hourly averaged O3 in selected urban areas  	AX3-71

AX3-43g-l    Diurnal variability in hourly averaged O3 in selected urban areas  	AX3-72

AX3-44a-f    Diurnal variability in 8 hour averaged O3 in selected urban areas  	AX3-74

AX3-44g-l    Diurnal variability in 8 hour averaged O3 in selected urban areas  	AX3-75

AX3-45a-d    Time of occurrence of daily maximum 1-h O3 concentration in four cities,
              averaged from April to October, 2000 to 2004	AX3-77

AX3-46a-d    Time of occurrence of daily maximum 8-h average O3 concentration in
              four cities, averaged from April to October, 2000 to 2004	AX3-78

AX3-47a-d    Diurnal variations in hourly averaged O3 on weekdays and weekends in
              four cities	AX3-80

AX3-47e-h    Diurnal variations in hourly averaged O3 on weekdays and weekends in
              four cities	AX3-81

AX3-48a-d    Diurnal variations in 8-h averaged O3 on weekdays and weekends in
              four cities	AX3-82

AX3-48e-h    Diurnal variations in hourly averaged O3 on weekdays and weekends in
              four cities	AX3-83

AX3-49a      Diurnal variations in hourly averaged O3 at a site in downtown Detroit, MI  .... AX3-84

AX3-49b      Diurnal variations in hourly averaged O3 at a site downwind of
              downtown Detroit 	AX3-84

AX3-50a      Diurnal variations in hourly averaged O3 at a site in downtown
              St. Louis, MO  	AX3-85

AX3-50b      Diurnal variations in hourly averaged O3 at a site downwind of downtown
              St. Louis	AX3-85

AX3-51a      Diurnal variations in hourly averaged O3 at a site in San Bernadino, CA	AX3-86

AX3-51b      Diurnal variations in hourly averaged O3 at a site in Riverside County
              well downwind of sources	AX3-86

-------
                                      List of Figures
                                          (cont'd)

Number                                                                                 Page

AX3-52a     Diurnal variations in 8-h average O3 at a site in downtown Detroit, MI	AX3-87

AX52-b      Diurnal variations in 8-h average O3 at a site downwind of downtown
             Detroit, MI	AX3-87

AX3-53a     Diurnal variations in 8-h average ozone at a site in downtown
             St. Louis, MO  	AX3-88

AX3-53b     Diurnal variations in 8-h average O3 at a site downwind of downtown
             St. Louis, MO
              	AX3-88

AX3-54a     Diurnal variations in 8-h average O3 at a site in San Bernadino, CA 	AX3-89

AX3-54b     Diurnal variations in 8-h average O3 at a site in Riverside County well
             downwind of sources	AX3-89

AX3-55      Composite diurnal variability in hourly O3 concentrations observed at
             CASTNET sites	AX3-90

AX3-56      Composite diurnal variability in 8-h average O3 concentrations observed
             at CASTNET sites	AX3-90

AX3-57      The comparison of the seasonal diurnal patterns for urban-influenced
             (Jefferson County, KY) and a rural-influenced (Oliver County, ND)
             monitoring sites using 2002 hourly data for April-October 	AX3-92

AX3-58a-d   Diurnal behavior of O3 at rural sites in the United States in July 	AX3-93

AX3-59      Composite diurnal O3 pattern at selected national forest sites in the
             United States using 2002 hourly average concentration data	AX3-94

AX3-60a,b   Composite diurnal pattern at (a) Whiteface Mountain, NY and (b) the
             Mountain Cloud Chemistry Program Shenandoah National Park site for
             May to  September 1987	AX3-95

AX3-61a-h   Seasonal variations in O3 concentrations as indicated by the 1-h maximum
             in each  month at selected sites, 2002	AX3-97

AX3-62a-f   Diurnal variability in 1-h average O3 concentrations in EPA's  12 cities  	AX3-98

AX3-62g-l   Diurnal variability in 1-h average O3 concentrations in EPA's  12 cities  	AX3-99

AX3-63a-f   Diurnal variability in 8-h average O3 concentrations in EPA's  12 cities  	AX3-100


                                           II-xxiv

-------
                                      List of Figures
                                           (cont'd)

Number                                                                                  Page

AX3-63g-l   Diurnal variability in 8-h average O3 concentrations in EPA's 12 cities  	AX3-101

AX3-64      Year-to-year variability in nationwide mean daily maximum 8-h
             O3 concentrations	AX3-104

AX3-65      Year-to-year variability in nationwide 95th percentile value of the daily
             maximum 8-h O3 concentrations	AX3-105

AX3-66a-h   Year-to-year variability in mean daily maximum 8-h O3 concentrations
             at selected national park (NP), national wildlife refuge (NWR), and
             national monument (NM) sites  	AX3-106

AX3-66i-p   Year-to-year variability in mean daily maximum 8-h O3 concentrations
             at selected national park (NP), national wildlife refuge (NWR), and
             national monument (NM) sites  	AX3-107

AX3-66q-v   Year-to-year variability in mean daily maximum 8-h O3 concentrations
             at selected national park (NP), national wildlife refuge (NWR), and
             national monument (NM) sites  	AX3-108

AX3-67a-h   Year-to-year variability in 95th percentile of daily maximum 8-h O3
             concentrations at selected national park (NP), national wildlife refuge
             (NWR), and national monument (NM) sites	AX3-109

AX3-67i-p   Year-to-year variability in 95th percentile of daily maximum 8-h O3
             concentrations at selected national park (NP), national wildlife refuge
             (NWR), and national monument (NM) sites	AX3-110

AX3-67q-v   Year-to-year variability in 95th percentile of daily maximum 8-h O3
             concentrations at selected national park (NP), national wildlife refuge
             (NWR), and national monument (NM) sites	AX3-111

AX3-68a-d   Measured O3 (ppbv) versus PAN (pptv) in Tennessee, including
             (a) aircraft measurements, and (b, c, and d) suburban sites near Nashville	AX3-116

AX3-69      Measured correlation between benzene and NOy at a measurement
             site in Boulder, CO 	AX3-117

AX3-70      Binned mean PM25 concentrations versus binned mean O3 concentrations
             observed at Fort Meade, MD from July 1999 to July 2001	AX3-120

AX3-71      The co-occurrence pattern for O3 and NO2 	AX3-123

AX3-72      The co-occurrence pattern for O3 and NO2 using 2001 data from the AQS  	AX3-124


                                            II-XXV

-------
                                     List of Figures
                                         (cont'd)

Number                                                                               Page

AX3-73      The co-occurrence pattern for O3 and SO2  	AX3-125

AX3-74      The co-occurrence pattern for O3 and SO2 using 2001 data from AQS	AX3-127

AX3-75      The co-occurrence pattern for O3 and PM2 5 using 2001 data from AQS  	AX3-127

AX3-76a     Monthly maximum hourly average O3 concentrations at Yellowstone
             National Park, Wyoming in 1998, 1999, 2000, and 2001	AX3-132

AX3-76b     Hourly average O3 concentrations at Yellowstone National Park,
             Wyoming for the period January to December 2001	AX3-132

AX3-77      (a) Contour plot of CO mixing ratios (ppbv) observed in and near the
             June 15, 1985, mesoscale convective complex in eastern Oklahoma	AX3-134

AX3-78      Maximum hourly average O3 concentrations at rural monitoring sites in
             Canada and the United States in February from 1980 to 1998  	AX3-142

AX3-79      Schematic diagram of a meteorological mechanism involved in high
             concentrations of O3 found in spring in the lower troposphere off the
             American East Coast	AX3-144

AX3-80      Ozone vertical profile at Boulder, Colorado on May 6, 1999 at
             1802UTC(1102LST)  	AX3-147

AX3-81      CASTNet stations in the continental United States for 2001  	AX3-151

AX3-82      Monthly mean afternoon (1300 to 1700 hours LT) concentrations (ppbv)
             in surface air averaged over the CASTNet stations (Figure AX3-81) in
             each U.S. quadrant for March to October 2001	AX3-152

AX3-83      Probability distributions of daily mean  afternoon (1300 to 1700 LT)
             O3 concentrations in surface air for March through October 2001 at
             U.S. CASTNet sites (Figure AX3-83):  observations (thick solid line)
             are compared with model results (thin solid line)	AX3-156

AX3-84      Daily mean afternoon (13 to 17 LT) O3 concentrations in surface air
             at Voyageurs National Park (NP), Minnesota in mid-May through
             June of 2001	AX3-157

AX3-85      Same as Figure AX3-85 but for Yellowstone National Park, Wyoming
             in March to May 2001	AX3-158
                                          II-xxvi

-------
                                      List of Figures
                                          (cont'd)

Number                                                                                Page

AX3-86      Same as Figure AX3-86 but for March of 2001 at selected western
             (left column) and southeastern (right column) sites  	AX3-159

AX3-87a     Detailed diagram illustrating components of an exposure model	AX3-170

AX3-87b     Detailed diagram illustrating components of an exposure model	AX3-171

AX3-88      Measured outdoor O3 concentrations (thin line) and modeled indoor
             concentrations (bold line) 	AX3-183

AX3-89      Air exchange rates and outdoor and indoor O3 concentrations during
             the summer at telephone switching station in Burbank, CA	AX3-199

AX3-90      Air exchange rates and outdoor and indoor O3 concentrations during
             the fall at a telephone switching station in Burbank, CA 	AX3-200

AX3-91      Diurnal variation of indoor and outdoor O3 and PAN concentrations
             measured in a private residence, Freising, Germany, August 11-12, 1995	AX3-202

AX3-92      Indoor and outdoor O3 concentration in moving cars 	AX3-203

AX3-93      Indoor/outdoor concentration ratios for PAN at 10 southern
             California museums	AX3-205

AX3-94      Ozone decay processes versus time measured for several indoor rooms 	AX3-211

AX6-1       FEV] decrements as a function of O3 concentration following a 2 h
             exposure with incremental exercise (15 min intervals) or rest  	AX6-8

AX6-2       Average FEVj decrements (±SE) for prolonged 6.6 h exposures to
             0.12 ppm O3 as afunction of exercise VE  	AX6-15

AX6-3       The forced expiratory volume in 1 s (FEVj) is shown in relation to
             exposure duration (hours) under three exposure conditions	AX6-18

AX6-4a,b     Recovery of spirometric responses following a 2 h exposure to 0.4 ppm
             O3 with IE  	AX6-22

AX6-5       Plot of the mean FEVj (% baseline) vs. time for ozone exposed cohorts	AX6-26

AX6-6       Frequency distributions of percent decrements in FEVj for 6.6-h exposure
             to four concentrations of ozone	AX6-42
                                          II-xxvii

-------
                                      List of Figures
                                           (cont'd)

Number                                                                                 Page

AX6-7        Effect of O3 exposure (0.42 ppm for 1.5 h with IE) on FEVj as a function
              of subject age	AX6-50

AX6-8        Regression curves were fitted to day-by-day postexposure FEV] values
              obtained after repeated daily acute exposures to O3 for 2 to 3 h with
              intermittent exercise at a VE of 24 to 43 L/min (adaptation studies)	AX6-81

AX7-1        Density curves of the O3-associated excess risk of cardiovascular mortality
              in the warm season per standardized increment (see Section 7.1.3.2)	AX7-122

-------
                       Authors, Contributors, and Reviewers


              CHAPTER 2 ANNEX (A TMOSPHERIC PHYSICS/CHEMISTR Y)

Principal Author

Dr. Joseph Pinto—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Dr. Russell Dickerson—University of Maryland, College Park, MD

Contributing Authors

Dr. Brooke Hemming—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Dr. Daniel Jacob—Harvard University, Cambridge, MA

Dr. William Keene—University of Virginia, Charlottesville, VA

Dr. Tadeusz Kleindienst—National Exposure Research Laboratory, U.S. Environmental
Protection Agency, Research Triangle Park, NC

Dr. Jennie Moody—University of Virginia, Charlottesville, VA

Mr. Charles Piety—University of Maryland, College Park, MD

Dr. Sandy Sillman—University of Michigan, Ann Arbor, MI

Dr. Jeffrey Stehr—University  of Maryland, College Park, MD

Dr. Bret Taubman—Pennsylvania State University, State College, PA

Contributors and Reviewers

Dr. Christoph Bruhl, Max Planck Institute for Atmospheric Chemistry, Mainz, Germany

Dr. Mohammed Elshahawy, Department of Meteorology and Astronomy, Cairo University, Giza,
Egypt.

Dr. Arlene Fiore, NOAA/GFDL, Princeton, NJ

Mr. Chris Geron, NRML, U.S. EPA, Research Triangle Park, NC

Dr. David Golden, Stanford University, Palo Alto, CA
                                        II-xxix

-------
                      Authors, Contributors, and Reviewers
                                       (cont'd)
Contributors and Reviewers
(cont'd)

Dr. John Merrill, University of Rhode Island, Kingston, RI

Dr. Sam Oltmans, NOAA, CMDL, Boulder, CO

Dr. David Parrish, NOAA/AL, Boulder, CO

Dr. Perry Samson, Depart. Atmos. Ocean, and Space Sciences, University of Michigan,
Ann Arbor, MI

Dr. Sandy Sillman, University of Michigan, Ann Arbor, MI

Dr. Melvin Shapiro, National Center for Atmospheric Research, Boulder, CO


                 CHAPTER 3 ANNEX (AIR QUALITY AND EXPOSURE)


Principal Authors

Ms. Beverly Comfort—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Dr. Joseph Pinto—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Dr. Arlene Fiore—NOAA/GFDL, Princeton, NJ

Dr. Daniel Jacob—Harvard University, Cambridge, MA

Dr. Alan S. Lefohn—ASL &Associates, Helena, MT

Dr. Clifford Weisel—Rutgers University, New Brunswick, NJ

Contributing Authors

Dr. Jee-Young Kim	National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Dr. Dennis Kotchmar—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
                                        II-XXX

-------
                       Authors, Contributors, and Reviewers
                                        (cont'd)


Contributing Authors
(cont'd)

Dr. Timothy Lewis—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Mr. Thomas McCurdy—U.S. EPA, NERL U.S. EPA, Research Triangle Park, NC

Contributors and Reviewers

Dr. Christoph Bruehl—Max Planck Institute for Atmospheric Chemistry, Mainz, Germany

Dr. Russell Dickerson—University of Maryland, College Park, MD

Dr. Judith Graham—American Chemistry Council, Washington, D.C.

Dr. Laszlo Horvath—Hungarian Meteorological Service, Budapest, Hungary

Dr. Ted Johnson—TRJ Associates, Durham, NC

Dr. John Merrill—University of Rhode Island, Kingston, RI

Dr. Jennie Moody—University of Virginia, Charlottesville, VA

Dr. Sam Oltmans—NOAA CMDL, Boulder, CO

Dr. Michiel G.M. Roemer, TNO, The Netherlands

Dr. Sandy Sillman—University of Michigan, Ann Arbor, MI

Dr. Tamas Weidinger—University of Budapest, Budapest, Hungary


                          CHAPTER 4 ANNEX (DOSIMETR Y)


Principal Authors

Dr. John Overton—U.S. Environmental Protection Agency, National Health and Environmental
Effects Research Laboratory-Research Triangle Park, NC 27711 (retired)

Dr. James S. Brown—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711


                                        II-xxxi

-------
                       Authors, Contributors, and Reviewers
                                        (cont'd)
Principal Authors
(cont'd)

Dr. Lori White—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC  27711

Contributors and Reviewers

Dr. Gary Hatch—U.S. Environmental Protection Agency, National Health and Environmental
Effects Research Laboratory, NC
                     CHAPTER 5 ANNEX (ANIMAL TOXICOLOGY)
Principal Authors
Dr. Lori White—National Center for Environmental Assessment (B243-01), U.S. Environmental
Protection Agency, Research Triangle Park, NC 27711

Mr. James Raub—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC  27711 (Retired)

Dr. Deepak Bhalla—Wayne State University, Detroit, MI

Dr. Carroll Cross—University of California, Davis, CA

Dr. Mitch Cohen—NYU School of Medicine, New York University, New York, NY

Contributors and Reviewers

Dr. Steven Kleeberger—National Institute of Environmental Health Sciences, Research
Triangle Park, NC 27711

Dr. George Liekauf—University of Cincinnati,  Cincinnati, OH

Dr. David Basset—Wayne State University, Detroit, MI

Dr. E.M. Postlethwait—University of Texas Medical Branch, Galveston, TX

Dr. Kent Pinkerton—University of California, Davis, CA
                                        II-xxxii

-------
                       Authors, Contributors, and Reviewers
                                        (cont'd)
Contributors and Reviewers
(cont'd)

Dr. Jack Harkema—Michigan State University, East Lansing, MI

Dr. Edward Schelegle—University of California, Davis, CA

Dr Judith Graham—American Chemical Council, Arlington, VA


               CHAPTER 6 ANNEX (CONTROLLED HUMAN EXPOSURE)


Principal Authors

Dr. James S. Brown—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Mr. James Raub—National Center for Environmental Assessment (B243-01), U.S. Environmental
Protection Agency, Research Triangle Park, NC  27711  (Retired)

Dr. William C. Adams—University of California, Davis, CA (Retired)

Dr. Milian J. Hazucha—University of North Carolina, Chapel Hill, NC

Dr. E. William Spannhake—Johns Hopkins University, Baltimore, MD

Contributors and Reviewers

Dr. Edward Avol—University of Southern California, Los Angeles, CA

Dr. Henry Gong—Ranches Los Amigos Medical Center, Los Angeles, CA

Dr. Jane Q. Koenig—University of Washington,  Seattle, WA

Dr. Michael Madden—National Health  and Environmental Effects Research Laboratory,
U.S. Environmental Protection Agency, Chapel Hill, NC

Dr.William McDonnell—National Health and Environmental Effects Research Laboratory,
U.S. Environmental Protection Agency, Chapel Hill, NC

-------
                       Authors, Contributors, and Reviewers
                                        (cont'd)
                        CHAPTER 7 ANNEX (EPIDEMIOLOGY)
Principal Authors
Dr. Dennis Kotchmar—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Dr. Jee-Young Kim—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Dr. David Svendsgaard—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Dr Kaz Ito—New York University, New York, NY

Dr. Pat Kinney—School of Public Health, Columbia University, New York, NY

Reviewers

Dr. Richard Burnett—Health Canada, Ottawa, Canada

Dr. Vic Hasselblad—Duke University, Durham, NC

Dr. Lucas Neas—National Health and Environmental  Effects Research Laboratory,
U.S. Environmental Protection Agency, Chapel Hill, NC
[Note: Any inadvertently omitted names of authors/reviewers will be inserted in the final draft of
this O3 AQCD, as will more complete addresses for all authors/reviewers.]
                                       II-xxxiv

-------
               U.S. Environmental Protection Agency Project Team
                for Development of Air Quality Criteria for Ozone
                       and Related Photochemical Oxidants
Executive Direction

Dr. Lester D. Grant (Director)—National Center for Environmental Assessment-RTF Division,
(B243-01), U.S. Environmental Protection Agency, Research Triangle Park, NC  27711

Scientific Staff

Dr. Lori White(Ozone Team Leader)—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC  27711

Dr. Joseph Pinto—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC  27711

Ms. Beverly Comfort—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC  27711

Dr. Brooke Hemming—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC  27711

Dr. James S. Brown—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC  27711

Dr. Dennis Kotchmar—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC  27711

Dr. Jee-Young Kim—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC  27711

Dr. David Svendsgaard—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC  27711

Dr. Srikanth Nadadur—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC  27711

Dr. Timothy Lewis—National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC  27711

Dr. William Hogsett—National Health and Environmental Effects Research Laboratory,
U.S. Environmental Protection Agency, Corvallis, OR
                                        II-XXXV

-------
               U.S. Environmental Protection Agency Project Team
                for Development of Air Quality Criteria for Ozone
                       and Related Photochemical Oxidants
                                        (cont'd)
Scientific Staff
(cont'd)

Dr. Christian Andersen—National Health and Environmental Effects Research Laboratory,
U.S. Environmental Protection Agency, Corvallis, OR

Dr. Jay Garner—National Center for Environmental Assessment (B243-01), U.S. Environmental
Protection Agency, Research Triangle Park, NC 27711 (retired)

Mr. Bill Ewald—National Center for Environmental Assessment (B243-01), U.S. Environmental
Protection Agency, Research Triangle Park, NC 27711 (retired)

Mr. James Raub—National Center for Environmental Assessment (B243-01), U.S. Environmental
Protection Agency, Research Triangle Park, NC 27711  (retired)

Technical Support Staff

Ms. Nancy Broom—Information Technology Manager, National Center for Environmental
Assessment (B243-01), U.S. Environmental Protection Agency, Research Triangle Park, NC  27711

Mr. Douglas B. Fennell—Technical Information Specialist, National Center for Environmental
Assessment (B243-01), U.S. Environmental Protection Agency, Research Triangle Park, NC  27711

Ms. Emily R. Lee—Management Analyst, National Center for Environmental Assessment
(B243-01), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Ms. Diane H. Ray—Program Specialist, National Center for Environmental Assessment
(B243-01), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711

Ms. Donna Wicker—Administrative Officer, National Center for Environmental Assessment
(B243-01), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711 (retired)

Mr. Richard Wilson—Clerk, National Center for Environmental Assessment (B243-01),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
                                        II-xxxvi

-------
               U.S. Environmental Protection Agency Project Team
                for Development of Air Quality Criteria for Ozone
                       and Related Photochemical Oxidants
                                       (cont'd)
Document Production Staff

Ms. Carolyn T. Perry—Manager, Computer Sciences Corporation, 2803 Slater Road, Suite 220,
Morrisville, NC 27560

Mr. John A. Bennett—Technical Information Specialist, Library Associates of Maryland,
11820 Parklawn Drive, Suite 400, Rockville, MD 20852

Mr. William Ellis—Records Management Technician, InfoPro, Inc., 8200 Greensboro Drive,
Suite 1450, McLean, VA  22102

Ms. Sandra L. Hughey—Technical Information Specialist, Library Associates of Maryland,
11820 Parklawn Drive, Suite 400, Rockville, MD 20852

Mr. Matthew Kirk—Graphic Artist, Computer  Sciences Corporation, 2803 Slater Road, Suite 220,
Morrisville, NC 27560

Dr. Barbara Liljequist—Technical Editor, Computer Sciences Corporation, 2803 Slater Road,
Suite 220, Morrisville, NC 27560

Ms. Faye Silliman—Word Processor, InfoPro, Inc., 8200 Greensboro Drive, Suite 1450,
McLean, VA  22102
                                       II-xxxvii

-------
      U.S. Environmental Protection Agency Science Advisory Board (SAB)
          Staff Office Clean Air Scientific Advisory Committee (CASAC)
                                 Ozone Review Panel
Chair

Dr. Rogene Henderson*, Scientist Emeritus, Lovelace Respiratory Research Institute, 2425
Ridgecrest Drive SE, Albuquerque, NM, 87108, Phone: 505-348-9464, Fax: 505-348-8541,
(rhenders@lrri.org) (FedEx: Dr. Rogene Henderson, Lovelace Respiratory Research Institute,
2425 Ridgecrest Drive SE, Albuquerque, NM, 87108, Phone: 505-348-9464)

Members

Dr. John Balmes, Professor, Department of Medicine, University of California San Francisco,
University of California - San Francisco, San Francisco, California, 94143, Phone: 415-206-8953,
Fax: 415-206-8949, (jbalmes@itsa.ucsf.edu)

Dr. Ellis Cowling*, University Distinguished Professor-at-Large, North Carolina State University,
Colleges of Natural Resources and Agriculture and Life Sciences, North Carolina State University,
1509 Varsity Drive, Raleigh, NC, 27695-7632, Phone: 919-515-7564 , Fax: 919-515-1700,
(ellis_cowling@ncsu.edu)

Dr. James D. Crapo*, Professor, Department of Medicine, National Jewish Medical and Research
Center. 1400 Jackson Street, Denver, CO, 80206, Phone: 303-398-1436, Fax: 303- 270-2243,
(crapoj @nj c. org)

Dr. William (Jim) Gauderman, Associate Professor, Preventive Medicine, University of
Southerm California, 1540 Alcazar #220, Los Angeles, CA, 91016, Phone: 323-442-1567,
Fax:  323-442-2349, (jimg@usc.edu)

Dr. Henry Gong, Professor of Medicine and Preventive Medicine, Medicine and Preventive
Medicine, Keck School of Medicine, University of Southern California, Environmental Health
Service, MSB 51, Rancho Los Amigos NRC, 7601 East Imperial Highway, Downey, CA, 90242,
Phone: 562-401-7561, Fax: 562-803-6883, (hgong@ladhs.org)

Dr. Paul J. Hanson, Senior Research and Development Scientist, Environmental  Sciences Division,
Oak Ridge National Laboratory (ORNL), Bethel Valley Road, Building 1062, Oak Ridge, TN,
37831-6422, Phone: 865-574-5361, Fax: 865-576-9939, (hansonpz@comcast.net)

Dr. JackHarkema, Professor, Department of Pathobiology, College of Veterinary Medicine,
Michigan State University, 212 Food Safety & Toxicology Center, East Lansing, MI, 48824,
Phone: 517-353-8627, Fax: 517-353-9902, (harkemaj@msu.edu)
                                       II-xxxviii

-------
      U.S. Environmental Protection Agency Science Advisory Board (SAB)
         Staff Office Clean Air Scientific Advisory Committee (CASAC)
                                Ozone Review Panel
                                        (cont'd)
Members
(cont'd)

Dr. Philip Hopke, Bayard D. Clarkson Distinguished Professor, Department of Chemical
Engineering, Clarkson University, Box 5708, Potsdam, NY, 13699-5708, Phone: 315-268-3861,
Fax: 315-268-4410, (hopkepk@clarkson.edu) (FedEx: 8 Clarkson Avenue, Potsdam, NY
136995708)

Dr. Michael T. Kleinman, Professor, Department of Community & Environmental Medicine, 100
FRF, University of California - Irvine, Irvine, CA, 92697-1825, Phone: 949-824-4765, Fax:
949-824-2070, (mtkleinm@uci.edu)

Dr. Allan Legge, President, Biosphere Solutions, 1601 11th Avenue NW, Calgary, Alberta,
CANADA, T2N 1H1, Phone: 403-282-4479, Fax: 403-282-4479, (allan.legge@shaw.ca)

Dr. Morton Lippmann, Professor, Nelson Institute of Environmental Medicine, New York
University School of Medicine, 57 Old Forge Road, Tuxedo, NY, 10987, Phone: 845-731-3558,
Fax: 845-351-5472, (lippmann@env.med.nyu.edu)

Dr. Frederick J. Miller*, Consultant, 911 Queensferry Road, Cary, NC, 27511, Phone:
919-467-3194, (fjmiller@nc.rr.com)

Dr. Maria Morandi, Assistant Professor of Environmental Science & Occupational Health,
Department of Environmental  Sciences, School of Public Health, University of Texas - Houston
Health Science Center,  1200 Herman Pressler Street, Houston, TX, 77030, Phone: 713-500-9288,
Fax: 713-500-9249, (mmorandi@sph.uth.tmc.edu) (FedEx: 1200 Herman Pressler, Suite 624)

Dr. Charles Plopper, Professor, Department of Anatomy, Physiology and Cell Biology, School of
Veterinary Medicine, University of California - Davis, Davis, California, 95616, Phone:
530-752-7065, (cgplopper@ucdavis.edu)

Mr. Richard L. Poirot*, Environmental Analyst, Air Pollution Control Division, Department of
Environmental Conservation, Vermont Agency of Natural Resources, Bldg. 3  South, 103 South
Main Street, Waterbury, VT, 05671-0402, Phone: 802-241-3807, Fax: 802-241-2590,
(rich.poirot@state.vt.us)

Dr. Armistead (Ted) Russell, Georgia Power Distinguished Professor of Environmental
Engineering, Environmental Engineering Group, School of Civil and Environmental Engineering,
Georgia Institute of Technology, 311 Ferst Drive, Room 3310, Atlanta, GA, 30332-0512, Phone:
404-894-3079, Fax: 404-894-8266, (trussell@ce.gatech.edu)
                                        II-xxxix

-------
      U.S. Environmental Protection Agency Science Advisory Board (SAB)
          Staff Office Clean Air Scientific Advisory Committee (CASAC)
                                 Ozone Review Panel

Members
(cont'd)

Dr. Elizabeth A. (Lianne) Sheppard, Research Associate Professor, Biostatistics and Environmental
& Occupational Health Sciences, Public Health and Community Medicine, University of
Washington, Box 357232, Seattle, WA, 98195-7232, Phone: 206-616-2722,  Fax: 206 616-2724,
(sheppard@u.washington.edu)

Dr. Frank Speizer*, Edward Kass Professor of Medicine, Channing Laboratory, Harvard Medical
School, 181 Longwood Avenue, Boston, MA, 02115-5804, Phone: 617-525-2275, Fax:
617-525-2066, (frank.speizer@channing.harvard.edu)

Dr. James Ultman, Professor, Chemical Engineering, Bioengineering program, Pennsylvania State
University, 106 Fenske Lab, University Park, PA, 16802, Phone: 814-863-4802, Fax:
814-865-7846, (jsu@psu.edu)

Dr. Sverre Vedal, Professor of Medicine, Department of Environmental and Occupational Health
Sciences, School of Public Health and Community Medicine, University of Washington, 4225
Roosevelt Way NE, Suite 100, Seattle, WA, 98105-6099, Phone: 206-616-8285, Fax:
206-685-4696, (svedal@u.washington.edu)

Dr. James (Jim) Zidek, Professor, Statistics, Science, University of British Columbia, 6856
Agriculture Rd., Vancouver, BC, Canada, V6T  1Z2, Phone: 604-822-4302, Fax: 604-822-6960,
(jim@stat.ubc.ca)

Dr. Barbara Zielinska*, Research Professor , Division of Atmospheric Science, Desert Research
Institute, 2215 Raggio Parkway, Reno, NV, 89512-1095, Phone:  775-674-7066, Fax:
775-674-7008, (barbz@dri.edu)

Science Advisory Board Staff

Mr. Fred Butterfield, CASAC Designated Federal Officer, 1200 Pennsylvania Avenue, N.W.,
Washington, DC, 20460, Phone: 202-343-9994, Fax: 202-233-0643 (butterfield.fred@epa.gov)
(Physical/Courier/FedEx Address: Fred A. Butterfield, III, EPA Science Advisory Board Staff
Office (Mail Code 1400F), Woodies Building, 1025 F Street, N.W., Room 3604, Washington,
DC 20004, Telephone: 202-343-9994)
*Members of the statutory Clean Air Scientific Advisory Committee (CASAC) appointed by
 the EPA Administrator
                                          II-xl

-------
                    ABBREVIATIONS AND ACRONYMS
AA
ACh
ADSS
AED
AER
AEROCE
AF
AGL
AHCs
AHR
AirPEX
AirQUIS
AIRS
ALI
AM
Ap
AP
AP-CIMS
APEX
APHEA
AQCD
ASC
A/V
BAL
BALF
BALT
B[a]P
BC
BEIS
ascorbic acid
acetylcholine
aged and diluted sidestream cigarette smoke
aerodynamic diameter
air exchange rate
Atmospheric/Ocean Chemistry Experiment
adsorbed fraction
above ground level
aromatic hydrocarbons
airway hyperreactivity
Air Pollution Exposure (model)
Air Quality Information System (model)
Aerometric Information Retrieval System
air-liquid interface
alveolar macrophage
peripheral lung
alkaline phosphatase
Atmospheric Pressure Chemical lonization Mass Spectrometer
Air Pollutants Exposure Model
Air Pollution on Health:  European Approach (study)
Air Quality Criteria Document
ascorbate
surface-to-volume ratio
bronchoalveolar lavage
bronchoalveolar lavage fluid
bronchus-associated lymphoid tissue
benzo[a]pyrene
black carbon
Biogenic Emission Inventory System

-------
BERLIOZ             Berlin Ozone Experiment
BHC                  biogenic hydrocarbons
BHR                  bronchial hyperresponsiveness
BME                  Bayesian Maxim Eutropy
BMZ                  basement membrane zone
BP                    blood pressure
BrdU                  bromodeoxyuridine
BS                    black smoke
BSA                  body surface area
BSA                  bovine serum albumin
C                     concentration
C x T                 concentration x time; concentration times duration of exposure
C2H2-H               ethane
C3a                   complement protein fragment
CAAA                Clean Air Act Amendments of 1990
CADS                Cincinnati Activity Diary Study
CAPs                 concentrated ambient particles
CAR                  centriacinar region
CASTNet, CASTNET   Clean Air Status and Trends Network
CAT                  cell antioxidant capacity
CBL                  convective boundary layer
CC16                 Clara cell secretory protein
CCh                  carbachol
CCSP                 Clara cell secretory protein
CE                    continuous exercise
CEPEX               Central Equatorial Pacific Experiment
CFD                  computational fluid dynamics
CG                   cloud-to-ground (flash)
CHAD                Consolidated Human Activities Database
CH2=C(CH3)-CHO     methacrolein

-------
CH3-CC13
CH3-CHO
CH3-CO
CH3-C(O)-CH=CH2
CH3-O(O)CH3
CH3-C(0)02,
CH3-O(O)OO
CH3O
CH3OOH
CH3O2
CH4
CI
CIMS
CINC
CIU, CBU
CL
CMAQ
CMDO
CMD
CMSA
CN
CNS
CO
CO2
COD
ConA
COPD
CTM
CYP
Cyt.
A
methyl chloroform
acetaldehyde
acetyl
methyl vinyl ketone
acetone
acetyl peroxy, peroxyacetyl

methoxy
methyl hydroperoxide
methyl peroxy
methane
confidence interval
Chemical lonization Mass Spectroscopy
cytokine-induced neutrophil chemoattractant
cumulative inhalation unit
chemiluminescence
Community Model for Air Quality
chl oromethy Ibutenone
count mean diameter
consolidated metropolitan statistical area
condensation nuclei
central nervous system
carbon monoxide
carbon dioxide
coefficient of divergence
concanavalin A
chronic obstructive pulmonary disease
chemistry transport model
cytochrome P-450
cytochrome
delta, mean change in a variable

-------
DA
DD
DHBA
DI
DIAL
DNA
DOAS
DOC
DPPC
DR
DTPA
EEC
ECG
EDMAS
ELF
EM
ENA
EOFs
EPA
EPEM
EPR
EPRI
ERAQS
ETS
EVR
F
F344
FA
FA
dry airstream
doubling dose
2,3-dehydroxybenzoic acid
dry intrusion
differential absorption lidar (system)
deoxyribonucleic acid
differential optical absorption spectroscopy
dissolved organic carbon
dipalmitoylglycero-3-phosphocholine
disulfide reductase
diethylenetriaminepentaacetic acid
exhaled breath condensate (fluid)
el ectrocardi ographi c
Exposure and Dose Modeling and Analysis System
epithelial lining fluid
electron microscopy
epithelial cell-derived neutrophil-activating peptide
empirical orthogonal  functions
U.S. Environmental Protection Agency
Event Probability Exposure Model
electron paramagnetic resonance
Electric Power Research Institute
Eastern Regional Air Quality Study
environmental tobacco smoke
equivalent ventilation rate
frequency of breathing
female
Fischer 344 (rat)
filtered air
fractional absorption; absorbed fraction

-------
FEF
FEF25.75
FEF
FIVC
Fn
FRC
FS
FTIR
FVC
GAM
GCE
GC-FID
GDI
GEE
GEOS-1 DAS
GEOS-CHEM

GLM
GM-CSF
G6PD
GR
GSH
GSHPx, GPx
GSSG
GST
GSTM1
GSTMlnull
FT
HCO
forced expiratory flow
forced expiratory flow between 25 and 75% of vital capacity
forced expiratory flow after X% vital capacity (e.g., after 50% vital
capcity)
forced expiratory volume in 1 second
forced inspiratory vital capacity
fibronectin
functional residual capacity
field stimulation
Fourier Transform Infrared Spectroscopy
forced vital capacity
General Additive Model
Goddard Cumulus Ensemble (model)
gas chromatography-flame ionization detection
glutathione-disulfide transhydrogenase
Generalized Estimating Equations
NASA Goddard Earth Orbiting System Data Assimilation System
three-dimensional model of atmospheric composition driven by
assimilated Goddard Earth Orbiting System observations
General Linear Model
granulocyte-macrophage colony stimulating factor
glucose-6-phosphate dehydrogenase
glutathione reductase
glutathione; reduced glutathione
glutathione peroxidase
glutathione disulfide
glutathione-^-transferase
glutathione S-transferase |i-l (genotype)
glutathione S-transferase |i-l null (genotype)
hydrogen ion; symbol for acid
formyl
                                        II-xlv

-------
H2CO, HCHO
HDMA
HF
H202
H2SO4
HCs
HHP-C9
HIST
HLA
HNE
HNO2
HNO3
HONO
HOONO
HR
HRV
5-HT
hv
IAS
IBM
1C
1C
ICAM
ICEM
ICS
ID#
IE
Ig
IL
IN
formaldehyde
house dust mite allergen
Rowland Forest
hydrogen peroxide
sulfuric acid
hydrocarbons
1 -hydroxy-1 -hydroperoxynonane
histamine
human lymphocyte antigen
4-hydroxynonenal
nitrous acid
nitric acid
nitrous acid
pernitrous acid
heart rate
heart rate variability
5 -hy droxytryptamine
solar ultraviolet proton
interalveolar septum
individual-based model or modeling
inspiratory capacity
intracloud (flash)
intracellular adhesion molecule
Indoor Chemistry and Exposure Model
inhaled steroids
identification number
intermittent exercise
immunoglobulin (e.g., IgA, IgE,  IgG, IgM)
interleukin (e.g., IL-1, IL-6, IL-8)
intranasal

-------
INF
inh
iNOS
I/O
ip
IPCC
IPMMI

IQR
ISCCP
IT
IU
iv
J(N02)
j(03)
Ka
KTB
LDH
LFHFR
LFT
LIF
LIS
LLJ
LM
LOESS
LPS
LT
LT
LSI
LWC
M
interferon
inhalation
inducible nitric oxide synthase
indoor-to-outdoor ratio
intraperitoneal
Intergovernmental Panel on Climate Change
International Photolysis Frequency Measurement and
Modeling Intercomparison
interquartile range
International Satellite Cloud Climatology Project
intratracheal
International Units
intravenous
photolysis rate coefficient for O3 to NO2
photolysis rate coefficient for O3 to O(1D)
intrinsic mass transfer coefficient/parameter
tracheobronchial region overall mass transfer coefficient
lactate dehydrogenase
low frequency/high frequency
lower free troposphere
laser-induced fluorescence
lateral intercellular space
low-level jet
light microscopy
locally estimated smoothing splines
lipopolysaccharide
leukotriene (e.g., LTB4, LTC4, LTD4, LTE4)
local time
local standard time
liquid water content
male

-------
MAQSIP
MBL
NBTH
MCCP
MCh
MCM
MCP
MENTOR
MET
MHC
MIESR
MIP
MLN
MM
MM5
MMAD
MoOx
MOZAIC

MPAN
mRNA
MS
MS
MSA
MS/MS
MT
n,N
N205
N/A
NAAQS
NADH
Multiscale Air Quality Simulation Platform
marine boundary layer
3-methyl-2-benzothiazolinone acetone azine
Mountain Cloud Chemistry Program
methacholine
master chemical mechanism
monocyte chemotactic protein
Modeling Environment for Total Risk Studies
metabolic equivalent of work
major histocompatibility
matrix isolation ESR spectroscopy
macrophage inflammatory protein
mediastinal lymph node
Mt. Mitchell
NCAR/Penn State Mesoscale Model
mass median aerodynamic diameter
molybdenum oxides
Measurement of Ozone and Water Vapor by Airbus
In-Service Aircraft
methacryloylperoxynitrate; peroxy-methacrylic nitric anhydride
messenger ribonucleic acid
mass spectrometry
Mt. Moosilauke
metropolitan statistical area
tandem mass spectrometry
metallothionein
number
dinitrogen pentoxide
not available
National Ambient Air Quality Standards
reduced nicotinamide adenine dinucleotide

-------
NADP
NADPH
NADPH-CR

NAG
NAPBM
NARE
NBS
NCAR
NCEA-RTP

NCICAS
NCLAN
ND
NEM
NESCAUM
NF
NF-KB
NH3
NH4+
NHAPS
NHBE
(NH4)2S04
NIST
NK
NL
NLF
NM
NMHCs
NMOCs
NMVOCs
National Atmospheric Deposition Program
reduced nicotinamide adenine dinucleotide phosphate
reduced nicotinamide adenine dinucleotide phosphate-
cytochrome c reductase
N-acetyl-p-J-glucosamine
National Air Pollution Background Network
North Atlantic Regional Experiment
National Bureau of Standards
National Center for Atmospheric Research
National Center for Environmental Assessment Division in
Research Triangle Park, NC
National Cooperative Inner-City Asthma Study
National Crop Loss Assessment Network
not detectable
National Ambient Air Quality Standards Exposure Model
Northeast States for Coordinated Air Use Management
national forest
nuclear factor kappa B
ammonia (gas)
ammonium ion
National Human Activity Pattern Survey
cultured human bronchial epithelial (cells)
ammonium sulfate
National Institute of Standards and Technology
natural killer (cells)
nasal lavage
nasal lavage fluid
national monument
nonmethane hydrocarbons
nonmethane organic compounds
nonmethane volatile organic compounds

-------
NO
NO2
NO3
NOAA
NOAELs
NOS
NOX
NOy
NOZ

NP
NQOlwt
NRC
NS
NS
NS
NSAID
NSBR
NTE
NTRMs
NWR
02
03
0(3P)
OAQPS
OEMs
OfD)
OH
8-OHdG
OLS
OPE
nitric oxide
nitrogen dioxide
nitrate
National Oceanic and Atmospheric Administration
non-observable-adverse-effect levels
nitric oxide synthase
nitrogen oxides
total reactive nitrogen; sum of NOX and NOZ; odd nitrogen species
nitrogen-containing species, the sum of the products of the
oxidation of NOX
national park
NAD(P)H-quinone oxidoreductase wild type (genotype)
National Research Council
national seashore
nonsignificant
nonsmoker
non-steroidal anti-inflammatory agent
nonspecific bronchial responsiveness
nasal trubinate epithelial (cells)
NIST Traceable Reference Materials
national wildlife refuge
superoxide
ozone
ground-state oxygen atom
Office of Air Quality Planning and Standards
observationally based methods
electronically excited oxygen atom
hydroxy
8-hydroxy-2'-deoxyguanosine
ordinary least squares
ozone production efficiency
                                         II-l

-------
OVA
Ox
P
PgO
PAF
PAHs
PAMS
PAN
Pa02
PAR
/7-ATP
PEL
PEL
PBM
PEN
PBPK
PC20
PC
   50
PCA
pCO2
PD,nnSRa,
PE
PEF
PEFR
PEM
PG
ovalbumin
odd oxygen species
probability
values of the 90th percentile absolute difference in concentrations
platelet-activating factor
polycyclic aromatic hydrocarbons
Photochemical Aerometric Monitoring System
peroxyacetyl nitrate
partial pressure of arterial oxygen
proximal alveolar region
/>ara-acetamidophenol
peripheral blood lymphocytes
planetary boundary layer
population-based model or modeling
C-phenyl N-tert-butyl nitrone
physiologically based pharmacokinetic (approach)
provocative concentration that produces a 20% decrease in forced
expiratory volume in 1 second
provocative concentration that produces a 50% decrease in forced
expiratory volume in 1 second
principal component analysis
partial pressure of carbon dioxide
provovative dose or concentration that produces a 20% decrease
in FEVj
provocative dose that produces a 100% increase in SRaw
provocative dose that produces a 20% decrease in forced expiratory
volume in 1 s
postexposure
peak expiratory flow
peak expiratory flow rate
personal exposure monitor
prostaglandin (e.g., PGD2, PGE, PGE^ PGE2 PGF1(X, PGF2a)
                                         Il-li

-------
6PGD
PHA
PIF
PM
PM
   10
PM25

PMNs
PND
pNEM

PNN50
POC
ppb
ppbv
pphm
ppm
PPN
PPPs
pptv
PRB
PTR-MS
PUL
PWM
r
R
 9
r
R2
RACM
6-phosphogluconate dehydrogenase
phytohemagglutinin
peak inspiratory flow
particulate matter
pressure at mouth at 0. 1 second of inspiration against a transiently
occluded mouthpiece, an index of inspiratory drive
combination of coarse and fine particulate matter (mass median
aerodynamic diameter <10 jim)
fine particulate matter (mass median aerodynamic diameter
<2.5
polymorphonuclear neutrolphil leukocytes; neutrophils
postnatal day
Probabilistic National Ambient Air Quality Standard Exposure
Model
proportion of adjacent NN intervals differing by more than 50 ms
particulate organic carbon
parts per billion
parts per billion by volume
parts per hundred million
parts per million
peroxypropionyl nitrate
power plant plumes
parts per trillion by volume
policy relevant background
proton-transfer-reaction mass spectroscopy
pulmonary
pokeweed mitogen
correlation coefficient
intraclass correlation coefficient
correlation coefficient
multiple correlation coefficient
Regional Air Chemistry Mechanism

-------
RADM
rALP
RAMS
RANTES
Raw
RB
RCO
RC(O)OO, RC(O)O2
RDBMS
REHEX
RER
RH
RIOPA
RL
RMR
rMSSD

RO2
ROI
RONO
RO2NO2
ROS
RR
RRMS
RT
RT
RT
PV
°g
s
SAC
Regional Acid Deposition Model
recombinant antileukoprotease
Regional Atmospheric Modeling System
regulated on activation, normal T cell-expressed and -secreted (cells)
airway resistance
respiratory bronchiole
acyl
peroxyacyl, acyl peroxy
Relational Database Management Systems
Regional Human Exposure Model
rough endoplasmic reticulum
relative humidity
Relationship of Indoor, Outdoor, and Personal Air (study)
total pulmonary resistance
resting metabolic rate
square root of the mean of the squared difference between adjacent
normal RR intervals
organic peroxy
reactive oxygen intermediate/superoxide anion
organic nitrate
peroxy nitrate
reactive oxygen species
relative risk
relatively remote monitoring sites
respiratory tract
total respiratory resistance
transepithelial resistance
potential vorticity
sigma-g, geometric standard deviation
smoker
Staphylococcus aureus Cowan 1 strain

-------
SAI
sao2
SAPRC

SAROAD

SAWgIp
sc
SC
SCAQS
SD, S-D
SD
SDNN
SE
SES
SGaw
SH
SHEDS
SO2
S042
SOD
SOS
SOX
SP
SP
SRBC
SRM
STE
STEP
Systems Applications International
oxygen saturation of arterial blood
Statewide Air Pollution Research Center, University of
California, Riverside
Storage and Retrieval of Aerometric Data (U.S. Environmental
Protection Agency centralized database; superseded by Aerometric
Information Retrieval System [AIRS])
small airway function
subcutaneous
stratum corneum
Southern California Air Quality Study
Sprague-Dawley
standard deviation
standard deviation around RR intervals
standard error
socioeconomic status
specific airway conductance
Shenandoah National Park
Simulation of Human Exposure and Dose System
sulfur dioxide
sulfate
superoxide dismutase
Southern Oxidant Study
sulfur oxides
substance P
surfactant protein (e.g., SP-A, SP-D)
specific airway resistance
sheep red blood cell
standard reference material
stratospheric-tropospheric exchange
Stratospheric-Tropospheric-Exchange Proj ect
                                        Il-liv

-------
STPD
STRF
SUM06
SUM07
SUM08
T
T3
T4
TAR
TB
TB
TEA
TEARS
99mTc-DTPA
T
xco
TOLAS
TEM
Tg
Ti
TLC
TLR
TNF
TOMS
TOPSE
TRIM
TRIM EXPO
TPLIF
TSH
TSP
TTFMS
TVA
standard temperature and pressure, dry
Spatio-Temporal Random Field
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
time (duration of exposure)
triiodothyronine
thyroxine
Third Assessment Report
terminal bronchiole
tracheobronchial (region)
thiobarbituric acid
thiobarbituric acid reactive substances
radiolabeled diethylenetriaminepentaacetic acid
core temperature
tunable-diode laser absorption spectroscopy
transmission electron microscopy
teragram
inspiratory time
total lung capacity
Toll-like receptor
tumor necrosis factor
Total Ozone Mapping Satellite; total ozone mapping spectrometer
Tropospheric Ozone Production About the Spring Equinox
Total Risk Integrated Methodology (model)
Total Risk Integrated Methodology Exposure Event (model)
two-photon laser-induced fluorescence
thyroid-stimulating hormone
total suspended particulate
two-tone frequency-modulated spectroscopy
Tennessee Valley Authority
                                        II-lv

-------
TWA
TX
UA
UAM
URT
UTC
UV
UV-A
UV-DIAL
VC
VCAM
VD
V,
  Emax
  max25%
  max50%
  max50%TLC
V;
v,
v,
v,
 * max75%
VMD
V02max
VOCs
VT
VT
vra
  Tmax
W126

WF, WFM
WT
WT
time-weighted average
thromboxane (A2, B2)
uric acid
Urban Airshed Model
upper respiratory tract
Coordinated Universal Time
ultraviolet
ultraviolet radiation of wavelengths 320 to 400 nm
Ultraviolet Differential Absorption Lidar
vital capacity
vascular cell adhesion molecule
anatomic dead space
minute ventilation; expired volume per minute
maximum minute ventilation
maximum expiratory flow at 25% of the vital capacity
maximum expiratory flow at 50% of the vital capacity
maximum expiratory flow at 50% of the total lung capacity
maximum expiratory flow at 75% of the vital capacity
volume mean diameter
maximal oxygen uptake (maximal aerobic capacity)
volatile organic  compounds
tidal volume
tracheal transepithelial potential
dose to tracheobronchial region
maximum tidal volume
cumulative integrated exposure index with a sigmoidal
weighting function
White Face Mountain
White Top Mountain
wild type

-------
 i       ANNEX AX2.  PHYSICS AND CHEMISTRY OF OZONE
 2                              IN THE ATMOSPHERE
 3
 4
 5     AX2.1   INTRODUCTION
 6          This annex (Annex AX2) provides detailed supporting information for Chapter 2 on the
 7     physics and chemistry of ozone (O3) in the atmosphere.  The organization of the material in this
 8     annex follows that used in prior Air Quality Criteria Documents, i.e., material is presented in
 9     sections and subsections. This annex provides material supporting Chapter 2 of the current draft
10     Air Quality Criteria Document for Ozone.
11          Section AX2.2 focuses on the chemistry of O3 formation. A very brief overview of
12     atmospheric structure is presented in Section AX2.2.1. An overview of O3 chemistry is given in
13     Section AX2.2.2. Information about reactive chemical species that initiate the oxidation of
14     VOCs is given in Section AX2.2.3. The chemistry of nitrogen oxides is then discussed briefly in
15     Section AX2.2.4. The oxidation of methane, the simplest hydrocarbon is outlined in
16     Section AX2.2.5.
17          The photochemical cycles leading to O3 production are best understood by considering the
18     oxidation of methane, structurally the simplest VOC.  The CH4 oxidation cycle serves as a model
19     which can be viewed as representing the chemistry of the relatively clean or unpolluted
20     troposphere (although this is a simplification because vegetation releases large quantities of
21     complex VOCs, such as isoprene, into the atmosphere).  Although the chemistry of the VOCs
22     emitted from anthropogenic and biogenic sources in polluted urban and rural areas is more
23     complex, a knowledge of the CH4 oxidation reactions aids in understanding the chemical
24     processes occurring in the polluted atmosphere because the underlying chemical principles are
25     the same.  The oxidation of more complex hydrocarbons (alkanes, alkenes, and aromatic
26     compounds) is discussed in Sections AX2.2.6,  AX2.2.7, and AX2.2.8, respectively. The
27     chemistry of oxygenated species is addressed in Section AX2.2.9.  Greater emphasis is placed on
28     the oxidation of aromatic hydrocarbons in this  section because of the large amount of new
29     information available since the last Air Quality Criteria for Ozone document (AQCD 96) was
30     published (U.S. Environmental Protection Agency, 1996) and because of their importance in O3
31     formation in polluted areas. Multiphase chemical processes influencing O3 are discussed in

       August  2005                             AX2-1      DRAFT-DO NOT QUOTE OR CITE

-------
 1      Section AX2.2.10. Meteorological processes that control the formation of O3 and other oxidants
 2      and that govern their transport and dispersion, and the sensitivity of O3 to atmospheric
 3      parameters are given in Section AX2.3. Greater emphasis is placed on those processes for which
 4      a large amount of new information has become available since AQCD 96. The role of
 5      stratospheric-tropospheric exchange in determining O3 in the troposphere is presented in Section
 6      AX2.3.1.  The importance of deep convection in redistributing O3 and its precursors and other
 7      oxidants throughout the troposphere is given in  Section AX2.3.2.  The possible importance of
 8      nocturnal  low-level jets in transporting O3 and other pollutants is presented in Section AX2.3.3.
 9      Information about the mechanisms responsible for the intercontinental transport of pollutants and
10      for the interactions between stratospheric-tropospheric exchange and convection is given in
11      Section AX2.3.4. Much of the material in this section is based on results of field programs
12      examining atmospheric chemistry over the North Atlantic ocean.  The sensitivity of O3 to solar
13      ultraviolet radiation and temperature is given in Section AX2.3.5. The relations of O3 to its
14      precursors and to other oxidants based on field and modeling studies are discussed in Section
15      AX2.4. Methods used to calculate relations between O3 its precursors and other oxidants are
16      given in Section AX2.5.  Chemistry-transport models are discussed in Section AX2.5.1.
17      Emissions of O3 precursors are presented in Section AX2.5.2. Issues related to the evaluation of
18      chemistry-transport models and emissions inventories are presented in Section AX2.5.3.
19      Measurement methods are summarized in Section AX2.6. Methods used to monitor ground-
20      level O3 are given in Section AX2.6.1, NO and NO2 in Section AX2.6.2, HNO3 in Section
21      AX2.6.3 and some important VOCs in Section AX2.6.4.
22
23
24      AX2.2   TROPOSPHERIC OZONE  CHEMISTRY
25      AX2.2.1  Atmospheric  Structure
26          The atmosphere can be divided into several distinct vertical layers, based  primarily on the
27      major mechanism by which that portion of the atmosphere is heated or cooled.  The lowest major
28      layer is the troposphere, which extends from the earth's surface to about 8 km above polar
29      regions and to about 16 km above tropical regions.  The troposphere is heated by convective
30      transport from the surface, and by the absorption of infrared radiation emitted by the surface,
31      principally by water vapor and CO2. The planetary boundary layer (PEL) is the sublayer of the

        August 2005                             AX2-2       DRAFT-DO NOT QUOTE OR CITE

-------
 1     troposphere that mixes with surface air on time scales of a few hours or less.  It typically extends
 2     to 1-2 km altitude and is often capped by a temperature inversion. The sublayer of the
 3     troposphere above the PEL is called the free troposphere.  Ventilation of the PEL with free
 4     tropospheric air takes place on a time scale of a week.  Vertical mixing of the whole troposphere
 5     takes place on a time scale of a month or two. The stratosphere extends from the tropopause, or
 6     the top of the troposphere, to about 50 km in altitude. The upper stratosphere is heated by the
 7     absorption of solar ultraviolet radiation by O3, while dissipation of wave energy transported
 8     upwards from the troposphere is a primary heating mechanism in the lower stratosphere.
 9     Heating of the stratosphere is balanced by radiative cooling due to infrared emissions to space
10     by CO2, H2O, and O3. As a result of heating of the upper stratosphere, temperatures increase
11     with height, inhibiting vertical mixing. A schematic overview of the major chemical cycles
12     involved in O3 formation and destruction in the stratosphere and troposphere is shown in Figure
13     AX2-1.  The figure emphasizes gas phase processes, but the importance of multiphase processes
14     is becoming apparent. The sequences of reactions shown in the lower right quadrant of the
15     figure will be discussed in Section AX2.2. The reader is referred to any of the large number of
16     texts on atmospheric chemistry, such as Wayne (2000) or Seinfeld and Pandis (1998), for an
17     introduction to stratospheric photochemistry, including the impact of O3-destroying compounds.
18
19     AX2.2.2  Overview of Ozone Chemistry
20           Ozone is found not only in polluted urban atmospheres but throughout the troposphere,
21     including remote areas of the globe.  Even without ground-level production, some O3 would be
22     found in the troposphere due to downward transport from the stratosphere.  Tropospheric
23     photochemistry leading to the formation of O3 and other photochemical air pollutants is
24     complex, involving thousands of chemical reactions and thousands of stable and reactive
25     intermediate products. Other photochemical oxidants, such as peroxyacetyl nitrate (PAN), are
26     among the reactive products. Ozone can be photolyzed in the presence of water to form
27     hydroxyl radical (OH), which is responsible for the oxidation of NOX and SOX to form
28     nitric (HNO3) and sulfuric acid (H2SO4), respectively.  Ozone participates directly in the
29     oxidation of unsaturated hydrocarbons, via the ozonolysis mechanism, yielding secondary
30     organic compounds that contribute to aerosol formation and mass, as well as formaldehyde
31     (H2CO) and other carbonyl compounds, such as aldehydes and ketones.

       August 2005                             AX2-3       DRAFT-DO NOT QUOTE OR  CITE

-------
                                           Stratosphere
                                             Non-Pola
                                             Regions
      Figure AX2-1.  Schematic overview of O3 photochemistry in the stratosphere
                      and troposphere.
1
2
3
4
5
6
7
     There is a rapid photochemical cycle in the troposphere that involves the photolysis of
nitrogen dioxide (NO2) by solar UV-A radiation to yield nitric oxide (NO) and a ground-state
oxygen atom, O(3P),
                           NO2 + hv -» NO + O(3P),
                          (AX2-1)
O(3P) then reacts with molecular oxygen to form O3:  A molecule from the surrounding air
collides with the newly-formed O3 molecule, removing excess energy to allow it to stabilize.
      August 2005
                                        AX2-4
DRAFT-DO NOT QUOTE OR CITE

-------
                       O(3P) + O2 + M -» O3 + M, where M = an air molecule.                (AX2-2)
 1
 2      Reaction AX2-2 is the only significant reaction forming O3 in the troposphere.
 3
 4      NO and O3 react to reform NO2:
 5
 6
                                                     +  2.                            (AX2-3)
 7
 8
 9      This reaction is responsible for O3 decreases found near sources of NO (e.g., highways)
10      especially at night.  The oxidation of reactive VOCs leads to the formation of reactive radical
1 1      species that allow the conversion of NO to NO2 without the participation of O3 (as in
12      reaction AX2-3).
13
                                         HO-,% RO2-
                                  NO - = - >  N02.                            (AX2.4)
14
15      O3 can, therefore, accumulate as NO2 photolyzes as in reaction AX2-1 followed by reaction
16      AX2-2.
17           It is often convenient to speak about families of chemical species, that are defined in terms
18      of members which interconvert rapidly among themselves on time scales that are shorter than
19      that for formation or destruction of the family as a whole. For example, an "odd oxygen" (Ox)
20      family can be defined as £ (O(3P) +O(1D)+ O3 + NO2) in much the same way as the NOX
21      (NO + NO2) family is defined.  We can then see that production of Ox occurs by the schematic
22      reaction AX2-4, and that the sequence of reactions given by reactions AX2-1 through AX2-3
23      represents no net production of Ox. Definitions of species families and methods for constructing
24      families are discussed in Jacobson (1999) and references therein.  Other families that include
25      nitrogen containing species, and will be referred to later in this chapter, are NOZ which is the
26      sum of the products of the oxidation of NOX = £ (HNO3 + PAN (CH3CHO-OO-NO2) + HNO4 +
27      other organic nitrates + particulate nitrate); and NOy, which is the sum of NOX and NOZ.

        August 2005                             AX2-5       DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX2.2.3   Initiation of the Oxidation of VOCs
 2           The key reactive species in the troposphere is the OH radical. OH radicals are
 3      responsible for initiating the photochemical oxidation of CO and most anthropogenic and
 4      biogenic VOCs, including those responsible for depleting stratospheric O3 (e.g., CH3Br,
 5      hydroclorofluorocarbons), and those which contribute to the greenhouse effect (e.g., CH4).
 6      Because of their role in removing so many potentially damaging species, OH radicals have
 7      sometimes been referred to as the atmosphere's detergent.  In the presence of NO, reactions of
 8      OH with VOCs lead to the formation of O3. In addition to OH radicals, there are several other
 9      atmospheric species such as NO3, Cl, and Br radicals and O3 that are capable of initiating VOC
10      oxidation.  Rate coefficients and estimated atmospheric lifetimes (the e-folding time) for
11      reactions of a number of alkanes, alkenes and dienes involved in O3 formation with these
12      oxidants at concentrations characteristic of the relatively unpolluted planetary boundary layer are
13      given in Table AX2-1. As can be seen from Table AX2-1, there is a wide range of lifetimes
14      calculated for the different species. However, under certain conditions the relative importance of
15      these oxidants can change from those shown in the table.  For hydrocarbons whose atmospheric
16      lifetime is  much longer than a day, diurnally averaged concentrations of oxidant concentrations
17      can be used, but for those whose lifetime is much shorter than a day it is more appropriate to use
18      either daytime or night-time averages depending on when the oxidant is at highest
19      concentrations.  During these periods, these averages are of the order of twice the values used in
20      Table AX2-1.
21           The main source of OH radicals is the photolysis of O3 by solar ultraviolet radiation at
22      wavelengths < 340 nm (solar radiation at wavelengths < 320 nm is also referred to as UV-B) to
23      generate electronically excited O(1D) atoms (Jet Propulsion Laboratory, 2003),
24
                                    O3 + hv -> O2 + O(1D).                             (AX2-5)

25
26      The O(JD) atoms can either be deactivated to the ground state O(3P) atom by collisions with N2
27      and O2, or they react with water vapor to form two OH radicals:
28
29                                  O(1D) + H2O ->• 2(-OH)                            (AX2-6)

        August 2005                             AX2-6      DRAFT-DO NOT QUOTE OR CITE

-------
        Table AX2-1. Comparison of the Atmospheric Lifetimes (u) of Low Molecular Weight Hydrocarbons Due to Reaction with
                                                   OH, NO3, Cl, Br and O3
r+
to
o
o





X
to

o
^
H
6
o
0
H
/O
o
H
W
O
A, cm3 molecule"1 s"1
OH
Hydrocarbon
Alkanes
Ethane
Propane
2-Methylpropane
w-Butane
2-Methylbutane
w-Pentane
2,2-Dimethylbutane
2,3 -Dimethylbutane
2-Methylpentane
3-Methylpentane
w-Hexane
2,2,4-Trimethylpentane


k x 1012

0.24
1.1
2.1
2.3
4
3.8
2.7
6.4
5.6
5.8
5.2
3.8


T

48 d
lid
5.6 d
5.2 d
2.9 d
3.0 d
4.3d
1.8d
2.1 d
2.0 d
2.2 d
3.0 d


NO3
k xlO12

< 1.0x10-'
0.00021
< 0.00007
0.000046
0.00016
0.000081
NA
0.00041
0.000017
0.00002
0.00011
0.000075


T

>13y
>0.60
>18y
2.8 y
0.79 y
1.6 y
NA
110 d
7.5 y
6.3 y
1.2 y
1.7 y


Cl
k x 1010

0.57
1.3
1.3
2.3
2
2.5
NA
2
2.5
2.5
3.1
2.3


T

6.7 mo
90 d
90 d
50 d
60 d
46 d
NA
60 d
47 d
46d
38 d
50 d


Br
k x 1012 T

3.1 x 1(T7 l.OxlOV
0 6.5xl03y2
<1.0x >3.2xl07y2
<1.0x >3.2xl07y2
NA NA
NA NA
NA NA
0.0064 50 y
NA NA
NA NA
NA NA
0.0068 47 y


03
k x 1018

<0.01
<0.01
<0.01
<0.01
NA
NA
NA
NA
NA
NA
NA
NA


T

>3.2y
>3.2y
>3.2y
>3.2y
NA
NA
NA
NA
NA
NA
NA
NA


O

-------
             Table AX2-1 (cont'd). Comparison of the Atmospheric Lifetimes (u) of Low Molecular Weight Hydrocarbons Due to
                                                     Reaction with OH, NO3, Cl, Br and O3
r+
to
o
o




X
to
oo

o
§
H
6
o
0
H
O
o
H
W
O
k, cm3 molecule"1 s"1
OH
Hydrocarbon
Ethene
Propene
2-Methylpropene
1-Butene
/ra«s-2-Butene
c/s-2-Butene
1,3 -Butadiene
Isoprene
2-Methyl-2-butene
1-Pentene
/ra«s-2-Pentene
c/s-2-Pentene
2,4,4-Trimethyl-l-pentene
k x 1012
8.5
26
51
31
64
56
67
100
87
31
67
65
65
T
33 h
llh
5.4 h
9.0 h
4.3 h
5.0 h
4.1 h
2.8 h
3.2 h
9.0 h
4.1 h
4.3 h
4.3 h
NO3
k x 1012
0
0.01
0.34
0.013
0.39
0.35
0.1
0.68
9.4
0.7
1.6
1.4
0.51
T k
230 d
4.9 d
3.3h
3.6 d
2.8 h
3.2 h
llh
1.6 h
0.12 h
1.6 h
0.69 h
0.79 h
2.2 h
Cl
xlO10
0.99
2.3
0.42
1.4
NA
NA
4.2
5.1
NA
NA
NA
NA
NA
T
3.8m
50 d
9.0m
65 d
NA
NA
28 d
23d
NA
NA
NA
NA
NA
Br
k x 1012
0.18
5.3
NA
3.4
0.23
6.3
57
74
19
NA
NA
NA
NA
T
1.8 y
22 d
NA
34 d
1.4 y
18 d
2.0 d
1.6 d
6.1 d
NA
NA
NA
NA
03
k x 1018
1.6
10
11
9.6
190
125
6.3
13
400
11
320
210
NA
Notes: NA = Reaction rate coefficient not available. Rate coefficients were calculated at 298k and 1 atmosphere, y = year, d = day.
OH = 1 x lOVcm3; NO3 = 2.5 x 108/cm3; Cl = 1 x 103/cm3; Br = 1 x 105/cm3; O3 = 1 x 1012/cm3. Value for Br calculated based on equilibrium with BrO
1 Rate Coefficients were Obtained from the NIST Online Kinetics Database for Reactions of Alkanes and for all Cl and Br Reactions.

T
7.2 d
1.2 d
Lid
1.2 d
1.5 h
2.3 h
1.8 d
21 h
0.69 h
Lid
0.86 h
1.3 h
NA
= 1 ppt.
         All Other Rate Coefficients were Obtained from the Evaluation
Q       ofCalvertetal. (2000).
d      2 Lifetimes should be regarded as lower limits.

        Sources: NIST online kinetics database (http://kinetics.nist.gov/index.php).

-------
 1      The O(3P) atoms formed directly in the photolysis of O3 in the Huggins and Chappuis bands or
 2      formed from deactivation of O(1D) atoms reform O3 through reaction AX2-2. Hydroxyl radicals
 3      produced by reactions AX2-5 and AX2-6 can react further with species such as carbon monoxide
 4      and with many hydrocarbons (for example, CH4) to produce HO2 radicals.
 5           Measurements of OH radical concentrations in the troposphere (Poppe et al., 1995; Eisele
 6      et al., 1997; Brune et al., 1999; Martinez et al., 2003; Ren et al., 2003) show that, as expected,
 7      the OH radical concentrations are highly variable in space and time, with daytime maximum
 8      concentrations of > 3 x io6 molecules /cm3 in urban areas. A global, mass-weighted mean
 9      tropospheric OH radical concentration also can be derived from the estimated emissions and
10      measured atmospheric concentrations of methyl chloroform (CH3CC13) and the rate constant for
11      the reaction of the OH radical with CH3CC13.  Krol et al. (1998) derived a global average OH
12      concentration  of 1.07 x IO6 molecules /cm3 for 1993 along with an upward trend of about
13      0.5%/yr between 1978 and 1993. Using an integrated data set of observed O3, H2O, NOy, CO,
14      VOCs, temperature and cloud optical depth, Spivakovsky et al. (2000) calculated a global annual
15      mean OH concentration of 1.16 x 106 molecules cm"3, consistent to within 10% of the value
16      obtained by Krol et al. (1998).
17           HO2 radicals do not initiate the oxidation of hydrocarbons, but serve to recycle OH mainly
18      by way of reaction with NO, O3, and itself (the latter produces H2O2, which can photolyze to
19      yield OH).  The HO2 radicals also react with organo-peroxy radicals produced during the
20      oxidation of VOCs to form organo-peroxides  (cf Section AX2.2.5, reaction AX2-20, e.g.).
21      Organo-peroxides undergo wet or dry deposition (Wesely and Hicks, 2000) or degrade further by
22      photolysis and reaction with OH (Jet Propulsion Laboratory, 2003).
23           At night, NO3 assumes the role of primary oxidant (Wayne,  1991).  Although it is generally
24      less reactive than OH, its high abundance in the polluted atmosphere compensates for its lower
25      reactivity.  For several VOCs, however, including dimethylsulfide, isoprene, some terpenes
26      (a-pinene, limonene, linalool) and some phenolic compounds (phenol, o-cresol), oxidation
27      by NO3 at night is competitive with oxidation by OH during the day, making it an important
28      atmospheric removal mechanism for these compounds (Wayne, 1991) (see Table AX2-1).  The
29      role of NO3 radicals in the chemistry of the remote marine boundary layer has been examined
30      recently by Allen et al., (2000) and in the polluted continental boundary layer by Geyer and
31      Platt(2002).

        August 2005                            AX2-9      DRAFT-DO NOT QUOTE OR CITE

-------
 1           Cl atoms, derived from products of multiphase processes can initiate the oxidation of most
 2      of the same VOCs as OH radicals, however, the rate coefficients for the reactions of alkanes with
 3      Cl atoms are usually much higher.  Cl will also oxidize alkenes and aromatic compounds, but
 4      with a significantly lower rate constant than for OH reactions.  Following the initial reaction
 5      with Cl, the degradation of the hydrocarbon proceeds as with OH and NO3, generating an
 6      enhanced supply of odd hydrogen radicals leading to O3 production in the presence of
 7      sufficient NOX.  The corresponding reactions of Br with hydrocarbons proceed in a similar
 8      manner, but with rate coefficients that can be substantially lower or higher.
 9           Chlorine and bromine radicals will also react directly with O3 to form CIO and BrO
10      radicals, providing a sink for odd oxygen if they do not react with NO to form NO2 (e.g.,
11      Pszenny et al., 1993). As  with other oxidants present in the atmosphere, Cl chemistry provides a
12      modest net sink for O3 when NOX is less than 20 pptv, and is a net source at higher NOX. Kasting
13      and Singh (1986) estimated that as much as 25% of the loss of nonmethane hydrocarbons in the
14      nonurban atmosphere can occur by reaction with Cl atoms, based on the production of Cl atoms
15      from gas phase photochemical reactions involving chlorine containing molecules (HC1, CH3C1,
16      CHC13, etc.).  Elevated concentrations of atomic Cl and other halogen radicals can be found in
17      polluted coastal cities where precursors are emitted directly from industrial sources and/or are
18      produced via acid-catalyzed reactions involving sea-salt particles (Tanaka et al., 2000; Spicer
19      etal., 2001).
20           Substantial chlorine-VOC chemistry has been observed in the cities of Houston and
21      Beaumont/Port Arthur, Texas (Tanaka et al., 2000; Chang et al., 2002; Tanaka et al., 2003a).
22      Industrial production activities in those areas frequently result in large releases of chlorine gas
23      (Tanaka et al., 2000). Chloromethylbutenone (CMBO), the product of the oxidation of isoprene
24      by atomic Cl and a unique marker for chlorine radical chemistry in the atmosphere (Nordmeyer
25      et al.,  1997), has been found at significant mixing ratios (up to  145 pptv) in ambient Houston  air
26      (Riemer and Apel, 2001).  However, except for situations in which there are strong local sources
27      such as these, the evidence for the importance of Cl as an oxidizing agent is mixed. Parrish et al.
28      (1992, 1993) argued that ratios of selected hydrocarbons measured at Pt. Arena, CA were
29      consistent with loss by reaction with OH radicals and that any deviations could be attributed to
30      mixing processes.  Finlayson-Pitts (1993), on the other hand had suggested that these deviations
31      could have been the result of Cl reactions. McKeen et al. (1996) suggested that hydrocarbon

        August 2005                             AX2-10      DRAFT-DO NOT QUOTE OR CITE

-------
 1      ratios measured downwind of anthropogenic source regions affecting the western Pacific Basin
 2      are consistent with loss by reaction with OH radicals only. Rudolph et al. (1997), based on data
 3      for several pairs of hydrocarbons collected during a cruise in the western Mediterranean Sea, the
 4      eastern mid- and North Atlantic Ocean and the North Sea during April and May of 1991, also
 5      found that ratios of hydrocarbons to each other are consistent with their loss given mainly by
 6      reaction with OH radicals without substantial contributions from reactions with Cl.  Their best
 7      estimate, for their sampling conditions was a ratio of Cl to OH of about 1CT3, implying a
 8      concentration of Cl of about 103/cm3 using the globally averaged OH concentration of
 9      about lOVcm3 given above. In contrast Wingenter et al. (1996) and Singh et al. (1996a) inferred
10      significantly higher concentrations of atomic Cl (104 to 10s cm"3) based on relative concentration
11      changes in VOCs measured over the eastern North Atlantic and Pacific Oceans, respectively.
12      Similar approaches employed over the high-latitude southern ocean yielded lower estimates of
13      Cl concentrations (103 cm"3; Wingenter et al., 1999). Taken at face value, these observations
14      indicate substantial variability in Cl concentrations and uncertainty in "typical" values.
15
16      AX2.2.4   Chemistry of Nitrogen Oxides in the Troposphere
17          In the troposphere, NO, NO2, and O3 are interrelated by the following reactions:
18
26
                                                       + U2                            (AX2-3)
                                      NO2 + hv -» NO + O(3P)                          (AX2-1)
                                 0(3P) + 02 + M -» 03 + M                              (AX2-2)
19
20      The reaction of NO2 with O3 leads to the formation of the nitrate (NO3) radical,
21
22                                  N02 + 03 -»N03- + 02,                             (AX2-7)
23
24      which in the lower troposphere is nearly in equilibrium with dinitrogen pentoxide (N2O5):
25
                                  NO3'  + NO2 <-^-» N2 O5.                           (AX2-8)
        August 2005                            AX2-11       DRAFT-DO NOT QUOTE OR CITE

-------
 1     However, because the NO3 radical photolyzes rapidly (with a lifetime of «5 s for an overhead
 2     sun [Atkinson et al., 1992a]),
 3
                              NCy + hv -> NO + O2      (10%)                        (Ax2-9a)
                                        -»NO2 + O(3P)  (90%)                       (AX2-9b)
 4
 5     its concentration remains low during daylight hours, but can increase after sunset to nighttime
 6     concentrations of < 5 x 107 to 1 x 1010 molecules cm"3 (< 2 to 430 ppt) over continental areas
 7     influenced by anthropogenic emissions of NOX (Atkinson et al., 1986). This leads to an increase
 8     of N2O5 concentrations during the night by reaction (AX2-8).
 9           The tropospheric chemical removal processes for NOX involve the reaction of NO2 with the
10     OH radical and the hydrolysis of N2O5 in aqueous aerosol solutions to produce HNO3.
11
                                  •OH + N02 -^-^ HN03                           (AX2-10)

12
                                   N2O5   H2°(1)  > HNO3                            (AX2-11)

13
14     The gas-phase reaction of the OH radical with NO2 initiates the major and ultimate removal
15     process for NOX in the troposphere. This reaction removes radicals (OH and NO2) and competes
16     with hydrocarbons for OH radicals in areas characterized by high NOX concentrations, such as
17     urban centers (see Section AX2.4). In addition to gas-phase nitric acid, Golden and Smith
18     (2000) have concluded that, pernitrous acid (HOONO) is also produced by the reaction of NO2
19     and OH radicals on the basis of theoretical studies.  However, a recent assessment (Jet
20     Propulsion Laboratory, 2003) has concluded that this channel represents a minor yield
21     (approximately 15% at the surface). HOONO will thermally decompose or photolyze.
22     Gas-phase HNO3 formed from reaction AX2-10 undergoes wet and dry deposition to the surface
23     and uptake by ambient aerosol particles.  The tropospheric lifetime of NOX due to reaction
24     AX2-10 ranges from a few hours to a few days. Geyer and Platt (2002) concluded that reaction
25     AX2-11 constituted about 10% of the removal of NOX at a site near Berlin, Germany during

       August 2005                            AX2-12      DRAFT-DO NOT QUOTE OR CITE

-------
 1      spring and summer. However, during winter the relative importance of reaction AX2-11 could
 2      be much higher because of the much lower concentration of OH radicals and the enhanced
 3      stability of N2O5 due to lower temperatures and intensity of sunlight. Note that reaction AX2-11
 4      surely proceeds as a heterogeneous reaction.
 5           OH radicals also can react with NO to produce nitrous acid (HNO2):
 6
                                   •OH + NO  M   >HN02.                           (AX2-12)

 7
 8      In the daytime, HNO2 is rapidly photolyzed back to the original reactants:
 9
10                                HNO2 + hv ->• 'OH + NO.                           (AX2-13)
11
12      At night, HNO2 can be formed by heterogeneous reactions of NO2 in aerosols or at the earth's
13      surface (Lammel and Cape, 1996; Jacob, 2000; Sakamaki et al., 1983; Pitts et al.,  1984a;
14      Svensson et al., 1987; Jenkin et al., 1988; Lammel and Perner, 1988; Notholt et al., 1992a,b).
15      This results  in accumulation of HNO2 during nighttime.  Modeling studies suggest that
16      photolysis of this HNO2 following sunrise, could provide an important early-morning source of
17      OH radicals to drive O3 formation (Harris et al., 1982).
18           Another important process controlling NOX concentrations is the formation of organic
19      nitrates. Oxidation of VOCs produces organic peroxy radicals (RO2), as discussed in the
20      hydrocarbon chemistry subsections to follow. Reaction of these RO2 radicals with NO and NO2
21      produces organic nitrates (RONO2) and peroxynitrates (RO2NO2):
22
23                               RO2' + NO  M  > RONO2                          (AX2-14)
24
25                               RO2' + NO2  M  > RO2NO2                         (AX2-15)
26
27           Reaction (AX2-14) is  a minor branch for the reaction of RO2 with NO (the major branch
28      produces RO and NO2,  as discussed in the next section). The organic nitrate yield increases with
29      carbon number (Atkinson, 2000).

        August 2005                            AX2-13      DRAFT-DO NOT QUOTE OR CITE

-------
 1           The organic nitrates may react further, depending on the functionality of the R group, but
 2     they will typically not return NOX and can therefore be viewed as a permanent sink for NOX.
 3     This sink is usually small compared to HNO3 formation, but the formation of isoprene nitrates
 4     may be a significant sink for NOX in the United States in summer (Liang et al., 1998).
 5           The peroxynitrates produced by (AX2-15) are thermally unstable and most have very short
 6     lifetimes (less than a few minutes) against thermal decomposition to the original reactants.  They
 7     are thus not effective sinks of NOX.  Important exceptions are the peroxyacylnitrates (PANs)
 8     arising from the peroxyacyl radicals RC(O)OO produced by oxidation and photolysis of
 9     carbonyl compounds.  PANs have lifetimes ranging from ~1 hour at room temperature to several
10     weeks at 250K. They can thus provide an effective sink of NOX at cold temperatures, but also a
11     reservoir allowing eventual release of NOX as air masses warm, in particular by subsidence. By
12     far the most important of these PANs compounds is peroxyacetylnitrate (PAN), with formula
13     CH3C(O)OONO2. PAN is a significant product in the oxidation of most VOCs.  It is now well
14     established that PAN decomposition provides a major source of NOx in the remote troposphere
15     (Staudt et al., 2003).  PAN decomposition in subsiding Asian air masses over the eastern Pacific
16     could make an important contribution to O3 enhancement in the U.S. from Asian pollution
17     (Hudman et al., 2004).
18
19     AX2.2.5  The Methane Oxidation Cycle
20           The photochemical cycles leading to O3 production are best understood by considering the
21     oxidation of methane, structurally the simplest VOC.  The CH4 oxidation cycle serves as a model
22     which describes the chemistry of the relatively  clean or unpolluted troposphere (although this is
23     a simplification because vegetation releases large quantities of complex VOCs into the
24     atmosphere). Although the chemistry of the VOCs emitted  from anthropogenic and biogenic
25     sources in polluted urban and rural areas is more complex, a knowledge of the CH4 oxidation
26     reactions aids in understanding the chemical processes occurring in the polluted atmosphere
27     because the underlying chemical principles are the same.
28           Methane is emitted into the atmosphere as the result of anaerobic microbial activity in
29     wetlands, rice paddies, the guts of ruminants, landfills, and from mining and combustion of
30     fossil fuels (Intergovernmental Panel on Climate Change, 2001).  The major tropospheric
31     removal process for CH4 is by reaction with the OH radical,

       August 2005                             AX2-14      DRAFT-DO NOT QUOTE OR CITE

-------
 1                               -OH + CH4 -» H2O + CH3.                           (AX2-16)
 2
 3     In the troposphere, the methyl radical reacts solely with O2 to yield the methyl peroxy (CH3O2»
 4     radical (Atkinson et al., 1992a):

                                CH3 + O2   M  > CH3O2-.                          (AX2-17)
 5
 6          In the troposphere, the methyl peroxy radical can react with NO, NO2, HO2 radicals, and
 7     other organic peroxy (RO^ radicals, with the reactions with NO and HO2 radicals being the most
 8     important (see, for example, World Meteorological Organization, 1990). The reaction with NO
 9     leads to the formation of the methoxy (CH36) radical,
10

                               CH3O2- + NO -» CH36 + NO2.                         (AX2-18)

11          The reaction with the HO2 radical leads to the formation of methyl hydroperoxide
12     (CH3OOH),

                             CH302- + H02- -» CH3OOH + O2,                       (AX2-19)

13     which can photolyze or react with the OH radical (Atkinson  et al., 1992a):

                               CH3OOH + hv -» CH36 + -OH.                        (AX2-20)
                             •OH + CH3OOH -> H2O + CH3O2'                      (AX2-21 a)
14     or
15
                                -» H2O + CH2OOH   fast decomposition
                                                                                 (AX2-21b)
                   CH2OOH + M -> H2CO + -OH
       August 2005                            AX2-15      DRAFT-DO NOT QUOTE OR CITE

-------
 1     Methyl hydroperoxide is much less soluble than hydrogen peroxide (H2O2), and so wet
 2     deposition after incorporation into cloud droplets is much less important as a removal process
 3     than it is for H2O2.  CH3OOH can also be removed by dry deposition to the surface or transported
 4     by convection to the upper troposphere. The lifetime of CH3OOH in the troposphere due to
 5     photolysis and reaction with the OH radical is estimated to be «2 days. Methyl hydroperoxide is
 6     then a temporary sink of radicals, with its wet or dry deposition representing a loss process for
 7     tropospheric radicals.
 8           The only important reaction for the methoxy radical (CH36) is
 9
                                CH3O + O2 -» H2CO + HCy.                        (AX2-22)
10
1 1                                HCy + NO -> -OH + N02                          (AX2-23)

12     The HO2 radicals produced in (AX2-22) can react with NO, O3, or other HO2 radicals according
13     to,
14
1 5                                H02-  + 03 -> -OH + 202                           (AX2-24)
16
17                               HO2- + HO2- -» H2O2 + O2.                         (AX2-25)

18     Formaldehyde (H2CO) produced in reaction AX2-22 can be photolyzed:
19
20                           H2CO+hv^H2 + CO  (55%)                       (AX2-26a)
21
                                                      (45%)                       (AX2-26b)
23     Formaldehyde also reacts with the OH radical,
24
25                              -OH + H2CO -> H2O + HCO.                         (AX2-27)
       August 2005                            AX2-16      DRAFT-DO NOT QUOTE OR CITE

-------
 1      The H atom and HCO (formyl) radical produced in these reactions react solely with O2 to form
 2      the HO2 radical:
 3
                                 •H + O2 + M -» HO2' + M                           (AX2-28)
                                   HCO + O2 ->• HO2- + CO.                         (AX2-29)
 4
 5      The lifetimes of H2CO due to photolysis and reaction with OH radicals are «4 h and 1.5 days,
 6      respectively, leading to an overall lifetime of slightly less than 4 hours for H2CO for overhead
 7      sun conditions (Rogers, 1990).
 8           The final step in the oxidation of CH4 involves the oxidation of CO by reaction with the
 9      OH radical to form CO2:
10
11                                CO +-OH-> C02 + «H                             (AX2-30)
12
1                                 HCO + 02 ->> H02' + CO.                           (AX2-29)
14
15      The lifetime of CO in  the lower troposphere is «2 months at midlatitudes.
16           NO and HO2 radicals compete for reaction with CH3O2 and HO2 radicals, and the reaction
17      route depends on the rate constants for these two reactions and the tropospheric concentrations
18      ofHO2andNO. The rate constants for the reaction of the CH3O2  radicals with NO (reaction
19      AX2-18) and HO2 radicals (reaction AX2-19) are of comparable magnitude (e.g., Jet Propulsion
20      Laboratory, 2003). Based on expected HO2 radical concentrations in the troposphere, Logan
21      et al.  (1981) calculated that the reaction of the CH3O2 radical with NO dominates for NO mixing
22      ratios of >30 ppt.  For NO mixing ratios <30 ppt, the reaction of the CH3O2 radical with HO2
23      dominates. The overall effects of methane oxidation on O3 formation for the case when
24      NO >30 ppt can be written as:
25
        August 2005                            AX2-17       DRAFT-DO NOT QUOTE OR CITE

-------
                         CH4 + -OH + 02 -> CH302- + H20               (AX2-16, AX2-17)
                         CH302- +NO -> CH30- +N02                          (AX2-18)
                         CH3O + O2 -> H2CO + HO2-                           (AX2-22)
                         HO2« + NO -» -OH + NO2                              (AX2-23)
                         2(NO2 + hv -» NO + O(3P))                              (AX2-1)
                         2(O(3P) + O2 + M -» O3 + M)                            (AX2-2)
                 Net:  CH4 + 4O2 + 2hv -> H2CO + 2O3 + H2O                    (AX2-31)
1
2     Further O3 formation occurs, based on the subsequent reactions of H2CO, e.g.,
3

                    H2CO + hv + 202 -» 2H02- + CO              (AX2-26b; AX2-28; AX2-29)
                    2(H02- + NO -> -OH + N02)                                  (AX2-23)
                                     0(3P))                                     (AX2-1)
                                     03+M)                                   (AX2-2)
               Net:  H2CO + 4O2 + hv -+ CO + 2O3 + 2-OH.                          (AX2-32)

4
5     Reactions in the above sequence lead to the production of two OH radicals which can further
6     react with atmospheric constituents (e.g.,  Crutzen, 1973). There is also a less important
7     pathway:
      August 2005                           AX2-18      DRAFT-DO NOT QUOTE OR CITE

-------
 2                               H2CO + hv->H2 + CO                              (AX2.26a)
 3                               CO + -OH -> C02 + -H                              (AX2.30)
 4
                                 •H + O2 + M -» HO2* + M                            (AX2-28)
 6                               H02- + NO^-OH + N02                            (AX2-23)
 7                               NO2 + hv -> NO + O(3P)                             (AX2-1)
 8
                                 O(3P) + O2 + M -» O3 + M                           (AX2-2)
10                        Net:   H2CO + 2O2 + hv-»CO2 + O3+H2                   (AX2-33)
11
12
13      These reaction sequences are important for tropospheric chemistry because formaldehyde is an
14      intermediate product of the oxidation of most VOCs. The reaction of O3 and HO2 radicals leads
15      to the net destruction of tropospheric O3:
16
                                  H02- + 03 -> -OH + 202                           (AX2-24)

17
18      Using the rate constants reported for reactions AX2-23 and AX2-24 (Atkinson et al., 1992a) and
19      the background tropospheric O3 mixing ratios given above, the reaction of HO2 radicals with NO
20      dominates over reaction with O3 for NO mixing ratios >10 ppt. The rate constant for
21      reaction AX2-25 is such that an NO mixing ratio of this magnitude also means that the HO2
22      radical reaction with NO will be favored over the self-reaction of HO2 radicals.
23           Consequently, there are two regimes in the "relatively clean" troposphere, depending on
24      the local NO concentration: (1) a "very low-NOx" regime in which HO2 and CH3O2 radicals
25      combine (reaction AX2-19), and HO2 radicals undergo self-reaction (to form H2O2) and react
26      with O3 (reactions AX2-25 and AX2-24), leading to net destruction of O3 and inefficient OH
27      radical regeneration (see also Ehhalt et al., 1991; Ayers et al.,  1992); and (2) a "low-NOx"
28      regime (by comparison with much higher NOX concentrations  found in polluted areas) in
29      which HO2 and CH3O2 radicals react with NO to convert NO to NO2, regenerate the OH radical,
       August 2005                            AX2-19      DRAFT-DO NOT QUOTE OR CITE

-------
 1      and, through the photolysis of NO2, produce O3.  In the "low NOX" regime there still may be
 2      significant competition from peroxy-peroxy reactions, depending on the local NO concentration.
 3           Nitric oxide mixing ratios are sufficiently low in the remote marine boundary layer
 4      relatively unaffected by transport of NOX from polluted continental areas (< 15 ppt) that
 5      oxidation of CH4 will lead to net destruction of O3, as discussed by Carroll et al. (1990) and
 6      Ayers et al. (1992). In continental and marine areas affected by transport of NOX from
 7      combustion sources, NO mixing ratios are high enough (of the order of-one to a few hundred
 8      ppt) for the oxidation of CH4, nonmethane hydrocarbons (NMHCs) and CO to lead to net O3
 9      formation (e.g., Carroll et al., 1990; Dickerson et al., 1995). Generally, NO mixing ratios
10      increase with altitude and can be of the order of fifty to a few hundred ppt in the upper
11      troposphere depending on location. The oxidation of peroxides, carbon monoxide and acetone
12      transported upward by convection, in the presence of this NO, can lead to local O3 formation
13      (e.g., Singh et al., 1995; McKeen et al., 1997; Wennberg et al.,  1998; Bruhl et al., 2000).
14
15      AX2.2.6  The Atmospheric Chemistry of Alkanes
16           The same basic processes by which CH4 is oxidized occur in the oxidation of other, even
17      more reactive and more complex VOCs. As in the CH4 oxidation cycle, the conversion of NO
18      to NO2 during the oxidation of VOCs results in the production of O3 and the efficient
19      regeneration of the OH radical, which in turn can react with other VOCs (Figure AX2-2).  The
20      chemistry of the major classes of VOCs important for O3 formation such  as alkanes, alkenes
21      (including alkenes from biogenic sources),  and aromatic hydrocarbons will be summarized in
22      turn.
23           Reaction with OH radicals represents the main loss process for alkanes and as also
24      mentioned earlier, reaction with nitrate and chlorine radicals are additional sinks for alkanes.
25      For alkanes having carbon-chain lengths of four or less, the chemistry is well understood and the
26      reaction rates are slow in comparison to alkenes  and other VOCs of similar structure and
27      molecular weight.  See Table AX2-1 for a comparison of reaction rate constants for several
28      small alkanes and their alkene and diene homologues.  For alkanes larger than C5, the situation
29      is more complex because the products generated during the degradation of these compounds are
30      usually not well characterized. Branched alkanes have rates of reaction that are highly
        August 2005                             AX2-20      DRAFT-DO NOT QUOTE OR CITE

-------
Anthropogenic
BiogenicVOC + OH

       ROOM  -*-

  Carbonyl    R°2
  Alcohol
                                                    R-
                                                    |02
                                                         NO2
                                                   R02 9    *  ROON02
                                                                  RON02
                                   Decomposition
                        02 ) Isomerization
                                                 Products
       Figure AX2-2. General chemical mechanism for the oxidative degradation of VOCs.
       Source: Atkinson (2000).
 1     dependent on carbon backbone structure. Stable products of alkane photooxidation are known to
 2     include carbonyl compounds, alkyl nitrates, and hydroxycarbonyls.
 3          Alkyl nitrates form primarily as an alternate product of reaction AX2-34 (below).  Several
 4     modeling studies have predicted that large fractions of NOy exist as alkyl and hydroxy alkyl
 5     nitrates (Calvert and Madronich, 1987; Atherton and Penner, 1988; Trainer et al., 1991).
 6     In NOX- and VOC-rich urban atmospheres, 100 different alkyl and 74 different hydroxy alkyl
 7     nitrate compounds have been predicted and identified (Calvert and Madronich, 1987; Schneider
 8     and Ballschmiter, 1999; Schneider et al., 1998).  Uncertainties in the atmospheric chemistry of
 9     the alkanes include the branching ratio of reaction AX2-34, i.e., the extent to which alkyl nitrates
10     form versus RO and NO2.  These uncertainties affect modeling predictions of NOX
11     concentrations, NO-to-NO2 conversion and O3 formation during photochemical degradation of
12     the VOCs. Discrepancies between observations and theory have been found in aircraft
13     measurements of NOy (Singh et al., 1996b).  Recent field studies conducted by Day et al. (2003)
14     have shown that large fractions of organic nitrates, which may be associated with isoprene
       August 2005
                  AX2-21
DRAFT-DO NOT QUOTE OR CITE

-------
 1     oxidation products, are present in urban and rural atmosphere that have not been previously
 2     measured and considered in NOy calculations to date.
 3          Alcohols and ethers in ambient air react almost exclusively with the OH radical, with the
 4     reaction proceeding primarily via H-atom abstraction from the C - H bonds adjacent to the
 5     oxygen-containing function group in these compounds (Atkinson and Arey, 2003).
 6          The following list of general reactions, analogous to those described for methane,
 7     summarizes the role of alkane oxidation in tropospheric O3 formation.
 8
10                               .OH + RH^H20 + .R                              (AX2_34)
12                               .R + 02 + M^R02- + M                            (AX2-35)
13
14                               R02' + NO -> RO + N02                            (AX2-3 6)
15
16                               H02- + NO-»«OH + N02                            (AX2-23)
17
18                               RO- + O2 -> R'CHO + HO2'                          (AX2-3 7)
19
20                               2(NO2 + hv-» NO + O)                              (AX2-1)
21
22                               2(O + O2 + M -^ O3  + M)                            (AX2-2)
23                      	:	
24                      Net:  RH + 4O2 + 2hv -> R'CHO  + 2O3 + H2O                 (AX2-38)
25
26
27     The oxidation of alkanes can also be initiated by other  oxidizing agents such as NO3 and Cl
28     radicals.  In this case, there is net production of an OH radical which can re-initiate the oxidation
29     sequence. The reaction of OH radicals with aldehydes forms acyl (R'CO) radicals, and acyl
30     peroxy radicals (R'C(O)O2) are formed  by the addition of O2. As an example, the oxidation of
31     ethane (C2H5-H) yields acetaldehyde (CH3-CHO). Acetyl (CH3-CO) and acetylperoxy
32     (CH3-C(O)O2) radicals can then be formed. Acetylperoxy radicals can combine with NO2 to
33     form peroxyacetyl nitrate (PAN) via:
34
                       CH3C(O)O2' + NO2 + M o CH3C(O)O2NO2 + M                (AX2-39)

35

       August 2005                            AX2-22     DRAFT-DO NOT QUOTE OR CITE

-------
 1     PAN can act as a temporary reservoir for NO2. Upon the decomposition of PAN, either locally
 2     or elsewhere, NO2 is released to participate in the O3 formation process again.  During the
 3     oxidation of propane, the relatively long-lived intermediate acetone (CH3 - C(O) CH3) is formed,
 4     as shown in Figure AX2-3.  The photolysis of acetone can be an important source of OH
 5     radicals, especially in the upper troposphere (e.g., Singh et al., 1995). Examples of oxidation
 6     mechanisms of more complex alkanes and other classes of hydrocarbons can be found in
 7     comprehensive texts such as Seinfeld and Pandis (1998).
 8
 9     AX2.2.7   The Atmospheric Chemistry of Alkenes
10           As shown in Figure AX2-3, the presence of a double carbon-carbon bond, i.e., > C = C <,
11     in a VOC can greatly increase the range of potential reaction intermediates and products,
12     complicating the prediction of O3 production. The alkenes emitted from anthropogenic  sources
13     are mainly  ethene, propene, and the butenes,  with lesser amounts of the > C5 alkenes. The major
14     biogenic alkenes emitted from vegetation are isoprene (2-methyl-1,3-butadiene) and C10H16
15     monoterpenes (Atkinson and Arey, 2003), and their tropospheric chemistry is currently the focus
16     of much attention (Zhang et al., 2002; Sauer  et al., 1999; Geiger et al., 2003; Sprengnether et al.,
17     2002; Witter et al., 2002; Bonn and Moortgat, 2003; Berndt et al., 2003; Fick et al., 2003;
18     Kavouras et al.,  1999; Atkinson and Arey, 2003).
19           Alkenes react in ambient air with OH and NO3 radicals and with O3. The mechanisms
20     involved in their oxidation have been discussed in detail by Calvert et al. (2000). All three
21     processes are important atmospheric  transformation processes, and all proceed by initial addition
22     to the > C = C < bonds or, to a much lesser extent, by H atom extraction. Products of alkene
23     photooxidation include carbonyl compounds, hydroxy alkyl nitrates and nitratocarbonyls, and
24     decomposition products  from the high energy biradicals formed in alkene-O3 reactions.
25     Table AX2-2 provides estimated atmospheric lifetimes for biogenic alkenes with respect to
26     oxidation by OH, NO3 and O3. The structures of most of the compounds given in Table AX2-2
27     are shown in Figure AX2-4.
28           Uncertainties in the atmospheric chemistry of the alkenes concern the products and
29     mechanisms of their reactions with O3, especially the yields of OH radicals, H2O2, and secondary
30     organic aerosol in both outdoor and indoor environments.  However, many product analyses of
31     important biogenic and anthropogenic alkenes in recent years have aided in the narrowing of

       August 2005                            AX2-23       DRAFT-DO NOT QUOTE OR CITE

-------
      a. Propane
          •OH +
                  HO
                              HO2« + CH2=O
             CH3O« + NO2
      b. Propene
             H3C —C
                                                               (02)
           H3C	CH  +HO2*
               CH2=O + HO2*
Figure AX2-3.  Hydro\y\ radical initiated oxidation of a) propane and b) propene.




Source:  Calvert et al. (2000).
August 2005
AX2-24      DRAFT-DO NOT QUOTE OR CITE

-------
      Table AX2-2. Calculated Atmospheric
                Compounds (adapted from
    Lifetimes of Biogenic Volatile Organic
    Atkinson and Arey, 2003)a
Lifetime for Reaction with
Biogenic VOC
Isoprene
Monoterpenes
Camphene
2-Carene
3-Carene
Limonene
Myrcene
cis-/trans-Qcimene
a-Phellandrene
p-Phellandrene
a-Pinene
p-Pinene
Sabinene
a-Terpinene
y-Terpinene
Terpinolene
Sesquiterpenes
p-Caryophyllene
a-Cedrene
a-Copaene
a-Humulene
Longifolene
Oxygenates
Acetone6
Camphor
1,8-Cineole
c/s-3-Hexen-l-ol
c/s-3-Hexenyl acetate
Linalool
OHb
1.4 h

2.6 h
1.7 h
16h
49min
39 min
33 min
27 min
50 min
2.6 h
1.8 h
1.2 h
23 min
47 min
37 min

42 min
2.1 h
1.5 h
28 min
2.9 h

61 df
2.5 dh
l.Od1
1.3 hk
18 hk
52 mink
03C
0.92 d

13d
1.2 h
8.0 h
1.4 h
35 min
31 min
5.6 min
5.9 h
3.2 h
0.77 d
3.4 h
0.7 min
2.0 h
9.1 min

1 .4 min
9.8 h
1.8 h
1 .4 min
>23d

>3.2y8
> 165 dh
>77dJ
4.3 hk
5.1hk
39 min k
NO3d
1.6 h

1.7 h
4 min
7 min
5 min
6 min
3 min
0.9 min
8 min
11 min
27 min
7 min
0.5 min
2 min
0.7 min

3 min
8 min
4 min
2 min
1.6h

>8yf
>300dh
1.5y'
4.1 hk
4.Shk
6 mink
August 2005
AX2-25
DRAFT-DO NOT QUOTE OR CITE

-------
           Table AX-2 (cont'd).  Calculated Atmospheric Lifetimes of Biogenic Volatile Organic
                          Compounds (adapted from Atkinson and Arey, 2003)a
Lifetime for Reaction with
Biogenic VOC
Oxygenates (cont'd)
Methanol
2-Methyl-3 -buten-2-ol
6-Methyl-5 -hepten-2-one
OHb

12 df
2.4 h1
53 min°
03C

>3.2yg
1.2dm
0.7 h°
NO3d

2.0yf
7.7 dn
9min°
         aRate coefficients rom Calvert et al. (2000) unless noted otherwise.
         b Assumed OH radical concentration: 1.0 x 106 molecule cm"3.
         0 Assumed O3 concentration: 1 * 1012 molecule cm"3, 24-h average.
         d Assumed NO3 radical concentration: 2.5 * 108 molecule cm"3, 12-h nighttime average.
         e Photolysis will also occur with a calculated photolysis lifetime of -60 day for the lower troposphere, July, 40° N
         (Meyrahnetal., 1986).
         f Atkinson etal. (1999).
         g Estimated.
         hReisselletal. (2001).
         1 Corchnoy and Atkinson (1990).
         J Atkinson etal. (1990).
         k Atkinson etal. (1995).
         'Papagnietal. (2001).
         mGrosjean and Grosjean (1994).
         "Rudichetal. (1996).
         "Smithetal. (1996).
 1      these uncertainties. The reader is referred to extensive reviews by Calvert et al. (2000) and

 2      Atkinson and Arey (2003) for detailed discussions of these products and mechanisms.

 3

 4      Oxidation by OH

 5           As noted above, the OH radical reactions with the alkenes proceed mainly by OH radical

 6      addition to the > C = C < bonds. As shown in Figure AX2-3, for example, the OH radical

 7      reaction with propene leads to the formation of two OH-containing radicals. The subsequent

 8      reactions of these radicals are similar to those of the alkyl radicals formed by H-atom abstraction

 9      from the alkanes. Under high NO conditions, CH3CHCH2OH continues to react — producing

10      several smaller, "second generation," reactive VOCs.

11



        August 2005                               AX2-26       DRAFT-DO NOT QUOTE OR CITE

-------
                     Camphene     2-Carene      3-Carene      Limonene      Myrcene
                     c;'s-Ocimene  a-Phellandrene  p-Phellandrene    a-Pinene      p-Pinene
                        Sabinene         a-Terpinene       y-Terpinene       Terpinolene
                                                       \
                         p-Caryophyllene
                                                u-Cedrene
                       u-Copaene
                                   a-Humulene
                                                            Longifolene
                                                                         OH
                          Camphor
  1,8-Cineole
                                                                      Linalool
Figure AX2-4.  Structures of a selected number of terpene and sesquiterpene compounds.




Source:  Atkinson and Arey (2003).
August 2005
AX2-27       DRAFT-DO NOT QUOTE OR CITE

-------
 1           For the simple  1 ppb.  The
25      values are much lower for lower NOX concentrations. The situation is much better for
26      methacrolein. Observed products can account for more than 90% of the reacted carbon.
27           The rates of formation of condensible, oxidation products of biogenic compounds that may
28      contribute to secondary organic aerosol formation is an important matter for the prediction of
29      ambient aerosol concentrations. Claeys et al. (2004) found that 2-methyltetrols are formed from
30      the oxidation of isoprene in yields of about 0.2%  on a molar basis, or 0.4%  on a mass basis.
        August 2005                            AX2-28      DRAFT-DO NOT QUOTE OR CITE

-------
 1      These are semivolatile compounds that can condense on existing particles. On the other hand,
 2      pinene oxidation leads to substantial organic aerosol formation.
 3
 4      Oxidation by Nitrate Radical
 5           NO3 radical reacts with alkenes mainly by addition to the double bond to form a
 6      b-nitrooxyalkyl radical (Atkinson 1991, 1994, 1997). The abstraction pathway may account for
 7      up to 20% of the reaction. For propene, the initial reaction is followed by a series of reactions
 8
 9
10                        NO3- + CH3CH=CH2 -» CH3CHCH2ONO2
                                                                                      (AX2-40)
11                                             -»CH3CH(ONO2)CH2
12
13
14      that (Atkinson, 1991)  to lead to the formation of, among others,  carbonyls and nitrato-carbonyls
15      including formaldehyde (HCHO), acetaldehyde (CH3CHO), 2-nitratopropanal
16      (CH3CH(ONO2)CHO), and 1-nitratopropanone (CH3C(O)CH2ONO2). By analogy to OH,
17      conjugated dienes like butadiene and isoprene will react with NO3 to form d-nitrooxyalkyl
18      radicals. (Atkinson, 2000).  If NO3 is available for reaction in the atmosphere, then NO
19      concentrations will be low, owing to the rapid reaction between NO3 and NO. Consequently,
20      nitrooxyalkyl peroxy radicals are expected to react primarily with NO2, yielding thermally
21      unstable peroxy nitrates, NO3, HO2, and organoperoxy radicals (Atkinson, 2000).
22           Several studies have undertaken the quantification of the products of NO3 initiated
23      degradation of several of the important biogenic alkenes in O3 and secondary  organic aerosol
24      formation, including isoprene, a- and b-pinene, 3-carene, limonene, linalool, and 2-methyl-3-
25      buten-2-ol.  See Figure AX2-4 for the chemical structures of these and other biogenic
26      compounds. The results of these studies have been tabulated by Atkinson and Arey (2003).
27
28      Oxidation by Ozone
29           Unlike other organic compounds in the  atmosphere, alkenes react at significant rates
30      with O3. Ozone initiates the oxidation of alkenes by addition across carbon-carbon double
31      bonds, at rates that are competitive with reaction with OH (see Table AX2-1). The addition

        August 2005                            AX2-29      DRAFT-DO NOT QUOTE OR CITE

-------
 1     of O3 across the double bond yields an unstable ozonide, a 5-member ring including a single
 2     carbon-carbon bond linked to the three oxygen atoms, each singly bound. The ozonide
 3     rearranges spontaneously and then fragments to form an aldehyde or ketone, depending on the
 4     original position of the double bond, and a high energy Criegee biradical. Collisional energy
 5     transfer may stabilize the radical, preventing it from decomposing. Low pressure studies of the
 6     decomposition  of the Criegee biradical have shown high yields of the OH radical.
 7     At atmospheric pressures, the rates of OH production have not been reliably established, due to
 8     complications arising from subsequent reactions of the OH produced with the ozonide fragments
 9     (Calvert et al., 2000).
10           The ozonolysis of larger biogenic alkenes yields high molecular weight oxidation products
11     with sufficiently low vapor pressures to allow condensation into the particle phase. Many
12     oxidation products of larger biogenic alkenes have been identified in ambient aerosol,
13     eliminating their further participation in O3 production. Figure AX2-5 shows the chemical
14     structures of the oxidation products of a-pinene and illustrates the complexity of the products.
15     Carbonyl  containing compounds are especially prevalent.  A summary of the results of product
16     yield studies for several biogenic alkenes can be found in Atkinson and Arey (2003).
                  CHO
                 CHO
                                                        CHO
                                                                            O-OH
                                      -CHO
                        CO2H
                                                                C02H
                                                    -CO2H
                               CO2H
                                                                                   CO9H
                                                                    CO2H
       Figure AX2-5. Products from the reaction of terpenes with O3.
       Source: Atkinson and Arey (2003).
       August 2005
AX2-30       DRAFT-DO NOT QUOTE OR CITE

-------
 1          NO2 also participates to a very small degree in the oxidation of alkenes by addition to
 2     double bonds in a manner similar to O3. Rate constants for reactions of this type range
 3     from 1CT18 to 1CT24 for dienes and monoalkenes (King et al., 2002). It should also be noted
 4     that O3 reacts with terpenoid compounds released from household products such as air fresheners
 5     and cleaning agents in indoor air to produce ultrafme particles (Wainman et al., 2000; Sarwar
 6     et al., 2002)
 7
 8     AX2.2.8  The Atmospheric Chemistry of Aromatic Hydrocarbons
 9          Aromatic hydrocarbons represent a major class of compounds found in gasoline and other
10     liquid fuels. Upon vaporization, most of these compounds react rapidly in the atmosphere
11     (Davis et al., 1975) and following a series of complex processes, involving molecular oxygen
12     and oxides of nitrogen, produce O3.  Reaction with OH radicals serves as the major atmospheric
13     loss process of aromatic hydrocarbons.  Atmospheric losses of alkyl aromatic compounds by O3
14     and nitrate radicals have been found to be minor processes for most monocyclic aromatic
15     hydrocarbons. (However, the reaction with of the nitrate radical with substituted
16     hydroxybenzenes, such as phenol or o-,m-,p-cresol, can be an important atmospheric loss
17     process for these compounds.) Much of the early work in this field focused on the temperature
18     dependence of the OH reactions (Perry et al., 1977; Tully et al., 1981) using absolute rate
19     techniques.  Typically two temperature regions were observed for a large number of aromatic
20     compounds and the complex temperature profile suggested that two mechanisms were operative.
21     In the high temperature region, hydrogen (H)-atom abstraction from the aromatic ring
22     dominates, and in the temperature regime less than 320K, OH addition to the aromatic ring is the
23     dominant process. Thus, at normal temperatures and pressures in the lower troposphere, ring
24     addition is the most important reactive process followed by H-atom abstraction from any alkyl
25     substituents.  The kinetics of monocyclic aromatic compounds are generally well understood and
26     there is generally broad consensus regarding the atmospheric lifetimes for these compounds. By
27     contrast, there is generally a wide range of experimental results from product studies of these
28     reactions.  This leads to a major problem in model development due to a general lack of
29     understanding of the product identities and yields for even the simplest aromatic compounds,
30     which is due to the complex reaction paths following initial reaction with OH, primarily by the
31     addition pathway.

       August 2005                            AX2-31       DRAFT-DO NOT QUOTE OR CITE

-------
 1           Two comprehensive reviews, which provide a detailed understanding of the current state-
 2      of-science of aromatic hydrocarbons have been written in the past five years.  Atkinson (2000)
 3      reviewed the atmospheric chemistry of volatile organic compounds, of which aromatic
 4      hydrocarbons are included in one section of the review.  More recently Calvert et al. (2002)
 5      conducted a highly comprehensive examination of the reaction rates, chemical mechanisms,
 6      aerosol formation, and contributions to O3 formation for monocyclic and polycyclic aromatic
 7      hydrocarbons.
 8
 9      AX2.2.8.1  Chemical Kinetics and Atmospheric Lifetimes of Aromatic Hydrocarbons
10           Rate constants for the reaction of species in the atmosphere with aromatic hydrocarbons
11      vary widely depending on the number of aromatic rings and substituent groups.  Reactions of O3
12      with aromatic hydrocarbons (AHCs) are generally slow except for monocyclic aromatic
13      hydrocarbons having unsaturated substituent groups. For example, indene and styrene have
14      atmospheric lifetimes of 3.3 h and 23 h with respect to reaction with O3, which are much longer
15      than that due to reactive loss with either  OH or NO3. Thus, the atmospheric lifetimes and
16      reaction products of O3 and aromatic hydrocarbons will be ignored in this discussion. In
17      addition to chemical reaction, some organic compounds photolyze in the lower atmosphere.
18      Virtually all aromatic precursors are not  subject to photolysis, although many of the ring
19      fragmentation products having multiple carbonyl  groups can photolyze in the troposphere.
20           The reaction rates and atmospheric lifetimes of monocyclic aromatic compounds due to
21      reaction with OH radicals are generally dependent on the number and types of substituent  groups
22      associated with the ring.  These reaction rates have been found to be highly temperature and
23      pressure dependent. The temperature regimes are governed by the processes involved and show
24      a quite complex appearance.  At room temperature (-300 K), both addition to the aromatic ring
25      and H-atom abstraction occur with the addition reaction being dominant. For the two smallest
26      monocyclic aromatic hydrocarbons, the initial addition adduct is not completely stabilized at
27      total pressures below 100 torr.
28           Numerous studies have been conducted to measure the OH + benzene rate constant over
29      a wide range of temperatures and pressures.  An analysis of absolute rate data taken at
30      approximately 100 torr argon and not at the high pressure limit yielded a value of 1.2 x 10~12 cm3
        August 2005                            AX2-32      DRAFT-DO NOT QUOTE OR CITE

-------
 1      molec"1 s"1. Atkinson (1989) recommended a value of 1.4 x 10"12 cm3molecfJ s"1 at room
 2      temperature and atmospheric pressure. This recommendation has been refined only slightly and
 3      is reflected in the recent value recommended by Calvert et al. (2002) which is given as
 4      1.39 x 10"12 cm3 molec'V1.  This recommended value for the reaction of OH + benzene together
 5      with values for other monocyclic aromatic hydrocarbons is given in Table AX2-3.
 6           In general, it is observed that the OH rate constants with monocyclic alkyl aromatic
 7      hydrocarbons are strongly influenced by the number of substituent groups found on the aromatic
 8      ring.  (That is, the identity of the alkyl substituent groups  has little influence on the overall
 9      reactions rate constant.) Single substituent single-ring aromatic compounds which include
10      toluene, ethyl benzene, n-propylbenzene, isopropylbenzene, and t-butylbenzene have average
11      OH reaction rate constants ranging from 4.5 to 7.0 x 10"12 cm3 molec"1 s"1 at room temperature
12      and atmospheric pressure. These rate constants lead to atmospheric lifetimes (see below) that
13      are still greater than 1 day. Rate constants for monocyclic aromatic compounds with greater
14      than 10 carbon atoms or more are generally not available.
15           The dominant monocyclic aromatic compounds with two  substituents are m-,o-, and
16      p-xylene.  Their recommended OH rate constants range from 1.4 to 2.4 x 10"11 cm3 molec"1 s"1.
17      Similarly, the three isomers of ethyltoluene have recommended OH rate constants ranging from
18      1.2 tol.9 x 1Q"11 cm3 molec"1 s"1. The only other two substituent single-ring aromatic compound
19      for which the OH rate constant has been measured is p-cymene (para-isopropyltoluene), giving a
20      value of 1.5 x 10"11 cm3 molec"1 s"1.
21           OH  rate constants for the C9 trimethyl substituted aromatic hydrocarbons (1,2,3-; 1,2,4-;
22      1,3,5-trimethylbenzene) are higher by a factor of approximately 2.6 over the di-substituted
23      compounds.  Rate constants for the three isomers range from 3.3 to 5.7 x 10"11 cm3 molec"1 s"1.
24      While concentrations for numerous other trisubstituted benzene compounds have been reported
25      (e.g., l,2-dimethyl-4-ethylbenzene),  OH rate constants for trimethylbenzene isomers are the only
26      tri substituted aromatic compounds that have been reported.
27           Aromatic hydrocarbons having substituent groups with unsaturated carbon groups have
28      much higher OH rate constants than  their saturated analogues.  The smallest compound in this
29      group is the C8 AHC, styrene. This compound reacts rapidly with OH and has a recommended
30      rate constant of 5.8 x 10"11 cm3 molec"1 s"1.  (Calvert, 2002).  Other methyl  substituted styrene-
        August 2005                            AX2-33      DRAFT-DO NOT QUOTE OR CITE

-------
     Table AX2-3. Hydroxyl Rate Constants and Atmospheric Lifetimes of Mono- and
              Di-cyclic Aromatic Hydrocarbons (adapted from Atkinson 2000)
Compound
Benzene
Toluene
Ethylbenzene
«-Propylbenzene
Isopropylbenzene
^-Butylbenzene
o-Xylene
m-Xylene
/•-Xylene
o-Ethyltoluene
m-Ethyltoluene
/•-Ethyltoluene
p-Cymene
1,2,3-Trimethylbenzene
1 ,2,4-Trimethylbenzene
1 , 3 ,5 -Trimethylbenzene
Indan
Styrene3
a-Methylstyrene
Napthalene
1 -Methylnapthalene
2-Methylnapthalene
OH Rate Constant
(xlO12)
1.4
5.6
7
5.8
6.3
4.5
14
23
14
12
19
12
14
33
33
57
19
58
51
23
53
52
TOH
(as indicated)
8.3d*
2.1 d
1.7 d
2.0 d
1.8 d
2.6 d
20 h
12 h
19 h
23 h
15 h
24 h
19 h
8.4 h
8.6 h
4.8 h
15 h
4.8 h
5.4 h
12 h
4.8 h
8.8 h
  1 Rate coefficients given as cmVmolec-sec.
  2Lifetime for zero and single alkyl substituted aromatic based on OH concentration of 1 x 106 molec cm"3.
  3 Lifetime for reaction of styrene with NO3 is estimated to be 44 min based on a nighttime NO3 concentration
  of 2.5 x 108 molec cm"3 and a rate coefficient of 1.5 x lCT12cm3/molec-sec.
August 2005
AX2-34
DRAFT-DO NOT QUOTE OR CITE

-------
 1      type compounds (e.g., a-methylstyrene) have OH rate constants within a factor of two of that
 2      with styrene. However, for unsaturated monocyclic aromatic hydrocarbons other processes
 3      including atmospheric removal by NO3 radicals can also be important, particularly at night when
 4      photolysis does not substantially reduce the NO3 radical concentration (see below).
 5           Polycyclic aromatic hydrocarbons are found to a much lesser degree in the atmosphere
 6      than are the monocyclic aromatic hydrocarbons. For example, measurements made in Boston
 7      during 1995 (Fujita et al., 1995) showed that a single PAH (napthalene) was detected in the
 8      ambient morning air at levels of approximately 1% (C/C) of the total monocyclic aromatic
 9      hydrocarbons. 1-methyl and 2-methylnaphthalene have sufficient volatility to be present in the
10      gas phase. Other higher molecular weight PAHs (< 3 aromatic rings), if present, are expected to
11      exist in the gas phase at much lower concentrations than napthalene and are not considered here.
12      OH rate constants for napthalene and the two methyl substituted napthalene compounds have
13      been reviewed by Calvert et al. (2002).  The values recommended (or listed) by Calvert et al.
14      (2002) are given in Table AX2-3.  As seen in the monocyclic aromatic hydrocarbons, the
15      substitution of methyl groups on the aromatic ring increases the OH rate constant, in this case by
16      a factor of 2.3.
17           Some data is available for the reaction of OH with aromatic oxidation products.  (In this
18      context, aromatic oxidation products refer to those products which retain the aromatic ring
19      structure.) These include the aromatic carbonyl compound, benzaldehyde, 2,4-; 2,5-; and
20      3,4-dimethyl-benzaldehyde, and t-cinnamaldehyde.  Room temperature rate constants for these
21      compounds range from 1.3  x 10~n cm3 molec"1 s"1 (benzaldehyde) to 4.8 x 10~n cm3 molec"1 s"1
22      (t-cinnamaldehyde). While the yields for these compounds are typically between 2 to 6%, they
23      can contribute to the aromatic reactivity for aldehydes having high precursor concentration (e.g.,
24      toluene, 1,2,4-trimethylbenzene).  OH also reacts rapidly with phenolic compounds. OH
25      reaction rates with phenols and o-, m, and p-cresol are typically rapid (2.7 to 6.8 x 10~n cm3
26      molec"1 s"1) at room temperature.  Five dimethylphenols and two trimethylphenols have OH
27      reaction rates ranging between 6.6 x 10~n and 1.25 x 10"10 cm3 molec"1 s"1.  Finally, unlike the
28      aromatic aldehydes and phenols, reaction rates for OH + nitrobenzene and OH + m-nitrotoluene
29      are much  lower than the parent molecules, given their electron withdrawing behavior from the
30      aromatic ring. The room temperature rate constants are 1.4 x 10"13 and 1.2 x 10~12, respectively.
        August 2005                            AX2-3 5      DRAFT-DO NOT QUOTE OR CITE

-------
 1           The NO3 radical is also known to react with selected AHCs and aromatic photooxidation
 2      products. Reaction can either occur by hydrogen atom abstraction or addition to the aromatic
 3      ring.  However, these reactions are typically slow for alkyl aromatic hydrocarbons and the
 4      atmospheric removal due to this process is considered negligible. For AHCs having substituent
 5      groups with double bonds (e.g., styrene, a-methylstyrene), the reaction is much more rapid, due
 6      to the addition of NO3 to the double bond. For these compounds, NO3 rate constants are on the
 7      order of 10"12 cm3 molec"1 s"1. This leads to atmospheric lifetimes on the order of about 1 h for
 8      typical night time atmospheric NO3 levels of 2.5 x 108 molec cm"3 (Atkinson,  2000).
 9           The most important reactions of NO3 with AHCs are those which involve phenol and
10      methyl, dimethyl, and trimethyl analogs.  These reactions can be of importance due to the high
11      yields of phenol for the atmospheric benzene oxidation and o-,m-,p-cresol from toluene
12      oxidation.  The NO3 + phenol rate has been given as 3.8 x 10~12 cm3 molec"1 s"1.  Similarly, the
13      cresol isomers each has an extremely rapid reaction rate with NO3 ranging from 1.1 to
14      1.4 x 1CT11 cm3 molec"1 s"1.  As a  result, these compounds, particularly the cresol isomers, can
15      show rapid nighttime losses due to reaction with NO3 with nighttime lifetimes on the order of a
16      few minutes. There is little  data for the reaction of NO3 with dimethylphenols or
17      trimethylphenols which have been found as products of the reaction of OH + m, p-xylene and
18      OH+1,2,4-; 1,3,5-trimethylbenzene.
19
20      AX2.2.8.2 Reaction Products and Mechanisms of Aromatic Hydrocarbon Oxidation
21           An understanding of the mechanism of the oxidation of AHCs is important 1 if O3 is to be
22      accurately predicted in urban atmospheres through modeling studies. As noted above, most
23      monocyclic aromatic hydrocarbons  are removed from the atmosphere through reaction with OH.
24      Thus, product studies of the OH + AHC should provide the greatest information regarding the
25      AHC oxidation products. However, the effort to study these reactions has been intractable over
26      the past two decades due to  a number of difficulties inherent in the OH-aromatic reaction
27      system.  There are several reasons for the slow progress in understanding these mechanisms.
28      (1) Product yields for OH-aromatic  systems are poorly understood; for the most studied system,
29      OH-toluene, approximately  50% of the reaction products have been identified under conditions
30      where NO2 reactions do not dominate the removal of the OH-aromatic adduct. (2) As noted, the
31      reaction mechanism can change as the ratio of NO2 to O2 changes in the system (Atkinson and

        August 2005                            AX2-36       DRAFT-DO NOT QUOTE OR CITE

-------
 1      Aschmann, 1994). Thus, reaction product distributions that may be measured in the laboratory
 2      at high NO2 (or NOX) concentrations may not be applicable to atmospheric conditions. This also
 3      limits the usefulness of models to predict O3 formation to the extent that secondary aromatic
 4      reactions are not completely parameterized in the system.  (3) Aromatic reactions produce highly
 5      polar compounds for which there are few calibration standards available. In most cases,
 6      surrogate compounds have to be used in GC/MS calibrations.  Moreover, it is not at all clear
 7      whether the present sampling techniques or analytical instruments are appropriate to measure the
 8      highly polar products produced in these systems.  (4) Finally for benzene and toluene in
 9      particular, reaction rates of the products are substantially faster than that of the parent
10      compounds.  Thus, it is difficult to measure yields accurately without substantial interferences
11      due to secondary  reactions. Even given these difficulties, over the past decade a body of
12      knowledge has been developed whereby the initial steps in the OH-initiated photooxidation have
13      been established and a  wide range of primary products from each of the major reaction systems
14      have been catalogued.
15           Benzene is one of the most important aromatic hydrocarbons released into the atmosphere
16      and is a recognized carcinogen. However, its reaction with OH  is extremely slow and its
17      contribution to urban O3 formation is generally recognized to be negligible (Carter, 1994).  As a
18      result, relatively few studies have been conducted on the OH reaction mechanism of benzene.
19      Major products of the oxidation of benzene have been found to be phenol and glyoxal (Berndt
20      et al., 1999; Tuazon et  al., 1986).
21           Most of the product analysis and mechanistic work on alkyl aromatic compounds in the gas
22      phase has focused on examining OH reactions with toluene. The primary reaction of OH with
23      toluene follows either of two paths, the first being an abstraction reaction from the methyl group
24      and the second being addition to the ring. It has previously been found that H-atom abstraction
25      from the aromatic ring  is of minor importance (Tully et al., 1981). A number of studies have
26      examined yields of the  benzyl radical formed following OH abstraction from the methyl group.
27      This radical forms the benzyl peroxy radical, which reacts with nitric oxide (NO) leading to the
28      stable products benzaldehyde, with an average yield of 0.06, and benzyl nitrate, with an average
29      yield less than 0.01 (Calvert et al., 2002). Thus, the overall yield for the abstraction channel is
30      less than approximately 7%.
        August 2005                             AX2-37      DRAFT-DO NOT QUOTE OR CITE

-------
 1           It is now generally recognized that addition of OH to the aromatic ring is the major process
 2      removing toluene from the atmosphere and appears to account for more than 90% of the reaction
 3      yield for OH + toluene.  The addition of OH to the ring leads to an intermediate OH-toluene
 4      adduct that can be stabilized or can redissociate to the reactant compounds. For toluene, OH
 5      addition can occur at any of the three possible positions on the ring (ortho, meta, or para) to form
 6      the adduct. Addition of OH to the toluene has been shown to occur predominately at the ortho
 7      position (yield of 0.81) with lesser amounts at the meta (0.05) and para (0.14) positions (Kenley
 8      et al.,  1981). The initial steps for both the abstraction and addition pathways in toluene have
 9      been shown in Figure AX2-6; only the path to form the ortho-adduct is shown, viz. reaction (2).
10           The OH-toluene adduct formed is an energy-rich intermediate that must be stabilized by
11      third bodies in the system to undergo further reaction.  Stabilization has been found to occur at
12      pressures above 100 Torr for most third bodies (Perry et al., 1977; Tully et al., 1981).  Therefore,
13      at atmospheric pressure, the adduct will not substantially decompose back to its reactants as
14      indicated by reaction (-2). The stabilized adduct (I) is removed by one of three processes:
15      H-atom abstraction by O2 to give  a cresol, as in reaction (5); an addition reaction with O2, as in
16      reaction (6); or reaction with NO2 to give m-nitrotoluene as in reaction (7).
17           The simplest fate for the adduct (I) is reaction with O2to form o-cresol.  Data from a
18      number of studies (e.g., Kenley et al., 1981; Atkinson et al., 1980; Smith et al., 1998; Klotz
19      et al.,  1998; summarized by Calvert et al., 2002) over a wide range of NO2 concentrations
20      (generally above 1 ppmv) show an average yield of approximately 0.15 for o-cresol. Most of the
21      measurements suggest the o-cresol yield is independent of total pressure, identity of the third
22      body,  and NO2 concentration (Atkinson and Aschmann, 1994; Moschonas et al.,  1999), but the
23      data tend to be scattered. This finding suggests that the addition of NO2 to the hydroxy
24      methylcyclo-hexadienyl radical does not contribute to the formation of phenolic-type
25      compounds.  Fewer studies have been conducted for m and p-cresol yields, but the results of two
26      studies indicate the yield is approximately 0.05 (Atkinson et al., 1980; Gery et al., 1985; Smith
27      et al.,  1998). The data suggests good agreement between the relative yields of the cresols from
28      the product studies at atmospheric pressure and studies at reduced pressures. Thus, H-atom
29      abstraction from adducts formed at all positions appears to represent approximately 20% of the
3 0      total yi el d for toluene.
31

        August 2005                             AX2-3 8       DRAFT-DO NOT QUOTE OR CITE

-------
         CH3
                  +      -OH
      Toluene

         CH3
                  +      -OH
         CH2OO'
                  +      NO
              OH
                                 02
                  +      O2
                  +      NO,
                                              CH200'
                                                        +      H2O
                                              CH3
                                                   OH
       (I)

       CHO
                                              CH2ON02
                                              CH3
                                                   OH
                                     (1)
                         NO9         (3)
                                                                            (4)
                 +       HOo         (5)
                                                                            (6)
                         H9O         (7)
Figure AX2-6.  Initial steps in the photooxidation mechanism of toluene initiated by its
               reaction with OH radicals.
August 2005
AX2-39      DRAFT-DO NOT QUOTE OR CITE

-------
 1           The OH-toluene adduct also reacts with O2 to form a cyclohexadienyl peroxy radical (III),
 2      shown as a product of reaction (6) after rearrangement. This radical can undergo a number of
 3      possible processes. Most of these processes lead to ring fragmentation products, many of which
 4      have been seen in several studies (Dumdei and O'Brien, 1984; Shepson et al., 1984).  Ring-
 5      fragmentation products are frequently characterized by multiple double bonds and/or multiple
 6      functional groups.  As such, these products are highly reactive and extremely difficult to detect
 7      and quantify.
 8           Klotz et al. (1997; 1998) have suggested that the intermediate could also follow through a
 9      mechanism where  toluene oxide/oxepin could be formed following the addition of O2 to the
10      OH-aromatic adduct. Recent experiments suggest that the formation of o-cresol through the
11      photolysis of toluene oxide/oxepin is only a minor contributor to the overall o-cresol that has
12      been measured (Klotz et al., 1998). This result contrasts to the high yield observed for the
13      formation of phenol from the photolysis of benzene oxide/oxepin (Klotz et al.,  1997). Recently,
14      Berndt et al. (1999) used a flow tube to test the hypothetical formation of benzene oxide/oxepin
15      from the OH + benzene reaction at pressures below 100 torr. They saw very little evidence for
16      its formation.
17           A few studies have been conducted to identify fragmentation products using a variety of
18      instruments.  Several approaches have been used that employ structural methods, particularly
19      mass spectrometry (MS), to identify individual products formed during the photooxidation.
20      In one approach (Dumdei and O'Brien, 1984), the walls of the reaction chamber were extracted
21      following an extended irradiation. In this study, the analysis was conducted by tandem mass
22      spectrometry (MS/MS), which allowed products to be separated without the use of a
23      chromatographic stationary phase. The investigators  reported 27  photooxidation products from
24      toluene, with 15 reportedly from ring fragmentation processes. However, the study was purely
25      qualitative and product yields could not be obtained.  No distinction could be made between
26      primary and secondary products from the reaction because extended irradiations and species in
27      various isotopic forms could not be differentiated. More refined approaches using atmospheric
28      pressure ionization-tandem mass spectrometry has been used to study toluene (Dumdei et al.,
29      1988) and m and p-xylene (Kwok et al., 1997) photooxidation.
30           In another study,  Shepson et al. (1984) demonstrated that a number of these fragmentation
31      products could be analyzed by gas chromatography. Fragmentation products detected in two

        August 2005                             AX2-40       DRAFT-DO NOT QUOTE OR CITE

-------
 1      investigations (Dumdei and O'Brien, 1984; Shepson et al., 1984) included glyoxal, methyl
 2      glyoxal, butenedial, 4-oxo-2-pentenal, hydroxybutenedial, l-pentene-3,4-dione, l-butene-3,4-
 3      dione, and methyl vinyl ketone. Additional evidence (Shepson et al., 1984) for fragmentation
 4      processes came from the detection of 2-methylfuran and furfural. These compounds, although
 5      cyclic in structure, result from a bridged oxygen intermediate. Yields of the detected
 6      fragmentation products were subsequently measured in a number of studies (e.g., Bandow et al.,
 7      1985a,b; Tuazon et al., 1986; Smith et al., 1998), were typically under 15% on a reacted carbon
 8      basis.
 9           An additional possible pathway for reaction of the OH-toluene adduct is by reaction
10      withNO2 to give isomers of nitrotoluene.  A yield of approximately 0.015 at NO2 concentrations
11      of about 1 ppmv has been measured (Atkinson et al., 1991).  Although this yield itself is fairly
12      minor, the investigators reported a positive intercept in plotting the nitrotoluene concentration
13      against the NO2 concentration; however, the data were considerably scattered. The positive
14      intercept has been interpreted as suggesting that the OH-toluene adduct does not add O2.  This
15      finding would require, therefore, another mechanism than that described above to be responsible
16      for the fragmentation  products.
17           The results of this study can be compared to experiments which directly examined the OH
18      radical loss in reactions of OH with toluene and other  aromatic compounds. Knipsel and co-
19      workers (Knispel et al., 1990) have found a double exponential decay for toluene loss in the
20      presence of added O2, a rapid decay reflective of the initial adduct formation and a slower decay
21      reflecting loss of the adduct by O2 or other scavengers. From the decay data in the presence
22      of O2, they determine  a loss rate for the OH-toluene of 5.4 x  10"16 cm3 molec"1 s"1.  Use of this
23      rate constant suggests that the loss rate of 2500 s"1 for the adduct in the presence of air at
24      atmospheric pressure. This loss rate compares to a loss due to NO2 (with a nominal  atmospheric
25      concentration of 0.1 ppmv) of about 100 s"1.  This finding suggests that removal of the OH-
26      toluene adduct by O2 is a far more important loss process than removal by NO2 under
27      atmospheric conditions which is in contrast other findings (Atkinson et al., 1991). This finding
28      was confirmed by the recent experiments from Moschonas et al. (1999).
29           Therefore, studies on the disposition of toluene following OH reaction can be summarized
30      as follows. It is generally accepted that H-atom abstraction from the methyl group by OH is a
31      relatively minor process accounting for a 6 to 7% yield in the OH reaction with toluene.

        August 2005                            AX2-41      DRAFT-DO NOT QUOTE OR CITE

-------
 1      Addition of OH to toluene to form an intermediate OH-toluene adduct is the predominant
 2      process. At atmospheric pressure, ring-retaining products such as the cresol and nitrotoluenes
 3      account for another 20% of the primary reaction products (Smith et al., 1999).  The remaining
 4      70 to 75% of the products are expected to be ring fragmentation products in the gas phase,
 5      having an uncertain mechanism for formation. Many of these fragmentation products have been
 6      detected, but appear to form at low yields, and relatively little quantitative information on their
 7      formation yields exists.  As noted earlier, some of these products contain multiple double bonds,
 8      which are likely to be highly reactive with OH or photolyze which enhances the reactivity of
 9      systems containing aromatics. Mechanisms that cannot adequately reflect the formation of
10      fragmentation products are likely to show depressed reactivity for the oxidation of toluene and
11      other aromatic compounds.
12           The number of studies of the multiple-substituted alkyl aromatics, such as the xylenes or
13      the trimethylbenzenes, is considerably smaller than for toluene.  Kinetic studies have focused on
14      the OH rate constants for these compounds. For the xylenes, this rate constant is typically a
15      factor of 2 to 5 greater than that for OH + toluene. Thus, the OH reactivity of the fragmentation
16      products is similar to that of the parent compounds, potentially making the study of the primary
17      products of the xylenes less prone to uncertainties from secondary reactions of the primary
18      products than is the case for toluene.
19           Products from the  OH reaction with the three xylenes have been studied most
20      comprehensively in a smog chamber using long-path FTIR (Bandow and Washida, 1985a) and
21      gas chromatography (Shepson et al., 1984; Atkinson and Aschmann, 1994; Smith et al., 1999).
22      Ring-fragmentation yields of 41, 55, and 36% were estimated for o-, m, and p-xylene,
23      respectively, based on the dicarbonyl compounds, glyoxal, methyl glyoxal, biacetyl, and 3-
24      hexene-2,5-dione detected during the photooxidation.  These values could be lower limits, given
25      that Shepson et al. (1984) report additional fragmentation products from o-xylene, including
26      l-pentene-4,5-dione, butenedial, 4-oxo-pentenal, furan, and 2-methylfuran. In the earlier
27      studies, aromatic concentrations were in the range of 5 to 10 ppmv with NOX at 2 to 5 ppmv.
28      At atmospheric ratios of NO2 and O2, the observed yields could be different. Smith et al. (1999)
29      examined most of the ring retaining products in the OH + m-xylene and OH + p-xylene systems.
30      In each case, tolualdehyde isomers, dimethylphenol isomers, and nitro xylene isomers specific
31      for each system were detected.  The total ring retaining yield for OH + m-xylene was 16.3%; the

        August 2005                             AX2-42      DRAFT-DO NOT QUOTE OR CITE

-------
 1      yield for OH + p-xylene it was 24.5%. A mass balance approach suggests that respective ring-
 2      fragmentation yields of 84% and 76%, respectively. Kwok et al. (1997) also measured products
 3      from the OH + m and p-xylene systems using atmospheric pressure ionization-tandem mass
 4      spectrometry. Complementary ring-fragmentation products to glyoxal, methylglyoxal, and
 5      biacetyl were detected from the parent ion peaks, although the technique did not permit the
 6      determination of reaction yields.
 7           Smith et al. (1999) also studied ring fragmentation products from the reaction of OH
 8      with 1,2,4- and 1,3,5-trimethylbenzene. Ring-retaining products from the  reaction with
 9      1,2,4-trimethylbenzene gave three isomers each of dimethylbenzaldehyde  and trimethylphenol
10      as expected by analogy with toluene. However, the ring-retaining products only accounted for
11      5.8% of the reacted carbon.  Seven additional ring-fragmentation products were also detected
12      from the reaction, although the overall carbon yield was 47%. For 1,3,5-trimethylbenzene, its
13      reaction with OH leads to only two ring-retaining products, 3,5-dimethyl-benzaldehyde and
14      2,4,6-trimethylphenol, given its molecular symmetry.  Only a single fragmentation product was
15      detected, methyl glyoxal, at a molar  yield of 90%.  The overall carbon yield in this case was
16      61%. The formation of relatively low yields of aromatic aldehydes and methylphenols suggests
17      that NOX removal by these compound in these reaction systems will be minimized (see below).
18           In recent years, computational  chemistry studies have been applied to reaction dynamics of
19      the OH-aromatic reaction systems. Bartolotti and Edney (1995) used density functional-based
20      quantum mechanical calculations to  help identify intermediates of the OH-toluene adduct.  These
21      calculations were consistent with the main addition of OH to the ortho position of toluene
22      followed by addition of O2 to the meta position of the adduct. The reaction energies suggested
23      the formation of a carbonyl epoxide  which was subsequently detected in aromatic oxidation
24      systems by Yu and Jeffries, (1997).  Andino et al. (1996) conducted ab initio calculations using
25      density functional theory with semiempirical intermediate geometries to examine the energies of
26      aromatic intermediates and determine favored product pathways.  The study was designed to
27      provide some insight into the fragmentation mechanism, although only a group additivity
28      approach to calculate AH^ was used to investigate favored reaction pathways. However, the
29      similarity in energies of the peroxy radicals formed from the  O2 reaction with the OH-aromatic
30      adduct were very similar in magnitude making it difficult to differentiate among structures.
        August 2005                             AX2-43      DRAFT-DO NOT QUOTE OR CITE

-------
 1           A detailed analysis of toluene oxidation using smog chamber experiments and chemical
 2     models (Wagner et al., 2003) shows that there are still large uncertainties in the effects of
 3     toluene on O3 formation. A similar situation is likely to be found for other aromatic
 4     hydrocarbons.
 5
 6     AX2.2.8.3 The Formation of Secondary Organic Aerosol as a Sink for Ozone Precursors
 7           Aromatic hydrocarbons are known to generate secondary organic aerosol (SOA) following
 8     their reaction with OH or other reactive oxidants. Secondary organic aerosol refers to the
 9     formation of fine particulate matter either through nucleation processes or through condensation
10     onto existing particles.  Over the last 12 years numerous experiments have been conducted in
11     environmental chambers to determine the yield of secondary organic aerosol as a function of the
12     reacted aromatic hydrocarbon.  A review of the results of these studies can be found in the latest
13     Air Quality for Parti culate Matter Document (U.S. Environmental Protection Agency, 2003).
14           The extent to which aromatic reaction products are removed from the gas phase and
15     become incorporated in the particle phase will influence the extent to which oxygenated organic
16     compounds will not be available for participation in the aromatic mechanisms that lead to O3
17     formation. However, this may be overstated to some degree for products of aromatic precursors.
18     First, at atmospheric loading levels of organic parti culate matter, the SOA yields  of the major
19     aromatic hydrocarbons are in the low percent range. Second, the aromatic products that are
20     likely to condense on particles are likely to be highly oxygenated and have OH reaction rates
21     that make them largely unreactive. Thus, while there may be some reduction of O3 formation, it
22     is not expected to be large.
23
24     AX2.2.9   Importance of Oxygenated VOCs
25           The role of oxygenated VOCs in driving O3 production has generated increasing interest
26     over the past decade.  These VOCs include carbonyls, peroxides, alcohols, and organic acids.
27     They are produced in the atmosphere by oxidation of hydrocarbons, as discussed  above, but  are
28     also directly emitted to the atmosphere, in particular by vegetation (Guenther et al., 2000).
29     In rural and remote atmospheres, oxygenated VOCs often dominate over nonmethane
30     hydrocarbons in terms of total organic carbon mass and reactivity (Singh et al., 2004).  The most
31     abundant by mass of these oxygenated VOCs is usually methanol, which is emitted by

       August 2005                            AX2-44      DRAFT-DO NOT QUOTE OR CITE

-------
 1      vegetation and is present in U.S. surface air at concentrations of typically 1-10 ppbv (Heikes
 2      et al., 2002).
 3          Most oxygenated VOCs react with OH to drive O3 production in a manner similar to the
 4      hydrocarbon chemistry discussed in the previous sections.  In addition, carbonyl compounds
 5      (aldehydes and ketones) photolyze to produce peroxy radicals that can accelerate O3 production,
 6      thus acting as a chemical amplifier (Jaegle et  al., 2001). Photolysis of formaldehyde by (A.26b)
 7      was discussed in section AX2.2.5. Also of particular importance is the photolysis of acetone
 8      (Blitz et al., 2004):
 9
                                               20->
                           (CH3)2C(O) + hv	?_» CH3C(O)O2-                    (AX2-41)

10
11      producing organic peroxy radicals that subsequently react with NO to produce O3. The
12      peroxyacetyl radical CH3C(O)OO can also react with NO2 to produce PAN, as discussed in
13      Section AX2.2.4. Photolysis of acetone are a minor but important source of HO2 radicals in the
14      upper troposphere (Arnold et al., 2004).
15
16      AX2.2.10   Influence of Multiphase Chemical Processes
17          In addition to reactions occurring in the gas phase, reactions occurring on the surfaces of or
18      within cloud droplets and airborne particles also occur. Their collective surface area is huge,
19      implying that collisions with gas phase species occur on very short time scales.  The integrated
20      aerosol surface area ranges from 4.2 x 1CT7 cm2/cm3 for clean continental conditions to
21      1.1 x 1CT5 cm2/cm3 for urban average conditions (Whitby, 1978).  There have been substantial
22      improvements in air quality especially in urban areas since the time these measurements were
23      made and so the U.S. urban values should be  scaled downward by roughly a factor of two to
24      four. The resulting  surface  area is still substantial and the inferred collision time scale of a
25      gaseous molecule with a particle ranges from a few seconds or less to a few minutes. These
26      inferred time scales imply that heterogenous reactions will generally be much less important than
27      gas phase reactions  for determining radical concentrations especially when reaction probabilities
28      much less then unity are considered. A large body of research has accumulated recently
        August 2005                             AX2-45      DRAFT-DO NOT QUOTE OR CITE

-------
 1      regarding chemical processes in cloud droplets, snow and ice crystals, wet (deliquesced)
 2      inorganic particles, mineral dust, carbon chain agglomerates and organic carbon-coated particles.
 3           Jacob's (2000) comprehensive review of the potential influences of clouds and aerosols on
 4      tropospheric O3 cycling provides the starting point for this section. Updates to that review will
 5      also be provided. Jacob's review evaluates the literature available through late 1999, discusses
 6      major areas of uncertainty, recommends experiments to reduce uncertainties, and (based on then
 7      current information) recommends specific multiphase pathways that should be considered in
 8      models of O3 cycling. In regard to the latter, Jacob's recommendations should be viewed as
 9      conservative. Specifically, only reasonably well-constrained pathways supported by strong
10      observational evidence are recommended for inclusion in models. Several poorly resolved
11      and/or controversial pathways that may be significant in the ambient troposphere lack sufficient
12      constraints for reliable modeling.  Some of these areas are discussed in more detail below.
13      It should be noted at the outset that many of the studies  described in this section involve either
14      aerosols that are not found commonly throughout the United States (e.g., marine aerosol) or
15      correspond to unaged particles (e.g., soot, mineral dust). In many areas of the United States,
16      particles accrete a layer of hydrated H2SO4 which will affect the nature of the multiphase
17      processes occurring on particle  surfaces.
18           Major conclusions from this review are summarized as follows (comments are given in
19      parentheses):
20       HOX Chemistry
21         (1)  Catalytic O3 loss via reaction of O2  + O3(aq) in clouds appears to be inefficient.
22         (2)  Aqueous-phase loss of HCHO in clouds appears to be negligible (see also Lelieveld and
                Crutzen, 1990).
23         (3)  Scavenging of HO2 by  cloud droplets is significant and can be acceptably parameterized
                with a reaction probability of YHo2 = 0.2, range 0.1 to 1, for HO2 -» 0.5 H2O2. However,
                this approach may overestimate HO2 uptake because the influence of HO2(aq) on the
                magnitude (and direction) of the flux is ignored.
24         (4)  The uptake of alkyl peroxy radicals by aerosols is probably negligible.
25         (5)  Hydrolysis of CH3C(O)OO in aqueous aerosols may be important at night in the
                presence of high PAN  and aerosol surface area; YcH3c(opo = 4 x io~3 is recommended.
        August 2005                             AX2-46       DRAFT-DO NOT QUOTE OR CITE

-------
 1       NOX Chemistry
 2         (6)  Hydrolysis of N2O5 to HNO3 in aqueous aerosols is important (Section AX2.2.4)
               (and can be parameterized with JHN03 = 0.01 to 0.1 [Schutze and Herrmann, 2002;
               Hallquist et al., 2003]).
 3         (7)  Although the mechanism is uncertain, heterogeneous conversion of NO2 to HONO
               on aerosol surfaces should be considered with YNo2 = 10 4 (range 10~6 to 10~3) for
               NO2 ->• 0.5 HONO + 0.5 HNO3. (This reaction also occurs on snow, Crawford et al.,
               2001). Wet and dry deposition sinks for HONO should also be considered although
               scavenging by aerosols appears to be negligible.
 4         (8)  There is no evidence for significant multiphase chemistry involving PAN.
 5         (9)  There is no evidence for significant conversion of HNO3 to NOX in aerosols.
 6       Heterogeneous ozone loss
 1        (10)  There is no evidence for significant loss of O3 to aerosol surfaces (except during
               dust storms observed in East Asia, e.g.,  Zhang and Carmichael, 1999).

 8       Halogen radical chemistry
 9        (11)  There is little justification for considering BrOx and C1OX chemistry (except perhaps
               in limited areas of the United States  and nearby coastal areas).
10          Most of the above conclusions remain valid but, as detailed below, some should be
11      qualified based on recently published findings and on reevaluation of results form earlier
12      investigations.
13
14      AX2.2.10.1  HOX and Aerosols
15          Field measurements of HOX reviewed by Jacob (2000) correspond to regions with
16      relatively low aerosol concentrations (e.g., Mauna Loa [Cantrell et al.,  1996]; rural Ontario
17      [Plummer et al., 1996]; and the upper troposphere [Jaegle et al., 1999]). In all cases, however,
18      significant uptake of HO2 or HO2 + RO2 radicals by aerosols was inferred based on imbalances
19      between measured concentrations of peroxy radicals and photochemical models of gas-phase
20      chemistry.  Laboratory studies using artificial aerosols (both deliquesced and solid) confirm
21      uptake but the actual mechanism remains unclear.  Several investigations report significant HOX
22      and H2O2 production in cloud water (e.g., Anastasio et al., 1994). However the potential
23      importance of this source is considered unlikely  because measurements in continental air show
        August 2005                             AX2-47      DRAFT-DO NOT QUOTE OR CITE

-------
 1      no evidence of missing sources for HOX or H2O2.  No investigations involving the potential
 2      influences of marine aerosols as sources or sinks for HOX were considered in the above analysis.
 3           Relative to conservative seawater tracers such as Mg2+ and Na+, organic C associated with
 4      sea-salt aerosols is typically enriched by 2 to 3 orders of magnitude in both polluted (e.g.,
 5      Hoffman and Duce, 1976, 1977; Turekian et al., 2003) and remote regions (Chesselet et al.,
 6      1981). This organic C originates from three major sources: 1) organic surfactants concentrated
 7      from bulk seawater on walls of subsurface bubbles (Tseng et al., 1992), 2) the surface microlayer
 8      of the ocean (Gershey, 1983), and 3) condensation of organic gases (Pun et al., 2000).
 9      Coagulation of chemically distinct aerosols (e.g., via cloud processing) may also contribute
10      under some conditions.
11           Resolving chemical processes involving particles in the marine boundary layer (MBL) is
12      constrained by the relative scarcity of measurements of paniculate organic carbon (POC)
13      (Penner, 1995) and its molecular composition (Saxena et al., 1995). In MBL regions impacted
14      by direct continental outflow, POC may constitute more that half of the total dry aerosol mass
15      (Hegg et al., 1997). Carbon isotopic compositions in the polluted North Atlantic MBL indicate
16      that, on average, 35% to 40% of POC originates from primary (direct injection) and secondary
17      (condensation of gases) marine sources (Turekian et al., 2003).
18           The photolysis of dissolved organic compounds is a major source for OH, H2O2,  and
19      C-centered radicals in both the surface ocean (e.g., Blough and Zepp, 1995; Blough, 1997;
20      Mopper  and Kieber, 2000) and in marine aerosols (e.g., McDow et al., 1996). Relative to the
21      surface ocean, however, production rates in the aerosol are substantially greater per unit volume
22      because  organic matter is highly enriched (Turekian et al., 2003) and aerosol pH is much lower
23      (Keene et al., 2002a). Lower pHs increase rates of many reactions including acid-catalyzed
24      pathways such as the breakdown of the HOC1" radical (King et al., 1995), the formation of H2O2
25      from the photolysis of phenolic compounds (Anastasio et al., 1997), and the photolysis of organic
26      acids.
27           To provide a semi-quantitative context for the potential magnitude of this source, we
28      assume a midday OH production rate in surface seawater of 10~n M sec"1 (Zhou and Mopper,
29      1990) and a dissolved organic carbon enrichment of 2 to 3 orders of magnitude in sea-salt
30      aerosols. This yields an estimated OH production rate in fresh (alkaline) sea-salt aerosols of 10~9
31      to 10~8 M sec"1.  As discussed above, rapid (seconds to minutes) acidification of the aerosol

        August 2005                             AX2-48      DRAFT-DO NOT QUOTE OR CITE

-------
 1      should substantially enhance these production rates. Consequently, the midday OH production
 2      rates from marine-derived organic matter in acidified sea-salt aerosols may rival or perhaps
 3      exceed midday OH scavenging rates from the gas phase (approximately 1CT7 M sec"1; [Chameides
 4      and Stelson, 1992]). Scavenging is the only significant source for OH in acidified sea-salt
 5      aerosols considered by many current models.
 6           Limited experimental evidence indicates that these pathways are important sources of HOX
 7      and ROX in marine air and possibly in coastal cities.  For example, the absorption of solar energy
 8      by organic species dissolved in cloud water (e.g., Faust et al., 1993; Anastasio et al., 1997) and in
 9      deliquesced sea-salt aerosols (Anastasio et al., 1999) produces OH, HO2, and H2O2. In addition,
10      Fe(III) complexation by oxalate and similar ligands to metal such as iron can greatly enhance
11      radical production through ligand to metal charge transfer reactions (Faust, 1994; Hoigne et al.,
12      1994). Oxalate and other dicarboxylic anions are ubiquitous components of MBL aerosols in
13      both polluted (e.g., Turekian et al., 2003) and remote regions (Kawamura et al., 1996).
14           Substantial evidence exists for washout of peroxy radicals. Near solar noon, mixing ratios
15      of total HOX plus ROX radicals generally fall in the 50 ppt range, but during periods of rain these
16      values dropped to below the detection limit of 3 to 5 ppt (Andres-Hernandez et al., 2001; Burkert
17      et al., 2001a; Burkert et al., 2001b; Burkert et al., 2003).  Such low concentrations cannot be
18      explained by loss of actinic radiation, because nighttime radical mixing ratios were higher.
19           Burkert et al.  (2003) investigated the diurnal behavior of the trace gases and peroxy radicals
20      in the clean and polluted MBL by comparing observations to a time dependant, zero-dimensional
21      chemical model.  They identified significant differences between the diurnal behavior of RO2*
22      derived from the model and  that  observed possibly attributable to multiphase chemistry.  The
23      measured HCHO concentrations differed from the model results and were best explained by
24      reactions involving low levels of Cl.
25           Finally, photolytic NO3  reduction is important in the surface ocean (Zafiriou and True,
26      1979) and could contribute to OH production in sea-salt aerosols. Because of the
27      pH-dependence of HNO3 phase partitioning, most total nitrate (HNO3 + particulate NO3 ) in
28      marine air is associated with sea  salt (e.g., Huebert et al., 1996; Erickson et al., 1999). At high
29      mM concentrations of NO3  in sea-salt aerosols under moderately polluted conditions (e.g.,
30      Keene et al., 2002) and with quantum yields for OH production of approximately 1% (Jankowski
31      et al., 2000), this pathway would be similar in magnitude to that associated with scavenging

        August 2005                             AX2-49      DRAFT-DO NOT QUOTE OR CITE

-------
 1      from the gas phase and with photolysis of dissolved organics. Experimental manipulations of
 2      marine aerosols sampled under relatively clean conditions on the California coast confirms that
 3      this pathway is a major source for OH in sea-salt solutions (Anastasio et al., 1999).
 4           Although largely unexplored, the potential influences of these poorly characterized radical
 5      sources on O3 cycling in marine air are probably significant. At minimum, the substantial
 6      inferred concentrations of HO2 in aerosol solutions would diminish and perhaps reverse HO2
 7      scavenging by marine aerosols and thereby increase O3 production relative to models based on
 8      Jacob's (2000) recommended reaction probability.
 9
10      AX2.2.10.2   NOX Chemistry
11           Jacob (2000) recommended as a best estimate, YN o = 0.1 for the reaction probability
12      of N2O5 on aqueous aerosol surfaces with conversion to HNO3. Recent laboratory studies on
13      sulfate and organic aerosols indicates that this reaction probability should be revised downward,
14      to a range 0.01-0.05 (Kane et al., 2001; Hallquist et al., 2003; Thornton et al., 2003). Tie et al.
15      (2003) found that a value of 0.04 in their global model gave the best simulation of observed NOX
16      concentrations over the Arctic in winter. A decrease in N2O5 slows down the removal of NOX
17      and thus increases O3 production. Based on the consistency between measurements of NOy
18      partitioning and gas-phase models, Jacob (2000) considers it unlikely that significant HNO3 is
19      recycled to NOX in the lower troposphere. However, only one of the reviewed studies (Schultz
20      et al., 2000) was conducted in the marine troposphere and none were conducted in the MBL.
21      An investigation over the equatorial Pacific reported discrepancies between observations and
22      theory (Singh et al., 1996b) that might be explained by HNO3 recycling. It is important to
23      recognize that both Schultz et al. (2000) and Singh et al. (1996b) involved aircraft sampling
24      which, in the MBL, significantly under represents sea-salt aerosols and thus most total NO3
25      (HNO3 + N(V) and large fractions of NOy in marine air (e.g., Huebert et al., 1996).
26      Consequently, some caution is warranted when interpreting constituent ratios and NOy budgets
27      based on such data.
28           Recent work in the Arctic has quantified significant photochemical recycling of NO3
29      to NOX and perturbations of OH chemistry in snow (Honrath et al., 2000; Dibb et al., 2002;
30      Domine and Shepson, 2002),  which suggests the possibility of similar multiphase pathways
31      occurring in aerosols. As mentioned above, recent evidence also indicates that NO3 is

        August 2005                             AX2-50       DRAFT-DO NOT QUOTE OR CITE

-------
 1      photolytically reduced to NO2 (Zafariou and True, 1979) in acidic sea-salt solutions (Anastasio
 2      et al., 1999). Further photolytic reduction of NO2" to NO (Zafariou and True, 1979) could
 3      provide a possible mechanism for HNO3 recycling. Early experiments reported production
 4      of NOX during the irradiation of artificial seawater concentrates containing NO3  (Petriconi and
 5      Papee, 1972). Based on the above, we believe that HNO3 recycling in sea-salt aerosols is
 6      potentially important and warrants further investigation. Other possible recycling pathways
 7      involving highly acidic aerosol solutions and soot are reviewed by Jacob (2000).
 8           Ammann et al. (1998) reported the efficient conversion of NO2 to HONO on fresh soot
 9      particles in the presence of water. They suggest that interaction between NO2 and soot particles
10      may account for high mixing ratios of HONO observed in urban environments. Conversion
11      of NO2 to HONO and subsequent photolysis to NO + OH would constitute an NOx-catalyzed O3
12      sink involving snow. High concentrations of HONO can lead to the rapid growth in OH
13      concentrations shortly after sunrise, giving a "jump start" to photochemical smog formation.
14      Prolonged exposure to ambient oxidizing agents appears to deactivate this process. Broske et al.
15      (2003) studied the interaction of NO2 on secondary organic aerosols and concluded that the
16      uptake coefficients were too low for this reaction to be an important source of HONO in the
17      troposphere.
18           Choi and Leu (1998) evaluated the interactions of nitric acid on a model black carbon soot
19      (FW2), graphite, hexane and kerosene soot. They found that HNO3 decomposed to NO2
20      and H2O at higher nitric acid surface coverages, i.e., P(HNO3) > = 10~4 Torr.  None of the soot
21      models used were reactive at low nitric acid coverages, at P(HNO3) = 5 x  io~7 Torr or at lower
22      temperatures (220K). They conclude that it is unlikely that aircraft soot in the upper
23      troposphere/lower stratosphere reduces HNO3.
24           Heterogeneous production on soot at night is believed to be the mechanism by which
25      HONO accumulates to provide an early morning source of HOX in high NOX environments
26      (Harrison et al., 1996; Jacob, 2000). HONO has been frequently observed to accumulate to
27      levels of several ppb over night, and has been attributed to soot chemistry (Harris et al., 1982;
28      Calvert et al., 1994; Jacob, 2000).
29           Longfellow et al. (1999) observed the formation of HONO when methane, propane, hexane
30      and kerosene soots were exposed to NO2.  They estimate that this reaction may account for some
31      part of the unexplained high levels of HONO observed in urban areas. They comment that

        August 2005                            AX2-51       DRAFT-DO NOT QUOTE OR CITE

-------
 1      without details about the surface area, porosity and amount of soot available for this reaction,
 2      reactive uptake values cannot reliably be estimated. They comment that soot and NO2 are
 3      produced in close proximity during combustion, and that large quantities of HONO have been
 4      observed in aircraft plumes.
 5           Saathoff et al. (2001) studied the heterogeneous loss of NO2, HNO3, NO3/N2O5,
 6      HO2/HO2NO2 on soot aerosol using a large aerosol chamber.  Reaction periods of up to several
 7      days were monitored and results used to fit a detailed model.  They derived reaction probabilities
 8      at 294 K and 50% RH for NO2, NO3, HO2 and HO2NO2 deposition to soot, HNO3 reduction
 9      to NO2, and N2O5 hydrolysis.  When these probabilities were included in photochemical box
10      model calculations of a 4-day smog event, the only noteworthy influence of soot was a 10%
11      reduction in the second day O3 maximum, for a soot loading of 20 jig m"3, i.e., a factor of 2 to
12      10 times observed black carbon loadings seen during extreme U.S. urban pollution events,
13      although such concentrations  are observed routinely in the developing world.
14           Mufioz and Rossi (2002) conducted Knudsen cell studies of HNO3 uptake on black and
15      grey decane soot produced in lean and rich flames, respectively.  They observed HONO as the
16      main species released following nitric acid uptake on grey soot, and NO and traces of NO2 from
17      black soot.  They conclude that these reactions would only have relevance in special situations in
18      urban settings where soot and HNO3 are present in high concentrations simultaneously.
19
20      AX2.2.10.3   Halogen Radical Chemistry
21           Barrie et al.  (1988) first suggested that halogen chemistry on snow surfaces in the Arctic
22      could lead to BrOx formation and  subsequent O3 destruction. More recent work suggests that
23      halogen radical reactions may influence O3 chemistry in mid latitudes as well.
24           The weight of available evidence supports the hypothesis that halogen radical chemistry
25      significantly influences O3 cycling over much of the marine boundary layer at lower latitudes
26      and in at least some other regions of the troposphere. However, proposed chemical mechanisms
27      are associated with substantial uncertainties and, based on available information,  it appears
28      unlikely that a simple parameterization (analogous to those recommended by Jacob (2000) for
29      other multiphase transformations)  would adequately capture major features of the underlying
30      transformations.
        August 2005                            AX2-52      DRAFT-DO NOT QUOTE OR CITE

-------
 1           Most of the Cl and Br in the marine boundary layer are produced in association with
 2      sea-salt aerosols by wind stress at the ocean surface (e.g., Gong et al., 1997).  Fresh aerosols
 3      rapidly dehydrate towards equilibrium with ambient water vapor and undergo other chemical
 4      processes involving the scavenging of reactive gases, aqueous-phase transformations, and
 5      volatilization of products. Many of these processes are strongly pH-dependent (Keene et al.,
 6      1998). Throughout most of the marine boundary layer, sea-salt alkalinity is tritrated rapidly
 7      (seconds to minutes) by ambient acids (Chameides and Stelson, 1992; Erickson et al., 1999) and,
 8      under a given set of conditions, the pHs of the super-jam, sea-salt size fractions are buffered to
 9      similar values via HC1 phase partitioning (Keene and Savoie, 1998; 1999; Keene et al., 2002).
10           Model calculations  based on the autocatalytic halogen activation mechanism (Vogt et al.,
11      1996; Keene et al., 1998; Sander et al., 1999; von Glasow et al., 2002a,b; Pszenny et al.,  2003;
12      Sander et al., 2003) predict that most paniculate Br- associated with acidified sea-salt aerosol
13      would react to form Br2 and BrCl, which subsequently volatilize and photolyze in sunlight to
14      produce atomic Br and Cl. Most Br atoms recycle in the gas phase via
15

                                   •Br + O3 -> BrO + O2                            (AX2-42)
16

                                  BrO- + HO2' ->> HOBr + O2                           (AX2-43)
17
                                   HOBr + hv -^ 'OH + 'Br                           (AX2-44)
18
19      and thereby catalytically  destroy O3, analogous  to Br cycling in the  stratosphere (e.g.,
20      Mozurkewitch, 1995; Sander and Crutzen, 1996).  Side reactions with HCHO and other
21      compounds produce HBr, which is either scavenged and recycled through the aerosol or  lost to
22      the surface via wet and dry deposition (Dickerson et al., 1999).
23           Cl-radical chemistry influences O3 in two  ways (e.g., Pszenny et al., 1993).  Some atomic
24      Cl in marine air reacts directly with O3 initiating a catalytic sequence analogous to that of Br
25      (AX2-42 through AX2-44 above). However, most atomic Cl in the MBL reacts with
26      hydrocarbons (which, relative to the stratosphere, are present at high concentrations) via
27      hydrogen extraction to form HC1 vapor.  The enhanced supply of odd hydrogen radicals from

        August 2005                             AX2-53       DRAFT-DO NOT QUOTE OR CITE

-------
 1      hydrocarbon oxidation leads to O3 production in the presence of sufficient NOX. Thus, Cl
 2      chemistry represents a modest net sink for O3 when NOX is less than 20 pptv and a net source at
 3      higher NOX.  Although available evidence suggest that significant Cl-radical chemistry occurs in
 4      clean marine air, its net influence on O3 appears to be small relative to that of Br and I.
 5           In addition to Br and Cl, several lines of recent evidence suggests that an autocatalytic
 6      cycle also sustains I-radical  chemistry leading to significant net O3 destruction in marine air
 7      (Vogt et al., 1996, 1999; von Glasow et al., 2002a).  The cycle is initiated by photolysis of
 8      organoiodine compounds emitted from the ocean surface to generate atomic I (Carpenter et al.,
 9      1999).  Iodine atoms react almost exclusively with O3 to form IO. Most IO photodissociates in
10      sunlight to generate I and atomic O, which rapidly recombines with O2 to form O3.
11      Consequently, this cycle has no net effect on O3 (Stutz et al., 1999). However, alternative
12      reaction pathways analogous to reactions AX2-42 through AX2-44 above lead to catalytic  O3
13      destruction. Model calculations suggest that HOI recycles via acid-catalyzed aerosol scavenging
14      to form IC1 and IBr, which subsequently volatilize and photolyze to form halogen atoms. The
15      net effect of this multiphase pathway is to increase concentrations of volatile reactive I. The self
16      reaction of IO to form I and OIO may further enhance O3 destruction (Cox et al., 1999;
17      Ashworth et al., 2002).  IO also reacts with NO2 to form INO3, which can be scavenged by
18      aqueous aerosols.  This pathway has been suggested as a potentially important sink for NOX in
19      the remote MBL and would, thus, contribute indirectly to net O3 destruction (McFiggans et al.,
20      2000).
21           Various lines of observational evidence support aspects of the above scenarios. Most
22      measurements of particulate Br in marine air reveal large depletions relative to conservative sea-
23      salt tracers (e.g., Sander et al., 2003) and, because HBr is highly soluble in acidic solution, these
24      deficits cannot be explained by simple acid-displacement reactions (e.g., Ayers et al., 1999).
25      Observed Br depletions are generally consistent with predictions based on the halogen activation
26      mechanism.  In contrast, available, albeit limited, data indicate that I is highly enriched in marine
27      aerosols relative to bulk seawater (e.g., Sturges and Barrie, 1988), which indicates active
28      multiphase iodine chemistry.
29           Direct measurements of BrO in marine air by differential  optical absorption spectroscopy
30      (DOAS) reveal mixing ratios that are near or below analytical detection limits of about 1 to 3 ppt
31      (Honninger,  1999; Pszenny  et al., 2003; Leser et al., 2003) but within the range of model

        August 2005                             AX2-54      DRAFT-DO NOT QUOTE OR CITE

-------
 1      predictions. Column-integrated DOAS observations from space reveal substantial mixing ratios
 2      of tropospheric BrO (e.g., Wagner and Platt, 1998). Although the relative amounts in the MBL
 3      cannot be resolved, these data strongly suggest active destruction of tropospheric O3 via the
 4      reaction sequence of AX2-42 through AX2-44.  Similarly, measurements of IO (McFiggans
 5      et al., 2000) and OIO (Allan et al., 2001) indicate active O3 destruction by an analogous pathway
 6      involving atomic I.  In addition, anticorrelations on diurnal time scales between total volatile
 7      inorganic Br and particulate Br and between volatile inorganic I and particulate I have been
 8      reported (e.g., Rancher and Kritz, 1980;  Pszenny et al., 2003). Although the lack of speciation
 9      precludes unambiguous interpretation, these relationships are also consistent with predictions
10      based on the halogen activation mechanism.
1 1           Large diurnal variabilities in O3 measured over the remote subtropical Atlantic and Indian
12      Oceans (Dickerson et al., 1999; Burkert  et al., 2003) and early morning depletions of O3
13      observed in the remote temperate MBL (Galbally et al., 2000) indicate that only about half of the
14      inferred O3 destruction in the MBL can be explained by conventional HOX/NOX chemistry.
15      Model calculations suggest that Br- and  I-radical chemistry could account for a "missing" O3
16      sink of this magnitude (Dickerson et  al., 1999; Stutz et al., 1999; McFiggans et al., 2000; von
17      Glasow et al., 2002b). In addition to the pathway for O3 destruction given by R AX2-39 to R
18      AX2-41, in areas with high concentrations of halogen radicals the following generic loss
19      pathways for O3 can occur in the Arctic at the onset of spring and also over  salt flats near the
20      Dead Sea (Hebestreit et al., 1999) and the Great Salt Lake (Stutz et al., 2002) analogous to their
21      occurrence in the lower stratosphere (Yung  et al., 1980).
22
                                                     X = Br,Cl,I)                      (AX2-45)
                               • Y + 03 -» YO + 02   (Y = Br, Cl, I)                      (AX2-46)
                                                     2                                 (AX2-47)
                                    Net:   203^302                                   (AX2-48)
23
24      Note that the self reaction of CIO radicals is likely to be negligible in the troposphere.  There are
25      three major reaction pathways involved in reaction AX2-47. Short-lived radical species are

        August 2005                             AX2-55       DRAFT-DO NOT QUOTE OR CITE

-------
 1      produced.  These radicals rapidly react to yield monoatomic halogen radicals. In contrast to the
 2      situation in marine air, where DO AS measurements indicate BrO concentrations of 1 to 3 ppt,
 3      Stutz et al. (2002) found peak BrO concentrations of about 6 ppt and peak CIO concentrations of
 4      about 15 ppt.  They also derived a correlation coefficient of -0.92 between BrO and O3 but much
 5      smaller values of r between CIO and O3.  Stutz et al. attributed the source of the reactive
 6      halogens to concentrated high molality solutions or crystalline salt around salt lakes, conditions
 7      that do not otherwise occur in more dilute or ocean salt water. They also suggest that halogens
 8      may be released from saline soils.  The inferred atmospheric concentrations of Cl are about
 9      105/cm3, or about a factor of 100 higher than found in the marine boundary layer by Rudolph
10      et al. (1997) indicating that, under these conditions, the Cl initiated oxidation of hydrocarbons
11      could be substantial.
12           Most of the well-established multiphase reactions tend to reduce the rate of O3 formation in
13      the polluted troposphere. Direct reactions of O3 and atmospheric particles appears to be too slow
14      to reduce smog significantly. Removal of HO2 onto hydrated particles will decrease the
15      production of O3 by the reaction of HO2 with NO.  The uptake of NO2 and HNO3 will also result
16      in the production of less O3. Conditions leading to high concentrations of Br, Cl, and I radicals
17      can lead to O3 loss. The oxidation  of hydrocarbons (especially alkanes) by Cl radicals,
18      in contrast, may lead to the rapid formation of peroxy radicals and faster smog production in
19      coastal environments where conditions are favorable for the release of gaseous Cl from the
20      marine aerosol.  There is still considerable uncertainty regarding the role of multiphase processes
21      in tropospheric photochemistry and so results should be viewed with caution and an appreciation
22      of their potential limitations.
23
24      AX2.2.10.4   Reactions on the Surfaces of Crustal Particles
25           Field studies have shown  that O3 levels are reduced in plumes containing high particle
26      concentrations (e.g., DeReus et al.; 2000; Berkowitz et al., 2001; Gaffney et al., 2002).
27      Laboratory studies of the uptake of O3 on un-treated mineral surfaces (Hanisch and Crowley,
28      2002; Michel  et al., 2002,2003) have shown that O3 is  lost by reaction on these surfaces and this
29      loss is catalytic. Values of y of 1.2 ± 0.4  x  10~4 were found for reactive uptake on a-A12O3
30      and 5 ± 1 x icr5 for reactive uptake on SiO2 surfaces.  Usher et al. (2003) found mixed behavior
31      for O3 uptake  on coated surfaces with respect to untreated surfaces.  They found that y drops

        August 2005                             AX2-56       DRAFT-DO NOT QUOTE OR CITE

-------
 1      from 1.2 ± 0.4 x icr4 to 3.4 ± 0.6 x io~5 when a-A12O3 surfaces are coated with NO3 derived
 2      from HNO3, whereas they found that y  increases to 1.6 ± 0.2 x 10~4 after these surfaces have
 3      been pre-treated with SO2.  Usher et al.  also pre-treated surfaces of SiO2 with either a C8-alkene
 4      or a C8-alkane terminated organotrichlorosilane.  They found that y increased to 7 ± 2 x 10~5 in
 5      the case of treatment with the alkene, but that it decreased to 3 ± 1 x icr5 for treatment with the
 6      alkane. Usher et al.  (2003) suggested, on the basis of these results that mineral  dust particles
 7      coated with nitrates or alkanes will affect O3 less than dust particles that have accumulated
 8      coatings of sulfite or alkenes. These studies indicate the importance of aging of airborne
 9      particles on their ability to take up atmospheric gases. Reactions such as these may also be
10      responsible for O3 depletions observed in dust clouds transiting the Pacific Ocean.
11           Underwood et  al. (2001) studied the uptake of NO2 and HNO3 on the surfaces of dry
12      mineral oxides (containing Al, Ca, Fe, Mg, Si and Ti) and naturally occurring mineral dust.
13      A wide range of values of y(NO2) were found, ranging from < 4 x 10"10 for SiO2 to 2 x 10~5 for
14      CaO, with most other values ~10~6.  Values of y for Chinese loess and Saharan  dust were also of
15      the order of 10~6. They found that as the reaction of NO2 proceeds on the surfaces that reduction
16      to NO occurs. They recommended a value of y for HNO3 of about 1 x 10~3. Not surprisingly,
17      the values of y increased from those given above if the surfaces were wetted. Underwood et al.
18      (2001) also suggested that the uptake of NO2 was likely to be  only of marginal importance but
19      that uptake of HNO3 could be of significance for photochemical oxidant cycles.
20           Li et al. (2001) examined the uptake of acetaldehyde,  acetone and propionaldehyde on the
21      same mineral oxide surfaces listed above. They found that these compounds weakly and
22      reversibly adsorb on SiO2 surfaces. However, on the other oxide surfaces, they irreversibly
23      adsorb and can form larger compounds.  They found values of y ranging from 10~6 to 10~4.
24      These reactions may reduce O3 production efficiency in areas of high mineral dust concentration
25      such as the American Southwest or in eastern Asia as noted earlier.
26
27      AX2.2.10.5   Reactions on the Surfaces of Aqueous H2SO4 Solutions
28           The most recent evaluation of Photochemical  and Chemical Data by the Jet Propulsion
29      Laboratory (Jet Propulsion Laboratory,  2003) includes recommendations for uptake coefficients
30      of various substances on a variety of surfaces including aqueous H2SO4 solutions.  Although
31      much of the data evaluated have been obtained mainly for stratospheric applications, there are

        August 2005                             AX2-57       DRAFT-DO NOT QUOTE OR CITE

-------
 1      studies in which the range of environmental parameters is compatible with those found in the
 2      troposphere.  In particular, the uptake of N2O5 on the surface of aqueous H2SO4 solutions has
 3      been examined over a wide range of values. Typical values of y are of the order of 0.1 (e.g., Jet
 4      Propulsion Laboratory, 2003). Values of y for NO2 are much lower (5 x 10"7to within a factor of
 5      three)  and thus the uptake of NO2 on the surface of aqueous H2SO4 solutions is unlikely to be of
 6      importance for oxidant cycles. The available data indicate that uptake of OH and HO2 radicals
 7      could be significant under ambient conditions with values of y of the order of 0.1 or higher for
 8      OH, and perhaps similar values for HO2.
 9
10      AX2.2.10.6  Oxidant Formation in Particles
11          Water is a major component of sub-micron particles in the atmosphere. However,
12      photochemical reactions in particles have not been studied to the same extent as they have in
13      hydrometeors (e.g., Lelieveld and Crutzen, 1991).  Friedlander and Yeh (1998) point out
14      that H2O2 and hydroxymethylhydroperoxide (HOCH2OOH) are especially likely to be found in
15      the aqueous component of atmospheric particles, based on observed gas-phase concentrations
16      and Henry's law solubility data; the concentrations in particles could be higher if the condensed
17      hydroperoxides form peroxyhydrate complexes (Wexler and Sarangapani,  1998). Laboratory
18      studies have found that UV irradiation of dissolved organic carbon (DOC) in collected
19      cloudwater samples is a source of free radicals to the aqueous phase (Faust et al., 1992, 1993)
20      but the mechanisms involved  and the atmospheric fate of these radicals are unclear. Chemical
21      reactions involving dissolved  transition metal ions could also provide significant sources of
22      radicals in particles (Jacob, 2000). However, only about 10 to 15% of the mass of organic
23      compounds in particles are quantified typically, but many of the compounds, in particular
24      aldehydes, could photolyze to produce free radicals. There are three basic mechanisms for the
25      formation of SOA (Pandis et al., 1992; Seinfeld and Pankow, 2003).  These are (1) condensation
26      of oxidized end-products of photochemical reactions (e.g., ketones, aldehydes, organic acids, and
27      hydroperoxides), (2) adsorption of semivolatile organic compounds (e.g., polycyclic aromatic
28      hydrocarbons) onto existing organic particles, and (3) dissolution of water-soluble gases that can
29      then undergo subsequent reactions in particles (e.g., aldehydes). The first and third mechanisms
30      are expected to be of major importance during the summer when photochemistry is at its peak.
31      Information about the chemistry of formation of secondary organic aerosol (SOA) was reviewed

        August 2005                             AX2-58      DRAFT-DO NOT QUOTE OR CITE

-------
 1     in Section 3.3.1 and available information about the composition of organic compounds in
 2     particles was summarized in Appendix 3C of the latest PM AQCD (U.S. Environmental
 3     Protection Agency, 2004).
 4          Recent measurements of aerosol-phase reactive oxygen species (ROS) in Rubidoux, CA
 5     and New York City have revealed relatively high concentrations, of the order of 5 to 6 x  10~7 in
 6     Rubidoux and 1 x 10~7 M nT3 in New York City, expressed as equivalent H2O2 (Venkatachari
 7     et al., 2005a,b). The ROS were found in particles of all sizes, with particularly high
 8     concentrations in the ultrafme range.  However, this finding could also be related to the
 9     condensation of vapors onto particles occurring during adiabatic expansion in the nano stages of
10     the sampler. A weak correlation was found with O3, but large ROS concentrations were still
11     found at night and in winter. The composition and  sources of the ROS are not clear. Millimolar
12     concentrations of hydroperoxides, as estimated by Friedlander and Yeh (1998), would contribute
13     only 10~12 M nT3 based on a typical liquid water volume fraction in air of 10~9. Formation of
14     peroxyhydrates would lead to higher values but would have to be very large to  account for the
15     ROS observations.  Ozone and PAN are orders of magnitude less water-soluble than the
16     hydroperoxides (Jacob, 2000) and would not contribute significantly to the ROS. Radical
17     oxidants (e.g., OH or the superoxide ion O2) do not seem to be present in sufficient abundance
18     in the atmosphere to possibly account for the ROS (Jacob, 2000). Low-volatility organic
19     peroxides produced from the oxidation of large substituted organic compounds could possibly
20     make a major contribution. Formation of these peroxides in the aerosol phase could be
21     facilitated by photochemical reactions of dissolved  organic components (Anastasio et al., 1997)
22     and by reactions of transition metals (Jacob, 2000).  Transition metals participate in the Haber-
23     Weiss set of reactions, including Fenton's reaction, generating free radicals from hydrogen
24     peroxide even in the dark.
25
26
27     AX2.3   PHYSICAL PROCESSES INFLUENCING THE ABUNDANCE
28               OF OZONE
29          The abundance and distribution of O3 in the atmosphere is determined by complex
30     interactions between meteorology and chemistry. This section will address these interactions,
31     based mainly on the results of field observations. The importance of a number of transport

       August 2005                           AX2-59      DRAFT-DO NOT QUOTE OR CITE

-------
 1      mechanisms, whose understanding has undergone significant advances since the last AQCD
 2      for O3, will be discussed in this section.
 3           Major episodes of high O3 concentrations in the eastern United States and in Europe are
 4      associated with slow moving, high pressure systems. High pressure systems during the warmer
 5      seasons are associated with the sinking of air, resulting in warm, generally cloudless conditions,
 6      with light winds. The sinking of air results in the development of stable conditions near the
 7      surface which inhibit or reduce the vertical mixing of O3 precursors. The combination of
 8      inhibited vertical mixing and light winds minimizes the dispersal of pollutants emitted in urban
 9      areas, allowing their concentrations to build up.  Photochemical activity involving these
10      precursors is enhanced because of higher temperatures  and the availability of sunlight.  In the
11      eastern United States, high O3 concentrations during a large scale episode can extend over a
12      hundred thousand square kilometers for several days. These conditions have  been described  in
13      greater detail in AQCD 96. The transport of pollutants downwind of major urban centers is
14      characterized by the development of urban plumes.  However, the presence of mountain barriers
15      can limit mixing as in Los Angeles and Mexico City and will result in even longer periods and a
16      higher frequency of days with high O3 concentrations.  Ozone concentrations  in southern urban
17      areas, such as Houston, TX and Atlanta, GA tend to follow this pattern and they tend to decrease
18      with increasing wind speed.  In northern cities, like Chicago, IL; New York, NY; and Boston,
19      MA the average O3 concentrations over the metropolitan areas increase with wind speed
20      indicating that transport of O3 and its precursors from upwind areas is important (Husar and
21      Renard, 1998; Schichtel and Husar, 2001).
22           Aircraft observations indicate that there can be substantial differences in mixing ratios of
23      key species between the surface and the atmosphere above (Fehsenfeld et al.,  1996a; Berkowitz
24      and Shaw, 1997). Convective processes and small scale turbulence transport  O3 and other
25      pollutants both upward and downward throughout the planetary boundary layer and the free
26      troposphere.  Ozone and its precursors were found to be transported vertically by convection into
27      the upper part of the mixed layer on  one day, then transported overnight as a layer of elevated
28      mixing ratios and then entrained into a growing convective boundary layer downwind and
29      brought back down to the surface. High concentrations of O3 showing large diurnal variations at
30      the surface in southern New England were associated with the presence of such layers
31      (Berkowitz et al., 1998). Because of wind shear, winds several hundred meters above the ground

        August 2005                            AX2-60      DRAFT-DO NOT QUOTE OR CITE

-------
 1      can bring pollutants from the west, even though surface winds are from the southwest during
 2      periods of high O3 in the eastern United States (Blumenthal et al., 1997). Low level nocturnal
 3      jets can also transport pollutants hundreds of kilometers. Turbulence associated with them can
 4      bring these pollutants to the surface and in many locations result in secondary O3 maxima in the
 5      early morning (Corsmeier et al., 1997).  Based on analysis of the output of model studies
 6      conducted by Kasibhatla and Chameides (2000), Hanna et al. (2001) concluded that O3 can be
 7      transported over thousands of kilometers in the upper boundary layer of the eastern half of the
 8      United States during specific O3 episodes.
 9           Stratospheric-tropospheric exchange (STE) will be discussed in Section AX2.3.1.  The
10      vertical redistribution of O3 and other pollutants by deep, or penetrating convection is discussed
11      in Section AX2.3.2.  The potential importance of transport of O3 and precursors by low-level jets
12      is the topic of Section AX2.3.3.  Issues related to the transport of O3 from North America are
13      presented in Section AX2.3.4. Relations of O3 to solar ultraviolet radiation and temperature will
14      then be discussed in Section AX2.3.5.
15
16      AX2.3.1  Stratospheric-Tropospheric Ozone Exchange (STE)
17           In the stratosphere, O3 formation is initiated by the photodissociation of molecular
18      oxygen (O2) by solar ultraviolet radiation at wavelengths less than 242 nm. Almost all of this
19      radiation is absorbed in the stratosphere (except for regions near the tropical tropopause),
20      preventing this mechanism from occurring in the troposphere. Some of the O3 in the
21      stratosphere is transported downward into the troposphere.  The potential importance of this
22      source of tropospheric O3 has been recognized since the early work of Regener (1941),  as cited
23      by  Junge (1963).  Stratospheric-tropospheric exchange (STE) of O3 and stratospheric
24      radionuclides produced by the nuclear weapons tests of the 1960s is at a maximum during late
25      winter and early spring (e.g., Ludwig et al., 1977 and references therein). Since AQCD 96 on O3
26      substantial new information from numerical models, field experiments and satellite-based
27      observations has become available. The following sections outline the basic  atmospheric
28      dynamics and thermodynamics of stratosphere/troposphere exchange and review these new
29      developments.
30
        August 2005                            AX2-61       DRAFT-DO NOT QUOTE OR CITE

-------
 1           There are several important mechanisms for injecting stratospheric O3 into the troposphere,
 2      they include tropopause folds (Reed, 1955; Danielsen, 1968), cut-off lows (Price and Vaughan,
 3      1993), clear air turbulence, mesoscale convective complexes and thunderstorms, breaking
 4      gravity waves (Poulida et al., 1996; Langford and Reid, 1998; Stohl et al., 2003) and streamers.
 5      Streamers are dry, stratospheric intrusions visible in satellite water vapor imagery that are
 6      sheared into long filamentary structures that often roll into vortices and exhibit visible evidence
 7      of the irreversible mixing of moist subtropical tropospheric and dry polar stratospheric air
 8      (Appenzeller et al., 1996; Wimmers et al., 2003). They are often present at a scale that eludes
 9      capture in large scale dynamical models of the atmosphere that cannot resolve features less than
10      1  degree (-100 km). Empirical evidence  for stratospheric intrusions comes from observations of
11      indicators of stratospheric air in the troposphere.  These indicators include high potential
12      vorticity, low water vapor mixing ratios, high potential temperature, enhancements in the ratio
13      of 7Be to 10Be in tropospheric aerosols, as well as enhancements in O3 mixing ratios and total
14      column amounts. These quantities can be observed with in situ aircraft and balloons, as well as
15      remotely sensed from aircraft and ground-based lidars and both geostationary and polar (low
16      earth orbiting) space platforms.
17           The exchange of O3 between the stratosphere and the troposphere in middle latitudes
18      occurs to a major extent by tropopause folding events (Reiter, 1963; Reiter and Mahlman, 1965;
19      Danielsen, 1968; Reiter,  1975; Danielsen and Mohnen, 1977; Danielsen, 1980).  The term,
20      tropopause folding is used to describe a process in which the tropopause intrudes deeply into the
21      troposphere along a sloping frontal zone bringing air from the lower stratosphere with it.
22      Tropopause folds occur with the formation of upper level fronts associated with transverse
23      circulations that develop around the core of the polar jet stream.  South of the jet stream core, the
24      tropopause is higher than to the north of it. The tropopause can be imagined as wrapping around
25      the jet stream core and folding beneath it  and extending into the troposphere (cf. Figure
26      AX2-7a). Although drawn as a heavy solid line, the tropopause should not be imagined as a
27      material surface, through which there is no exchange. Significant intrusions of stratospheric air
28      occur in "ribbons" -200 to 1000 km in length, 100 to 300 km wide and about 1 to 4 km thick
29      (Hoskins, 1972; Wimmers et al., 2003). These events occur throughout the year and their
30      location follows the seasonal displacement of the polar jet stream.
31

        August 2005                             AX2-62       DRAFT-DO NOT QUOTE OR CITE

-------
          o>

          (A
          (A
          0)
            150
            200
 300



 400


 500

 600

 700

 800

 900
1000
                  330
                  320

                                                                    360

                                                                    350

                                                                    340

                                                                    330
                          13

                          12

                          11

                          10

                          9

                          8 §

                          7 £
                             O)
                          6 '55
                            I
                          5

                          4

                          3

                          2
                          1
                  121° VBG
                                119°
                                          SAN 117°
                                                      UCC 115°   ELY
                                                                       113°
      Figure AX2-7a.
            Cross section through a tropopause folding event on March 13,1978 at
            0000 GMT. Potential temperatures (K) are represented by thin solid
            lines. Wind speeds (m s'1) are given by thin dashed lines.  The hatched
            area near the center of the figure indicates the location of the jet stream
            core. The tropopause defined by a potential vorticity of 100 x 10~7 K
            mb"1 s"1 is shown as the heavy solid line.  The two Sabreliner flight tracks
            through this cross section are shown as a heavy solid line with filled
            arrows and heavy dashed line with open arrows.  Longitude is shown
            along the x-axis . Upper air soundings were taken at Vandenberg AFB,
            CA (VBG); San Diego, CA (SAN); Winnemucca, NV (UCC); and Ely, NV
            (ELY).
      Source: Adapted from Shapiro (1980).
1          The seasonal cycle of O3 exchange from the stratosphere into the troposphere is not caused

2     by a peak in the seasonal cycle of upper tropospheric cyclone activity. Instead, it is related to the

3     large scale pattern of tracer transport in the stratosphere. During winter in the Northern

4     Hemisphere, there is a maximum in the poleward, downward transport of mass, which moves O3

5     from the tropical upper stratosphere to the lower stratosphere of the polar- and midlatitudes.
      August 2005
                                  AX2-63
DRAFT-DO NOT QUOTE OR CITE

-------
 1      This global scale pattern is controlled by the upward propagation of large-scale and small-scale
 2      waves generated in the troposphere. As the energy from these disturbances dissipates, it drives
 3      this stratosphere circulation. As a result of this process, there is a springtime maximum in the
 4      total column abundance of O3 over the poles.  The concentrations of O3 (and other trace
 5      substances) build up in the lower stratosphere until their downward fluxes into the lower
 6      stratosphere are matched by increased fluxes into the troposphere. Thus, there would be a
 7      springtime maximum in the flux of O3 into the troposphere even if the flux of stratospheric air
 8      through the tropopause by tropopause folding remained constant throughout the year (Holton
 9      et al., 1995). Indeed, cyclonic activity in the upper troposphere is active throughout the entire
10      year in transporting air from the lower stratosphere into the troposphere (Mahlman, 1997; and
11      references therein). Oltmans et al.  (1996) and Moody et al. (1996) provide evidence that
12      stratospheric intrusions contribute to the O3 abundance in the upper troposphere over the North
13      Atlantic even during the summer.
14           There are a number of techniques that have been used to quantify the amount of O3 in the
15      free troposphere or even the amount of O3 reaching the surface that can be attributed to
16      downward transport from the stratosphere. Earlier work, cited in AQCD 96 relied mainly on the
17      use of 7Be as a tracer of stratospheric air.  However, its use is ambiguous because it is also
18      formed in the upper troposphere. Complications also arise because its production rate is also
19      sensitive to solar activity (Lean, 2000).  The ratio of 7Be to 10Be provides a much more sensitive
20      tracer of stratospheric air than the use of 7Be alone (Jordan et al., 2003 and references therein).
21      More recent work than cited in AQCD 96 has focused on the use of potential vorticity (PV) as  a
22      tracer of stratospheric air.  Potential vorticity is a dynamical tracer used in meteorology.
23      Generally, PV is calculated from wind and temperature  observations and represents the
24      rotational tendency of a column of air weighted by the static stability, which is just the distance
25      between isentropic surfaces. This quantity is a maximum in the lower stratosphere where static
26      stability is great and along the jet stream where wind shear imparts significant rotation to air
27      parcels. As air moves from the stratosphere to the troposphere, PV is conserved, and therefore it
28      traces the motion of O3. The static stability is lower in the troposphere, so to preserve PV, fluid
29      rotation will increase. This is why STE is associated with cyclogenesis, or the formation of
30      storms along the polar jet stream.  Dynamical models clearly capture this correspondence
31      between the location of storm tracks and preferred regions for  STE.  However, because PV is

        August 2005                             AX2-64      DRAFT-DO NOT QUOTE OR CITE

-------
 1      destroyed at a faster rate with increasing depth, it is not useful as a tracer of stratospheric air
 2      reaching the surface.  Appenzeller et al. (1996) found that maps of PV coupled with satellite
 3      images of humidity can provide indications of the intrusion of stratospheric air into the
 4      troposphere, however, they had no measurements of O3. Even if measurements of O3 were
 5      available, the extrapolation of any relations to other events would still be problematic as Olsen
 6      et al. (2002) have noted that there are seasonal and geographic variations in the relation between
 7      O3 and PV. Recent flights of the NCAR C130 during the TOPSE campaign measured in situ O3,
 8      and curtains of O3 above and below the aircraft observed with a lidar and clearly showed a
 9      correspondence between high PV and stratospheric levels  of O3 and satellite depictions of dry air
10      indicating the presence of tropopause folding (Wimmers and Moody, 2004a,b).
11           Detailed cross sections through a tropopause folding event showing atmospheric structure,
12      O3 mixing ratios and condensation nuclei (CN) counts are given in Figures AX2-7a, AX2-7b,
13      and AX2-7c (Shapiro, 1980). Flight tracks of an NCAR Sabreliner obtaining data through the
14      tropopause fold are also shown.  The core of the jet stream is indicated by the hatched area near
15      the center of Figure AX2-7a. As can be seen from Figure AX2-7 a and b, there is a strong
16      relation between the folding of the tropopause, indicated by the heavy solid line and O3. CN
17      counts during the portions of the flights in the lower troposphere were tropospheric were
18      typically of the order  of several x 10 3 cm"3 and 100 or less in the stratospheric  portion.
19      However, it is clear that CN counts in the fold are much higher than in the stratosphere proper,
20      suggesting that there was active mixing between tropospheric and stratospheric air in the fold.
21      Likewise, it can also be seen form Figure AX2-7b that O3  is  being mixed outside the fold into the
22      middle and upper troposphere. The two data sets shown in Figures AX2-7b and 7c indicate that
23      small scale turbulent processes were occurring to mediate this exchange and that the folds are
24      mixing regions whose chemical characteristics lie between those of the stratosphere and the
25      troposphere (Shapiro, 1980). Chemical interactions between stratospheric and  tropospheric
26      constituents are also possible within tropopause folds. These considerations also imply that  in
27      the absence of turbulent mixing, tropopause folding can be a reversible process.
28           Several recent papers have attempted to demonstrate that the atmosphere  is a fluid
29      composed  of relatively distinct airstreams with characteristic three-dimensional motions and
30      corresponding trace gas signatures.  Based on aircraft observations, satellite imagery, and back
31      trajectories, it has been shown that dry airstreams, or dry intrusions (DA or DI) always advect

        August 2005                             AX2-65      DRAFT-DO NOT QUOTE OR CITE

-------
           150
          1000
                 121° VBG
                                119°
                                          SAN 117°
                                                       UCC 115°
                                                                ELY
                                                                        113°
       Figure AX2-7b.  Ozone mixing ratios pphm (parts per hundred million) corresponding
                       to Figure AX2-7A.  The two Sabreliner flight tracks through this cross
                       section are shown as a heavy solid line with filled arrows and a heavy
                       dashed line with open arrows. Longitude is shown along the x-axis.
                       Upper air soundings were taken at Vandenberg AFB, CA (VBG);
                       San Diego, CA (SAN); Winnemucca, NV (UCC); and Ely, NV (ELY).
       Source:  Adapted from Shapiro (1980).
 1     stratospheric O3 into the middle and upper troposphere (Cooper et al, 2001; Cooper et al.,
 2     2002a), however the seasonal cycle of O3 in the lowermost stratosphere allows greater quantities
 3     of O3 to enter the troposphere during spring (Cooper et al., 2002b). Other work has focused on
 4     the signatures of PV to show specific instances of STE (Olsen and Stanford, 2001). This
 5     correlation between TOMS gradients and PV was also used to derive the annual mass flux of O3
 6     from STE and generated an estimate somewhat higher (500 Tg/yr over the Northern
 7     Hemisphere) than the estimates of most general circulation models.  The IPCC has reported a
 8     large range of model estimates of STE, expressed as the net global flux of O3 in Tg/yr, from a
 9     low of 390 to a high of 1440  (reproduced as Table AX2-3C-1).  A few other estimates have been
10     made based on chemical observations in the lower stratosphere, or combined chemistry and
       August 2005
AX2-66
DRAFT-DO NOT QUOTE OR CITE

-------
          150
                 121° VBG
                               119°
                                         SAN 117°
                                                                         113°
      Figure AX2-7c.  Condensation nuclei concentrations (particles cm"3) corresponding to
                       Figure AX2-7a. The two flight Sabreliner flight tracks through this cross
                       section are shown as a heavy solid line with filled arrows and a heavy
                       dashed line with open arrows. Longitude is shown along the x-axis.
                       Upper air soundings were taken at Vandenberg AFB, CA (VBG);
                       San Diego, CA (SAN); Winnemucca, NV (UCC); and Ely, NV (ELY).
      Source: Adapted from Shapiro (1980).
1     dynamics (450 Tg/yr Murphy and Fahey, 1994; 510 Tg/yr extratropics only, Gettleman et al.,
2     1997; and 500 Tg/yr midlatitude NH only (30 to 60N) (Olsen et al., 2002).  These values
3     illustrate the large degree of uncertainty that remains in quantifying this important source of O3.
4          Based on the concept of tracing airstream motion, a number of Lagrangian model studies
5     have resulted in climatologies that have addressed the spatial and temporal variability in
6     stratosphere to troposphere transport (Stohl, 2001; Wernli and Borqui, 2002; Seo and Bowman,
7     2002; James et al., 2003a,b; Sprenger and Wernli, 2003; Sprenger et al., 2003). Both Stohl
8     (2001) and  Sprenger et al. (2003) produced one year climatologies of tropopause folds based on
      August 2005
AX2-67
DRAFT-DO NOT QUOTE OR CITE

-------
 1      a 1° by 1° gridded meteorological data set.  They each found the probability of deep folds
 2      (penetrating to the 800 hPa level) was a maximum during winter (December through February).
 3      The highest frequency of folding extended from Labrador down the east coast of North America.
 4      However, these deep folds occurred less than 1% of the six hour intervals for which
 5      meteorological data is assimilated for grid points in the United States. They observed a higher
 6      frequency of more shallow folds (penetrating to the upper troposphere) and medium folds
 7      (penetrating to levels between 500 and 600  hPa) of about 10% and 1 to 2% respectively. These
 8      events occur preferentially across the subtropics and the southern United States. At higher
 9      latitudes other mechanisms such as the erosion of cut-off lows and the breakup of stratospheric
10      streamers are likely to play an important role in STE. Stohl (2001) also described the region of
11      strong stirring in the upper extratropical troposphere related to the  midlatitude storm tracks.
12      Stohl (2001) demonstrated that airstreams with strong vertical motion are all highly incoherent,
13      they stir their air parcels into a new environment, producing filamentary tracer structures and
14      paving the way for subsequent mixing. A 15-year climatology by  Sprenger and Wernli (2003)
15      shows the consistent pattern of STE occurring over the primary storm tracks in the Pacific and
16      Atlantic along the Asian and North American coasts.  This climatology, and the one of James
17      et al. (2003a,b) both found that recent stratospheric air associated with deep intrusions are
18      relatively infrequent occurrences in these models. Thus, stratospheric intrusions are most likely
19      to directly affect the middle and upper troposhere and not the planetary boundary layer.
20      However, this O3 can still exchange with the planetary boundary layer through convection as
21      described later in this sub-section and in Section AX2.3.2, AX2.3.3 and AX2.3.4.
22           Interannual variations in STE are related to anomalies in large-scale circulation such as the
23      North Atlantic Oscillation which causes changes in storm track positions and intensities, and the
24      El Nino-Southern Oscillation, which results in anomalous strong convection over the eastern
25      Pacific (James et al., 2003a,b). It should also be remembered that the downward flux of O3 into
26      the troposphere is related to the depletion of O3 within the wintertime stratospheric polar
27      vortices.  The magnitude of this depletion and the transport of O3 depleted air to midlatitudes in
28      the stratosphere (Mahlman et al., 1994; Hadjinicolaou and Pyle, 2004) shows significant
29      interannual variability which may also be reflected in the downward flux of O3 into the
30      troposphere. All of these studies, from the analysis of individual events to multiyear
31      climatologies are based on the consideration of the three-dimensional motion of discrete

        August 2005                             AX2-68       DRAFT-DO NOT QUOTE OR CITE

-------
 1      airstreams in the atmosphere. However, there is a significant body of work that reports that
 2      airstreams are not entirely independent of each other (Cooper et al., 2004a,b). Midlatitude
 3      cyclones typically form in a sequential manner, some trailing in close proximity along a quasi-
 4      stationary frontal boundary, with each system influenced by remnants of other systems. For
 5      example, a rising stream of air ahead of a cold front (also known as a warm conveyor belt or
 6      WCB) on the back (western) side of a surface anticyclone may entrain air that has subsided
 7      anticyclonically into the surface high pressure system from the upper troposphere and the lower
 8      stratosphere (also known as a Dry Airstream or DA) that intruded into the mid-troposphere in a
 9      cyclone that is further downstream.  Convective mixing of the boundary layer in the WCB will
10      distribute this enhanced O3  throughout the lower troposphere and down to the surface (Davies
11      and Shuepbach, 1994; Cooper and Moody, 2000).  The net effect is that the DA of one cyclone
12      may feed into the WCB of the system immediately upwind. Similarly, the lofting of warm moist
13      air in the WCB may inject surface emissions into the upper troposphere adjacent to the western
14      side of the subsiding Dry Airstream of the storm system immediately downwind, with
15      subsequent interleaving of these two airstreams (Prados et al., 1999; Parrish et al., 2000; Cooper
16      et al., 2004a,b) as illustrated schematically in Figure AX2-8. The ultimate mixing of these
17      airstreams, which inevitably occurs at a scale that is not resolved by current models confounds
18      our ability to attribute trace gases to their sources.
19           These studies suggest that both downward transport from the stratosphere  and upward
20      transport from the atmospheric boundary layer act in concert with their relative roles determined
21      by the balance between the  amount of O3 in the lower stratosphere and the availability of free
22      radicals to initiate the photochemical processes forming O3 in the boundary layer. Dickerson
23      et al. (1995) pointed out that springtime maxima in O3 observed in Bermuda correlate well with
24      maxima in carbon monoxide. Carbon monoxide, O3 and its photochemical precursors may have
25      been transported into the upper troposphere from the polluted continental boundary layer by
26      deep convection.  The photochemical processes involve the buildup of precursors during the
27      winter at Northern mid- and high latitudes. Parrish et al. (1999) have noted that reactions
28      occurring during the colder months may tend to titrate O3.  However, as NOX and its reservoirs
29      are transported sourthward they can initiate O3 formation through reactions described in
30      Section AX2.2 (see also Stroud et al., 2003).
31

        August 2005                            AX2-69      DRAFT-DO NOT QUOTE OR CITE

-------
                 Altitude
                             Stratospheric Air
                               »--
                                Cold Front (A)
               Stratospheric Air

                  Cold Front (B)
                   West
          -1500
Distance (km)
                 East
       Figure AX2-8.  Schematic diagram of a meteorological mechanism involved in high
                       concentrations of O3 found in spring in the lower troposphere off the
                       American east coast.  Subsidence behind the first cold front meets
                       convection ahead of a second cold front such that polluted air and O3 from
                       the upper troposphere/lower stratosphere are transported in close
                       proximity (or mixed) and advected over the north Atlantic Ocean.  The
                       vertical scale is about 10 km; the horizontal scale about 1500 km. (Note
                       that not all cold fronts are associated with squall lines and that mixing
                       occurs even in their absence.)

       Source: Prados (2000).
 1     AX2.3.2  Deep Convection in the Troposphere

 2          Much of the upward motion in the troposphere is driven by convergence in the boundary
 3     layer and deep convection. Deep convection, as in developing thunderstorms can transport

 4     pollutants rapidly to the middle and upper troposphere (Dickerson et al., 1987).  The outflow

 5     from these systems results in the formation of layers with distinctive chemical properties in the

 6     middle troposphere. In addition, layers are formed as the result of stratospheric intrusions.

 7     Layers ranging in thickness typically from 0.3 to about 2 km in the middle troposphere (mean

 8     altitudes between 5 and 7 km) are ubiquitous and occupy up to 20% of the troposphere to 12 km

 9     (Newell et al., 1999).  The origin of these layers can be judged by analysis of their chemical

10     composition (typically by comparing ratios of H2O, O3 and CO to each other) or dynamical

11     properties (such as potential vorticity). Thus, pollutants that have been transported into the
       August 2005
  AX2-70
DRAFT-DO NOT QUOTE OR CITE

-------
 1      middle and upper troposphere at one location can then be transported back down into the
 2      boundary layer somewhere else.
 3           Crutzen and Gidel (1983), Gidel (1983), and Chatfield and Crutzen (1984) hypothesized
 4      that convective clouds played an important role in rapid atmospheric vertical transport of trace
 5      species and first tested simple parameterizations of convective transport in atmospheric chemical
 6      models.  At nearly the same time, evidence was shown of venting of the boundary layer by
 7      shallow fair weather cumulus clouds (e.g., Greenhut et al., 1984; Greenhut, 1986).  Field
 8      experiments were conducted in 1985, which resulted in verification of the hypothesis that deep
 9      convective clouds are instrumental in atmospheric transport of trace constituents (Dickerson
10      et al., 1987; Luke et al.,  1997). Once pollutants are  lofted to the middle and upper troposphere,
11      they typically have a much longer chemical lifetime and with the generally stronger winds at
12      these altitudes they can be transported large distances from their source regions.  Photochemical
13      reactions occur during this long-range transport. Pickering et al. (1990) demonstrated that
14      venting of boundary layer pollutants by convective clouds (both shallow and deep) causes
15      enhanced O3  production in the free troposphere.  Therefore, convection aids in the
16      transformation of local pollution into a contribution to global atmospheric pollution.  Downdrafts
17      within thunderstorms tend to bring air with less pollution  from the middle troposphere into the
18      boundary layer.
19           Field studies have  established that downward transport of larger O3 and NOX mixing ratios
20      from the free troposphere to the boundary layer is an important process over the remote oceans
21      (e.g., Piotrowicz et al., 1991), as well as the upward transport  of very low O3 mixing ratios from
22      the boundary layer to the upper troposphere (Kley et al., 1996).  Global modeling by Lelieveld
23      and Crutzen (1994) suggests that the downward mixing of O3 into the boundary layer (where it is
24      destroyed) is the dominant global effect of deep convection.  Some indications of downward
25      transport of O3 from higher altitudes (possibly from the stratosphere) in the anvils of
26      thunderstorms have been observed (Dickerson et al., 1987; Poulida et al., 1996; Suhre et al.,
27      1997). Ozone is most effective as a greenhouse gas in the vicinity of the tropopause.  Therefore,
28      changes  in the vertical profile of O3 in the upper troposphere caused by deep convection have
29      important radiative forcing implications for climate.
30           Other effects of deep convection include perturbations to photolysis rates, which include
31      enhancement of these rates in the upper portion of the thunderstorm anvil. In addition,

        August 2005                             AX2-71      DRAFT-DO NOT  QUOTE OR CITE

-------
 1     thunderstorms are effective in the production of NO by lightning and in wet scavenging of
 2     soluble species.
 3
 4     AX2.3.2.1  Observations of the Effects of Convective Transport
 5           Some fraction of shallow fair weather cumulus clouds actively vent boundary layer
 6     pollutants to the free troposphere (Stull, 1985). The first airborne observations of this
 7     phenomenon were conducted by Greenhut et al. (1984) over a heavily urbanized area, measuring
 8     the in-cloud flux of O3 in a relatively large cumulus cloud. An extension of this work was
 9     reported by Greenhut (1986) in which data from over 100 aircraft penetrations of isolated
10     nonprecipitating cumulus clouds over rural and suburban areas were obtained. Ching and
11     Alkezweeny (1986) reported tracer (SF6) studies associated with nonprecipitating cumulus (fair
12     weather cumulus and cumulus congestus). Their experiments showed that the active  cumulus
13     clouds transported mixed layer air upward into the overlying free troposphere and suggested that
14     active cumuli can also induce rapid downward transport from the free troposphere into the mixed
15     layer.  A UV-DIAL (Ultraviolet Differential Absorption Lidar) provided space-height cross
16     sections of aerosols and O3 over North Carolina in a study of cumulus venting reported by Ching
17     et al. (1988).  Data collected on evening flights showed regions of cloud debris containing
18     aerosol and O3 in the lower free troposphere in excess of background, suggesting that significant
19     vertical exchange had  taken place during afternoon cumulus cloud activity. Efforts have also
20     been made to estimate the vertical transport by ensembles of nonprecipitating cumulus clouds in
21     regional chemical transport models (e.g., Vukovich and Ching, 1990).
22           The first unequivocal observations of deep convective transport of boundary layer
23     pollutants to the upper troposphere were documented by Dickerson et al. (1987).
24     Instrumentation aboard three research aircraft measured CO, O3, NO, NOX, NOy, and
25     hydrocarbons in the vicinity of an active mesoscale convective system near the
26     Oklahoma/Arkansas border during the 1985 PRE-STORM experiment. Anvil penetrations about
27     two hours after maturity found greatly enhanced mixing ratios of all of the aforementioned
28     species compared with outside of the cloud. Among the species measured, CO is the  best tracer
29     of upward convective transport because it is produced primarily in the boundary layer and has an
30     atmospheric lifetime much longer than the timescale of a thunderstorm. In the observed storm,
31     CO measurements exceeded 160 ppbv as high as 11 km, compared with -70 ppbv outside of the

       August 2005                            AX2-72      DRAFT-DO NOT QUOTE OR CITE

-------
 1      cloud (Figure AX2-9a). Cleaner middle tropospheric air appears to have descended in
 2      downdrafts forming a pool of lower mixing ratio CO beneath the cloud.  Nonmethane
 3      hydrocarbons (NMHC) with moderate lifetimes can also serve as tracers of convective transport
 4      from the boundary layer. Ozone can also be an indicator of convective transport. In the polluted
 5      troposphere large O3 values will indicate upward transport from the boundary layer, but in the
 6      clean atmosphere such  values are indicative of downward transport from the uppermost
 7      troposphere or lowermost stratosphere. In this case measured O3 in the upper rear portion of the
 8      anvil peaked at 98 ppbv, while boundary layer values were only -65 ppbv (Figure AX2-9b). It is
 9      likely that some higher-O3 stratospheric air mixed into the anvil.
10          The large amount of vertical trace gas transport noted by Dickerson et al. (1987) cannot,
11      however, be extrapolated to all convective cells.  Pickering et al. (1988)  reported airborne
12      measurements of trace  gases taken in the vicinity of a line of towering cumulus and
13      cumulonimbus clouds that also occurred during PRE-STORM.  In this case trace gas mixing
14      ratios in the tops of these clouds were near ambient levels. Meteorological analyses showed that
15      these clouds were located above a cold front, which prevented entry of air from the boundary
16      layer directly below or  near the clouds. Instead, the air entering these clouds likely originated in
17      the layer immediately above the boundary layer which was quite clean. Luke et al. (1992)
18      summarized the air chemistry data from all  18 flights during PRE-STORM by categorizing each
19      case according to synoptic flow patterns.  Storms in the maritime tropical flow regime
20      transported large amounts of CO, O3, and NOy into the upper troposphere with the
21      midtroposphere remaining relatively clean.  During frontal passages a combination of stratiform
22      and convective clouds mixed pollutants more uniformly into the middle and upper levels; high
23      mixing ratios of CO were found at all altitudes.
24          Prather and Jacob (1997) and Jaegle et al. (1997) noted that in addition to the primary
25      pollutants (e.g., NOX, CO, VOCs), precursors of HOX are also transported to the upper
26      troposphere by deep convection.  Precursors of most importance are water vapor, formaldehyde,
27      hydrogen peroxide, methylhydroperoxide, and acetone. HOX is critical for oxidizing NO to NO2
28      in the O3 production process.
29          Over remote marine areas the effects of deep convection on trace gas distributions differ
30      from that over moderately polluted continental regions. Chemical measurements taken by the
31      NASA ER-2 aircraft during the Stratosphere-Troposphere Exchange Project (STEP) off the

        August 2005                            AX2-73      DRAFT-DO NOT QUOTE OR CITE

-------
                   a. Carbon Monoxide
            Surface
                 98°Wf    97'
                  Oklahoma City
96°      95°      94°
     Longitude
                   b. Ozone
            Surface
                          97C
                  Oklahoma City
96°      95°      94°
     Longitude
         93°
92°
Figure AX2-9a,b.  (a) Contour plot of CO mixing ratios (ppbv) observed in and near the
                  June 15,1985, mesoscale convective complex in eastern Oklahoma.
                  Heavy line shows the outline of the cumulonimbus cloud. Dark shading
                  indicates high CO and light shading indicates low CO. Dashed contour
                  lines are plotted according to climatology since no direct measurements
                  were made in that area, (b) Same as (a) but for O3 (ppbv).

Source:  Dickerson et al. (1987).
August 2005
   AX2-74
DRAFT-DO NOT QUOTE OR CITE

-------
 1      northern coast of Australia show the influence of very deep convective events. Between
 2      14.5 and 16.5 km on the February 2 to 3, 1987 flight, perturbations in the chemical profiles were
 3      noted that included pronounced maxima in CO, water vapor, and CCN and minima of NOy,
 4      and O3 (Pickering et al., 1993). Trajectory analysis showed that these air parcels likely were
 5      transported from convective cells 800 to 900 km upstream.  Very low boundary layer mixing
 6      ratios of NOy and O3 in this remote region were apparently transported upward in the convection.
 7      A similar result was noted in CEPEX (Central Equatorial Pacific Experiment; Kley et al., 1996)
 8      where a series of ozonesonde ascents showed very low upper tropospheric O3 following deep
 9      convection.  It is likely that similar transport of low-O3 tropical marine boundary layer air to the
10      upper troposphere occurs in thunderstorms along the east coast of Florida. Convection over the
11      Pacific will likely transport halogens to the upper troposphere where they may aid in the
12      destruction of O3. This low-O3 convective outflow will likely descend in the subsidence region
13      of the eastern Pacific, leading to some of the cleanest air that arrives at the west coast of the
14      United States.
15
16      AX2.3.2.2  Modeling the Effects of Convection
17          The effects of deep convection may be simulated using cloud-resolving models, or in
18      regional or global models in which the convection is parameterized.  The Goddard Cumulus
19      Ensemble (GCE) model (Tao and Simpson,  1993) has been used by Pickering et al. (1991;
20      1992a,b; 1993; 1996), Scala et al. (1990) and Stenchikov et al. (1996) in the analysis of
21      convective transport of trace gases.  The cloud model is nonhydrostatic and contains detailed
22      representation of cloud microphysical processes.  Two  and three dimensional versions of the
23      model have been applied in transport analyses.  The initial conditions for the model are usually
24      from a sounding of temperature, water vapor and winds representative of the region of storm
25      development.  Model-generated wind fields can be used to perform air parcel trajectory analyses
26      and tracer advection calculations. Once transport calculations are performed for O3 precursors, a
27      1-D photochemical model was employed to estimate O3 production rates in the outflow air from
28      the convection. These rates were then compared with those prior to convection to determine an
29      enhancement factor due to convection.
30          Such methods were used by Pickering et al. (1992b) to examine transport of urban plumes
31      by deep convection. Transport of the Oklahoma City plume by the 10 - 11 June 1985

        August 2005                            AX2-75      DRAFT-DO NOT QUOTE OR CITE

-------
 1     PRE-STORM squall line was simulated with the 2-D GCE model. In this event forward
 2     trajectories from the boundary layer at the leading edge of the storm showed that almost 75% of
 3     the low level inflow was transported to altitudes exceeding 8 km.  Over 35% of the air parcels
 4     reached altitudes over 12 km.  Tracer transport calculations were performed for CO, NOX, O3,
 5     and hydrocarbons.  The 3-D version of the GCE model has also been run for the 10-11 June
 6     1985 PRE-STORM case.  Free tropospheric O3 production enhancement of a factor of 2.5 for
 7     Oklahoma rural air and -4 for the Oklahoma City case were calculated.
 8          Stenchikov et al. (1996) used the 2-D GCE model to simulate the North Dakota storm
 9     observed by Poulida et al. (1996). This storm showed the unusual feature of an anvil formed
10     well within the stratosphere. The increase of CO and water vapor above the altitude of the
11     preconvective tropopause was computed in the model.  The total mass of CO across the model
12     domain above this level increased by almost a factor of two during the convective event. VOCs
13     injected into the lower stratosphere could enhance O3 production there. Downward transport of
14     O3 from the stratosphere was noted in the simulation in the rear anvil.
15          Regional estimates of deep convective transport have been made through use of a traveling
16     1-D model, regional transport models driven by parameterized convective mass fluxes from
17     mesoscale meteorological models, and a statistical-dynamical approach. Pickering et al. (1992c)
18     developed a technique which uses a combination of deep convective cloud cover statistics from
19     the International Satellite Cloud Climatology Project (ISCCP) and convective transport statistics
20     from GCE model simulations of prototype storms to estimate the amount of CO vented from the
21     planetary boundary layer (PEL) by deep convection. This statistical-dynamical approach was
22     used by Thompson et al. (1994) to estimate the convective transport component of the boundary
23     layer CO budget for the central United States (32.5° - 50° N, 90° - 105° W) for the month of
24     June.  They found that the net upward deep convective flux (-18 x 105 kg-CO/month) and the
25     shallow convective flux (-16 x 10s kg-CO/month) to the free troposphere accounted for about
26     80% of the loss of CO from the PEL. These losses roughly balanced horizontal transport of CO
27     (-28 x 10s kg-CO/month), the oxidation of hydrocarbons (-8 x 10s kg-CO/month) and
28     anthropogenic and  biogenic emissions (-8 + -1  x 10s kg-CO/month) into the PEL in the central
29     United States. In this respect the central United States acts as a "chimney" for venting CO and
30     other pollutants.
       August 2005                            AX2-76       DRAFT-DO NOT QUOTE OR CITE

-------
 1          Regional chemical transport models have been used for applications such as simulations of
 2     photochemical O3 production, acid deposition, and fine particulate matter. Walcek et al. (1990)
 3     included a parameterization of cloud-scale aqueous chemistry, scavenging, and vertical mixing
 4     in the chemistry model of Chang et al. (1987).  The vertical distribution of cloud microphysical
 5     properties and the amount of subcloud-layer air lifted to each cloud layer are determined using a
 6     simple entrainment hypothesis (Walcek and Taylor, 1986).  Vertically-integrated O3 formation
 7     rates over the northeast U.S. were enhanced by -50% when the in-cloud vertical motions were
 8     included in the model.
 9          Wang et al. (1996) simulated the 10-11 June 1985 PRE-STORM squall line with the
10     NCAR/Penn State Mesoscale Model (MM5; Grell et al., 1994; Dudhia et al., 1993).  Convection
11     was parameterized as a subgrid-scale process in MM5 using the Kain and Fritsch (1993) scheme.
12     Mass fluxes and detrainment profiles from the convective parameterization were used along with
13     the 3-D wind fields in CO tracer transport calculations for this convective event.  The U.S.
14     Environmental Protection Agency has developed a Community Multiscale Air Quality (CMAQ)
15     modeling system that uses MM5 with the Kain-Fritsch convective scheme as the dynamical
16     driver (Ching et al.,  1998).
17          Convective transport in global chemistry and transport models is treated as a subgrid-scale
18     process that is parameterized typically using cloud mass flux information from a general
19     circulation model or global data assimilation system. While GCMs can provide data only for a
20     "typical" year, data assimilation systems can provide "real" day-by-day meteorological
21     conditions, such that CTM output can be compared directly with observations of trace gases.
22     The NASA Goddard Earth Observing System Data Assimilation System (GEOS-1 DAS and
23     successor systems; Schubert et al., 1993; Bloom et al., 1996) provides archived global data sets
24     for the period 1980 to present, at 2° x 2.5°  or better resolution with 20 layers or more in the
25     vertical. Convection is parameterized with the Relaxed Arakawa-Schubert scheme (Moorthi and
26     Suarez, 1992).  Pickering et al. (1995) showed that the cloud mass fluxes from GEOS-1 DAS are
27     reasonable for the 10-11 June 1985 PRE-STORM squall line based on comparisons with the
28     GCE model (cloud-resolving model) simulations of the same storm.  In addition, the GEOS-1
29     DAS cloud mass fluxes compared favorably with the regional estimates of convective transport
30     for the central U.S. presented by Thompson et al. (1994). However, Allen et al. (1997) have
       August 2005                            AX2-77      DRAFT-DO NOT QUOTE OR CITE

-------
 1      shown that the GEOS-1 DAS overestimates the amount and frequency of convection in the
 2      tropics and underestimates the convective activity over midlatitude marine storm tracks.
 3
 4      AX2.3.3  Nocturnal  Low-Level Jets
 5           Nocturnal low-level jets (LLJ) are coincident with synoptic weather patterns involved with
 6      high O3 episodes implying that they may play an important role in the formation of severe O3
 7      events (Rao and Zurbenko, 1994).  LLJ can transport pollutants hundreds of kilometers from
 8      their sources.  Figure AX2-10 shows the evolution of the planetary boundary layer (PEL) over
 9      land during periods when high-pressure weather patterns prevail (Stull, 1999). During synoptic
10      weather patterns with stronger zonal flow, a schematic of the boundary layer could look quite
11      different with generally more uniform mixing present. As  can be seen from Figure AX2-10, the
12      PEL can be divided into three sublayers: a turbulent mixed layer (typically present during
13      daylight hours), a less turbulent residual layer which occupies space that was formerly the mixed
14      layer, and a nocturnal, stable boundary layer that has periods of sporadic turbulence (Stull,
15      1999).  The LLJ forms in the residual layer. It is important to note, that during the nighttime, the
16      PEL often comprises thin, stratified layers with different physical and chemical properties (Stull,
17      1988).
18           At night, during calm conditions, the planetary boundary  layer is stably stratified and as a
19      result verticle mixing is inhibited. On cloud-free evenings the LLJ begins to form shortly after
20      sunset.  The wedge of cool air in the stable nocturnal boundary layer decouples the surface layer
21      from the residual layer and acts like a smooth surface allowing the air just above it (in the
22      residual layer) to flow rapidly past the inversion mostly unencumbered by surface friction (Stull,
23      1999).  As the sun rises, its energy returns to heat the land  and the lower atmosphere begins to
24      mix as the warm air rises. The jet diminishes as the nocturnal temperature inversion erodes and
25      surface friction slows winds speeds. If stable synoptic conditions persist, the same conditions
26      the next night could allow the low-level jet to reform with  equal strength and similar
27      consequences. LLJ formation results in vertical wind shear that induces mixing between the
28      otherwise stratified layers.
29           LLJs are often associated with mountain ranges. Mountains and pressure gradients on
30      either side of a developing LLJ help concentrate the flow of air into a corridor or horizontal
31      stream (Hobbs et al., 1996). Figure AX2-11 shows that LLJs commonly form east of the Rocky

        August 2005                            AX2-78      DRAFT-DO NOT QUOTE OR CITE

-------
                   2-
                o>
                •53 1
                   0-
           Free Atmosphere

     Cloud Layer
                                                             Cloud Layer
                                      Stable Nocturnal Boundary Layer
        Surface Layer
                    Afternoon
                Sunset
Midnight
Sunrise
                                                                 Mixed
                                                                 Layer
  I
Noon
Figure AX2-10.
The diurnal evolution of the planetary boundary layer while high
pressure prevails over land. Three major layers exist (not including the
surface layer):  a turbulent mixed layer; a less turbulent residual layer
which contains air formerly in the mixed layer; and a nocturnal, stable
boundary layer which is characterized by periods of sporadic turbulence.
 Source: Adapted from Stall (1999) Figures 1.7 and 1.12.
Figure AX2-11.   Locations of low-level jet occurrences in decreasing order of prevalence
                 (most frequent, common, observed).  These locations are based on 2-years
                 radiosonde data obtained over limited areas. With better data coverage,
                 other low-level jets may well be observed elsewhere in the United States.

Source: Bonner (1968).
August 2005
                       AX2-79
          DRAFT-DO NOT QUOTE OR CITE

-------
 1     Mountains and east of the Appalachian Mountains (Bonner, 1968). There may be other locations
 2     in the U.S. where LLJs  occur. The width of the jet can vary from location to location and from
 3     one weather pattern to another, but is typically less than several hundred km not greater than
 4     1000 km long. In extreme cases, winds in a LLJ can exceed 60 ms"1 but average speeds are
 5     typically in the range of 10 to 20 ms"1.
 6          Nocturnal low-level jets are not unique to the United States; they have been detected in
 7     many other parts of the world (Corsmeier, 1997, Reitebuch, et al., 2000). Corsmeier et al.
 8     (1997) observed secondary maxima in surface O3 at nighttime at a rural site in Germany,
 9     supporting the notion that downward transport from the residual layer was occurring.  The
10     secondary maxima were, on average, 10% of the next day's O3 maximum but at times could be
11     as much as 80% of the maximum (Corsmeier et al., 1997). The secondary O3 maxima were well
12     correlated with an increase in wind speed and wind shear. The increased vertical shear over the
13     very thin layer results in mechanical mixing that leads a downward flux of O3 from the residual
14     to the near surface layer (see Low-level jets AX2-12 and AX2-13). Analysis of wind profiles
15     from aerological stations in northeastern Germany revealed the spatial  extent of that particular
16     LLJ was up to 600 km in length  and 200 km in width. The study concluded the importance of O3
17     transport by low-level jets was twofold: O3 and other pollutants could  be transported hundreds
18     of kilometers at the j et core level during the night and then mixed to the ground far from their
19     source region. Salmond and McKendry (2002) also observed secondary O3 maxima (in the
20     Lower Fraser Valley, British Columbia) associated with low-level jets  that occasionally
21     exceeded half the previous day's maximum O3 concentration. The largest increases in surface
22     O3 concentration occurred when boundary layer turbulence coincided with O3 levels greater than
23     80 parts per billion were observed aloft. In addition, the study suggests horizontal transport
24     efficiency during a low-level jet event could be as much as six times greater than transport with
25     light winds without an LLJ. Reitebuch et al. (2000) observed secondary O3 maxima associated
26     with low-level jet evolution in an urban area in Germany.  The notion that O3 was transported
27     downward from the residual layer to the surface was supported by observed decreases in
28     concentrations of NO, NO2 and CO in the residual layer during secondary O3 maxima. Unlike
29     O3 in the residual layer, concentrations of NO, NO2, and CO should be lower than those found
        August 2005                            AX2-80      DRAFT-DO NOT QUOTE OR CITE

-------
             o
            "•5
             (0
2  1.0-
                    Surface Ozone During Low Level Jet Periods
o
O


O
•o

N


I

o
                0.5-
                0.0-
                                                  Secondary

                                                  63 Maxima
                    o
                    o

                    CD
                    O
             O

             O
o
o
o
o
o
o
66
o
o
                                               CM
o
o
o
o
o
o

CO
o
o
o

CD
o
Figure AX2-12.
                    Time of Day (Local)



    Schematic diagram showing the diurnal behavior of O3 and the

    development of secondary O3 maxima resulting from downward

    transport from the residual layer when a low-level jet is present.
Source: Adapted from Reitbuch et al. (2000); Corsmeier et al. (1997); and Salmond and McKendry (2002).
                            Boundary Layer Winds During

                              a Nocturnal Low Level Jet
                   1.0-
                f 0.5-
                     0-
                              M
                              £ O
                              m N
                   10   15   20
                        Wind Speed (ms-1)    Wind Direction




Figure AX2-13.  The nocturnal low-level jet occupies a thin slice of the atmosphere near

                the Earth's surface. Abrupt changes in wind speed and wind direction

                with height associated with the low-level jet create conditions favorable

                for downward transport of air to the surface layer.



Source: Singh et al. (1997); Corsmeier et al. (1997).
August 2005
                         AX2-81
                   DRAFT-DO NOT QUOTE OR CITE

-------
 1      nearer the surface (Reitebuch et al., 2000; Seinfeld and Pandis, 1998). As in other studies, wind
 2      speed and directional shear were detected during these events. Calculations of the average wind
 3      speed and duration of the LLJ suggested that pollutants were transported several hundred
 4      kilometers.  A study of the PEL and the vertical structure of O3 observed at a costal site in Nova
 5      Scotia described how temperature and differences of surface roughness in a marine environment
 6      can induce LLJ formation and pollution transport (Gong  et al., 2000).  In this case, rather strong
 7      horizontal sea surface temperature gradients provided the necessary baroclinic forcing.
 8           While the studies mentioned above have shed light on the possible role of the LLJ in the
 9      transport of O3 and its precursors, quantitative statements about the significance of the LLJ in
10      affecting local and regional O3 budgets can not yet be made. This inability reflects the lack of
11      available data for wind profiles in the planetary boundary layer in areas where LLJ are likely to
12      occur and because of the inadequacy of numerical models in simulating their occurrence.
13
14      AX2.3.4   Intercontinental Transport of Ozone and Other Pollutants
15      AX2.3.4.1  The Atmosphere/Ocean Chemistry Experiment, AEROCE
16           The AEROCE experiment, initiated in the early 1990s set out to examine systematically
17      the chemistry and meteorology leading to the trace gas and aerosol composition over the North
18      Atlantic Ocean.  One particular focus area was to determine the relative contribution of
19      anthropogenic and natural processes to the O3 budget and oxidizing capacity of the troposphere
20      over the North Atlantic Ocean. Early results using isentropic back trajectories suggested that
21      periodic pulses of O3 mixing ratios up to  80 ppb were associated with large-scale subsidence
22      from the mid-troposphere, favoring a natural source (Oltmans and Levy, 1992). Moody et al.
23      (1995) extended this work with a five-year seasonal climatology and found the highest
24      concentrations of O3 were always associated with synoptic scale post-frontal subsidence off the
25      North American continent behind cold fronts, and this pattern was most pronounced in the
26      April-May time frame.  These post-frontal air masses had uniformly low humidity and high
27      concentrations of 7Be, a  cosmogenic tracer produced in the upper troposphere and lower
28      stratosphere. However,  the pulsed occurrence of these postfrontal air masses also frequently
29      delivered enhanced concentrations of species such as SO4=, NO3 , 210Pb, etc. suggesting a
30      component originating in North America. In a subsequent analysis of data from one year (1992)
31      when CO observations were available, Dickerson et al. (1995) concluded that anthropogenic

        August 2005                            AX2-82      DRAFT-DO NOT QUOTE OR CITE

-------
 1      sources made a significant contribution to surface O3, and using a simple mixing model they
 2      determined that 57% of the air had a continental boundary layer origin.
 3           Based on these observations of the synoptically modulated concentrations, AEROCE
 4      conducted an aircraft and ozonesonde intensive in the spring of 1996. The intention was to
 5      adopt a meteorologically informed sampling strategy to clearly distinguish the characteristics of
 6      air masses ahead of and behind eastward progressing cold fronts.  Sixteen research flights were
 7      conducted with the University of Wyoming King Air research aircraft.  The goal was to
 8      differentiate the sources of enhanced O3 mixing ratios observed on Bermuda after the passage of
 9      cold fronts, and to identify the major processes controlling the highly variable O3 mixing ratios
10      in the mid-to-upper troposphere over eastern North America and the North Atlantic Ocean
11      during April and May. In addition to aircraft flights, near-daily ozonesondes were launched in a
12      quasi-zonal transect from Purdue, Indiana, to Charlottesville, Virginia to Bermuda. An effort
13      was made to time the release of ozonesondes to cleanly differentiate pre and post-frontal air
14      masses.
15           In several aircraft flights, the presence at altitude of distinct layers of air with elevated
16      concentrations of nonmethane hydrocarbons (NMHCs) attested to the dynamic vertical mixing
17      associated with springtime frontal activity.  Layers  of mid-tropospheric air of high O3 (140 ppb)
18      and low background NMHC mixing ratios (1.44 ppbv ethane, 0.034 ppbv propene, 0.247 ppbv
19      propane, and 0.034 ppbv isobutene, 0.041 ppbv n-butane, 0.063 ppbv benzene, 0.038 ppbv
20      toluene) were indicative of descending, stratospherically influenced air on a flight to the east of
21      Norfolk, VA on April 24 (alt 4600m). However layers of elevated NMHC concentrations
22      (1.88 ppbv ethane,  0.092 ppbv propene, 0.398 ppbv propane, 0.063 ppbv isobutene, 0.075 ppbv
23      n-butane, 0.106 ppbv benzene, 0.0102 ppbv toluene) occurred along with 60-70 ppbv of O3 on a
24      flight west of Bermuda April 28 (alt. 4100m), indicating air had been lofted from the continental
25      boundary layer. Meteorological evidence, supported by ozonesonde observations and  earlier
26      King Air  flights,  indicated that stratosphere/troposphere exchange associated with an upstream
27      frontal system had injected and advected dry, O3-rich air into the mid-troposhere region over the
28      continent. This subsiding air mass provided deep layers of enhanced O3 in the offshore,
29      postfrontal area.  Convection from a developing (upwind) system lifted continental boundary
30      layer air into the  proximity of the dry, subsiding air layer (Prados, et al., 1999).  This resulted in
31      a mixture of high concentrations of anthropogenic pollutants along with naturally enhanced O3.

        August 2005                             AX2-83      DRAFT-DO NOT QUOTE OR CITE

-------
 1      Ozone mixing ratios exceeded those attributable to boundary layer venting or in-transit
 2      photochemical production.  These meteorological processes led to pollution and stratospherically
 3      enhanced O3 co-occurring in post-frontal air masses over the North Atlantic Ocean.  A similar
 4      event in February 1999 was observed by Parrish et al. (2000). It confirmed the occurrence of
 5      thin layers of anthropogenic and stratospheric air that subsequently mix.  These results, along
 6      with recent modeling studies suggest that North American pollution clearly does contribute to
 7      the periodic influx of less-than-pristine air observed in the marine boundary layer over Bermuda
 8      (e.g., Li et al., 2002) and yet these incursions are not inconsistent with observing enhancements
 9      in O3 due to stratospheric exchange.
10          The ozonesonde climatology of AEROCE clearly established that O3 mixing ratios were
11      always enhanced and increased with height in post-frontal air masses. Postfrontal O3 in the
12      lower troposphere over Bermuda originates in the postfrontal midtroposphere over the continent,
13      supporting the hypothesis that naturally occurring stratospheric O3 makes a contribution to air in
14      the marine boundary layer (Cooper et al., 1998). A schematic of the meteorological processes
15      responsible for the close proximity of natural and man-made O3 can be seen in Figure AX2-8
16      from Prados (2000). Cold fronts over North America tend to be linked in wave-like patterns
17      such that the subsidence behind one front may occur above with intrusions of convection ahead
18      of the next cold front. Pollutants, including VOC and NOx, precursors to O3, may be lofted into
19      the mid-to-upper troposphere where they have the potential to mix with layers of air descending
20      from O3-rich but relatively unpolluted upper troposphere and lower stratosphere. Through this
21      complex mechanism, both stratospheric and photochemically produced O3 may be transported to
22      the remote marine environment where they have large-scale impacts on the radiative and
23      chemical properties of the atmosphere. Recent three-dimensional modeling studies of air mass
24      motion over the Pacific  provide further evidence that these complex mechanisms are indeed
25      active  (Cooper et al., 2004b).
26
27      AX2.3.4.2  The North Atlantic Regional Experiment, NARE
28          NARE was established by the International Global Atmospheric Chemistry Project to study
29      the chemical processes occurring in the marine troposphere of the North Atlantic, the marine
30      region expected to be the most impacted by industrial emissions from eastern North America and
31      western Europe. Surface measurements from several surface sites were initiated in 1991, with

        August 2005                            AX2-84      DRAFT-DO NOT QUOTE OR CITE

-------
 1      major field intensives in summer 1993, spring 1996, early autumn 1997 and a few winter flights
 2      in 1999. In the summer of 1993, airborne and ground-based measurements of O3 and O3
 3      precursors were made in the North Atlantic region by an international team of scientists to
 4      determine how the continents that rim the North Atlantic are affecting atmospheric composition
 5      on a hemispheric scale  (Fehsenfeld et al., 1996a; Fehsenfeld et al., 1996b). The focus of NARE
 6      was to investigate the O3 budget of the North Atlantic region.  Previous observations indicated
 7      that the O3 produced from anthropogenic sources is greater than that reaching the lower
 8      troposphere from the stratosphere and that O3 derived from anthropogenic pollution has a
 9      hemisphere wide effect at northern mid latitudes. This study was performed to better quantify
10      the contribution of continental sources to the O3 levels over the North Atlantic.
11           Buhr et al. (1996) measured O3, CO, NO, and NOy as well as meteorological parameters
12      aboard the NCAR King Air in August 1993 during 16 flights over and near the  Gulf of Maine.
13      They found that O3 produced from anthropogenic precursors was dominant throughout the
14      experimental region below 1500 m, in altitude.
15           The National Research Council of Canada Twin Otter aircraft was used to measure the O3
16      and related compounds in the summertime atmosphere over southern Nova Scotia (Kleinman
17      et al., 1996a; Kleinman et al., 1996b). Forty-eight flights were performed, primarily over the
18      surface sampling site in Chebogue Point, Nova Scotia, or over the Atlantic Ocean. They found
19      that a wide variety of air masses with varying chemical content impact Nova Scotia.  The effect
20      depends on flow conditions relative to the locations of upwind emission regions and the degree
21      of photochemical processing associated with transport times ranging from about 1 - 5 d.  Moist
22      continental boundary layer air with high concentrations of O3 and other anthropogenic pollutants
23      was advected to Nova Scotia in relatively thin vertical layers, usually with a base altitude of
24      several  hundred meters. Dry air masses with high concentrations of O3 often had mixed
25      boundary layer and upper atmosphere source regions.  When a moist and dry air mass with the
26      same photochemical age and O3 concentration were compared, the dry air mass had lower
27      concentrations of NOy and aerosol particles, which was interpreted as evidence for the selective
28      removal of soluble constituents during vertical lifting.
29           Due to strong, low-level temperature inversions over the North Atlantic, near surface air is
30      often unrepresentative of the eastward transport of the North American plume because  of a
31      decoupling from the air transported aloft (Kleinman et al., 1996a; Daum et al., 1996; Angervine

        August 2005                             AX2-85       DRAFT-DO NOT QUOTE OR CITE

-------
 1      et al., 1996). Pollution plumes were observed in distinct strata up to 1 km. Plume chemical
 2      compositions were consistent with the occurrence of considerable photochemical processing
 3      during transit from source regions over the eastern seaboard of the U.S.  Ozone concentrations
 4      reached 150 ppbv, NOX conversion to its oxidation products exceeded 85%, and high hydrogen
 5      peroxide concentrations were observed (median 3.6 ppbv, maximum  11 ppbv). CO and O3
 6      concentrations were well  correlated (R2 = 0.64) with a slope (0.26) similar to previous
 7      measurements in photochemically aged air (Parrish et al., 1998). Ozone depended nonlinearly
 8      on the NOX oxidation product concentration, but there was a correlation (r2 = 0.73) found
 9      between O3  and the concentration of radical sink species as represented by the quantity
10      ((NOy-NOx) + 2H2O2)
11          Banic  et al. (1996) determined that the average mass of O3 transported through an area
12      1 m in horizontal extent and 5 km in the vertical over the ocean near Nova Scotia to be 2.8 g s"1,
13      moving from west to east. Anthropogenic O3 accounted for half of the transport below 1 km,
14      35 to 50% from  1 to 3 km, 25 to 50% from 3 to 4 km, and only 10% from 4 to 5 km. Merrill and
15      Moody (1996) analyzed the meteorological conditions during the NARE intensive period
16      (August 1 to September 13, 1993).  They determined the ideal meteorological scenario for
17      delivering pollution plumes from the U.S. East Coast urban areas over the Gulf of Maine to the
18      Maritime  Provinces of Canada to be warm sector flow ahead of an advancing cold front. In the
19      winter phase of NARE, O3 and CO were measured from the NOAA WP-3D Orion aircraft from
20      St. John's, Newfoundland, Canada, and Keflavik,  Iceland, from February 2 to 25, 1999 (Parrish
21      et al., 2000). In the lower troposphere over the western North Atlantic Ocean, the close
22      proximity of air masses with contrasting source signatures was remarkable. High levels of
23      anthropogenic pollution immediately adjacent to elevated O3 of stratospheric origin were
24      observed, similar to those reported by Prados et al. (1999). In air masses with differing amounts
25      of anthropogenic pollution, O3  was negatively correlated with CO, which indicates that
26      emissions from surface anthropogenic sources had reduced O3, in this wintertime period, even in
27      air masses transported into the  free troposphere.
28          The influence of the origin and evolution of airstreams on trace gas mixing ratios has been
29      studied in great detail for NARE aircraft data.  The typical midlatitude cyclone is composed of
30      four major component  airstreams, the warm conveyor belt, the cold conveyor belt, the post cold-
31      front airstream and the dry airstream (Cooper et al., 2001). The physical and chemical

        August 2005                            AX2-86      DRAFT-DO NOT QUOTE OR CITE

-------
 1      processing of trace species was characterized for each airstream, and a conceptual model of a
 2      midlatitude cyclones was developed (Cooper et al., 2002a). This showed how airstreams within
 3      midlatitude cyclones drew and exported trace gases from the polluted continental boundary
 4      layer, and the stratospherically enhanced mid-troposphere. Using back trajectories, airstream
 5      composition was related to the origin and transport history of the associated air mass.  The
 6      lowest O3 values were associated with airstreams originating in Canada or the Atlantic Ocean
 7      marine boundary layer; the highest O3 values were associated with airstreams of recent
 8      stratospheric origin.  The highest NOy values were seen in polluted outflow from New England
 9      in the lower troposphere.  A steep and positive O3/NOy slope was found for all airstreams in the
10      free troposphere regardless of air mass origin. Finally, the seasonal variation of photochemistry
11      and meteorology and their impact on trace gas mixing ratios in the conceptual cyclone model
12      was determined (Cooper et al., 2002b). Using a positive O3/CO slope as an indicator of
13      photochemical  O3  production, O3 production during late summer-early autumn is associated with
14      the lower troposphere post-cold-front airstream and all levels of the WCB, especially the lower
15      troposphere.  However, in the early spring, there is no significant photochemical O3 production
16      for airstreams at any level, and negative slopes in the dry air airstream indicate STE causes the
17      O3 increase in the mid- and upper troposphere.
18           Stohl et al. (2002) analyzed total odd nitrogen (NOy) and CO data taken during NARE in
19      spring 1996 and fall 1997. They studied the removal timescales of NOy originating from surface
20      emissions of NOX and what fraction reached the free troposphere. NOX limits O3 production in
21      the free troposphere and can be regenerated from NOy after the primary NOx has been
22      exhausted. It was  determined that < 50% of the NOy observed above 3 km came from
23      anthropogenic surface emissions.  The rest had to have been emitted in situ.
24           Several studies (e.g., Stohl and Trickl, 1999; Brunner et al., 1998; Schumann et al., 2000;
25      Stohl et al., 2003; Traub et al., 2003) have identified plumes that have originated in North
26      America over Europe and over the eastern Mediterranean basin (e.g., Roelofs et al., 2003; Traub
27      et al., 2003).  Modeling studies indicate that North American emissions contribute roughly 20%
28      to European CO levels and 2 to 4 ppb to surface O3, on average. Episodic events, such as forest
29      fires in North America have  also been found to result in elevated CO and O3 levels and visible
30      haze layers in Europe (Volz-Thomas, et al., 2003). The O3 is either transported from North
31      America or formed during transport across the North Atlantic Ocean,  perhaps as the result of

        August 2005                             AX2-87      DRAFT-DO NOT QUOTE OR CITE

-------
 1      interactions between the photochemical degradation products of acetone with emissions of NOX
 2      from aircraft (Bruhl et al., 2000; Arnold et al., 1997). In addition, North American and European
 3      pollution is exported to the Arctic.  Eckhardt et al. (2003) show that this transport is related to
 4      the phase of the North Atlantic Oscillation which has a period of about 20 years.
 5
 6      AX2.3.5   The Relation of Ozone to Solar Ultraviolet Radiation, Aerosols,
 7                 and Air Temperature
 8      AX2.3.5.1   Solar Ultraviolet Radiation and Ozone
 9           The effects of sunlight on photochemical oxidant formation, aside from the role of solar
10      radiation in meteorological processes, are related to its intensity and its spectral distribution.
11      Intensity varies diurnally, seasonally, and with latitude, but the effect of latitude is strong only in
12      the winter.  Ultraviolet radiation from the sun plays a key role in initiating the photochemical
13      processes leading to O3 formation and affects individual photolytic reaction steps. However,
14      there is little empirical evidence in the literature, directly linking day-to-day variations in
15      observed UV radiation levels with variations in O3 levels.
16           In urban environments the rate of O3 formation is sensitive to the rate  of photolysis of
17      several species including H2CO, H2O2, O3, and especially NO2. Monte Carlo calculations
18      suggest that model calculations of photochemical O3 production are most sensitive to uncertainty
19      in the photolysis rate coefficient for NO2 (Thompson and Stewart, 1991; Baumann et al., 2000).
20      The International Photolysis Frequency Measurement and Modeling Intercomparison (IPMMI)
21      hosted recently by NCAR in Boulder, CO brought together more than 40 investigators from
22      8 institutions from around the world (Bais et al.,  2003; Cantrell et al., 2003  and Shelter et al.,
23      2003).  They compared direct actinometric measurements, radiometric measurements, and
24      numerical models of photolysis rate coefficients, focusing on O3 to O(JD) and NO2, referred  to as
25      j(O3)andj(NO2).
26           The combination of direct measurements and comparisons to model calculations indicated
27      that for clear skies, zenith angles less than 70°, and low aerosol loadings, the absolute value  of
28      the j(NO2)  at the Earth's surface is known to better than 10% with 95% confidence.  The results
29      suggest that the cross sections of Harder et al. (1997a) may yield more accurate values when
30      used in model calculations of j(NO2). Many numerical models agreed among themselves and
31      with direct measurements (actinometers) and semi-direct measurements (radiometers) when

        August 2005                             AX2-88      DRAFT-DO NOT QUOTE OR CITE

-------
 1      using ATLAS extraterrestrial flux from Groebner and Kerr (2001). The results of IPMMI
 2      indicate numerical models are capable of precise calculation of photolysis rates at the surface
 3      and that uncertainties in calculated chemical fields arise primarily from uncertainties in the
 4      variation of actinic flux with altitude in addition to the impact of clouds and aerosols on
 5      radiation.
 6
 7      AX2.3.5.2  Impact of Aerosols on Radiation and Photolysis Rates and
 8                 Atmospheric Stability
 9           Because aerosol particles influence the UV flux there is a physical link between particles
10      and gases that depends on the concentration, distribution, and refractive index of the particles.
11      Scattering of UV radiation by tropospheric aerosol particles can strongly impact photolysis rates
12      and thus photochemical O3 production or destruction. The effect shows high sensitivity to the
13      properties of the aerosol. Particles in the boundary layer can accelerate photochemistry if the
14      single scattering albedo is near unity, such as for sulfate and ammoniated sulfate aerosols, or
15      inhibit O3 production if the single scattering albedo is low, such as for mineral dust or soot
16      (Dickerson et al., 1997; Jacobson, 1998; Liao et al., 1999; Castro et al., 2001; Park et al.,  2001).
17      Any aerosol layer in the free troposphere will reduce photolysis rates in the boundary layer.
18           The interaction of aerosols, photochemistry, and atmospheric thermodynamic processes
19      can impact radiative transport, cloud microphysics, and atmospheric stability with respect to
20      vertical mixing. Park et al. (2001) developed a single-column chemical transport model that
21      simulates vertical transport by convection, turbulent mixing, photochemistry, and interactive
22      calculations of radiative fluxes and photolysis rates. Results from simulations of an episode over
23      the eastern United States showed strong sensitivity to convective mixing and aerosol optical
24      depth. The aerosol optical properties observed during the episode produced a surface cooling of
25      up to 120 W/m2 and  stabilized the atmosphere suppressing convection. This suggests two
26      possible feedbacks mechanisms between aerosols and O3-reduced vertical mixing would  tend to
27      increase the severity of O3 episodes, while reduced  surface temperatures would decrease it.
28
29      AX2.3.5.3  Temperature and Ozone
30           An association between surface O3 concentrations and temperature has been demonstrated
31      from measurements in outdoor smog chambers and from measurements in ambient air.
32      Numerous ambient studies done over more than a decade have reported that successive

        August 2005                             AX2-89      DRAFT-DO NOT QUOTE OR CITE

-------
 1     occurrences or episodes of high temperatures characterize high O3 years (Clark and Karl, 1982;
 2     Kelly et al., 1986). The relation of daily maximum 8-h average O3 concentration to daily
 3     maximum temperature from May to September 1994 to 2004 is illustrated in Figure AX2-14 for
 4     the Baltimore Air Quality Forecast Area.  The relation, based on daily maximum 1-h average O3
 5     concentration is illustrated in Figure AX2-15.  The relations are very similar in the two figures,
 6     reflecting the high degree of correlation (r = 0.98) between the daily maximum 1-h and 8-h O3
 7     concentrations.  The relation of daily maximum 8-h average O3 to daily maximum temperature
 8     from May to September 1994 to 2004 is illustrated in Figure AX2-15 for the three sites
 9     downwind of Phoenix, AZ on high O3 days (cf, Figure AX3-32).  As can be seen from a
10     comparison of Figures AX2-14 and AX2-16, O3 concentrations in the Phoenix area are not as
11     well correlated with daily maximum temperature  (r = 0.14)  as they are in the Baltimore Area
12     (r = 0.74).  There appears to be an upper-bound on O3 concentrations that increases with
13     temperature. Likewise, Figure AX2-16 shows that a similar qualitative relationship exists
       between O3 and temperature even at a number of nonurban locations.
                       160'
                    .a
                    a.
                    a.
                     c
                     O
                    O
                    CO
                    E
                    3
                    E
                    'x
                       100'
                        80
60
                        40

                                    10    15    20    25    30    35    40    45    50
       Figure AX2-14.  A scatter plot of daily maximum 8-h O3 concentration versus daily
                        maximum temperature in the Baltimore, MD Air Quality Forecast Area.
       Source: Piety (2005)
       August 2005
                       AX2-90
DRAFT-DO NOT QUOTE OR CITE

-------
                200
             O.

             O.
             c
             
-------
 1           The notable trend in these plots is the apparent upper-bound to O3 concentrations as a
 2      function of temperature.  It is clear that, at a given temperature, there is a wide range of possible
 3      O3 concentrations because other factors (e.g., cloudiness, precipitation, wind speed) can reduce
 4      O3 production rates.  The upper edge of the curves may represent a practical upper bound on the
 5      maximum O3 concentration achieved under the most favorable conditions. Relationships
 6      between peak O3 and temperature also have been recorded by Wunderli and Gehrig (1991) for
 7      three locations in Switzerland.  At two sites near Zurich, peak O3 increased 3 to 5 ppb/°C for
 8      diurnal average temperatures between  10 and 25 °C, and little change in peak O3 occurred for
 9      temperatures below 10 °C.  At the third site, a high-altitude location removed from
10      anthropogenic influence, a much smaller variation of O3 with temperature was observed.
11           Some possible  explanations for the correlation of O3 with temperature include:
12       (1)   Increased photolysis rates under meterological conditions associated with higher
              temperatures;
13       (2)   Increased H2O concentrations with higher temperatures as this will lead to greater OH
              production via R(2-6);
14       (3)   Enhanced thermal decomposition of PAN and similar compounds to release NOX at
              higher temperatures;
15       (4)   Increase of anthropogenic hydrocarbon (e.g., evaporative losses) emissions or NOX,
              emissions with temperature or both;
16       (5)   Increase of natural hydrocarbon emissions (e.g., isoprene) with temperature; and
17       (6)   Relationships  between high temperatures and stagnant circulation patterns.
18       (7)   Advection of warm air enriched with O3.
19           Cardelino and Chameides (1990) and Sillman and Samson (1995) both identified the
20      temperature-dependent thermal decomposition of PAN as the primary cause of the observed
21      O3-temperature relationship. When temperatures are low PAN is relatively stable. Formation of
22      PAN represents  a significant sink for NOX (in low NOX rural areas) and radicals (in high NOX
23      urban areas).  This has the effect of slowing the rate of O3 production. Sillman and Samson
24      found that the impact of the PAN decomposition rate could explain roughly half of the observed
25      correlation between O3 and temperature.  Jacob et al. (1993) found that warm events in summer
26      in the United States were likely to occur during stagnant meteorological conditions, and the
27      concurrence between warm temperatures and meteorological stagnation also explained roughly

        August 2005                             AX2-92       DRAFT-DO NOT QUOTE OR CITE

-------
 1     half of the observed O3-temperature correlation.  Other possible causes include higher solar
 2     radiation during summer, the strong correlation between biogenic emission of isoprene and
 3     temperature, and the somewhat weaker tendency for increased anthropogenic emissions
 4     coinciding with warmer temperatures.
 5          However, it should also be noted that a high correlation of O3 with temperature does not
 6     necessarily imply a causal relation.  Extreme episodes of high temperatures (a heat wave) are
 7     often multiday events — high O3 episodes are also multiday events, concentrations build,
 8     temperatures rise, but both are being influenced by larger-scale regional  or synoptic
 9     meteorological conditions. It also seems apparent, that while there is a trend for higher O3
10     associated with higher temperatures, there is also much greater variance  in the range of O3
11     mixing ratios at higher temperatures.
12
13
14     AX2.4   THE RELATION OF OZONE TO ITS PRECURSORS AND
15               OTHER OXIDANTS
16          Ozone is unlike many other species whose rates of formation vary  directly with the
17     emissions of their precursors. Ozone changes in a nonlinear fashion with the concentrations of
18     its precursors.  At the low NOX concentrations found in most environments, ranging from remote
19     continental areas to rural and suburban areas downwind of urban centers the net production of O3
20     increases with increasing NOX.  At the high NOX concentrations found in downtown metropolitan
21     areas, especially near busy streets and roadways, and in power plant plumes there is net
22     destruction (titration) of O3 by reaction with NO.  In between these two regimes there is a
23     transition stage in which O3 shows only a weak dependence on NOX concentrations. In the high
24     NOX regime, NO2 scavenges OH radicals which would otherwise oxidize VOCs to produce
25     peroxy radicals, which in turn would oxidize NO to NO2. In the low NOX regime, the oxidation
26     of VOCs generates,  or at least does not consume, free radicals and O3 production varies directly
27     with NOX. Sometimes the terms VOC limited and NOX limited  are used to describe these two
28     regimes. However, there are difficulties with this usage because (1) VOC measurements are not
29     as abundant as they are for nitrogen oxides, (2) rate coefficients for reaction of individual VOCs
30     with free radicals vary over an extremely wide range, and (3) consideration is not given to CO
31     nor to reactions that can produce free radicals without invoking VOCs. The terms NOx-limited

       August 2005                             AX2-93      DRAFT-DO NOT QUOTE OR CITE

-------
 1      and NOx-saturated (e.g., Jaegle et al., 2001) will be used wherever possible to describe these two
 2      regimes more adequately. However, the terminology used in original articles will also be kept.
 3      The chemistry of OH radicals, which are responsible for initiating the oxidation of hydrocarbons,
 4      shows behavior similar to that for O3 with respect to NOX concentrations (Hameed et al., 1979;
 5      Pinto et al., 1993; Poppe et al., 1993; Zimmerman and Poppe, 1993).  These considerations
 6      introduce a high degree  of uncertainty into attempts to relate changes in O3 concentrations to
 7      emissions of precursors.
 8           Various analytical techniques have been proposed that use ambient NOX and VOC
 9      measurements to derive information about O3 production and O3-NOX-VOC sensitivity. It has
10      been suggested that O3 formation in individual urban areas could be understood in terms of
11      measurements of ambient NOX and VOC concentrations during the  early morning (e.g., National
12      Research Council, 1991). In this approach, the ratio of summed (unweighted by chemical
13      reactivity) VOC to NOX is used to determine whether conditions were NOx-sensitive or VOC
14      sensitive. This procedure is inadequate because it omits many factors that are recognized as
15      important for O3 production:  the impact of biogenic VOCs (which  are not present in urban
16      centers during early morning); important individual differences in the ability of VOCs to
17      generate free radicals (rather than just total VOC) and other differences  in O3 forming potential
18      for individual VOCs (Carter,  1995); the impact of multiday transport; and general changes in
19      photochemistry as air moves downwind from urban areas (Milford  et al., 1994).
20           Jacob et al. (1995) used a combination of field measurements and  a chemistry-transport
21      model (CTM) to show that the formation of O3 changed from NOx-limited to NOx-saturated as
22      the season changed from summer to fall at a monitoring site in Shenandoah National Park, VA.
23      Photochemical production of O3 generally occurs simultaneously with the production of various
24      other species: nitric acid (HNO3), organic nitrates, and hydrogen peroxide.  The relative rate of
25      production of O3 and other species varies depending on photochemical conditions, and can be
26      used to provide information about O3-precursor sensitivity.
27           There are no hard and fast rules governing the levels of NOX at which the transition from
28      NOx-limited to NOx-saturated conditions occurs. The transition between these two regimes  is
29      highly spatially and temporally dependent. Similar responses to NOX additions from commercial
30      aircraft have also been found  for the upper troposphere (Bruhl et al., 2000).  Bruhl et al. (2000)
31      found that the NOX levels for  O3 production versus loss are highly sensitive to the radical sources

        August 2005                             AX2-94       DRAFT-DO NOT QUOTE OR  CITE

-------
 1      included in model calculations. They found that the inclusion of only CH4 and CO oxidation
 2      leads to a decrease in net O3 production in the North Atlantic flight corridor due to NO emissions
 3      from aircraft.  However, the inclusion of acetone photolysis was found to shift the maximum in
 4      O3 production to higher NOX mixing ratios, thereby reducing or eliminating areas in which there
 5      is a decrease in O3 production rates due to aircraft emissions.
 6           Trainer et al. (1993) suggested that the slope of the regression line between O3 and
 7      summed NOX oxidation products (NOZ, equal to the difference between measured total reactive
 8      nitrogen, NOy, and NOX) can be used to estimate the rate of O3 production per NOX (also known
 9      as the O3 production efficiency, or OPE).  Ryerson et al. (1998,  2001) used measured
10      correlations between O3 and NOZ to identify different rates of O3 production in plumes from
11      large point sources.
12           Sillman (1995) and Sillman and He (2002) identified several secondary reaction products
13      that show different correlation patterns for NOx-limited conditions and NOx-saturated  conditions.
14      The most important correlations are for O3 versus NOy, O3 versus NOZ, O3 versus HNO3, and
15      H2O2 versus HNO3. The correlations between O3 and NOy, and  O3 and NOZ are especially
16      important because measurements of NOy and NOX are widely available. Measured O3  versus
17      NOZ (Figure AX2-17) shows distinctly different patterns in different locations.  In rural areas and
18      in urban areas such as Nashville, TN, O3 shows a strong correlation with NOZ and a relatively
19      steep slope to the regression line. By contrast, in Los Angeles O3 also increases with NOZ, but
20      the rate of increase of O3 with NOZ is lower and the O3 concentrations for a given NOZ value are
21      generally lower.
22           The difference between NOx-limited and NOx-saturated regimes is also reflected in
23      measurements of hydrogen peroxide (H2O2).  Hydrogen peroxide production is highly sensitive
24      to the abundance of free radicals and is thus favored in the NOx-limited regime, typical of
25      summer conditions. Differences between these two regimes are also related to the preferential
26      formation of sulfate during summer and to the inhibition of sulfate and hydrogen  peroxide
27      during winter (Stein and Lamb, 2003).  Measurements in the rural eastern United States (Jacob
28      et al., 1995) Nashville (Sillman et al., 1998), and Los Angeles (Sakugawa and Kaplan, 1989)
29      show large differences in H2O2 concentrations between likely NOx-limited and NOx-saturated
30      locations.
        August 2005                             AX2-95       DRAFT-DO NOT QUOTE OR CITE

-------
                                                        x
                                                       X
                                                              x
                                                              X
                                                      X X
                                                     X
                                                                 X
                                       10
                                            20
                                        NOZ (ppb)
     30
                                                                 x
                                                                         X
                                                                     X
40
       Figure AX2-17.  Measured values of O3 and NOZ (NOy - NOX) during the afternoon at
                       rural sites in the eastern United States (gray circles) and in urban areas
                       and urban plumes associated with Nashville, TN (gray dashes), Paris, FR
                       (black diamonds) and Los Angeles, CA (X's).
       Sources: Trainer etal. (1993), Sillmanetal. (1997, 1998, 2003).
 1
 2
 3
 4
 5
 9
10
11
12
13
     The discussion in Section AX2.4.1 centers mainly on the relations among O3, NOX and its
oxidation products, represented as NOZ (NOy - NOX) and VOCs derived from the results of field
studies.  Most of these studies examined processes occurring in power plant and urban plumes.

AX2.4.1   Summary of Results for the Relations Among Ozone, its Precursors
           and Other Oxidants from Recent Field Experiments
AX2.4.1.1  Results from the Southern Oxidant Study and Related Experiments
     The Southern Oxidant Study (SOS) was initiated to describe the sources, variation, and
distribution of O3 and its precursors in the southeastern United States during the summer season
(Hiibler et al., 1998; Meagher et al.,  1998; Goldan et al., 2000). Specific issues that were
addressed included: (1) the role of biogenic VOC and NOX emissions on local and regional O3
production, (2) the effect of urban-rural exchange/interchange on local and regional O3
production, (3) sub-grid-scale photochemical and meteorological  processes, and (4) the
       August 2005
                                      AX2-96
DRAFT-DO NOT QUOTE OR CITE

-------
 1      provision of a high-quality chemical and meteorological data set to test and improve observation
 2      and emission-based air quality forecast models. Some of the more significant findings of the
 3      1994 to 1995 studies include the following: (1) Ozone production in Nashville was found to be
 4      close to the transition between NOx-limited and NOx-saturated regimes.  (2) The number of
 5      molecules of O3 produced per molecule of NOX oxidized in power plant plumes, or the O3
 6      production efficiency (OPE) was found to be inversely proportional to the NOX emission rate,
 7      with the plants having the highest NOX emissions exhibiting the lowest OPE. (3) During
 8      stagnant conditions, winds at night dominated pollutant transport and represent the major
 9      mechanism for advecting urban pollutants to rural areas—specific findings follow.
10           As part of SOS, the Tennessee Valley Authority's instrumented helicopter conducted
11      flights over Atlanta, Georgia to investigate the evolution of the urban O3 plume (Imhoff et al.,
12      1995). Ozone peak lev els occurred at 20 - 40 km downwind of the city center. The OPE
13      obtained from five afternoon flights ranged between 4 and 10 molecules of O3 per molecule
14      ofNOx.
15           Berkowitz and Shaw (1997) measured O3 and its precursors at several altitudes over a
16      surface site near Nashville during SOS to determine the effects of turbulent mixing on
17      atmospheric chemistry. Early morning measurements of O3 aloft revealed values near 70 ppb,
18      while those measured at the surface were closer to 25 ppb.  As the daytime mixed layer
19      deepened, surface O3 values steadily increased until they reached 70 ppb. The onset of
20      turbulence increased isoprene mixing ratios aloft by several orders of magnitude and affected the
21      slope of O3 as a function of NOy for each of the flight legs.  Measurements from nonturbulent
22      flight legs yielded slopes  that were considerably steeper than those from measurements made in
23      turbulence.  This study shows that the concentration of O3 precursors aloft is dependent on the
24      occurrence of turbulence, and turbulent mixing could explain the evolution of O3 concentrations
25      at the surface. In general, conclusions regarding pollutant concentrations must account for both
26      chemical  and local dynamic processes.
27           Gillani et al. (1998) analyzed data from instrumented aircraft during SOS that flew through
28      the plumes of three large, tall-stack, base-load, Tennessee Valley Authority (TVA) coal-fired
29      power plants in northwestern Tennessee. They determined that plume chemical maturity and
30      peak O3 and NOZ production occurred within 30 to 40 km and 4 hours of summer daytime
31      convective boundary layer (CBL) transport time for a coal-fired power plant in the Nashville,

        August 2005                            AX2-97      DRAFT-DO NOT QUOTE OR CITE

-------
 1      TN urban O3 nonattainment area (Gallatin). For a rural coal-fired power plant in an isoprene-
 2      rich forested area about 100 km west of Nashville (Cumberland), plume chemical maturity and
 3      peak O3 and NOZ production were realized within approximately 100 km and 6 hours of CBL
 4      transport time.  Their findings included approximately 3 molecules of O3 and more than
 5      0.6 molecules of NOZ may be produced in large isolated rural power plant plumes (PPPs) per
 6      molecule of NOX release; the corresponding peak yields of O3 and NOZ may be significantly
 7      greater in urban PPPs. Both power plants can contribute as much as 50 ppb of excess O3 to the
 8      Nashville area,  raising the local levels to well above 100 ppb. Also using aircraft data collected
 9      during SOS, Ryerson et al. (1998) concluded that the lower and upper limits to O3 production
10      efficiency in the Cumberland and Paradise PPPs (located in rural Tennessee) were 1 and
11      2 molecules of O3 produced per molecule of NOX emitted.  The estimated lower and upper limits
12      to O3 production efficiency in the Johnsonville PPP (also located in rural Tennessee) were
13      higher, at 3 and 7.
14          The NOAA airborne O3 lidar provided detailed,  three-dimensional lower tropospheric O3
15      distribution information during June and July 1995 in the Nashville area (Senff et al., 1998;
16      Alvarez et al., 1998). The size and shape of power plant plumes as well as their impacts on O3
17      concentration levels as the plume is advected downwind were studied.  Specific examples
18      include:  the July 7 Cumberland plume that was symmetrical and confined to the boundary layer,
19      and the July 19  Cumberland plume that was irregularly shaped with two cores, one above and
20      the other within the boundary layer. The disparate plume characteristics on these two days were
21      the result of distinctly different meteorological conditions. Ozone in the plume was destroyed at
22      a rate of 5 to 8 ppbv h'1 due to NOX titration close to the power plant, while farther downwind,
23      O3 was produced at rates between 1.5 and 4 ppbv h"1.  The lidar O3 measurements compared
24      reasonably well with in situ values, with the average magnitude of the offsets over all the flights
25      at 4.3 ppbv (7%).
26          The highest O3 concentrations observed during the 1995 SOS in middle Tennessee
27      occurred during a period of strong, synoptic-scale stagnation from July 11 through July  15. This
28      massive episode covered most of the eastern United States (e.g., Ryan et al., 1998). During this
29      time, the effects of vertical wind profiles on the buildup and transport of O3 were studied by
30      Banta et al. (1998) using an airborne differential absorption lidar (DIAL) system. Vertical cross
31      sections showed O3 concentrations exceeding 120 ppb extending to nearly 2 km above ground

        August 2005                            AX2-98      DRAFT-DO NOT QUOTE OR CITE

-------
 1      level, but that O3 moved little horizontally.  Instead, it formed a dome of pollution over or near
 2      Nashville.  Due to the stagnant daytime conditions (boundary layer winds ~1 to 3 m s"1),
 3      nighttime transport of O3 became disproportionately important.  At night, in the layer between
 4      100 and 2000 m AGL (which had been occupied by the daytime mixed layer), the winds could
 5      be accelerated to 5 to 10 m s"1 as a result of nocturnal decoupling from surface friction. Data
 6      from surface and other aircraft measurements taken during this period suggest that the
 7      background air and the edges of the urban plume were NOX sensitive and the core of the urban
 8      plume was hydrocarbon sensitive (Valente et al., 1998). Also revealed was the fact that the
 9      surface monitoring network failed to document the maximum surface O3 concentrations. Thus,
10      monitoring networks, especially in medium-sized urban areas under slow transport conditions,
11      may underestimate the magnitude and frequency of urban O3 concentrations greater than
12      120 ppb.
13          Nunnermacker et al. (1998) used both aircraft and surface data from SOS to perform a
14      detailed kinetic analysis of the chemical evolution of the Nashville urban plume. The analysis
15      revealed OH concentrations around 1.2 x 107 cm"3 that consumed 50% of the NOX within
16      approximately 2 hours, at an OPE of 2.5 to 4 molecules for each molecule of NOX.
17      Anthropogenic hydrocarbons provided approximately 44% of the fuel for O3 production by the
18      urban plume.
19           Surface and aircraft observations of O3 and O3 precursors were compared during SOS to
20      assess the degree to which mid-day surface measurements may be considered representative of
21      the larger planetary boundary layer (PEL) (Luke et al., 1998). Overall agreement between
22      surface and aircraft O3 measurements was excellent in the well-developed mixed layer
23      (r2 = 0.96), especially in rural-regional background air and under stagnant conditions, where
24      surface concentrations change only slowly.  Vertical variations in trace gas concentrations were
25      often minimal in the well-mixed PEL, and measurements at the surface always agreed well with
26      aircraft observations up to the level of measurements (460 m above ground level). Under
27      conditions of rapidly varying surface concentrations (e.g., during episodes of power plant plume
28      fumigation and early morning boundary layer development), agreement between surface and
29      aloft was dependent upon the spatial (aircraft) and temporal (ground) averaging intervals used in
30      the comparison.  Under these conditions, surface sites were representative of the PEL only to
31      within a few kilometers horizontally.

        August 2005                            AX2-99       DRAFT-DO NOT QUOTE OR CITE

-------
 1          On four days during SOS, air samples were taken in the plume of the Cumberland Power
 2     Plant in central Tennessee using an instrumented helicopter to investigate the evolution of
 3     photochemical smog (Luria et al.,  1999, 2000).  Twelve crosswind air-sampling traverses were
 4     made between 35 and 116 km from this Power Plant on 16 July 1995. Winds, from the west-
 5     northwest during the sampling period, directed the plume toward Nashville.  Ten of the traverses
 6     were performed upwind of Nashville, where the plume was isolated, and two were made
 7     downwind of the city.  The results indicated that even six hours after the plume left the stacks,
 8     excess O3 production was limited to the edges of the plume.  Excess O3 production within the
 9     plume was found to vary from 20 ppb up to 55 ppb. It was determined that this variation
10     corresponded to differences in ambient isoprene levels. Excess O3 (up to 109 ppbv, 50 to
11     60 ppbv above background), was produced in the center of the plume when there was sufficient
12     mixing upwind of Nashville. The power plant plume apparently mixed with the urban plume
13     also, producing O3 up to  120 ppbv 15 to 25 km downwind of Nashville.
14          Nunnermacker et al. (2000) used data from the DOE G-l aircraft to characterize emissions
15     from a small power plant plume (Gallatin) and a large power plant plume (Paradise) in the
16     Nashville region.  Observations made on July 3, 7, 15, 17, and 18,  1995, were compiled, and a
17     kinetic analysis of the chemical  evolution of the power plant plumes was performed. OPEs were
18     found to be 3 in the Gallatin and 2 in the Paradise plumes. Lifetimes for NOX (2.8 and 4.2 hours)
19     and NOy (7.0 and 7.7 hours) were determined in the Gallatin and Paradise plumes, respectively.
20     These NOX and NOy lifetimes imply rapid loss of NOZ (assumed to be primarily HNO3), with a
21     lifetime determined to be 3.0 and 2.5 hours for the Gallatin and Paradise plumes, respectively.
22
23     AX2.4.1.2 Results from Studies on Biogenic and Anthropogenic Hydrocarbons and
24                Ozone Production
25          Williams et al. (1997) made the first airborne measurements of peroxy-methacrylic nitric
26     anhydride (MPAN), which is formed from isoprene-NOx chemistry and is an indicator of recent
27     O3 production from isoprene and therefore biogenic hydrocarbons (BHC).  They also measured
28     peroxyacetic nitric anhydride (PAN), peroxypropionic nitric anhydride (PPN), and O3 to estimate
29     the contributions of anthropogenic hydrocarbons (AHC) and BHC to regional tropospheric O3
30     production.
31
       August 2005                            AX2-100     DRAFT-DO NOT QUOTE OR CITE

-------
 1          Airborne measurements of MPAN, PAN, PPN, and O3 were made during the 1994 and
 2      1995 Nashville intensive studies of SOS to determine the fraction of O3 formed from
 3      anthropogenic NOX and BHC (Roberts et al., 1998). It was found that PAN, a general product of
 4      hydrocarbon-NOx photochemistry, could be well represented as a simple linear combination of
 5      contributions from BHC and AHC as indicated by MPAN and PPN, respectively. The
 6      PAN/MPAN ratios, characteristic of BHC-dominated chemistry, ranged from 6 to 10. The
 7      PAN/PPN ratios, characteristic of AHC-dominated chemistry, ranged from 5.8 to 7.4. These
 8      ratios were used to estimate the contributions of AHC and BHC to regional tropospheric O3
 9      production.  It was estimated that substantial O3 (50 to 60 ppbv) was produced from BHC when
10      high NOX from power plants was present in areas of high BHC emission.
11
12      AX2.4.1.3   Results of Studies on Ozone Production in Mississippi and Alabama
13          Aircraft flights made in June 1990 characterized the variability of O3 and reactive nitrogen
14      in the lower atmosphere over Mississippi and Alabama. The variety and proximity of sources
15      and the photochemical production and loss of O3 were found to be contributing factors (Ridley
16      et al., 1998).  Urban, biomass burning, electrical power plant, and paper mill plumes were all
17      encountered during these flights. Urban plumes from Mobile, AL had OPEs as high as 6 to
18      7 ppbv O3 per ppbv of NOX. Emissions measured from biomass burning had lower efficiencies
19      of 2 to 4  ppbv O3 per ppbv of NOX, but the average rate of production of O3 was as high as
20      58 ppbv hr1 for one fire where the plume was prevented from vertical mixing. Near-source
21      paper mill and power plant plumes showed O3 titration, while far-field observations of power
22      plant plumes showed net O3 production. Early morning observations below a nocturnal
23      inversion provided evidence for the nighttime  oxidation of NOX to reservoir species.
24          Aircraft measurements of O3 and oxides  of nitrogen were made downwind of Birmingham,
25      AL to estimate the OPE in the urban plume (Trainer et al., 1995). NOX emission rates were
26      estimated at 0.6 x 1025 molecules s"1 with an uncertainty of a factor of 2. During the
27      summertime it was determined that approximately seven O3 molecules could be formed for
28      every molecule of NOX  emitted by the urban and proximately located power plant plumes.  The
29      regional O3, the photochemical production of O3 during the oxidation of the urban emissions, and
30      wind speed and direction all combined to dictate the magnitude and location of the peak O3
31      concentrations observed in the vicinity of the Birmingham metropolitan area.

        August 2005                           AX2-101      DRAFT-DO NOT QUOTE OR CITE

-------
 1           Aircraft observations of rural U.S. coal-fired power plant plumes in the middle Mississippi
 2      and Tennessee Valleys were used to quantify the nonlinear dependence of tropospheric O3
 3      formation on plume NOX concentration, determined by plant NOX emission rate and atmospheric
 4      dispersion (Ryerson et al., 2001). The ambient availability of reactive VOCs, primarily biogenic
 5      isoprene, was also found to affect O3 production rate and yield in these rural plumes. Plume O3
 6      production rates and yields as a function of NOX and VOC concentrations differed by a factor of
 7      2 or more. These large differences indicate that power plant NOX emission rates and geographic
 8      locations play a large role in tropospheric O3 production.
 9
10      AX2.4.1.4  The Nocturnal Urban Plume Over Portland, Oregon
11           Aircraft observations of aerosol surface area, O3, NOy and moisture were made at night in
12      the Portland, Oregon urban plume (Berkowitz et al., 2001). Shortly after sunset, O3, relative
13      humidity, NOy and aerosol number density were all positively correlated. However, just before
14      dawn, O3 mixing ratios were highly anti-correlated with aerosol number density, NOy and
15      relative humidity. Back-trajectories showed that both samples came from a common source to
16      the northwest of Portland.  The pre-dawn parcels passed directly over Portland, while the other
17      parcels passed to the west of Portland. Several hypotheses were put forward to explain the loss
18      of O3 in the parcels that passed over Portland, including homogeneous gas-phase mechanisms
19      and a heterogeneous mechanism on the aerosol particle surface.
20
21      AX2.4.1.5  Effects of VOCs in Houston on Ozone Production
22           Aircraft Observations of O3 and O3 precursors over Houston, TX, Nashville, TN; New
23      York, NY; Phoenix, AZ, and Philadelphia, PA showed that despite similar NOX concentrations,
24      high concentrations of VOCs in the lower atmosphere over Houston led to calculated O3
25      production rates that were 2 to 5 times higher than in the other 4 cities (Kleinman et al., 2002).
26      Concentrations of VOCs and O3 production rates are highest in the Ship Channel region of
27      Houston, where one of the largest petrochemical complexes in the world is located. As a result,
28      Houston lays claim to the highest recorded hourly average O3 concentrations  in the United States
29      within the last 5 years (in excess of 250 ppb).
30
        August 2005                            AX2-102      DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX2.4.1.6  Chemical and Meteorological Influences on the Phoenix Urban Ozone Plume
 2           The interaction of chemistry and meteorology for western cities can contrast sharply with
 3     that of eastern cities. A 4-week field campaign in May and June of 1998 in the Phoenix area
 4     comprised meteorological and chemical measurements (Fast et al., 2000). Data from models and
 5     observations revealed that heating of the higher terrain north and east of Phoenix produced
 6     regular, thermally driven circulations during the afternoon from the south and southwest through
 7     most of the boundary layer, advecting the urban O3 plume to the northeast. Deep  mixed layers
 8     and moderate winds aloft ventilated the Phoenix area during the study period so that multiday
 9     buildups of locally produced O3 did not appear to contribute significantly to O3 levels.
10     Sensitivity simulations estimated that 20% to 40% of the afternoon surface O3 mixing ratios
11     (corresponding to 15 to 35 ppb) was due to the entrainment of O3 reservoirs into the growing
12     convective boundary layer.  The model results also indicated that O3 production in this arid
13     region is NOx-saturated, unlike most eastern U.S. sites.
14
15     AX2.4.1.7  Transport of Ozone and Precursors on the Regional Scale
16           Instrumented aircraft flights by the University of Maryland in a Cessna 172  and Sonoma
17     Technology, Inc. in a Piper Aztec measured the vertical profiles of trace gases and
18     meteorological parameters in Virginia, Maryland, and Pennsylvania on July 12 -  15,  1995 during
19     a severe O3 episode in the mid-Atlantic region (Ryan et al., 1998). Ozone measured upwind of
20     the urban centers reached 80 to 110 ppbv.  Layers of high O3 aloft were associated with local
21     concentration maxima of SO2 and NOy, but not CO or NOX.  This, together with a  back trajectory
22     analysis, implicated coal-fired power plants in the industrialized Midwest as the source of the
23     photochemically aged air in the upwind boundary of the urban centers.  When the PEL over the
24     Baltimore-Washington area deepened, the O3 and O3 precursors that had been transported from
25     the west and northwest mixed with the local  emissions and O3  in excess of 125  ppbv was
26     measured at the surface.
27           During the blackout of August 14, 2003 Marufu et al. (2004) measured profiles of O3, SO2
28     and CO over areas in western Pennsylvania,  Maryland and Virginia. They found  notable
29     decreases in O3, SO2, and NOX, over areas affected by the blackout but not over those that were
30     not affected. They also found that CO concentrations aloft were comparable over areas affected
31     and not affected by the blackout.  They attributed the differences in concentrations between what

       August 2005                           AX2-103      DRAFT-DO NOT QUOTE OR CITE

-------
 1     was observed and what was expected to the reduction in emissions from power plants mainly in
 2     the Ohio Valley. They also reasoned that the CO concentrations were relatively unaffected
 3     because they arise from traffic emissions, which may have been largely unaffected by the
 4     blackout.  However, the blackout also disrupted many industries, small scale emission sources,
 5     and rail and air transportation.
 6           The Department of Energy G-l aircraft flew in the New York City metropolitan area in
 7     the summer of 1996 as part of the North American Research Strategy for Tropospheric
 8     Ozone-Northeast effort to ascertain the causes leading to high O3 levels in the northeastern
 9     United States (Kleinman et al., 2000). Ozone, O3 precursors, and other photochemically active
10     trace gases were measured upwind and downwind of New York City to characterize the O3
11     formation process and its dependence on NOX and VOCs.  During two flights, the wind was
12     south southwesterly and O3 levels reached 110 ppb.  On two other flights, the wind was from the
13     north-northwest and O3 levels were not as high.  When the G-l observed O3 around 110 ppb, the
14     NOx/NOy ratio measured at the surface was between 0.20 and 0.30,  indicating an aged plume.
15
16     AX2.4.1.8  Model Calculations and Aircraft Observations of Ozone Over Philadelphia
17           Regional-scale transport and local O3 production over Philadelphia was estimated using a
18     new meteorological-chemical model (Fast et al., 2002). Surface and airborne meteorological and
19     chemical measurements made during a 30-day period in July and August of 1999 as  part of the
20     Northeast Oxidant and Particulate Study were used to evaluate the model performance.  Both
21     research aircraft and ozonesondes, during the morning between 0900 and 1100 LST, measured
22     layers of O3 above the convective boundary layer. The model accounted for these layers through
23     upwind vertical mixing the previous day, subsequent horizontal transport aloft, and NO titration
24     of O3 within the stable boundary layer at night.  Entrainment of the  O3 aloft into the growing
25     convective boundary contributed to surface O3 concentrations. During the study period, most of
26     the O3 appeared to result from local emissions in the vicinity of Philadelphia and the Chesapeake
27     Bay area, but during high O3 episodes, up to 30 to 40% of the O3 was due to regional transport
28     from upwind sources.
29
        August 2005                            AX2-104     DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX2.4.1.9  The Two-Reservoir System
 2           Studies described above and aircraft observations made in August 2002 over the mid-
 3     Atlantic region show that a two-reservoir system illustrated schematically in Figure AX2-18 may
 4     realistically represent both the dynamics and photochemistry of severe, multiday haze and O3
 5     episodes over the eastern United States (Taubman et al., 2004).  The first reservoir is the PEL,
 6     where most precursor species are injected, and the second is the lower free troposphere (LFT),
 7     where photochemical processes are accelerated and removal via deposition is rare. Bubbles of
 8     air lifted from urban and industrial sources were rich in CO and SO2, but not O3, and contained
 9     greater numbers of externally mixed primary sulfate and black carbon (BC) particles.
10     Correlations among O3, air parcel altitude, particle size, and relative humidity suggest that
11     greater O3 concentrations and relatively larger particles are produced in the LFT and mix back
12     down into the PEL. Backward trajectories indicated source regions in the Midwest and
13     mid-Atlantic urban corridor, with southerly transport up the urban corridor augmented by the
14     Appalachian lee trough and nocturnal low-level jet (LLJ). This concept of two-reservoirs may
15     facilitate the numerical simulation of multiday events in the eastern United States. A relatively
16     small number of vertical layers will be required if accurate representation of the sub-gridscale
17     transport can be parameterized to represent the actual turbulent exchange of air between the PEL
18     and lower free troposphere.
19
20     AX2.5   METHODS USED TO  CALCULATE RELATIONS BETWEEN
21               OZONE AND ITS PRECURSORS
22           Atmospheric chemistry and transport models are the major tools used to calculate the
23     relations between O3, its precursors, and other oxidation products.  Other techniques, involving
24     statistical relations between O3 and other variables have also been used. Chemistry-transport
25     models  (CTM)  are driven by emissions inventories for O3 precursor compounds and by
26     meterological fields.  Emissions of precursor compounds can be divided into anthropogenic and
27     natural source categories. Natural sources can be further divided into biotic (vegetation,
28     microbes, animals) and abiotic (biomass burning, lightning) categories.  However, the distinction
29     between natural sources and anthropogenic  sources is often difficult to make as human activities
30     affect directly, or indirectly, emissions from what would have been considered natural sources
31     during the pre-industrial era.  Emissions from plants and animals used in agriculture are usually

       August  2005                            AX2-105     DRAFT-DO NOT QUOTE OR CITE

-------
                                         Two Reservoir Model
                     >
                                          QJ
                               NO
             SO2
O
              CO
                   0
        BC
                                                             0
                                                                  0
                                        > Actinic
                                         Flux
                                        a £1.9
                         .•' < Actinic
                           Flux
                          a a 1.9
                           0
VOCs    Primary
        Particles

      Figure AX2-18.  Conceptual two-reservoir model showing conditions in the PEL and
                       in the lower free troposphere during a multiday O3 episode.  The dividing
                       line, the depth of the mixed layer, is about 1000 m. Emissions occur in
                       the PEL, where small, unmixed black carbon, sulfate, and crustal
                       particles in the PM2 5 size range are also shown. Ozone concentrations as
                       well as potential temperature (0) and actinic flux are lower in the PEL
                       than in the lower free troposphere, while RH is higher. Larger, mixed
                       sulfate and carbonaceous particles (still in the PM2 5 size range) and more
                       O3 exist in the lower free troposphere.

      Source: Taubman et al. (2004).
1     referred to as anthropogenic. Wildfire emissions may be considered natural, except that forest
2     management practices may have led to the buildup of fuels on the forest floor, thereby altering
3     the frequency and severity of forest fires. Needed meteorological quantities such as winds and
4     temperatures are taken from operational analyses, reanalyses, or circulation models. In most
5     cases, these are off-line analyses, i.e., they are not modified by radiatively active species such as
6     O3 and particles generated by the model.
      August 2005
             AX2-106
             DRAFT-DO NOT QUOTE OR CITE

-------
 1           A brief overview of atmospheric chemistry-transport models is given in Section AX2.5.1.
 2      A discussion of emissions inventories of precursors that are used by these models is given in
 3      Section AX2.5.2. Uncertainties in emissions estimates have also been discussed in Air Quality
 4      Criteria for Particulate Matter (U.S. Environmental Protection Agency, 2000). So-called
 5      "observationally based models" which rely more heavily on observations of the concentrations
 6      of important species are discussed in Section AX2.5.3. Chemistry-transport model evaluation
 7      and an evaluation of the reliability of emissions inventories are presented in Section AX2.5.4.
 8
 9      AX2.5.1   Chemistry-Transport Models
10           Atmospheric chemistry-transport models (CTMs) are used to obtain better understanding
11      of the processes controlling the formation, transport, and destruction of O3 and other air
12      pollutants; to understand the relations between O3 concentrations and concentrations of its
13      precursors such as NOX and VOCs; and to understand relations among the concentration patterns
14      of O3 and other oxidants that may also exert health effects. Detailed examination of the
15      concentrations of short-lived species in a CTM can provide important insights into how O3 is
16      formed under certain conditions and can suggest likely avenues for data analysis and future
17      experiments and field campaigns. The dominant processes leading to the formation of O3 in a
18      particular time period, questions about whether NOX or VOCs were more important, the
19      influence of meteorology and of emissions from a particular geographic region, and the
20      transformation or formation  of other pollutants could be examined using a CTM.
21           CTMs are also used for determining control strategies for O3 precursors. However, this
22      application has met with varying degrees of success because of the highly nonlinear relations
23      between O3 and emissions of its precursors.  CTMs include mathematical descriptions of
24      atmospheric transport, emissions, the transfer of solar radiation through the atmosphere,
25      chemical reactions, and removal to the surface by turbulent motions and precipitation for
26      chemical species of interest.  Increasingly, the trend is for these processes to be broken down and
27      handled by other models  or sub-models, so a CTM will likely use emissions and meteorological
28      data from at least two other models.
29           There are two major formulations of CTMs in current use. In the first approach,
30      grid-based, or Eulerian, air quality models, the region to be modeled (the modeling domain) is
31      subdivided into a three-dimensional array of grid cells. Spatial derivatives in the species

        August 2005                            AX2-107      DRAFT-DO NOT QUOTE OR CITE

-------
 1      continuity equations are cast in finite-difference form over this grid, and a system of equations
 2      for the concentrations of all the chemical species in the model are solved numerically at each
 3      grid point. The modeling domain may be limited to a particular airshed or provide global
 4      coverage and extend through several major atmospheric layers.  Time dependent continuity
 5      (mass conservation) equations are solved for each species including terms for transport, chemical
 6      production and destruction, and emissions and deposition (if relevant), in each cell. Chemical
 7      processes are simulated with ordinary differential equations, and transport processes are
 8      simulated with partial differential equations. Because of a number of factors such as the
 9      different time scales inherent in different processes, the coupled, nonlinear nature of the
10      chemical process terms, and computer storage limitations, all of the terms in the equations are
11      not solved simultaneously in three dimensions. Instead, a technique known as operator splitting,
12      in which terms involving individual processes are solved sequentially, is used. In the second
13      application of CTMs, trajectory or Lagrangian models, a large number of hypothetical air parcels
14      are specified as following wind trajectories. In these models, the original system of partial
15      differential equations is transformed into a system of ordinary differential equations.
16           A less common approach is to use a hybrid Lagrangian/Eulerian model, in which certain
17      aspects of atmospheric chemistry and transport are treated with a Lagrangian approach and
18      others are treaded in a Eulerian manner (e.g., Stein et al., 2000). Both modeling approaches have
19      their advantages and disadvantages. The Eulerian approach is more general in that it includes
20      processes that mix air parcels and allows integrations to be carried out for long periods during
21      which individual air parcels lose  their identity.  There are, however, techniques for including the
22      effects of mixing in Lagrangian models such as FLEXPART (e.g., Zanis et al., 2003), ATTILA
23      (Reithmeir and Sausen, 2002), and  CLaMS (McKenna et al., 2002).
24           Major modeling efforts within the U.S. Environmental Protection Agency center on the
25      ModelsS/Community Modeling for Air Quality (CMAQ, Byun et al.,  1998) and the Multi Scale
26      Air Quality  Simulation Platform  (MAQSIP, Odman and Ingram, 1996) whose formulations are
27      based on the regional acid deposition model (RADM, Chang et al., 1987).  A number of other
28      modeling platforms using the Lagrangian and Eulerian frameworks have been reviewed in
29      AQCD  96.  CTMs currently in use  are summarized in the review by Russell and Dennis (2000).
30      The domains of MAQSIP and CMAQ  are flexible and can extend from several hundred km to
31      the hemispherical scale. In addition, both of these  classes of models allow the resolution of the

        August 2005                            AX2-108      DRAFT-DO NOT QUOTE OR CITE

-------
 1      calculations over specified areas to vary.  CMAQ and MAQSIP are both driven by the MM5
 2      mesoscale meteorological model (Seaman, 2000 and references therein), though both may be
 3      driven by other meteorological models (e.g., RAMS and Eta).  Simulations of regional O3
 4      episodes have been performed with a horizontal resolution of 4 km. In principle, calculations
 5      over limited domains can be accomplished to even finer scales. However, simulations at these
 6      higher resolutions require better parameterizations of meteorological processes such as boundary
 7      layer fluxes, deep convection and clouds (Seaman, 2000), and knowledge of emissions.
 8      Resolution at finer scales will likely be necessary to resolve smaller-scale features such as the
 9      urban heat island; sea, bay, and land breezes; and the nocturnal low-level jet.
10           Currently, the most common approach to setting up the horizontal domain is to nest a finer
11      grid within a larger domain of coarser resolution. However, a number of other strategies are
12      currently being developed, such as the stretched grid (e.g., Fox-Rabinowitz et al., 2002) and the
13      adaptive grid.  In a stretched grid, the  grid's resolution continuously varies throughout the
14      domain, thereby eliminating any potential problems with the sudden change from one resolution
15      to another at the boundary. One must be careful in using such a formulation, because certain
16      parameterizations that are valid on a relatively  coarse grid scale (such as convection, for
17      example) are not valid or should not be present on finer scales. Adaptive grids are not set at the
18      start of the simulation, but instead adapt to the  needs of the simulation as it evolves (e.g., Hansen
19      et al., 1994).  They have the advantage that, if the algorithm is properly set up, the resolution is
20      always sufficient to resolve the process at hand. However, they can be very slow if the situation
21      to be modeled is complex. Additionally, if one uses adaptive grids for  separate meteorological,
22      emissions, and photochemical models, there is  no reason a priori why the resolution of each grid
23      should match; and the gains realized from increased resolution in one model will be  wasted in
24      the transition to another model. The use of finer and finer horizontal resolution in the
25      photochemical model will necessitate  finer-scale inventories of land use and better knowledge of
26      the exact paths of roads,  locations of factories,  and, in general, better methods for locating
27      sources. The present practice of locating a source in the middle of a county or distributing its
28      emissions throughout a county if its location is unknown will likely not be adequate  in the future.
29           The vertical resolution of these models continues to improve as more layers are added to
30      capture atmospheric processes and structures.  This trend will likely continue because a model
31      with 25 vertical layers, for example, may have  layers that are 500 m thick at the top  of the

        August 2005                             AX2-109      DRAFT-DO NOT QUOTE OR CITE

-------
 1      planetary boundary layer.  Though the boundary layer height is generally determined through
 2      other methods, the chemistry in the model is necessarily confined by such layering schemes.
 3      Because the height of the boundary layer is of critical importance in simulations of air quality,
 4      improved resolution of the boundary layer height would likely improve air quality simulations.
 5      The difficulty of properly  establishing the boundary layer height is most pronounced when
 6      considering tropopause folding events, which are important in determining the chemistry of the
 7      background atmosphere.  In the vicinity of the tropopause, the vertical resolution of most any
 8      large scale model is quite unlikely to be able to capture such a feature. Additionally, any current
 9      model is likely to have trouble adequately resolving fine scale features such as the low-level jet.
10      Finally, models must be able to treat emissions, meteorology,  and photochemistry differently in
11      different areas. Emissions models are likely to need better resolution near the surface and
12      possibly near any tall stacks. Photochemical models, on the other  hand, may  need better
13      resolution away from the surface and be more interested in resolving the planetary boundary
14      layer height,  terrain differences, and other higher altitude features. Meteorological models share
15      some of the concerns of photochemical models, but are less likely  to need sufficient resolution to
16      adequately treat a process such as dry deposition beneath a stable nocturnal boundary layer.
17      Whether the  increased computational power necessary for such increases in resolution will be
18      ultimately justified by improved results in the meteorological  and subsequent photochemical
19      simulations remains to be  seen.
20           CTMs require time dependent, three-dimensional wind fields for the time period of
21      simulation.  The winds may be either generated by a model using initial fields alone or four
22      dimensional  data assimilation can be used to improve the model's  meteorological fields (i.e.,
23      model equations can be updated periodically [or "nudged"] to bring results into agreement with
24      observations).  Most modeling efforts have focused on simulations of several days duration (a
25      typical time scale for individual O3 episodes), but there have been  several attempts at modeling
26      longer periods. For example, Kasibhatla and Chameides (2000) simulated a four month period
27      from May to September of 1995 using MAQSIP.  The current trend appears to be toward
28      simulating longer time periods. This will impose additional strains on computational resources,
29      as most photochemical modeling until recently has been performed with an eye toward
30      simulating only summertime episodes of peak O3.  With the shift toward modeling an entire year
31      being driven by the desire to understand observations of periods of high wintertime PM (e.g.,

        August 2005                            AX2-110     DRAFT-DO NOT QUOTE OR CITE

-------
 1     Blanchard et al., 2002), models will be further challenged to simulate air quality under
 2     conditions for which they may not have been used previously.
 3           Chemical kinetics mechanisms (a set of chemical reactions) representing the important
 4     reactions that occur in the atmosphere are used in air quality models to estimate the net rate of
 5     formation of each pollutant simulated as a function of time. Chemical mechanisms that
 6     explicitly treat the chemical reactions of each individual reactive species are too lengthy and
 7     demanding of computer resources to be incorporated into three-dimensional atmospheric models.
 8     As an example, a master chemical mechanism includes approximately 10,500 reactions
 9     involving 3603 chemical species (Derwent et al., 2001 and references therein).  Instead,
10     "lumped" mechanisms, that group compounds of similar chemistry together, are used. The
11     chemical mechanisms used in existing photochemical O3 models contain significant uncertainties
12     that may limit the accuracy of their predictions; the accuracy of each of these mechanisms is also
13     limited by missing chemistry. Because of different approaches to the lumping of organic
14     compounds into surrogate groups, chemical mechanisms, can produce somewhat different results
15     under similar conditions.  The CB-IV chemical mechanism (Gery et al., 1989),  the RADM II
16     mechanism (Stockwell et al.,  1990), the SAPRC (e.g., Wang et al., 2000a; Wang et al., 2000b;
17     Carter, 1990) and the RACM mechanisms can be used in CMAQ. Jimenez et al. (2003)
18     provide brief descriptions of the features of the main mechanisms in use and they compared
19     concentrations of several key species predicted by seven chemical mechanisms in a box model
20     simulation over 24 h.  The average deviation from the average of all mechanism predictions for
21     O3 and NO over the daylight period was less than 20%, and 10% for NO2 for all mechanisms.
22     However, much larger deviations were found for HNO3, PAN, HO2, H2O2, C2H4 and C5H8
23     (isoprene).  An analysis for OH radicals was not presented. The large deviations shown for most
24     species imply differences between the calculated lifetimes  of atmospheric species and the
25     assignment of model simulations to either NOX limited or radical limited regimes between
26     mechanisms. Gross and Stockwell (2003) found small differences between mechanisms for
27     clean conditions with differences becoming more significant for polluted conditions, especially
28     for NO2 and organic peroxy radicals. They caution modelers to consider carefully the
29     mechanisms they are using.
30           As CTMs incorporate more processes and knowledge of aerosol- and gas-phase chemistry
31     improves, a "one atmosphere" approach is evolving. For example, CMAQ and PM-CAMx now

       August 2005                           AX2-111      DRAFT-DO NOT QUOTE OR CITE

-------
 1      incorporate some aerosol processes, and several attempts are currently underway to study
 2      feedbacks of chemistry on atmospheric dynamics using meteorological models, usually MM5
 3      (e.g., Grell et al., 2000; Liu et al., 2001; Lu et al., 1997; and Park et al., 2001). This coupling
 4      may be necessary to accurately simulate cases such as the heavy aerosol loading found in forest
 5      fire plumes (Lu et al., 1997 and Park et al., 2001).
 6           Spatial and temporal characterizations of anthropogenic and biogenic precursor emissions
 7      must be specified as inputs to a CTM.  Emissions inventories have been compiled on grids of
 8      varying resolution for many hydrocarbons, aldehydes, ketones, CO, NH3, and NOX. Emissions
 9      inventories for many species require the application of some algorithm for calculating the
10      dependence of emissions on physical variables such as temperature.  For many species,
11      information concerning the temporal variability of emissions is lacking, so long term (e.g.,
12      annual or O3-season) averages are used in short term, episodic simulations.  Annual emissions
13      estimates are often modified by the emissions model to produce emissions more characteristic of
14      the time of day and season. Significant errors in emissions can occur if an inappropriate time
15      dependence or a default profile is used. Additional complexity arises in model calculations
16      because different chemical mechanisms are based on different species, and inventories
17      constructed for use with another mechanism must be adjusted to reflect these differences. This
18      problem also complicates comparisons of the outputs of these models because one chemical
19      mechanism will necessarily produce species that are different from those in another and neither
20      output will necessarily agree with the measurements.
21           The effects of clouds on atmospheric chemistry are complex and introduce considerable
22      uncertainty into CTM calculations. Thunderstorm clouds are optically very thick and have
23      major effects on radiative fluxes and thus on photolysis rates. Madronich (1987) provided
24      modeling estimates of the effects of clouds of various optical depths on photolysis rates. In the
25      upper portion of a thunderstorm anvil, photolysis is likely to be enhanced (as much as a factor of
26      2 or more) due to multiple reflections off the ice crystals. In the lower portion of the cloud and
27      beneath the cloud, photolysis is substantially decreased. Thunderstorm updrafts, which contain
28      copious amounts of water, are regions where efficient scavenging of soluble species occurs
29      (Balkanski et al., 1993).  Direct field measurements of the amounts of specific trace gases
30      scavenged in observed storms are sparse.  Pickering et al. (2001) used a combination of model
31      estimates of soluble species that did not include wet scavenging and observations  of these

        August 2005                            AX2-112     DRAFT-DO NOT QUOTE OR CITE

-------
 1      species from the upper tropospheric outflow region of a major line of convection observed near
 2      Fiji.  Over 90% of the nitric acid and hydrogen peroxide in the outflow air appeared to have been
 3      removed by the storm. Walcek et al. (1990) included a parameterization of cloud-scale aqueous
 4      chemistry, scavenging, and vertical mixing in the regional scale, chemistry-transport model of
 5      Chang et al. (1987). The vertical distribution of cloud microphysical properties and the amount
 6      of subcloud-layer air lifted to each cloud layer were determined using a simple entrainment
 7      hypothesis (Walcek and Taylor, 1986). Vertically-integrated O3 formation rates over the
 8      northeastern United States were enhanced by -50% when the in-cloud vertical motions were
 9      included in the model.
10           In addition to wet deposition, dry deposition (the removal of chemical species from the
11      atmosphere by interaction with ground-level surfaces) is an important removal process for
12      pollutants on both urban and regional scales and must be included in CTMs.  The general
13      approach used in most models is the three-resistance method, in which where dry deposition is
14      parameterized with a deposition velocity, which is represented as vd = (ra + rb + rc)-1 where ra, rb,
15      and rc represent the resistance due to atmospheric turbulence, transport in the fluid sublayer very
16      near the elements of surface such as leaves or soil, and the resistance to uptake of the surface
17      itself. This approach works for a range of substances although it is inappropriate for species
18      with  substantial emissions from the surface or for species whose deposition to the surface
19      depends on its concentration at the surface itself. The approach is also modified somewhat for
20      aerosols: the terms rb and rc are replaced with a surface  deposition velocity to account for
21      gravitational settling. In their review, Wesley and Hicks (2000) point out several shortcomings
22      of current knowledge of dry deposition.  Among those shortcomings are difficulties in
23      representing dry deposition over varying terrain where horizontal advection plays a significant
24      role in determining the magnitude of ra and difficulties in adequately determining a deposition
25      velocity for extremely stable conditions such as those occurring at night (e.g., Mahrt, 1998).
26      Under the best of conditions, when a model is exercised over a relatively small area where dry
27      deposition measurements have been made, models still commonly show uncertainties at least as
28      large as ± 30% (e.g., Massman  et al., 1994; Brook et al., 1996; Padro, 1996). Wesley and
29      Hicks (2000) state that an important result of these  comparisons is that the current level of
30      sophistication of most dry deposition models is relatively low and relies heavily on empirical
31      data.   Still larger uncertainties exist when the surface features are not well  known or when the

        August 2005                             AX2-113      DRAFT-DO NOT QUOTE OR CITE

-------
 1      surface comprises a patchwork of different surface types, as is common in the eastern United
 2      States.
 3           The initial conditions, i.e., the concentration fields of all species computed by a model, and
 4      the boundary conditions, i.e., the concentrations of species along the horizontal and upper
 5      boundaries of the model domain throughout the simulation must be specified at the beginning of
 6      the simulation. It would be best to specify initial and boundary  conditions according to
 7      observations.  However, data for vertical profiles of most species of interest are sparse.
 8      Ozonesonde data have been used to specify O3 fields, but the  initial and boundary values of
 9      many other species are often set equal to zero because of a lack  of observations. Further,
10      ozonesondes are thought to be subject to errors in measurement and differences arising from
11      improper corrections for pump efficiency and the solutions  used (e.g., Hilsenrath et al., 1986;
12      Johnson et al., 2002). The results of model simulations over larger, preferably global, domains
13      can also be used. As may be expected, the influence of boundary conditions depends on the
14      lifetime of the species under consideration and the time scales for transport from the boundaries
15      to the interior of the model domain (Liu et al., 2001).
16           Each of the model components described above has an associated uncertainty, and the
17      relative importance of these uncertainties varies with the modeling application. The largest
18      errors in photochemical modeling are still thought to arise from the meteorological and
19      emissions inputs to the model (Russell and Dennis, 2000). Within the model itself, horizontal
20      advection algorithms are still thought to be significant source  of uncertainty (e.g., Chock and
21      Winkler, 1994) though more recently those errors are thought to have been reduced (e.g., Odman
22      et al., 1996). There are also indications that problems with  mass conservation continue to be
23      present in photochemical and meteorological models (e.g., Odman and Russell, 1999); these can
24      result in significant simulation errors. Uncertainties in meteorological variables and emissions
25      can be large enough that they would lead one to make the wrong decision when considering
26      control strategies (e.g., Russell and Dennis, 2000; Sillman et al., 1995). The effects of errors in
27      initial conditions can be minimized by including several days  "spin-up" time in a simulation to
28      allow species to come to chemical equilibrium with each other before the simulation of the
29      period of interest begins.
30           While the effects of poorly specified boundary conditions  propagate through the model's
31      domain, the effects of these errors remain undetermined.  Many regional models specify constant

        August 2005                            AX2-114      DRAFT-DO NOT QUOTE OR CITE

-------
 1      O3 profiles (e.g., 35 ppb) at their lateral and upper boundaries; ozonesonde data, however,
 2      indicate that the mixing ratio of O3 increases vertically in the troposphere (to over 100 ppb at the
 3      tropopause) and into the stratosphere (e.g., Newchurch et al., 2003). The practice of using
 4      constant O3 profiles strongly reduces the potential effects of vertical mixing of O3 from above
 5      the planetary boundary layer (via mechanisms outlined in Section AX2.3) on surface O3 levels.
 6      The use of an O3 climatology (e.g., Fortuin and Kelder, 1998) might reduce the errors that would
 7      otherwise be incurred.  Because many meteorological processes occur on spatial scales which
 8      are smaller than the grid spacing (either horizontally or vertically) and thus are not calculated
 9      explicitly, parameterizations of these processes must be used and these introduce additional
10      uncertainty.
11           Uncertainty also arises in modeling the chemistry of O3 formation because it is highly
12      nonlinear with respect to NOX concentrations.  Thus, the volume of the grid cell into which
13      emissions are injected is important because the nature of O3 chemistry (i.e., O3 production or
14      titration) depends in a complicated way on the concentrations of the precursors and the OH
15      radical. The use of ever-finer grid spacing allows regions of O3 titration to be more clearly
16      separated from regions of O3 production.  The use of grid spacing fine enough to resolve the
17      chemistry in individual power-plant plumes is too demanding of computer resources for this to
18      be attempted in most simulations.  Instead, parameterizations of the effects of subgrid scale
19      processes such as these must be developed; otherwise serious errors can result if emissions are
20      allowed to mix through  an excessively large grid volume before the chemistry step in a model
21      calculation is performed. In light  of the significant differences between atmospheric chemistry
22      taking place inside and outside of a power plant plume (e.g., Ryerson et al.,  1998 and Sillman,
23      2000), inclusion of a separate, meteorological  module for treating large, tight plumes is
24      necessary.  Because the photochemistry of O3  and many other atmospheric species is nonlinear,
25      emissions correctly modeled in a tight plume may be incorrectly modeled in a more dilute
26      plume. Fortunately, it appears that the chemical mechanism used to follow a plume's
27      development need not be as detailed as that used to simulate the rest of the domain, as the
28      inorganic reactions  are the most important in the plume (e.g., Kumar and Russell, 1996). The
29      need to include explicitly plume-in-grid chemistry disappears if one uses the adaptive grid
30      approach mentioned previously, though such grids are more computationally intensive. The
        August 2005                            AX2-115     DRAFT-DO NOT QUOTE OR CITE

-------
 1      differences in simulations are significant because they can lead to significant differences in the
 2      calculated sensitivity of O3 to its precursors (e.g., Sillman et al., 1995).
 3           Because the chemical production and loss terms in the continuity equations for individual
 4      species are coupled, the chemical calculations must be performed iteratively until calculated
 5      concentrations converge to within some preset criterion. The number of iterations and the
 6      convergence criteria chosen also can introduce error.
 7           The importance of global transport of O3 and its contribution to regional O3 levels in the
 8      United States is slowly becoming apparent. There are presently on the order of 20
 9      three-dimensional global models that have been developed by various groups to address
10      problems in tropospheric chemistry.  These models resolve synoptic meteorology, O3-NOX-CO-
11      hydrocarbon photochemistry, wet and dry deposition, and parameterize sub-grid scale vertical
12      mixing such as convection.  Global models have proven useful  for testing and advancing
13      scientific understanding beyond what is possible with observations alone.  For example, they can
14      calculate quantities of interest that we do not have the resources to measure directly, such as
15      export of pollution from one continent to the global atmosphere or the response of the
16      atmosphere to future perturbations to anthropogenic emissions.
17           The finest horizontal resolution at which global simulations are typically conducted is
18      -200 km2 although rapid advances in computing power continuously change what calculations
19      are feasible.  The next generation of models will consist of simulations that link multiple
20      horizontal resolutions from the global to the local scale. Finer resolution will only improve
21      scientific understanding to the extent that the governing processes are more accurately described
22      at that scale. Consequently there is a critical need for observations at the appropriate scales to
23      evaluate the scientific understanding represented by the models.
24           Observations of specific chemical species have been useful for testing transport schemes.
25      Radon-222 simulations in sixteen global models have been evaluated with observations to show
26      that vertical mixing is captured to within the constraints offered by the mean observed
27      concentrations (Jacob et al., 1997). Tracers such as cosmogenic 7Be and terrigenic 210Pb have
28      been used to test and constrain model transport and wet deposition (e.g., Liu et al., 2001).
29           Other chemical species obtained from various platforms (surface measurements, aircraft,
30      satellites) are useful for evaluating the simulation of chemical and dynamical processing in
31      global models. For example, Emmons et al. (2000) compiled available measurements of

        August 2005                            AX2-116      DRAFT-DO NOT QUOTE OR CITE

-------
 1      12 species relevant to O3 photochemistry from a number of aircraft campaigns in different
 2      regions of the world and used this data composite to evaluate two global models. They
 3      concluded that one model (MOZART) suffered from weak convection and an underestimate of
 4      nitrogen oxide emissions from biomass burning, while another model (IMAGES) transported too
 5      much O3 from the stratosphere to the troposphere (Emmons et al., 2000).  The global coverage
 6      available from satellite observations offers new information for testing models.  Recent efforts
 7      are using satellite observations to evaluate the emission inventories of O3 precursors that are
 8      included in global models; such observations should help to constrain the highly uncertain
 9      natural emissions of isoprene and nitrogen oxides (e.g., Palmer et al., 2003; Martin et al., 2003).
10          A comparison of numerous global chemistry-transport models developed by groups around
11      the world was included in Section 4.4 of the recent report of the Intergovernmental Panel on
12      Climate Change (Prather and Ehhalt, 2001). In that report, monthly mean O3 (O3) and carbon
13      monoxide (CO) simulated by the various models was evaluated with O3 observations from global
14      ozonesonde stations at 700, 500, and 300 hPa  and with surface CO measurements from
15      17 selected NOAA/CMDL sites. The relevant figures (Figures AX2-4-10 and AX2-4-11) are
16      reproduced here (as Figures AX2-19 for O3 and AX2-20 for CO) along with the references in
17      their Table AX2-10 (as Table AX2-4). Overall, the models capture the general features of the O3
18      and CO seasonal cycles but meet with varying levels of success at matching the observed
19      concentrations and the amplitude of the observed seasonal cycle. For O3, the models show less
20      disagreement in the lower troposphere than in the upper troposphere, reflecting the difficulty of
21      representing the exchange between the stratosphere and troposphere  and the loose constraints on
22      the net O3 flux that are provided by observations.
23          An evaluation of five global models with data from the Measurement of Ozone and Water
24      Vapor by Airbus In-Service Aircraft (MOZAIC) project over New York City and Miami
25      indicates that the models tend to underestimate the summer maximum in the middle and lower
26      troposphere over northern mid-latitude cities such as New York City and to underestimate the
27      variability over coastal cities such as Miami which are strongly influenced by both polluted
28      continental and clean marine air masses (Law et al., 2000). Local maxima and minima are
29      difficult to reproduce with global models because processes are averaged over an entire model
30      grid cell.  Much of the spatial and temporal variability in surface O3 over the United States is
31      modulated by synoptic meteorology (e.g., Logan, 1989; Eder et al., 1993; Vukovich, 1995, 1997;

        August 2005                           AX2-117      DRAFT-DO NOT QUOTE OR CITE

-------
       J FMAM J J ASON D
 J FMAM J J ASON D
J FMAM J J ASON D
       J FMAMJ JASOND
 J FMAMJ JASOND
J FMAMJ JASOND
       J FMAM J JASOND
 J FMAM J JASOND
J FMAM J JASOND
               	 observ.
               	GISS
                    HGEO
HGIS   	 MOZ1       UIO    -*- UCAM
IASB       MOZ2   	 UKMO       MPIC
KNMI       UCI1	ULAQ
Figure AX2-19.  Seasonal variability in O3 concentrations observed at a number of
                pressure surfaces at six ozonesonde sites and the predictions of 13 global
                scale chemistry-transport models.

Source:  IPCC Third Assessment Report (2001).
August 2005
     AX2-118     DRAFT-DO NOT QUOTE OR CITE

-------
             J FMAM J J ASON D
J FMAM J J ASON D
         J FMAM J  J ASON D
              J FMAM J J ASON D
J FMAM J J ASON D
         J FMAM J  J ASON D
                                                         observ.
                                                         GISS
                                                         HGEO
                                                         IASB
                                                         KNMI
                                                         MOZ1
                               MOZ2
                           	ucn
                               UIO
                           	 UKMO
                           	ULAQ
                           -*- UCAM
                           	 MPIC
             J FMAM J J ASON D
      Figure AX2-20.   Seasonal variability in O3 concentrations observed at a number of
                       pressure surfaces at six ozonesonde sites and the predictions of 13 global
                       scale chemistry-transport models.

      Source: IPCC Third Assessment Report (2001).
1     Cooper and Moody, 2000) which is resolved in the current generation of global models.

2     For example, an empirical orthogonal function analysis on observed and simulated fields over

3     the eastern United States in summer has shown that a 2° x 2.5° horizontal resolution global

4     model (GEOS-CHEM) captures the synoptic-scale processes that control much of the observed

5     variability (Fiore et al., 2003).  Further evaluation of the same model showed that it can also
      August 2005
    AX2-119
DRAFT-DO NOT QUOTE OR CITE

-------
           Table AX2-4. Chemistry-Transport Models (CTM) Contributing to the Oxcomp
             Evaluation of Predicting Tropospheric O3 and OH (Prather and Ehhalt, 2001)
CTM
GISS
HGEO
HGIS
IASB
KNMI
MOZ1
MOZ2
MPIC
UCI
UIO
UIO2
UKMO
ULAQ
UCAM
Institute
GISS
Harvard U.
Harvard U.
lAS/Belg.
KNMI/Utrecht
NCAR/CNRS
NCAR
MPI/Chem
UC Irvine
U. Oslo
U. Oslo
UK Met Office
U. L. Aquila
U. Cambridge
Contributing Authors
Shindell/Grenfell
Bey/Jacob
Mickley/Jacob
Mulller
van Weele
Hauglustaine/Brasseur
Horowitz/Brasseur
Kuhlmann/Lawrence
Wild
Berntsen
Sundet
Stevenson
Pitari
Plantevin/Johnson
(TOMCAT)
References
Hansenetal. (1997)
Beyetal. (200 la)
Mickley etal. (1999)
Mtiller and Brasseur
(1995, 1999)
Jeuken etal. (1999),
Houweling et al. (2000)
Brasseur etal. (1998),
Hauglustaine et al. (1998)
Brasseur etal. (1998),
Hauglustaine et al.(1998)
Crutzen etal. (1999),
Lawrence et al. (1999)
Hannegan etal. (1998),
Wild and Prather (2000)
Berntsen and Isaksen (1997),
Fuglestvedt etal. (1999)
Sundet (1997)
Collins etal. (1997),
Johnson etal. (1999)
Pitari etal. (1997)
Law etal. (1998, 2000)
1     capture many of the salient features of the observed distributions of O3 as well as its precursors
2     in surface air over the United States in summer, including formaldehyde concentrations and
3     correlations between O3 and the oxidation products of nitrogen oxides (O3: NOy-NOx), all of
4     which indicate a reasonable photochemical simulation (Fiore et al., 2002).
5           A significant amount of progress in evaluating the performance of three-dimensional
6     global models with surface, aircraft, and satellite data has been made in recent years.
7     Disagreement among model simulations mainly stems from differences in the driving
8     meteorology and emissions. The largest discrepancies amongst models and between models and
      August 2005
AX2-120
DRAFT-DO NOT QUOTE OR CITE

-------
 1      observations occur in the upper troposphere and likely reflect uncertainties in exchange between
 2      the stratosphere and troposphere and photochemical processes there; the models agree better
 3      with observations closer to the surface. Synoptic-scale meteorology is resolved in these models,
 4      enabling them to simulate much of the observed variability in pollutants in the lower

 5      troposphere.
 6

 7      AX2.5.2  Emissions of Ozone Precursors

 8           Estimated annual emissions of nitrogen oxides, VOCs, CO, and NH3 for 1999 (U.S.
 9      Environmental Protection Agency, 2001) are shown in Tables AX2-5, AX2-6, AX2-7,  and

10      AX2-8.  Methods for estimating emissions of criteria pollutants, quality assurance procedures
11      and examples of emissions calculated by using data are given in U.S. Environmental Protection
12      Agency (1999).
                  Table AX2-5. Emissions of Nitrogen Oxides by Various Sources in the
                                          United States in 1999
         Source
Emissions1
 (1012 g/y)
      Notes
         On-road vehicle exhaust         7.8

         Non-road vehicle exhaust       5


         Fossil fuel combustion          9.1


         Industrial Processes           0.76



         Biomass burning              0.35


         Waste disposal                0.053

         Natural sources2                3.1

         Total                         26
             Gasoline (58%) and diesel (42%) vehicles.

             Diesel (49%) and gasoline (3%) vehicles; railroads (22%);
             marine vessels (18%); other sources (8%).

             Electric utilities (57%); industry (31%); commercial,
             institutional and residential combustion (12%).

             Mineral products (43%); petrochemical products (17%);
             chemical mfg. (16%); metal processing (11%); misc.
             industries (12%).

             Residential wood burning (11%);
             open burning (8%); wildfires (81%).

             Non-biomass incineration.

             Lightning (50%); soils(50%).
         'Emissions are expressed in terms of NO2.
         2Estimated on the basis of data given in Guenther et al. (2000).

         Source: U.S. Environmental Protection Agency (2001).
        August 2005
              AX2-121
DRAFT-DO NOT QUOTE OR CITE

-------
    Table AX2-6. Emissions of Volatile Organic Compounds by Various Sources in the
                                    United States in 1999
Source
On-road vehicles
Non-road vehicles
Fossil fuel combustion
Chemical industrial
processes
Petroleum industrial
processes
Other industrial
processes
Emissions
(1012 g/y)
4.8
2.9
0.27
0.36
0.39

0.48
Notes
Exhaust and evaporative losses from gasoline (95%)
diesel (5%) vehicles.
Exhaust and evaporative losses from gasoline (80%)
diesel (12%) vehicles; aircraft and other sources (8%

and
and
).
Electrical utilities; industrial, commercial, institutional,
and residential sources.
Mfg. of organic chemicals, polymers and resins, and
products.
Oil and gas production (64%); refining (36%).

Metal processing (15%); wood processing (32%);
agricultural product processing (21%); misc.
misc.



 Solvent volatilization
          processes (18%).

4.4       Surface coatings (44%); other industrial uses (20%); non-
          industrial uses (e.g., pesticide application, consumer
          solvents) (36%).
Storage and transport
of volatile compounds
Biomass burning
Waste disposal
Biogenic sources1
Total
1.1
1.2
0.53
4.4
21
Evaporative losses from petroleum products and other
organic compounds.
Residential wood combustion (37%); open burning
(22%); agricultural burning (22%); wildfires (19%).
Residential burning (63%); waste water (23%); landfills
(6%); non-biomass incineration (8%).
Approximately 98% emitted by vegetation. (Isoprene
[35%], monoterpenes [25%], and all other reactive and
non-reactive compounds [40%]).

 'Estimated on the basis of data given in Guenther et al. (2000).

 Source: U.S. Environmental Protection Agency (2001).
August 2005
           AX2-122       DRAFT-DO NOT QUOTE OR CITE

-------
          Table AX2-7. Emissions of Ammonia by Various Sources in the United States in 1999
         Source
Emissions
 (1012 g/y)    Notes
         Exhaust from on-road
         and non-road engines and
         vehicles
         Fossil fuel combustion
         Industry
   0.25      Exhaust from on-road (96%) and non-road (4%) vehicles.
  0.044     Combustion by electric utilities, industry, commerce,
            institutions, residences.
   0.18      Chemical manufacturing (67%); petroleum refining (9%);
            other industries (25%).
Agriculture
Waste disposal and
recycling
Natural sources
Total
3.9
0.08
0.032
4.5
Livestock (82%); fertilizer application (18%).
Wastewater treatment (99%).
Unmanaged soils;

wild animals.

         Source: U.S. Environmental Protection Agency (2001).
 1           Emissions of nitrogen oxides associated with combustion arise from contributions from
 2     both fuel nitrogen and atmospheric nitrogen.  Sawyer et al. (2000) have reviewed the factors
 3     associated with NOX emissions by mobile sources. Estimates of NOX emissions from mobile
 4     sources are generally regarded as fairly reliable although further work is needed to clarify this
 5     point (Sawyer et al., 2000). Both nitrifying and denitrifying bacteria in the soil can produce
 6     NOX, mainly in the form of NO.  Emission rates depend mainly on fertilization levels and soil
 7     temperature. About 60% of the total NOX emitted by soils occurs in the central corn belt of the
 8     United States. The oxidation of NH3 emitted mainly by livestock and soils, leads to the
 9     formation of NO.  Estimates of emissions from natural sources are less certain than those from
10     anthropogenic sources.
11           Natural sources of oxides of nitrogen include lightning, oceans, and soil. Of these, as
12     reviewed in AQCD 96,  only soil emissions appear to have the potential to impact surface O3 over
13     the U.S.  On a global scale, the contribution of soil emissions to the oxidized nitrogen budget is
14     on the order of 10% (van Aardenne et al., 2001; Finlayson-Pitts and Pitts, 2000; Seinfeld and
15     Pandis, 1998), but attempts to quantify emissions of NOX from fertilized fields show great
        August 2005
              AX2-123
DRAFT-DO NOT QUOTE OR CITE

-------
                Table AX2-8.  Emissions of Carbon Monoxide by Various Sources in the
                                          United States in 1999
        Source
Emissions
 (1012 g/y)
Notes
        On-road vehicle exhaust
        Non-road engines and
         vehicle exhaust
        Fossil fuel combustion
        Industrial Processes
        Biomass burning
    50
    25
   3.7
    16
Gasoline-fueled light-duty cars (54%) and trucks (32%),
heavy-duty trucks (9%); diesel vehicles (5%);
motorcycles (0.4%).

Gasoline-fueled (lawn and garden [44%], light
commercial [17%], recreational [14%], logging [4%],
industry and construction [6%, other [1%]); diesel-fueled
(5%); aircraft (4%); other (5%).

Electric utilities (22%); industry (58%); commercial,
institutional and residential combustion (20%).

Metal processing (45%); chemical mfg. (29%);
petrochemical production; (10%); mineral products (5%);
wood products (10%); misc. industries (1%).
Residential wood burning (21%); open burning (21%);
agricultural burning (41%); wildfires (17%).
Waste disposal
Other
Biogenic emissions1
Total
0.42
0.19
4.7+
702+
Non-biomass incineration.
Structural fires (45%); storage and transport (38%);
misc. sources (17%).
Primary emissions from vegetation and soils;
secondary formation (?).
        'Estimated on the basis of data given in Guenther et al. (2000).

        Source: U.S. Environmental Protection Agency (2001).
1      variability. Soil NO emissions can be estimated from the fraction of the applied fertilizer

2      nitrogen emitted as NOX, but the flux varies strongly with land use and temperature. The fraction

3      nitrogen. Estimated globally averaged fractional applied nitrogen loss as NO varies from 0.3%

4      (Skiba et al.,  1997) to 2.5% (Yienger and Levy, 1995).  Variability within biomes to which

5      fertilizer is applied, such as shortgrass versus tallgrass prairie, accounts for a factor of

6      three in uncertainty (Williams et al., 1992; Yienger and Levy, 1995; Davidson and Kingerlee,
7      1997).
       August 2005
              AX2-124
                DRAFT-DO NOT QUOTE OR CITE

-------
 1           The local contribution can be much greater than the global average, particularly in summer
 2      especially where corn is grown extensively. Williams et al. (1992) estimated that contributions
 3      from soils in Illinois contribute about 26% of the emissions from industrial and commercial
 4      processes in that State. In Iowa, Kansas, Minnesota, Nebraska, and South Dakota soil emissions
 5      may dominate. Conversion of ammonium to nitrate (nitrification) in aerobic soils appears to be
 6      the dominant pathway to NO. The mass and chemical form of nitrogen (reduced or oxidized)
 7      applied to soils, the vegetative cover, temperature, soil moisture, and agricultural practices such
 8      as tillage all influence the amount of fertilizer nitrogen released as NO.
 9           As pointed out in the previous AQCD for O3, emissions of NO from soils peak in summer
10      when O3 formation is at a maximum. A recent NRC report outlined the role of agricultural in
11      emissions of air pollutants including NO and NH3 (NRC,  2002). That report recommends
12      immediate implementation of best management practices  to control these emissions, and further
13      research to quantify the magnitude of emissions and the impact of agriculture on air quality.
14      Civerolo and Dickerson (1998) report that use of the no-till cultivation technique on a fertilized
15      cornfield in Maryland reduced NO emissions by a factor of seven.
16           Annual global production of NO by lightning is the  most uncertain source of reactive
17      nitrogen.  In the last decade literature values of the production rate range from 2 to 20 Tg-N per
18      year. However, the most likely range is from 3  to 8 Tg-N per year, because the majority of the
19      recent estimates fall in this range. The large uncertainty stems from several factors: (1) a large
20      range of NO  production rates per flash (as much as two orders of magnitude); (2) the open
21      question of whether cloud-to-ground (CG) flashes and intracloud flashes (1C) produce
22      substantially different amounts of NO; (3) the global flash rate; and (4) the ratio of the number of
23      1C flashes to the number of CG flashes. Estimates of the  amount of NO produced per flash have
24      been made based on theoretical considerations (e.g., Price et al., 1997), laboratory experiments
25      (e.g., Wang et al.,  1998); field experiments (e.g., Stith et al., 1999; Huntrieser et al., 2002), and
26      through a combination of cloud-resolving model simulations, observed lightning flash rates, and
27      anvil measurements of NO (e.g., DeCaria et al., 2000).  The latter method was also used by
28      Pickering et al. (1998), who showed that only -5% to 20% of the total NO production by
29      lightning in a given storms exists in the boundary layer at the end of a thunderstorm.  Therefore,
30      the direct contribution to boundary layer O3 production by lightning NO is thought to be small.
31      However, lightning NO production can contribute substantially to  O3 production in the middle

        August 2005                            AX2-125     DRAFT-DO NOT QUOTE OR CITE

-------
 1      and upper troposphere. DeCaria et al. (2000) estimated that up to 7 ppbv of O3 were produced in
 2      the upper troposphere in the first 24 hours following a Colorado thunderstorm due to the
 3      injection of lightning NO.  A major uncertainty in mesoscale and global chemical transport
 4      models is the parameterization of lightning flash rates. Model variables such as cloud top height,
 5      convective precipitation rate, and upward cloud mass flux have been used to estimate flash rates.
 6      Allen and Pickering (2002) have evaluated these methods against observed flash rates from
 7      satellite, and examined the effects on O3 production using each method.
 8          Literally tens of thousands of organic compounds have been identified in plant tissues.
 9      However, most of these compounds either have sufficiently low volatility or are constrained so
10      that they are not emitted in significant quantities.  Less than 40 compounds have been identified
11      by Guenther et al. (2000) as being emitted in large enough quantities to affect atmospheric
12      composition. These compounds include terpenoid compounds (isoprene, 2-methyl-3-buten-2-ol,
13      monoterpenes), compounds in the hexanal family, alkenes, aldehydes, organic acids, alcohols,
14      ketones and alkanes.  As can be seen from Table AX2-6, the major species emitted by plants are
15      isoprene (35%),  19 other terpenoid compounds (25%) and 17 non-terpenoid compounds (40%)
16      (Guenther et al., 2000). Of the latter, methanol contributes 12% of total emissions.
17          Because isoprene has been identified as the most abundant of biogenic VOCs (Guenther
18      et al., 1995, 2000; Geron et al., 1994), it has been the focus of air quality model analyses in
19      many published  studies (Roselle, 1994; Sillman et al., 1995).  The original Biogenic Emission
20      Inventory System (BEIS) of Pierce et al. (1991) used a branch-level isoprene emission factor of
21      14.7 jig (g-foliar dry mass)"1 h"1 for high isoprene emitting species (e.g.,  oaks, or North
22      American Quercus species). When considering self-shading of foliage within branch enclosures,
23      this is roughly equivalent to a leaf level emission rate of 20 to 30 |ig-C (g-foliar dry mass)"1 h"1
24      (Guenther at al,  1995). Geron et al (1994) reviewed studies between 1990 and 1994 and found
25      that a much higher leaf-level rate of 70 |ig-C (g-foliar dry mass)"1 h"1 + 50% was more realistic,
26      and this rate was used in BEIS2 for high isoprene emitting tree species. BEIS3 (Guenther  et al.,
27      2000) applied similar  emission factors at tree species levels (Geron et al 2000a, 2001) and  more
28      recent canopy environment models to estimate isoprene fluxes.
29          The results from several studies of isoprene emission measurements made at leaf, branch,
30      tree, forest stand, and  landscape levels have been used to test the accuracy of BEIS2 and BEIS3.
31      These comparisons are documented in Geron et al. (1997) and Guenther  et al. (2000).  The

        August 2005                           AX2-126      DRAFT-DO  NOT QUOTE OR CITE

-------
 1     results of these studies support the higher emission factors used in BEIS2 and BEIS3. Typically,
 2     leaf emission factors (normalized to standard conditions of PAR = 1000 jimol nT2 and leaf
 3     temperature of 30 °C) measured at the top of tree canopies equal or exceed those used in
 4     BEIS2/3, while those in more shaded portions of the canopy tend to be lower than those assumed
 5     in the models, likely due to differences in developmental environments of leaves within the
 6     canopy (Monson et al., 1994; Sharkey et al., 1996; Harley et al., 1996; Geron  et al., 2000b).
 7     Uncertainty in isoprene emissions due to variability in forest composition and leaf area remain in
 8     BVOC emission models and inventories.  Seasonality and moisture stress also impact isoprene
 9     emission, but algorithms to simulate these effects are currently fairly crude (Guenther et al,
10     2000). The bulk of biogenic emissions occur during the summer, because of their dependence
11     on temperature and incident sunlight.  Biogenic emissions are also higher in southern states than
12     in northern states for these reasons.  The uncertainty associated with natural emissions ranges
13     from about 50% for isoprene under midday summer conditions to about a factor often for other
14     compounds (Guenther et al., 2000).  In assessing the relative importance of these compounds, it
15     should be borne in mind that the oxidation of many of the classes of compounds result in the
16     formation of secondary organic aerosol and that many of the intermediate products may be
17     sufficiently long lived to affect O3 formation in areas far removed from where they were emitted.
18     The oxidation of isoprene can also contribute about 10% of the source of CO (U.S.
19     Environmental Protection Agency, 2000). Direct emissions of CO by vegetation is  of much
20     smaller importance.  Soil microbes both emit and take up atmospheric CO, however, soil
21     microbial activity appears to represent a net sink for CO.
22          Emissions from biomass burning depend strongly on the stage of combustion.  Smoldering
23     combustion, especially involving  forest ecosystems favors the production of CH4, NMHC and
24     CO at the expense of CO2, whereas active combustion produces more CO2 relative to the  other
25     compounds mentioned above. Typical emissions ratios (defined as moles of compound per
26     moles of emitted CO2 expressed as a percentage) range from 6 to 14% for CO, 0.6 to 1.6% for
27     CH4, and 0.3 to 1.1% for NMHCs (Andreae, 1991).  Most NMHC emissions are due to
28     emissions of lighter compounds, containing 2  or 3 carbon atoms.
29
       August 2005                           AX2-127     DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX2.5.3  Observationally-Based Models
 2           As an alternative to chemistry-transport models, observationally-based methods (OEMs),
 3      which seek to infer O3-precursor relations by relying more heavily on ambient measurements,
 4      can be used.  Observationally-based methods are intuitively attractive because they provide an
 5      estimate of the O3-precursor relationship based directly on observations of the precursors.  These
 6      methods rely on observations as much as possible to avoid many of the uncertainties associated
 7      with chemistry /transport models (e.g., emission inventories and meteorological processes).
 8      However, these methods have large uncertainties with regards to photochemistry.  As originally
 9      conceived, the observation-based approaches were intended to provide an alternative method for
10      evaluating critical issues associated with urban O3 formation. The proposed OEMs include
11      calculations driven by ambient measurements (Chameides et al., 1992; Cardelino et al., 1995)
12      and proposed "rules of thumb" that seek to show whether O3 is primarily sensitive to NOX  or to
13      VOC concentrations (Sillman,  1995; Chang  et al., 1997; Tonnesen and Dennis, 2000a,b;
14      Blanchard et al., 1999; Blanchard, 2000).  These methods are controversial when used as
15      "stand-alone" rules, because significant uncertainties and possible errors have been identified for
16      all the methods (Chameides et al., 1988, Lu and Chang, 1998, Sillman and He, 2002; Blanchard
17      and Stockenius, 2001). Methods such as these are most promising for use in combination  with
18      chemistry/transport models principally for evaluating the accuracy of model predictions.
19           Recent results (Tonnesen and Dennis, 2000a; Kleinman et al.,  1997; 2000, 2001;
20      Kleinman, 2000) suggest that ambient VOC  and NOX data can be used to identify the
21      instantaneous production rate for O3 and how the production rate varies with concentrations of
22      NOX and VOCs.  The instantaneous production rate for O3 is only one of the factors that affect
23      the total O3 concentration, because O3 concentrations result from photochemistry and transport
24      over time periods ranging from several hours to several days in regional pollution events.  Ozone
25      concentrations can be affected by distant emissions and by photochemical conditions at upwind
26      locations, rather than instantaneous production at the site. Despite this limitation, significant
27      information can be obtained by interpreting ambient NOX and VOC measurements. Kleinman
28      et al.  (1997, 2000,  2001) and Tonnesen  and Dennis (2000a) both derived simple expressions that
29      relate the NOX-VOC sensitivity of instantaneous O3 production to ambient VOC and NOX.  These
30      expressions usually involve summed VOC weighted by reactivity.
        August 2005                            AX2-128      DRAFT-DO NOT QUOTE OR CITE

-------
 1           Cardelino et al. (1995, 2000) developed a method that seeks to identify O3-NOX-VOC
 2      sensitivity based on ambient NOX and VOC data. Their method involves an area-wide sum of
 3      instantaneous production rates over an ensemble of measurement sites, which serve to represent
 4      the photochemical conditions associated with O3 production in metropolitan areas. Their
 5      method, which relies on routine monitoring methods, is especially useful because it permits
 6      evaluation for a full season rather than just for individual episodes.
 7
 8      AX2.5.4 Chemistry-Transport Model Evaluation
 9           The comparison of model predictions with ambient measurements represents a critical task
10      for establishing the accuracy of photochemical models and evaluating their ability to serve as the
11      basis for making effective  control strategy decisions. The evaluation of a model's performance,
12      or its adequacy to perform the tasks for which it was designed can only be conducted within the
13      context of measurement errors and artifacts. Not only are there analytical problems, but there
14      are also problems in assessing the representativeness of monitors at ground level for comparison
15      with model values which represent typically an average over the volume of a grid box.
16           Chemistry-transport models for O3 formation at the urban/regional scale have traditionally
17      been evaluated based on their ability to correctly simulate O3. A series of performance statistics
18      that measure the success of individual model simulations to represent the observed distribution
19      of ambient O3, as represented by a network of surface measurements were recommended in U.S.
20      Environmental Protection Agency (1991; see also Russell and Dennis, 2000). These statistics
21      consist of the following:
22       •  Unpaired peak O3 within a metropolitan region (typically for a single day).
23       •  Normalized bias equal to the summed difference between model and measured hourly
            concentrations divided by the sum of measured hourly concentrations.
24       •  Normalized gross error, equal to the summed unsigned (absolute value) difference
            between model and measured hourly concentrations divided by the sum of measured
            hourly concentrations.
25
        August 2005                            AX2-129      DRAFT-DO NOT QUOTE OR CITE

-------
        Unpaired peak prediction accuracy, Au;
                                                  J /max
                                                           *100%,                   (AX2-49)
                                               max
 2
 3     Normalized bias, D;
 4
 5
 7
 8      Gross error, Ed (for hourly observed values of O3 >60 ppb)
                                                            t= 1, 24.                 (AX2-51)
10
11           The following performance criteria for regulatory models were recommended in U.S.
12      Environmental Protection Agency (1991): unpaired peak O3 to within ±15% or ±20%;
13      normalized bias within ±5% to ±15%; and normalized gross error less than 30% to 35%, but
14      only when O3 >60 ppb. This can lead to difficulties in evaluating model performance since
15      nighttime and diurnal cycles are ignored. A major problem with this method of model
16      evaluation is that it does not provide any information about the accuracy of O3-precursor
17      relations predicted by the model. The process of O3 formation is sufficiently complex that
18      models can predict O3 correctly without necessarily representing the O3 formation process
19      properly.  If the O3 formation process is incorrect, then the modeled source-receptor relations
20      will also be incorrect.
21           Studies by Sillman et al. (1995, 2003), Reynolds et al. (1996) and Pierce et al. (1998) have
22      identified instances in which different model scenarios can be created with very different
23      O3-precursor sensitivity, but without significant differences in the predicted O3 fields.
24      Figures AX2-21a,b provide an example.  Referring to the O3-NOX-VOC isopleth plot
25      (Figure AX2-22), it can be seen that similar O3 concentrations can be found for photochemical
26      conditions that have very different sensitivity to NOX and VOCs.

        August 2005                            AX2-130      DRAFT-DO NOT QUOTE OR CITE

-------
^.o
*>" 9n
£ 2D
Q.
tt 1S
C
.2 10
o
3
01
co n
0 °
c.
±J







^€
*


X
* •
• X
•
XXX*
**9\
1
•

, x •,
V


1
•

, .
<



100 120 140 160 180 200
Ozone (ppb)
otr

.Q
Q. OR _
3
C on
.2
o m
3 15
O -in

O 5 "
n
u








4



X X

.
' *•
**

X
XX
X
X
1
\ ,
. "
•.
X* '
X
X
X

<
X


100 120 140 160 180 200
Ozone (ppb)
      Figure AX2-21a,b.  Impact of model uncertainty on control strategy predictions for O3
                          for two days (August 10 [a] and ll[b], 1992) in Atlanta, GA.  The
                          figures show the predicted reduction in peak O3 resulting from 35%
                          reductions in anthropogenic VOC emissions (crosses) and from 35%
                          reductions in NOX (solid circles) in a series of model scenarios with
                          varying base case emissions, wind fields, and mixed layer heights.

      Source: Results are plotted from tabulated values published in Sillman et al. (1995, 1997).
1           Global-scale chemistry-transport models have generally been evaluated by comparison
2     with measurements for a wide array of species, rather than just for O3 (e.g., Wang et al., 1998;
3     Emmons et al., 2000; Bey et al., 2001b; Hess, 2001; Fiore et al., 2002).  These have included
4     evaluation of major primary species (NOX, CO, and selected VOCs) and an array of secondary
      August 2005
AX2-131
DRAFT-DO NOT QUOTE OR CITE

-------
                             -   10.00
                             E
                             o
                                 3.16
                             o
                             E
                            CM
                            O
                            ^,     1.0
                             0)
                            •5,  0.316
                            (A
                            E
                            UJ
                            5    0.1
                           o.
                           a.
                           CO
                           o
                                     1.0     3.16     10.0     31.6     100.0
                                   VOC Emission Rate (1012 molec. cm-2 s-1)
       Figure AX2-22.  Ozone isopleths (ppb) as a function of the average emission rate for
                        NOX and VOC (1012 molec. cm~2 s'1) in zero dimensional box model
                        calculations.  The isopleths (solid lines) represent conditions during
                        the afternoon following 3-day calculations with a constant emission
                        rate, at the hour corresponding to maximum O3.  The ridge line
                        (shown by solid circles) lies in the transition from NOx-saturated to
                        NO-limited conditions.
 1     species (HNO3, PAN, H2O2) that are often formed concurrently with O3. Models for
 2     urban/regional O3 have also been evaluated against a broader ensemble of measurements in a
 3     few cases, often associated with measurement intensives (e.g., Jacobson et al., 1996, Lu et al.,
 4     1997; Sillman et al., 1998). The results of a comparison between observed and computed
 5     concentrations from Jacobson et al. (1996) for the Los Angeles Basin are shown in
 6     Figures AX2-23a,b.
 7          The highest  concentrations of primary species usually occur in close proximity to emission
 8     sources (typically in urban centers) and at times when dispersion rates are low. The diurnal
 9     cycle includes high concentrations at night, with maxima during the morning rush hour, and low
10     concentrations during the afternoon (Figure AX2-23a). The afternoon minima are driven by the
11     much greater rate  of vertical mixing at that time. Primary species also show a seasonal
       August 2005
AX2-132
DRAFT-DO NOT QUOTE OR CITE

-------
               Q.
               a.
               o
              44
               (0
               O)
               c
               Q.
               Q.
               (Q
              o:
               D)
               C
              'x
               Q.
               Q.
(0
cr
_c
'R
0.30

0.25

0.20

0.15

0.10

0.05

0.00


0.30

0.25

0.20

0.15

0.10

0.05

0.00


  6

  5

  4

  3

  2

  1

  0
                          Reseda
                          O3(g)
                                     	  Predicted
                                     	Observed
                           8    16   24    32    40   48   56   64    72
                                 Hour After First Midnight
                          Reseda
                          N0x(g)
                                     	  Predicted
                                     	Observed
                           8    16   24    32    40   48   56   64    72
                                 Hour After First Midnight
                          Riverside
                          C0(g)
                                     	  Predicted
                                     	Observed
                                                                 \   /
                           8    16   24    32    40   48   56   64    72
                                 Hour After First Midnight
Figure AX2-23a.   Time series for measured gas-phase species in comparison with
                  results from a photochemical model.  The dashed lines represent
                  measurements, and solid lines represent model predictions (in parts per
                  million, ppmv) for August 26 - 28,1988 at sites in southern California.
                  The horizontal axis represents hours past midnight, August 25. Results
                  represent O3 and NOX at Reseda and CO at Riverside.

Source:  Jacobson et al. (1996).
August 2005
                    AX2-133
                                     DRAFT-DO NOT QUOTE OR CITE

-------
 1      maximum during winter, and are often high during fog episodes in winter when vertical mixing ,
 2      is suppressed. By contrast, secondary species such as O3 are typically highest during the
 3      afternoon (the time of greatest photochemical activity), on sunny days and during summer.
 4      During these conditions concentrations of primary species may be relatively low.  Strong
 5      correlations between primary and secondary species are generally observed only in downwind
 6      rural areas where all anthropogenic species are high simultaneously. The difference in the
 7      diurnal cycles of primary species (CO, NOX and ethane)and secondary species (O3, PAN and
 8      HCHO) is evident in Figure AX2-23b.
 9          Models for urban/regional O3 have been evaluated less extensively than global-scale
10      models in part because the urban/regional context presents a number of difficult challenges.
11      Global-scale models typically represent continental-scale events and can be evaluated
12      effectively against a sparse network of measurements. By contrast, urban/regional models are
13      critically dependent on the accuracy of local emission inventories and event-specific
14      meteorology, and must be evaluated separately for each urban area that is represented.
15          The evaluation of urban/regional models is also limited by the availability of data.
16      Measured NOX and speciated VOC concentrations are widely available through the EPA PAMs
17      network, but questions have been raised about the accuracy of those measurements and the data
18      have not yet been analyzed thoroughly. Evaluation of urban/regional models versus
19      measurements has generally relied on results from a limited number of field  studies in the United
20      States. Short term research-grade measurements for species relevant to O3 formation, including
21      VOCs, NOX, PAN, nitric acid (HNO3) and hydrogen peroxide (H2O2) are also widely available at
22      rural and remote sites (e.g., Daum et al., 1990, 1996; Martin et al., 1997; Young et al., 1997;
23      Thompson et al., 2000; Hoell et al., 1996, 1997; Fehsenfeld et al., 1996a; Emmons et al., 2000;
24      Hess, 2001; Carroll et al., 2001). The equivalent measurements are available for some polluted
25      rural sites in the eastern United States (e.g.) but only at a few urban locations (Meagher et al.,
26      1998; Hubler et  al.,  1998; Kleinman et al., 2000, 2001; Fast et al., 2002; new SCAQS-need
27      reference). Extensive measurements have also been made in Vancouver (Steyn et al., 1997) and
28      in several European cities (Staffelbach et al., 1997; Prevot et al., 1997, Dommen et al., 1999;
29      Geyer et al., 2001; Thielman et al., 2001; Martilli et al., 2002; Vautard et al., 2002).
30          The results of straightforward comparisons between observed and predicted concentrations
31      of O3 can be misleading because of compensating errors, although this possibility is diminished

        August 2005                            AX2-134     DRAFT-DO NOT QUOTE OR CITE

-------
                   0.060

                |  0.050
                Q.
                3  0.040
                o
                '-§  0.030
                OL
                0)  0.020
                _c
                ~  0.010

                   0.000
                   0.030

                |  0.025
                Q.
               3  0.020
                O
               ~  0.015
               a:
                0)  0.010

               ~  0.005

                   0.000
                   0.020
                £  0.016
                Q.
                a.
                "g"  0.012
                «rf
                (0
               o:
                   0.008
                O)
                _c
                !B  0.004
                §
                   0.000
             Claremont
             Ethane (g)
  	  Predicted
  O    Observed
               _L
                                 16   24   32   40   48   56
                                  Hour After First Midnight
                                             64   72
          Claremont
          Formaldehyde (g)
 	  Predicted
  O    Observed
                I
                                 16   24   32   40   48   56
                                  Hour After First Midnight
                                             64   72
          Los Angeles
          PAN (g)
	  Predicted
	Observed
               _L
                                 16   24   32   40   48   56
                                  Hour After First Midnight
                                             64   72
Figure AX2-23b.
Time series for measured gas-phase species in comparison with results
from a photochemical model.  The circles represent measurements, and
solid lines represent model predictions (in parts per million, ppmv) for
August 26 - 28,1988 at sites in southern California.  The horizontal axis
represents hours past midnight, August 25. Results represent ethane
and formaldehyde at Claremont, and PAN at Los Angeles.
Source:  Jacobsonetal. (1996).
August 2005
                     AX2-135
DRAFT-DO NOT QUOTE OR CITE

-------
 1      when a number of species are compared. Ideally, each of the main modules of a chemistry-
 2      transport model system (for example, the meteorological model and the chemistry and radiative
 3      transfer routines) should be evaluated separately.  However, this is rarely done in practice.
 4      To better indicate how well physical and chemical processes are being represented in the model,
 5      comparisons of relations between concentrations measured in the field and concentrations
 6      predicted by the model can be made.  These comparisons could involve ratios and correlations
 7      between species.  For example, correlation coefficients could be calculated between primary
 8      species as a means of evaluating the accuracy of emission inventories; or between secondary
 9      species as a means of evaluating the treatment of photochemistry in the model.  In addition,
10      spatial relations involving individual species (correlations, gradients) can also be used as a
11      means of evaluating the accuracy of transport parameterizations.  Sillman and He (2002)
12      examined differences in correlation patterns between O3 and NOZ in Los Angeles, CA, Nashville,
13      TN and various sites in the rural United States. Model calculations (Figure AX2-24)  show
14      differences in correlation patterns associated with differences in the sensitivity of O3 to NOX
15      and VOCs.  Primarily NOx-sensitive ( NOx-limited) areas in models show a strong correlation
16      between O3 and NOZ with a relatively steep slope, while primarily VOC-sensitive (NOX-
17      saturated) areas in models show lower O3 for a given NOZ and a lower O3-NOZ slope.  They
18      found that differences found in measured data ensembles were  matched by predictions from
19      chemical transport models. Measurements in rural areas in the eastern U.S. show differences in
20      the pattern of correlations for O3 versus NOZ between summer and autumn (Jacob et al., 1995;
21      Hirsch  et al., 1996), corresponding to the transition from NOx-limited to NOx-saturated patterns,
22      a feature which is also matched by chemistry-transport models.
23           The difference in correlations between secondary species in NOx-limited to NOx-saturated
24      environments can also be used to evaluate the accuracy of model predictions in individual
25      applications. Figures AX2-25a and AX2-25b show results for two different model scenarios for
26      Atlanta. As shown in the figures, the first model scenario predicts an urban plume with high
27      NOy and O3 formation apparently suppressed by high NOy. Measurements show much lower
28      NOy in  the Atlanta plume. This error was especially significant because the model locations
29      with high NOy were not sensitive to NOX, while locations with lower NOy were primarily
30      sensitive to NOX.  The second model scenario (with primarily NOx-sensitive conditions) shows
31      much better agreement with measured values.  Figure AX2-26a,b shows model-measurement

        August 2005                            AX2-136     DRAFT-DO NOT QUOTE OR CITE

-------
                      250
                                                      20
                                                  NOZ (ppb)
       Figure AX2-24.  Correlations for O3 versus NOZ (NOy - NOX) in ppb from chemical
                        transport models for the northeast corridor, Lake Michigan, Nashville,
                        the San Joaquin Valley and Los Angeles. Each location is classified as
                        NOx-limited or NOx-sensitive (circles), NOx-saturated or VOC-sensitive
                        (crosses), mixed or with near-zero sensitivity (squares), and dominated by
                        NOX titration (asterisks) based on the model response to reduced NOX and
                        VOC.
       Source: Sillman and He (2002).
 1     comparisons for secondary species in Nashville, showing better agreement with measured
 2     conditions. Greater confidence in the predictions made by chemistry-transport models will be
 3     gained by the application of techniques such as these on a more routine basis.
 4          The ability of chemical mechanisms to calculate the concentrations of free radicals under
 5     atmospheric conditions was tested in the Berlin Ozone Experiment, BERLIOZ (Volz-Thomas
 6     et al., 2003) during July and early August at a site located about 50 km NW of Berlin.  (This
 7     location was chosen as O3 episodes in central Europe are often associated with SE winds.)
 8     Concentrations of major compounds such as O3, hydrocarbons, etc., were fixed at observed
 9     values. In this regard, the protocol used in this evaluation is an example  of an observationally
10     based method. Figure AX2-27 compares the concentrations of RO2 (organic peroxy), HO2
       August 2005
AX2-137
DRAFT-DO NOT QUOTE OR CITE

-------
                     .0
                     CL
                     Q.
                                       10         20          30
                                              NOy (Ppb)
                             40
                                       10         20          30
                                              NOy (Ppb)
                             40
      Figure AX2-25a,b.   Evaluation of model versus measured O3 versus NOy for two
                          model scenarios for Atlanta. The model values are classified as NOX
                          limited (circles), NOx-saturated (crosses), or mixed or with low
                          sensitivity to NOX (squares). Diamonds represent aircraft
                          measurements.

      Source: Sillmanetal. (1997).
1     (hydroperoxy) and OH (hydroxyl) radicals predicted by RACM (regional air chemistry

2     mechanism; Stockwell et al., 1997) and MCM (master chemical mechanism; Jenkin et al, 1997

3     with updates) with observations made by the laser induced fluorescence (LIF) technique and by
      August 2005
AX2-138
DRAFT-DO NOT QUOTE OR CITE

-------
                        160
                        140
                        160
                                       10          20          30
                                              NOZ (ppb)
                            40
                                       10          20          30
                                          2H202 + NOZ (ppb)
                            40
      Figure AX2-26a,b.  Evaluation of model versus: (a) measured O3 versus NOZ and (b) O3
                         versus the sum 2H2O2 + NOZ for Nashville, TN. The model values are
                         classified as NOx-limited (gray circles), NOx-saturated (xs), mixed or
                         with near-zero sensitivity (squares), or dominated by NOX titration
                         (filled circles).  Diamonds represent aircraft measurements.

      Source:  Sillmanetal. (1998).
1     matrix isolation ESR spectroscopy (MIESR). Also shown are the production rates of O3

2     calculated using radical concentrations predicted by the mechanisms and those obtained by

3     measurements, and measurements of NOX concentrations.  As can be seen, there is good
      August 2005
AX2-139
DRAFT-DO NOT QUOTE OR CITE

-------
CO
 I

 o
CO    4
O
£.   2
 CM
O
x   0
           o
           CD
           O
               10
           ^   5

                0
               10
           Q.
           Q.
           X   5
           CO
           o
           Q.
           a.
            X
           O
                0
              20
    10
                0
                     O   LIF
                     •   MIESR
                                                  J(d1D)*106(s-1)
                       8
                      10        12        14

                           UT 20.7.98
              16
Figure AX2-27.  Time series of concentrations of RO2, HO2, and OH radicals, local
               O3 photochemical production rate and concentrations of NOX from
               measurements made during BERLIOZ.  Also shown are comparisons
               with results of photochemical box model calculations using the RACM
               and MCM chemical mechanisms.

Source: Volz-Thomas et al. (2003).
August 2005
                         AX2-140
DRAFT-DO NOT QUOTE OR CITE

-------
 1      agreement between measurements of organic peroxy, hydroperoxy and hydroxyl radicals with
 2      values predicted by both mechanisms at high concentrations of NOX (> 10 ppb). However, at
 3      lower NOX concentrations, both mechanisms substantially overestimate OH concentrations and
 4      moderately overestimate HO2 concentrations.  Agreement between models and measurements is
 5      generally better for organic peroxy radicals, although the MCM appears to overestimate their
 6      concentrations somewhat. In general, the mechanisms reproduced the HO2 to OH and RO2 to
 7      OH ratios better than the individual measurements.  The production of O3 was found to increase
 8      linearly with NO (for NO <0.3 ppb) and to decrease with NO (for NO >0.5 ppb).
 9          OH and HO2 concentrations measured during the PM2 5 Technology Assessment and
10      Characterization Study conducted at Queens College in New York City in the summer of 2001
11      were also compared with those predicted by RACM (Ren et al., 2003). The ratio of observed to
12      predicted HO2 concentrations over a diurnal cycle was 1.24 and the ratio of observed to
13      predicted OH concentrations was about 1.10 during the day, but the mechanism significantly
14      underestimated OH concentrations during the night.
15
16      AXA.5.4.1  Evaluation of Emissions Inventories
17          Comparisons of emissions model predictions with observations have been performed in a
18      number of environments. A number of studies of ratios of concentrations of CO to NOX and
19      NMOC to NOX during the early 1990s in tunnels and ambient air (summarized in Air Quality
20      Criteria for Carbon Monoxide [U.S. Environmental Protection Agency, 2000]) indicated that
21      emissions of CO and NMOC were systematically underestimated in emissions inventories.
22      However, the results of more recent studies have been mixed in this regard, with many studies
23      showing agreement to within ± 50% (U.S. Environmental Protection Agency, 2000).
24      Improvements in many areas have resulted from the process of emissions model development,
25      evaluation, and further refinement. It should be remembered that the conclusions from these
26      reconciliation studies depend on the assumption that NOX emissions are predicted correctly by
27      emissions factor models. Road side remote sensing data indicate that over 50% of NMHC and
28      CO emissions are produced by less than about 10% of the vehicles (Stedman et al.,  1991). These
29      "super-emitters" are typically poorly maintained vehicles.  Vehicles of any age engaged in off-
30      cycle operations (e.g., rapid accelerations) emit much more than if operated in normal driving
       August 2005                           AX2-141      DRAFT-DO NOT QUOTE OR CITE

-------
 1     modes. Bishop and Stedman (1996) found that the most important variables governing CO
 2     emissions are fleet age and owner maintenance.
 3          Emissions inventories for North America can be evaluated with comparisons to measured
 4     long-term trends and or ratios of pollutants in ambient air.  A decadal field study of ambient CO
 5     at a rural cite in the Eastern U.S. (Hallock-Waters et al., 1999) indicates a downward trend
 6     consistent with the downward trend in estimated emissions over the period 1988 to 1999 (U.S.
 7     Environmental Protection Agency, 1997), even when a global downward trend is accounted for.
 8     Measurements at two urban areas in the United States confirmed the decrease in CO emissions
 9     (Parrish et al.,  2002). That study also indicated that the ratio of CO to NOX emissions decreased
10     by almost a factor of three over 12 yr (such a downward trend was noted in AQCD 96).
11     Emissions estimates (U.S. Environmental Protection Agency, 1997) indicate a much smaller
12     decrease in this ratio, suggesting that NOX emissions from mobile sources may be
13     underestimated and/or increasing. The authors conclude that O3 photochemistry in U.S. urban
14     areas may have become more NOx-limited over the past decade.
15          Pokharel et al. (2002) employed remotely-sensed emissions from on-road vehicles and fuel
16     use data to estimate emissions in Denver. Their calculations indicate a continual decrease in CO,
17     HC, and NO emissions from mobile sources over the 6 yr study period.  Inventories based on the
18     ambient data were 30 to 70% lower for CO, 40% higher for HC, and 40 to 80% lower for NO
19     than those predicted by the recent MOBILE6 model.
20          Stehr et al. (2000) reported simultaneous measurements of CO, SO2 and NOy at an East
21     Coast site. By taking advantage of the nature of mobile sources (they emit NOX and CO but little
22     SO2) and power plants (they emit NOX and SO2 but little CO), the authors evaluated emissions
23     estimates for the eastern United States.  Results indicated that coal combustion contributes 25 to
24     35% of the total NOX emissions in agreement with emissions inventories (U.S. Environmental
25     Protection Agency, 1997).
26          Parrish et al. (1998) and Parrish and Fehsenfeld (2000) proposed methods to derive
27     emission rates by examining measured ambient ratios among individual VOC, NOX and NOy.
28     There is typically a strong correlation among measured values for these species (e.g., Figure
29     AX2-14) because emission sources are geographically collocated, even when individual sources
30     are different. Correlations can be used to derive emissions ratios between species, including
31     adjustments for the impact of photochemical  aging.  Investigations of this type  include

       August 2005                           AX2-142      DRAFT-DO NOT QUOTE OR CITE

-------
 1      correlations between CO and NOy (e.g., Parrish et al., 1991), between individual VOC species
 2      and NOy (Goldan et al., 1995,1997, 2001; Harley et al., 1997) and between various individual
 3      VOC (Goldan et al.,  1995, 1997; McKeen and Liu, 1993; McKeen et al., 1996).  Buhr et al.
 4      (1992) derived emission estimates from principal component analysis (PCA) and other statistical
 5      methods.  Many of these studies are summarized in Trainer et al. (2000), Parrish et al. (1998),
 6      and Parrish and Fehsenfeld (2000). Goldstein and Schade (2000) also used species correlations
 7      to identify the relative impacts of anthropogenic and biogenic emissions.  Chang et al. (1996,
 8      1997) and Mendoza-Dominguez and Russell (2000, 2001) used the more formal techniques of
 9      inverse modeling to derive emission rates, in conjunction with results from chemistry-transport
10      models.  Another concern regarding the use of emissions inventories is that emissions from all
11      significant sources have been included.  This may not always be the case. As an example,
12      hydrocarbon seeps from off-shore oil fields may represent a significant source of reactive
13      organic compounds in near by coastal areas (Quigley et al., 1999).
14
15      AX2.5.4.2  Availability and Accuracy of Ambient Measurements
16           The use of methods such as observationally based methods or source apportionment
17      models, either as stand-alone methods or as a basis for evaluating chemistry/transport models,
18      is often limited by the availability and accuracy of measurements.  Measured speciated VOC and
19      NOX are widely available in the United States through the PAMS network. However, challenges
20      have been raised about both the  accuracy of the measurements and their applicability.
21           Parrish et al. (1998) and Parrish and Fehsenfeld (2000) developed a series of quality
22      assurance tests for speciated VOC measurements.  Essentially these tests used ratios  among
23      individual VOC with common emission sources to identify whether the variations in species
24      ratios were consistent with the relative photochemical lifetimes of individual species. These
25      tests were based on a number of assumptions: the ratio between ambient concentrations of
26      long-lived species should show relatively little variation among measurements affected by a
27      common emissions sources; and the ratio between ambient concentrations of long-lived and
28      short-lived species should vary in a way that reflects photochemical aging at sites more different
29      from source regions. Parrish et al. used these expectations to establish criteria for rejecting
30      apparent errors in measurements.  They found that the ratios among alkenes at many  PAMS sites
31      did not show variations that would be expected due to photochemical aging.

        August 2005                            AX2-143      DRAFT-DO NOT QUOTE OR CITE

-------
 1          The PAMs network currently includes measured NO and NOX. However, Cardelino and
 2     Chameides (2000) reported that measured NO during the afternoon was frequently at or below
 3     the detection limit of the instruments (1 ppb), even in large metropolitan regions (Washington,
 4     DC; Houston, TX; New York, NY). NOX measurements are made with commercial
 5     chemilluminescent detectors with molybdenum converters. However these measurements
 6     typically include some organic nitrates in addition to NOX, and cannot be interpreted as a "pure"
 7     NOX measurement (see summary in Parrish and Fehsenfeld, 2000).
 8          Total reactive nitrogen (NOy) is included in the PAMS network only at a few sites. The
 9     possible expansion of PAMS to include more widespread NOy measurements has been suggested
10     (McClenny, 2000). A major issue concerning measured NOy is the possibility that HNO3,
11     a major component of NOy, is sometimes lost in inlet tubes and not measured (Luke et al., 1998;
12     Parrish and Fehsenfeld, 2000). This problem is especially critical if measured NOy is used to
13     identify NOx-limited versus NOx-saturated conditions. The correlation between O3 and NOy
14     differs for NOx-limited versus NOx-saturated locations, but this difference is driven primarily by
15     differences in the ratio of O3 to HNO3. If HNO3 were omitted from the NOy measurements, than
16     the measurements would represent a biased estimate and their use would be problematic.
17
18
19     AX2.6   TECHNIQUES FOR MEASURING OZONE AND ITS
20               PRECURSORS
21     AX2.6.1  Sampling and Analysis of Ozone
22          Numerous techniques have been developed for sampling and measurement of O3 in the
23     ambient atmosphere at ground level.  As noted above,  sparse surface networks tend to
24     underestimate maximum O3 concentrations. Today, monitoring is conducted almost exclusively
25     with UV absorption spectrometry with commercial short path instruments, a method that has
26     been thoroughly evaluated in clean air. The ultimate reference method is a relatively long-path
27     UV absorption instrument maintained under carefully controlled conditions at NIST (e.g., Fried
28     and Hodgeson, 1982).  Episodic measurements are made with a variety of other techniques based
29     on the principles of chemiluminescence, electrochemistry, DOAS, and LIDAR. The rationale,
30     history, and calibration of O3 measurements were summarized in AQCD 96, so this section will
       August 2005                           AX2-144      DRAFT-DO NOT QUOTE OR CITE

-------
 1      focus on the current state of ambient O3 measurement, tests for artifacts, and on new
 2      developments.
 3           Several reports in the reviewed scientific literature have investigated interferences in O3
 4      detection via UV radiation absorption. Kleindienst et al. (1993) investigated the effects of water
 5      vapor and VOCs on instruments based on both UV absorption and chemiluminescence. They
 6      concluded that water vapor had no significant impact on UV absorption-based instruments, but
 7      could cause a positive interference of up to 9% in chemiluminescence-based detectors at high
 8      humidities (dew point of 24 C). In smog chamber studies, aromatic compounds and their
 9      oxidation products were found to generate a positive but small interference in the UV absorption
10      instruments.  Kleindienst et al. concluded that "when the results are scaled back to ambient
11      concentrations of toluene and NOX, the effect appears to be very minor (ca. 3 percent under the
12      study conditions)." Narita et al. (1998) tested organic and inorganic compounds and found
13      response to several, but not at levels likely to interfere with accurate determination of O3 in an
14      urban environment. More recently, Arshinov et al. (2002) reported a positive interference in UV
15      absorption instruments from ambient aerosols, but this interference is eliminated by use of
16      appropriate particle filters. The possibility for substantive interferences in O3 detection exists,
17      but such interferences have not been observed even in urban plumes. Ryerson et al. (1998)
18      measured O3 with UV absorption and chemiluminescence instruments operated  off a common
19      inlet on the NOAA WP-3 research aircraft. As reported by Parrish and Fehsenfeld (2000)
20      "Through five field missions over four years, excellent correlations were found between the
21      measurements of the two instruments, although the chemiluminescence instrument was
22      systematically low (5%) throughout some flights. The data sets include many passes through the
23      Nashville urban plume.  There was never any indication (< 1%) that the UV instrument
24      measured systematically higher in the urban plume." The same group tested the air of Houston,
25      El Paso, Nashville, Los Angeles, San Francisco, and the East Coast. They observed only one
26      instance of substantive positive interference defined as the UV absorption technique showing
27      more than a few ppb  more than the CL. This occurred in Laporte, TX under heavily polluted
28      conditions and a low inversion, at night (Jobson et al., 2004).
29           Leston et al. (2005) reported positive and negative interferences in UV absorption
30      techniques for measuring O3 (relative to the CL technique) in Mexico City and in a smog
31      chamber study.  They suggested that O3 measured in ambient air could be too high by 20 to

        August 2005                            AX2-145      DRAFT-DO NOT QUOTE OR CITE

-------
 1      40 ppb under specific conditions due to positive interference by a number of organic compounds,
 2      mainly those produced during the oxidation of aromatic hydrocarbons and some primary
 3      compounds such as styrene and naphthalene.  However, they did not collocate CL and UV
 4      instruments at any ambient air monitoring sites in the United States. In addition, the
 5      concentrations of these compounds were many times higher in both of these environments than
 6      are typically found at ambient air monitoring sites in the United States. Although Hg is also
 7      potentially a strong interfering agent, because the Hg resonance line is used in this technique, its
 8      concentration would  also have to be many times higher than is typically found in ambient air.
 9      Thus, it seems unlikely that such interferences would amount to more  than one or two ppb
10      (within the design specifications of the FEM), except under conditions conducive to producing
11      high concentrations of the substances they identified as causing interference. These conditions
12      might be found next to a plant producing styrene, for example. Leston et al. (2005) also noted
13      that the use of alternative materials in the scrubber could alleviate many potential problems
14      under these conditions.
15           Ozone can also be detected by differential optical absorption  spectroscopy (DOAS) at a
16      variety of wavelengths in the UV and visible parts of the spectrum.  Prior comparisons of DOAS
17      results to those from  a UV absorption instrument showed good agreement, on the order of 10%
18      (Stevens et al., 1993). Reisinger (2002) reported a positive interference due to an unidentified
19      absorber in the 279 to 289 nm spectral region used by many commercial short-path DOAS
20      systems for the measurement of O3. Results of that study suggest that effluent from wood
21      burning, used for  domestic heating, may be responsible. Vandaele et al. (2002) reported good
22      agreement with other methods in the detection of O3 (and SO2) over the course of several years
23      in Brussels. While the DOAS method remains attractive due to its  sensitivity and speed of
24      response further intercomparisons and interference tests are recommended.
25           Electrochemical methods are commonly employed where sensor weight is a problem,  such
26      as in balloon borne sondes, and these techniques have been investigated for ambient monitoring.
27      Recent developments include changes in the electrodes and electrolyte solution (Knake and
28      Hauser, 2002) and selective removal of O3 for a chemical zero (Penrose et al.,  1995).
29      Interferences from other oxidants such as NO2 and HONO remain potential problems and further
30      comparisons with UV absorption are necessary. Because of potential interferences from water
        August 2005                            AX2-146      DRAFT-DO NOT QUOTE OR CITE

-------
 1      vapor (ASTM, 2003 a,b), all instruments should be either calibrated and zeroed with air humidity
 2      near ambient or demonstrated to be insensitive to humidity.
 3           Change in the vibration frequency of a piezoelectric quartz crystal has been investigated as
 4      a means of detecting O3.  Ozone reacts with polybutadiene coated onto the surface of a crystal,
 5      and the resulting change in mass is detected as a frequency change (Black et al., 2000). While
 6      this sensor has advantages of reduced cost power consumption and weight, it is lacks the lifetime
 7      and absolute accuracy for ambient monitoring.
 8           In summary, new techniques are being developed, but UV absorption remains the method
 9      of choice for ambient O3 monitoring near the Earth's surface. These commercial UV absorption
10      detectors are available at a moderate price. They show good absolute accuracy with only minor
11      cross sensitivity in clean to moderately  polluted environments; they are stable, reliable, and
12      sensitive.
13
14      AX2.6.2   Sampling and Analysis of Nitrogen Oxides
15           The role of nitrogen oxides in tropospheric O3 formation was reviewed thoroughly in the
16      previous AQCD and will be only briefly summarized here.  Reactive nitrogen is generally
17      released as NO but quickly converted in ambient air to NO2 and back again, thus these two
18      species are often referred to together as NOX (NO + NO2).  The photochemical interconversion of
19      NO and NO2 leads to O3 formation. Because NO2 is a health hazard at sufficiently high
20      concentrations, it is itself a criteria pollutant.  In EPA documents, emissions of NOX are
21      expressed in units of mass of NO2 per unit time, i.e., the total mass of NOX that would be emitted
22      if all the NO were converted to NO2.  Ambient air monitors have been required to demonstrate
23      compliance with the standard for NO2 and thus have focused on measuring this gas or
24      determining an upper limit for its concentration.
25           NOX can be further oxidized to species including nitrous acid (HNO2), nitric acid (HNO3),
26      aerosol nitrate (NO3 ), and organo-nitrates such as alkyl nitrates (RONO2) and peroxy acetyl
27      nitrate, PAN, (CH3C(O)O2NO2). The sum of these species (explicitly excluding N2, N2O, and
28      reduced N such as NH3 and HCN) is called NOy. Nitrates play important roles in acid rain, and
29      nutrient cycling including over nitrification of surface ecosystems and in the formation of fine
30      particulate matter, but are generally inactive photochemically.  Some studies refer specifically to
31      the oxidized or processed NOy species, NOy-NOx, as NOZ because this quantity is related to the

        August 2005                            AX2-147     DRAFT-DO NOT QUOTE OR CITE

-------
 1      degree of photochemical aging in the atmosphere.  Several NOZ species such as PAN and HONO
 2      can be readily photolyzed or thermally dissociated to NO or NO2 and thus act as reservoirs for
 3      NOX. This discussion focuses on current methods and on promising new technologies, but no
 4      attempt is made here to cover the extensive development of these methods or of methods such as
 5      wet chemical techniques, no longer in widespread use. More detailed discussions of the histories
 6      of these methods may be found elsewhere (U.S. Environmental Protection Agency, 1993,  1996).
 7
 8      AX2.6.2.1  Calibration Standards
 9           Calibration gas standards of NO, in nitrogen (certified at concentrations of approximately
10      5 to 40 ppm) are obtainable from the Standard Reference Material (SRM) Program of the
11      National Institute of Standards and Technology (NIST),  formerly the National Bureau of
12      Standards (NBS), in Gaithersburg, MD.  These SRMs are supplied as compressed gas mixtures
13      at about 135 bar (1900 psi) in high-pressure aluminum cylinders containing 800 L of gas at
14      standard temperature and pressure, dry (STPD; National Bureau of Standards, 1975; Guenther
15      et al., 1996). Each cylinder is supplied with a certificate stating concentration and uncertainty.
16      The concentrations are certified to be accurate to ±1 percent relative to the stated values.
17      Because of the resources required for their certification,  SRMs are not intended for use as daily
18      working standards, but rather as primary standards against which transfer standards can be
19      calibrated.
20           Transfer stand-alone calibration gas standards of NO in N2 (in the concentrations indicated
21      above) are obtainable from specialty gas companies. Information as to whether a company
22      supplies such mixtures is obtainable from the company, or from the SRM Program of NIST.
23      These NIST Traceable Reference Materials (NTRMs) are purchased directly from industry and
24      are supplied as compressed gas mixtures at approximately  135 bars (1,900 psi) in high-pressure
25      aluminum cylinders containing 4,000 L of gas at STPD.  Each cylinder is supplied with a
26      certificate stating concentration and uncertainty. The concentrations are certified to  be accurate
27      to within ±1 percent of the stated values (Guenther et al., 1996). Additional details can be found
28      in the previous AQCD for O3 (U.S. Environmental Protection Agency, 1996).
29
30
        August 2005                            AX2-148      DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX2.6.2.2  Measurement of Nitric Oxide
 2      Gas-Phase Chemiluminescence (CL) Methods
 3           Nitric oxide, NO, can be measured reliably using the principle of gas-phase
 4      chemiluminescence induced by the reaction of NO with O3 at low pressure. Modern commercial
 5      NOX analyzers have sufficient sensitivity and specificity for adequate measurement in urban and
 6      many rural locations (U.S. Environmental Protection Agency, 1996).  The physics of the method,
 7      detection limits, interferences, and comparisons under field comparisons have been thoroughly
 8      reviewed in the previous AQCD. Research grade CL instruments have been compared under
 9      realistic field conditions to spectroscopic instruments, and the results indicate that both methods
10      are reliable (at concentrations relevant to smog studies) to better than  15 percent with 95 percent
11      confidence. Response times are on the order of 1 minute. For measurements meaningful for
12      understanding O3 formation, emissions modeling, and N deposition, special care must be taken
13      to frequently zero and calibrate the instrument.  A chemical zero, by reacting the NO up stream
14      and out of view of the PMT, is preferred because it accounts for unsaturated hydrocarbon or
15      other interferences. Calibration should be performed with NIST-traceable reference material of
16      compressed NO in N2.  Standard additions of NO at the inlet will account for NO loss or
17      conversion to NO2 in the lines. In summary CL methods, when operated in an appropriate
18      manner, can be suitable for measuring or monitoring NO (e.g., Crosley,  1996).
19
20      Spectroscopic Methods for Nitric Oxide
21           Nitric oxide has also been successfully measured in ambient air with direct spectroscopic
22      methods; these include two-photon laser-induced fluorescence (TPLIF), tunable diode laser
23      absorption spectroscopy (TDLAS), and two-tone frequency-modulated spectroscopy (TTFMS).
24      These were reviewed thoroughly in the previous AQCD and will be only briefly summarized
25      here. The spectroscopic methods demonstrate excellent sensitivity and selectivity for NO with
26      detection limits on the order of 10 ppt for integration times of 1 min.  Spectroscopic methods
27      compare well with the CL method for NO in controlled laboratory air, ambient air, and heavily
28      polluted air (e.g., Walega et al.,  1984; Gregory et al., 1990; Kireev et al., 1999). These
29      spectroscopic methods remain in the research arena due to their complexity, size, and cost,  but
30      are essential for demonstrating that CL methods are reliable for monitoring NO  concentrations
31      involved in O3 formation — from 100's of ppb to around 20 ppt.

        August 2005                            AX2-149     DRAFT-DO NOT QUOTE OR CITE

-------
 1           Atmospheric pressure laser ionization followed by mass spectroscopy has also been
 2      reported for detection of NO and NO2. Garnica et al. (2000) describe a technique involving
 3      selective excitation at one wavelength followed by ionization at a second wavelength. They
 4      report good selectivity and detection limits well below 1 ppb. The practicality of the instrument
 5      for ambient monitoring has yet to be demonstrated.
 6
 7      AX2.6.2.3  Measurements of Nitrogen Dioxide
 8      Gas-Phase Chemiluminescence Methods
 9           Since the previous AQCD, photolytic reduction followed by CL has been improved and the
10      method of laser-induced fluorescence has been developed. Ryerson et al. (2000) developed a
11      photolytic converter based on a Hg lamp with increased radiant intensity in the region of peak
12      NO2 photolysis (350 to 400 nm) and producing conversion efficiencies  of 70% or more in less
13      than 1 s. Because the converter produces little radiation at wavelengths less than 350 nm,
14      interferences from HNO3 and PAN are minimal.
15           Alternative methods to photolytic reduction followed by CL are desirable to test the
16      reliability of this widely used technique. In any detector based on conversion to another species
17      interferences can be a problem.  Several atmospheric species, PAN and HO2NO2 for example,
18      dissociate to NO2 at higher temperatures.
19           Laser induced fluorescence for NO2 detection involves excitation  of atmospheric NO2 with
20      laser light emitted at wavelengths too long to induce photolysis.  The resulting excited molecules
21      relax in a photoemissive mode and the fluorescing photons are counted. Because collisions
22      would rapidly quench electronically excited NO2, the reactions are conducted at low pressure
23      (Cohen,  1999; Thornton et al., 2000; Day et al., 2002).  For example Cleary et al. (2002)
24      describe field tests of a system that uses continuous, supersonic expansion followed by
25      excitation at 640 nm with a commercial cw external-cavity tunable diode laser.  Sensitivity is
26      adequate for measurements in most continental environments (145 ppt in 1 min) and no
27      interferences have been identified.
28           Matsumi et al. (2001) describe a comparison of laser-induced fluorescence with a
29      photofragmentation chemiluminescence instrument. The laser-induced fluorescence system
30      involves excitation at 440 nm with a multiple laser system. They report sensitivity of 30 ppt in
31      10s and good agreement between the two methods under  laboratory conditions at mixing ratios

        August 2005                            AX2-150      DRAFT-DO NOT QUOTE OR CITE

-------
 1     up to 1.0 ppb. This high-sensitivity laser-induced fluorescence system has yet to undergo long-
 2     term field tests.
 3           NO2 can be detected by differential optical absorption spectroscopy (DOAS) in an open,
 4     long-path system (Kim and Kim, 2001). Vandaele et al. (2002) reported that the DOAS
 5     technique measured higher NO2 concentrations than were reported by other techniques in a
 6     three-year study conducted in Brussels.  Harder et al. (1997b) conducted an experiment in rural
 7     Colorado involving simultaneous measurements of NO2 with DOAS and photolysis followed by
 8     chemiluminescence.  The found differences of as much as 110% in clean air from the west, but
 9     for NO2 mixing ratios in excess of 300 ppt, the two methods agreed to better than  10%.  Stutz
10     and Platt (1996) report less uncertainty.
11
12     AX2.6.2.4  Monitoring for NO2 Compliance Versus Monitoring for Ozone Formation
13           Observations of NO2 have been focused on demonstrating compliance with the NAAQS for
14     NO2.  Today, few locations violate that standard, but NO2 and related NOy compounds remain
15     among the most important atmospheric trace gases to measure and understand.  Commercial
16     instruments for NO/NOX detection are generally constructed with an internal converter for
17     reduction of NO2 to NO, and generate a signal referred to as NOX.  These converters, generally
18     constructed of molybdenum oxides (MoOx), reduce not only NO2 but also most other NOy
19     species (Fehsenfeld et al., 1987; Crosley, 1996; Nunnermacker et al., 1998). Thus the NOX
20     signal is more accurately referred to as NOy. Unfortunately with an internal  converter, the
21     instruments may not give a faithful indication of NOy either — reactive species such as  HNO3
22     will adhere to the walls of the inlet system. Most recently, commercial vendors such as Thermo
23     Environmental (Franklin, MA) have offered NO/NOy detectors with external Mo converters.
24     If such instruments are calibrated through the inlet with a reactive nitrogen species such as
25     propyl nitrate, they should give accurate measurements of total NOy, suitable for evaluation of
26     photochemical models. States should be encouraged to make these NOy measurements where
27     ever possible.
28
29     AX2.6.3  Measurements of Nitric Acid Vapor, HNO3
30           Accurate measurement of nitric acid vapor, HNO3,  has presented a long-standing analytical
31     challenge to the atmospheric chemistry community.  In this context, it is useful to consider the

       August 2005                            AX2-151     DRAFT-DO NOT QUOTE  OR CITE

-------
 1      major factors that control HNO3 partitioning between the gas and deliquesced-particulate phases
 2      in ambient air. In equation form,
 3
 4                                  K            K
 5                          HN03g d [HN03aq] ^  [H+] + [N03-]                    (AX2-52)
 6
 7      where KH is the Henry's Law constant in M atm-1 and Ka is the acid dissociation constant in M.
 8           Thus, the primary controls on HNO3 phase partitioning are its thermodynamic properties
 9      (KH, Ka, and associated temperature corrections), aerosol liquid water content (LWC), solution
10      pH, and kinetics. Aerosol LWC and pH are controlled by the relative mix of different acids and
11      bases in the system, hygroscopic properties of condensed compounds, and meteorological
12      conditions (RH, temperature, and pressure).  It is evident from relationship XX that, in the
13      presence of chemically distinct aerosols of varying acidities (e.g., super-Jim predominantly sea
14      salt and sub-jam predominantly S aerosol), HNO3 will partition preferentially with the less-acidic
15      particles, which is consistent with observations (e.g., Huebert et al., 1996; Keene and Savoie,
16      1998;  Keene et al.,  2002). Kinetics are controlled by atmospheric concentrations of HNO3 vapor
17      and particulate NO3 and the size distribution and corresponding atmospheric lifetimes of
18      particles against deposition.  Sub-jim-diameter aerosols typically equilibrate with the gas phase
19      in seconds to minutes while super-um aerosols require hours to a day or more (e.g., Meng and
20      Seinfeld, 1996; Erickson et al., 1999. Consequently, smaller aerosol size fractions are typically
21      close to thermodynamic equilibrium  with respect to HNO3 whereas larger size fractions (for
22      which atmospheric lifetimes against deposition range from hours to a few days) are often
23      undersaturated (e.g., Erickson et al.,  1999; Keene and Savioe, 1998).
24           Many sampling techniques for HNO3 (e.g., standard filterpack and mist-chamber samplers)
25      employ upstream prefilters to remove particulate species from sample air. However, when
26      chemically distinct aerosols with different pHs (e.g., sea salt and S aerosols) mix together on a
27      bulk filter, the acidity of the bulk mixture will be greater than that of the less acidic aerosols with
28      which most NO3 is associated.  This change in pH may  cause the bulk mix to be supersaturated
29      with respect to HNO3 leading to volatilization and, thus,  positive measurement bias in HNO3
30      sampled downstream.  Alternatively, when undersaturated super-um size fractions (e.g., sea salt)
31      accumulate on a bulk filter and chemically interacts over time with HNO3 in the  sample air

        August 2005                             AX2-152      DRAFT-DO NOT QUOTE OR CITE

-------
 1      stream, scavenging may lead to negative bias in HNO3 sampled downsteam.  Because the
 2      magnitude of both effects will vary as functions of the overall composition and thermodynamic
 3      state of the multiphase system, the combined influence can cause net positive or net negative
 4      measurement bias in resulting data.  Pressure drops across particle filters can also lead to artifact
 5      volatilization and associated positive bias in HNO3 measured downstream.
 6          Widely used methods for measuring HNO3 include standard filterpacks configured with
 7      nylon or alkaline-impregnated filters (e.g., Goldan et al., 1983; Bardwell et al., 1990;
 8      respectively) and standard mist chambers (Talbot et al., 1990). Samples are typically analyzed
 9      by ion chromatography. Intercomparisons of these measurement techniques (e.g., Hering et al.,
10      1988; Tanner et al.,  1989; Talbot et al., 1990) report differences of a factor of two or more.
11          More recently, sensitive HNO3 measurements based on the principle of Chemical
12      lonization Mass  Spectroscopy (CIMS) have been reported (e.g., Huey et al., 1998; Mauldin
13      et al., 1998; Furutani and Akimoto, 2002; Neuman et al., 2002). CIMS relies on selective
14      formation of ions such as SiF5"-HNO3 or HSO4"-HNO3 followed by detection via mass
15      spectroscopy.  Two  CIMS techniques and a filter pack technique were intercompared in Boulder,
16      CO (Fehsenfeld et al.,  1998). Results indicated excellent agreement (within 15%) between the
17      two CIMS instruments and between the CIMS and filterpack methods under relatively clean
18      conditions with HNO3 mixing ratios between 50 and 400 pptv. In more polluted air, the
19      filterpack technique generally yielded higher values than the CIMS  suggesting that interactions
20      between chemically distinct particles on bulk filters is a more important source of bias in
21      polluted continental air. Differences were also greater at lower temperature when particulate
22      NO3 corresponded to relatively greater fractions of total NO3 .
23
24      AX2.6.4   Sampling and Analysis of Volatile Organic Compounds
25          Hydrocarbons can be measured with gas chromatography followed by flame ionization
26      detection (GC-FID). Detection by mass spectroscopy is sometimes used to confirm species
27      identified by retention time (Westberg and Zimmerman, 1993; Dewulf and Van Langenhove,
28      1997). Preconcentration is typically required for less abundant species.  Details are available in
29      AQCD 96.
30          Because of their variety, nonmethane hydrocarbons pose special analytical problems,
31      and several laboratory and field studies have recently addresses the uncertainty of VOC

        August 2005                           AX2-153      DRAFT-DO NOT QUOTE OR CITE

-------
 1      measurements.  An intercomparison conducted with 16 components among 28 laboratories,
 2      showed agreement on the order of 10s of percents (Apel et al., 1994). In a more recent
 3      intercomparison (Apel et al., 1999) 36 investigators from around the world were asked to
 4      identify and quantify C2 to C10 hydrocarbons (HCs) in a mixture in synthetic air.  Calibration was
 5      based on gas standards of individual compounds, such as propane in air, and a 16-compound
 6      mixture of C2 to C16 -alkanes, all prepared by NIST and certified to ± 3 percent. The
 7      top-performing laboratories, including several in the United States, identified all the compounds
 8      correctly, and obtained agreement of generally better than 20 percent for the 60 compounds.
 9      Intercomparison of NMHCs in ambient air has only recently been reported by a European group
10      of 12-14 laboratories (Slemr et al., 2002). Some compounds gave several groups difficulties,
11      including isobutene, butadiene, methyl pentanes, and trimethyl benzenes.  These
12      intercomparisons illustrated the need for reliable, multicomponent calibration standards.
13
14      AX2.6.4.1  Polar Volatile Organic Compounds
15          Many of the more reactive oxygen- and nitrogen-containing organic compounds play a role
16      in O3 formation and are included among list of 189 hazardous air pollutants specified in the 1990
17      CAAA (U.S. Congress, 1990).  These compounds are emitted directly from a variety of sources
18      including biogenic processes, biomass burning, industry, vehicles, and consumer products.
19      Some can also be formed in the atmosphere by photochemical oxidation of hydrocarbons.
20      Although these  compounds have been referred to collectively as PVOCs, their reactivity and
21      water solubility, rather than just polarity, make sampling and measurement challenging. As
22      indicated in the earlier AQCD, few ambient data exist for these species, but that database has
23      grown. The previous AQCD discusses two analytical methods for PVOCs — cryogenic trapping
24      techniques similar to those used for the nonpolar hydrocarbon species, and adsorbent material
25      for sample preconcentration. Here we discuss recently developed methods.
26          Several techniques for sampling, preconcentrating and detecting oxygenated volatile
27      organic compounds were inter-compared during the 1995 Southern Oxidants Study Nashville
28      Intensive (Apel et al., 1998). Both chemical traps and derivatization followed by HPLC and
29      pre-concentration and gas chromatography followed by mass spectrometric of flame ionization
30      were investigated. Both laboratory and field tests were conducted for formaldehyde,
31      acetaldehyde, acetone, and propanal. Substantial differences were observed indicating that

        August 2005                            AX2-154     DRAFT-DO NOT QUOTE OR CITE

-------
 1     reliable sampling and measurement of PVOCs remains an analytical challenge and high research
 2     priority.
 3          Chemical ionization-mass spectroscopy, such as proton-transfer-reaction mass
 4     spectroscopy (PTR-MS) can also be used for fast-response measurement of volatile organic
 5     compounds including acetonitrile (CH3CN), methanol (CH3OH), acetone (CH3COCH3),
 6     acetaldehyde (CH3CHO), benzene (C6H6) and toluene (C6H5CH3) (e.g., Hansel et al., 1995a,b;
 7     Lindinger et al., 1998; Leibrock and Huey, 2000; Warneke et al., 2001). The method relies on
 8     gas phase proton transfer reactions between H3O+ primary ions and volatile trace gases with a
 9     proton affinity higher than that of water. Into a flow drift tube continuously flushed with
10     ambient air, H3O+ ions (from a hollow cathode ion source) are injected.  On collisions between
11     H3O+ ions and organic molecules protons H+  are transferred thus charging the reagent.  Both
12     primary and product ions are analyzed in a quadrupole mass spectrometer and detected by a
13     secondary electron multiplier/pulse counting  system. The instrument has been successfully
14     employed in several field campaigns and compared to other techniques including gas
15     chromatography  and Atmospheric Pressure Chemical lonization Mass Spectrometer (AP-CIMS)
16     (Crutzen et al., 2000; Sprung et al., 2001).  Sufficient sensitivity was observed for urban and
17     rural measurements; no interferences were  discovered, although care must be exercised to avoid
18     sampling losses.  Commercial  instruments  are becoming available, but their price still precludes
19     widespread monitoring.
       August 2005                            AX2-155      DRAFT-DO NOT QUOTE OR CITE

-------
  1       REFERENCES

  2       Allan, B. I; McFiggans, G.; Plane, J. M. C. (2000) The nitrate radical in the remote marine boundary layer.
  3             J. Geophys. Res. [Atmos.] 105(D19): 24,191-24,204.
  4       Allan, B. J.; Plane, J. M. C.; McFiggans, G. (2001) Observations of OIO in the remote marine boundary layer.
  5             Geophys. Res. Lett. 28: 1945-1948.
  6       Allen, D. J.; Pickering, K. E. (2002) Evaluation of lightning flash rate parameterizations for use in a global chemical
  7             transport model. J. Geophys. Res. [Atmos.] 107(D23): 10.1029/2002JD002066.
  8       Allen, D. J.; Pickering, K. E.; Molod, A. (1997) An evaluation of deep convective mixing in the Goddard Chemical
  9             Transport Model using ISCCP cloud parameters. J. Geophys. Res. [Atmos.] 102: 25,467-25,476.
10       Allen, D.; Pickering, K.; Stenchikov, G.; Thompson, A.; Kondo, Y. (2000) A three-dimensional total odd nitrogen
11             (NOy) simulation during SONEX using a stretched-grid chemical transport model. J. Geophys.  Res. [Atmos.]
12             105: 3851-3876.
13       Altshuller, A. P. (1986) The role of nitrogen oxides in nonurban ozone formation in the planetary boundary layer
14             over N America, W Europe and adjacent areas of ocean. Atmos. Environ.  20: 245-268.
15       Altshuller, A. P.; Lefohn, A.  S. (1996) Background ozone in the planetary boundary layer over the United States.
16             J. Air Waste Manage.  Assoc. 46: 134-141.
17       Alvarez, II; Senff, C. J.; Hardesty, R. M.; Parrish, D. D.; Luke, W. T.; Watson, T. B.; Daum, P. H.; Gillani, N.
18             (1998) Comparisons of airborne lidar measurements of ozone with airborne in situ measurements during the
19             1995 SouthernOxidants study. J. Geophys. Res.  [Atmos.] 103: 31155-35171.
20       American Society for Testing and Materials (ASTM) International. (2003a) Standard test method for ozone in the
21             atmosphere: continuous measurement by ethylene chemiluminescence. In: Annual Book of ASTM Standards
22             2003, v.  11.03: D 5149-02. West Conshohocken, PA: American Society for Testing and Materials;
23             pp. 443-447.
24       American Society for Testing and Materials (ASTM) International. (2003b) Standard test methods for continuous
25             measurement of ozone in ambient, workplace, and indoor atmospheres (ultraviolet absorption). In: Annual
26             Book of ASTM Standards 2003, v. 11.03: D 5156-02. West Conshohocken, PA: American Society for
27             Testing and Materials; pp. 448-454.
28       Ammann, M.; Kalberer, M.; Jost, D. T.; Tobler, L.; Rossler, E.; Piguet, D.; Gaggeler, H. W.; Baltensperger, U.
29             (1998) Heterogeneous production of nitrous acid on soot in polluted air masses. Nature (London) 395:
30             157-160.
31       Anastasio, C.; Faust, B. C.; Allen, J. M. (1994) Aqueous phase photochemical formation of hydrogen peroxide in
32             authentic cloud waters. J. Geophys. Res.  [Atmos.] 99: 8231-8248.
33       Anastasio, C.; Faust, B.C.; Rao, C. J. (1997) Aromatic carbonyl compounds as aqueous-phase photochemical
34             sources of hydrogen peroxide in acidic sulfate aerosols, fogs, and clouds. 1. Non-phenolic
35             methoxybenzaldehydes and methoxyacetophenones with reductants (phenols). Environ. Sci. Technol.
36             31:218-232.
37       Anastasio, C.; Newberg, J. T.; Williams, D. K.; Chu, G. B.; Mathew,  B. M. (1999) Photoformation of hydroxyl
38             radical in sea salt particles. EOS Trans. Am. Geophys. Union 80: F147.
39       Andino, J. M.; Smith, J. N.; Flagan, R. C.; Goddard, W. A., Ill; Seinfeld, J. H. (1996) Mechanism of the atmospheric
40             photooxidation of aromatics: a theoretical study.  J. Phys. Chem. 100: 10,967-10,980.
41       Andreae, M. O. (1991) Biomass burning: its history, use, and distribution and its impact on environmental quality
42             and global climate. In: Levine, J. S., ed. Global biomass burning: atmospheric, climatic, and biospheric
43             implications. Cambridge, MA: MIT Press; pp.  1-21.
44       Andres Hernandez, M. D.; Burkert, J.; Reichert, L.; Stobener, D.; Meyer-Arnek, J.; Dickerson, R. R.;  Doddridge,
45             E.G.; Burrows, J. P. (2001) Marine boundary layer peroxy radical chemistry during the AEROSOLS99
46             campaign: measurements and analysis. J. Geophys. Res. (Atmos.) 106(D18): 20,833-20,846.
47       Andronache, C.; Chameides, W. L.; Rodgers, M. O.; Martinez, J. E.; Zimmerman, P.; Greenberg, J. (1994) Vertical
48             distribution of isoprene in the lower boundary layer of the rural and urban southern United States. J. Geophys.
49             Res. [Atmos.]) 99: 16,989-17,000.
50       Angevine, W. M.; Buhr, M. P.; Holloway, J.  S.; Trainer, M.; Parrish,  D. D.; MacPherson, J. I.; Kok, G. L.;
51             Schillawski, R. D.; Bowlby, D. H.  (1996) Local meteorological features affecting chemical measurements at a
52             North Atlantic coastal site. J.  Geophys. Res. (Atmos.) 101(D22): 28,935-28,946.
53       Apel, E. C.; Calvert, J. G.; Fehsenfeld, F. C. (1994) The nonmethane  hydrocarbon intercomparison experiment
54             (NOMHICE): tasks 1 and 2. J. Geophys. Res. [Atmos.] 99(D8): 16,651-16,664.
         August 2005                                 AX2-156      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Apel, E. C.; Calvert, J. G.; Riemer, D.; Pos, W.; Zika, R.; Kleindienst, T. E.; Lonneman, W. A.; Fung, K.; Fujita, E.;
  2             Shepson, P. B.; Starn, T. K.; Roberts, P. T. (1998) Measurements comparison of oxygenated volatile organic
  3             compounds at a rural site during the 1995 SOS Nashville Intensive. J. Geophys. Res. (Atmos.)
  4             103(D17): 22,295-22,316.
  5       Apel, E. C.; Calvert, J. G.; Gilpin, T. M; Fehsenfeld, F. C.; Parrish, D. D.; Lonneman, W. A. (1999) The
  6             Nonmethane Hydrocarbon Intercomparison Experiment (NOMHICE): task 3. J. Geophys. Res. [Atmos.]
  7             104(021): 26,069-26,086.
  8       Appenzeller, C.; Davies, H. C. (1992) Structure of stratospheric intrusions into the troposphere. Nature
  9             358: 570-572.
10       Appenzeller, C.; Davies, H. C.; Norton, W. A. (1996) Fragmentation of stratospheric intrusions. J. Geophys. Res.
11             [Atmos.] 101: 1435-1456.
12       Arnold, F.; Burger, V.; Droste-Fanke, B.; Grimm, F.; Krieger, A.; Schneider, J.; Stilp, T. (1997) Acetone in the
13             upper troposphere and lower stratosphere: impact on trace gases and aerosols. Geophys. Res. Lett.
14             24:3017-3020.
15       Arnold, S. R.; Chipperfield, M. P.; Blitz, M. A.; Heard, D. E.; Pilling, M. J. (2004) Photodissociation of acetone:
16             atmospheric implications of temperature-dependent quantum yields. Geophys. Res. Lett.
17             31(L07110): 10.1029/2003GL019099.
18       Arshinov, M. Y.; Belan, B. D.; Krasnov, O. A.; Kovalevskii, V. K.; Pirogov, V. A.; Plotnikov, A. P.; Tolmachev,
19             G. N.; Fofonov, A. V. (2002) Comparison of ultraviolet and chemiluminescent ozonometers. Atmos. Oceanic
20             Opt. 15: 656-658.
21       Ashworth, S. H.;  Allan, B. J.; Plane, J. M. C. (2002) High resolution spectroscopy of the OIO radical: implications
22             for the ozone-depleting potential of iodine. Geophys. Res. Lett. 29: 65-1 - 65-4.
23       Atherton, C.  S.; Penner, J.  E. (1988) The transformation of nitrogen oxides in the polluted troposphere. Tellus B
24             40:380-392.
25       Atkinson, R. (1989) Kinetics and mechanisms of the gas-phase reactions of the hydroxyl radical with organic
26             compounds. Washington, DC: American Chemical Society. (J. Phys. Chem. Ref. Data Monograph no. 1).
27       Atkinson, R. (1990) Gas-phase tropospheric chemistry of organic compounds: a review. Atmos. Environ. Part A
28             24: 1-41.
29       Atkinson, R. (1991) Kinetics and mechanisms of the gas-phase reactions of the NOX radical with organic
30             compounds. J. Phys. Chem. Ref. Data 20: 459-507.
31       Atkinson, R. (1994) Gas-phase tropospheric chemistry of organic compounds. Washington, DC: American Chemical
32             Society. (J. Phys. Chem. Ref. Data Monograph no. 2).
33       Atkinson, R. (1997) Atmospheric reactions of alkoxy and p-hydroxyalkoxy radicals. Int. J. Chem. Kinet. 29: 99-111.
34       Atkinson, R. (2000) Atmospheric chemistry of VOCs and NOX. Atmos. Environ. 34: 2063-2101.
35       Atkinson, R.; Arey, J. (2003) Gas-phase tropospheric chemistry of biogenic volatile organic compounds: a review.
36             Atmos. Environ.: 10.1016/81352-2310(03)00391-1.
37       Atkinson, R.; Aschmann, S. M. (1994) Products of the gas-phase reactions of aromatic hydrocarbons: effect of NO2
38             concentration. Int. J. Chem. Kinet. 26: 929-944.
39       Atkinson, R.; Carter, W. P. L.; Darnall, K. R.; Winer, A. M.; Pitts, J. N., Jr. (1980) A smog chamber and modeling
40             study of the gas phase NOx-air photooxidation of toluene and the cresols. Int.  J. Chem. Kinet. 12: 779-836.
41       Atkinson, R.; Winer, A. M.; Pitts, J. N., Jr. (1986) Estimation of night-time N2O5 concentrations from ambient NO2
42             and NO3 radical concentrations and the role of N2O5 in night-time chemistry. Atmos. Environ. 20: 331-339.
43       Atkinson, R.; Aschmann, S. M.; Arey, J.; Carter, W. P. L. (1989) Formation of ring-retaining products from the OH
44             radical-initiated reactions of benzene and toluene. Int. J. Chem. Kinet. 21: 801-827.
45       Atkinson, R.; Arey, J.; Zielinska,  B.; Aschmann, S. M. (1990) Kinetics and nitro-products of the gas-phase OH and
46             NO3 radical-initiated reactions of naphthalene-d8, fluoranthene-dlO and pyrene. Int. J. Chem. Kinet.
47             22:999-1014.
48       Atkinson, R.; Baulch, D. L.;  Cox, R. A.; Hampson, R. F., Jr.; Kerr, J. A.; Troe, J. (1992a) Evaluated kinetic and
49             photochemical data for atmospheric chemistry: supplement IV, IUPAC Subcommittee on Gas Kinetic Data
50             Evaluation for Atmospheric Chemistry. J. Phys. Chem. Ref. Data 21:  1125-1568.
51       Atkinson, R.; Aschmann, S. M.; Arey, J.; Shorees, B. (1992b) Formation of OH radicals in the gas phase reactions
52             of O3 with a series of terpenes.  J. Geophys. Res. [Atmos.] 97: 6065-6073.
53       Atkinson, R.; Kwok, E. S.  C.; Arey, J.; Aschmann, S. M. (1995) Reactions of alkoxyl radicals in the atmosphere.
54             Faraday Discuss. 100: 23-37.
         August 2005                                 AX2-157       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Atkinson, R.; Baulch, D. L.; Cox, R. A.; Hampson, R. F., Jr.; Kerr, J. A.; Rossi, M. I; Troe, J. (1999) Evaluated
  2             kinetic and photochemical data for atmospheric chemistry, organic species: supplement VII. J. Phys. Chem.
  3             Ref. Data 28:  191-393.
  4       Ayers, G. P.; Penkett, S. A.; Gillett, R. W.; Bandy, B.; Galbally, I. E.; Meyer, C. P.; Elsworth, C. M.; Bentley, S. T.;
  5             Forgan, B. W. (1992) Evidence for photochemical control of ozone concentrations in unpolluted marine air.
  6             Nature (London) 360: 446-449.
  7       Ayers, G. P.; Gillett, R. W.; McAiney, J. M.; Dick, A. L. (1999) Chloride and bromide loss from sea-salt particles in
  8             the southern ocean air. J. Atmos. Chem. 33: 299-319.
  9       Bais, A. F.; Madronich, S.;  Crawford, J.; Hall, S. R.; Mayer, B.; Van Weele, M.; Lenoble, J.; Calvert, J. G.; Cantrell,
10             C. A.; Shelter, R. E.; Hofzumahaus, A.; Koepke, P.; Monks, P. S.; Frost, G.; McKenzie, R.; Krotkov, N.;
11             Kylling, A.; Swartz,  W.  H.; Lloyd, S.; Pfister, G.; Martin, T. J.; Roeth, E.-P.; Griffioen, E.; Ruggaber, A.;
12             Krol, M.; Kraus, A.;  Edwards, G. D.; Mueller, M.; Lefer, B. L. (2003) International Photolysis Frequency
13             Measurement and Model Intercomparison (IPMMI):  Spectral actinic solar flux measurements and modeling.
14             J. Geophys. Res. [Atmos.] 108(D16): 10.1029/2002JD002891.
15       Balkanski, Y. J.; Jacob, D. J.; Gardener, G. M.; Graustein, W. C.; Turekian, K. K. (1993) Transport and residence
16             times of tropospheric aerosols inferred  from a global three-dimensional simulation of 210Pb. J. Geophys. Res.
17             [Atmos.] 98: 20,573-20,586.
18       Bandow, H.; Washida, N. (1985a)  Ring-cleavage reactions of aromatic hydrocarbons studied by FT-IR
19             spectroscopy. II photooxidation of o-, m-, and^-xylenes in the NOx-air system. Bull. Chem. Soc. Jpn.
20             58:2541-2548.
21       Bandow, H.; Washida, N. (1985b)  Ring-cleavage reactions of aromatic hydrocarbons studied by FT-IR
22             spectroscopy. Ill photooxidation of 1,2,3-, 1,2,4-, and 1,3,5-trimethylbenzene in the NOx-air system. Bull.
23             Chem. Soc. Jpn. 58: 2549-2555.
24       Banic, C. M.; Leaitch, W. R.; Isaac, G. A.; Couture, M. D.;  Kleinman, L. I.; Springston, S. R.; MacPherson, J. I.
25             (1996) Transport of ozone and sulfur to the North Atlantic atmosphere during the North Atlantic Regional
26             Experiment. J. Geophys. Res. [Atmos.] 101: 29,091-29,104.
27       Banta, R. M.; Senff, C. J.; White, A. B.; Trainer, M.; McNider, R.  T.; Valente, R. J.; Mayor, S. D.; Alvarez, R. J.;
28             Hardesty, R. M.; Parrish, D.; Fehsenfeld, F. C. (1998) Daytime buildup and nighttime transport of urban
29             ozone in the boundary layer during a stagnation episode. J. Geophys. Res.  [Atmos.] 103: 22,519-22,544.
30       Bardwell, C. A.; Maben, J. R.;  Hurt, J. A.; Keene, W. C.; Galloway, J. N.; Boatman, J. F.; Wellman, D.  (1990)
31             A technique using high-flow, dichotomous filter packs for measuring major atmospheric chemical
32             constituents. Glob. Biogeochem. Cycles 4: 151-163.
33       Barrie, L. A.; Bottenheim, J. W.; Crutzen, P. J.; Rasmussen, R. A.  (1988) Ozone destruction at polar sunrise in the
34             lower Arctic atmosphere. Nature 334: 138-141.
3 5       Bartolotti, L. J.; Edney, E. O. (1995) Density functional theory derived intermediates from the OH initiated
36             atmospheric oxidation of toluene. Chem. Phys. Lett. 245: 119-122.
3 7       Baugues, K. (1986) A review of NMOC, NOX and NMOC/NOX ratios measured in 1984 and 1985. Research
3 8             Triangle Park, NC: U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards;
39             report no. EPA-450/4-86-015. Available from: NTIS, Springfield, VA; PB87-166963/HSU.
40       Baumann, K.; Williams, E.  J.; Angevine, W. M.; Roberts, J. M.; Norton, R. B.; Frost, G. J.; Fehsenfeld, F. C.;
41             Springston, S. R.; Bertman,  S. B.; Hartsell, B. (2000) Ozone production and transport near Nashville,
42             Tennessee: results from the  1994 study at New Hendersonville. J. Geophys. Res. [Atmos.]
43             105(D7): 9137-9153.
44       Behnke, W.; George, C.; Sheer, V.; Zetzsch, C. (1997) Production and decay of C1NO2 from the reaction of gaseous
45             N2O5 withNaCl solution: bulk and aerosol experiments. J. Geophys. Res. 102: 3795-3804.
46       Bell, G.  B. (1960) Meteorological  conditions during oxidant episodes in coastal San Diego County in October and
47             November, 1959. Sacramento, CA: State of California, Department of Public Health.
48       Benjamin,  S. G.; Carlson, T. N. (1986) Some effects of surface heating and topography on the region severe storm
49             environment. Part I: three dimensional  simulations. Mon. Weather Rev.  114: 307-329.
50       Bergin, M. H.; Pandis, S. N.; Davidson, C. I.;  Jaffrezo, J.-L.; Dibb, J. E.; Russell, A. G.; Kuhns, H. D. (1996)
51             Modeling of the processing  and removal of trace gas and aerosol  species by Arctic radiation fogs and
52             comparison with measurements. J. Geophys. Res. [Atmos.]  101: 14,465-14,478.
53       Berkowitz, C. M.; Shaw, W. J. (1997) Airborne measurements of boundary layer chemistry during the Southern
54             Oxidant Study: a case study. J. Geophys. Res. [Atmos.] 102: 12,795-12,804.
         August 2005                                 AX2-158       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Berkowitz, C. M; Fast, J. D.; Sprinston, S. R.; Larsen, R. J.; Spicer, C. W.; Doskey, P. V.; Hubbe, J. M;
  2             Plastridge, R. (1998) Formation mechanisms and chemical characteristics of elevated photochemical layers
  3             over the northeast United States. J. Geophys. Res. [Atmos.] 103:  10,631-10,647.
  4       Berkowitz, C. M.; Fast, J. D.; Easter, R. C. (2000) Boundary layer vertical exchange processes and the mass budget
  5             of ozone:  observations and model results. J.  Geophys. Res. [Atmos.] 105: 14,789-14,805.
  6       Berkowitz, C. M.; Zaveri, R. A.; Bian, X.; Zhong, S.; Disselkamp, R. S.; Laulainen, N. S.; Chapman, E. G. (2001)
  7             Aircraft observations of aerosols, O3 and NO,, in a nighttime urban plume. Atmos. Environ. 35: 2395-2404.
  8       Berndt, T.; Boge, O.; Herrmann, H. (1999) On the formation of benzene oxide/oxepin in the gas-phase reaction of
  9             OH radicals with benzene.  Chem. Phys. Lett. 314: 435-442.
10       Berndt, T.; Boge, O.; Stratmann, F. (2003) Gas-phase ozonolysis of a-pinene: gaseous products and particle
11             formation. Atmos. Environ. 37: 3933-3945.
12       Berntsen, T.; Isaksen, I. S. A. (1997) A global three-dimensional chemical transport model for the troposphere.
13             1. Model description and CO and ozone results. J. Geophys. Res. [Atmos.] 102: 21,239-21,280.
14       Bey, I.; Jacob, D. J.; Logan, J. A.; Yantosca, R. M. (2001a) Asian chemical outflow to the Pacific in spring: origins,
15             pathways, and budgets. J. Geophys. Res. [Atmos.] 106: 23,097-23,113.
16       Bey, I.; Jacob, D. J.; Yantosca, R. M.; Logan, J. A.; Field, B.; Fiore, A. M.; Li, Q.; Liu, H.; Mickley, L. J.; Schultz,
17             M. G. (200Ib) Global modeling of tropospheric chemistry with assimilated meteorology: model description
18             and evaluation. J. Geophys. Res. [Atmos.] 106: 23,073-23,095.
19       Bierbach, A.; Barnes, I.; Becker, K. H.; Wiesen, E. (1994) Atmospheric chemistry of unsaturated carbonyls:
20             butenedial, 4-Oxo-2-Pentenal, 3-Hexene-2,5-Dione, Maleic Anhydride, 3H-Furan-2-one, and
21             5-Methyl-3H-Furan-2-one. Environ. Sci. Technol. 28: 715-729.
22       Bishop, G. A.; Stedman, D. H. (1990) On-road carbon monoxide emission measurement comparisons for the
23             1988-1989 Colorado oxy-fuels program. Environ. Sci. Technol. 24: 843-847.
24       Bishop, G. A.; Stedman, D. H. (1996) Measuring the emissions of passing cars.  Ace. Chem. Res. 29: 489-495.
25       Black, D. R.; Harley, R. A.; Hering, S. V.; Stolzenburg, M. R. (2000) A new, portable, real-time monitor. Environ.
26             Sci. Technol. 34:3031-3040.
27       Blanchard, C. L. (2000) Ozone process insights from field experiments.  Part III: extent of reaction and ozone
28             formation. Atmos. Environ. 34: 2035-2043.
29       Blanchard, C. L.; Stoeckenius, T.  (2001) Ozone response to precursor control: comparision of data analysis methods
30             with the predictions of photochemical air quality simulation models. Atmos. Environ. 35: 1203-1215.
31       Blanchard, C. L.; Lurmann, F. W.; Roth, P. M.; Jeffries, H. E.; Korc, M. (1999) The use of ambient data to
32             corroborate analyses of ozone control strategies. Atmos. Environ. 33: 369-381.
33       Blanchard, P.; Brook, J. R.; Brazil, P. (2002) Chemical characterization of the organic fraction of atmospheric
34             aerosol at two sites in Ontario, Canada. J. Geophys. Res. [Atmos.] 107(D21):  10.1029/2001JD000627.
35       Bloom, S. C.; Takacs, L. L.; Da Silva, A. M.; Ledvina, D. (1996) Data assimilation using incremental analysis
36             updates. Mon. Weather Rev. 124: 1256-1271.
37       Blough, N. V. (1997) Photochemistry in the sea-surface microlayer. In: Liss, P.  S.; Duce, R. A., eds. The sea surface
38             and global change. New York, NY: Cambridge University Press, pp. 383-424.
39       Blough, N. V.; Zepp. R. G. (1995) Reactive oxygen species in natural waters. In: Foote, C. S.; Valentine, A.;
40             Greenberg, A.; Liebman, J. F., eds. Active Oxygen in Chemistry. New York, NY: Chapman and Hall, Ch.8,
41             pp. 280-333.
42       Blumenthal, D. L.; Lurmann, F. W.; Kumar, N.; Dye, T. S.; Ray, S. E.; Korc, M. E.;  Londergan, R.; Moore, G.
43             (1997) Transport and mixing phenomena related to ozone exceedances in the northeast U.S. (analysis based
44             on NARSTO-northeast data). Available:  http://capita.wusti.edu/otag/reports/otagrept/otagrept.html
45             (30 October 2003).
46       Bonn, B.; Moortgat, G. K. (2003)  Sesquiterpene ozonolysis: origin of atmospheric new particle formation from
47             biogenic hydrocarbons. Geophys. Res. Lett.  30: 10.1029/2003GL017000.
48       Bonner, W. D. (1968) Climatology of the low level jet. Mon. Weather Rev. 96: 833-850.
49       Bottenheim, J. W.; Barrie, L. A.; Atlas, E.; Heidt, L. E.; Niki, H.; Rasmussen, R. A.; Shepson, P. B. (1990)
50             Depletion of lower tropospheric ozone during Arctic spring: The  Polar Sunrise Experiment 1988. J. Geophys.
51             Res. (Atmos.) 95: 18,555-18,568.
52       Brasseur, G. P.; Hauglustaine, D. A.; Walters, S.; Rasch, P. J.; Mtiller, J.-F.; Granier, C.; Tie,  X. X. (1998)
53             MOZART, a global transport model for ozone and related chemical tracers: 1. model description. J. Geophys.
54             Res. [Atmos.]  103: 28,265-28,289.
55       Brook, J. R.; Sirois, A.; Clarke, J.  F. (1996) Comparison of dry deposition velocities for SO2, HNO3, and SO42"
56             estimated with two inferential models.  Water Air Soil Pollut. 87:  205-218.
         August 2005                                 AX2-159      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Broske, R.; Kleffmann, J.; Wiesen, P. (2003) Heterogeneous conversion of NO2 on secondary organic aerosol
  2             surfaces: a possible source of nitrous acid (HONO) in the atmosphere? Atmos. Chem. Phys. 3: 469-474.
  3       Browell, E. V.; Fenn, M. A.; Butler, C. F.; Grant, W. B.; Ismail, S.; Ferrare, R. A.; Kooi, S. A.; Brackett, V. G.;
  4             Clayton, M. B.; Avery, M. A.; Barrick, J. D. W.; Fuelberg, H. E.; Maloney, J. C.; Newell, R. E.; Zhu, Y.;
  5             Mahoney, M. J.; Anderson, B. E.; Blake, D. R.; Brune, W. H.; Heikes, B. G.;  Sachse, G. W.; Singh, H. B.;
  6             Talbot, R. W. (2001) Large-scale air mass characteristics observed over the remote tropical Pacific Ocean
  7             during March-April 1999: results from PEM-Tropics B field experiment. J. Geophys. Res. [Atmos.]
  8             106: 32,481-32,501.
  9       Brueggemann, N.; Schnitzler, J.-P. (2002) Comparison of isoprene emission, intercellular isoprene concentration and
10             photo synthetic performance with water-limited oak (Quercus pubescens Willd. and Quercus robur L.)
11             saplings. Plant Biol. 4: 456-463.
12       Briihl, C.; Poschl, U.; Crutzen, P. J.; Steil, B. (2000) Acetone and PAN in the upper troposphere: impact on ozone
13             production from aircraft emissions. Atmos. Environ. 34: 3931-3938.
14       Brune, W. H.; Tan, D.; Faloona, I. F.; Jaegle, L.; Jacob, D. J.; Heikes, B. G.; Snow, J.; Kondo, Y.; Shelter, R.;
15             Sachse, G. W.; Anderson, B.; Gregory, G. L.; Vay, S.; Singh, H. B.; Davis, D. D.; Crawford, J. H.; Blake, D.
16             R. (1999) OH and HO2 chemistry in the North Atlantic free troposphere. Geophys. Res. Lett. 26: 3077-3080.
17       Brunner,D.; Staehelin, I;  Jeker, D. (1998) Large-scale nitrogen oxide plumes in the tropopause region and
18             implications for ozone. Science (Washington, DC) 282: 1305-1309.
19       Buhr, M. P.;  Trainer, M.; Parrish, D. D.; Sievers, R. E.; Fehsenfeld, F. C. (1992) Assessment of pollutant emission
20             inventories by principal component analysis of ambient air measurements. Geophys.  Res. Lett.
21             19:  1009-1012.
22       Buhr, M.; Parrish, D.; Elliot, J.; Holloway, J.; Carpenter, J.; Goldan, P.; Kuster, W.; Trainer, M.; Montzka, S.;
23             McKeen, S.; Fehsenfeld, F. C. (1995) Evaluation of ozone precursor source types using principal component
24             analysis of ambient air measurements in rural Alabama. J. Geophys. Res. [Atmos.] 100: 22,853-22,860.
25       Buhr, M.; Sueper, D.; Trainer, M.; Goldan, P.; Kuster, B.; Fehsenfeld, F. (1996) Trace gas and aerosol
26             measurements using aircraft data from the North Atlantic Regional Experiment (NARE 1993). J. Geophys.
27             Res. [Atmos.] 101: 29,013-29,027.
28       Burkert, J.; Behmann, T.; Andres-Hernandez, M. D.; Stobener, D.; WeiBenmeyer, M.; Perner, D.; Burrows, J. P.
29             (200 la) Measurements of peroxy radicals in a forested area of Portugal. Chemosphere Glob. Change Sci.
30             3:327-338.
31       Burkert, J.; Andres-Hernandez, M. D.; Stobener, D.; Burrows, J. P.; Weilknmeyer, M.; Kraus, A. (200Ib) Peroxy
32             radical and related trace gas measurements in the boundary layer above the Atlantic Ocean. J. Geophys. Res.
33             (Atmos.) 106: 5457-5477.
34       Burkert, J.; Andres-Hernandez, M. D.; Reichert, L.; Meyer-Arnek, J.; Burrows, J. P.; Doddridge, B. G.; Dickerson,
35             R. R.; Muehle, J.; Zhan, A.; Carsey, T. (2003) Trace gas and radical diurnal behavior in the marine boundary
36             layer during INDOEX 1999. J. Geophys. Res. (Atmos.)  108(D8): 10.1029/2002JD002790.
37       Byun, D.; Young, J.; Gipson, G.; Schere,  K.; Godowitch, J. (1998) Urban air quality simulation with community
38             multi-scale air quality (CMAQ) modeling system. Research Triangle Park, NC: U.S.  Environmental
39             Protection Agency,  National Exposure Research Laboratory; report no., EPA--600/A-98/115. Available from:
40             NTIS, Springfield, VA; PB99-102246.
41       Calvert, J.  G.; Madronich,  S. (1987) Theoretical study of the initial products of the atmospheric oxidation of
42             hydrocarbons. J. Geophys. Res. [Atmos.] 92: 2211-2220.
43       Calvert, J.  G.; Heywood, J. B.; Sawyer, R. F.; Seinfeld, J. H. (1993) Achieving acceptable air quality: some
44             reflections on controlling vehicle emissions. Science (Washington, DC) 261: 37-45.
45       Calvert, J.  G.; Yarwood, G.;  Dunker, A. M. (1994) An evaluation of the mechanism of nitrous acid formation in the
46             urban atmosphere. Res. Chem.  Intermed. 20: 463-502.
47       Calvert, J.  G.; Atkinson, R.; Kerr, J. A.; Madronich, S.; Moortgat, G. K.; Wallington, T. J.; Yarwood,  G.  (2000)
48             The mechanisms of atmospheric oxidation of the alkenes. New York, NY: Oxford University Press, Inc.
49       Calvert, J.  G.; Atkinson, R.; Becker, K. H.; Kamens, R. M.; Seinfeld, J. H.; Wallington, T. J.; Yarwood, G. (2002)
50             The mechanisms of atmospheric oxidation of aromatic hydrocarbons. New York, NY: Oxford University
51             Press, Inc.
52       Cantrell, C. A.; Shelter, R. E.; Gilpin, T. M.; Calvert, J. G.; Eisele, F. L.; Tanner, D.  J. (1996) Peroxy radical
53             concentrations measured and calculated from trace gas measurements in the Mauna Loa Observatory
54             Photochemistry Experiment 2. J. Geophys. Res. (Atmos.) 101: 14,653-14,664.
         August 2005                                 AX2-160       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Cantrell, C. A.; Calvert, J. G.; Bais, A.; Shelter, R. E.; Lefer, B. L.; Edwards, G. D. (2003) Overview and
  2             conclusions of the International Photolysis Frequency Measurement and Modeling Intercomparison (IPMMI)
  3             study. J. Geophys.  Res. (Atmos.) 108(D16): 10.1029/2002JD002962.
  4       Cardelino, C. A.; Chameides, W. L. (1990) Natural hydrocarbons, urbanization, and urban ozone. J. Geophys. Res.
  5             [Atmos.] 95: 13,971-13,979.
  6       Cardelino, C. A.; Chameides, W. L. (1995) An observation-based model for analyzing ozone precursor relationships
  7             in the urban atmosphere. J. Air Waste Manage. Assoc. 45: 161-180.
  8       Cardelino, C. A.; Chameides, W. L. (2000) The application of data from photochemical assessment monitoring
  9             stations to the observation-based model. Atmos. Environ. 34: 2325-2332.
10       Carpenter, L. J.; Stages, W. T. Liss, P. S.; Penkett, S. A.; Alicke, B.; Hebestreit, K.; Platt, U. (1999) Short-lived
11             alkyl iodides and bromides at Mace Head, Ireland: links to biogenic sources and halogen oxide production.
12             J. Geophys. Res. (Atmos.) 104: 1679-1689.
13       Carroll, M. A.; Hastie, D. R.; Ridley, B. A.; Rodgers, M. O.; Torres, A. L.; Davis, D. D.; Bradshaw, J. D.;
14             Sandholm, S. T.; Schiff, H. L; Karecki, D. R.; Harris, G. W.; Mackay, G. L; Gregory, G. L.; Condon, E. P.;
15             Trainer, M.; Hubler, G.; Montzka, D. D.; Madronich, S.; Albritton, D. L.; Singh, H. B.; Beck, S. M.;
16             Shipham, M. C.; Bachmeier, A. S. (1990) Aircraft measurement of NOX over the eastern Pacific and
17             continental United States and implications for ozone production. J. Geophys. Res. [Atmos.] 95:
18             10,205-10,233.
19       Carroll, M. A.; Bertman, S. B.; Shepson, P. B. (2001) Overview of the program for research on oxidants,
20             photochemistry, emissions and transport (PROPHET) summer 1998 measurements intensive. J. Geophys.
21             Res. [Atmos.] 106: 24,275-24,288.
22       Carter, W. P. L. (1990) A detailed mechanism for the gas-phase atmospheric reactions of organic compounds.
23             Atmos. Environ. Part A 24: 481-518.
24       Carter, W. P. L. (1994) Development of ozone reactivity scales for volatile organic compounds. J. Air Waste
25             Manage. Assoc. 44: 881-899.
26       Carter, W. P. L. (1995) Computer modeling of environmental chamber studies of maximum incremental reactivities
27             of volatile organic  compounds. Atmos. Environ. 29: 2513.
28       Castro, T; Madronich, S; Rivale, S; Muhlia, A; Mar, B. (2001) The influence of aerosols on photochemical smog in
29             Mexico City. Atmos. 35: 1765-1772..
30       Chameides, W. L.; Stelson, A. W. (1992) Aqueous-phase chemical processes in deliquescent sea-salt aerosols:
31             A mechanism that couples the atmospheric cycles of S and sea salt. J. Geophys. Res. [Atmos.]
32             97(018): 20,565-20,580.
33       Chameides, W. L.; Lindsay, R. W.; Richardson, J.; Kiang,  C. S. (1988) The role of biogenic hydrocarbons in urban
34             photochemical smog: Atlanta as a case study. Science (Washington, DC) 241: 1473-1475.
35       Chameides, W. L.; Fehsenfeld, F.; Rodgers, M. O.; Cardelino, C.; Martinez, J.; Parrish, D.; Lonneman, W.; Lawson,
36             D. R.; Rasmussen, R. A.; Zimmerman, P.; Greenberg, J.; Middleton, P.; Wang, T. (1992) Ozone precursor
37             relationships in the ambient atmosphere. J.  Geophys. Res. [Atmos.] 97: 6037-6055.
38       Chang, J. S.; Brost, R. A.; Isaken, I. S. A.; Madronich, S.; Middleton, P.; Stockwell, W. R.; Walcek, C. J. (1987)
39             A three-dimensional Eulerian acid deposition model: physical concepts and formulation. J. Geophys. Res.
40             [Atmos.] 92: 14,681-14,700.
41       Chang, M. E.; Cardelino,  C.; Hartley, D.; Chang, W. L. (1996) Inverse modeling of biogenic emissions. Geophys.
42             Res. Lett. 23:  3007ff.
43       Chang, M. E.; Hartley, D. E.; Cardelino, C.; Hass-Laursen, D.; Chang, W. L. (1997) On using inverse methods for
44             resolving emissions with large spatial inhomogeneities. J. Geophys. Res. (Atmos.) 102: 16,023-16,036.
45       Chang, S.; McDonand-Buller, E. C.; Kimura, Y.; Russell, M. M.; Tanaka, P. L.; Allen, D. T. (2002) Spatial and
46             temporal impacts of chlorine chemistry on ozone formation in southeastern Texas. Austin, TX: The Texas
47             Natural Resource Conservation Commission; Contract No. 9880077600-18. Available:
48             ftp://ftp.tnrcc.state.tx.us/pub/OEP AA/TAD/Modeling/HGAQSE/Contract_Reports/Others/Impacts_Chlorine_
49             on_Ozone_Formation.pdf (17 December 2003).
50       Chatfield, R. B.; Crutzen, P. J. (1984) Sulfur dioxide in remote oceanic air: cloud transport of reactive precursors.
51             J. Geophys. Res. 89:7111-7132.
52       Chen, L.-W. A.; Chow, J. C.; Doddridge, B. G.; Dickerson, R. R.; Ryan, W. F.; Mueller, P. K. (2003) Analysis of
53             summertime PM25 and haze episode in the  mid-Atlantic region. J. Air Waste Manage. Assoc. 53: 946-956.
54       Cheng, M.-D.; Hopke, P. K.; Zeng, Y. (1993) A receptor-oriented methodology for determining source regions of
55             paniculate sulfate observed at Dorset, Ontario. J. Geophys. Res. [Atmos.] 98: 16,839-16,849.
         August 2005                                 AX2-161       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Chesselet, R.; Fontugne, M; Buat-Menard, P.; Ezat, U.; Lambert, C. E. (1981) The origin of paniculate organic
  2             carbon in the marine atmosphere as indicated by its stable carbon isotopic composition. Geophys. Res. Lett.
  3             8:345-348.
  4       Ching, J. K. S.; Alkezweeny, A. J. (1986) Tracer study of vertical exchange by cumulus clouds. J. Clim. Appl.
  5             Meteorol. 25: 1702-1711.
  6       Ching, J. K. S.; Shipley, S. T.; Browell, E. V. (1988) Evidence for cloud venting of mixed layer ozone and aerosols.
  7             Atmos. Environ. 22:  225-242.
  8       Ching, J. K. S.; Byun, D.; Young, J.; Binkowski, F. S.; Pleim, J.; Roselle, S.; Godowitch, J.; Benjey, W.; Gipson, G.
  9             (1998) Science features inModels-3 Community Multiscale Air Quality System. In: Preprints of the 10th
10             Joint AMS/AWMA Conference on Applications of Air Pollution Meteorology; Phoenix, AZ. Pittsburgh, PA:
11             Air & Waste Management Association.
12       Chock, D. P.; Winkler, S. L. (1994) A comparison of advection algorithms coupled with chemistry. Atmos. Environ.
13             28:2659-2675.
14       Choi, W.; Leu, M.-T. (1998) Nitric acid uptake and decomposition on black carbon (soot) surfaces: its implications
15             for the upper troposphere and lower stratosphere. J. Phys. Chem. A 102: 7618-7630.
16       Choi, Y. J.; Ehrman, S. H. (2004) Investigation of sources of volatile organic carbon in the Baltimore area using
17             highly time-resolved measurements. Atmos. Environ. 38: 775-791.
18       Civerolo, K.; Dickerson, R. R. (1998) Nitric oxide soil emissions from tilled and untilled cornfields. Agric. For.
19             Meteorol.  90:307-311.
20       Civerolo, K. L.; Mao, H. T.; Rao, S.  T. (2003) The airshed for ozone and fine paniculate pollution in the eastern
21             United States. Pure Appl. Geophys. 160: 81-105.
22       Claeys, M.; Wang, W.; Ion, A. C.; Kourtchev, I.; Gelencser, A.; Maenhaut, W. (2004) Formation of secondary
23             organic aerosols from isoprene and its gas-phase oxidation products through reaction with hydrogen peroxide.
24             Atmos. Environ. 38:  4093-4098.
25       Clark, T. L.; Karl, T. R. (1982) Application of prognostic meteorological variables to forecasts of daily maximum
26             one-hour ozone concentrations in the northeastern United States. J. Appl. Meteorol. 21:  1662-1671.
27       Cleary, P. A.; Wooldridge, P. J.; Cohen, R. C. (2002) Laser-induced fluorescence detection of atmospheric NO2 with
28             a commercial diode laser and a supersonic expansion. Appl. Optics 41: 6950-6956.
29       Cleveland, W. S.; Guarino,  R.; Kleiner, B.; McRae, J. E.; Warner, J. L. (1976) The analysis of the ozone problem in
30             the northeast United  States. In: Specialty conference on: ozone/oxidants-interactions  with the total
31             environment; March; Dallas, TX. Pittsburgh, PA: Air Pollution Control Association;  pp. 109-119.
32       Cleveland, W. S.; Kleiner, B.; McRae, J. E.; Warner, J. L. (1976) Photochemical air pollution: transport from
33             New York City area  into Connecticut and Massachusetts. Science (Washington, DC) 191: 179-181.
34       Cohen, R. C. (1999) Laser-induced fluorescence detection of atmospheric NO2 at parts per trillion mixing ratios:
3 5             implications for nitrogen oxide photochemistry in the stratosphere and troposphere. Abstr. Pap. Am. Chem.
36             Soc. (Phys.) 218(Pt. 2): 262.
37       Collins, W. J.; Stevenson, D. S.; Johnson, C. E.; Derwent, R. G. (1997) Tropospheric ozone in a global-scale
3 8             three-dimensional Lagrangian model and its response to NO^ emission controls. J. Atmos. Chem.
39             26:223-274.
40       Comes, F. J.; Armerding, W.; Grigonis, R.; Herbert, A.; Spiekermann, M.; Walter, J. (1992) Tropospheric OH:  local
41             measurements and their interpretations. Ber. Bunsen-Ges. Phys. Chem. 96: 284-286.
42       Cooper, O. R.; Moody, J. L. (2000) Meteorological controls on ozone at an elevated eastern United States regional
43             background monitoring site. J. Geophys. Res. [Atmos.] 105: 6855-6869.
44       Cooper, O. R.; Moody, J. L.; Davenport, J. C.; Oltmans, S. J.; Johnson, B. J.; Chen, X.; Shepson, P. B.; Merrill, J. T.
45             (1998) Influence of springtime weather systems on vertical ozone distributions over three North American
46             sites. J. Geophys. Res. [Atmos.] 103: 22,001-22,013.
47       Cooper, O. R.; Moody, J. L.; Parrish, D. D.; Trainer, M.; Ryerson, T. B.; Holloway, J. S.; Htibler, G.; Fehsenfeld,
48             F. C.; Oltmans, S. J.; Evans, M. J. (2001) Trace gas signatures of the airstreams within North Atlantic
49             cyclones: case studies from the North Atlantic Regional Experiment (NARE '97) aircraft intensive. J.
50             Geophys. Res. [Atmos.]  106: 5437-5456.
51       Cooper, O. R.; Moody, J. L.; Parrish, D. D.; Trainer, M.; Ryerson, T. B.; Holloway, J. S.; Htibler, G.; Fehsenfeld,
52             F. C.; Evans, M. J. (2002a) Trace gas composition of mid-latitude cyclones  over the western North Atlantic
53             Ocean: a conceptual  model. J. Geophys. Res. [Atmos.]  107(D7): 10.1029/2000JD000901.
54       Cooper, O. R.; Moody, J. L.; Parrish, D. D.; Trainer, M.; Holloway, J. S.; Htibler, G.; Fehsenfeld, F. C.; Oltmans,
55             S. J.; Evans, M. J. (2002b) Trace gas composition of mid-latitude cyclones over the western North Atlantic
56             Ocean: a seasonal comparison of O3 and CO. J. Geophys. Res. [Atmos.] 107(D7):  10.1029/2001JD000902.
         August 2005                                 AX2-162       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Cooper, O.; Forster, C.; Parrish, D.; Dunlea, E.; Htibler, G.; Fehsenfeld, F.; Holloway, J.; Oltmans, S.; Johnson, B.;
  2             Wimmers, A.; Horowitz, L. (2004a) On the life cycle of a stratospheric intrusion and its dispersion into
  3             polluted warm conveyor belts. J. Geophys. Res. (Atmos.) 109(D23S09): 10.1029/2003 JD004006.
  4       Cooper, O. R.; Forster, C.; Parrish, D.; Trainer, M; Dunlea, E.; Ryerson, T.; Htibler, G.; Fehsenfeld, F.; Nicks, D.;
  5             Holloway, J.; De Gouw, J.; Warneke, C.; Roberts, J. M.; Flocke, F.; Moody, J. (2004b) A case study of
  6             transpacific warm conveyor belt transport: the influence of merging airstreams on trace gas import to North
  7             America. J. Geophys. Res. (Atmos.) 109(D23S08): 10.1029/2003JD003624.
  8       Corchnoy, S. B.; Atkinson, R. (1990) Kinetics of the gas-phase reactions of OH and NO3 radicals with 2-Carene,
  9             l,8-cineole,/>-cymene, and terpinolene. Environ. Sci. Technol. 24: 1497-1502.
10       Cornell, S.; Rendell, A.; Jickells, T. (1995) Atmospheric inputs of dissolved organic nitrogen to the oceans. Nature
11             376:  243-246.
12       Cornell, S. E.; Jickells, T. D.; Thornton, C. A. (1998) Urea in rainwater and atmospheric aerosols. Atmos. Environ.
13             32: 1903-1910.
14       Cornell, S.; Mace, K.; Coeppicus, S.; Duce, R.; Huebert, B.; Jickells, T.; Zhuang, L. Z. (2001) Organic nitrogen in
15             Hawaiian rain and aerosols. J. Geophys. Res. (Atmos.) 106: 7973-7983.
16       Corsmeier, U.; Kalthhoff, N.; Kolle, O.; Motzian, M.; Fiedler, F. (1997) Ozone concentration jump in the stable
17             nocturnal boundary layer during a LLJ-event. Atmos. Environ. 31: 1977-1989.
18       Cowling, E. B.;  Chameides, W. L.; Kiang, C.  S.; Fehsenfeld, F. C.; Meagher, J. F. (2000) Introduction to special
19             section: Southern Oxidant Study Nashville/Middle Tennessee Ozone Study, part 2. J. Geophys. Res. (Atmos.)
20             105:  9075-9077.
21       Cox, R. A.; Bloss, W. J.; Jones, R. L.; Rowley, D. M. (1999) OIO and the atmospheric cycle of iodine. Geophys.
22             Res.  Lett. 26:  1857-1860.
23       Crawford, J. H.; Davis, D. D.; Chen, G.; Buhr, M.; Oltmans, S.; Weller, R.; Mauldin, L.; Eisele, F.; Shelter, R.;
24             Lefer, B.; Arimoto, R.; Hogan, A. (2001) Evidence for photochemical production of ozone at the South Pole
25             surface. Geophys. Res. Lett. 28: 3641-3644.
26       Crosley, D. R. (1996) NOyblue ribbon panel.  J. Geophys. Res. [Atmos.] 101(D1): 2049-2052.
27       Crutzen, P.  (1973) A discussion of the chemistry of some minor constituents in the stratosphere and troposphere.
28             Pure Appl. Geophys. 106-108: 1385-1399.
29       Crutzen, P.  J.; Gidel, L. T. (1983) A two-dimensional photochemical model of the atmosphere. 2. The tropospheric
30             budgets of the anthropogenic chlorocarbons, CO, CH4, CH3C1 and the effect of various NOX sources on
31             tropospheric ozone. J. Geophys. Res. [Atmos.] 88: 6641-6661.
32       Crutzen, P.  J.; Lawrence, M. G.; Poschl, U. (1999) On the background photochemistry of tropospheric ozone. Tellus
33             51A-B: 123-146.
34       Crutzen, P.  J.; Williams, J.; Poschl, U.; Hoor, P.; Fischer, H.; Warneke, C.; Holzinger, R.; Hansel, A.; Lindinger,
35             W.; Scheeren, B.; Lelieveld, J. (2000) High spatial and temporal resolution measurements of primary organics
36             and their oxidation products over the tropical forest of Surinam. Atmos. Environ. 34: 1161-1165.
37       Cziczo, D. J.; Abbatt, J. P. D. (1999) Deliquescence, efflorescence, and supercooling of ammonium sulfate aerosols
38             at low temperature: implications for cirrus cloud formation and aerosol phase in the atmosphere. J. Geophys.
39             Res.  [Atmos.] 104:  13,781-13,790.
40       Czoschke, N. M.; Jang, M.; Kamens, R. M. (2003) Effect of acidic seed onbiogenic secondary organic aerosol
41             growth. Atmos. Environ. 37: 4287-4299.
42       Danielsen, E. F. (1968) Stratospheric-tropospheric exchange based on radioactivity, ozone and potential vorticity.
43             J. Atmos. Sci. 25:502-518.
44       Danielsen, E. F. (1980) Stratospheric source for unexpectedly large values of ozone measured over the Pacific Ocean
45             during Gametag, August 1977. J. Geophys. Res. C: Oceans Atmos. 85: 401-412.
46       Danielsen, E. F. (1993) In situ evidence of rapid, vertical, irreversible transport of lower tropospheric air into the
47             lower tropical stratosphere by convective cloud turrets and by larger-scale upwelling in tropical cyclones. J.
48             Geophys. Res. [Atmos.]  98(D5): 8665-8681.
49       Danielsen, E. F.; Mohnen, V. A. (1977) Project Dustorm report: ozone transport, in situ measurements, and
50             meteorological analyses of tropopause  folding. J. Geophys. Res. [Atmos.] 82:  5867-5877.
51       Danielsen, E. F., Hipskind, R. S.; Gaines, S. E.; Sachse, G. W.; Gregory, G. L.; Hill, G. F. (1987) Three-dimensional
52             analysis of potential vorticity  associated with tropopause folds and observed variations of ozone and carbon
53             monoxide. J. Geophys. Res. [Atmos.]  92: 2103-2111.
54       Daum, P. H.; Kleinman, L. I.; Hills, A. J.; Lazrus,  A. L.; Leslie, A. C. D.; Busness, K.; Boatman, J. (1990)
5 5             Measurement and interpretation of concentrations of H2O2 and related species in the upper midwest during
56             summer. J. Geophys. Res. [Atmos.] 95: 9857-9871.
         August 2005                                  AX2-163       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Daum, P. H.; Kleinman, L. I.; Newman, L.; Luke, W. T.; Weinstein-Lloyd, J.; Berkowitz, C. M; Busness, K. M.
  2             (1996) Chemical and physical properties of plumes of anthropogenic pollutants transported over the North
  3             Atlantic during the North Atlantic Regional Experiment. J. Geophys. Res. [Atmos.] 101: 29,029-29,042.
  4       Davidson, E. A.; Kingerlee, W. (1997) A global inventory of nitric oxide emissions from soils. Nutr. Cycling
  5             Agroecosyst. 48: 37-50.
  6       Davies, T. D.; Schuepbach, E. (1994) Episodes of high ozone concentrations at the earth's surface resulting from
  7             transport down from the upper troposphere/lower stratosphere: a review and case studies. Atmos. Environ.
  8             28:53-68.
  9       Davis, D. D.; Bellinger, W.; Fischer, S. (1975) A kinetics study of the reaction of the OH free radical with aromatic
10             compounds I. Absolute rate constants for reaction with benzene and toluene at 300°K. J.  Phys. Chem.
11             79: 293-294.
12       Day, D. A.; Wooldridge, P. J.; Dillon, M. B.; Thornton, J. A.; Cohen, R. C. (2002) A thermal dissociation
13             laser-induced fluorescence instrument for in situ detection of NO2, peroxy nitrates, alkyl nitrates, and HNO3.
14             J. Geophys. Res. (Atmos.) 107(D5-D6): 10.1029/2001JD000779.
15       Day, D. A.; Dillon, M. B.; Wooldridge, P.  J.; Thornton, J. A.; Rosen, R. S.; Wood, E. C.; Cohen, R. C. (2003) On
16             alkyl nitrates, O3, and the "missing NOy". J. Geophys. Res. [Atmos.] 108(D16): 10.1029/2003JD003685.
17       DeCaria,  A. J.; Pickering, K. E.; Stenchikov, G. L.; Scala, J. R.; Stith, J. L.; Dye, J.  E.; Ridley, B. A.; Laroche, P.
18             (2000) A cloud-scale model study of lightning-generate NOX in an individual thunderstorm during
19             STERAO-A. J. Geophys. Res. [Atmos.] 105: 11601-11616.
20       DeMore,  W. B.; Sander, S. P.; Golden, D.  M.; Hampson, R. F.; Kurylo, M. J.; Howard, C. J.; Ravishankara, A. R.;
21             Kolb, C. E.; Molina, M. J. (1992) Chemical kinetics and photochemical data for use in stratospheric
22             modeling: evaluation number 10. Pasadena, CA: National Aeronautics and Space Administration, Jet
23             Propulsion Laboratory; JPL publication no. 92-20.
24       Dennis, R. L.; Novak, J. H. (1992) EPA's third generation modeling system (Models-3): an overview. In: Berglund,
25             R.  L., ed. Tropospheric ozone and the environment II: effects, modeling and control [papers from an
26             international specialty conference];  November; Atlanta, GA. Pittsburgh, PA: Air & Waste Management
27             Association; pp. 137-147. (A&WMA transactions series no. 20).
28       De Reus,  M.; Dentener, F.; Thomas, A.; Borrmann, S.; Strom, J.; Lelieveld, J. (2000) Airborne observations of dust
29             aerosol over the North Atlantic Ocean during ACE 2: indications for heterogeneous ozone destruction. J.
30             Geophys. Res. [Atmos.]  105: 15,263-15,275.
31       Derwent,  R. G.; Jenkin, M. E. (1991) Hydrocarbons and the long-range transport of ozone and PAN across Europe.
32             Atmos. Environ. Part A 25: 1661-1678.
33       Derwent,  R. G.; Jenkin, M. E.; Saunders, S. M. (1996) Photochemical ozone  creation potentials for a large number
34             of reactive hydrocarbons under European conditions. Atmos. Environ. 30: 181-199.
35       Derwent,  R. G.; Jenkin, M. E.; Saunders, S. M.; Pilling, M. J. (1998) Photochemical ozone creation potentials for
36             organic compounds in northwest Europe calculated with a master chemical mechanism. Atmos. Environ.
37             32:2429-2441.
38       Derwent,  R. G.; Collins, W. J.; Johnson, C. E.; Stevenson, D. S. (2001) Transient behaviour of tropospheric ozone
39             precursors in a global 3-D CTM and their indirect greenhouse effects.  Climatic Change 49: 463-487.
40       Dewulf, J.; Van Langenhove, H. (1997) Analytical techniques for the determination and measurement data of
41             7 chlorinated Cl- and C2-hydrocarbons and 6 monocyclic aromatic hydrocarbons in remote air masses:
42             an overview. Atmos. Environ. 31: 3291-3307.
43       Dibb, J. E.; Arsenault, M.; Peterson, M. C. (2002) Fast nitrogen oxide photochemistry in Summit, Greenland snow.
44             Atmos. Environ. 36:2501-2511.
45       Dickerson, R. R.; Huffman, G. J.; Luke, W. T.; Nunnermacker, L. J.; Pickering, K. E.; Leslie, A. C. D.; Lindsey, C.
46             G.; Slinn, W. G. N.; Kelly, T. J.; Daum, P. H.; Delany, A. C.;  Greenberg, J. P.; Zimmerman, P. R.; Boatman,
47             J. F.; Ray, J. D.; Stedman, D. H. (1987) Thunderstorms: an important mechanism in the transport of air
48             pollutants. Science (Washington, DC) 235: 460-465.
49       Dickerson, R. R.; Doddridge, E.G.; Kelley, P.; Rhoads, K. P. (1995) Large-scale pollution of the atmosphere over
50             the remote Atlantic Ocean: evidence from Bermuda. J. Geophys. Res.  [Atmos.] 100: 8945-8952.
51       Dickerson, R. R.; Kondragunta, S.; Stenchikov, G.; Civerolo, K. L.; Doddridge, B. G.; Holben, B. N. (1997)
52             The impact of aerosols on solar ultraviolet radiation and photochemical smog. Science (Washington, DC)
53             278: 827-830.
54       Dickerson, R. R.; Rhoads, K. P.; Carsey, T. P.; Oltmans, S. J.; Burrows, J. P.; Crutzen, P. J. (1999) Ozone in the
55             remote marine boundary layer: a possible role for halogens. J. Geophys. Res. [Atmos.] 104: 21,385-21,395.
         August 2005                                 AX2-164       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Djuric, D.; Damiani, M. S., Jr. (1980) On the formation of the low-level jet over Texas. Mon. Weather Rev. Vol.
  2             108: 1854-1865.
  3       Dodge, M. C. (1977) Combined use of modeling techniques and smog chamber data to derive ozone-precursor
  4             relationships. In: Dimitriades, B., ed. International conference on photochemical oxidant pollution and its
  5             control - proceedings: volume II; September 1976; Raleigh, NC. Research Triangle Park, NC: U.S.
  6             Environmental Protection Agency, Environmental Sciences Research Laboratory; pp. 881-889; report no.
  7             EPA-600/3-77-001b. Available from: NTIS, Springfield, VA; PB-264233.
  8       Dodge, M. C. (1977) Effect of selected parameters on predictions of a photochemical model. Research Triangle
  9             Park, NC: U.S. Environmental Protection Agency, Environmental Sciences Research Laboratory; EPA report
10             no. EPA-600/3-77-048. Available from: NTIS, Springfield, VA; PB-269858.
11       Domine, F.; Shepson, P. B. (2002) Air-snow interactions and atmospheric chemistry. Science (Washington, DC)
12             297: 1506-1510.
13       Dommen, I; Prevot, A. S. H.; Hering, A. M.; Staffelbach, T.; Kok, G. L.; Schilawsky, R. D. (1999) Photochemical
14             production and aging of an urban air mass. J. Geophys. Res. (Atmos.) 104: 5493-5506.
15       Dudhia, J. (1993) A nonhydrostatic version of the Perm State-NCAR mesoscale model: validation tests and
16             simulation of an Atlantic  cyclone and cold front. Mon. Weather Rev. 121: 1493-1513.
17       Dumdei, B. E.; O'Brien, R. J. (1984) Toluene's degradation products under simulated atmospheric conditions.
18             Nature (London) 311: 248-250.
19       Dumdei, B. E.; Kenny, D. V; Shepson, P. B.; Kleindienst, T. E.; Nero, C. M.; Cupitt, L. T.; Claxton, L. D. (1988)
20             MS/MS analysis of the products of toluene photooxidation and measurement of their mutagenic activity.
21             Environ. Sci. Technol. 22: 1493-1498.
22       Eckhardt, S.; Stohl, A.; Beirle, S.; Spichtinger, N.; James, P.; Forster, C.; Junker, C.; Wagner, T.; Platt, U.; Jennings,
23             S. G. (2003) The North Atlantic oscillation  controls air pollution transport to the Arctic. Atmos. Chem. Phys.
24             3:  1769-1778.
25       Eder, B. K.; Davis, J. M.; Bloomfield, P. (1993) A characterization of the spatiotemporal variability of non-urban
26             ozone concentrations over the Eastern United States. Atmos. Environ. 27A: 2645-2668.
27       Eder, B. K.; Davis, J. M.; Bloomfield, P. (1994) An automated classification scheme designed to better elucidate the
28             dependence of ozone on meteorology. J. Appl. Meteorol. 33: 1182-1199.
29       Edney, E. O.; Driscoll, D. J.; Weathers, W. S.; Kleindienst, T. E.; Conver, T. S.; Mclver, C. D.; Li, W. (2001)
3 0             Formation of polyketones in irradiated toluene/propylene/NOx/air mixtures. Aerosol Sci. Technol.
31             35:998-1008.
32       Ehhalt, D. H.; Dorn, H.-P.; Poppe, D. (1991) The chemistry of the hydroxyl radical in the troposphere. In: Last, F.
33             T.; Watling, R., eds. Acidic deposition: its nature and impacts.  Proceedings of the international symposium;
34             September 1990; Glasgow, United Kingdom. Proc. R. Soc. Edinburgh, Sect. B: Biol. Sci. 97:  17-34.
35       Eisele, F. L.; Mount, G. H.; Tanner, D.; Jefferson,  A.; Shelter, R.; Harder, J. W.;  Williams, E. J. (1997)
3 6             Understanding the production and interconversion of the hydroxyl radical  during the tropospheric OH
37             photochemistry experiment. J. Geophys. Res. 102: 6457-6465.
38       Emmons, L. K.; Carroll, M. A.; Hauglustaine, D. A.; Brasseur, G. P.;  Atherton, C.; Penner, J.; Sillman, S.; Levy, H.,
39             II; Rohrer, F.; Wauben, W. M. F.; van Velthoven, P. F. J.; Wang, Y.; Jacob, D.; Bakwin, P.; Dickerson, R.;
40             Doddridge, B.; Gerbig, C.; Honrath, R.; Htibler, G.; Jaffe, D.; Kondo, Y.; Munger, J. W.; Torres, A.;
41             Volz-Thomas, A. (1997) Climatologies of NOX and NOy: a comparison of data and models. Atmos. Environ.
42             31: 1837-1850.
43       Emmons, L. K.; Hauglustaine, D. A.; Mtiller,  J.-F.; Carroll, M. A.; Brasseur, G. P.; Brunner, D.; Staehelin, J.;
44             Thouret, V.; Marenco, A. (2000) Data composites of airborne observations of tropospheric ozone and its
45             precursors. J. Geophys. Res. [Atmos.] 105:  20,497-20,538.
46       Erickson, D. J.; Seuzaret, C.; Keene, W. C.; Gong, S.-L. (1999) A general circulation model based calculation of
47             HC1 and CINO2 production from sea-salt dechlorination: reactive chlorine emissions inventory. J. Geophys.
48             Res. [Atmos.] 104: 8347-8372.
49       Fast, J. D.; Zhong, S. Y. (1998) Meteorological factors associated with inhomogeneous ozone concentrations within
50             the Mexico City basin. J.  Geophys. Res. (Atmos.) 103: 18,927-18,946.
51       Fast, J. D.; Doran, J.; Shaw, W. J.; Coulter, R. L.; Martin, T. J. (2000) The evolution of the boundary layer and its
52             effect on air chemistry in the Phoenix area. J. Geophys. Res. [Atmos.] 105: 22,833-22,848.
53       Fast, J. D.; Zaveri, R. A.; Bian, R. X.; Chapman, E. G.; Easter, R. C. (2002) Effect of regional-scale transport on
54             oxidants in the vicinity of Philadelphia during the 1999 NE-OPS Field Campaign. J. Geophys. Res. [Atmos.]
55             107(016): 10.1029/2001JD000980.
         August 2005                                 AX2-165       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Faust, B. C. (1994) A review of the photochemical redox reactions of iron(III) species in atmospheric, oceanic, and
  2             surface waters: influences on geochemical cycles and oxidant formation. In: Helz, G. R.; Zepp, R. G.; Crosby,
  3             D. G., eds. Aquatic and Surface Photochemistry. Boca Raton, FL: Lewis Publishers; pp. 3-37.
  4       Faust, B. E.; Allen, J. M; Anastasio, C. (1992) Aqueous-phase photochemical formation of peroxides peroxyl
  5             radicals and singlet molecular oxygen in authentic cloud waters [abstract]. Presented at: 203rd national
  6             meeting of the American Chemical Society; April; San Francisco, CA. Abstr. Pap. Am. Chem. Soc.
  7             203: ENVR94.
  8       Faust, B. C.; Anastasio, C.; Allen, J. M.; Arakaki, T. (1993) Aqueous-phase photochemical formation of peroxides
  9             in authentic cloud and fog waters. Science 260: 73-75.
10       Fehsenfeld, F. C.; Dickerson, R. R.; Htibler, G.; Luke, W. T.; Nunnermacker, L. J.; Williams, E. J.; Roberts, J. M.;
11             Calvert, J. G.; Curran, C. M.; Delany, A. C.; Eubank, C. S.; Fahey, D. W.; Fried, A.; Gandrud, B. W.;
12             Langford, A. O.; Murphy, P. C.; Norton, R. B.; Pickering, K. E.; Ridley, B. A. (1987) A ground-based
13             intercomparison of NO, NOX, and NOy measurement techniques. J. Geophys. Res. [Atmos.] 92:
14             14,710-14,722.
15       Fehsenfeld, F.; Calvert, J.; Fall, R.; Goldan, P.; Guenther, A. B.; Hewitt, C. N.; Lamb, B.; Liu, S.; Trainer, M.;
16             Westberg, H.; Zimmerman, P. (1992a) Emissions of volatile organic compounds from vegetation and the
17             implications for atmospheric chemistry. Global Biogeochem. Cycles 6:  389-430.
18       Fehsenfeld, F. C.; Trainer, M.; Parrish, D. D.; Volz-Thomas, A.; Penkett, S. (1996a) North Atlantic Regional
19             Experiment (NARE) 1993 summer intensive: foreword. J. Geophys. Res. [Atmos.] 101: 28,869-28,875.
20       Fehsenfeld, F. C.; Damn, P.; Leaitch, W. R.; Trainer, M.; Parrish, D. D.; Htibler, G. (1996b) Transport and
21             processing of O3 and O3 precursors over the North Atlantic: an overview of the 1993 North Atlantic Regional
22             Experiment (NARE) summer intensive. J. Geophys. Res. [Atmos.] 101(D22): 28877-28891.
23       Fehsenfeld, F. C.; Huey, L. G.; Sueper, D. T.; Norton, R. B.; Williams, E. J.; Eisele, F. L.; Mauldin, R. L., Ill;
24             Tanner, D. J. (1998) Ground-based intercomparison of nitric acid measurement techniques. J. Geophys. Res.
25             [Atmos.] 103(03): 3343-3353.
26       Ferman, M. A.; Monson, P. R. (1978) Comparison of rural and urban air quality. Presented at: 71st annual meeting
27             of the Air Pollution Control Association; June; Houston, TX. Pittsburgh, PA: Air Pollution Control
28             Association; paper no. 78-10.4.
29       Pick, J.; Pommer, L.; Nilsson, C.; Andersson, B. (2003) Effect of OH radicals, relative humidity, and time on the
30             composition of the products formed in the ozonolysis of a-pinene. Atmos. Environ. 37: 4087-4096.
31       Finlayson-Pitts, B. J. (1993) Comment on "Indications of photochemical histories of Pacific air masses from
32             measurements of atmospheric trace species at Point Arena, California" by D. D. Parrish et al. J. Geophys.
33             Res. [Atmos.]  98(D8): 14,991-14,993.
34       Finlayson-Pitts, B. J.; Pitts, J. N., Jr. (1993) Atmospheric chemistry of tropospheric ozone formation: scientific and
35             regulatory implications. Air Waste 43: 1091-1100.
36       Finlayson-Pitts, B. J.; Pitts, J. N., Jr. (2000) Chemistry of the upper and lower atmosphere theory, experiments and
37             applications. San Diego, CA: Academic Press.
38       Fiore, A. M.; Jacob, D. J.; Bey, L; Yantosca, R. M.; Field, B. D.; Fusco, A. C.; Wilkinson, J. G.  (2002) Background
39             ozone over the United States in  summer: origin, trend, and contribution to pollution episodes. J. Geophys.
40             Res. (Atmos.) 107(D15): 10.1029/2001JD000982.
41       Fiore, A. M.; Jacob, D. J.; Mathur, R.;  Martin, R. V. (2003) Application of empirical orthogonal functions to
42             evaluate ozone simulations with regional and global models. J. Geophys. Res.  (Atmos.)
43             108(D14): 10.1029/2002JD003151.
44       Fishman, J.; Wozniak, A. E.; Creilson, J. K. (2003) Global distribution of tropospheric ozone from satellite
45             measurements using the empirically corrected tropospheric ozone residual technique: identification of the
46             regional aspects of air pollution. Atmos. Chem. Phys. 3: 893-907.
47       Forstner, H. J. L.; Flagan, R. C.; Seinfeld, J. H. (1997) Secondary organic aerosol from the photooxidation of
48             aromatic hydrocarbons: molecular composition. Environ. Sci. Technol.  31: 1345-1358.
49       Fortuin, J. P. F.; Kelder, H. (1998) An  ozone climatology based on ozonesonde and satellite measurements. J.
50             Geophys. Res. 103: 31,709-31,734.
51       Fox-Rabinovitz, M. S.; Takacs, L. L.; Govindaraju, R. C. (2002) A variable-resolution stretched-grid general
52             circulation model and data assimilation system with multiple areas of interest:  studying the anomalous
53             regional climate events of 1998. J. Geophys. Res. [Atmos.] 107(D24): 10.1029/2002JD002177.
54       Fried, A.; Hodgeson, J. (1982) Laser photoacoustic detection of nitrogen dioxide in the gas-phase titration of nitric
55             oxide with ozone. Anal. Chem.  54:  278-282.
         August 2005                                  AX2-166       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Friedlander, S. K.; Yeh, E. K. (1998) The submicron atmospheric aerosol as a carrier of reactive chemical species:
  2             case of peroxides. Appl. Occup. Environ. Hyg. 13: 416-420.
  3       Fuglestvedt, J. S.; Berntsen, T. K.; Isaksen, I. S. A.; Mao, H.; Liang, X.-Z.; Wang, W.-C. (1999) Climatic forcing of
  4             nitrogen oxides through changes in tropospheric ozone and methane; global 3D model studies. Atmos.
  5             Environ. 33: 961-978.
  6       Fujita, E. M; Croes, B. E.; Bennett, C. L.; Lawson, D. R.; Lurmann, F. W.; Main, H. H. (1992) Comparison of
  7             emission inventory and ambient concentration ratios of CO, NMOG, and NOX in California's south coast air
  8             basin. J. Air Waste Manage. Assoc. 42: 264-276.
  9       Fujita etal. (1995).
10       Furutani, H.; Akimoto, H. (2002) Development and characterization of a fast measurement system for gas-phase
11             nitric acid with a chemical ionization mass spectrometer in the marine boundary layer. J. Geophys. Res.
12             (Atmos.) 107(D1-D2): 10.1029/2000JD000269.
13       Fusco, A. C.; Logan, J. A. (2003) Analysis of 1970-1995 trends in tropospheric ozone at Northern Hemisphere
14             midlatitudes with the GEOS-CHEM model. J. Geophys.  Res. (Atmos.) 108: 10.1029/2002JD002742.
15       Gaffney, J. S.; Marley, N. A.; Drayton, P. J.; Doskey, P. V.; Kotamarthi, V. R.; Cunningham, M. M.; Baird, J. C.;
16             Dintaman, J.; Hart, H. L. (2002) Field observations of regional and urban impacts on NO2, ozone, UVB,  and
17             nitrate radical production rates in the Phoenix air basin. Atmos. Environ. 36: 825-83 3.
18       Galbally, I. E.; Bentley, S. T.; Meyer, C. P. (2000) Mid-latitude marine boundary-layer ozone destruction at visible
19             sunrise observed at Cape Grim, Tasmania, 41°S. Geophys. Res. Lett. 27: 3841-3844.
20       Galloway, J. N.; Whelpdale, D. M. (1980) An atmospheric sulfur budget for eastern North America. Atmos.
21             Environ. 14: 409-417.
22       Galmarini, S.; Duynkerke, P. G.; de Arellano, J. V. G. (1997) Evolution of nitrogen oxide chemistry in the nocturnal
23             boundary layer. J. Appl. Meteorol. 36: 943-957.
24       Garnica, R. M.; Appel, M. F.; Eagan, L. (2000) A REMPI method for the ultrasensitive detection of NO and NO2
25             using atmospheric pressure laser ionization mass spectrometry. Anal. Chem. 72: 5639-5646.
26       Gauss, M.; Myhre, G.; Pitari, G.; Prather, M. J.; Isaksen, I. S. A.; Berntsen, T. K.; Brasseur, G. P.; Dentener, F. J.;
27             Derwent, R. G.; Hauglustaine, D. A.; Horowitz, L. W.; Jacob, D. J.; Johnson, M.; Law, K. S.; Mickley, L. J.;
28             Mtiller, J.-F.; Plantevin, P. H.; Pyle, J. A.; Rogers, H. L.; Stevenson, D. S.; Sundet, J. K.; Van Weele, M.;
29             Wild, O. (2003) Radiative forcing in the 21st century due to ozone changes in the troposphere and the lower
30             stratosphere. J. Geophys. Res. (Atmos.) 108(D9): 10.1029/2002JD002624.
31       Geiger, H.; Barnes, I.; Bejan, I.; Benter, T.; Spittler, M. (2003)  The tropospheric degradation of isoprene: an updated
32             module for the regional atmospheric chemistry mechanism. Atmos. Environ. 37: 1503-1519.
33       Geron, C. D.;  Guenther, A. B.; Pierce, T. E. (1994) An improved model for estimating emissions of volatile organic
34             compounds from forests in the eastern United States. J. Geophys. Res. [Atmos.] 99: 12,773-12,791.
35       Geron, C. D.; Nie, D.; Arnts, R. R.; Sharkey, T. D.; Singsaas, E. L.; Vanderveer, P. J.; Guenther, A.; Sickles, J. E.,
36             II; Kleindienst, T. E. (1997) Biogenic isoprene emission: model evaluation in a southeastern United States
37             bottomland deciduous forest. J. Geophys. Res.  [Atmos.]  102(D15): 18,889-18,901.
38       Geron, C.; Guenther, A.; Sharkey, T.; Arnts, R.R.  (2000a) Temporal variability in basal isoprene emission factor.
39             Tree Physiol. 20: 799-805.
40       Geron, C.; Rasmussen, R.; Arnts, R. R.; Guenther, A. (2000b) A review and synthesis of monoterpene speciation
41             from forests in the United States. Atmos. Environ. 34: 1761-1781.
42       Geron, C. D.; Harley, P.; Guenther, A. (2001) Isoprene emission capacity for US tree species. Atmos. Environ.
43             35:  3341-3352.
44       Gershenzon, Y. M.; Grigorieva, V. M.; Ivanov, A. V.; Remorov, R.  G.  (1995) O3 and OH sensitivity to
45             heterogeneous sinks of HOX and CH3O2 on aerosol particles. Faraday Discuss. 100: 83-100.
46       Gershenzon, M. Y.; Il'in, S.; Fedotov, N. G.; Gershenzon, Y. M.; Aparina, E. V.; Zelenov, V. V. (1999) The
47             mechanism of reactive NO3 uptake on dry NaX (X=C1, Br). J. Atmos. Chem. 34: 119-135.
48       Gershey, R. M. (1983) Characterization of seawater organic matter carried by bubble-generated aerosols. Limnol.
49             Oceanogr. 28:309-319.
50       Gertler, A. W.; Sagebiel, J. C.; Dippel, W. A.; O'Connor, C. M. (1999) The impact of California phase 2
51             reformulated gasoline on real-world vehicle emissions. J. Air Waste Manage. Assoc. 49: 1339-1346.
52       Gery, M. W.; Fox, D. L.; Jeffries, H. E.; Stockburger, L.; Weathers, W. S. (1985) A continuous stirred tank reactor
53             investigation of the gas-phase reaction of hydroxyl radicals and toluene. Int. J. Chem. Kinet. 17: 931-955.
54       Gery, M. W.; Whitten, G. Z.; Killus, J. P. (1988) Development  and testing of the CMB-IV for urban and regional
55             modeling. Research Triangle Park, NC: U.S. Environmental Protection Agency, Atmospheric Sciences
56             Research Laboratory; report no. EPA-600/3-88-012. Available from: NTIS, Springfield, VA; PB88-180039.
         August 2005                                 AX2-167       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Gery, M. W.; Whitten, G. Z.; Killus, J. P.; Dodge, M. C. (1989) A photochemical kinetics mechanism for urban and
  2             regional scale computer modeling. J. Geophys. Res. [Atmos.] 94: 12,925-12,956.
  3       Gettelman, A.; Holton, J. R.; Rosenlof, K. H. (1997) Mass fluxes of O3, CH4, N2O, and CF2C12 in the lower
  4             stratosphere calculated from observational data. J. Geophys. Res. 102(D15): 19,149-19,159.
  5       Geyer, A.; Alicke, B.; Konrad, S.; Schmitz, T.; Stutz, J.; Platt, U. (2001) Chemistry and oxidation capacity of the
  6             nitrate radical in the continental boundary layer near Berlin. J. Geophys. Res. (Atmos.) 106: 8013-8025.
  7       Geyer, A.; Platt, U.  (2002) Temperature dependence of the NO3 loss frequency: A new indicator for the contribution
  8             of NO3 to the oxidation of monoterpenes and NOX removal in the atmosphere. J. Geophys. Res. (Atmos.)
  9             107(020): 10.1029/2001JD001215.
10       Gidel, L. T. (1983) Cumulus cloud transport of transient tracers. J. Geophys. Res. [Atmos.]  88: 6587-6599.
11       Gillani, N. V.; Pleim, J. E. (1996) Sub-grid-scale features of anthropogenic emissions  of NOX and VOC in the
12             context of regional Eulerian models. Atmos. Environ. 30: 2043-2059.
13       Gillani, N. V.; Meagher, J. F.; Valente, R. J.; Imhoff, R. E.; Tanner, R. L.; Luria, M. (1998) Relative production of
14             ozone and nitrates in urban and rural power plant plumes.  1. Composite results  based on data from 10 field
15             measurement days. J. Geophys. Res. [Atmos.] 103: 22,593-22,615.
16       Gilpin, T.; Apel, E.; Fried, A.; etal. (1997) Intercomparison of six ambient (CH2O) measurement techniques.
17             J. Geophys. Res. (Atmos.) 102(D17): 21,161-21,188.
18       Goldan,  P. D.; Kuster, W. C.; Albritton, D. L.;  Fehsenfeld, F. C.; Connell, P. S.; Norton, R. B.; Huebert, B. J. (1983)
19             Calibration and tests of the filter-collection method for measuring clean-air, ambient levels of nitric acid.
20             Atmos. Environ. 17: 1355-1364.
21       Goldan,  P. D.; Trainer, M.; Kuster, W. C.; Parrish, D. D.; Carpenter, J.; Roberts, J. M.; Yee,  J. E.; Fehsenfeld, F. C.
22             (1995) Measurements of hydrocarbons, oxygenated hydrocarbons, carbon monoxide, and nitrogen oxides in
23             an urban basin in Colorado: implications for emission inventories. J. Geophys.  Res. [Atmos.] 100:
24             22,771-22,783.
25       Goldan,  P. D.; Kuster, W. C.; Fehsenfeld, F. C. (1997) Non-methane hydrocarbon measurements during the
26             tropospheric  OH photochemistry experiment.  J. Geophys. Res.  (Atmos.) 102: 6315-6324.
27       Goldan,  P. D.; Parrish, D. D.; Kuster, W. C.; Trainer, M.; McKeen, S. A.; Holloway, J.; Jobson, B. T.;  Sueper, D. T.;
28             Fehsenfeld, F. C. (2000) Airborne measurements of isoprene, CO, and anthropogenic  hydrocarbons and their
29             implications. J. Geophys. Res. [Atmos.]  105: 9091-9105.
30       Goldan etal. (2001).
31       Golden,  D. M.; Smith,  G. P. (2000) Reaction of OH + NO2 + M: a new view. J. Phys. Chem. A 104: 3991-3997.
32       Goldstein, A. H.; Schade, G. W. (2000) Quantifyingbiogenic and anthropogenic contributions to acetone mixing
33             ratiosin a rural environment. Atmos. Environ. 34: 4997-5006.
34       Gong, S. L.; Barrie, L.  A.; Blanchet, J.-P. (1997) Modeling sea-salt aerosols in the environment: 1. model
35             development. J.  Geophys. Res. 102: 3805-3818.
36       Gong, W.; Mickle, R. E.; Bottenheim, F. F.; Beauchamp, S.; Waugh, D. (2000) Marine/coastal boundary layer and
37             vertical structure of ozone observed at a coastal site in Nova Scotia during the 1996 NARSTO-CE field
38             campaign. Atmos. Environ. 34: 4139-4154.
39       Gorzelska, K.; Galloway, J. N.; Watterson, K.;  Keene, W. C. (1992) Water-soluble primary amine compounds  in
40             rural continental precipitation. Atmos. Environ. 26: 1005-1018.
41       Goult, J. (1938) Vents en alititude a Fort Lamy (Tach). Ann. Phys. (Paris, France) 5: 373-379.
42       Grassi, B.; Pitari, G.; Visconti, G. (1997) Results of a chemical-transport model with interactive aerosol
43             microphysics. In: Bojkov, R.; Visconti, G. Proceedings of the XVIII Quadrennial Ozone Symposium;
44             September, 1996; L'Aquila, Italy.
45       Greenhut, G. K. (1986) Transport of ozone between boundary layer and cloud layer by cumulus clouds. J. Geophys.
46             Res.  [Atmos.] 91: 8613-8622.
47       Greenhut, G. K.; Ching, J. K. S.; Pearson, R., Jr.; Repoff, T. P. (1984) Transport of ozone by turbulence and clouds
48             in an urban boundary layer. J. Geophys.  Res. [Atmos.]  89:4757-4766.
49       Gregg, J. W.; Jones, C. G.; Dawson, T. E. (2003) Urbanization effects on tree growth  in the vicinity of New York
50             City  [letter]. Nature 424:  183-187.
51       Gregory, G. L.; Hoell, J. M., Jr.; Torres, A. L.;  Carroll, M. A.; Ridley, B. A.; Rodgers, M. O.; Bradshaw, J.;
52             Sandholm, S.; Davis, D. D. (1990) An intercomparison of airborne nitric oxide  measurements: a second
53             opportunity. J. Geophys. Res. [Atmos.] 95: 10,129-10,138.
         August 2005                                 AX2-168       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Grell, G. A.; Dudhia, I; Stauffer, D. R. (1994) Description of the fifth-generation Perm State/NCAR mesoscale
  2             model (MM5). Boulder, CO: National Center for Atmospheric Research, Mesoscale and Microscale
  3             Meteorology Division; report no. NCAR/TN-398+STR. Available from: NTIS, Springfield, VA;
  4             PB95-206348.
  5       Grell, G. A.; Emeis, S.; Stockwell, W. R.; Schoenemeyer, T.; Forkel, R.; Michalakes, I; Knoche, R.; Seidl, W.
  6             (2000) Application of a multiscale, coupled MM5/chemistry model to the complex terrain of the VOTALP
  7             valley campaign. Atmos. Environ. 34: 1435-1453.
  8       Grobner, J.; Kerr, J. B.  (2001) Ground-based determination of the spectral ultraviolet extraterrestrial solar irradiance:
  9             providing a link between space-based and ground-based solar UV measurements. J. Geophys. Res. (Atmos.)
10             106(D7): 7211-7217.
11       Grosjean, E.; Grosjean, D. (1994) Rate constants forthe gas-phase reactions of ozone with unsaturated aliphatic
12             alcohols. Int. J. Chem. Kinetics 26: 1185-1191.
13       Grosjean, D.; Williams, E. L., II; Grosjean, E.; Novakov, T. (1994) Evolved gas  analysis of secondary organic
14             aerosols. Aerosol Sci. Technol. 21: 306-324.
15       Gross, A.; Stockwell, W. R. (2003) Comparison of the EMEP, RADM2 and RACM Mechanisms. J. Atmos. Chem.
16             44: 151-170.
17       Guenther, A.; Hewitt, C. N.; Erickson, D.; Fall, R.; Geron, C.; Graedel, T.; Harley, P.; Klinger, L.; Lerdau, M.;
18             McKay, W. A.; Pierce, T.; Scholes, B.; Steinbrecher, R.; Tallamraju, R.; Taylor, J.; Zimmerman, P. (1995) A
19             global model of natural volatile organic compound emissions J. Geophys. Res. (Atmos.) 100(D5): 8873-8892.
20       Guenther, A.; Baugh, W.; Davis, K.; Hampton, G.; Harley, P.; Klinger, L.; Vierling, L.; Zimmerman, P.; Allwine,
21             E.; Dilts, S.; Lamb, B.; Westberg, H.; Baldocchi, D.; Geron, C.; Pierce, T. (1996) Isoprene fluxes measured
22             by enclosure, relaxed eddy accumulation, surface layer gradient, mixed layer gradient, and mixed layer mass
23             balance techniques. J. Geophys. Res. (Atmos.) 101: 18,555-18,567.
24       Guenther, F. R.; Dorko, W. D.; Miller, W. R.; Rhoderick, G. C. (1996) The NIST traceable reference material
25             program for gas  standards. Washington, DC:  U.S. Department of Commerce, National Institute of Standards
26             and Technology; NIST special publication 260-126.
27       Guenther, A.; Geron, C.; Pierce, T.; Lamb, B.; Harley, P.; Fall, R. (2000) Natural emissions of non-methane volatile
28             organic compounds, carbon monoxide, and oxides of nitrogen from North America. Atmos. Environ.
29             34:2205-2230.
30       Hadjinicolaou, P.; Pyle, J. A. (2004) The impact of Arctic ozone depletion on northern middle latitudes: interannual
31             variability  and dynamical control. J. Atmos. Chem. 47: 25-43.
32       Hallock-Waters, K. A.; Doddridge, B. G.; Dickerson, R. R.; Spitzer, S.; Ray, J. D. (1999) Carbon monoxide in the
33             U.S. mid-Atlantic troposphere: evidence for a decreasing trend. Geophys. Res. Lett. 26: 2861-2864.
34       Hallquist, M.; Stewart,  D. J.; Stephenson, S. K.; Cox, R. A. (2003) Hydrolysis of N2O5 on sub-micron sulfate
35             aerosols. Phys. Chem. Chem. Phys. 5: 3453-3463.
36       Hameed,  S.; Pinto, J. P.; Stewart,  R. W. (1979) Sensitivity of the predicted CO-OH-CH4 perturbation to tropospheric
37             NO-j concentrations. J. Geophys. Res. C: Oceans Atmos. 84: 763-768.
38       Hanisch, F.; Crowley, J. N. (2003) Ozone decomposition on Saharan dust: an experimental investigation. Atmos.
39             Chem. Phys. 3: 119-130.
40       Hanna, A. F.; Mathur, R.; Pinto, J.; Jang, C. (2001)  Seasonal modeling of the export of pollutants from North
41             America using the Multi Scale Air Quality Simulation Pattern (MAQSIP). Presented at: Workshop on
42             photooxidants, particles, and haze across the  Arctic and North Atlantic: observations and models; June;
43             New York, NY. Palisades, NY: Center for International Earth Science Information Network (CIESIN).
44             Available:  http://www.ciesin.org/pph/uploads/H_A_Seasonal_Modeling_PDF.pdf (29 December 2003).
45       Hannegan, B.; Olsen, S.; Prather,  M.; Zhu, X.; Rind, D.; Lerner, J. (1998) The dry stratosphere: a limit on cometary
46             water influx. Geophys. Res. Lett. 25: 1649-1652.
47       Hansel, A.; Jordan, A.;  Taucher, J. (1995a) Analysis of volatile organic compounds using
48             proton-transfer-reaction-mass spectrometry (PTR-MS) in the ppb range. In: Becker, K. H.; Carr, W. E.;
49             Kunhardt, E. E., eds. Proceedings of the XXII International conference  on phenomena in ionized gases;
50             July-August; Hoboken, NJ. Hoboken, NJ: Stevens Institute of Technology; pp.  181-182.
51       Hansel, A.; Jordan, A.;  Holzinger, R.; Prazeller, P.;  Vogel, W.; Lindinger, W.  (1995b) Proton transfer reaction mass
52             spectrometry: on-line trace gas analysis at the ppb level. Int. J. Mass  Spectrom. Ion Proc. 149/150: 609-619.
53       Hansen, K.; Draaijers, G. P. J.; Ivens, W. P. M. F.; Gundersen, P.; vanLeeuwen, N. F. M. (1994) Concentration
54             variations in rain and canopy throughfall collected sequentially during individual rain events. Atmos. Environ.
55             28: 3195-3205.
         August 2005                                 AX2-169       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Hansen, I; Sato, M; Ruedy, R.; Lacis, A.; Asamoah, K.; Beckford, K.; Borenstein, S.; Brown, E.; Cairns, B.;
  2             Carlson, B.; Curran, B.; de Castro, S.; Druyan, L.; Etwarrow, P.; Ferede, T.; Fox, M.; Gaffen, D.; Glascoe, J.;
  3             Gordon, H.; Hollandsworth, S.; Jiang, X.; Johnson, C.; Lawrence, N.; Lean, J.; Lerner, J.; Lo, K.; Logan, J.;
  4             Luckett, A.; McCormick, M. P.; McPeters, R.; Miller, R.; Minnis, P.; Ramberran, L; Russell, G.; Russell, P.;
  5             Stone, P.; Tegen, I.; Thomas, S.; Thomason, L.; Thompson, A.; et al. (1997) Forcings and chaos in
  6             interannual to decadal climate change. J.  Geophys. Res. [Atmos.]  102: 25,679-25,720.
  7       Hara, K.; Osada, K.; Hayashi, M.; Matsunaga, K.; Shibata, T.; Iwasaka, Y.; Furuya, K. (1999) Fractionation of
  8             inorganic nitrates in winter Arctic troposphere: coarse aerosol particles containing inorganic nitrates.
  9             J. Geophys. Res. (Atmos.) 104: 23,671-23,679.
10       Hard, T. M.; Mehrabzadeh, A. A.; Chan, C. Y.;  O'Brien, R. J. (1992) PAGE measurements of tropospheric HO with
11             measurements and model of interferences. J. Geophys. Res. [Atmos.] 97: 9795-9817.
12       Harder, J. W.; Brault, J. W.; Johnston, P. V.; Mount, G. H. (1997a) Temperature dependent NO2 cross sections at
13             high spectral resolution. J. Geophys. Res. [Atmos.] 102: 3861-3879.
14       Harder, J. W.; Williams, E. J.; Baumann, K.; Fehsenfeld, F. C. (1997b) Ground-based comparison of NO2, H2O, and
15             O3 measured by long-path and in situ techniques during the 1993 Tropospheric OH Photochemistry
16             Experiment. J. Geophys. Res. 102: 6227-6243.
17       Harley, P.; Guenther, A.; Zimmerman, P. (1996) Effects of light, temperature and canopy position on net
18             photosynthesis and isoprene emission from sweetgum (Liquidambar styraciflua) leaves. Tree Physiol.
19             16:25-32.
20       Harley etal. (1997).
21       Harris, G. W.; Carter, W. P. L.; Winer, A. M.; Pitts, J. N., Jr.; Platt, U.; Perner, D. (1982) Observations of nitrous
22             acid in the Los Angeles atmosphere and implications for predictions of ozone-precursor relationships.
23             Environ.  Sci. Technol. 16: 414-419.
24       Harris, J. M.; Oltmans, S. J.; Bodeker, G. E.; Stolarski, R.; Evans, R. D.; Quincy, D. M. (2003) Long-term variations
25             in total ozone derived from Dobson and satellite data. Atmos. Environ. 37: 3167-3175.
26       Harrison, R. M.; Peak, J. D.; Collins, G. M. (1996) Tropospheric cycle of nitrous acid. J.  Geophys. Res. (Atmos.)
27             101: 14,429-14,439.
28       Hauglustaine, D. A.; Brasseur, G. P.; Walters, S.; Rasch, P. J.; Mtiller, J.-F.; Emmons, L. K.; Carroll, M. A. (1998)
29             MOZART, a global chemical transport model for ozone and related chemical tracers 2. model results and
30             evaluation. J. Geophys. Res. (Atmos.) 103:  28291-28335.
31       Hebestreit, K.; Stutz, J.; Rosen, D.; Matveiv, V.; Peleg, M.; Luria, M.; Platt, U. (1999) DOAS measurements of
32             tropospheric bromine oxide in mid-latitudes. Science 283: 55-57.
33       Hegg, D. A.; Livingston, J.; Hobbs, P. V.; Novakov,  T.; Russell, P. (1997) Chemical apportionment of aerosol
34             column optical depth off the mid-Atlantic coast of the United States. J. Geophys. Res. (Atmos.) 102:
35             25,293-25,303.
36       Heikes, B. G.; Chang, W.; Pilson, M. E. Q.; Swift, E.; Singh, H. B.; Guenther, A.; Jacob, D. J.; Field, B. D.;  Fall, R.;
37             Riemer, D.; Brand,  L. (2002) Atmospheric methanol budget and ocean implication. Glob. Biogeochem.
38             Cycles 16: 10.1029/2002GB001895.
39       Henry, R. C.; Lewis, C. W.; Hopke, P. K.; Williamson, H. J. (1984) Review of receptor model fundamentals. Atmos.
40             Environ.  18: 1507-1515.
41       Hering, S. V.; Lawson, D. R.; Allegrini, I.; Febo, A.; Perrino, C.; Possanzini, M.; Sickles, J. E., II; Anlauf, K. G.;
42             Wiebe, A.; Appel, B. R.; John, W.; Ondo, J.; Wall, S.; Braman, R. S.; Sutton, R.; Cass, G. R.; Solomon, P.
43             A.; Eatough, D. J.; Eatough, N. L.; Ellis,  E. C.; Grosjean, D.; Hicks, B. B.; Womack, J. D.; Horrocks, J.;
44             Knapp, K. T.; Ellestad, T. G.; Paur, R. J.; Mitchell, W. J.; Pleasant, M.; Peake, E.; MacLean, A.; Pierson,
45             W. R.; Brachaczek, W.; Schiff, H. I.; Mackay, G. I.; Spicer, C. W.; (1988) The nitric acid Shootout: field
46             comparison of measurement methods. Atmos. Environ. 22: 1519-1539.
47       Herlihy, L. J.; Galloway, J. N.; Mills, A. L. (1987) Bacterial utilization of formic and acetic acid in rainwater.
48             Atmos. Environ. 21: 2397-2402.
49       Hess, P. G. (2001) Model and measurement analysis of springtime transport and chemistry of the Pacific Basin.
50             J. Geophys. Res. (Atmos.) 106: 12,689-12,717.
51       Hilsenrath, E.; Attmannspacher, W.; Bass, A.; Evans, W.; Hagemeyer, R.; Barnes, R. A.; Komhyr, W.;
52             Mauersberger, K.; Mentall, J.; Proffitt, M.; Robbins, D.; Taylor, S.; Torres, A.; Weinstock, E. (1986) Results
53             from the ballon ozonr intercomparison campaign (BOIC). J. Geophys. Res.  91: 13,137-13,152.
54       Hirsch, A. I.; Munger, J. W.; Jacob, D. J.; Horowitz,  L. W.; Goldstein, A. H. (1996) Seasonal variation of the ozone
55             production efficiency per unit NOX at Harvard Forest, Massachusetts. J. Geophys. Res. [Atmos.]
56             101: 12,659-12,666.
         August 2005                                  AX2-170       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Hobbs, P. V.; Locatelli, J. D.; Martin, J. E. (1996) A new conceptual model for cyclones generated in the lee of the
  2             Rocky Mountains. Bull. Am. Meteorol. Soc. 77: 1169-1178.
  3       Hoell, J. M.; Davis, D. D.; Jacob, D. J.; Rodgers, M. O.; Newell, R. E.; Fuelberg, H. E.; McNeal, R. J.; Raper, J. L.;
  4             Bendura, R. J. (1996) Pacific exploratory mission in the tropical Pacific: PEM-Tropics A, August-September
  5             1996. J.  Geophys. Res. (Atmos.) 104: 5567-5583.
  6       Hoell, J. M.; Davis, D. D.; Liu, S. C.; Newell, R. E.; Akimoto, H.; McNeal, R. J.; Bendura, R. J. (1997) The Pacific
  7             exploratory mission A west phase B: February-March,  1994. J. Geophys. Res. [Atmos.] 102(D23):
  8             28223-28239.
  9       Hoffman, E. J.; Duce, R. A. (1976) Factors influencing the organic carbon content of marine aerosols: a laboratory
10             study. J. Geophys. Res. (Atmos.) 81: 3667-3670.
11       Hoffman, E. J.; Duce, R. A. (1977) Organic carbon in marine atmospheric paniculate matter: concentration and
12             particle size distribution. Geophys. Res. Lett. 4: 449-452.
13       Hoigne, J.; Zuo, Y.; Nowell, L. (1994) Photochemical reactions in atmospheric waters: role of dissolved iron
14             species.  In: Helz, G. R.; Zepp, R. G.; Crosby, D. C. Aquatic and Surface Photochemistry. Boca Raton, FL:
15             Lewis Publishers.
16       Holton, J. R. (1967) The diurnal boundary layer wind oscillation above sloping terrain. Tellus 19: 199-205.
17       Holton, J. R.; Haynes, P. H.; Mclntyre, M. E.; Douglass, A. R.; Rood, R. B.; Pfister, L. (1995)
18             Stratosphere-troposphere exchange. Rev. Geophys. 33: 403-439.
19       Honninger, G. (1999) Referenzspektren reaktiver halogenverbindungen fur DOAS-messungen. Heidelberg,
20             Germany: Ruprecht Karls University.
21       Honrath, R. E.; Guo, S.; Peterson, M. C. (2000) Photochemical production of gas phase NOX from ice crystal NO3".
22             J. Geophys. Res. (Atmos.) 105: 24,183-24,190.
23       Hopke, P. K. (1985) Receptor modeling in environmental chemistry. New York, NY: John Wiley & Sons.
24       Horowitz, L. W.;  Liang, J.; Gardner, G. M.; Jacob, D.  J . (1998) Export of reactive nitrogen from North America
25             during summertime: sensitivity to hydrocarbon chemistry. J. Geophys. Res. 103(D11): 13451-13476.
26       Hoskins, B. J. (1972) Non-boussinesq effects and further development in a model of upper tropospheric
27             frontogenesis. Q. J. Roy. Meteorol.  Soc. 98: 532-541.
28       Houweling, S.; Dentener, F.; Lelieveld, J.; Walter, B.; Dlugokencky, E. (2000) The modeling of tropospheric
29             methane: how well can point measurements be reproduced by a global model? J. Geophys. Res. (Atmos.)
30             105:8981-9002.
31       Htibler, G.; Alvarez, R., II; Daum, P.; Dennis, R.; Gillani, N.;  Kleinman, L.; Luke, W.; Meagher, J.; Rider, D.;
32             Trainer, M.; Valente, R. (1998) An overview of the airborne activities during the Southern Oxidants Study
33             (SOS) 1995 Nashville/Middle Tennessee ozone study. J. Geophys. Res. [Atmos.] 103: 22,245-22,259.
34       Hudman, R. C.; Jacob, D. J.; Cooper, O. C.; Evans, M. J.; Heald, C. L.; Park, R. J.; Fehsenfeld, F.; Flocke, F.;
35             Holloway, J.; Htibler, G.; Kita, K.; Koike, M.; Kondo, Y.; Neuman, A.; Nowak, J.; Oltmans, S.; Parrish, D.;
36             Roberts, J. M.; Ryerson, T. (2004) Ozone production in transpacific Asian pollution plumes and implications
37             for ozone air quality in California. J. Geophys. Res. (Atmos.) 109(D23): 10.1029/2004JD004974.
38       Huebert, B. J.; Zhuang, L.; Howell, S.; Noone, K.; Noone, B. (1996) Sulfate, nitrate, methanesulfonate, chloride,
39             ammonium, and sodium measurements from ship, island, and aircraft during the Atlantic Stratocumulus
40             Transition Experiment/Marine Aerosol Gas Exchange. J. Geophys. Res. (Atmos.) 101: 4413-4423.
41       Huey, L. G.; Dunlea, E. J.; Lovejoy, E. R.; Hanson,  D. R.; Norton, R. B.; Fehsenfeld, F. C.; Howard, C. J.  (1998)
42             Fast time response measurements of HNO3 in air with a chemical ionization mass spectrometer. J. Geophys.
43             Res. 103(D3): 3355-3360.
44       Hunsaker,  C.; Graham, R.; Turner, R. S.; Ringold, P. L.; Holdren, G. R., Jr.; Strickland, T. C. (1993) A national
45             critical loads framework for atmospheric deposition effects assessment: II. Defining assessment end points,
46             indicators, and functional subregions. Environ. Manage. 17: 335-341.
47       Huntrieser, H.; Feigl, C.; Schlager, H.; Schroder, F.; Gerbig, C.; Van Velthoven, P.; Flatey, F.; Thery, C.; Petzold,
48             A.; Holler, H.; Schumann, U.  (2002) Airborne measurements of NOX, tracer species, and small particles
49             during the  European Lightning Nitrogen Oxides Experiment. J. Geophys. Res. 107(D11):
50             10.1029/2000JD000209.
51       Husain, L.; Coffey,  P. E.; Meyers, R. E.; Cederwall, R. T. (1977) Ozone transport from stratosphere to troposphere.
52             Geophys. Res. Lett. 4: 363-365.
53       Husar, R. B.; Renard, W. P. (1998) Ozone as a function of local wind speed and direction: Evidence of local and
54             regional transport. Presented at: 91st annual meeting and exhibition of the Air & Waste Management
55             Association;  June; San Diego, CA. Pittsburgh, PA: Air & Waste Management Association; online paper no.
56             98-A922. Available: http://capita.wustl.edu/capita/CapitaReports/REPORTSl.HTML (13 November 2003).
         August 2005                                 AX2-171       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Imhoff, R. E.; Valente, R.; Meagher, J. F.; Luria, M. (1995) The production of O3 in an urban plume - airborne
  2             sampling of the Atlanta urban plume. Atmos. Environ. 29: 2349-2358.
  3       Intergovernmental Panel on Climate Change (IPCC). (1995) Climate change 1994: radiative forcing of climate
  4             change and an evaluation of the IPCC IS92 emission scenarios. Cambridge, United Kingdom: University
  5             Press.
  6       Intergovernmental Panel on Climate Change (IPCC). (1996) Climate change 1995: the science of climate change:
  7             contribution of working group I to the second assessment of the Intergovernmental Panel on Climate Change.
  8             Cambridge, United Kingdom: Cambridge University Press.
  9       Intergovernmental Panel on Climate Change (IPCC). (2001) Climate change 2001: the scientific basis. Contribution
10             of working group I to the third assessment report of the Intergovernmental Panel on Climate Change.
11             Cambridge, United Kingdom: Cambridge University Press.
12       Jacob, D. J. (2000) Heterogeneous chemistry and tropospheric ozone. Atmos. Environ. 34: 2131-2159.
13       Jacob, D. J.; Logan, J. A.; Gardner, G.  M.; Yevich, R. M.; Spivakovsky, C. M.; Wofsy, S. C.; Sillman, S.; Prather,
14             M. J. (1993) Factors regulating ozone over the United States and its export to the global atmosphere. J.
15             Geophys. Res. [Atmos.] 98: 14,817-14,826.
16       Jacob, D. J.; Horowitz, L. W.; Munger, J. W.; Heikes, B. G.; Dickerson, R. R.; Artz, R. S.; Keene, W. C. (1995)
17             Seasonal transition from NOX- to hydrocarbon-limited conditions for ozone production over the eastern
18             United States  in September. J. Geophys. Res. 100: 9315-9324.
19       Jacob, D. J.; Prather, M. J.; Rasch, P. J.; Shia, R.-L.; Balkanski, Y. J.; Beagley, S. R.; Bergman, D. J.; Blackshear,
20             W. T.; Brown, M.; Chiba, M.; Chipperfield, M.; De Grandpre, J.; Dignon, J. E.; Feichter, J.; Genthon, C.;
21             Grose, W.L.; Kasibhatla,  P. S.; Kohler, L; Kritz, M. A.; Law, K.; Penner, J. E.;  Ramonet, M.; Reeves, C. E.;
22             Rotman, D. A.; Stockwell, D. Z.; VanVelthoven, P. F. J.; Verver, G.; Wild, O.; Yang, H.; Zimmermann, P.
23             (1997) Evaluation and intercomparison of global atmospheric transport models using  222Rn and other
24             short-lived tracers. J. Geophys. Res. [Atmos.] 102: 5953-5970.
25       Jacobi, C.; Becker, C.; Kratzsch, T.; Siemer, A.; Roth, R. (1994) Experimental investigations of nocturnal low-level
26             jets with some remarks on flight safety. Contrib. Atmos. Phys. 67: 71-82.
27       Jacobson, M. Z. (1997) Development and application of a new air pollution modeling system - Part III.
28             aerosol-phase simulations. Atmos. Environ. 31: 587-608.
29       Jacobson, M. Z. (1998) Studying the effects of aerosols on vertical photolysis rate coefficient and temperature
30             profiles over an urban airshed. J. Geophys. Res. [Atmos.] 103: 10,593-10,604.
31       Jacobson, M. Z. (1999) Isolating nitrated and aromatic aerosols and nitrated aromatic gases as sources of ultraviolet
32             light absorption. J. Geophys. Res. (Atmos.) 104: 3527-3542.
33       Jacobson, M. Z.; Lu, R.; Turco, R. P.; Toon, O.  P. (1996) Development and application of a new air pollution
34             modeling system-Part I: gas-phase simulations. Atmos. Environ. 30: 1939-1963.
35       Jacobson, M. C.; Hansson, H.-C.; Noone, K. J.;  Charlson, R. J. (2000) Organic atmospheric  aerosols: review and
36             state of the science. Rev.  Geophys. 38: 267-294.
37       Jaegle, L.; Jacob, D.  J.; Wennberg, P. O.; Spivakovsky, C. M.; Hanisco, T. F.; Lanzendorf, E. J.; Hintsa, E. J.;
38             Fahey, D. W.; Keim, E. R.; Proffitt, M. H.; Atlas, E. L.; Flocke, F.; Schauffler, S.; McElroy, C. T.;
39             Midwinter, C.; Pfister, L.; Wilson, J. C. (1997) Observed OH and HO2 in the upper troposphere suggests a
40             major source from convective injection of peroxides. Geophys. Res. Lett. 24:  3181-3184.
41       Jaegle, L.; Jacob, D.  J.; Brune, W. H.; Faloona,  I. C.; Tan, D.; Kondo, Y.; Sachse, G. W.; Anderson, B.; Gregory,
42             G. L.; Vay, S.; Singh, H. B.; Blake, D. R.; Shelter, R. (1999) Ozone production in the upper troposphere and
43             the influence of aircraft during SONEX: approach of NOx-saturated conditions. Geophys. Res. Lett.
44             26:  3081-3084.
45       Jaegle, L.; Jacob, D.  J.; Brune, W. H.; Faloona,  I.; Tan, D.; Heikes, B. G.; Kondo, Y.; Sachse, G. W.; Anderson, B.;
46             Gregory, G. L.; Singh, H. B.; Pueschel, R.; Ferry, G.; Blake, D. R.; Shelter, R. E. (2000) Photochemistry of
47             HOX in the upper troposphere at northern midlatitudes. J. Geophys. Res. (Atmos.)  105(D3): 3877-3892.
48       Jaegle, L.; Jacob, D.  J.; Brune, W. H.; Wennberg, P. O.  (2001) Chemistry of HOX radicals in the upper troposphere.
49             Atmos. Environ. 35: 469-489.
50       Jaffe, D.; Price, H.; Parrish, D.; Goldstein, A.; Harris, J. (2003) Increasing background ozone during spring on the
51             west coast of North America. Geophys. Res. Lett. 30: 10,1029/2003GL017024.
52       Jaffe, D.; McKendry, I.; Anderson, T.;  Price, H. (2003) Six 'new'episodes of trans-Pacific transport of air pollutants.
53             Atmos. Environ. 37: 391-404.
54       James, P.;  Stohl, A.;  Forster, C.; Eckhardt, S.; Seibert, P.; Frank, A. (2003a) A 15-year climatology of
55             stratosphere-troposphere exchange with a Lagrangian particle dispersion model: 1. Methodology and
56             validation. J. Geophys. Res. 108(D12): 10.1029/2002JD002637.
         August 2005                                  AX2-172       DRAFT-DO NOT QUOTE OR CITE

-------
  1       James, P.; Stohl, A.; Forster, C.; Eckhardt, S.; Seibert, P.; Frank, A. (2003b) A 15-year climatology of
  2             stratosphere-troposphere exchange with a Lagrangian particle dispersion model. 2. Mean climate and seasonal
  3             variability. J. Geophys. Res. (Atmos.) 108(D12): 10.1029/2002JD002639.
  4       Jang, M.; Kamens, R. M. (2001) Characterization of secondary aerosol from the photooxidation of toluene in the
  5             presence of NOX and 1-propene. Environ. Sci. Technol. 35: 3626-3639.
  6       Jankowski, J. J.; Kieber, D. J.; Mopper, K.; Neale, P. J. (2000) Development and intercalibration of ultraviolet solar
  7             actinometers. Photochem. Photobiol. 71: 431-440.
  8       Jenkin, M. E.; Cox, R. A.; Williams, D. J. (1988) Laboratory studies of the kinetics of formation of nitrous acid from
  9             the thermal reaction of nitrogen dioxide and water vapour. Atmos. Environ. 22: 487-498.
10       Jenkin, M. E.; Saunders, S. M.; Pilling, M. J. (1997) The tropospheric degradation of volatile organic compounds:
11             a protocol for mechanism development. Atmos. Environ. 31: 81-104.
12       Jet Propulsion Laboratory. (2003) Chemical kinetics and photochemical data for use  in atmospheric studies.
13             Pasadena, CA: California Institute of Technology; JPL publication no. 02-25.  Available:
14             http://jpldataeval.jpl.nasa.gov/pdf/JPL_02-25_l_intro_rev0.pdf (18 December 2003).
15       Jeuken, A. B. M.; Eskes, H. J.; Van Velthoven, P.  F. J.; Kelder, H. M.; Holm,  E. V. (1999) Assimilation of total
16             ozone satellite measurements in a three-dimensional tracer transport model. J.  Geophys. Res. (Atmos.)
17             104: 5551-5563.
18       Jimenez, P.; Baldasano, J. M.; Dabdub, D. (2003) Comparison of photochemical mechanisms for air quality
19             modeling. Atmos. Environ.  37: 4179-4194.
20       Jobson, B. T.; Berkowitz, C. M.; Kuster, W. C.; Goldan, P. D.; Williams, E. J.; Fehsenfeld, F. C.; Apel, E. C.;
21             Karl, T.; Lonneman, W. A.; Riemer, D. (2004) Hydrocarbon source signatures in Houston, Texas: influence
22             of the petrochemical industry. J. Geophys. Res. 109: D24305: 10.1029/2004JD004887.
23       Johnson, W. B.; Viezee, W. (1981) Stratospheric ozone in the lower troposphere -1. presentation and interpretation
24             of aircraft measurements. Atmos. Environ. 15: 1309-1323.
25       Johnson, C. E.; Collins, W. J.; Stevenson, D. S.; Derwent,  R. G. (1999) The relative roles of climate and emissions
26             changes on future oxidant concentrations. J. Geophys. Res. (Atmos.) 104: 18,631-18,645.
27       Johnson, B. J.; Oltmans, S. J.; Vomel, H.; Smit, H. G. J.; Deshler, T.; Kroger,  C. (2002) Electrochemical
28             concentration cell (ECC) ozonesonde pump efficiency measurements and tests on the sensitivity to ozone of
29             buffered and unbuffered ECC sensor cathode solutions. J. Geophys. Res. 107(D19): 10.1029/2001JD000557.
30       Jordan, C. E.; Dibb, J. E.; Finkel, R. C. (2003) 10Be/7Be tracer of atmospheric transport and stratosphere-troposphere
31             exchange. J. Geophys. Res.  108(D8): 10.1029/2002JD002395.
32       Junge, C. E. (1962) Global ozone budget and exchange between stratosphere and troposphere. Tellus 14: 363-377.
33       Junge, C. E. (1963) Air chemistry and radioactivity. New York, NY: Academic Press. (Van Mieghem, J.; Hales,
34             A. L., eds. International geophysics series: v. 4).
35       Kain,  J. S.; Fritsch, J. M.  (1993) Convective parameterization in mesoscale models: the Kain-Fritsch scheme. In:
36             Emanuel, K.  A.; Raymond,  D. J., eds. The Representation of Cumulus Convection in Numerical Models.
37             Boston, MA: American Meteorological Society; pp. 165-170. (Meteorological Monographs, v. 24, no. 46).
38       Kane, M. M.; Rendell, A. R.; Jickells, T. D. (1994) Atmospheric scavenging processes over the North Sea. Atmos.
39             Environ. 28:  2523-2530.
40       Kane, S. M.; Caloz, F.; Leu, M.-T. (2001) Heterogeneous uptake of gaseous N2O5 by (NH4)2SO4 NH4HSO4, and
41             H2SO4 aerosols.  J.  Phys. Chem. A 105: 6465-6470.
42       Karamchandani, P.; Santos, L.; Sykes, I.; Zhang, Y.; Tonne, C.; Seigneur, C. (2000) Development and evaluation of
43             a state-of-the-science reactive plume model. Environ. Sci. Technol. 34: 870-880.
44       Karamchandani, P.; Seigneur, C.; Vijayaraghavan, K.; Wu, S.-Y. (2002) Development and application of a
45             state-of-the-science plume-in-grid model. J. Geophys. Res. (Atmos.)  107: 10.1019/2002D002123.
46       Kasibhatla, P.; Chameides, W. L. (2000) Seasonal modeling of regional ozone pollution in the eastern United  States.
47             Geophys. Res. Lett. 27: 1415-1418.
48       Kasting, J. F.; Singh, H. B. (1986)  Nonmethane hydrocarbons in the troposphere: impact on the odd hydrogen and
49             odd nitrogen chemistry, J. Geophys. Res. 91: 13,239-13,256.
50       Kavouras, I. G.; Mihalopoulus, N.; Stephanou, E. G. (1999) Secondary organic aerosol formation vs primary organic
51             aerosol emission: in situ evidence for the chemical coupling between monoterpene acidic photooxidation
52             products and new particle formation over forests. Environ. Sci. Technol. 33: 1028-1037.
53       Kawamura, K.; Semere, R.; Imai, Y.; Fuji, Y.; Hayashi, M. (1996) Water soluble dicarboxylic acids and related
54             compounds in Antarctic aerosols. J. Geophys. Res.  101: 18,721-18,728.
55       Keene, W. C.; Savoie, D. L. (1998) The pH of deliquesced sea-salt aerosol in polluted marine air. Geophys. Res.
56             Lett. 25: 2181-2194.
         August 2005                                 AX2-173       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Keene, W. C.; Savoie, D. L. (1999) Correction to "The pH of deliquesced sea-salt aerosol in polluted marine air."
  2             Geophys. Res. Lett. 26: 1315-1316.
  3       Keene, W. C.; Sander, R.; Pszenny, A. A. P.; Vogt, R.; Crutzen, P. I; Galloway, J. N. (1998) Aerosol pH in the
  4             marine boundary layer: a review and model evaluation. J. Aerosol Sci. 29: 339-356.
  5       Keene, W. C.; Pszenny, A. A. P.; Maben, J. R.; Sander, R. (2002) Variation of marine aerosol acidity with particle
  6             size. Geophys. Res. Lett. 29(7): 10.1029/2001GL013881.
  7       Kelly, N. A.; Wolff, G. T.; Ferman, M.  A. (1982) Background pollutant measurements in air masses affecting the
  8             eastern half of the United States -1. air masses arriving from the northwest. Atmos. Environ. 16: 1077-1088.
  9       Kelly, N. A.; Ferman, M. A.; Wolff, G. T. (1986) The chemical and meteorological conditions associated with high
10             and low ozone concentrations in southeastern Michigan and nearby areas of Ontario. J. Air Pollut. Control
11             Assoc. 36: 150-158.
12       Kenley, R. A.; Davenport, J. E.; Hendry, D. G. (1981) Gas-phase hydroxyl radical reactions. Products and pathways
13             for the reaction of OH with aromatic hydrocarbons. J. Phys. Chem. 85: 2740-2746.
14       Kim, B. M.; Henry, R. C. (2000) Application of SAFER model to the Los Angeles PM10 data. Atmos. Environ.
15             34:  1747-1759.
16       Kim, K. -H.; Kim, M. -Y. (2001) Comparison of an open path differential optical absorption spectroscopy system and
17             a conventional in situ monitoring system on the basis of long-term measurements of SO2, NO2, and O3.
18             Atmos. Environ. 35: 4059-4072.
19       King, D. W.; Lounsbury, H. A.; Millero, F. J. (1995) Rates and mechanisms of Fe(II) oxidation at nonmolar total
20             iron concentrations. Environ. Sci. Technol. 29: 818-824.
21       King, M. D.; Canosa-Mas, C. E.; Wayne, R. P. (2002) A structure activity relationship (SAR) for predicting rate
22             constants for the reaction of nitrogen dioxide (NO2) with alkenes. Phys. Chem. Chem. Phys.  4: 295-303.
23       Kireev, S. V.; Shnyrev, S. L.; Zhiganov, A. A. (1999) A laser fluorimeter for NO and NO2 in atmosphere. Instrum.
24             Exp. Tech. 42: 701-703.
25       Kleindienst, T. E.; Hudgens, E. E.; Smith, D. F.; McElroy, F. F.; Bufalini, J. J. (1993) Comparison of
26             chemiluminescence and ultraviolet ozone monitor responses in the presence of humidity and photochemical
27             pollutants. Air Waste 43: 213-222.
28       Kleindienst, T. E.; Smith, D. F.; Li, W.; Edney, E. O.; Driscoll, D. J.; Speer, R. E.; Weathers, W. S. (1999)
29             Secondary organic aerosol formation from the oxidation of aromatic hydrocarbons in the presence of dry
30             submicronammonium sulfate aerosol. Atmos. Environ. 33: 3669-3681.
31       Kleinman, L. I. (2000) Ozone process insights from field experiments - part II: observation-based analysis for ozone
32             production. Atmos. Environ. 34: 2023-2034.
33       Kleinman, L. L; Daum, P. H.; Lee, Y.-N.; Springston, S. R.; Newman, L.; Leaitch, W. R.; Banic, C. M.; Isaac,
34             G. A.; MacPherson, J. I. (1996a) Measurement of O3 and related compounds over southern Nova Scotia. 1.
35             Vertical distributions. J. Geophys. Res. [Atmos.] 101: 29,043-29,060.
36       Kleinman, L. I.; Daum, P. H.; Springston, S. R.; Leaitch, W. R.; Banic, C. M.; Isaac, G. A.; Jobson, T. B.; Niki, H.
37             (1996b) Measurement of O3 and related compounds over southern Nova Scotia. 2. Photochemical age and
38             vertical transport. J. Geophys. Res. [Atmos.] 101: 29,061-29,074.
39       Kleinman, L. I.; Daum, P. H.; Lee, J. H.; Lee, Y.-N.; Nunnermacker, S. R.; Springston, L.; Newman, J.;
40             Weinstein-Lloyd, J.; Sillman, S. (1997) Dependence of ozone production on NOand hydrocarbons in the
41             troposphere. Geophys. Res. Lett. 24: 2299-2302.
42       Kleinman, L. I.; Daum, P. H.; Imre, D. G.; Lee, J. H.; Lee, Y.-N.; Nunnermacker, L. J.; Springston, S. R.;
43             Weinstein-Lloyd, J.; Newman, L. (2000) Ozone production in the New York City urban plume. J. Geophys.
44             Res. [Atmos.]  105: 14,495-14,511.
45       Kleinman, L. L; Daum, P. H.; Lee, Y.-N.; Nunnermacker, L. J.; Springston, S. R.; Weinstein-Lloyd, J.; Rudolph, J.
46             (2001) Sensitivity of ozone production rate to ozone precursors. Geophys. Res. Lett. 28: 2903-2906.
47       Kleinman, L. L; Daum, P. H.; Imre, D.; Lee, Y.-N.; Nunnermacker, L. J.; Springston, S. R.; Weinstein-Lloyd, J.;
48             Rudolph, J. (2002) Ozone production rate and hydrocarbon reactivity in 5 urban areas: a cause of high ozone
49             concentration in Houston. Geophys. Res. Lett. 29:  10.1029/2001GL014569.
50       Kley, D.; Crutzen, P. J.; Smit, H. G. J.; Vomel, H.; Oltmans, S. J.; Grassl, H.; Ramanathan, V (1996) Observations
51             of near-zero ozone concentrations over the convective Pacific:  effects on air chemistry. Science 274: 230-233.
52       Klotz, B.; Barnes, I.; Becker, K. H.; Golding, B. T. (1997) Atmospheric chemistry of benzene oxide/oxepin. J.
53             Chem. Soc. Faraday Trans. 93: 1507-1516.
54       Klotz, B; Barnes, L; Becker, K. H. (1998) New results on the atmospheric photooxidation of simple alkylbenzenes.
55             Chem. Phys. 231:289-301.
56       Knake, R.; Hauser, P. C. (2002) Sensitive electrochemical detection of ozone. Anal. Chim. Acta 459: 199-207.
         August 2005                                 AX2-174      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Knispel, R.; Koch, R.; Siese, M.; Zetzsch, C. (1990) Adduct formation of OH radicals with benzene, toluene, and
  2             phenol and consecutive reactions of the adducts with NO^ and O2. Ber. Bunsen-Ges. Phys. Chem.
  3             94: 1375-1379.
  4       Koch, T. G.; Banham, S. F.; Sodeau, J. R.; Horn, A. B.; McCoustra, M. R. S.; Chesters, M. A. (1997) Mechanisms
  5             for the heterogeneous hydrolysis of hydrogen chloride, chlorine nitrate and dinitrogen pentoxide on water-rich
  6             atmospheric particle surfaces. J. Geophys Res. (Atmos.) 102: 1513-1522.
  7       Konrad, S.; Schmitz, T.; Buers, H.-J.; Houben, N.; Mannschreck, K.; Mihelcic, D.; Mtisgen, P.; Patz, H.-W.;
  8             Holland, F.; Hofzumahaus, A.; Schafer, H.-J.; Schroder, S.; Volz-Thomas, A.; Bachmann, K.; Schlomski, S.;
  9             Moortgat, G.; GroBmann, D. (2003) Hydrocarbon measurements at Pabstthum during the BERLIOZ
10             campaign and modeling of free radicals. J. Geophys. Res. [Atmos.] 108(D4): 10.1029/2001JD000866.
11       Kotamarthi, V. R.; Gaffney, J. S.; Marley, N. A.; Doskey, P. V. (2001) Heterogeneous NOX chemistry in the polluted
12             PEL. Atmos. Environ. 35: 4489-4498.
13       Krol, M.; Van Leeuwen, P. J.; Lelieveld, J. (1998) Global OH trend inferred from methylchloroform measurements.
14             J. Geophys. Res. [Atmos.]  103: 10,697-10,711.
15       Kumar, N.; Russell, A. G.  (1996) Development of a computationally efficient, reactive subgrid-scale plume model
16             and the impact in the northeastern United States using increasing levels of chemical detail. J. Geophys. Res.
17             [Atmos.] 101: 16,737-16,744.
18       Kwok, E. S. C.; Aschmann, S. M.; Atkinson, R.; Arey, J. (1997) Products of the gas-phase reactions of o-, m- and
19             p-xylens with the OH radical in the presence and absence of NOX. J. Chem. Soc. Faraday Trans.
20             93:2847-2854.
21       Lamb, B.; Gay, D.; Westberg, H.; Pierce, T. (1993) Abiogenic hydrocarbon emission inventory for the U.S.A. using
22             a simple forest canopy model. Atmos. Environ. Part A 27:  1673-1690.
23       Lammel, G.; Cape, J. N. (1996) Nitrous acid and nitrite in the atmosphere. Chem. Soc. Rev. 25: 361-369.
24       Lammel, G.; Perner, D. (1988) The atmospheric aerosol as a source of nitrous acid in the polluted atmosphere.
25             J. Aerosol  Sci. 19: 1199-1202.
26       Langford, A. O.; Reid, S. J. (1998) Dissipation and mixing of a small-scale stratospheric intrusion in the upper
27             troposphere. J. Geophys. Res. 103(D23): 31,265-31,276.
28       Lantz, K. O.; Shelter, R. E.; Cantrell, C. A.; Flocke, S. J.; Calvert, J. G.; Madronich, S. (1996) Theoretical,
29             actinometric, and radiometric determinations of the photolysis rate coefficient of NO2 during the Mauna Loa
30             Observatory Photochemistry Experiment 2. J. Geophys. Res. (Atmos.) 101(D9): 14,613-14629.
31       Law, K. S.; Plantevin, P. H.; Shallcross, D. E.; Rogers, H. L.; Pyle, J. A.; Grouhel, C.; Thouret, V.; Marenco, A.
32             (1998) Evaluation of modeled O3 using measurement of ozone by airbus in-service aircraft (MOZAIC) data.
33             J. Geophys. Res. (Atmos.)  103: 25,721-25,737.
34       Law, K. S.; Plantevin, P.-H.; Thouret, V.; Marenco, A.; Asman, W. A. H.; Lawrence, M.; Crutzen, P. J.; Muller,
35             J. F.; Hauglustaine,  D. A.; Kanakidou, M. (2000) Comparison between global chemistry transport model
3 6             results and measurement of ozone and water vapor by airbus in-service aircraft (MOZAIC) data. J. Geophys.
37             Res. (Atmos.) 105:  1503-1525.
38       Lawrence, M. G.; Crutzen, P. J.; Rasch, P. J.;  Eaton, B. E.; Mahowald, N. M. (1999) A model for studies of
39             tropospheric photochemistry: description, global distributions and evaluation. J. Geophys. Res.  (Atmos.)
40             104:26,245-26,277.
41       Lean, J. L. (2000) Short term, direct indices of solar variability. Space Science Reviews 94: 39-51.
42       Lefohn, A. S.; Oltmans, S. J.; Darin, T.; Singh, H. B. (2001) Present-day variability of background ozone in the
43             lower troposphere. J. Geophys. Res. [Atmos.] 106: 9945-9958.
44       Leibrock, E.; Huey, L. G. (2000) Ion chemistry for the detection of isoprene and other volatile organic compounds in
45             ambient air. Geophys. Res. Lett. 27: 1719-1722.
46       Lelieveld, J.; Crutzen, P. J. (1990) Influences  of cloud photochemical processes on tropospheric ozone. Nature
47             (London) 343: 227-233.
48       Lelieveld, J.; Crutzen, P. J. (1991) The role of clouds in tropospheric photochemistry.  J. Atmos. Chem. 12: 229-267.
49       Lelieveld, J.; Crutzen, P. J. (1994) Role of deep cloud convection in the ozone budget of the troposphere. Science
50             264: 1759-1761.
51       Lelieveld, J.; Dentener, F.  J. (2000) What controls tropospheric ozone? J. Geophys. Res. [Atmos.] 105: 3531-3551.
52       Leser, H.; Honninger, G.; Platt, U. (2003) MAX-DOAS measurements of BrO and NO2 in the marine boundary
53             layer. Geophys. Res. Lett. 30: 10.1029/2002GL015811.
54       Leston, A. R.; Ollinson, W. M.; Spicer, C. W.; Satola, J. (2005) Potential interference bias in ozone standard
5 5             compliance monitoring. J. Air Waste Manage. Assoc.: in press.
         August 2005                                 AX2-175       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Leu, M.-T. (2003) Laboratory studies of interaction between trace gases and sulphuric acid or sulphate aerosols
  2             using flow-tube reactors. Int. Rev. Phys. Chem. 22: 341-376.
  3       Li, Q.; Jacob, D. I; Logan, J. A.; Bey, I.; Yantosca, R. M; Liu, H.; Martin, R. V.; Fiore, A. M; Field, B. D.;
  4             Duncan, B. N. (2001) A tropospheric ozone maximum over the Middle East. Geophys. Res. Lett.
  5             28: 3235-3238.
  6       Li, Q.; Jacob, D. J.; Fairlie, T. D.; Liu, H.; Martin, R. V.; Yantosca, R. M. (2002) Stratospheric versus pollution
  7             influences on ozone at Bermuda:  reconciling past analyses. J. Geophys. Res. 107(D22):
  8             10.1029/2002JD002138.
  9       Liang, J.; Horowitz, L. W.; Jacob, D. J.; Wang, Y.; Fiore, A. M.; Logan, J. A.; Gardner, G. M.; Munger,  J. W. (1998)
10             Seasonal variations of reactive nitrogen species and ozone over the United States, and  export fluxes to the
11             global atmosphere. J. Geophys. Res. (Atmos.) 103: 13,435-13,450.
12       Liao, H.;  Yung, Y. L.;  Seinfeld, J. H.  (1999) Effects of aerosols on tropospheric photolysis rates in clear and cloudy
13             atmospheres. J. Geophys. Res.  (Atmos.) 104: 23,697-23,707.
14       Lillis, D.; Cruz, C. N.;  Collett,  J.; Richards, L. W.; Pandis, S. N. (1999) Production and removal of aerosol in a
15             polluted fog layer: model evaluation and fog effect on PM.  Atmos. Environ. 33: 4797-4816.
16       Lin, C.; Milford, J. B. (1994) Decay-adjusted chemical mass balance receptor modeling for volatile organic
17             compounds. Atmos. Environ. 28: 3261-3276.
18       Lin, X.; Trainer, M.; Liu, S. C. (1988) On the nonlinearity of the tropospheric ozone production. J. Geophys. Res.
19             [Atmos.] 93: 15,879-15,888.
20       Lindinger, W.; Hansel, A.; Jordan, A. (1998) Proton-transfer-reaction mass spectrometry (PTR-MS): on-line
21             monitoring of volatile organic compounds at pptv levels. Chem. Soc. Rev. 27: 347-354.
22       Liu, M.-K.; Seinfeld, J. H. (1975) On the validity of grid and trajectory models of urban air pollution. Atmos.
23             Environ. 9: 555-574.
24       Liu, S. C.; Trainer, M.; Fehsenfeld, F. C.; Parrish, D. D.; Williams, E. J.; Fahey, D. W.; Htibler, G.; Murphy, P. C.
25             (1987) Ozone production in the rural troposphere and the implications for regional and global ozone
26             distributions. J. Geophys. Res.  [Atmos.] 92: 4191-4207.
27       Liu, X. H.; Hegg, D. A.; Stoelinga, M. T. (2001a) Numerical simulation of new particle formation over the
28             northwest Atlantic using the MM5 mesoscale model coupled with sulfur chemistry. J.  Geophys. Res.
29             [Atmos.] 106: 9697-9715.
30       Liu, H.; Jacob, D. J.; Bey, I.; Yantosca, R. M. (200Ib) Constraints from 210Pb and 7Be on wet deposition and
31             transport in a global three-dimensional chemical tracer model driven by assimilated meteorological fields.
32             J. Geophys. Res. [Atmos.] 106: 12,109-12,128.
33       Logan, J. A. (1989) Ozone in rural areas of the United States. J. Geophys. Res. [Atmos.] 94: 8511-8532.
34       Logan, J. A.; Prather, M. J.; Wofsy, S. C.; McElroy, M. B. (1981) Tropospheric chemistry: a  global perspective.
35             J. Geophys. Res. C: Oceans Atmos. 86: 7210-7254.
36       Logan, J. A. (1994) Trends in the vertical distribution of ozone: an analysis of ozonesonde data. J. Geophys. Res.
37             (Atmos.) 99: 25,553-25,585.
38       Longfellow, C. A.; Ravishankara, A. R.; Hanson, D. R.  (1999) Reactive uptake on hydrocarbon soot: focus on NO2.
39             J. Geophys. Res. (Atmos.) 104: 13,833-13,840.
40       Lonneman, W. A.; Seila, R. L.  (1993) Hydrocarbon compositions  in Los Angeles and New York 20 years later.
41             Presented at: International symposium  on measurement of toxic and related air pollutants. Pittsburgh, PA:
42             Air & Waste Management Association.
43       Lu, C.-H.; Chang, J. S. (1998) On the indicator-based approach to assess ozone sensitivities and emissions features.
44             J. Geophys. Res. (Atmos.) 103: 3453-3462.
45       Lu, R.; Turco, R. P.; Jacobson, M. Z.  (1997) An integrated air pollution modeling system for  urban and regional
46             scales: 1. Structure and performance. J. Geophys. Res. [Atmos.] 102: 6063-6079.
47       Ludwig, F. L.; Reiter, E.; Shelar, E.; Johnson, W. B. (1977) The relation of oxidant levels to precursor emissions and
48             meteorological features: v. 1, analysis and findings. Research Triangle Park, NC: U.S.  Environmental
49             Protection Agency, Office of Air Quality Planning and Standards; report no. EPA-450/3-77-022a. Available
50             from: NTIS, Springfield, VA; PB-275 001.
51       Luke, W. T. (1997) Evaluation of a commercial pulsed fluorescence detector for the measurement of low-level SO2
52             concentrations during the Gas-Phase Sulfur Intercomparison Experiment. J. Geophys. Res. (Atmos.)
53             102: 16,255-16,265.
54       Luke, W. T.; Dickerson, R. R.; Ryan, W. F.; Pickering, K. E.; Nunnermacker, L. J. (1992) Tropospheric  chemistry
5 5             over the lower Great Plains of the United States. 2.  Trace gas profiles and distributions. J.  Geophys. Res.
56             97(018): 20647-20670.
         August 2005                                  AX2-176      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Luke, W. T.; Watson, T. B.; Olszyna, K. J.; Gunter, R. L.; McMillen, R. T.; Wellman, D. L.; Wilkison, S. W. (1998)
  2             A comparison of airborne and surface trace gas measurements during the Southern Oxidants Study (SOS).
  3             J. Geophys. Res. [Atmos.] 103: 22,317-22,337.
  4       Luria, M; Valente, R. J.; Tanner, R. L.; Gillani, N. V.; Imhoff, R. E.; Mueller, S. F.; Olszyna, K. J.; meagher, J. F.
  5             (1999) The evolution of photochemical smog in a power plant plume. Atmos. Environ. 33: 3023-3036.
  6       Luria, M.; Tanner, R. L.; Imhoff, R. E.; Valente, R. J.; Bailey, E. M.; Mueller, S. F. (2000) Influence of natural
  7             hydrocarbons on ozone formation in an isolated power plant plume. J. Geophys. Res. [Atmos.]
  8             105:9177-9188.
  9       Madronich, S. (1987) Photodissociation in the atmosphere. 1. Actinic flux and the effects of ground reflections and
10             clouds. J. Geophys. Res. [Atmos.] 92: 9740-9752.
11       Madronich, S.; McKenzie, R. L.; Caldwell, M. M.; Bjorn, L. O. (1995) Changes in ultraviolet radiation reaching the
12             Earth's surface. Ambio 24: 143-152.
13       Mahlman, J. D. (1997) Dynamics of transport processes in the upper troposphere. Science 276: 1079-1083.
14       Mahlman, J. D.; Pinto, J. P.; Umscheid, L. J. (1994) Transport, radiative, and dynamical effects of the Antarctic
15             ozone hole: a GFDL '' SKYHI" model experiment. J. Atmos. Sci. 51: 489-508.
16       Mahrt, L. (1998) Stratified atmospheric boundary layers and breakdown of models.  Theor. Comput. Fluid Dyn.
17             11:263-279.
18       Martilli, A.; Neftel, A.; Favaro, G.; Kirchner, F.; Sillman, S.; Clappier, A. (2002) Simulation of the ozone formation
19             in the northern part of the Po Valley. J. Geophys. Res. (Atmos.) 107(D22): 10.1029/2001JD000534.
20       Martin, D.; Tsivou, M.; Bonsang, B.; Abonnel, C.; Carsey, T.;  Springer-Young, M.; Pszenny, A.; Suhre, K. (1997)
21             Hydrogen peroxide in the marine atmospheric boundary layer during the Atlantic Stratocumulus Transition
22             Experiment/Marine Aerosol and Gas Exchange experiment in the eastern subtropical North Atlantic.
23             J. Geophys. Res. (Atmos.) 102: 6003-6015.
24       Martin, R. V.; Jacob, D. J.; Chance, K. V.; Kurosu, T. P.; Palmer, P. I.; Evans, M. J. (2003) Global inventory of
25             nitrogen oxide emissions constrained by space-based observations of NO2 columns. J. Geophys. Res.
26             108(D17):  10.1029/2003 JD003453.
27       Martinez, M.; Harder, H.; Kovacs, T. A.; Simpas, J. B.; Bassis, J.; Lesher, R.; Brune, W. H.;  Frost, G. J.; Williams,
28             E. J.; Stroud, C. A.; Jobson, B. T.; Roberts, J. M.; Hall, S. R.; Shelter, R. E.;  Wert, B.; Fried, A.; Alicke, B.;
29             Stutz, J.; Young, V. L.; White, A. B.; Zamora, R. J. (2003) OH and HO2 concentrations, sources, and loss
30             rates during the Southern Oxidants study in Nashville, Tennessee, summer 1999. J. Geophys. Res. (Atmos.)
31             108(019):  10.1029/2003 JD003551.
32       Marufu, L. T.; Taubman, B. F.; Bloomer, B.; Piety, C. A.; Doddridge, B. G.; Stehr,  J. W.; Dickerson, R. R. (2004)
33             The 2003 North American electrical blackout: an accidental experiment in atmospheric chemistry. Geophys.
34             Res. Lett. 31(L13106):  10.1029/2004GL019771.
35       Massman, W. J.; Pederson, J.; Delany, A.; Grantz, D.; Denhartog, G.; Neumann, H. H.; Oncley, S. P.; Pearson, R.;
36             Shaw, R. H. (1994) An evaluation of the regional acid deposition model surface module for ozone uptake at
37             3 sites in the San-Joaquin valley  of California. J. Geophys. Res. [Atmos.] 99: 8281-8294.
38       Matsumi, Y.; Murakami, S.-L; Kono, M. (2001) High-sensitivity instrument for measuring atmospheric NO2. Anal.
39             Chem. 73:5485-5493.
40       Mauldin, R. L., Ill; Tanner, D. J.; Eisele, F. L. (1998) A new chemical ionization mass spectrometer technique for
41             the fast measurement of gas phase nitric acid in the atmosphere. J. Geophys.  Res. 103(D3): 3361-3367.
42       McClenny, W. A., ed. (2000) Recommended methods for ambient air monitoring of NO, NO2, NOy, and individual
43             NOZ species. Research Triangle Park, NC: U.S. Environmental Protection Agency, National Exposure
44             Research Laboratory; report no. GPRA APM #442.
45       McDow, S. R.; Jang, M.; Hong, Y.; Kamens, R. M. (1996) An approach to studying the effect of organic
46             composition on atmospheric aerosol photochemistry. J. Geophys. Res. (Atmos.) 101(D14): 19,593-19,600.
47       McFiggans, G.; Allan, B.; Coe, H.; Plane, H. M. C.; Carpenter, L. J.; O'Dowd, C. (2000) A modeling study of iodine
48             chemistry in the marine boundary layer. J. Geophys. Res. (Atmos.) 105: 14,371-14,385.
49       McFiggans, G.; Cox, R. A.; Mossinger, J. C.; Allan, B. J.; Plane, J. M. C. (2002) Active chlorine release from
50             marine aerosols: roles for reactive iodine and nitrogen species. J. Geophys. Res.
51             17(D15): 10.1029/2001JD000383.
52       McKeen, S. A.; Liu, S. C. (1993) Hydrocarbon ratios and photochemical history of  air masses. Geophys. Res. Lett.
53             20:2363-2366.
54       McKeen, S. A.; Liu, S. C. (1997) Hydrocarbon ratios and the photochemical history of air masses. Geophys. Res.
55             Lett. 20: 2363-2366.
         August 2005                                 AX2-177       DRAFT-DO NOT QUOTE OR CITE

-------
  1       McKeen, S. A.; Liu, S. C.; Hsie, E.-Y.; Lin, X.; Bradshaw, J. D.; Smyth, S.; Gregory, G. L.; Blake, D. R. (1996)
  2             Hydrocarbon ratios during PEM-WEST A: a model perspective. J. Geophys. Res. (Atmos.)
  3             101(D1): 2087-2109.
  4       McKeen, S. A.; Mount, G.; Eisele, F.; Williams, E.; Harder, J.; Goldan, P.; Kuster, W.; Liu, S. C.; Baumann, K.;
  5             Tanner, D.; Fried, A.; Sewell, S.; Cantrell, C.; Shelter, R. (1997) Photochemical modeling of hydroxyl and its
  6             relationship to other species during the Tropospheric OH Photochemistry Experiment. J. Geophys. Res.
  7             (Atmos.) 102: 6467-6493.
  8       McKenna, D. S.; Konopka, P.; GrooB, J.-U.; Gtinther, G.; Mtiller, R.; Spang, R.; Offermann, D.; Orsolini, Y. (2002)
  9             A new chemical Lagrangian model of the stratosphere (CLaMS)  1. formulation of advection and mixing.
10             J. Geophys. Res. 107(D16): 10.1029/2000JD000114.
11       Meagher, J. F.; Cowling, E. B.; Fehsenfeld, F. C.; Parkhurst, W. J. (1998) Ozone formation and transport in
12             southeastern United States: overview of the SOS Nashville/Middle Tennessee Ozone Study. J. Geophys. Res.
13             [Atmos.] 103:22,213-22,223.
14       Mendoza-Dominguez, A.; Russell, A. G. (2000) Iterative inverse modeling and direct sensitivity analysis of a
15             photochemical air quality model. Environ. Sci. Technol. 3: 4974-4981.
16       Mendoza-Dominguez, A.; Russell, A. G. (2001) Estimation of emission adjustments from the application of
17             four-dimensional data assimilation to photochemical air quality modeling. Atmos. Environ. 35: 2879-2894.
18       Meng, Z.; Seinfeld, J. H. (1996) Time scales to achieve atmospheric gas-aerosol equilibrium for volatile species.
19             Atmos. Environ. 30: 2889-2900.
20       Meng, Z.; Dabdub, D.; Seinfeld, J. H. (1997) Chemical coupling between atmospheric ozone and paniculate matter.
21             Science (Washington, DC) 277: 116-119.
22       Mentel, T. F.; Sohn, M.; Wahner, A. (1999) Nitrate effect in the heterogeneous hydrolysis of dinitrogen pentoxide
23             on aqueous aerosols. Phys. Chem. Chem. Phys.  1: 5451-5457.
24       Merrill, J. T.; Moody, J. L. (1996) Synoptic meteorology and transport during the North Atlantic Regional
25             Experiment (NARE) intensive: overview. J. Geophys. Res. [Atmos.] 101: 28,903-28,921.
26       Meyers, T.; Sickles, J.; Dennis, R.; Rusell, K.; Galloway, J.; Church, T. (2001) Atmospheric nitrogen deposition to
27             coastal estuaries and their watersheds. In: Valigura, R. A.; Alexander, R. B.; Castro, M. S.; Meyers, T. P.;
28             Paerl, H. W.; Stacey, P. E.; Turner, R. E., eds. Nitrogen Loading in Coastal Water Bodies: an Atmospheric
29             Perspective. Washington, DC: AGU Press.
3 0       Meyrahn, H.; Pauly, J.; Schneider, W.; Warneck, P. (1986) Quantum yields for photodissociation of acetone in air
31             and an estimate for the life time of acetone in the lower troposphere. J. Atmos.  Chem. 4: 277-291.
32       Michel, A. E.; Usher, C. R.; Grassian, V. H. (2002) Heterogeneous  and catalytic uptake of ozone on mineral oxides
33             and dusts: a Knudsen cell investigation. Geophys. Res. Lett.  29: DOI: 10.1029/2002GLO14896.
34       Michel, A. E.; Usher, C. R.; Grassian, V. H. (2003) Reactive uptake of ozone on mineral oxides and mineral dusts.
35             Atmos. Environ. 37: 3201-3211.
36       Mickley, L. J.; Murti, P. P.; Jacob, D. J.; Logan, J. A.; Koch, D. M.; Rind, D. (1999) Radiative forcing from
37             tropospheric ozone calculated with a unified chemistry-climate model. J. Geophys. Res. [Atmos.]
38             104: 30,153-30,172.
39       Mihelcic, D.; Holland, F.; Hofzumahaus, A.; Hoppe, L.; Konrad, S.; Mtisgen, P.; Patz, H.-W.; Schafer, H.-J.;
40             Schmitz, T.; Volz-Thomas, A.; Bachmann, K.; Schlomski, S.; Platt, U.;  Geyer, A.; Alicke, B.; Moortgat,
41             G. K.  (2003) Peroxy radicals during BERLIOZ and Pabstthum: measurements, radical budgets and ozone
42             production. J. Geophys. Res. [Atmos.]  108(D4): 10.1029/2001JD001014.
43       Milanchus, M. L.; Rao, S. T.; Zurbenko. I. G. (1998) Evaluating the effectiveness of ozone management efforts in
44             the presence of meteorological variability. J. Air Waste Manage. Assoc. 48: 201-215.
45       Milford, J. B.; Russell, A. G.; McRae, G. J. (1989) A new  approach to photochemical pollution control: implications
46             of spatial patterns in pollutant responses to reductions in nitrogen oxides and reactive organic gas emissions.
47             Environ. Sci. Technol. 23: 1290-1301.
48       Milford, J. B.; Gao, D.; Sillman, S.; Blossey, P.; Russell, A. G. (1994) Total reactive nitrogen (NOy) as an indicator
49             of the sensitivity of ozone to reductions in hydrocarbon and NOX emissions. J. Geophys. Res. 99: 3533-3542.
50       Milne, P. J.; Zika, R. G. (1993) Amino acid nitrogen in atmospheric aerosols: occurrence, sources, and
51             photochemical modification. J. Atmos. Chem. 16: 361-398.
52       Milne, P. J.; Prados, A. I.; Dickerson, R. R.; Doddrige, B. G.; Riemer, D. D.; Zika, R. G.; Merrill, J. T.; Moody, J. L.
53             (2000) Nonmethane hydrocarbon mixing ratios in continental outflow air from Eastern North America: export
54             of ozone precursors to Bermuda.  J. Geophys. Res. [Atmos.] 105: 9981-9990.
55       Mochida, M.; Hirokawa, J.; Kajii, Y.; Akimoto, H. (1998) Heterogeneous reactions of C12 with sea salts at ambient
56             temperature: implications for halogen exchange in the atmosphere. Geophys. Res. Lett. 25: 3927-3930.
         August 2005                                 AX2-178       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Mohnen, V. A.; Reiter, E. R. (1977) International conference on oxidants, 1976 - analysis of evidence and
  2             viewpoints: part III. the issue of stratospheric ozone intrusion. Research Triangle Park, NC: U.S.
  3             Environmental Protection Agency, Environmental Sciences Research Laboratory; EPA report no.
  4             EPA-600/3-77-115. Available from: NTIS, Springfield, VA; PB-279 010.
  5       Monks, P. S. (2000) A review of the observations and origins of the spring ozone maximum. Atmos. Environ.
  6             34: 3545-3561.
  7       Monson, R. K.; Harley, P. C.; Litvak, M. E.; Wildermuth, M; Guenther, A. B.; Zimmerman, P. R.; Fall, R. (1994)
  8             Environmental and developmental controls over the seasonal pattern of isoprene emission from aspen leaves.
  9             Oecologia 99: 260-270.
10       Moody,  J. L.; Oltmans, S. J.; Levy, H., II; Merrill, J. T. (1995) Transport climatology of tropospheric ozone:
11             Bermuda, 1988-1991. J. Geophys. Res. [Atmos.] 100: 7179-7194.
12       Moody,  J. L.; Davenport, J. C.; Merrill, J. T.; Oltmans, S. J.; Parrish, D. D.; Holloway, J. S.; Levy, H., II; Forbes,
13             G. L.; Trainer, M.; Buhr, M. (1996) Meteorological mechanisms for transporting O3 over the  western
14             North Atlantic Ocean: a case study for August 24-29, 1993. J. Geophys. Res. 101(D22):  29,213-29,227.
15       Moody,  J. L.; Wimmers, A. J.; Davenport, J. C. (1999) Remotely sensed specific humidity: development of a
16             derived product from the GOES Imager Channel 3. Geophys. Res. Lett. 26: 59-62.
17       Moorthi, S.; Suarez, M. J. (1992) Relaxed Arakawa-Schubert: a parameterization of moist convection for general
18             circulation models. Mon. Weather Rev. 120: 978-1002.
19       Mopper, K.; Kieber, D. J. (2000) Marine photochemistry and its impact on carbon cycling. In: Demers, S.; Vernet,
20             M., eds. Effects of UV Radiation on Marine Ecosystems. New York, NY: Cambridge University Press, Ch. 4,
21             pp. 101-129.
22       Morales-Morales, R. (1998) Carbon monoxide, ozone, and hydrocarbons in the Baltimore metropolitan area
23             [dissertation]. College Park, MD: University of Maryland, Department of Meteorology. Available from:
24             University Microfilms International, Ann Arbor, MI; publication no. 9908993.
25       Moschonas, N.; Danalatos, D.;  Glavas, S. (1999) The effect of O2 and NO2 on the ring retaining products of the
26             reaction of toluene withhydroxyl radicals. Atmos. Environ. 33: 111-116.
27       Mount, G. H.; Eisele, F. L. (1992) An intercomparison of tropospheric OH measurements at Fritz Peak Observatory,
28             Colorado. Science (Washington, DC) 256: 1187-1190.
29       Mount, G. H.; Williams, E.  J. (1997) An overview of the tropospheric OH photochemistry experiment, Fritz
30             Peak/Idaho Hill, Colorado, fall 1993. J. Geophys. Res. [Atmos.] 102: 6171-6186.
31       Mozurkewich, M. (1995) Mechanisms for the  release of halogens from sea-salt particles by free radical reactions.
32             J. Geophys. Res. (Atmos.) 100: 14,199-14,207.
33       Mtiller, J.-F.; Brasseur, G. (1995) IMAGES: a three-dimensional chemical transport model of the global troposphere.
34             J. Geophys. Res. (Atmos.) 100: 16,445-16,490.
35       Mtiller, J.-F.; Brasseur, G. (1999) Sources of upper tropospheric HOX: a three-dimensional study.  J. Geophys. Res.
36             (Atmos.) 104: 1705-1715.
37       Munger, J. W.; Fan, S.-M.; Bakwin, P. S.; Goulden, M. L.; Goldstein, A. H.; Colman, A. S.; Wofsy, S. C. (1998)
3 8             Regional budgets of nitrogen oxides from continental sources: variations of rates for oxidation and deposition
39             with season and distance from source regions. J. Geophys. Res. (Atmos.) 103: 8355-8368.
40       Munoz,  M. S. S.; Rossi, M. J. (2002) Heterogeneous reactions of HNO3 with flame soot generated under different
41             combustion conditions. Reaction mechanism and kinetics. Phys. Chem. Chem. Phys. 4:5110-5118.
42       Murphy, D.; Fahey, D. (1994) An estimate of the flux of stratospheric reactive nitrogen and ozone into the
43             troposphere. J. Geophys. Res.  (Atmos.) 99: 5325-5332.
44       Narita S.; Uchiyama S.; Takemi K. (1998) The interference and cross sensitivity for an UV ozone monitor. Ozone
45             Sci. Eng. 20: 495-497.
46       National Bureau of Standards. (1975) Catalog of NBS standard reference materials, 1975-76 edition. Washington,
47             DC: U.S. Department of Commerce, National Bureau of Standards; NBS special publication no. 260.
48       National Research Council. (1991) Rethinking the ozone problem in urban and regional air pollution. Washington,
49             DC: National Academy Press.
50       National Research Council. (1999) Ozone-forming potential of reformulated gasoline. Washington, DC: National
51             Academy of Sciences.
52       National Research Council. (2002) Estimating the public health benefits of proposed air pollution regulations.
53             Washington, DC: National Academy of Sciences. Available at:
54             http://books.nap.edu/books/0309086094/html/index.html [7 November, 2002].
55       National Research Council. (2003) Air emissions from animal feeding operations: current knowledge and future
56             needs. Available: http://www.nap.edu/books/0309087058/html/ (5 February 2004).
         August 2005                                 AX2-179       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Neff, J. C.; Holland, E. A.; Dentener, F. I; McDowell, W. H.; Russell, K. M. (2002) Atmospheric organic nitrogen:
  2             implications for the global N cycle. Biogeochemistry 57/58: 99-136.
  3       Neuman, J. A.; Huey, L. G.; Dissly, R. W.; Fehsenfeld, F. C.; Flocke, F.; Holecek, J. C.; Holloway, J. S.; Htibler, G.;
  4             Jakoubek, R.; Nicks, D. K., Jr.; Parrish, D. D.; Ryerson, T. B.; Sueper, D. T.; Weinheimer, A. J. (2002)
  5             Fast-response airborne in-situ measurements of HNO3 during the Texas Air Quality Study. J. Geophys. Res.
  6             (Atmos.) 107: 10.1029/2001JD001437.
  7       Newberg, J.; Williams, D.; Chu, G.; Matthew, B.; Anastasio, C. (2000) Photoformation of hydroxyl radical in
  8             sea-salt particles. In: Preprints of Extended Abstracts presented at the 219th American Chemical Society
  9             National Meeting; March; San Francisco, CA. Columbus, OH: American Chemical Society; ENVR 176.
10       Newchurch, M. J.; Ayoub, M. A.; Oltmans, S.; Johnson, B.; Schmidlin, F. J. (2003) Vertical distribution of ozone at
11             four sites in the United States. J. Geophys. Res. 108(D1): 10.1029/2002JD002059.
12       Newell, R. E.; Thouret, V.; Cho, J. Y. N.; Stoller, P.; Marenco, A.; Smit, H. G.  (1999) Ubiquity of quasi-horizontal
13             layers in the troposphere. Nature (London) 398: 316-319.
14       Nordmeyer, T.; Wang, W.; Ragains, M. L.; Finlayson-Pitts, B. J.; Spicer, C. W.; Plastridge, R. A. (1997) Unique
15             products of the reaction of isoprene with atomic chlorine: potential markers. Geophys. Res. Lett.
16             24: 1615-1618.
17       North American Research Strategy for Tropospheric Ozone (NARSTO) Synthesis Team. (2000) An assessment of
18             tropospheric ozone pollution: a North American perspective. Palo Alto, CA: Electric Power Research
19             Institute.
20       Notholt, J.; Hjorth, J.; Raes, F. (1992a) Formation of HNO2 on aerosol surfaces during foggy periods in the presence
21             of NO and NO2. Atmos. Environ. Part A 26: 211-217.
22       Notholt, J.; Hjorth, J.; Raes, F.; Schrems,  O. (1992b) Simultaneous long path field measurements of HNO2, CH2O
23             and aerosol. Ber. Bunsen Ges. Phys. Chem. 96: 290-293.
24       Nunnermacker, L. J.; Imre, D.; Daum, P. H.; Kleinman, L.; Lee, Y.-N.; Lee, J.  H.; Springston, S. R.; Newman, L.;
25             Weinstein-Lloyd, J.; Luke, W. T.; Banta, R.; Alvarez, R.; Senff, C.; Sillman, S.; Holdren, M.; Keigley, G. W.;
26             Zhou, X. (1998) Characterization of the Nashville urban plume on July 3 and July 18, 1995. J. Geophys. Res.
27             103(D21): 28,129-28,148.
28       Nunnermacker, L. J.; Kleinman, L. I.; Imre, D.; Daum, P. H.; Lee, Y.-N.; Lee, J. H.; Springston, S. R.; Newman, L.;
29             Gillani, N. (2000) NOy lifetimes and O3 production efficiencies in urban and power plant plumes: analysis of
30             field data. J. Geophys. Res. [Atmos.] 105: 9165-9176.
31       O'Brien, R. J.; Green, P. J.; Nguyen, N.-L.; Doty, R. A.; Dumdel, B. E. (1983) Carbon balances in simulated
32             atmospheric reactions: aromatic hydrocarbons. Environ. Sci. Technol. 17: 183-186.
33       Odman, M. T.; Ingram, C. I. (1996) Multiscale air quality simulation platform (MAQSIP): source code
34             documentation and validation. Research Triangle Park, NC: MCNC; technical report ENV-96TR002-vl.O.
35             Available: http://www.ce.gatech.edu/~todman/maqsip.pdf (9 September  2003).
36       Odman, M. T.; Russell, A. G. (1999) Mass conservative coupling of non-hydrostatic meteorological models with air
37             quality models. In: Crying, S.-E.; Batchvarova, E., eds. Air Pollution Modeling and its Application XIII.
3 8             New York, NY: Plenum Press.
39       Odum, J. R.; Hoffmann, T.; Bowman, F.;  Collins, D.; Flagan, R. C.; Seinfeld, J. H. (1996) Gas/particle partitioning
40             and secondary organic aerosol yields. Environ. Sci. Technol. 30: 2580-2585.
41       Olsen, M. A.; Stanford, J. L. (2001) Evidence of stratosphere-to-troposphere transport within a mesoscale model and
42             total ozone mapping spectrometer total ozone. J. Geophys. Res. [Atmos.] 106: 27,323-27,334.
43       Olsen, M. A.; Douglass, A. R.; Schoeberl, M. R. (2002) Estimating downward cross-tropopause flux using column
44             ozone and potential vorticity. J. Geophys. Res. 107(D22): 10.1029/2001JD002041.
45       Olszyna, K. J.; Bailey, E. M.;  Simonaitis, R.; Meagher, J. F. (1994) O3 and NOy relationships at a rural site. J.
46             Geophys. Res.  [Atmos.] 99: 14,557-14,563.
47       Oltmans, S. J.;  Levy, H., II. (1992) Seasonal cycle of surface ozone over the North Atlantic. Nature 358: 392-394.
48       Oltmans, S. J.;  Levy, H., II; Harris, J. M.; Merrill, J. T.; Moody, J. L.; Lathrop,  J. A.; Cuevas, E.; Trainer, M.;
49             O'Neill,  M. S.;  Prospero, J. M.; Vomel, H.; Johnson, B. J. (1996) Summer and spring ozone profiles over the
50             North Atlantic from ozonesonde observations. J. Geophys. Res. 101: 29,179-29,200.
51       Oltmans, S. J.;  Lefohn, A. S.;  Scheel, H. E.; Harris, J. M.; Levy, H., II; Galbally, I. E.; Brunke, E.-G.; Meyer, C. P.;
52             Lathrop, J. A.; Johnson, B. J.; Shadwick, D. S.; Cuevas, E.; Schmidlin, F. J.; Tarasick, D. W.; Claude, H.;
53             Kerr, J. B.; Uchino, O.; Mohnen, V. (1998) Trends of ozone in the troposphere. Geophys. Res.  Lett.
54             25: 139-142.
55       Orlando, J. J.; Tyndall, G. S.; Paulson, S.  E. (1999) Mechanism of the OH-initiated oxidation of methacrolein.
56             Geophys. Res. Lett. 26: 2191-2194.
         August 2005                                 AX2-180       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Padro, J. (1996) Summary of ozone dry deposition velocity measurements and model estimates over vineyard,
  2             cotton, grass and deciduous forest in summer. Atmos. Environ. 30: 2363-2369.
  3       Palmer,  P. I.; Jacob, D. J.; Fiore, A. M; Martin, R. V.; Chance, K.; Kuruso, T. P. (2003) Mapping isoprene
  4             emissions over North America using formaldehyde column observations from space. J. Geophys. Res.
  5             108(D6): 10.1029/2002JD002153.
  6       Pandis,  S. N.; Paulson, S. E.; Seinfeld, J. H.; Flagan, R. C. (1991) Aerosol formation in the photooxidation of
  7             isoprene and p-pinene. Atmos. Environ. Part A 25: 997-1008.
  8       Pandis,  S. N.; Harley, R. A.; Cass, G. R.; Seinfeld, J. H. (1992) Secondary organic aerosol formation and transport.
  9             Atmos. Environ. Part A 26: 2269-2282.
10       Papagni, C.; Arey, J.; Atkinson, R. (2001) Rate constants for the gas-phase reactions of OH radicals with a series of
11             unsaturated alcohols. Int. J. Chem. Kinet. 33: 142-147.
12       Park, R. J.; Stenchikov,  G. L.; Pickering; Dickerson, R. R.; Allen, D. J.; Kondragunta, S. (2001) Regional air
13             pollution and its radiative forcing: studies with a single column chemical and radiation transport model. J.
14             Geophys. Res. [Atmos.] 106: 28,751-28,770.
15       Parrish,  D. D.; Fehsenfeld, F. C. (2000) Methods for gas-phase measurements of ozone, ozone precursors and
16             aerosol precursors. Atmos.  Environ. 34: 1921-1957.
17       Parrish,  D. D.; Trainer, M.; Buhr,  M. P.; Watkins, B. A.; Fehsenfeld, F. C. (1991) Carbon monoxide concentrations
18             and their relatin to concentrations of total reactive oxided nitrogen at two rural U.S. sites. J. Geophys. Res.
19             (Atmos.) 96: 9309-9320.
20       Parrish,  D. D.; Hahn, C. J.; Williams, E. J.; Norton, R. B.; Fehsenfeld, F.  C.; Singh, H. B.; Shelter, J. D.; Gandrud,
21             B. W.; Ridley, B. A. (1992) Indications of photochemical histories of Pacific air masses from measurements
22             of atmospheric trace species at Point Arena, California. J. Geophys. Res. [Atmos.] 97: 15,883-15,901.
23       Parrish,  D. D.; Hahn, C. J.; Williams, E. J.; Norton, R. B.; Fehsenfeld, F.  C.; Singh, H. B.; Shelter, J. D.; Gandrud,
24             B. W.; Ridley, B. A. (1993) Reply [to Comment on "Indications of photochemical histories of Pacific air
25             masses from measurements of atmospheric trace species at Point Arena,  California" by D. D. Parrish et al.].
26             J. Geophys. Res.  (Atmos.) 98(D8): 14,995-14,997.
27       Parrish,  D. D.; Trainer, M.; Young, V.; Goldan, P. D.; Kuster, W. C.; Jobson, B. T.; Fehsenfeld, F. C.; Lonneman,
28             W. A.; Zika, R. D.; Farmer, C. T.; Riemer,  D. D.;  Rodgers, M. O. (1998) Internal consistency tests for
29             evaluation of measurements of anthropogenic hydrocarbons in the  troposphere. J. Geophys. Res. (Atmos.)
30             103(017): 22,339-22,359.
31       Parrish,  D. D.; Ryerson, T. B.; Holloway, J. S.; Trainer, M.; Fehsenfeld, F. C. (1999) Does pollution increase or
32             decrease tropospheric ozone in winter-spring? Atmos. Environ. 33: 5147-5149.
33       Parrish,  D. D.; Holloway, J.  S.; Jakoubek, R.; Trainer, M.; Ryerson, T. B.; Htibler, G.; Fehsenfeld, F. C.; Moody,
34             J. L.; Cooper. O.  R. (2000) Mixing of anthropogenic pollution with stratospheric ozone: a case study from the
35             North Atlantic wintertime troposphere. J. Geophys. Res. [Atmos.]  105: 24,363-24,374.
36       Parrish,  D. D.; Trainer, M.; Hereid, D.; Williams,  E. J.; Olszyna, K. J.; Harley, R. A.; Meagher, J. F.; Fehsenfeld,
37             F. C. (2002) Decadal change in carbon monoxide to nitrogen oxide ratio in U.S. vehicular emissions
38             J. Geophys. Res.  (Atmos.)  107(D13): 10.1029/2001JD000720.
39       Paul, J. W.; Beauchamp, E. G. (1993) Nitrogen availability for corn in soils amended with urea, cattle slurry, and
40             solid and composted manures. Can. J. Soil  Sci. 73: 253-266.
41       Paul, J. W.; Beauchamp, E. G.; Zhang, X. (1993) Nitrous and nitric oxide emissions during nitrification and
42             denitrification from manure-amended soil in the laboratory. Can. J. Soil Sci. 73: 539-553.
43       Penner,  J. E. (1995) Carbonaceous aerosols influencing atmospheric radiation: black and organic carbon. In:
44             Charlson, R. J.; Heintzenberg, J., eds. Aerosol Forcing of Climate. New York, NY:  John Wiley and Sons;
45             pp. 91-108.
46       Penrose, W. R..; Pan, L.; Stetter, J. R.; Ollison, W. M. (1995) Sensitive measurement of ozone using amperometric
47             gas sensors. Anal. Chim. Acta 313: 209-219.
48       Perry, R. A.; Atkinson, R.; Pitts, J. N., Jr. (1977) Kinetics and mechanism of the gas phase reaction of OH radicals
49             with aromatic hydrocarbons over the temperature range 296-473 K. J. Phys. Chem.  81: 296-304.
50       Petriconi, G. L.; Papee, H. M. (1972) On the photolytic separation of halogens from sea water concentrates. Water
51             Air Soil Pollut. 1: 117-131.
52       Pickering, K. E.; Dickerson, R. R.; Huffman, G. J.; Boatman, J. F.; Schanot, A. (1988) Trace gas transport in the
53             vicinity of frontal convective clouds. J. Geophys. Res. 93: 759-773.
54       Pickering, K. E.; Thompson, A. M.; Dickerson, R. R.; Luke, W. T.; McNamara, D. P.; Greenberg, J. P.;
55             Zimmerman, P. R. (1990) Model calculations of tropospheric ozone production potential following observed
56             convective events. J. Geophys. Res. [Atmos.] 95:  14,049-14,062.
         August 2005                                  AX2-181       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Pickering, K. E.; Thompson, A. M; Scala, J. R.; Tao, W.-K.; Simpson, J.; Garstang, M. (1991) Photochemical ozone
  2             production in tropical squall line convection during NASA Global Tropospheric Experiment/Amazon
  3             Boundary Layer Experiment 2A. J. Geophys. Res. 96: 3099-3114.
  4       Pickering, K. E.; Thompson, A. M.; Scala, J. R.; Tao, W.-K.; Simpson, J. (1992a) Ozone production potential
  5             following convective redistribution of biomass burning emissions. J. Atmos. Chem. 14: 297-313.
  6       Pickering, K. E.; Thompson, A. M.; Scala, J. R.; Tao, W.-K.; Dickerson, R. R.; Simpson, J. (1992b) Free
  7             tropospheric ozone production following entrainment of urban plumes into deep convection. J.  Geophys. Res.
  8             97(016):  17985-18,000.
  9       Pickering, K. E.; Scala, J. R.; Thompson, A. M.; Tao, W.-K.; Simpson, J. (1992c) A regional estimate of the
10             convective transport of CO from biomass burning. Geophys. Res. Lett. 19: 289-292.
11       Pickering, K. E.; Thompson, A. M.; Tao, W.-K.; Kucsera, T. L. (1993) Upper tropospheric ozone production
12             following mesoscale convection during STEP/EMEX. J. Geophys. Res. 98: 8737-8749.
13       Pickering, K. E.; Thompson, A. M.; Tao, W.-K.; Rood, R. B.; McNamara, D. P.; Molod, A. M. (1995) Vertical
14             transport by convective clouds: comparisons of three modeling approaches. Geophys. Res. Lett. 22:
15             1089-1092.
16       Pickering, K. E.; Thompson, A. M.; Wang,  Y.; Tao, W.-K.; McNamara, D. P.; Kirchhoff, W. J. H.; Heikes, B. G.;
17             Sachse, G. W.; Bradshaw, J. D.; Gregory, G. L.; Blake, D. R. (1996) Convective transport of biomass burning
18             emissions over Brazil during TRACE A. J. Geophys. Res. [Atmos.] 101: 23,993-24,012.
19       Pickering, K. E.; Wang, Y.; Tao, W.-K.; Price, C.; Mtiller, J.-F. (1998) Vertical distributions of lightning NOX for
20             use in regional and global chemical transport models. J. Geophys. Res. [Atmos.]  103: 31,203-31,216.
21       Pickering, K. E.; Thompson, A. M.; Kim, H.; DeCaria, A. J.; Pfister, L.; Kucsera, T. L.; Witte, J. C.; Avery, M. A.;
22             Blake, D. R.; Crawford, J. H.; Heikers, B. G.; Sachse, G. W.; Sandholm, S. T.; Talbot, R. W. (2001) Trace
23             gas transport and scavenging ini PEM-Tropics B South Pacific Convergence Zone convection.  J. Geophys.
24             Res. [Atmos.] 106: 32,591-32,602.
25       Pierce, T. E.; Waldruff, P. S. (1991) PC-BEIS: a personal computer version of the biogenic emissions inventory
26             system. J. Air Waste Manage. Assoc. 41: 937-941.
27       Pierce, T; Geron, C.; Bender, L.; Dennis, R.; Tonnesen, G.; Guenther, A. (1998) Influence of increased isoprene
28             emissions on regional ozone modeling. J. Geophys. Res. (Atmos.) 103: 25,611-25,630.
29       Pierson, W. R.; Gertler, A. W.; Bradow, R.  L. (1990) Comparison of the SCAQS tunnel study with historical data.
30             Presented at: 83rd annual meeting & exhibition of the Air and Waste Management Association; June;
31             Pittsburgh, PA.  Pittsburgh, PA: Air and Waste Management Association; paper no. 90-175.3.
32       Piety, C. A. (2005) The relation between daily maximum ozone and daily maximum temperature [memorandum to
33             Dr. Joseph Pinto]. Research Triangle Park, NC: U.S. Environmental Protection Agency; July 18.
34       Pinto, J. P.; Bruhl, C.;  Thompson, A. M. (1993) The current and future envirionmental role of atmospheric methane.
35             In: Khalil, M. A. K., ed. Atmospheric methane sources, sinks, and role in global change, p. 514-531.
36             (NATO ASI Series, v. 113).
37       Piotrowicz, S. R.; Bezdek, H. F.; Harvey, G. R.; Springer-Young, M.; Hanson, K. J. (1991) On the ozone minimum
38             over the equatorial Pacific Ocean. J.  Geophys. Res. 96: 18,679-18,687.
39       Pippin, M.  R.; Bertman, S.; Thornberry, T.; Town, M.; Carroll, M. A.; Sillman, S. (2001) Seasonal variation of
40             PAN, PPN, MPAN and O3 at the upper midwest PROPHET site. J. Geophys. Res. (Atmos.) 106:
41             24,451-24,463.
42       Pitari, G.; Grassi, B.; Visconti, G. (1997) Results of a chemical transport model with interactive aerosol
43             microphysics. In: Bojkov, R.; Visconti, G., eds. Proceedings of the XVIII Quadrennial Ozone Symposium;
44             September 1996; L'Aquila, Italy, pp. 759-762.
45       Pitts, J. N., Jr.; Sanhueza, E.; Atkinson, R.;  Carter, W. P. L.; Winer, A. M.; Harris, G. W.; Plum, C. N. (1984a)
46             An investigation of the dark formation of nitrous acid in environmental chambers. Int.  J. Chem. Kinet.
47             16:919-939.
48       Pitts, J. N., Jr.; Biermann, H. W.; Atkinson, R.; Winer, A. M. (1984b) Atmospheric implications of simultaneous
49             nighttime measurements of NO3 radicals and HONO. Geophys. Res. Lett. 11: 557-560.
50       Placet, M.; Mann, C. O.;  Gilbert, R. O.; Niefer, M. J. (2000) Emissions of ozone precursors from stationary sources:
51             a critical review. Atmos.  Environ. 34: 2183-2204.
52       Plummer, D. A.; McConnell, J.  C.; Shepson, P. B.; Hastie, D. R.; Niki, H. (1996) Modeling of ozone formation at a
53             rural site in southern Ontario. Atmos. Environ. 30: 2195-2217.
54       Pokharel, S. S.; Bishop, G. A.; Stedman, D. H. (2002) An on-road motor vehicle emissions inventory  for Denver:
55             an efficient alternative to modeling. Atmos. Environ. 36: 5177-5184.
         August 2005                                AX2-182       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Poppe, D.; Wallasch, M.; Zimmermann, J. (1993) The dependence of the concentration of OH on its precursors
  2             under moderately polluted conditions: a model study. J. Atmos. Chem. 16: 61-78.
  3       Poppe, D.; Zimmermann, J.; Bauer, R.; Brauers, T.; Briining, D.; Callies, J.; Dorn, H.-P.; Hofzumahaus, A.; Johnen,
  4             F. -J.; Khedim, A.; Koch, H.; Koppmann, R.; London, H.; Mtiller, K. -P.; Neuroth, R.; Plass-Dtilmer, C.;
  5             Platt, U.; Rohrer, F.; Roth, E.-P.; Rudolph, J.; Schmidt, U.; Wallasch, M.; Ehhalt, D. H. (1994) Comparison
  6             of measured OH concentrations with model calculations. J. Geophys. Res. [Atmos.] 99: 16,633-16,642.
  7       Poppe, D.; Zimmerman, J.; Dorn, H. P. (1995) Field data and model calculations for the hydroxyl radical. J. Atmos.
  8             Sci. 52: 3402-3407.
  9       Poulida, O.; Dickerson, R. R.; Heymsfield, A. (1996) Troposphere-stratosphere exchange in a midlatitude mesoscale
10             convective complex: 1. Observations. J. Geophys. Res. [Atmos.] 101: 6823-6836.
11       Prados, A. I. (2000) The export of ozone and other atmospheric trace species from North America: observations and
12             numerical simulations [dissertation]. College Park, MD: The University of Maryland.
13       Prados, A. I.; Dickerson, R. R.; Doddridge, B.  G.; mime, P. A.; Moody, J. L.; Merrill, J. T. (1999) Transport of
14             ozone and pollutants from North America to the North Atlantic Ocean during the 1996 AEROCE intensive
15             experiment. J. Geophys. Res. 104: 26,219-26,233.
16       Prather, M.; Ehhalt, D. (2001) Atmospheric chemistry and greenhouse gases. In: Houghton, J. T.; Ding, Y.; Griggs,
17             D. J.; Noguer, M.; van der Linden, P. J.; Dai, X.; Maskell, K.; Johnson, C. A., eds. Climate Change 2001:
18             The Scientific Basis; pp. 239-287. Cambridge, MA: Cambridge University Press. Available:
19             http://www.grida.no/climate/ipcc_tar/wgl/ (20 January 2004).
20       Prather, M. J.; Jacob, D. J. (1997) A persistent imbalance in HOX and NOX photochemistry in the upper troposphere
21             driven by deep tropical convection. Geophys. Res. Lett. 24: 3189-3192.
22       Prenni, A. J.; DeMott, P. J.; Kreidenweis, S. M. (2003) Water uptake of internally mixed particles containing
23             ammomium sulfate and dicarboxylic acids. Atmos. Environ.  37: 4243-4251.
24       Prevot, A. S. H.; Staehelin, J.; Kok, G. L.; Schillawski, R. D.; Neininger, B.; Staffelbach, T.; neftel, A.; Wernli, H.;
25             Dommen, J. (1997) The Milan photooxidant plume. J. Geophys. Res. (Atmos.) 102: 23,375-23,388.
26       Price, J. D.; Vaughan, G. (1993) The potential for stratosphere-troposphere exchange in cut-off-low systems. Q. J. R.
27             Meteorol. Soc. 119:343-365.
28       Price, C.; Penner, J.; Prather, M. (1997) NOX from lightning: 1. Global distribution based on lightning physics.
29             J. Geophys. Res. [Atmos.] 102: 5929-5941.
30       Pszenny, A. A. P.; Keene, W. C.; Jacob, D. J.;  Fan, S.; Maben, J. R.; Zetwo, M. P.; Springer-Young, M.; Galloway,
31             J. N. (1993) Evidence of inorganic chlorine gases other than hydrogen chloride in marine surface air.
32             Geophys. Res. Lett. 20:  699-702.
33       Pszenny, A. A. P.; Moldanova, J.; Keene, W. C.; Sander, R.; Maben, J. R.; Martinez, M.; Crutzen, P. J.; Perner, D.;
34             Prinn, R. G. (2004) Halogen cycling and aerosol pH in the Hawaiian marine boundary  layer. Atmos. Chem.
35             Phys. 4: 147-168.
36       Pun, B. K.; Seigneur, C.; Grosjean,  D.; Saxena, P. (2000) Gas-phase formation of water-soluble organic compounds
37             in the atmosphere: a retrosynthetic analysis. J. Atmos. Chem. 35: 199-223.
38       Quigley, D. C.; Hornafius, J. S.; Luyendyk, B. P.; Francis, R. D.; Clark, J.; Washburn, L. (1999) Decrease in natural
39             marine hydrocarbon seepage near Coal Oil Point, California, associated with offshore oil production. Geol.
40             27: 1047-1050.
41       Rancher, J.; Kritz, M. A. (1980) Diurnal fluctuations of Brand I in the tropical marine atmosphere. J. Geophys. Res.
42             85: 5581-5587.
43       Rao, S. T.; Zurbenko, I. G. (1994) Detecting and tracking changes in ozone air quality. J. Air Waste Manage. Assoc.
44             44: 1089-1092.
45       Rao, S. T.; Zurbenko, I. G.; Neagu, R.; Porter, P. S.; Ku, J. Y.; Henry, R. F. (1997) Space and time scales in ambient
46             ozone data. Bull. Am. Meteorol. Soc. 78: 2153-2166.
47       Rao, S. T.; Ku, J. Y.; Berman, S. (2003) Summertime characteristics of the atmospheric boundary layer and
48             relationships to ozone levels  over the eastern United States. Pure Appl. Geophys. 160:  21-55.
49       Reed, R. J. (1955) A study of a characteristic type of upper-level frontogenesis. J. Meteorol. 12: 226-237.
50       Regener, E. (1941) Ozonschicht und atmospharische turbulenz. Ber. Dtsch. Wetterdienstes US-Zone No. 11: 45-57.
51       Reid, J. P.; Sayer, R. M. (2003) Heterogeneous atmospheric aerosol chemistry: laboratory studies of chemistry on
52             water droplets. Chem. Soc. Rev. 32: 70-79.
53       Reisinger, A. R. (2000) Unidentified interference inDOAS measurements of ozone. Appl. Spectrosc. 54: 72-79.
54       Reissell, A.; Arey, J.; Atkinson, R. (2001) Atmospheric chemistry of camphor. Int. J. Chem. Kinet. 33: 56-63.
55       Reitebuch, O.; Strassburger, A.; Emeis, S.; Kuttler, W. (2000) Nocturnal  secondary ozone concentration maxima
56             analysed by sodar observation and surface measurements. Atmos. Environ. 34: 4315-4329.
         August 2005                                  AX2-183       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Reiter, E. R. (1963) A case study of radioactive fallout. J. Appl. Meteorol. 2: 691-705.
  2       Reiter, E. R. (1975) Stratospheric-tropospheric exchange processes. Rev. Geophys. Space Phys. 13: 459-474.
  3       Reiter, E. R.; Mahlman, J. D. (1965) Heavy radioactive fallout over the southern United States, November 1962. J.
  4             Geophys. Res. 70: 4501-4520.
  5       Reithmeier, C.; Sausen, R. (2002) ATTILA: atmospheric tracer transport in a Lagrangian model. Tellus 54B:
  6             278-299.
  7       Ren, X. R.; Harder, H.; Martinez, M; Lesher, R. L.; Oliger, A.; Shirley, T.; Adams, J.; Simpas, J. B.; Brune, W. H.
  8             (2003) HOx concentrations and OH reactivity observations in New York City during PMTACS-NY2001.
  9             Atmos. Environ. 37: 3627-3637.
10       Reynolds, S.; Michaels, H.; Roth, P.; Tesche, T. W.; McNally, D.; Gardner, L.; Yarwood, G. (1996) Alternative base
11             cases in photochemical modeling: their construction, role, and value. Atmos. Environ. 30: 1977-1988.
12       Ridley, B. A.; Carroll, M. A.; Dunlap, D. D.; Trainer, M.; Sachse, G. W.; Gregory, G. L.; Condon, E. P. (1989)
13             Measurements of NOX over the eastern Pacific ocean and southwestern United States during the  spring 1984
14             NASA GTE aircraft program. J. Geophys. Res. [Atmos.] 94: 5043-5067.
15       Ridley, B. A.; Walega, J. G.; Lamarque, J.-F.; Grahek, F. E.; Trainer, M.; Htibler, G.; Lin, X.; Fehsenfeld, F. C.
16             (1998) Measurements of reactive nitrogen and ozone to 5-km altitude in June 1990 over the southeastern
17             United States. J. Geophys. Res. 103: 8369-8388.
18       Riemer, D. D.; Apel, E. C. (2001) Confirming the presence and extent of oxidation by Cl in the Houston, Texas
19             urban area using specific isoprene oxidation products as tracers. Miami, FL: University of Miami, Rosenstiel
20             School of Marine and Atmospheric  Science. Available:
21             http://clearinghouse.ces.utexas.edu/TexAQS/DB_Files/ConfirmingPresenceandExtentOfOxidationByCI.pdf
22             (5 February 2004).
23       Riemer, N.; Vogel, H.; Vogel, B.; Schell, B.; Ackermann, L; Kessler, C.; Hass, H. (2003) Impact of the
24             heterogeneous hydrolysis of N2O5 on chemistry and nitrate aerosol formation in the lower troposphere under
25             photosmog conditions. J. Geophys. Res. (Atmos.) 108(D4): 10.1029/2002JD002436.
26       Roberts, J. M.; Williams, J.; Baumann, K.; Buhr, M. P.; Goldan, P. D.; Holloway, J.; Htibler, G.; Kuster, W. C.;
27             McKeen, S. A.; Ryerson, T. B.; Trainer, M.; Williams, E. J.; Fehsenfeld, F. C.; Bertman, S. B.; Nouaime, G.;
28             Seaver, C.; Grodzinsky, G.; Rodgers, M.; Young, V. L. (1998) Measurements of PAN, PPN, and MPAN
29             made during the 1994 and 1995 Nashville Intensives of the Southern Oxidant Study: implications for regional
30             ozone production frombiogenic hydrocarbons. J.  Geophys. Res. [Atmos.] 103: 22,473-22,490.
31       Robinson, G. N.; Worsnop, D. R.; Jayne, J. T.; Kolb, C. E.; Davidovits, P. (1997) Heterogeneous uptake of C1ONO2
32             and N2O5 by sulfuric acid solutions. J. Geophys. Res. (Atmos.) 102: 3583-3601.
33       Rodgers, M. O.; Davis, D. D. (1989) A UV-photofragmentation/laser-induced fluorescence sensor for the
34             atmospheric detection of HONO. Environ. Sci. Technol. 23:  1106-1112.
35       Roelofs, G. J.; Scheeren, H. A.; Heland, J.; Ziereis, H.; Lelieveld, J. (2003) A model study of ozone in  the eastern
36             Mediterranean free troposphere during MINOS (August 2001). Atmos. Chem. Phys. 3: 1199-1210.
37       Rogers, J. D. (1990) Ultraviolet absorption cross sections and atmospheric photodissociation rate constants of
38             formaldehyde. J. Phys. Chem. 94: 4011-4015.
39       Roselle, S. J. (1994) Effects of biogenic emission uncertainties on regional photochemical modeling of control
40             strategies. Atmos. Environ. 28: 1757-1772.
41       Rudich, Y.; Talukdar, R. K.; Fox, R. W.; Ravishankara, A. R. (1996) Rate coefficients for reactions of NO3 with a
42             few olefins and oxygenated olefins. J. Phys. Chem.  100: 5374-5381.
43       Rudolph, J.; Ramacher, B.; Plass-Dulmer, C.; Mtiller, K.-P.; Koppmann, R. (1997) The indirect determination of
44             chlorine atom concentration in the troposphere from changes in the patterns of non-methane hydrocarbons.
45             Tellus 649:592-601.
46       Ruppert, L.; Becker, K. H. (2000) A product study of the OH radical-initiated oxidation of isoprene: formation of
47             C5-unsaturated diols. Atmos. Environ. 34: 1529-1542.
48       Russell, K. M. (2002) Characterization and dry deposition of atmospheric nitrogen at the mid-Atlantic United States
49             coast [dissertation]. Richmond, VA: University of Virginia. Available from: University Microfilms,
50             Ann Arbor, MI;  AADAA-13044880.
51       Russell, A.; Dennis, R. (2000) NARSTO critical review of photochemical models and modeling. Atmos. Environ.
52             34:2283-2324.
53       Ryan, W. F.; Doddridge, B. G.; Dickerson, R. R.; Morales, R. M.; Hallock, K. A.; Roberts, P. T.; Blumenthal, D. L.;
54             Anderson, J. A.; Civerolo, K. (1998) Pollutant transport during a regional O3 episode in the mid-Atlantic
55             states. J. Air Waste Manage. Assoc. 48: 786-797.
         August 2005                                 AX2-184       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Ryerson, T. B.; Buhr, M. P.; Frost, G. I; Goldan, P. D.; Holloway, J. S.; Htibler, G.; Jobson, B. T.; Kuster, W. C.;
  2             McKeen, S. A.; Parrish, D. D.; Roberts, J. M.; Sueper, D. T.; Trainer, M.; Williams, J.; Fehsenfeld, F. C.
  3             (1998) Emissions lifetimes and ozone formation in power plant plumes. J. Geophys. Res.  103(D17):
  4             22,569-22,583.
  5       Ryerson, T. B.; Williams, E J.; Fehsenfeld, F. C. (2000) An efficient photolysis system for fast-response NO2
  6             measurements. J. Geophys. Res. (Atmos.) 105(D21): 26,447-26,461.
  7       Ryerson, T. B.; Trainer, M.; Holloway, J. S.; Parrish, D. D.; Huey, L. G.; Sueper, D. T.; Frost, G. J.; Donnelly,
  8             S. G.; Schauffler, S.; Atlas, E. L.; Kuster, W. C.; Goldan, P. D.; Htibler, G.; Meagher, J. F.; Fehsenfeld, F. C.
  9             (2001) Observations of ozone formation in power plant plumes and implications for ozone control strategies.
10             Science (Washington, DC) 292: 719-723.
11       Saathoff, H.; Naumann, K. H.; Riemer, N.; Kamm, S.; Mohler, O.; Schurath, U.; Vogel, H.; Vogel, B. (2001) The
12             loss of NO2, HNO3,,NO3/N2O5 and HO2/HOONO2 on soot aerosol: a chamber and modeling study. Geophys.
13             Res. Lett. 28:  1957-1960.
14       Sakamaki, F.; Hatakeyama, S.; Akimoto, H. (1983) Formation of nitrous acid and nitric oxide in the heterogeneous
15             dark reaction of nitrogen dioxide and water vapor in a smog chamber. Int. J. Chem. Kinet. 15: 1013-1029.
16       Sakugawa, H.; Kaplan, I. R. (1989) H2O2 and O3 in the atmosphere of Los Angeles and its vicinity: factors
17             controlling their formation and their role as oxidants of SO2. J. Geophys. Res. [Atmos.] 94: 12,957-12,973.
18       Salmond, J. A.; McKendry, I. G. (2002) Secondary ozone maxima in a very stable nocturnal boundary layer:
19             observations from the Lower Fraser Valley, BC. Atmos. Environ. 36: 5771-5782.
20       Samson, P. J.; Shi, B. (1988) A meteorological investigation of high ozone values in American cities.  Ann Arbor,
21             MI: University of Michigan, Space Physics Research Laboratory, Department of Atmospheric,  Oceanic, and
22             Space Sciences.
23       Sander, R.; Crutzen, P. J. (1996) Model study indicating halogen activation and ozone destruction in polluted air
24             masses transported to the sea. J. Geophys. Res. [Atmos.] 101: 9121-9138.
25       Sander, R.; Rudich, Y.; Von Glasow, R.; Crutzen, P. J. (1999) The role of BrNO3 in marine tropospheric chemistry:
26             a model study. Geophys. Res. Lett. 26: 2858-2860.
27       Sander, R.; Keene, W. C.; Pszenny, A. A. P.; Arimoto, R. Ayers, G. P.; Chainey, J.  M.; Crutzen, P. J.; Duce, R. A.;
28             Huebert, B. J.; Maenhaut, W.; Turekian, V. C.; VanDingenan, R. (2003) Inorganic bromine in the marine
29             boundary layer: a critical review. Atmos.  Chem. Phys. 3: 1301-1336.
30       Santer, B. D.; Wehner, M. F.; Wigley, T. M. L.; Sausen, R.; Meehl, G. A.; Taylor, K. E.;  Ammann, C.; Arblaster, J.;
31             Washington, W. M.; Boyle, J. S.; Bruggemann, W. (2003) Contributions of anthropogenic and  natural forcing
32             to recent tropopause height changes. Science 301: 479-483.
33       Sarwar, G.; Corsi, R.; Kumura, Y.; Allen, D.; Weschler, C. J. (2002) Hydroxyl radicals in indoor environments.
34             Atmos. Environ. 36: 3973-3988.
35       Sauer, F.; Schaefer, C.; Neeb, P.; Horie, O.; Moortgat, G. (1999) Formation of hydrogen peroxide in the ozonolysis
36             of isoprene and simple alkenes. Atmos. Environ. 33: 229-241.
37       Sawyer, R. F.; Harley, R. A.; Cadle,  S. H.; Norbeck, J. M.; Slott, R.; Bravo, H. A. (2000) Mobile sources critical
38             review: 1998 NARSTO assessment. Atmos. Environ. 34: 2161-2181.
39       Saxena, P.; Hildemann, L. M.; McMurry, P. H.;  Seinfeld, J. H.  (1995) Organics alter hygroscopic behavior of
40             atmospheric particles. J. Geophys. Res. [Atmos.] 100: 18,755-18,770.
41       Saylor, R. D. (1997) An estimate of the potential significance of heterogeneous loss to aerosols as an additional sink
42             for hydroperoxy radicals in the troposphere. Atmos. Environ. 31: 3653-3658.
43       Scala, J. M.; Garstang, M.; Tao, W.-K.; Pickering, K. E.; Thompson, A. M.; Simpson, J.;  Kirchhoff, V W. J. H.;
44             Browell, E. V; Sachse, G. W.; Torres. A. L.; Gregory, G. L.; Rasmussen, R. W.; Khalil, M. A. K. (1990)
45             Cloud draft structure and trace gas transport.  J. Geophys. Res. 95: 17,015-17,030.
46       Schere, K. L.; Wayland, R. A. (1989) Development and evaluation of the regional oxidant model forthe northeastern
47             United States. In: Schneider, T.; Lee, S. D.; Wolters, G.  J. R.; Grant, L. D., eds. Atmospheric ozone research
48             and its policy implications: proceedings of the 3rd US-Dutch international symposium; May 1988; Nijmegen,
49             The Netherlands. Amsterdam, The Netherlands: Elsevier Science Publishers; pp. 613-622. (Studies in
50             environmental science 35).
51       Schere, K. L.; Wayland, R. A. (1989) EPA regional oxidant model (ROM2.0): evaluation on 1980 NEROS data
52             bases. Research Triangle Park, NC: U.S. Environmental Protection Agency, Atmospheric Research and
53             Exposure Assessment Laboratory; report  no. EPA/600/3-89/057. Available from: NTIS, Springfield, VA;
54             PB89-200828/HSU.
55       Schichtel, B. A.; Husar, R. B. (1998) Eastern U.S. transport climatology during average, high and low ozone days.
56             Online paper 98-A884. Available: http://capita.wustl.edu/capita/CapitaReports (30 October 2003).
         August 2005                                 AX2-185       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Schichtel, B. A.; Husar, R. B. (2001) Eastern North American transport climatology during high- and low-ozone
  2             days. Atmos. Environ. 35: 1029-1038.
  3       Schneider, M; Ballschmiter, K. (1999) C3-C14-alkyl nitrates in remote South Atlantic air. Chemosphere 38: 233-244.
  4       Schneider, M.; Luxenhofer, O.; Deissler, A.; Ballschmiter, K. (1998) C{-C15 alkyl nitrates, benzyl nitrate, and
  5             bifunctional nitrates: Measurements in California and South Atlantic air and global comparison using C2C14
  6             and CHBr3 as marker molecules. Environ. Sci. Technol. 32: 3055-3062.
  7       Schubert, S. D.; Rood, R. B.; Pfaendtner, J. (1993) An assimilated dataset for earth science applications. Bull. Am.
  8             Meteorol. Soc.  74: 2331-2342.
  9       Schultz, M. G.; Jacob, D. J.; Bradshaw, J. D.; Sandholm, S. T.; Dibb, J. E.; Talbot, R. W.; Singh, H. B. (2000)
10             Chemical NOX budget in the upper troposphere over the tropical South Pacific. J. Geophys. Res.
11             105: 6669-6679.
12       Schumann, U.; Schlager, H.; Arnold, F.; Ovarlez, J.; Kelder, H.; Hov, 0.; Hayman, G.; Isaksen, I. S.  A.;
13             Staehelin, J.; Whitefield, P. D. (2000) Pollution from aircraft emissions in the North Atlantic flight corridor:
14             overview on the POLINAT projects. J. Geophys. Res. 105(D3): 10.1029/1999JD900941.
15       Schutze, M.; Herrmann, H. (2002) Determination of phase transfer parameters for the uptake of HNO3, N2O5 and O3
16             on single aqueous drops. Phys. Chem. Chem. Phys.  4: 60-67.
17       Scudlark, J. R.; Russell, K. M.; Galloway, J. N.; Church, T. M.; Keene, W. C. (1998) Dissolved organic nitrogen in
18             precipitation at the mid-Atlantic U.S. coast—methods evaluation and preliminary data. Atmos. Environ.
19             32:  1719-1728.
20       Seaman, N. L. (2000)  Meteorological modeling for air quality assessments. Atmos. Environ.  34: 2231-2259.
21       Seila, R. L.; Lonneman, W. A.; Meeks, S. A. (1989) Determination of C2 to C12 ambient air hydrocarbons in 39 U.S.
22             cities from 1984 through 1986. Research Triangle Park, NC: U.S. Environmental Protection Agency,
23             Atmospheric Research and Exposure Assessment Laboratory; report no. EPA/600/3-89/058.
24       Seinfeld, J. H. (1988)  Ozone air quality models: a critical review. JAPCA 38: 616-645.
25       Seinfeld, J. H. (1989)  Urban air pollution: state of the science. Science (Washington, DC) 243: 745-752.
26       Seinfeld, J. H.; Pandis, S. N. (1998) Atmospheric chemistry and physics: from air pollution to climate change.
27             New York, NY: John Wiley & Sons, Inc.
28       Seinfeld, J. H.; Pankow, J. F. (2003) Organic atmospheric paniculate material. Ann. Rev. Phys. Chem. 54: 121-140.
29       Senff, C. J.; Hardesty, R. M.; Alvarez, R. J., II; Mayor, S. D. (1998) Airborne lidar characterization of power plant
30             plumes during the 1995 Southern Oxidants Study. J. Geophys. Res. [Atmos.] 103: 31,173-31,189.
31       Seo, K.-H.; Bowman,  K. P.  (2001) A climatology of isentropic cross-tropopause exchange. J. Geophys. Res.
32             [Atmos.] 106: 28,159-28,172.
33       Seo, K.-H.; Bowman,  K. P.  (2002) Lagrangian estimate of global stratosphere-troposphere mass exchange.
34             J. Geophys. Res. (Atmos.) 107(D21): 10.1029/2002JD002441.
3 5       Sexton, K. (1982) Evidence of an additive effect for small  city plumes. Presented at: 75th annual  meeting of the Air
36             Pollution Control Association; June; New Orleans, LA. Pittsburgh, PA: Air Pollution Control Association;
37             paper no. 82-31.4.
38       Sexton, K.; Westberg, H.  (1980) Elevated ozone concentrations measured downwind of the Chicago-Gary urban
39             complex. J. Air Pollut. Control Assoc. 30: 911-914.
40       Sexton, F. W.; Michie, R. A., Jr.; McElroy, F. F.; Thompson, V. L. (1981) Technical assistance document for the
41             calibration and operation of automated ambient non-methane organic compound analyzers. Research Triangle
42             Park, NC: U.S. Environmental Protection Agency, Environmental Monitoring Systems Laboratory; report no.
43             EPA-600/4-81-015. Available from: NTIS, Springfield, VA; PB82-147406.
44       Sexton, F. W.; Michie, R. M., Jr.; McElroy, F. F.; Thompson, V. L. (1982) A comparative evaluation of seven
45             automated ambient nonmethane organic compound analyzers. Research Triangle Park, NC: U.S.
46             Environmental Protection Agency, Environmental Monitoring Systems Laboratory; report no.
47             EPA-600/4-82-046. Available from: NTIS, Springfield, VA; PB82-230798.
48       Shafran, P. C.; Seaman, N. L.; Gayno, G. A. (2000) Evaluation of numerical  predictions of boundary layer structure
49             during the Lake Michigan ozone study. J. Appl. Meteorol. 39: 412-420.
50       Shapiro, R. (1977) Genetic effects of bisulfite (sulfur dioxide). Mutat. Res. 39: 149-175.
51       Shapiro, M. A. (1980) Turbulent mixing within tropopause folds as a mechanism for the exchange of chemical
52             constituents between the stratosphere and troposphere. J. Atmos. Sci. 37:  994-1004.
53       Sharkey, T. D.; Singsaas, E. L.; Vanderveer, P. J.; Geron, C. D. (1996) Field measurements of isoprene emission
54             from trees in response to temperature and light. Tree Physiol. 16: 649-654.
55       Shepson, P. B.; Edney, E. O.; Corse, E. W. (1984) Ring fragmentation reactions on the photooxidations of toluene
56             and o-xylene. J. Phys. Chem. 88: 4122-4126.
         August 2005                                 AX2-186       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Shepson, P. B.; Mackay, E.; Muthuramu, K. (1996) Henry's Law constants and removal processes for several
  2             atmospheric 6-hydroxy alkyl nitrates. Environ. Sci. Technol. 30: 3618-3623.
  3       Shelter, R. E.; Junkermann, W.; Swartz, W. H.; Frost, G. I; Crawford, J. H.; Lefer, B. L.; Barrick, J. D.; Hall, S. R.;
  4             Hofzumahaus, A.; Bais, A.; Calvert, J. G.; Cantrell, C. A.; Madronich, S.; Mtiller, M.; Kraus, A.; Monks,
  5             P. S.; Edwards, G. D.; McKenzie, R.; Johnston, P.; Schmitt, R.; Griffioen, E.; Krol, M.; Kylling, A.;
  6             Dickerson, R. R.; Lloyd, S. A.; Martin, T.; Gardiner, B.; Mayer, B.; Pfister, G.; Roeth, E. P.; Koepke, P.;
  7             Ruggaber, A.; Schwander, H.; Van Weele, M. (2003) Photolysis frequency of NO2: Measurement and
  8             modeling during the International Photolysis Frequency Measurement and Modeling intercomparison
  9             (IPMMI). J. Geophys. Res. (Atmos.) 108(D16): 10.1029/2002JD002932.
10       Sillman, S. (1995) The use of NOy, H2O2 and HNO3 as indicators for O3-NOX-ROG sensitivity in urban locations.
11             J. Geophys. Res. 100:  14,175-14,188.
12       Sillman, S. (1999) The relation between ozone, NOX and hydrocarbons in urban and polluted rural environments.
13             Atmos. Environ. 33: 1821-1845.
14       Sillman, S. (2000) Ozone production efficiency and loss of NOX in power plant plumes: photochemical model and
15             interpretation of measurements in Tennessee. J. Geophys. Res. [Atmos.] 105: 9189-9202.
16       Sillman, S.; He, D.-Y. (2002) Some theoretical results concerning O3-NOX-VOC chemistry and NOX-VOC
17             indicators. J.  Geophys. Res. (Atmos.) 107: 10.1029/2001JDOO1123.
18       Sillman, S.; Samson, P. J. (1995)  Impact of temperature on oxidant photochemistry in urban, polluted rural and
19             remote environments.  J. Geophys. Res. [Atmos.] 100: 11,497-11,508.
20       Sillman, S.; Logan, J. A.; Wofsy, S. C.  (1990) The sensitivity of ozone to nitrogen oxides and hydrocarbons in
21             regional ozone episodes. J. Geophys. Res. [Atmos.] 95: 1837-1851.
22       Sillman, S.; Al-Wali, K. I.; Marsik, F. J.; Nowacki, P.; Samson, P. J.; Rodgers, M. O.; Garland, L. J.; Martinez, J. E.;
23             Stoneking, C.; Imhoff, R.;  Lee, J.-H.; Newman, L.; Weinstein-Lloyd, J.; Aneja, V. P.  (1995) Photochemistry
24             of ozone formation in  Atlanta, GA-models and measurements. Atmos. Environ. 29: 3055-3066.
25       Sillman, S.; He, D.;  Cardelino, C.; Imhoff, R. E. (1997) The use of photochemical indicators to evaluate
26             ozone-NOx-hydrocarbon sensitivity: case studies from Atlanta, New York, and Los Angeles. J. Air Waste
27             Manage. Assoc. 47: 1030-1040.
28       Sillman, S.; He, D.;  Pippin, M. R.; Daum, P. H.; Imre, D. G.; Kleinman, L. I.; Lee, J. H.; Weinstein-Lloyd, J.  (1998)
29             Model correlations for ozone, reactive nitrogen, and peroxides for Nashville in comparison with
30             measurements: implications for O3-NOx-hydrocarbon chemistry. J. Geophys. Res. [Atmos.]
31             103:22,629-22,644.
32       Sillman, S.; Vautard, R.; Menut, L.; Kley, D. (2003) O3-NOX-VOC sensitivity and NOX-VOC indicators in Paris:
33             results from models and atmospheric pollution over the Paris area (ESQUIF) measurements. J.  Geophys. Res.
34             (Atmos.) 108: 10.1029/2002JD001561.
35       Singh, H. B.; Ludwig, F. L.; Johnson, W. B. (1977) Ozone in clean remote atmospheres: concentrations and
36             variabilities. New York, NY: Coordinating Research Council, Inc.; report no. CRC-APRAC CAPA-15-76.
37             Available from: NTIS, Springfield, VA; PB-272290.
3 8       Singh, H. B.; Viezee, W.; Johnson, W. B.; Ludwig, F. L. (1980) The impact of stratospheric ozone on tropospheric
39             air quality. J. Air Pollut. Control Assoc. 30:  1009-1017.
40       Singh, H. B.; Kanakidou, M.; Crutzen, P. J.; Jacob, D. J. (1995) High concentrations and photochemical fate of
41             oxygenated hydrocarbons in the  global troposphere. Nature (London) 378: 50-54.
42       Singh, H. B.; Gregory, G. L.; Anderson, B.; Browell, E.; Sachse, G. W.; Davis, D. D.; Crawford, J.; Bradshaw, J. D.;
43             Talbot, R.; Blake, D. R.; Thornton, D.; Newell, R.; Merrill, J. (1996a) Low ozone in the marine boundary
44             layer of the tropical Pacific Ocean: photochemical loss, chlorine atoms, and entrainment. J. Geophys. Res.
45             [Atmos.] 101: 1907-1917.
46       Singh, H. B.; Herlth, D.; Kolyer, R.; Salas, L.; Bradshaw, J. D.; Sandholm, S. T.; Davis, D. D.; Crawford, J.;
47             Kondo, Y.; Koike, M.; Talbot, R.; Gregory,  G. L.; Sachse, G. W.; Browell, E.; Blake, D. R.; Rowland, F. S.;
48             Newell, R.; Merrill, J.; Heikes, B.; Liu, S. C.; Crutzen, P. J.; Kanakidou, M. (1996b) Reactive nitrogen and
49             ozone over the western Pacific: distribution, partitioning, and sources. J. Geophys. Res. 101: 1793-1808.
50       Singh, M. P.; McNider, R. T.; Meyers, R.; Shekhar, G. (1997) Nocturnal wind structure and dispersion of pollutants:
51             an analytical  study. Atmos. Environ. 31:  105-115.
52       Singh, H. B.; Salas,  L. J.; Chatfield, R.  B.; Czech, E.; Fried, A.; Walega, J.; Evans, M. J.; Field, B. D.; Jacob, D.  J.;
53             Blake, D.; Heikes, B.; Talbot, R.; Sachse, G.; Crawford, J. H.; Avery, M. A.;  Sandholm, S.; Fuelberg, H.
54             (2004) Analysis of the atmospheric distribution, sources, and sinks of oxygenated volatile organic chemicals
55             based on measurements over the Pacific during TRACE-P. J. Geophys. Res. (Atmos.)
56             109(D15S07): 10.1029/2003JD003883.
         August 2005                                 AX2-187      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Sirju, A. P.; Shepson, P. B. (1995) Laboratory and field investigation of the DNPH cartridge technique for the
  2             measurement of atmospheric carbonyl compounds. Environ. Sci. Technol. 29: 384-392.
  3       Skiba, U.; Fowler, D.; Smith, K.A. (1997) Nitric oxide emissions from agricultural soils in temperate and tropical
  4             climates: sources, controls and mitigation options. Nutrient Cycling Agroecosystems. 48 (1-2): 139-153.
  5       Slemr, F.; Seiler, W. (1984) Field measurements of NO and NO2 emissions from fertilized and unfertilized soils.
  6             J. Atmos. Chem. 2: 1-24.
  7       Slemr, J.; Slemr, F.; Partridge, R.; D'Souza, H.; Schmidbauer, N. (2002) Accurate Measuements of Hydrocarbons in
  8             the Atmosphere (AMOHA): three European intercomparisons. J. Geophys. Res. [Atmos.]
  9             107(019): 10.1020/2001JD001357.
10       Smith, F.; Strong, R. (1977) Independent performance audits on rams station sensors - audit #4. interim report.
11             Research Triangle Park, NC: Research Triangle Institute; EPA contract no. 68-02-2407.
12       Smith, A. M; Rigler, E.; Kwok, E. S. C.; Atkinson, R. (1996) Kinetics and products of the gas-phase reactions of
13             6-methyl-5-hepten-2-one and /rans-cinnamaldehyde with OH and NO3 radicals and O3 at 29672K. Environ.
14             Sci. Technol. 30:  1781-1785.
15       Smith, D. F.; Mclver, C. D.; Kleindienst, T. E. (1998) Primary product distribution from the reaction of hydroxyl
16             radicals with toluene at ppb NOX mixing ratios. J. Atmos. Chem. 30: 209-228.
17       Smith, D. F.; Kleindienst, T. E.; Mclver, C. D. (1999) Primary product distributions from the reaction of OH with
18             m-, />-xylene,  1,2,4- and 1,3,5-trimethylbenzene. J. Atmos. Chem. 34: 339-364.
19       Spicer, C. W.; Chapman, E. G.; Finlayson-Pitts, B. J.; Plastridge, R. A.; Hubbe, J. M.; Fast, J. D.; Berkowitz, C. M.
20             (1998) Unexpectedly high concentrations of molecular chlorine in coastal air (letter). Nature (London)
21             394: 353-356.
22       Spicer, C. W.; Billick, I. H.; Yanagisawa, Y. (2001) Nitrous acid interference with passive NO2 measurement
23             methods and the impact on indoor NO2 data. Indoor Air 11:  156-161.
24       Spivakovsky, C. M.; Logan, J. A.; Montzka, S. A.; Balkanski, Y. J.; Foreman-Fowler, M.; Jones, D. B. A.;
25             Horowitz, L. W.; Fusco, A. C.; Brenninkmeijer, C. A. M.; Prather, M. J.; Wofsy, S. C.; McElroy, M. B.
26             (2000) Three-dimensional climatological distribution of tropospheric OH: update and evaluation. J. Geophys.
27             Res. (Atmos.) 105:8931-8980.
28       Spokes, L. J.; Yeatman, S. G.; Cornell, S. E.; Jickells, T. D. (2000) Nitrogen deposition to the eastern Atlantic
29             Ocean, the importance  of south-easterly flow.  Tellus Ser. B  52B: 37-49.
30       Sprenger, M.; Wernli, H. (2003) A northern hemispheric climatology of cross-tropopause exchange for the ERA 15
31             time period (1979-1993). J. Geophys. Res. 108(D12): 10.1029/2002JD002636.
32       Sprenger, M.; Maspoli, M. C.; Wernli, H. (2003)  Tropopause folds and cross-tropopause exchange: a global
3 3             investigation based upon ECMWF analyses for the time period March 2000 to February 2001. J. Geophys.
34             Res. (Atmos.) 108(D12): 10.1029/2002JD002587.
35       Sprengnether, M.; Demerjian, K. L.; Donahue, N. M.; Anderson, J. G. (2002) Product analysis of the OH oxidation
36             of isoprene and 1,3-butadiene in the presence of NO J. Geophys. Res. (Atmos.) 107(D15):
37             10.1029/2001JD000716.
38       Sprung, D.; Jost, C.; Reiner, T.; Hansel, A.; Wisthaler, A. (2001) Acetone and acetonitrile in the tropical Indian
3 9             ocean boundary layer and free troposphere: aircraft-based intercomparison of AP-CIMS and PTR-MS
40             measurements. J. Geophys. Res. 106: 28,511-28,528.
41       Staffelbach, T.; Neftel, A.; Blatter, A.; Gut, A.; Fahrni, M.; Stahelin, J.; Prevot, A.; Hering, A.; Lehning, M.;
42             Neininger, B.; Baumle, M.; Kok, G. L.; Dommen, J.; Hutterli, M.; Anklin, M. (1997) Photochemical oxidant
43             formation over southern Switzerland 1. results from summer 1994. J. Geophys. Res. (Atmos.)
44             102:23,345-23,362.
45       Staudt, A. C.; Jacob, D. J.; Ravetta, F.; Logan, J. A.;  Bachiochi, D.; Krishnamurti, T. N.; Sandholm, S. T.; Ridley,
46             B. A.; Singh, H. B.; Talbot, R. W. (2003) Sources and chemistry of nitrogen oxides over the tropical Pacific.
47             J. Geophys. Res. (Atmos.) 108(8239): 10.1029/2002JD002139.
48       Stedman, D.  H.; Bishop, G.; Peterson, J. E.; Guenther, P. L. (1991) On-road CO remote sensing in the Los Angeles
49             Basin: final report. Sacramento, CA: California Air Resources Board, ARE Contract No.  A932-189.
50       Stehr, J. W.;  Dickerson, R. R.; Hallock-Waters, K. A.; Doddridge, B. G.; Kirk, D. (2000) Observations of NOy, CO,
51             and SO2 and the origin of reactive nitrogen J. Geophys. Res. 105(D3): 3553-3563.
52       Stein, A. F.; Lamb, D. (2002)  Chemical indicators of sulfate sensitivity to nitrogen oxides an volatile organic
53             compounds. J. Geophys. Res. (Atmos.) 107: 10.1029/2001JDOO1088.
54       Stein, A. F.; Lamb, D. (2003)  Empirical evidence for the low- and high-NOx photochemical regimes of sulfate and
55             nitrate formation. Atmos. Environ. 37: 3615-3625.
         August 2005                                 AX2-188       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Stein, A. F.; Lamb, D.; Draxler, R. R. (2000) Incorporation of detailed chemistry into a three-dimensional
  2             Lagrangian-Eulerian hybrid model: application to regional tropospheric ozone. Atmos. Environ.
  3             34:4361-4372.
  4       Stenchikov, G.; Dickerson, R.; Pickering, K.; Ellis, W. Jr.; Doddridge, B.; Kondragunta, S.; Poulida, O.; Scala, J.;
  5             Tao, W.-K. (1996) Stratosphere-troposphere exchange in a midlatitude mesoscale convective complex. 2.
  6             Numerical simulation. J. Geophys. Res. [Atmos.] 101: 6837-6851.
  7       Stevens, R. K.; Drago, R. J.; Mamane, Y. (1993) A long path differential optical absorption spectrometer and
  8             EPA-approved fixed-point methods intercomparison. Atmos. Environ. Part B 27: 231-236.
  9       Steyn, D. J.; Bottenheim, J. W.; Thomson, R. B. (1997) Overview of tropospheric ozone in the Lower Fraser Valley,
10             and the Pacific '93 field study. Atmos. Environ. 31: 2025-2036.
11       Stith, J.; Dye, J.; Ridley, B.; Laroche, P.; Defer, E.; Baumann, K.; Htibler, G.; Zerr, R.; Venticinque, M. (1999) NO
12             signatures from lightning flashes. J. Geophys. Res.  [Atmos.] 101: 6837-6851.
13       Stockwell, W. R.; Middleton, P.; Chang, J. S.;  Tang, X. (1990) The second generation Regional Acid Deposition
14             Model chemical mechanism for regional air quality modeling. J. Geophys. Res. [Atmos.] 95: 16,343-16,367.
15       Stockwell, W. R.; Kirchner,  F.; Kuhn, M.; Seefeld, S. (1997) A new mechanism for regional atmospheric chemistry
16             modeling. J. Geophys. Res. 102(D22): 25847-25879.
17       Stohl, A. (2001) A 1-yearLagrangian "climatology" of airstreams in the Northern Hemisphere troposphere and
18             lowermost stratosphere. J. Geophys. Res. (Atmos.)  106: 7263-7279.
19       Stohl, A.; Trickl, T. (1999) A textbook example of long-range transport: simultaneous observation of ozone maxima
20             of stratospheric and North American origin in the free troposphere of Europe. J. Geophys. Res.
21             104(D23): 30,445-30,462.
22       Stohl, A.; Spichtinger-Rakowsky,  N.; Bonasoni, P.; Feldmann, H.;  Memmesheimer, M.;  Scheel, H. E.; Trickly, T.;
23             Htibener, S.; Ringer, W.; Mandl, M. (2000) The influence of stratospheric intrusions  on alpine ozone
24             concentrations. Atmos. Environ. 34: 1323-1354.
25       Stohl, A.; Trainer, M.; Ryerson, T. B.; Holloway, J. S.; Parrish, D. D. (2002) Export of NOy from the North
26             American boundary layer during 1996 and 1997 North Atlantic Regional Experiments. J. Geophys. Res.
27             107(D11): 10.1029/2001JD000519.
28       Stohl, A.; Bonasoni, P.; Cristofanelli, P.; Collins, W.; Feichter, J.; Frank, A.; Forster, C.; Gerasopoulos, E.;
29             Gaggeler, H.; James, P.; Kentarchos, T.; Kromp-Kolb, H.; Kriiger, B.; Land, C.; Meloen, J.; Papayannis, A.;
30             Priller, A.; Seibert, P.; Sprenger, M.; Roelofs, G. J.; Scheel, H. E.; Schnabel, C.; Siegmund, P.; Tobler, L.;
31             Trickl, T.; Wernli, H.; Wirth, V.; Zanis, P.; Zerefos, C. (2003) Stratosphere-troposphere exchange: a review
32             and what we have learned from STACCATO. J. Geophys. Res. 108(D12): 10.1029/2002JD002490.
33       Stroud, C.; Madronich, S.; Atlas, E.; Ridley, B.; Flocke, F.; Weinheimer, A.; Talbot, B.;  Fried, A.; Wert, B.;
34             Shelter,  R.; Lefer, B.; Coffey, M.; Heikes, B.; Blake, D. (2003) Photochemistry in the arctic free troposphere:
3 5             NOX budget and the role of odd nitrogen reservoir recycling. Atmos. Environ. 37:  3351-3364.
36       Stull, R. B. (1985) A fair-weather  cumulus cloud classification scheme for mixed layer studies.  J. Clim. Appl.
37             Meteorol. 24: 49-56.
3 8       Stull, R. B. (1988) An introduction to boundary layer meteorology. Dordrecht, The Netherlands: Kluwer Academic
39             Publishers, pp. 9-16; 499-506. (Atmospheric Sciences Library, v. 13).
40       Stull, R.B. (1999) An introduction to boundary layer meteorology. Dordrecht, The Netherlands: Kluwer Academic
41             Publishers; pp. 9-15, 500-505.
42       Stull, R. B. (2000) Meteorology for scientists and engineers: a technical companion book to C. Donald Ahrens'
43             Meteorology Today. 2nd ed. New York,  NY: Brooks Cole.
44       Stages, W. T.; Barrie, L. A. (1988) Chlorine Bromine and Iodine in Arctic aerosols. Atmos. Environ.
45             22: 1179-1194.
46       Stutz, J.; Platt,  U. (1996) Numerical analysis and estimation of the statistical error of differential optical absorption
47             spectroscopy measurements with least-squares methods. Appl. Opt. 35: 6041-6053.
48       Stutz, J.; Hebestreit, K.; Alicke, B.; Platt, U. (1999) Chemistry of halogen ozides in the troposphere: comparison of
49             model calculations with recent field data. J. Atmos. Chem. 34: 65-85.
50       Stutz, J.; Ackermann, R.; Fast, J. D.; Barrie, L. (2002) Atmospheric reactive chlorine and bromine at the Great Salt
51             Lake, Utah. Geophys. Res.  Lett. 29: 10.1029/2002GL014812.
52       Suhre, K.; Gammas, J.-P.; Nedelec, P.; Rosset, R.; Marenco, A.; Smit, H. G. J. (1997) Ozone-rich transients in the
53             upper equatorial Atlantic troposphere (letter). Nature 388: 661-663.
54       Sundet, J. K. (1997) Model studies with a 3-D  global CTM using ECMWF data [thesis].  Oslo, Norway: University
55             of Oslo, Department of Geophysics.
         August 2005                                  AX2-189      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Svensson, R.; Ljungstroem, E.; Lindqvist, O. (1987) Kinetics of the reaction between nitrogen dioxide and water
  2             vapour. Atmos. Environ. 21: 1529-1539.
  3       Talbot, R. W.; Vijgen, A. S.; Harris, R. C. (1990) Measuring tropospheric HNO3: problems and prospects for nylon
  4             filter and mist chamber. J. Geophys. Res. [Atmos.] 95: 7553-7542.
  5       Tan, D.; Faloona, I.; Simpas, J. B.; Brune, W.; Shepson, P. B.; Couch, T. L.; Sumner, A. L.; Carroll, M. A.;
  6             Thornberry, T.; Apel, E.; Riemer, D.; Stockwell, W. (2001) HOX budgets in a deciduous forest: results from
  7             the PROPHET summer 1998 campaign. J. Geophys. Res. (Atmos.) 106: 24,407-24,427.
  8       Tanaka, P. L.; Oldfield, S.; Neece, J. D.; Mullins, C. B.; Allen, D. T. (2000) Anthropogenic sources of chlorine and
  9             ozone formation in urban atmospheres. Environ. Sci. Technol. 34: 4470-4473.
10       Tanaka, P. L.; Riemer, D. D.; Chang, S.; Yarwood, G.; McDonald-Buller, E. C.; Apel, E. C.; Orlando, J. J.; Silva,
11             P. J.; Jimenez, J. L.; Canagaratna, M. R.;  Neece, J. D.; Mullins, C. B.; Allen, D. T. (2003a) Direct evidence
12             for chlorine-enhanced urban ozone formation in Houston, Texas.  Atmos. Environ. 37: 1393-1400.
13       Tanaka, P. L.; McDonald-Buller, E. C.; Chang, S.; Yarwood, G.; Kimura, Y.; Neece, J. D.; Mullins, C. B.; Allen,
14             D. T. (2003b) Development of a chlorine mechanism for use in the Carbon Bond IV chemistry  model.
15             J. Geophys. Res. [Atmos.]  108(D4): art. no. 4145.
16       Tanner, R. L.; Kelly, T. J.; Dezaro, D. A.; Forrest, J. (1989) A comparison of filter, denuder, and real-time
17             chemiluminescence techniques for nitric acid determination in ambient air. Atmos. Environ. 23: 2213-2222.
18       Tao, W.-K.; Simpson, J. (1993) The Goddard Cumulus Ensemble Model. Part I: model description. Terr. Atmos.
19             Oceanic Sci. (TAO) 4:35-71.
20       Taubman, B. F.; Marufu, L. T.; Piety, C. A.; Doddridge, E.G.; Stehr, J.  W.; Dickerson, R. R. (2004) Airborne
21             characterization of the chemical, optical, and meteorological properties, and origins of a combined
22             ozone-haze episode over the eastern United States. J. Atmos. Sci. 61: 1781-1793.
23       Tesche, T. W.; Roth, P. M.; Reynolds, S. D.; Lurmann, F. W. (1993) Scientific assessment of the urban  airshed
24             model (UAM-IV). Washington, DC: American Petroleum Institute; API publication no. 4556.
25       Thielman, A.; Prevot, A. S. H.; Griiebler, F. C.;  Staelhelin, J. (2001) Empirical ozone isopleths as a tool to identify
26             ozone production regimes. Geophys. Res. Lett. 28: 2369-2372.
27       Thielman, A.; Prevot, A. S. H.; Staelhelin, J. (2002) Sensitivity of ozone production derived from field
28             measurements in the Italian Po basin. J. Geophys. Res. (Atmos.) 107(D22): 10.1029/2000JD000119.
29       Thompson, A. M. (1992) The oxidizing capacity of the Earth's atmosphere:  probable past and future changes.
30             Science (Washington, DC) 256:  1157-1165.
31       Thompson, A. M.; Stewart, R. W. (1991) Effect of chemical kinetics uncertainties on calculated constituents in a
32             tropospheric photochemical model. J. Geophys. Res. 96(D7): 13,089-13,108.
33       Thompson, A. M.; Pickering, K. E.; Dickerson, R. R.; Ellis, W. G., Jr.; Jacob, D. J.; Scala, J. R.; Tao,  W.-K.;
34             McNamara, D. P.; Simpson, J. (1994) Convective-transport over the  central United States and its role in
35             regional CO and O3 budgets. J. Geophys. Res. [Atmos.] 99: 18,703-18,711.
36       Thompson, A. M.; Singh, H. B.; Schlager, H.  (2000) Subsonic assessment ozone and nitrogen oxide experiment
3 7             (SONEX) and pollution from aircraft emissions in the North Atlantic Flight Corridor (POLINAT 2).
38             J. Geophys. Res. 105: 3595-3603.
39       Thornton, J. A.; Wooldridge, P. J.; Cohen, R.  C. (2000) Atmospheric NO2: in situ laser-induced fluorescence
40             detection at parts per trillion mixing ratios. Anal. Chem. 72: 528-539.
41       Thornton, J. A.; Braban, C. F.; Abbatt, J. P. D. (2003) N2O5 hydrolysis on sub-micron organic aerosol: the effect of
42             relative humidity, particle phase and particle size. Phys. Chem. Chem. Phys. 5: 4593-4603.
43       Tie, X. X.; Emmons, L.; Horowitz, L.; Brasseur, G.; Ridley, ,B.; Atlas, E.; Stround, C.; Hess, P.; Klonecki, A.;
44             Madronich, S.; Talbot, R.; Dibb, J. (2003) Effect of sulfate aerosol on tropospheric NOX and ozone budgets:
45             model simulations and TOPSE evidence.  J. Geophys. Res. (Atmos.)  108(D4): 10.1029/2001JD001508.
46       Tonnesen, G. S.; Dennis, R. L. (2000a) Analysis of radical propagation efficiency to assess ozone sensitivity to
47             hydrocarbons and NOX 1. Local indicators of odd oxygen production sensitivity. J.  Geophys. Res. (Atmos.)
48             105: 9213-9225.
49       Tonnesen, G. S.; Dennis, R. L. (2000b) Analysis of radical propagation efficiency to assess ozone sensitivity to
50             hydrocarbons and NOX 2. Long-lived species as indicators of ozone concentration sensitivity. J. Geophys.
51             Res. 105: 9227-9241.
52       Trainer, M.; Buhr, M. P.; Curran, C. M.; Fehsenfeld, F. C.; Hsie,  E. Y.; Liu, S. C.; Norton, R. B.; Parrish, D. D.;
53             Williams, E. J.; Gandrud, B. W.; Ridley, B. A.; Shelter, J.  D.; Allwine, E. J.; Westberg, H. H. (1991)
54             Observations and modeling of the reactive nitrogen photochemistry at a rural site. J. Geophys. Res. [Atmos.]
55             96: 3045-3063.
         August 2005                                 AX2-190       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Trainer, M; Parrish, D. D.; Buhr, M. P.; Norton, R. B.; Fehsenfeld, F. C.; Anlauf, K. G.; Bottenheim, J. W.; Tang,
  2             Y. Z.; Wiebe, H. A.; Roberts, J. M.; Tanner, R. L.; Newman, L.; Bowersox, V. C.; Meagher, J. F.; Olszyna,
  3             K. J.; Rodgers, M. O.; Wang, T.; Berresheim, H.; Demerjian, K. L.; Roychowdhury, U. K. (1993) Correlation
  4             of ozone with NOy in photochemically aged air. J. Geophys. Res. [Atmos.] 98: 2917-2925.
  5       Trainer, M.; Ridley, B. A.; Buhr, M. P.; Kok, G.; Walega, J.; Htibler, G.; Parrish, D. D.; Fehsenfeld, F. C. (1995)
  6             Regional ozone and urban plumes in the southeastern United States: Birmingham, a case study. J. Geophys.
  7             Res. [Atmos.] 100: 18,823-18,834.
  8       Trainer, M.; Parrish, D. D.; Golday, P. D.; Roberts, J.; Fehsenfeld, F. C. (2000) Review of observation-based
  9             analysis of the regional factors influencing ozone concentrations. Atmos. Environ.  34: 2045-2061.
10       Traub, M.; Fischer, H.; De Reus, M.; Kormann, R.; Heland, J.; Ziereis, H.; Schlager, H.; Holzinger, R.; Williams, J.;
11             Warneke, C.; De Gouw, J.; Lelieveld, J. (2003) Chemical characteristics assigned to trajectory clusters during
12             the MINOS campaign. Atmos. Chem. Phys. 3: 459-468.
13       Tseng, R.-S.; Viechnicki, J. T.; Skop, R. A.; Brown, J. W. (1992) Sea-to-air transfer of surf ace-active organic
14             compounds by bursting bubbles. J. Geophys. Res. (Atmos.) 97: 5201-5206.
15       Tuazon, E. C.; MacLeod, H.; Atkinson, R.; Carter, W. P. L. (1986) a-Dicarbonyl yields from the NOx-air
16             photooxidations of a series of aromatic hydrocarbons in air. Environ. Sci. Technol. 20: 383-387.
17       Tully, F. P.; Ravishankara, A. R.; Thompson, R. L.; Nicovich, J. M.; Shah, R. C.; Kreutter, N. M.; Wine, P. H.
18             (1981) Kinetics of the reactions of hydroxyl radical with benzene and toluene. J. Phys. Chem.  85: 2262-2269.
19       Turekian, V. C.; Macko, S. A.; Keene, W. C. (2003) Concentrations, isotopic compositions, and sources of
20             size-resolved, paniculate organic carbon and oxalate in near-surface marine air at Bermuda during spring.
21             J. Geophys. Res. (Atmos.) 108(D5): 10.1029/2002JD002053.
22       U.S. Congress. (1990) Clean Air Act amendments of 1990, PL 101-549, November 15. Washington, DC: U.S.
23             Government Printing Office.
24       U.S. Environmental Protection Agency. (1991) Guideline for regulatory application of the urban airshed model.
25             Research Triangle Park, NC: Office of Air Quality Planning and Standards; report  no. EPA-450/4-91-013.
26             Available from: NTIS, Springfield, VA; PB92-108760/HSU.
27       U.S. Environmental Protection Agency. (1993a) Air quality criteria for oxides of nitrogen. Research Triangle Park,
28             NC: Office of Health and Environmental Assessment, Environmental Criteria and Assessment Office; report
29             nos. EPA/600/8-9!/049aF-cF. 3v. Available from: NTIS, Springfield, VA; PB95-124533, PB95-124525, and
30             PB95-124517.
31       U.S. Environmental Protection Agency. (1993) Regional interim emission inventories (1987-1991), v. II: emission
32             summaries. Research Triangle Park, NC: Office of Air Quality Planning and Standards; report no.
33             EPA-454/R-93-021b. Available from: NTIS, Springfield, VA; PB93-236115.
34       U.S. Environmental Protection Agency. (1996) Air quality criteria for ozone and related photochemical oxidants,
35             v. I-III. Research Triangle Park, NC: Office of Research and Development; report no.
36             EPA/600/P-93/004aF-cF. Available from: NTIS, Springfield, VA; PB96-185582. PB96-185590, and
37             PB96-185608. Available: www.epa.gov/ncea/ozone.htm [25 February 2002].
38       U.S. Environmental Protection Agency. (1997) National air pollutant emission trends 1990-1996. Research Triangle
39             Park, NC: Office of Air Quality Planning and Standards; report no. EPA/454/R-97/011. Available from:
40             NTIS, Springfield, VA.
41       U.S. Environmental Protection Agency. (1999) Control measure evaluations: the Control Measure Data Base for the
42             National Emissions Trends Inventory (ControlNET) [draft report]. Research Triangle Park, NC: Office of Air
43             Quality Planning and Standards; report no. EPA/452/D-01/001. Available from: NTIS, Springfield, VA;
44             PB2001-105812.
45       U.S. Environmental Protection Agency. (2000) Air quality criteria for carbon monoxide. Research Triangle Park,
46             NC: National Center for Environmental Assessment; report no. EPA/600/P-99/00 IF. Available:
47             http://www.epa.gov/ncea/pdfs/coaqcd.pdf (7 May 2003).
48       U.S. Environmental Protection Agency. (2001) National air quality and emissions trends report,  1999. Research
49             Triangle Park, NC: Office of Air Quality Planning and Standards; report no. EPA 454/R-01-004. Available:
50             http://www.epa.gov/oar/aqtrnd99/ [7 May 2003].
51       U.S. Environmental Protection Agency. (2003) Air quality criteria for ozone and related photochemical oxidants.
52             Research Triangle Park, NC: Office of Research and Development: manuscript in preparation.
53       U.S. Environmental Protection Agency. (2004) Air quality criteria for paniculate matter. Research Triangle Park,
54             NC: National Center for Environmental Assessment; report no. EPA/600/P-99/002aF-bF. 2v. Available:
55             http://cfpub.epa.gov/ncea/ [9 November,  2004].
         August 2005                                 AX2-191       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Underwood, G. M; Song, C. H.; Phadnis, M.; Carmichael, G. R.; Grassian, V. H. (2001) Heterogeneous reactions of
  2             NO2 and HNO3 on oxides and mineral dust: a combined laboratory and modeling study. J. Geophys. Res.
  3             [Atmos.]  106: 18,055-18,066.
  4       Usher, C. R.; Michel, A. E.; Stec, D.; Grassian, V. H. (2003) Laboratory studies of ozone uptake on processed
  5             mineral dust. Atmos. Environ. 37: 5337-5347.
  6       Valente, R. J.; Imhoff, R. E.; Tanner, R. L.; Meagher, J. F.; Damn, P. H.; Hardesty, R. M.; Banta, R. M.; Alvarez,
  7             R. J.; McNider, R. T.; Gillani, N. V.  (1998) Ozone production during an urban air stagnation episode over
  8             Nashville, Tennessee. J. Geophys. Res. [Atmos.] 103: 22,555-22,568.
  9       Van Aardenne, J. A.; Dentener, F. J.; Olivier, J. G. J.; Goldewijk, C. G. M. K.; Lelieveld, J. (2001) A 1 ° x 1°
10             resolution data set of historical anthropogenic trace gas emissions for the period 1980-1990. Glob.
11             Biogeochem. Cycles 15: 909-928.
12       Vandaele, A. C.; Tsouli, A.; Carleer, M.; Colin, R. (2002) UV Fourier transform measurements of tropospheric O3,
13             NO2, SO2, benzene, and toluene. Environ. Pollut. 116: 193-201.
14       Vaughan, W. M.; Chan, M.; Cantrell, B.; Pooler, F. (1982) A study of persistent elevated pollution episodes in the
15             northeastern United States. Bull. Am. Meteorol. Soc. 63: 258-266.
16       Vautard, R.; Martin, D.; Beekman, M.; Drobinski, P.; Friedrich, R.; Jaubertie, A.; Kley, D.; Lattuati, M.; Moral, P.;
17             Neininger, B.; Theloke, J. (2002) Paris emission inventory diagnostics from the ESQUIF airborne
18             measurements and a chemistry transport model. J. Geophys. Res. (Atmos.) 108(D17):
19             10.1029/2002JD002797.
20       Venkatachari, P.; Hopke, P. K.; Grover, B. D.; Eatough, D. J. (2005a) Measurement of particle-bound reactive
21             oxygen species inRubidoux aerosols. J. Atmos. Chem. 50: 49-58.
22       Venkatachari, P.; Hopke, P .K.; Brune, W. H.; Ren, X.; Lesher, R.; Mao, J.; Mitchell, M. (2005b) Characterization
23             of reactive oxygen species trends in Flushing, New York. Atmos. Environ.: in press.
24       Viezee, W.; Johnson, W. B.; Singh, H. B. (1979) Airborne measurements of stratospheric ozone intrusions into the
25             troposphere over the United States [final report]. Menlo Park, CA: SRI International; SRI project 6690.
26       Viezee, W.; Singh, H. B.  (1982) Contribution of stratospheric ozone to ground-level ozone concentrations -
27             a scientific review of existing evidence. Research Triangle Park, NC:  U.S. Environmental Protection Agency,
28             Environmental Science Research Laboratory; grant CR-809330010.
29       Vogt, R.; Crutzen, P. J.; Sander, R. (1996) A mechanism for halogen release from sea-salt aerosol in the remote
30             marine boundary layer. Nature 383: 327-330.
31       Vogt, R.; Sander, R.; Von Glasow, R.; Crutzen, P. J. (1999) Iodine chemistry and its role in halogen activation and
32             ozone loss in the marine boundary layer: a model study. J. Atmos. Chem. 32: 375-395.
33       Volz-Thomas, A.; Geiss, H.; Hofzumahaus, A.; Becker, K.-H. (2003) Introduction to special section: photochemistry
34             experiment in BERLIOZ. J. Geophys. Res. [Atmos.] 108(D4): 10.1029/JD002029.
35       Von Glasow, R.; Sander, R.; Bott, A.; Crutzen, P. J. (2002a) Modeling halogen chemistry in the marine boundary
36             layer. 1. Cloud-free MBL. J. Geophys. Res. 107(D17): 10.1029/2001JD000942.
37       Von Glasow, Sander, R.; Bott, A.; Crutzen, P. J. (2002b) Modeling halogen chemistry in the marine boundary layer.
38             2. Interactions with sulfur and cloud-covered MBL. J. Geophys. Res.  107(D17): 10.1029/2001JD000943.
39       Vuilleumier, L.; Earner, J. T.; Harley, R. A.; Brown, N. J.  (2001) Evaluation of nitrogen dioxide photolysis rates in
40             an urban area using data from the 1997 Southern California Ozone Study.  Atmos. Environ. 35: 6525-6537.
41       Vukovich, F. M. (1995) Regional-scale boundary layer ozone variations in the eastern United States and their
42             association with meteorological variations. Atmos. Environ. 29: 2259-2273.
43       Vukovich, F. M. (1997) Time scales of surface ozone variations in the regional, non-urban environment. Atmos.
44             Environ. 31:1513-1530.
45       Vukovich, R. M.; Ching, J. K.  S. (1990) A semi-impirical  approach to estimate vertical transport by nonprecipitating
46             convective clouds on a regional scale. Atmos. Environ. 24A: 2153-2168.
47       Vukovich, F. M.; Bach, W. D., Jr.; Crissman, B. W.; King, W. J. (1977) On the relationship between high ozone in
48             the rural surface layer and high pressure systems. Atmos. Environ. 11: 967-983.
49       Wagner, T.; Platt, U. (1998) Satellite mapping of enhanced BrO concentrations in the troposphere. Nature (London)
50             395:486-490.
51       Wagner, V.; Jenkin, M. E.; Saunders, S. M.; Stanton, K.; Wirtz, M.; Pilling,  J.  (2003) Modelling of the
52             photooxidation of toluene: conceptual ideas for validating detailed mechanisms. Atmos. Chem. Phys.
53             3:89-106.
54       Wahner, A.; Mentel, T. F.; Sohn, M.; Stier, J. (1998) Heterogeneous reaction of N2O5 on sodium nitrate aerosol.
55             J. Geophys. Res. (Atmos.) 103: 31,103-31,112.
         August 2005                                  AX2-192       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Wainman, T.; Zhang, J.; Weschler, C. J.; Lioy, P. J. (2000) Ozone and limonene in indoor air: a source of submicron
  2             particle exposure. Environ. Health Perspect. 108: 1139-1145.
  3       Walcek, C. J.; Taylor, G. R. (1986) A theoretical method for computing vertical distributions of acidity and sulfate
  4             production within cumulus clouds. J. Atmos. Sci. 43: 339-355.
  5       Walcek, C. J.; Stockwell, W. R.; Chang, J. S. (1990) Theoretical estimates of the dynamic, radiative and chemical
  6             effects of clouds on tropospheric trace gases. Atmos. Res. 25: 53-69.
  7       Walega, J. G.; Stedman, D. H.; Shelter, R. E.; Mackay, G.  I.; Iguchi, T.; Schiff, H. I. (1984) Comparison of a
  8             chemiluminescent and a tunable diode laser absorption technique for the measurement of nitrogen oxide,
  9             nitrogen dioxide, and nitric acid. Environ. Sci. Technol.  18: 823-826.
10       Wang, Y.; Tao, W.-K.; Pickering, K. E.; Thompson, A. M; Kain, J. S.; Adler, R. F.; Simpson, J.; Keehn, P. R.;
11             Lai, G. S. (1996) Mesoscale model simulations of TRACE A and preliminary regional experiment for
12             storm-scale operational and research meteorology convective systems and associated tracer transport.
13             J. Geophys. Res. [Atmos.] 101: 24,013-24,027.
14       Wang, Y.; DeSilva, A. W.; Goldenbaum, G. C.; Dickerson, R. R. (1998a) Nitric oxide production by simulated
15             lightning: dependence on current, energy, and pressure. J. Geophys. Res. [Atmos.] 103: 19,149-19,159.
16       Wang, Y.; Logan, J. A.; Jacob, D. J. (1998b) Global simulation of tropospheric O3-NOx-hydrocarbon chemistry, 2.
17             Model evaluation and global ozone budget. J. Geophys. Res. (Atmos.) 103:  10,727-10,755.
18       Wang, L. H.; Milford, J. B.; Carter, W. P. L. (2000a) Reactivity estimates for aromatic compounds. Part I:
19             uncertainty in chamber-derived parameters. Atmos.  Environ. 34: 4337-4348.
20       Wang, L. H.; Milford, J. B.; Carter, W. P. L. (2000b) Reactivity estimates for aromatic compounds. Part 2.
21             uncertainty in incremental reactivities. Atmos. Environ. 4349-4360.
22       Warneke, C.; Van Der Veen, C.; Luxembourg, S.; De Gouw, J. A.; Kok, A. (2001) Measurements of benzene and
23             toluene in ambient air using proton-transfer-reaction mass spectrometry: calibration, humidity dependence,
24             and field intercomparison. Int. J. Mass Spectrom. 207: 167-182.
25       Watson, J. G.; Chow, J. C.; Fujita, E. M. (2001) Review of volatile organic compound source apportionment by
26             chemical mass balance. Atmos. Environ. 35: 1567-1584.
27       Wayne, R. P.  (2000) Chemistry of Atmospheres: an introduction to the chemistry of the atmospheres of Earth, the
28             planets, and their satellites. 3rd ed. New York, NY:  Oxford University Press, Inc.
29       Wayne, R. P.; Barnes, I.; Biggs, P.; Burrows, J. P.; Canosa-Mas, C. E.; Hjorth, J.; Le Bras, G.; Moortgat, G. K.;
30             Perner, D.; Poulet, G.; Restelli, G.; Sidebottom, H. (1991) The nitrate radical: physics, chemistry, and the
31             atmosphere. Atmos. Environ. Part A 25: 1-203.
32       Wennberg, P. O.; Cohen, R. C.; Stimpfle, R. M.; Koplow,  J. P.; Anderson, J. G.; Salawitch, R. J.; Fahey, D. W.;
33             Woodbridge, E. L.; Keim, E. R.; Geo, R. S.; Webster, C. R.; May, R. D.; Toohey, D. W.; Avallone, L. M.;
34             Proffitt, M. H.; Loewenstein, M.; Podolske, J. R.; Chan, K.  R.; Wofsy, S. C. (1994) Removal of stratospheric
35             O3 by radicals - in situ measurements of OH, HO2, NO, NO2, CLO, and BRO. Science 266: 398-404.
36       Wennberg, P. O.; Hanisco, T. F.; Jaegle, L.; Jacob, D. J.; Hintsa, E. J.; Lanzendorf, E. J.; Anderson, J. G.; Gao,
37             R.-S.; Keim, E. R.; Donnelly, S. G.; Del Negro, L. A.; Fahey, D. W.; McKeen, S. A.; Salawitch, R. J.;
38             Webster, C. R.; May, R. D.; Herman, R. L.; Proffitt, M. H.; Margitan, J. J.; Atlas, E. L.; Schauffler, S. M.;
39             Flocke, F.; McElroy, C.  T.; Bui, T. P. (1998) Hydrogen radicals, nitrogen radicals, and the production of O3 in
40             the upper troposphere. Science (Washington, DC) 279: 49-53.
41       Wernli, H.; Bourqui, M. (2002) A Lagrangian " 1-year climatology" of (deep) cross-tropopause exchange in the
42             extratropical Northern Hemisphere. J. Geophys. Res. (Atmos.) 107(D1-D2): 10.1029/2001JD000812.
43       Wesely, M. L.; Hicks, B. B. (2000) A review of the current status of knowledge on dry deposition. Atmos. Environ.
44             34:2261-2282.
45       Westberg, H.; Zimmerman, P. (1993) Analytical methods used to identify nonmethane organic compounds in
46             ambient atmospheres. In: Newman,  L., ed. Measurement challenges in atmospheric chemistry. Washington,
47             DC: American Chemical Society, pp. 275-291.
48       Wexler, A. S.; Sarangapani, R. (1998) Particles do not increase vapor deposition in human airways. J. Aerosol  Sci.
49             29: 197-204.
50       Whitby, K. T. (1978) The physical characteristics of sulfur aerosols. Atmos. Environ. 12: 135-159.
51       Wild, O.; Prather, M. J. (2000) Excitation of the primary tropospheric chemical mode in a global CTM. J. Geophys.
52             Res. (Atmos.) 105: 24,647-24,660.
53       Williams, E. J.; Guenther, A.; Fehsenfeld, F. C. (1992) An inventory of nitric oxide emissions from soils in the
54             United States. J. Geophys. Res.  [Atmos.] 97: 7511-7519.
         August 2005                                 AX2-193       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Williams, I; Roberts, J. M; Fehsenfeld, F. C; Bertman, S. B.; Buhr, M. P.; Goldan, P. D.; Hubler, G.; Kuster, W.
  2             C.; Ryerson, T. B.; Trainer, M.; Young, V. (1997) Regional ozone frombiogenic hydrocarbons deduced from
  3             airborne measurements of PAN, PPN, and MPAN. Geophys. Res. Lett. 24: 1099-1102.
  4       Williams, E. J.; Baumann, K.; Roberts, J. M.; Bertman, S. B.; Norton, R. B.; Fehsenfeld, F. C.; Springston, S. R.;
  5             Nunnermacker, L. J.; Newman, L.; Olszyna, K.; Meagher, J.; Hartsell, B.; Edgerton, E.; Pearson, J. R.;
  6             Rodgers, M. O. (1998) Intercomparison of ground-based NOy measurement techniques. J. Geophys. Res.
  7             103:22,261-22,280.
  8       Wimmers, A. J. (2003) (dissertation). Charlottesville, VA: University  of Virginia, R. J. Reynolds Institute for
  9             Environmental Science.
10       Wimmers, A. J.; Moody, J. L. (2001) A fixed-layer estimation of upper-tropospheric specific humidity from the
11             GOES water vapor channel: parameterization and validation of the altered brightness temperature product.
12             J. Geophys. Res. [Atmos.] 106: 17,115-17,132.
13       Wimmers, A. J.; Moody, J. L. (2004a) Tropopause folding at satellite-observed spatial gradients: 1. Verification of
14             an empirical relationship. J. Geophys. Res. [Atmos.] 109(D19306): 10.1029/2003JD004145.
15       Wimmers, A. J.; Moody, J. L. (2004b) Tropopause folding at satellite-observed spatial gradients: 2. Development of
16             an empirical model. J. Geophys. Res. [Atmos.] 109(D19307): 10.1029/2003JD004146.
17       Wimmers, A. J.; Moody, J. L.; Browell, E. V; Hair,  J. W.; Grant, W. B.; Butler, C.  F.; Fenn, M. A.; Schmidt, C. C.;
18             Li, J.; Ridley, B. A. (2003)  Signatures of tropopause folding in satellite imagery. J. Geophys. Res. (Atmos.)
19             108(D4): 10.1029/2001JD001358.
20       Winer, A. M.; Peters, J. W.; Smith, J. P.;  Pitts, J. N, Jr. (1974) Response of commercial chemiluminescent NO-NO2
21             analyzers to other nitrogen-containing compounds. Environ. Sci. Technol. 8:  1118-1121.
22       Winer, A. M.; Arey, J.; Atkinson, R.; Aschmann, S. M.; Long, W. D.; Morrison, C. L.; Olszyk, D. M. (1992)
23             Emission rates of organics from vegetation in California's Central Valley. Atmos. Environ. Part A
24             26:2647-2659.
25       Wingenter, O. W.; Kubo, M. K.; Blake, N. J.; Smith, T. W., Jr.; Blake, D. R.; Rowland, F. S. (1996) Hydrocarbon
26             and halocarbon measurements as photochemical and dynamical indicators of atmospheric hydroxyl, atomic
27             chlorine, and vertical mixing obtained during Langrangian flights.  J. Geophys. Res. (Atmos.)
28             101(D1): 4331-4340.
29       Wingenter, O. W.; Blake, D. R.; Blake, N. J.; Sive, B. C.; Rowland, F. S.; Atlas, E.; Flocke, F. (1999) Tropospheric
3 0             hydroxyl and atomic chlorine concentrations,  and mixing timescales determined from hydrocarbon and
31             halocarbon measurements made over the Southern Ocean. J. Geophys. Res. (Atmos.)
32             104(017): 21819-21828.
33       Winner, D. A.; Cass, G. R. (2001) Modeling the long-term frequency distribution of regional ozone concentrations
34             using synthetic meteorology. Environ. Sci. Technol. 35: 3718-3726.
35       Witter, M.; Berndt, T.; Boge, O.; Stratmann, F.; Heintzenberg, J. (2002) Gas-phase ozonolysis: rate coefficients for a
36             series of terpenes and rate coefficients and OH yields for 2-methyl-2-butene and 2,3-dimethyl-2-butene. Int.
37             J. Chem. Kinet. 34: 394-403.
3 8       Wolff, G. T. (1993) On a NOx-focused control strategy to reduce O3. J. Air Waste Manage.  Assoc. 43:  1593-1596.
39       Wolff, G. T.; Korsog, P. E. (1992) Ozone control strategies based on the ratio of volatile organic compounds to
40             nitrogen oxides. J. Air Waste Manage. Assoc. 42: 1173-1177.
41       Wolff, G. T.; Lioy, P. J. (1978) An empirical model for forecasting maximum daily  ozone levels in the northeastern
42             U.S. J. Air Pollut. Control Assoc. 28: 1034-1038.
43       Wolff, G. T.; Lioy, P. J. (1980) Development of an ozone river associated with synoptic scale episodes in the eastern
44             United States. Environ. Sci. Technol. 14: 1257-1260.
45       Wolff, G. T.; Lioy, P. J.; Meyers, R. E.; Cederwall, R. T.; Wight, G. D.; Pasceri, R. E.; Taylor, R.  S. (1977)
46             Anatomy of two ozone transport episodes in the Washington, D.C., to Boston, Mass., corridor. Environ. Sci.
47             Technol. 11:506-510.
48       Wolff, G. T.; Lioy, P. J.; Wight, G. D.; Meyers, R. E.; Cederwall, R. T. (1977) An investigation of long-range
49             transport of ozone across the midwestern and  eastern United States. Atmos. Environ. 11: 797-802.
50       Wolff, G. T.; Lioy, P. J.; Wight, G. D.; Pasceri, R.  E. (1977) Aerial investigation of the ozone plume phenomenon.
51             J. Air Pollut. Control Assoc. 27: 460-463.
52       Wolff, G. T.; Lioy, P. J.; Wight, G. D.; Pasceri, R.  E. (1977) Aerial investigation of photochemical oxidants over the
53             northeast. In: Bufalini, J. J.; Lonneman, W. A., eds. Proceedings of symposium on 1975 northeast oxidant
54             transport study; January 1976; Research Triangle Park, NC. Research Triangle Park, NC: U.S. Environmental
55             Protection Agency, Environmental Sciences Research Laboratory; pp. 70-86; EPA report no.
56             EPA-600/3-77-017. Available from: NTIS, Springfield, VA; PB-265370.
         August 2005                                 AX2-194       DRAFT-DO NOT QUOTE OR CITE

-------
  1       World Meteorological Organization. (1990) Scientific assessment of stratospheric ozone: 1989. Volume II.
  2             Appendix: AFEAS report. Geneva, Switzerland: Global Ozone Research Monitoring Project; report no. 20.
  3       World Meteorological Organization. (1992) Scientific assessment of ozone depletion: 1991 [preprint]. Geneva,
  4             Switzerland: World Meteorological Organization; report no. 25.
  5       Wright, E. S.; Vang, M. I; Finkelstein, J. N.; Mavis, R.  D. (1982) Changes inphospholipid biosynthetic enzymes in
  6             type II cells and alveolar macrophages isolated from rat lungs after NO2 exposure. Toxicol. Appl. Pharmacol.
  7             66:305-311.
  8       Wunderli, S.; Gehrig, R. (1991) Influence of temperature on formation and stability of surface PAN and ozone.
  9             A two year field study in Switzerland. Atmos. Environ. Part A 25: 1599-1608.
10       Yienger, J. J.; Levy, H., II. (1995) Empirical model of global soil-biogenic NOX emissions. J. Geophys. Res.
11             100(D6): 11,447-11,464.
12       Young, V. L.; Kieser, B. N.; Chen, S. P.; Nii, H. (1997)  Seasonal trends and local influences on nonmethane
13             hydrocarbon concentrations in the Canadian boreal forest. J. Geophys. Res. (Atmos.) 102:  5913-5918.
14       Yu, J.; Jeffries, H. E. (1997) Atmospheric photooxidation of alkylbenzenes II. evidence of formation of epoxide
15             intermediates. Atmos. Environ. 31: 2281-2287.
16       Yu, J.; Jeffries, H. E.; Sexton, K. G. (1997) Atmospheric photooxidation of alkylbenzenes I. carbonyl product
17             analyses. Atmos. Environ. 31: 2261-2280.
18       Yung, Y. L.; Pinto, J. P.; Watson, R. T.; Sander, S. P. (1980) Atmospheric bromine and ozone perturbations in the
19             lower stratosphere. J. Atmos. Sci. 37: 339-353.
20       Zafiriou, O. C.; True, M. B. (1979) Nitrate photolysis in seawater by sunlight. Mar. Chem. 8: 33-42.
21       Zanis, P.; Trickl, T.; Stohl, A.; Wernli, H.; Cooper, O.; Zerefos, C.; Gaeggeler, H.; Schnabel, C.;  Tobler, L.; Kubik,
22             P. W.; Priller, A.; Scheel, H. E.; Kanter, H. J.; Cristofanelli, P.; Forster,  C.; James, P.; Gerasopoulos, E.;
23             Delcloo, A.;  Papayannis, A.; Claude, H. (2003) Forecast, observation and modelling of a deep stratospheric
24             intrusion event over Europe. Atmos. Chem. Phys. 3: 763-777'.
25       Zenker, T.; Fischer, H.; Nikitas, C.; Parchatka, U.; Harris, G. W.; Mihelcic, D.; Mtisgen, P.; Patz, H. W.; Schultz,
26             M.; Volz-Thomas, A.; Schmitt, R.; Behmann, T.; Wiessenmayer, M.; Burrows, J. P. (1998) Intercomparison
27             of NO, NO2, NOy, O3, and ROX measurements during the oxidizing capacity of the tropospheric atmosphere
28             (OCTA) campaign 1993 at Izana. J. Geophys. Res. [Atmos.] 103: 13,615-13,634.
29       Zhang, Y.; Carmichael, G. R.  (1999) The role of mineral aerosol in tropospheric chemistry in East Asia~a model
30             study. J. Appl. Meteorol. 38: 353-366.
31       Zhang, J. S.; Rao, S. T. (1999) The role of vertical mixing in the temporal evolution of ground-level ozone
32             concentrations. J. Appl. Meteorol. 38: 1674-1691.
33       Zhang, D.-L.; Zheng, W.-Z. (2004) Diurnal cycles of surface winds and temperatures as simulated by five  boundary
34             layer parameterizations. J. Appl. Meteorol. 43: 157-169.
35       Zhang, S.-H.; Shaw,M.; Seinfeld, J. H.; Flagan, R. C. (1992) Photochemical aerosol formation from a-pinene and
36             p-pinene. J. Geophys. Res. [Atmos.] 97: 20,717-20,729.
37       Zhang, K.; Juiting, M.; Civerolo, K. C.; Berman, S.; Ku, J.; Rao, S. T.; Doddridge, B.; Philbrick,  R. C.; Clark, R.
3 8             (2001) Numerical investigation of boundary layer evolution and nocturnal low-level jets: local versus
39             non-local PEL schemes. Environ. Fluid Mech. 1:  171-208.
40       Zhang, D.; Lei, W. F.; Zhang, R. Y. (2002) Mechanism of OH formation from ozonolysis of isoprene: kinetics and
41             product yields. Chem. Phys. Lett. 358:  171-179.
42       Zhang, D.; Zang, J.; Shi, G.; Iwasaka, Y.; Matsuki, A.; Trochkine, D. (2003) Mixture state of individual Asian dust
43             particles at a coastal site of Qingdao, China. Atmos. Environ. 37: 3895-3901.
44       Zhou, X.; Mopper, K. (1990) Determination of photochemically produced hydroxyl radicals in seawater and
45             freshwater. Mar. Chem. 30: 71-88.
46       Zimmermann, J.; Poppe, D. (1993) Nonlinear chemical couplings in the tropospheric NOX-HOX gas phase  chemistry.
47             J. Atmos. Chem. 17: 141-155.
48       Zondlo, M. A.; Barone, S. B.; Tolbert, M. A. (1998) Condensed-phase products in heterogeneous reactions: N2O5,
49             C1ONO2, and HNO3 reacting on ice films at 185 K. J. Phys. Chem. A 102: 5735-5748.
50
51
         August 2005                                 AX2-195       DRAFT-DO NOT QUOTE OR CITE

-------
 i      ANNEX AX3.  ENVIRONMENTAL CONCENTRATIONS,
 2               PATTERNS, AND EXPOSURE ESTIMATES
 3
 4
 5     AX3.1  INTRODUCTION
 6     Identification and Use of Existing Air Quality Data
 1          Topics discussed in this annex include the characterization of ambient air quality data for
 8     ozone (O3), the uses of these data in assessing the exposure of vegetation to O3, concentrations
 9     of O3 in microenvironments, and a discussion of the currently available human exposure data and
10     exposure model development.  The information contained in this chapter pertaining to ambient
11     concentrations is taken primarily from the U.S. Environmental Protection Agency (EPA) Air
12     Quality System (AQS; formerly the AIRS database). The AQS contains readily accessible
13     detailed, hourly data that has been subject to EPA quality control and assurance procedures.
14     Data available in AQS were collected from 1979 to 2001. As discussed in previous versions of
15     the O3 Air Quality Criteria Document or AQCD (U.S. Environmental Protection Agency, 1986,
16     1996), the data available prior to 1979 may be unreliable due to calibration problems and
17     measurement uncertainties.
18          As indicated in the 1996 O3 AQCD (U.S. Environmental Protection Agency, 1996a), O3  is
19     the only photochemical oxidant other than nitrogen dioxide (NO2) that is routinely monitored
20     and for which a comprehensive database exists. Data for peroxyacetyl nitrate (PAN), hydrogen
21     peroxide (H2O2), and other oxidants either in the gas phase or particle phase typically have been
22     obtained only as part of special field studies.  Consequently, no data on nationwide patterns of
23     occurrence are available for these non-O3 oxidants; nor are extensive data available on the
24     relationships of levels  and patterns of these oxidants to those of O3.  However, available data for
25     gas-phase and particle-phase oxidants will be discussed.
26
27     Characterizing Ambient Ozone Concentrations
28          The "concentration" of a specific air pollutant is typically defined as the amount (mass) of
29     that material per unit volume of air.  However, most of the data presented in this annex are
30     expressed as "mixing ratios" in terms of a volume-to-volume ratio (parts per million [ppm] or
31     parts per billion [ppb]). Data expressed this way are often referred to as concentrations, both in

       August 2005                            AX3-1      DRAFT-DO NOT QUOTE OR CITE

-------
 1      the literature and in the text, following common usage. Human exposures are expressed in units
 2      of mixing ratio times time.
 3           Several different types of indicators are used for evaluating exposures of vegetation to O3.
 4      The peak-weighted, cumulative exposure indicators used in this chapter for characterizing
 5      vegetation exposures are SUM06 and SUM08 (the sums of all hourly average concentrations
 6      >0.06 and  0.08 ppm, respectively) and W126 (the sum of the hourly average concentrations that
 7      have been  weighted according to a sigmoid function that is based on a hypothetical vegetation
 8      response [see Lefohn and Runeckles, 1987]). Further discussion of these exposure indices is
 9      presented in Chapter 9.
10           The U.S. Environmental Protection Agency (U.S. EPA) has established "ozone seasons"
11      for the required monitoring of ambient O3 concentrations for different locations within the
12      United States and U.S. territories (CFR, 2000).  Table AX3-1 shows the O3 seasons during which
13      continuous, hourly averaged O3 concentrations must be monitored. Note that O3 monitoring is
14      optional outside of the "ozone season" and is monitored in many locations throughout the year.
15           In  Section AX3.2, surface O3 concentrations are characterized and the difficulties of
16      characterizing background O3 concentrations for controlled exposure studies and for assessing
17      the health benefits associated with setting the NAAQS are discussed. In addition,  hourly
18      averaged concentrations obtained by several monitoring networks have been characterized for
19      urban and  rural areas.  Spatial variations that occur within urban areas, between rural and urban
20      areas, as well as variations with elevation are discussed in Section AX3.3. The diurnal
21      variations  for the various urban and rural locations are found in Section AX3.4, where urban and
22      rural patterns are described.  In Section AX3.5  seasonal variations in 1-h and 8-h average
23      concentrations are discussed.  Section AX3.6 of this annex summarizes the historical trends for
24      1980 to 2001 on a national scale and for selected cities. The most recent U.S. EPA trends results
25      are also  presented. Section AX3.7 describes available information for the concentrations and
26      patterns  of related photochemical oxidants.  Section AX3.8 describes the co-occurrence patterns
27      of O3 with NO2, sulfur dioxide (SO2), and 24-h PM25.  Indoor O3 concentrations, including
28      sources and factors affecting indoor O3 concentrations, are described in Section AX3.9. Section
29      AX3.10  describes human population exposure measurement methods, factors influencing
30      exposure, and exposure models.
        August 2005                              AX3-2      DRAFT-DO NOT QUOTE OR CITE

-------
                    Table AX3-1. Ozone Monitoring Seasons by State
State
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
District of Columbia
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Start Month — End
March -
April —
January
March -
January
March -
April —
April —
April —
March -
March -
January
April —
April —
April —
April —
April —
March -
January
April —
April —
April —
April —
April —
March -
April —
June —
April —
- October
October
— December
- November
— December
September
September
October
October
- October
- October
— December
October
October
September
October
October
- October
— December
September
October
September
September
October
- October
October
September
October
State
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Puerto Rico
Rhode Island
South Carolina
South Dakota
Tennessee
Texas1
Texas1
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
American Samoa
Guam
Virgin Islands
Start Month — End
January
April —
April —
January
April —
April —
May —
April —
— December
September
October
— December
October
October
September
October
March — November
May —
April —
January
April —
April —
June —
September
October
— December
September
October
September
March — October
January
— December
March — October
May —
April —
April —
May —
April —
April 15
April —
January
January
January
September
September
October
September
October
— October 15
October
— December
— December
— December
     ozone season is defined differently in different sections of Texas.




 Source: CFR(2000a).
August 2005
AX3-3
DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX3.2  SURFACE OZONE CONCENTRATIONS
 2          Data for O3 concentrations in a number of different environments, ranging from urban to
 3     remote, are summarized and characterized in this section. The main emphasis is placed on the
 4     characterization of the variability of O3 concentrations in these different environments. Another
 5     important issue relates to the determination of background concentrations. There are a number
 6     of different uses of the term background depending on the context in which it is used. Various
 7     definitions of background have been covered in the 1996 O3 AQCD (U.S. Environmental
 8     Protection Agency, 1996a) and in Air Quality Criteria for Particulate Matter (PM AQCD; U.S.
 9     Environmental Protection Agency, 1996b).  This section deals with the characterization of
10     background O3 concentrations that are used for two main purposes:  (1) performing experiments
11     relating the effects of exposure to O3 on humans, animals, and vegetation; and (2) assessing the
12     health benefits associated with setting different levels of the NAAQS for O3.  Ozone background
13     concentrations used for NAAQS setting purposes are referred to as policy relevant background
14     (PRB) concentrations. PRB concentrations are defined by the U.S. EPA Office of Air Quality
15     Programs and Standards (OAQPS) as those concentrations that would be observed in the United
16     States if anthropogenic sources of O3 precursors were turned off in continental North America
17     (the United States, Canada and Mexico), i.e., the definition includes O3 formed from natural
18     sources everywhere in the world and from anthropogenic O3 precursors outside of North
19     America.  The 1996 O3 AQCD considered two possible methods for quantifying background O3
20     concentrations for the two purposes mentioned  above. The first method relied on mathematical
21     models and historical data. The second method used the distribution of hourly  average O3
22     concentrations observed at clean, relatively remote monitoring sites (RRMS) in the United States
23     (i.e., those which experience low maximum hourly concentrations). At the time of the  1996 O3
24     AQCD, simulations of mathematical models were limited; therefore, the second method was
25     employed to quantify background O3 concentrations.
26           Sections AX3.2.1 and AX3.2.2 review data for O3 concentrations in urban and nonurban
27     (but influenced by urban emissions) environments. Section AX3.2.3 reviews the data from
28     relatively clean remote sites, addresses the issue of how to use these data to help set background
29     levels for controlled exposure  studies,  and presents evidence of trends in O3 concentrations at
30     these sites. The characterization of PRB O3 concentrations will be the subject of Section
31     AX3.2.4.  Two  alternative approaches for establishing PRB concentrations are presented: the

       August 2005                              AX3-4      DRAFT-DO NOT QUOTE OR CITE

-------
 1      first uses data from relatively clean, remote monitoring sites and the second uses numerical
 2      models. The strengths and weaknesses of each approach are presented in the hopes of
 3      stimulating discussion that will resolve issues related to the use of either of these alternative
 4      methods.
 5
 6      Ozone Air Quality at Urban, Suburban, andNonurban Sites
 1           Figure AX3-1 shows the mean daily maximum 8-h O3 concentrations and Figure AX3-2
 8      shows the 95th percentile values of the daily maximum 8-h O3 concentrations, based on
 9      county wide averages across the United States from May to September 2000 to 2004.  The
10      locations of the monitoring sites used to calculate background O3 concentrations are shown in
11      Figure AX3-3.  The period from May to September was chosen because, although O3 was
12      monitored for different lengths of time across the country, all O3 monitors should be operational
13      during these months.  Data flagged because of quality control issues were removed with
14      concurrence by the local monitoring agency.  Only days for which there were 75% complete data
15      (i.e.,  18 of 24 hours) were kept, and a minimum of 115 of 153 days were required in each year.
16      Cut points for the tertile distributions on each map were chosen at the median and 95th
17      percentile values.  These cut points were chosen as they represent standard metrics for
18      characterizing important aspects of human exposure used by the EPA. Any other percentiles or
19      statistics that are believed to be helpful for characterizing human exposures could also be used.
20      Blank areas on the maps indicate no data coverage. It should be noted that county areas can be
21      much larger in the West than in the East, but monitors are not spread evenly within a county. As
22      a result, the assigned concentration range might not represent conditions throughout a particular
23      county and so large areas in western counties where there are not any monitors were blanked out.
24           As shown in Figure AX3-1, the median of the county wide, mean daily maximum 8-h O3
25      concentration across the United States is 49 ppb, and the  corresponding 95th  percentile value is
26      57 ppb. Although the median and 95th percentile value of the countywide means are fairly
27      close, these results cannot be taken to imply that mean O3 concentrations lie in a relatively
28      narrow range throughout the United States, because data coverage is not as complete in the West
29      as it is in the East. High mean  daily maximum 8-h O3 concentrations are found in California
30
        August 2005                             AX3-5      DRAFT-DO NOT QUOTE OR CITE

-------
                 Seasonal (May-September) Mean of Daily Maximum 8-Hour Values, 2002-2004
                    Concentration PPM
                                         X < 0.049
                                                     0.049 £ X < 0.057
                                                                        0.057 < X
      Figure AX3-1. Countywide mean daily maximum 8-h O3 concentrations, May to
                     September 2002 to 2004.
      Source: Fitz-Simons et al. (2005).
1     and states in the Southwest as well as in several counties in the East. As shown in
2     Figure AX3-2, the nationwide median of the county wide, 95th percentile value of the daily
3     maximum 8-h O3 concentration is 73 ppb and 5% of these values are above 85 ppb.  High values
4     for the 95th percentiles are found in California, Texas, and in the East, but not necessarily in the
5     same counties as shown for the mean daily maximum 8-h concentrations in Figure AX3-1.
6           Although mean O3 concentrations in Houston, TX were below the nationwide median, its
7     95th percentile value ranks in the highest 5% nationwide. Conversely, mean O3 concentrations
8     in southwestern states are among the highest in the United States, but peak values (i.e., 95th or
      August 2005
AX3-6
DRAFT-DO NOT QUOTE OR CITE

-------
             Seasonal (May-September) 95th Percentile of Daily Maximum 8-Hour Values, 2002-2004
                     Concentration PPM
                                         X < 0.073
                                                      0.073 < X < 0.085
                                                                          0.085 < X
       Figure AX3-2.  Countywide 95th percentile value of daily maximum 8-h O3 concentrations,
                      May to September 2002 to 2004.
       Source: Fitz-Simons et al. (2005).
1      98th percentile values) in those counties are not among the highest peak values in the United
2      States.  In other areas where the highest mean O3 concentrations occurred, such as California;
3      Dallas-Fort Worth, TX; and the Northeast Corridor, the highest peak values also were observed.
4           Although countywide averages are shown, it should be noted that considerable spatial
5      variability can exist within a county, especially within urban areas as described in Section
6      AX3.3. In addition, there can also be differences in the diurnal profile of O3 among monitors
7      within counties.
8           Box plots showing the percentile distribution of nationwide O3 concentrations for
9      different averaging periods (1-h daily maximum, 8-h daily maximum and 24-h daily average) are
       August 2005
AX3-7
DRAFT-DO NOT QUOTE OR CITE

-------
       Figure AX3-3.  Locations of monitoring sites used for calculating countywide averages
                       across the United States.
       Source: Fitz-Simons et al. (2005).
 1     given in Figures AX3-4 to AX3-6 and the numerical values are given in Table AX3-2.  The
 2     differences between the 50th and 95th percentile values can be used to provide indications of
 3     differences in O3 levels between "typical" O3 days and "high" O3 days.  These differences are
 4     approximately 40, 30, and 25 ppb for the daily 1-h, 8-h, maxima and daily averaged O3
 5     concentrations.  As might be expected, the daily maximum 1-h and 8-h O3 concentrations are
 6     highly correlated.
 7          Lehman et al. (2004) have shown that the eastern United States can be divided into five
 8     regions, each of which exhibit spatial, relatively coherent patterns of O3 properties at nonurban
 9     sites. Only sites classified as being rural or suburban and with land usage of forest, agriculture,
10     or residential were included in the analyses. These criteria were chosen to avoid sites where O3
11     is scavenged by NO that can be found in high concentrations near major sources, such as traffic
       August 2005
AX3-8
DRAFT-DO NOT QUOTE OR CITE

-------
                       1 Hour Daily Ozone Maximum Concentrations
0.24 r
0.22 -.
| 0.20 -
Q.
& 0.18-

0 0.16 -!
« 0.14 -
+•« :
g 0.10-

C 0.06 '-
O
N 0.04 -
0.02 \
0.00 '-
691012 367121 323891
X X
£ x_








_jj

p






t- -C

X




1- M








T t f
                  All Values
                           In a CSA
              Not in a CSA
Figure AX3-4.
Distribution of nationwide daily maximum 1-h average O3 concentrations
from May to September 2000 to 2004. Medians, interquartile ranges,
minima and maxima and means (as dots are shown). Values above box
plots give number of observations.
Source: Fitz-Simons et al. (2005).
                       8 Hour Daily Ozone Maximum Concentrations
0.16 :
•g- 0.14 :
Q. :
3 0.12 :
C ;
5 0.10 :
E ;
g 0-08 :
o
O 0 06
o
c 0.04 -
O :
O 0.02 -
0.00 -
6908
X

^

44 367
£

i_ -•

329 323

L. _jm

815



t T f
                  All Values
                           In a CSA
              Not in a CSA
Figure AX3-5. Distribution of nationwide daily maximum 8-h average O3 concentrations
              from May to September 2000 to 2004. Medians, interquartile ranges,
              minima and maxima and means (as dots are shown). Values above box
              plots give number of observations.

Source: Fitz-Simons et al. (2005).
August 2005
                       AX3-9
DRAFT-DO NOT QUOTE OR CITE

-------
                                  24 Hour Average Ozone Concentrations
                 E
                 Q.
                 C
                 o
                 C
                 0)
                 o
                 C
                 o
                 o
                 0)
                 C
                 o
                 N
                 o
0.12 7
0,11 \
0.10 -
0.09 -
0.08 -!
0.07 -
0.06 -i
0.05 ^
0.04 ~
0.03 ^
0.02 -i
0.01 ^
0.00 '-
691
s


012 367
1



t


J
1
121 323
>
|

i_ j

891
<
I

l_

r f
                          All Values
    In a CSA
               Not in a CSA
       Figure AX3-6. Distribution of nationwide 24-h average O3 concentrations from May to
                      September 2000 to 2004.  Medians, interquartile ranges, minima and
                      maxima and means (as dots are shown). Values above box plots give
                      number of observations.
       Source: Fitz-Simons et al. (2005).
 1     in urban cores.  The five regions, shown in Figure AX3-7, are characterized by different patterns
 2     of O3 properties such as temporal persistence and seasonal variability.  Figure AX3-7 shows
 3     nonurban, monthly average, daily maximum 8-h O3 concentrations in the five regions in the
 4     eastern United States from April to October 1993 to 2002.
 5           Regional differences are immediately apparent.  Highest concentrations among all the
 6     regions are generally found in the Mid-Atlantic region (mean of 52 ppb) with highest values
 7     throughout the percentile distribution except for the overall maximum. Lowest mean
 8     concentrations (42 ppb) are found in Florida. In the northern regions (the Northeast, Great
 9     Lakes) and the Mid-Atlantic region, highest median and peak concentrations are found in July,
10     whereas in the Southwest, highest median concentrations are found in August, with highest
11     peaks in June and September, i.e., outside the warmest summer months. In Florida, highest
12     monthly averaged median and peak concentrations are found during the spring.  High O3
13     concentrations tend  to be most persistent (3-4 days of persistence) in the southern regions, less
14     persistent in the Mid-Atlantic region (2-3 days) and least persistent in the northern regions  (1 or
       August 2005
AX3-10
DRAFT-DO NOT QUOTE OR CITE

-------
Table AX3-2. Summary of Percentiles of Pooled Data Across Monitoring Sites for May to September 2000-2004
S-
to
o








>
X
1
^

O
t?
H
6
o
2|
0
H
O
O
H
W
O
O
H
W


Pooled Group/ Number
Avg. Time of Values Mean 1
Daily 1-h Maximum Concentrations
Monitors in 367,121 58 20
CSAs
Monitors not 323,891 55 20
in CSAs
8-h Daily Maximum Concentrations
Monitors in 367,029 50 16
CSAs

Monitors not 323,815 49 16
in CSAs

24-h Average Concentrations
Monitors in 367,121 33 10
CSAs

Monitors not 323,891 34 10
in CSAs









Concentrations are in ppb.
Percentiles

5 10 25 30 50 70 75 90 95 99

29 34 44 46 56 66 70 84 94 116

28 33 43 45 54 64 67 79 87 104


23 28 37 40 49 58 61 73 81 98


23 28 37 39 48 57 59 70 77 91



15 18 24 26 32 39 41 50 56 68


15 18 25 27 33 39 41 50 56 68











-------
 "
Q.
ri
~

(8
C


O
     180-


     160-


     140-


     120'


     100-


     80-


     60.


     40-


     20-


      0-
                   Great Lakes Region
                             R
        APR
              MAY
                    JUN    JUL

                       Month
                               AUG
                                     SEP
                                           OCT
   180-


   160


   140-


S 120-
Q.
rt
— 100-

a>
C  80-


|  60.


   40-


   20-


    0-
R
                                                        APR
                                                                 Northeast Region
                                                              MAY
                                                                 JUN    JUL

                                                                    Month
                                                                               AUG
                                                                                     SEP
                                                                                           OCT
     180 :


     160-


     140.


  S 120.
  Q.
  n
  •s 100-

  a>
  C 80.


  O 60.


     40-


     20.
                   Southwest Region
                             R
                    JUN    JUL

                       Month
                               AUG
                                     SEP
                                           OCT
                                                0)
                                                C
   180-


   160-


   140-


   120-


   100-


   80.


   60.


   40-


   20.


    0-
                                                                Mid-Atlantic Region
                                                                 JUN    JUL

                                                                    Month
                                                                               AUG
                                                                                     SEP
                                                                                           OCT
                            180-


                            160-


                            140.


                         2" 120-
                         o.

                         •2= 100-
                         0>
                            80-


                            60-


                            40.


                            20.


                             0-
                                           Florida Region
                                           JUN    JUL

                                              Month
                                                       AUG
                                                             SEP
                                                                   OCT
Figure AX3-7.   Box plots showing O3 averaged by month from 1993 to 2002 in the five

                  regions in the eastern United States derived by Lehman et al. (2004). The

                  boxes define the interquartile range and the whiskers, the extreme values.


Source: Lehman et al. (2004).
August 2005
                                          AX3-12
         DRAFT-DO NOT QUOTE OR CITE

-------
 1      2 days). Analyses, such as these, are not available for the western United States, in part because
 2      of the difficulty in defining regions with relatively coherent O3 properties.
 3           Box plots showing the percentile distribution of hourly average O3 concentrations for
 4      different types of rural sites for 2001 are given in Figures AX3-8  (rural-agricultural), AX3-9
 5      (rural-forest) and AX3-10 (rural-residential or commercial). Shown below the figures are the
 6      number of observations and various metrics for characterizing vegetation exposures. Note that
 7      high O3 concentrations are found at sites that are classified as rural, as in Anne Arundel Co.,
 8      MD; Yosemite NP, CA; and Crestline, CA. Land use designations might not give an accurate
 9      picture of exposure regimes in rural areas, because the land use characterization of "rural" does
10      not necessarily mean that a specific location is isolated from anthropogenic influences. Rather,
11      the characterization refers only to the current use of the land, not to the presence of sources.
12      Since O3 produced from emissions in urban areas is transported to more rural downwind
13      locations, elevated O3 concentrations can occur at considerable distances from urban centers.
14      In addition, major sources of O3 precursors such as power plants and highways are located in
15      nonurban areas and also produce O3 in these areas.  Due to lower rates of chemical scavenging in
16      nonurban areas, O3 tends to persist longer in nonurban than in  urban areas, also tending to lead to
17      higher exposures in nonurban areas influenced by anthropogenic precursor emissions.
18
19      Ozone Air Quality Data at Relatively Remote Monitoring Sites (RRMS)
20           RRMS are sites that are located in the national parks that tend to be less affected by
21      obvious pollution sources than other sites. This does not mean that they are completely
22      unaffected by local pollution, as evidenced by the number of visitors to these national parks.
23      It is important to characterize hourly average O3 concentrations at RRMS so that assessments of
24      the possible effects of O3 on human health and vegetation use ranges of concentrations in their
25      experiments that mimic the range that is found in the United States. Hourly average
26      concentrations used as controls in controlled O3 exposures for both human health and vegetation
27      studies appear to be lower than those experienced at RRMS in the United States or in other parts
28      of the world (see Chapter 9).  Typically, ambient air is filtered to remove O3 before being
29      admitted into the exposure chambers. As a result, O3 concentrations might only be a few ppb
30      within these chambers.
31

        August 2005                             AX3-13      DRAFT-DO NOT QUOTE OR CITE

-------
                                                 Rural Agriculture
0.18-
0.17-
0.16-
—. 0.15-
| 0.14-
3 0.13-
c 0.12-
.2 0.11 -
2 0.10-
e 0.09 -
S 0.08-
§ 0.07 -
U 0.06 -
g 0.05 -
S 0.04-
O 0.03 -
0.02-
0.01 -
0.00-
No. of Hours -
Hours > 0.08 -
Hours a 0.10-
SUM06 (ppm-h) -
W126(ppm-h)-





1 1 1 n i
r~i i~i
JL
_ - _ U ~
I I T-*
y y y i x b







I I
H
3 i
5076 4367 5059 5116 5080 5055 4345
21 37 142 1 156 48 59
0 0 34 0 6
1
24.7 34.2 43.5 8.6 63.2 27.4 24.8
20.7 25.8 35.7 10.6 50.0 22.7 22.8
                                   0°''


       Figure AX3-8.  Hourly average O3 concentrations observed at selected rural-agricultural
                       sites from April to October 2001. The whiskers represent minimum and
                       maximum concentrations. The boxes represent the 10th and 90th
                       percentile concentrations.
       Source: Fitz-Simons et al. (2005).
 1          Box plots showing the percentile distribution of annual hourly averaged O3 concentrations
 2     at four relatively remote monitoring sites (RRMS) are given in Figures AX3-1 la-d. As can be
 3     seen from Figures AX3-1 la-d, annual mean values of the daily 8-h maximum O3 concentration
 4     have not changed much over the past 10 years of available data.  Mean values range typically
 5     from about 0.020 ppm to about 0.040 ppm. Concentrations only rarely exceed 0.080 ppm, in
 6     contrast to observations at other "rural" sites shown in Figures AX3-8 to AX3-10.
 7          The extent to which distributions found at sites with low maximum hourly average
 8     concentrations in the western United States are representative of sites in the eastern and
 9     midwestern United States is debatable because of regional differences in sources of precursors
10     and transport patterns. Given the high density of sources in the eastern and midwestern United
11     States, it is unclear whether a site could be found in either of these regions that would not be
       August 2005
AX3-14
DRAFT-DO NOT QUOTE OR CITE

-------
                                                  Rural Forest
0.18-
0.17-
0.16-
—. 0.15-
| 0.14-
0. 0.13-
c 012-
~ 0.11-
2 o. ID-
'S 0.09-
0 0.08 -
0 0.07 -
O 0.06 -
g 0.05-
g 0.04 -
O 0.03 -
0.02-
0.01 -
0.00-
No. of Hours -
HoursaO.08-
Hours>0.10-
SUM06(ppm-h)-
W126 (ppm-h)-





T T T T

I I I
I I i | I
T P I V
— I—

4619 4536 5074 4813
106 77 267 157
6454
86.0 44.7 149.2 98.5
65.6 37.7 106.5 73.6
       Figure AX3-9.   Hourly average O3 concentrations observed at selected rural-forest sites
                       from April to October 2001. The whiskers represent minimum and
                       maximum concentrations. The boxes represent the 10th and 90th
                       percentile concentrations.
       Source: Fitz-Simons et al. (2005).
 1     influenced by the transport of O3 from nearby urban areas. Thus, with the exception of the
 2     Voyageurs NP site in Minnesota, observations at RRMS are limited to those obtained in the
 3     western United States.  However, not all national park sites in the West can be considered to
 4     be free of strong regional pollution influences, e.g., Yosemite NP (CA), as can be seen from
 5     Figure AX3-9.
 6          The 1996 O3 AQCD (U.S. Environmental Protection Agency, 1996a) concluded that the
 7     annual average "background" concentration of O3 near sea level ranged from 0.020 to 0.035 ppm
 8     and that, during the summer, the 1-h daily maximum ranged from 0.03 to 0.05 ppm. The 1996
 9     O3 AQCD also included O3 hourly average concentrations measured at several clean, RRMS
10     mostly located in the western United States. Table AX3-3 provides a summary of the
       August 2005
AX3-15
DRAFT-DO NOT QUOTE OR CITE

-------
                                                  Rural Other
0.18-
0.17-
016-
^ 0.15-
i 0.14-
Q.
& 0.13-
c 0.12-
.2 0.11 -
5 0.10-
c 0.09-
8 008-
0 0.07 -
O 0.06 -
§ 005-
8 0.04-
O 0.03-
002-
0.01-
0.00-
No. of Hours -
Hours > 0.08 -
Hours > 0.10-
SUM06 (ppm-h) -
W126(ppm-h)-



















-,—






—
i D
4922 5088
933 9
369 0
174.8 21.4
144.9 20.3







_^





-*-|


	
J
4879
111
10
45.4
334.4
       Figure AX3-10.   Hourly average O3 concentrations observed at selected rural-commercial
                        or -residential sites from April to October 2001. The whiskers represent
                        minimum and maximum concentrations.  The boxes represent the 10th
                        and 90th percentile concentrations.
       Source: Fitz-Simons et al. (2005).
 1     characterization of the hourly average concentrations recorded from 1988 to 2001 at some of the
 2     monitoring sites previously analyzed. The percentile distribution of the hourly average
 3     concentrations (April to October), number of hourly average occurrences >0.08 and >0.10 ppm,
 4     seasonal 7-h average concentrations, the SUM06, and W126 values were characterized for those
 5     site years with a data capture of >75%. From 1988 to 2001, no hourly average concentrations
 6     >0.08 ppm were observed at monitoring sites in Redwood NP (CA), Olympic NP (WA), Glacier
 7     NP (MT), Denali NP (AK), Badlands (SD), and Custer NF (MT) during the months of April to
 8     October.  There were eight occurrences of hourly average O3 concentrations >0.08 ppm from
 9     April to October of 1997 at the monitoring site in Theodore Roosevelt NP (ND). However, no
10     hourly average concentrations >0.08 ppm were observed from April to October in any other year
11
       August 2005
AX3-16
DRAFT-DO NOT QUOTE OR CITE

-------
           a. Theodore Roosevelt National Park
          b. Yellowstone National Park
        0.16-  3525 3162 3629 1808 3649 3651 3652 3672 3199 3667
        0.14
        0.12
        0.06-
        0.02-
        0.00-
             I    I   I    I   I    I   I    I   I    I
            1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
                                                    0.16- 3276 3263 3408 3444 3385 3333 3478 3311 3343 3402
                                                    0.12-
                                                    0.04-
             I   I    I   I    I   I    I   I    I   I
            1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
           c. Glacier National Park
        0.16-  3365 3317 3169 3277 3232 3230 3320 3311 3157 3333
        0.12
        0.02-
        0.00
             I    I   I    I   I    I   I    I   I    I
            1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
          d. Olympic National Park
                                                    0.16- 3321 3438 3095 3416 3347 3351 3304 3361 3361 3276
                                                    0.10-
                                                    0.04-
                                                    0.02-
                                                    0.00-
             I   I    I   I    I   I    I   I    I   I
            1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Figure AX3-lla-d.  Daily 8-h maximum O3 concentrations observed at selected national
                      park sites.  The whiskers on the box plot represent the 10th and
                      90th percentile concentrations.  The "X"s above and below the
                      whiskers are the values that fall below and above the 10th and 90th
                      percentile concentrations. The dots inside the box represent the
                      mean. The number of observations is shown above each box plot.

Source: Fitz-Simons et al. (2005).
August 2005
AX3-17
DRAFT-DO NOT QUOTE OR CITE

-------
   Table AX3-3. Seasonal (April to October) Percentile Distribution of Hourly Ozone Concentrations (ppm), Number of
Hourly Mean Ozone Occurrences >0.08 and >0.10, Seasonal 7-h Average Concentrations, SUM06, and W126 Values for Sites
to
o
o









X
OJ
oo



o
§
H
6
o
0
H
O
o
H
W
O
O
H
W
MLl
A|JCI ici
ivlllg J-
t\J TV -LTid
LA1111U11I
xx IF ui i
y r-ivcn
igt^ V^LFI
ii^iiu a
Percentiles
Site Year
Redwood NP 1988
060150002
(California) 1989
235m
1990
1991
1992
1993
1994
Olympic NP 1989
(Washington)
530090012 1990
125m
1991

1993
1994
1995
1996
1997
1998
1999
2000
2001




Min.
0.002

0.000

0.000
0.001
0.000
0.000
0.001
0.000

0.000
0.000

0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.002




10
0.011

0.010

0.011
0.012
0.010
0.010
0.011
0.003

0.005
0.006

0.004
0.006
0.006
0.006
0.005
0.008
0.006
0.006
0.009




30
0.018

0.017

0.018
0.019
0.017
0.017
0.018
0.010

0.012
0.014

0.010
0.013
0.014
0.013
0.010
0.014
0.014
0.013
0.017




50
0.023

0.022

0.023
0.025
0.021
0.022
0.024
0.015

0.018
0.019

0.016
0.019
0.020
0.019
0.015
0.019
0.019
0.019
0.023




70
0.029

0.027

0.028
0.031
0.026
0.027
0.028
0.022

0.023
0.024

0.021
0.025
0.027
0.025
0.022
0.025
0.026
0.025
0.028




90
0.038

0.034

0.035
0.038
0.035
0.035
0.035
0.030

0.030
0.033

0.029
0.033
0.037
0.034
0.035
0.033
0.036
0.035
0.036




95
0.041

0.038

0.038
0.041
0.039
0.038
0.038
0.035

0.034
0.036

0.034
0.038
0.040
0.038
0.040
0.037
0.039
0.039
0.041




99
0.046

0.042

0.043
0.045
0.045
0.042
0.043
0.046

0.043
0.044

0.041
0.043
0.048
0.043
0.046
0.044
0.044
0.045
0.046




iiuiia TV 11.11 xr
-------
Table AX3-3 (cont'd).  Seasonal (April to October) Percentile Distribution of Hourly Ozone Concentrations (ppm), Number
of Hourly Mean Ozone Occurrences >0.08 and >0.10, Seasonal 7-h Average Concentrations, SUM06, and W126 Values for
to
o











;>
X
OJ

VO



O
?
"T1
H
6
o

0
H
O
o
H
W
O
O
H
W
k^n-cz
i^A|JCl
l^llvlll
£ AJLFYV J
TX
-------
Table AX3-3 (cont'd).  Seasonal (April to October) Percentile Distribution of Hourly Ozone Concentrations (ppm), Number
of Hourly Mean Ozone Occurrences >0.08 and >0.10, Seasonal 7-h Average Concentrations, SUM06, and W126 Values for
to
o













;>
X
OJ

to
o



O
§
H
6
o

0
H
O
o
NW'
H
W
O
O
H
W

k^ii-cz
i^A|JCl
i^iiviii
£ AJLFYV J
TX
-------
Table AX3-3 (cont'd).  Seasonal (April to October) Percentile Distribution of Hourly Ozone Concentrations (ppm), Number
of Hourly Mean Ozone Occurrences >0.08 and >0.10, Seasonal 7-h Average Concentrations, SUM06, and W126 Values for
to
o











>
X
OJ
to
^^




o

c
'•Tj
H
6
o

H
O
o
H
W
O
O
H
W
k^n-cz
i^A|JCl
l^llvlll
£ AJLFYV J
TX
-------
 1      at this site. Except for 1988, the year in which there were major forest fires at Yellowstone NP
 2      (WY), the monitoring site located there experienced no hourly average concentrations
 3      >0.08 ppm.  Logan (1989) noted that O3 hourly average concentrations rarely exceed 0.08 ppm
 4      at remote monitoring sites in the western United States. In almost all cases for the above sites,
 5      the maximum hourly average concentration was < 0.075 ppm.  The top 10 daily maximum 8-h
 6      average concentrations for sites experiencing low maximum hourly average concentrations with
 7      a data capture of >75% are summarized in Table AX3-4.  The highest 8-h daily maximum
 8      concentrations do not necessarily all occur during the summer months. For example, at the
 9      Yellowstone National Park site, the first three highest 8-h daily maximum concentrations
10      occurred in April and May in 1998, and the fourth highest, 8-h daily maximum concentration did
11      not occur until July of that year. In 1999, the first three highest, 8-h daily maximum
12      concentrations were observed in March and May, and the fourth highest value occurred in April.
13      In 2000, the four highest values occurred in May, June, July, and August.
14           The 1996 O3 AQCD (U.S. Environmental Protection Agency, 1996a) noted that the
15      7-month (April to October) average of the 7-h daily average concentrations (0900 to 1559 hours)
16      observed at the Theodore Roosevelt National Park monitoring site in North Dakota were 0.038,
17      0.039, and 0.039 ppm, respectively, for 1984, 1985, and 1986 and concluded that the range of
18      7-h seasonal averages for the Theodore Roosevelt National Park site was representative of the
19      range of maximum daily 8-h average O3 concentrations that may occur at other fairly clean sites
20      in the United States and other locations in the Northern Hemisphere.  However, as shown in
21      Table AX3-4, the representative (as given by the fourth highest) daily maximum 8-h average O3
22      concentrations at fairly clean sites in the United States are higher than the 0.038 and 0.039 ppm
23      values cited in the 1996 O3 AQCD, and more appropriate values should be used.
24           As described in the  1996  O3 AQCD, the O3 monitoring site in the Ouachita National Forest,
25      AR experienced distributions of hourly average concentrations similar to some of the western
26      sites. However, since 1993, this site has seen significant shifts, both increases and decreases, in
27      hourly average concentrations.  Figure AX3-12 shows the changes that have occurred from 1991
28      to 2001. The large changes in hourly average O3 concentrations observed at the Ouachita
29      National Forest may indicate that this rural site is influenced by the transport of pollution.  Given
30      the high density of sources in the eastern and midwestern United States, it is unclear whether a
31

        August 2005                            AX3-22      DRAFT-DO NOT QUOTE OR CITE

-------
Table AX3-4. The Top 10 Daily Maximum 8-h Average Concentrations (ppm) for Sites Experiencing Low Maximum
                         Hourly Average Concentrations with Data Capture of >75%
g Site
^ Redwood NP
060150002
(California)
235m




Olympic NP
530090012
s> (Washington)
X 125m
OJ
to
OJ


o
H
6
o
0
H
O
o
H
w
o
o
H
W
Year
1988
1989
1990
1991
1992
1993
1994
1989
1990
1991
1993
1994
1995
1996
1997
1998
1999
2000
2001



1
0.061
0.044
0.051
0.048
0.060
0.049
0.048
0.054
0.056
0.050
0.055
0.050
0.064
0.046
0.052
0.051
0.045
0.051
0.051



2
0.058
0.043
0.048
0.047
0.053
0.046
0.048
0.052
0.048
0.048
0.052
0.046
0.063
0.046
0.051
0.050
0.044
0.051
0.050



3
0.053
0.043
0.048
0.046
0.045
0.043
0.046
0.047
0.046
0.045
0.044
0.042
0.050
0.046
0.046
0.049
0.044
0.048
0.047



4
0.052
0.043
0.047
0.045
0.045
0.043
0.046
0.044
0.046
0.043
0.042
0.042
0.049
0.046
0.045
0.046
0.043
0.047
0.045



5
0.049
0.042
0.047
0.045
0.045
0.043
0.045
0.044
0.043
0.042
0.040
0.042
0.045
0.043
0.045
0.044
0.043
0.045
0.045



6
0.047
0.042
0.046
0.045
0.044
0.042
0.044
0.044
0.040
0.041
0.039
0.042
0.045
0.042
0.045
0.043
0.042
0.044
0.044



7
0.046
0.042
0.045
0.044
0.044
0.042
0.044
0.042
0.040
0.041
0.038
0.041
0.044
0.041
0.044
0.042
0.042
0.043
0.044



8
0.046
0.042
0.044
0.044
0.043
0.042
0.043
0.042
0.039
0.041
0.038
0.041
0.044
0.041
0.043
0.041
0.042
0.042
0.044



9
0.045
0.041
0.043
0.043
0.043
0.041
0.043
0.038
0.038
0.041
0.037
0.040
0.044
0.041
0.042
0.041
0.042
0.042
0.043



10
0.045
0.041
0.043
0.043
0.042
0.041
0.043
0.038
0.038
0.041
0.037
0.040
0.044
0.040
0.042
0.041
0.041
0.042
0.043




-------
Table AX3-4 (cont'd). The Top 10 Daily Maximum 8-h Average Concentrations (ppm) for Sites Experiencing Low
                   Maximum Hourly Average Concentrations with Data Capture of >75%
r+
to
o
o









X
OJ
to
•^

o
>
H
6
o
0
H
O
o
H
W
O
O
HH
H
W
Site Year
Glacier NP 1989
300298001
(Montana) 199°
963m
1992
1993
1994
1995
1996
1997
1998

1999
2000
2001
Yellowstone NP 1988
560391010
(Wyoming) 1989
2484 m
1991
1992
1993
1994
1995

1
0.062

0.058
0.060
0.062
0.055
0.057
0.061
0.059
0.056
0.060

0.065
0.059
0.054
0.068

0.067
0.057
0.059
0.066
0.057
0.067
0.064

2
0.061

0.057
0.057
0.056
0.052
0.057
0.055
0.059
0.054
0.059

0.065
0.058
0.052
0.068

0.065
0.056
0.058
0.064
0.054
0.063
0.062

3
0.060

0.055
0.057
0.055
0.051
0.056
0.053
0.058
0.052
0.058

0.060
0.058
0.049
0.067

0.064
0.054
0.058
0.064
0.054
0.063
0.061

4
0.059

0.054
0.057
0.054
0.051
0.056
0.052
0.058
0.052
0.058

0.058
0.056
0.049
0.066

0.063
0.054
0.057
0.063
0.054
0.061
0.060

5
0.058

0.053
0.056
0.054
0.050
0.055
0.052
0.057
0.052
0.056

0.056
0.054
0.049
0.066

0.063
0.053
0.056
0.063
0.053
0.061
0.059

6
0.057

0.053
0.055
0.054
0.050
0.055
0.052
0.057
0.051
0.056

0.055
0.052
0.048
0.066

0.061
0.052
0.056
0.061
0.053
0.061
0.059

7
0.056

0.052
0.055
0.053
0.049
0.055
0.051
0.055
0.050
0.055

0.055
0.051
0.047
0.064

0.061
0.050
0.056
0.061
0.053
0.061
0.059

8
0.056

0.052
0.054
0.053
0.049
0.055
0.051
0.056
0.050
0.055

0.055
0.050
0.047
0.064

0.061
0.050
0.055
0.059
0.052
0.061
0.059

9
0.056

0.052
0.054
0.053
0.049
0.054
0.051
0.055
0.050
0.055

0.055
0.050
0.047
0.063

0.061
0.049
0.055
0.059
0.052
0.059
0.059

10
0.056

0.052
0.053
0.053
0.048
0.053
0.050
0.055
0.050
0.054

0.054
0.050
0.047
0.061

0.060
0.048
0.055
0.058
0.052
0.059
0.058


-------
Table AX3-4 (cont'd). The Top 10 Daily Maximum 8-h Average Concentrations (ppm) for Sites Experiencing Low
                   Maximum Hourly Average Concentrations with Data Capture of >75%
r+
to
o
o







X
OJ
to


o
^
H
6
o
0
H
O
o
H
W
O
O
H
W
Site
Yellowstone NP
560391011
(Wyoming)
2468m


Denali NP
022900003
(Alaska)
640m









Badlands NP
460711001
(South Dakota)
730m


Year
1997
1998
1999
2000
2001
1988
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
1989
1990
1991


1
0.065
0.069
0.078
0.070
0.068
0.055
0.049
0.054
0.053
0.053
0.053
0.058
0.058
0.054
0.057
0.056
0.046
0.061
0.069
0.061
0.058


2
0.065
0.068
0.074
0.069
0.068
0.054
0.048
0.054
0.052
0.053
0.051
0.056
0.053
0.053
0.056
0.056
0.046
0.058
0.066
0.059
0.058


3
0.062
0.066
0.073
0.067
0.066
0.054
0.048
0.050
0.052
0.051
0.049
0.056
0.053
0.052
0.056
0.054
0.044
0.057
0.064
0.055
0.056


4
0.061
0.066
0.071
0.065
0.066
0.053
0.048
0.050
0.051
0.048
0.049
0.054
0.053
0.051
0.055
0.054
0.044
0.055
0.063
0.055
0.056


5
0.061
0.063
0.070
0.065
0.065
0.053
0.048
0.047
0.050
0.048
0.049
0.051
0.052
0.051
0.054
0.054
0.044
0.055
0.060
0.054
0.056


6
0.060
0.063
0.070
0.065
0.064
0.053
0.047
0.046
0.050
0.047
0.048
0.050
0.052
0.050
0.054
0.053
0.043
0.055
0.058
0.052
0.055


7
0.057
0.061
0.070
0.064
0.064
0.052
0.047
0.046
0.049
0.047
0.048
0.049
0.052
0.050
0.054
0.053
0.043
0.053
0.057
0.052
0.055


8
0.056
0.061
0.069
0.064
0.064
0.052
0.046
0.046
0.049
0.046
0.048
0.046
0.052
0.049
0.054
0.053
0.042
0.053
0.057
0.051
0.054


9
0.056
0.061
0.068
0.063
0.064
0.052
0.046
0.045
0.049
0.046
0.048
0.046
0.051
0.049
0.053
0.052
0.042
0.053
0.057
0.051
0.054


10
0.056
0.060
0.067
0.063
0.063
0.052
0.046
0.044
0.049
0.046
0.048
0.046
0.051
0.049
0.053
0.051
0.042
0.053
0.057
0.050
0.053



-------
Table AX3-4 (cont'd). The Top 10 Daily Maximum 8-h Average Concentrations (ppm) for Sites Experiencing Low
                   Maximum Hourly Average Concentrations with Data Capture of >75%
r+
to
o
o









X
OJ
to
ON


O
J>
H
6
o
0
H
O
o
H
W
O
O
H
W
Site
Theod. Roos. NP
380530002
(North Dakota)
730m







Theod. Roos. NP
380070002
(North Dakota)
808m
Custer NF, MT
300870101
(Montana)
1006m






Year
1984

1985
1986
1989
1992
1993
1994
1995
1996
1997
1999

2000
2001
1978

1979
1983





1
0.064

0.058
0.059
0.073
0.060
0.062
0.066
0.060
0.060
0.080
0.063

0.062
0.060
0.069

0.073
0.069
0.064





2
0.062

0.055
0.058
0.069
0.059
0.059
0.064
0.059
0.059
0.073
0.060

0.061
0.059
0.065

0.066
0.069
0.061





3
0.062

0.055
0.057
0.066
0.058
0.056
0.058
0.058
0.059
0.072
0.059

0.060
0.059
0.063

0.066
0.069
0.060





4
0.062

0.054
0.056
0.065
0.058
0.056
0.058
0.058
0.059
0.071
0.058

0.059
0.058
0.062

0.065
0.068
0.060





5
0.059

0.054
0.055
0.065
0.056
0.055
0.057
0.058
0.058
0.069
0.057

0.059
0.058
0.061

0.063
0.067
0.059





6
0.058

0.054
0.055
0.064
0.056
0.053
0.056
0.058
0.058
0.068
0.057

0.057
0.057
0.061

0.060
0.067
0.058





7
0.057

0.053
0.054
0.063
0.056
0.052
0.056
0.057
0.057
0.066
0.057

0.057
0.057
0.060

0.060
0.066
0.058





8
0.057

0.053
0.053
0.063
0.054
0.052
0.056
0.057
0.057
0.066
0.056

0.057
0.056
0.060

0.060
0.066
0.058





9
0.057

0.053
0.053
0.063
0.054
0.052
0.056
0.056
0.057
0.063
0.056

0.057
0.055
0.060

0.059
0.064
0.056





10
0.057

0.052
0.052
0.063
0.054
0.052
0.055
0.055
0.056
0.063
0.056

0.057
0.055
0.058

0.059
0.064
0.056






-------
                          1991  1992  1993  1994  1995  1996  1997  1998  1999  2000  2001
                                                   Year

       Figure AX3-12.  Seasonal SUM06 and W126 exposure indices for the Ouachita National
                        Forest for the period of 1991 to 2001.
 1
 2
 3
 4
 5
 9
10
11
12
13
14
15
16
17
18
site could be found in either of these regions that would not be influenced by the transport of O3
from urban areas.  Thus, with the exception of the Voyageurs National Park site, observations in
this section are limited to those obtained at relatively clean, remote sites in western North
America.

AX3.2.1     Nationwide Distribution of Metrics for Characterizing Exposures
             of Vegetation to Ozone
     The previous O3 AQCD (U.S. Environmental Protection Agency, 1996a) concluded that
higher hourly average concentrations (>0.10 ppm) should be provided greater weight than
mid-level (0.06 to 0.099 ppm) and lower hourly average concentrations in predicting injury and
yield reduction for agricultural crops and forests. The most recent findings concerning the
importance of the higher hourly average concentrations in comparison to the mid-level and lower
values will be discussed in Chapter 9. Because of a lack of air quality data collected at rural and
remote locations, interpolation techniques, such as kriging, have been applied to the estimation
of O3 exposures across the United States (Reagan, 1984; Lefohn et al., 1987; Knudsen and
Lefohn, 1988). "Kriging" (Matheron, 1963) has been used in the analyses of air quality data
(Grivet, 1980; Faith and Sheshinski, 1979) and was used to provide estimates of seasonal O3
values for  the National Crop Loss Assessment Network (NCLAN) for 1978 through 1982 (May
       August 2005
                                       AX3-27
DRAFT-DO NOT QUOTE OR CITE

-------
 1      to September of each year) (Reagan, 1984).  These values, along with updated values, coupled
 2      with exposure-response models, were used to predict agriculturally related economic benefits
 3      anticipated by lower O3 levels in the United States (Adams et al., 1985, 1989).
 4           For 2001, ordinary kriging was used to estimate the seasonal W126, SUM06, and number
 5      of hours >0.10 ppm (N100), using hourly average concentrations accumulated over a 24-h
 6      period. As discussed in Chapter 9, the correlation between the number of occurrences of hourly
 7      average concentrations >0.10 ppm and the magnitude of the W126 and SUM06 values is not
 8      strong. Because of this, the N100 was also estimated, along with the W126 and SUM06
 9      exposure indices. For the period of April through September, the estimates of the seasonal
10      W126, SUM06, and N100 exposure index values were made for each 0.5° by 0.5° cell in the
11      continuous United States. The kriged values, the variance, and the 95% error bound for each
12      0.5° by 0.5° cell were estimated. Because of the concern for inner-city depletion caused by NOX
13      scavenging, data from specific monitoring stations located in large metropolitan areas were not
14      included in the analysis.
15           Figure AX3-13 shows the kriged values for the 24-h cumulative seasonal W126 exposure
16      index and the N100 index for 2001 for the eastern United States. Note that for some of the areas
17      with elevated W126 values (e.g., >35 ppm-h), the number of hourly average concentrations was
18      estimated to be <22.  Figure AX3-14 illustrates the kriged values using the 24-h cumulative
19      seasonal SUM06 exposure index and the N100 index for 2001 for the eastern United  States.
20      Figures AX3-15 and AX3-16 show the W126 and SUM06 values, respectively, with the N100
21      values for the central United States region. For 2001, the number of hourly average
22      concentrations >0.10 ppm was usually <22 for the 6-month period. Figures AX3-17 and
23      AX3-18 illustrate the W126 and SUM06 values, respectively, for the western United States
24      region. Note that in the Southern California  and Central California areas, the number of hourly
25      average concentrations >0.10 ppm was in the range of 48 to 208 for the 6-month period. This is
26      considerably greater than the frequency  of occurrences for the higher hourly average
27      concentrations observed in the eastern and central United States.
28           Due to the scarcity of monitoring sites across the United States, especially in the Rocky
29      Mountain region, the uncertainty in the estimates for the various exposure indices vary.
30      Figures AX3-19 through AX3-27 illustrate the 95%  confidence intervals associated with the
        August 2005                            AX3-28      DRAFT-DO NOT QUOTE OR CITE

-------
                                  » * *
                                 -V»
                          0    125    250
                                                500
                                                           750
                                                                      1,000
                                                                     ^^ Miles
                                            1:13,936,494
Figure AX3-13.
Six-month (April to September) 24-h cumulative W126 exposure index
with the number of hourly average concentrations ^0.10 ppm (N100)
occurring during 2001 for the eastern United States.
August 2005
                      AX3-29
DRAFT-DO NOT QUOTE OR CITE

-------
                  fc .* J • • L .~

                                     i^jjjjjii



                                       	   	
                                     J^a^|Kl»^P^E^^^S
                       f'«S|%^S^=t|B|J^^|i|»_ ^r-JpUpftiSp «
      N100 (hours)
                            0    125   250
                                              1:13,934,062
                                                                         1,000
                                                                           Miles
Figure AX3-14.   Six-month (April to September) 24-h cumulative SUM06 exposure index
                 with the number of hourly average concentrations >0.10 ppm (N100)
                 occurring during 2001 for the eastern United States.
August 2005
AX3-30
DRAFT-DO NOT QUOTE OR CITE

-------
      N100 (hours)
        •   1-2
        »   >2-7
        •   >7-21
        •   >21 - 48
        A   >48 - 208
      W126 (ppm-h)
      I	i 1-19
      j	j >19-35
      |    I >35 - 68
           >63- 146
                                                                  v
                                                      i:







I • ••!!••***
i 	 +...-• • • • • Vwiwi^ipui,



* * » • • » *••• * 	 *
•»»•***»+•••

\ * * * * •







...#
JL



*

*






.

*^



• •
* • 1

» V
• * *





\ ' *





:^:
"*>


*






_*







*






»












M-y






-.-.—^


                      N
100   200
               400
                         600
                                   800
                                  ^3 Miles
            1:11,998,827
Figure AX3-15.  Six-month (April to September) 24-h cumulative W126 exposure index
                with the number of hourly average concentrations ^0.10 ppm (N100)
                occurring during 2001 for the central United States.
August 2005
     AX3-31
DRAFT-DO NOT QUOTE OR CITE

-------

"X
N1 00 (hours)
• 1-2
» >2-7
• >7-21
• >21 - 48
A >48 - 208
SUM06 (ppm-h)
[~] 1-18
>55-110
I ' ' * *1
! x >
I "•"-' 	 ^/ V_
I "l ''•-.
I \ ^

f 1 	 /•
[ ' * \ Y
. . , , -a . . v
'*** ** » % f # f 9 * \
' * * * * * •f^=|= *****/
f » » J| f • • * * ***X*,,
• • SJL_*_4 'J








» • • » »>-• » + . . »f
i • • • «BipaB|^Pip*iiSP » » * •»»»»«!

t**HH**jHHH*4ttHH**\t* f


%•_ i -x • • • » • »••••• %?^ *| A» ?'^''
4^-4
0 100 200 400 600 800
1:11,998,827
Figure AX3-16.  Six-month (April to September) 24-h cumulative SUM06 exposure index
               with the number of hourly average concentrations ^0.10 ppm (N100)
               occurring during 2001 for the central United States.
August 2005
AX3-32
DRAFT-DO NOT QUOTE OR CITE

-------
                                                                  800
                                                                   Miles
                                           1:13,052,120
Figure AX3-17.  Six-month (April to September) 24-h cumulative W126 exposure index
               with the number of hourly average concentrations ^0.10 ppm (N100)
               occurring during 2001 for the western United States.
August 2005
AX3-33
DRAFT-DO NOT QUOTE OR CITE

-------
                                                                  800
                                                                    Miles
                                            1:13,052,120
Figure AX3-18.
Six-month (April to September) 24-h cumulative SUM06 exposure index
with the number of hourly average concentrations ^0.10 ppm (N100)
occurring during 2001 for the western United States.
August 2005
                     AX3-34
DRAFT-DO NOT QUOTE OR CITE

-------
                                                W126 95% Conf. Interval
                                                       3-10 ppm-h
                                                      | >10-15 ppm-h
                                                      j >15-25 ppm-h
                                                       >25 - 90 ppm-h
Figure AX3-19.  The 95% confidence interval for the 6-month (April to September) 24-h
                cumulative W126 exposure index for 2001 for the eastern United States.
                                               SUM06 95% Conf. Interval
                                                   i   12-12 ppm-h
                                                   |   i >12 -15 ppm-h
                                                       >15 -20 ppm-h
                                                      [ > 20 - 73 ppm-h
Figure AX3-20.  The 95% confidence interval for the 6-month (April to September)
                 24-h cumulative SUM06 exposure index for 2001 for the eastern
                 United States.
August 2005
AX3-35
DRAFT-DO NOT QUOTE OR CITE

-------
                                               N100 95% Conf. Interval
                                                   1	 0-lOh
                                                       >10-15h
                                                       >15-20h
                                                      [ >20 - 95 h
                                                      i >95-12?h
Figure AX3-21.  The 95% confidence interval for the 6-month (April to September) 24-h
                cumulative N100 exposure index for 2001 for the eastern United States.
                 W126 95% Conf. Interval
                        : 3- 10ppm-h
                        | >10 -15 ppm-h
                        i >15 - 25 ppm-h
                         25 - 90 ppm-h
Figure AX3-22.  The 95% confidence interval for the 6-month (April to September) 24-h
                cumulative W126 exposure index for 2001 for the central United States.
August 2005
AX3-36
DRAFT-DO NOT QUOTE OR CITE

-------




SUM06 95% Conf. Interval |
; 2- 12ppm-h !
,__: >12- 15ppm-h ;
>1 5 - 20 pprn-h '•
> 20 - 73 ppnn-h !


x-
^-ss%


B
list
B





iffis,
=


S

w«

ftJASS
i
^
i
IS?
Si

I f
I ---

V 	 -v-?
1 \
	

''~~-~-\s.^,^_ . 1 ,'
^s ^'^ ^J-| 	 e
[ 	 }
\ /
i
Figure AX3-23.  The 95% confidence interval for the 6-month (April to September)
                24-h cumulative SUM06 exposure index for 2001 for the central
                United States.
                 N100 95% Conf. Interval
                       I o- lOh
                       ; >10-1Sh
                        >20 - 95 h

                        >95 -127 h
                                                   -^
Figure AX3-24.  The 95% confidence interval for the 6-month (April to September) 24-h
                cumulative N100 exposure index for 2001 for the central United States.
August 2005
AX3-37
DRAFT-DO NOT QUOTE OR CITE

-------
                W126 95% Conf. Interval

                   'y^vi 3- 10ppm-h
                   	j >10 -15 ppm-h
                       >15 - 25 ppm-h
                       >25 - 90 ppm-h
Figure AX3-25.  The 95% confidence interval for the 6-month (April to September)
                24-h cumulative W126 exposure index for 2001 for the western
                United States.
               SUM06 95% Conf. Interval
                      I >12-15 ppm-ti
                       >15 - 20 ppm-h
                       > 20 - 73 ppm-h
Figure AX3-26.  The 95% confidence interval for the 6-month (April to September)
                24-h cumulative SUM06 exposure index for 2001 for the western
                United States.
August 2005
AX3-38
DRAFT-DO NOT QUOTE OR CITE

-------
                         N100 95% Conf. Interval
                                0- 10 h
                                >15-20h
                                >20 - 95 h
                                >95-127h
       Figure AX3-27.  The 95% confidence interval for the 6-month (April to September)
                        24-h cumulative N100 exposure index for 2001 for the western
                        United States.
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
indices by region.  In some cases, the uncertainty in the estimates of the exposure indices is
large. However, based on the actual hourly average concentrations measured, the pattern of
distinct differences across the regions in the United States for the number of hourly average
concentrations >0.10 ppm is real even though the uncertainty in the kriged estimates may be
large.
AX3.3   SPATIAL VARIABILITY IN OZONE CONCENTRATIONS
     The spatial variability of O3 concentrations in different environments in the United States
occurring across a variety of spatial scales is characterized in this section. This information will
be useful for understanding the influence of regional or altitudinal differences in O3 exposure on
vegetation and for establishing the spatial variations in O3 concentrations as they are used in
epidemiologic studies.  Intracity variations in O3 concentrations are described in Section
       August 2005
                                       AX3-39
DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX3.3.1. Small scale horizontal and vertical variations in O3 concentrations are discussed in
 2      Section AX3.3.2.  Ozone concentrations at high elevations are characterized in Section AX3.3.3.
 3
 4      AX3.3.1 Spatial Variability of Ozone Concentrations in Urban Areas
 5           A number of processes can contribute to spatial variability in O3 concentrations in urban
 6      areas. Ozone formation occurs more or less continuously downwind of sources of precursors,
 7      producing a gradient in O3 concentrations. Ozone "titration" by reaction with NO can deplete O3
 8      levels near NO sources such as highways and busy streets. Differences in surface characteristics
 9      affect the rate of deposition of O3.  Mixing of O3 from aloft can also lead to local  enhancements
10      in O3 concentration.
11           The spatial variability in O3 concentrations in 24 MSAs across the United States is
12      characterized in this section.  These areas were chosen to provide analyses to help guide in risk
13      assessments, to provide a general overview of the spatial variability of O3 in different regions of
14      the country, and also to provide insight in to the spatial distribution of O3 in cities where health
15      outcome studies have been conducted. Statistical analyses of the human health effects of
16      airborne pollutants based on aggregate population time-series data have often relied on ambient
17      concentrations of pollutants measured at one or more central sites in a given metropolitan area.
18      In the particular case of ground-level O3 pollution, central-site monitoring has been justified as a
19      regional measure of exposure partly on grounds that  correlations between concentrations at
20      neighboring sites measured over time are usually high (U.S. Environmental Protection Agency,
21      1996a).  In analyses where multiple monitoring sites provide ambient O3 concentrations, a
22      summary measure such as an averaged concentration has often been regarded as adequately
23      characterizing the exposure distribution. Indeed, a number of studies have referred to
24      multiple-site averaging as the method for estimating  O3 exposure (U.S. Environmental
25      Protection Agency, 1996a).  It is hoped that the analyses presented here will shed some light on
26      the suitability of this practice.  Earlier analyses were reported in the previous O3 AQCD (U.S.
27      Environmental Protection Agency, 1996a).  The analyses presented there concluded that the
28      extent of spatial homogeneity is specific to the MSA under study. In particular, cities with low
29      traffic densities that are located downwind of major sources of precursors are heavily influenced
30      by long range transport and tend to show smaller spatial variability (e.g., New Haven, CT) than
31      those source areas with high traffic densities located  upwind (e.g., New York, NY).

        August 2005                             AX3-40      DRAFT-DO NOT QUOTE OR CITE

-------
 1           Metrics for characterizing spatial variability include the use of Pearson correlation
 2      coefficients (r), values of the 90th percentile (P90) absolute difference in concentrations, and
 3      coefficients of divergence (COD)1.  These methods of analysis follow those used for
 4      characterizing PM25 and PM10_25 concentrations in Pinto et al. (2004) and in the latest edition of
 5      the PM AQCD (U.S. Environmental Protection Agency, 2004a).  Data were aggregated over the
 6      local O3 season as indicated in Table AX3-1.  The length of the O3 season varies across the
 7      country. In several southwestern states, it lasts all year long. In other areas, such as in New
 8      England, the mid-Atlantic states, the Midwest and the Northwest it can be 6 months long, but
 9      typically it lasts from April through October.
10           Table AX3-5 shows the urban areas chosen, the range of 24-h average O3 concentrations
11      over the O3 season, the range of intersite correlation coefficients, the range of P90 differences
12      in O3 concentrations between site pairs, and the range in COD values.  A COD of zero implies
13      that values in both data sets are identical, and a COD of one indicates that two data sets are
14      completely different. In general, statistics were calculated for partial MSAs.  This was done so
15      as to obtain reasonable lower estimates of the spatial variability that is present, as opposed to
16      examining the consolidated MSAs.  In Boston, MA and New York, NY, this could not be readily
17      done, and  so statistics were  calculated for the consolidated MSAs. More detailed calculations
18      for a subset of nine MSAs are given in Figures AX3-28 through AX3-36.
19           As can be seen, there are no clearly discernible regional trends in the  ranges of parameters
20      shown.  Additional urban areas would need to be examined to discern broadscale patterns. The
21      data indicate considerable variability in the concentration fields. Mean O3  concentrations vary
22      within individual urban areas from factors of 1.4 to 4.0.
23           The highest annual mean  O3 concentration (0.058 ppm) is found in the Phoenix, AZ MSA
24      at a site which is located in the  mountains well downwind of the main urban area. The lowest
25      annual mean O3 concentration (0.010 ppm) was found in Lynwood in the urban core of the
26      Los Angeles MSA.  CO and NOX monitors at this  site recorded the highest  concentrations in
        1 The COD is defined as follows:
                                            I  I  % i [  Xjj  •%}& 1
                                                                                          (AX3-1)

        where xv and xik represent the 24-h average PM2 5 concentration for day /' at site y and site k andp is the number of
        observations.

        August 2005                              AX3-41      DRAFT-DO NOT QUOTE OR CITE

-------
&
^
S-
to
o
o









X
OJ
to

o
>
H
6
o
0
H
O
o
H
W
O
O
HH
H
W
Table

Urban Area
Boston, MA
New York, NY
Philadelphia, PA
Washington, DC
Charlotte, NC
Atlanta, GA
Tampa, FL
Detroit, MI
Chicago, IL
Milwaukee, WI
St. Louis, MO
Baton Rouge, LA
Dallas, TX
Houston, TX
Denver, CO
El Paso, TX
Salt Lake City, UT
Phoenix, AZ
Seattle, WA
Portland, OR
Fresno, CA
Bakersfield, CA
Los Angeles, CA
Riverside, CA
"P90 = 90th nercentile abs<
AX3-5. Summary Statistics for Ozone (in ppm) Spatial Variability in Selected U.S. Urban Areas
Number of
Sites
18
29
12
20
8
12
9
7
24
9
17
7
10
13
8
4
8
15
5
5
6
8
14
18
ilute difference in co
Minimum
Mean Cone.
0.021
0.015
0.020
0.022
0.031
0.023
0.024
0.022
0.015
0.027
0.022
0.018
0.028
0.016
0.022
0.022
0.029
0.021
0.015
0.015
0.030
0.028
0.010
0.018
ncentrations.
Maximum
Mean Cone.
0.033
0.041
0.041
0.041
0.043
0.047
0.035
0.037
0.039
0.038
0.038
0.031
0.043
0.036
0.044
0.032
0.048
0.058
0.038
0.036
0.047
0.047
0.042
0.054

Minimum
Corr. Coeff.
0.46
0.45
0.79
0.72
0.48
0.63
0.74
0.74
0.38
0.73
0.78
0.81
0.67
0.73
0.60
0.81
0.52
0.29
0.63
0.73
0.90
0.23
0.42
0.38

Maximum
Corr. Coeff.
0.93
0.96
0.95
0.97
0.95
0.94
0.94
0.96
0.96
0.96
0.96
0.95
0.95
0.96
0.92
0.94
0.92
0.95
0.94
0.91
0.97
0.96
0.95
0.95

Minimum
P a
^90
0.012
0.0080
0.011
0.010
0.012
0.013
0.011
0.0090
0.0080
0.0090
0.0090
0.0090
0.011
0.0090
0.013
0.012
0.012
0.011
0.0080
0.011
0.0090
0.013
0.010
0.013

Maximum
p
^90
0.041
0.044
0.036
0.032
0.038
0.045
0.025
0.027
0.043
0.025
0.031
0.029
0.033
0.027
0.044
0.023
0.043
0.057
0.024
0.025
0.027
0.052
0.053
0.057

Minimum
COD"
0.17
0.17
0.23
0.17
0.17
0.24
0.20
0.19
0.16
0.18
0.15
0.23
0.16
0.20
0.16
0.24
0.13
0.15
0.16
0.20
0.17
0.20
0.22
0.15

Maximum
COD
0.45
0.55
0.46
0.45
0.32
0.55
0.35
0.36
0.50
0.33
0.41
0.41
0.36
0.38
0.46
0.31
0.51
0.61
0.46
0.50
0.40
0.58
0.59
0.64

bCOD = coefficient of divergence for different site pairs.

-------
                           Charlotte - Gastonia - Rock Hill, NC - SC MSA
             a.
                            c.
              40    20
             b.
 o
40
                           Kilometers
AIRS Site ID
Site A
SiteB
SiteC
SiteD
Site E
SiteF
SiteG
SiteH
37-1
37-1
37-1
37-1
37-1
37-1
37-1
09-0004
19-0041
19-1005
19-1009
59-0021
59-0022
79-0003
45-091-0006
Site
mean
(ppm)
A


A
0.034
1


B
0.034
0.87
0.020
0.28
4508
C
0.031
0.86
0.021
0.29
4824
D
0.035
0.88
0.019
0.28
4834
E
0.039
0.88
0.021
0.28
4821
F
0.040
0.86
0.023
0.31
4677
G
0.038
0.84
0.023
0.29
4664
H
0.035
0.85
0.021
0.29
4749
 80
	i
D
                                                        H
1   0.93  0.95  0.91   0.87  0.90  0.82
   0.015  0.012 0.018 0.020 0.017 0.021
   0.20  0.17  0.23   0.26  0.24  0.27
   4706  4720 4503  4366  4341 4621
     1   0.91  0.85   0.82  0.87  0.82
        0.017 0.023 0.025 0.020 0.022
        0.21  0.28   0.30  0.27  0.30
        5045 4821  4677  4657 4955
         1    0.93   0.89  0.87  0.81
            0.015 0.018 0.019 0.022
             0.21   0.25  0.25  0.27
            4832  4688  4664 4968
              1    0.89  0.88  0.83
                 0.016 0.018 0.021
                  0.22  0.22  0.24
                 4684  4651 4744
                   1   0.81  0.81
                      0.022 0.021
                      0.26  0.23
                      4510 4608
                       1    0.83
                          0.019
                           0.26
                          4582
                                              Key
                                    Pearson correlation coefficient
                                    90th %-ile difference in concentration
                                    Coefficient of divergence
                                    Number of paired observations
                                                                                             1
       Figure AX3-28.
Locations of O3 sampling sites (a) by AQS ID# (b) and intersite
correlation statistics (c) for the Charlotte, NC-Gastonia-Rock Hill, SC
MSA. The mean observed O3 concentration at each site is given above
its letter code.  For each data pair, the Pearson correlation coefficient,
90th percentile difference in  absolute concentrations, the coefficient of
divergence, and number of observations are given.
1      California, indicating that titration of O3 by NO freshly emitted from tail pipes of motor vehicles

2      is responsible for the low O3 values that are found.  Ratios of highest to lowest mean O3

3      concentrations in these two MS As are among the highest shown in Table AX3-5. Both of these

4      MS As are characterized by sunny, warm climates; sources of precursors that are associated with

5      O3 titration to varying degrees in their urban centers; and with maximum O3 found well
       August 2005
                         AX3-43
                           DRAFT-DO NOT QUOTE OR CITE

-------
                                     Baton Rouge, LA MSA
                                                    c.
                 25
                  i	
             b.
     o
     i
                            Kilometers
                         AIRS Site ID
                      Site A 22
                      Site B 22
                      Site C 22
                      Site D 22
                      Site E 22
                      Site F 22
                      Site G 22
         005-0004
         033-0003
         033-0009
         033-0013
         033-1001
         063-0002
         121-0001
 50
	i
Site
mean
(ppm)

A



B



C



D



E



F



G


A B
0.022 0.023
1 0.88
0.015
0.32
8378
1












C
0.023
0.86
0.016
0.33
8242
0.95
0.009
0.23
8159
1








D
0.025
0.90
0.015
0.32
8342
0.87
0.015
0.29
8266
0.87
0.015
0.30
8224
1




E
0.026
0.87
0.017
0.36
8381
0.92
0.012
0.27
8301
0.92
0.012
0.27
8171
0.93
0.011
0.24
8266
1
F

0.025
0.94
0.011
0.28
8382
0.86
0.016
0.32
8302
0.86
0.017
0.32
8260
0.93
0.012
0.24
8379
0.89
0.015
0.29
8306












1


Key
Pearson correlation coefficient
90th %-ile difference in concentration
Coefficient of divergence
Number of paired observations








G
0.023
0.86
0.016
0.33
8360
0.93
0.010
0.26
8281
0.94
0.010
0.23
8157
0.88
0.015
0.28
8251
0.94
0.011
0.25
8290
0.86
0.016
0.31
8292
1



      Figure AX3-29.
Locations of O3 sampling sites (a) by AQS ID# (b) and intersite
correlation statistics (c) for the Baton Rouge, LA MSA. The mean
observed O3 concentration at each site is given above its letter code.
For each data pair, the Pearson correlation coefficient, 90th percentile
difference in absolute concentrations, the coefficient of divergence, and
number of observations are given.
1     downwind of the urban centers. Intersite correlation coefficients show mixed patterns, i.e.,

2     in some urban areas all pairs of sites are moderately to highly correlated, while other areas show

3     a very large range of values. As may be expected, those areas which show smaller ratios of

4     seasonal mean concentrations also exhibit a smaller range of intersite correlation coefficients.

5     Within the examined urban areas, P90 values were evenly distributed between all site pairs

6     considered. The CODs indicate variability among site pairs. However, there are a number of
      August 2005
                       AX3-44
            DRAFT-DO NOT QUOTE OR CITE

-------
                              Detroit -Ann Arbor - Flint, Ml  CMSA
                                                               b.
                                          Kilometers
                                                                      AIRS Site ID
                                                                    Site A
                                                                    SiteB
                                                                    SitaC
                                                                    SiteD
                                                                    SiteE
                                                                    SiteF
                                                                    SiteG
                                                                    SiteM
                                                                    Site I
                                                                    SiteJ
                                                                    SiteK
                                                   26-049-
                                                   26-049-
                                                   26-091-
                                                   26-099-
                                                   26-099-
                                                   26-125-
                                                   26-147-
                                                   26-161-
                                                   26-163-
                                                   26-163-
                                                   26-163-
          0021
          2001
          0007
          0009
          1003
          0001
          0005
          0008
          0001
          0016
          0019
                           50
                                                         100
                          C.   Site
                               mean
                               (ppm)
                                              H    I
                                    0.033 0.035  0.037 0.032  0.031  0.028 0.031  0.032 0.026  0.028 0.028
                                     1   0.93  0.84  0.84  0.86  0,86  0.81  0.86  0.82  0.84 0.85
                                        0.012  0,019 0.019  0.019 0.020  0.020 0.019 0,024  0.022 0.020
                                        0.20  0.24  0.26  0.27  030  0.26  0.27  035  0.32 0.29
                                        4333  4151  4253  4341 4103  4338 4163 4228  4324 4247
                                          1
                                             O.B4  0.86  0.85  0,84  0.82  0.84  0.81  0.81  0.84
                                             0019 0.017  0.020 0.023  0.019 0.021 0025 0.024 0.021
                                             0.22  0.25  0.28  0,33  0.25  0.28  0.37  0.36  0.32
                                             4152  4252  4341 4104  4339 4184 4230  4323  4248

                                              1   0.85  0.86  087  0.78  0.92  0.84  0.86  0.86
                                                 0.020  0.020 0.022  0.023 0.015 0,025 0.023 0.021
                                                 027  0.28  032  0.27  0.25  0.37  0.35  032
                                                 4072  4159 3930  4156 4164 4049  4143  4068

                                                  1    0.91  0,88  0.86  0.86  0.84  0.87  0.90
                                                      0.014 0.019  0.017 0.019 0022 0.020 0.016
                                                      0.26  0,31  0.25  0.27  0.34  0.31  0.27
                                                      4260 4022  4259 4086 4148  4245  4193

                                                       1    0,94  0.83  0.89  0.88  0.94  0.96
                                                          0.014  0.020 0017 0020 0.013 0.010
                                                           024  0.27  0.25  0.31  0.24  0.19
                                                          4112  4346 4172 4235  4330  4255
                                                            1   0.83  0.91  0.85  0.93  0.93
                                                               0.021 0.015 0020 0.013 0.014
                                                               0.29  0.25  0.31  0.23  0.24
                                                               4109 3944 4000  4093  4025
                                                                1   0.79  0.76  0.80  0.83
                                                                   0.022 0,024 0.022 0.020
                                                                   0.28  0.36  0.33  0.29
                                                                   41 S3 4233  4329  4255
                                                                    1
                                               Key
                                     Pearson correlation coeflcient
                                     90th %-ile difference in concentration
                                     Coefficient of divergence
                                     Number of paired observations
                                                                        0.88  0.89 0.89
                                                                       0,019  0.017 0.017
                                                                        0.34  0.29 0.27
                                                                       4062  4153 4081

                                                                         1   0.89 0.90
                                                                            0.016 0.018
                                                                            0.29 0.29
                                                                            4217 4144

                                                                             1   0.95
                                                                                0.011
                                                                                0.21
                                                                                4242
Figure AX3-30.
Locations of O3 sampling sites (a) by AQS ID# (b) and intersite
correlation statistics (c) for the Detroit-Ann Arbor-Flint, MI CMSA.
The mean observed O3 concentration at each site is given above its letter
code.  For each data pair, the Pearson correlation coefficient, 90th
percentile difference in absolute concentrations, the coefficient of
divergence, and number  of observations are given.
August 2005
                               AX3-45
DRAFT-DO NOT QUOTE OR CITE

-------
                         St. Louis, MO - IL CMSA
    a.
                                          b.
    60
     i	
30
0
i
 60
	i	
 120
	i
                     Kilometers
                                                         AIRS Site ID
Site A
SiteB
SiteC
SiteD
SiteE
SiteF
SiteG
SiteH
Site I
Site J
SiteK
SiteL
SiteM
SiteN
SiteO
SiteP
SiteQ
17-083-
17-119-
17-119-
17-119-
17-119-
17-163-
29-099-
29-183-
29-183-
29-189-
29-189-
29-189-
29-189-
29-189-
29-510-
29-510-
29-510-
1001
0008
1009
2007
3007
0010
0012
1002
1004
0004
0006
3001
5001
7003
0007
0072
0086
Figure AX3-31.
    Locations of O3 sampling sites (a) by AQS ID# (b) and intersite
    correlation statistics (c) for the St. Louis, MO-IL MSA.  The mean
    observed O3 concentration at each site is given above its letter code.
    For each data pair, the Pearson correlation coefficient, 90th percentile
    difference in absolute concentrations, the coefficient of divergence, and
    number of observations are given.
August 2005
                        AX3-46
                             DRAFT-DO NOT QUOTE OR CITE

-------
                  c.
                                              St.  Louis, MO  -  IL CMSA
Site
mean
(ppm)
A B
0.037 0.035
C
0.031
D
0.029
E
0.031
F G
0.028 0.035
H 1
0.034 0.035
J K
0.032 0.031
L
0.025
M N
0.030 0.030
0
0.028
P Q
0.024 0.030
                   M
                   O
                   Q
                         1    0.88  0.84  0.84   0.85   0.83  0.81   0.89   0.91   0.82  0.84  0.80  0.86  0.85  0.80   0.82   0.85
                             0.015 0.020  0.022  0.021  0.025 0.021 0.016 0.014 0.022  0.021  0.029  0.022 0.021 0.026 0.027 0.022
                              0.19  0.26  0.29   0.29   0.38  0.28  0.21   0.20   0.28  0.29  0.37  0.31  0.32  0.36   0.40   0.34
                             5055  4935  5032  5021   5045 5049 5052  4735  5077  5073  5075  5077  5071  5035  5034 5065
                               1
                                   0.87  0.90   0.93   0.86  0.82   0.93   0.90   0.83  0.85  0.84  0.87  0.84  0.84   0.87   0.88
                                   0.017  0.016  0.013 0.021 0.019 0.012 0.014 0.020  0.018  0.024  0.019 0.020 0.022 0.022 0.018
                                   0.26  0.26   0.23   0.34  0.28   0.18   0.19   0.27  0.27  0.34  0.30  0.30  0.33   0.35   0.31
                                   4963  5061  5048   5077 5078  5082  4766  5106  5102  5105  5106  5101  5064 5064 5094
                                     1
                                         0.92   0.88   0.92  0.87   0.90   0.88   0.88  0.87  0.85  0.90  0.88  0.86   0.88   0.91
                                         0.012  0.015 0.014 0.018 0.015 0.016 0.016  0.016  0.020  0.015 0.016 0.018 0.017 0.014
                                         0.24   0.26   0.29  0.26   0.23   0.24   0.25  0.26  0.32  0.26  0.27  0.30   0.32   0.27
                                         4941  4933   4952 4958  4963  4680  4985  4981  4983  4985  4979  4943  4942  4973

                                           1     0.93   0.90  0.85   0.91   0.88   0.86  0.84  0.84  0.87  0.85  0.86   0.88   0.90
                                               0.012 0.014 0.019 0.016 0.017 0.018  0.018  0.020  0.016 0.017 0.017 0.016 0.015
                                                0.22   0.29  0.26   0.23   0.25   0.27  0.27  0.30  0.27  0.28  0.28   0.29   0.27
                                               5024   5050 5055  5058  4742  5083  5079  5081  5085  5077  5042  5040  5073
                                                 1
               Key
Pearson correlation coefficient
90th %-ile difference in concentration
Coefficient of divergence
Number of paired observations
 0.90  0.84  0.93  0.90  0.85  0.84  0.88   0.88   0.86   0.88  0.89  0.90
0.015 0.019 0.013  0.015  0.018 0.018 0.018 0.016 0.017  0.016  0.017  0.015
 0.29  0.27  0.22  0.23  0.27  0.29  0.32   0.29   0.29   0.29  0.30  0.28
 5039  5045  5046  4728  5069  5065 5067  5069   5065  5027  5027  5057

  1   090  090  089  091  088  090   092   091   092  092  094
      0.018 0.018  0.018  0.016 0.017 0.016 0.014 0.015  0.013  0.015  0.012
      0.28  0.32  0.32  0.29  0.30  0.32   0.27   0.27   0.26  0.28  0.23
      5067  5070  4754  5095  5091 5093  5095   5091  5053  5053  5083

        1   0.87  0.86  0.95  0.91  0.88   0.88   0.88   0.89  0.86  0.91
            0.017  0.017  0.012 0.015 0.022 0.018 0.017  0.020  0.023  0.016
            0.23  0.24  0.22  0.25  0.33   0.27   0.26   0.29  0.35  0.26
            5075  4756  5100  5096 5098  5100   5094  5058  5057  5088

              1    0.94  0.87  0.88  0.87   0.92   0.90   0.86  0.87  0.91
                  0.011  0.017 0.016 0.022 0.015 0.015  0.020  0.022  0.015
                  0.15  0.24  0.24  0.32   0.26   0.27   0.31  0.35  0.29
                  4759  5103  5099 5101  5103   5097  5062  5060  5091

                    1    0.86  0.90  0.88   0.91   0.90   0.85  0.86  0.90
                        0.017 0.015 0.022 0.016 0.016  0.021  0.022  0.017
                        0.28  0.25  0.33   0.27   0.28   0.32  0.35  0.30
                        4784  4782 4782  4784   4778  4742  4743  4773

                          1    0.92  0.91   0.91   0.91   0.90  0.88  0.93
                             0.013 0.018 0.015 0.014  0.016  0.019  0.013
                              0.26  0.30   0.24   0.23   0.25  0.30  0.22
                              5124 5126  5128   5122  5086  5085  5116

                                1    0.87   0.90   0.91   0.86  0.87  0.90
                                   0.019 0.015 0.014  0.018  0.020  0.015
                                    0.29   0.23   0.23   0.28  0.30  0.25
                                   5122  5124   5118  5082  5081  5112

                                     1    0.88   0.90   0.90  0.88  0.92
                                         0.018 0.017  0.015  0.016  0.016
                                          0.28   0.28   0.27  0.27  0.26
                                         5126   5120  5084  5083  5114

                                           1    0.96   0.87  0.87  0.95
                                               0.009  0.018  0.019  0.011
                                                0.18   0.26  0.29  0.21
                                                5122  5086  5085  5116
                                                 1    0.87  0.86  0.94
                                                     0.017  0.019  0.011
                                                      0.26  0.31  0.20
                                                     5080  5079  5110

                                                       1   0.92  0.92
                                                           0.014  0.014
                                                           0.26  0.25
                                                           5045  5074

                                                             1    0.92
                                                                 0.016
                                                                 0.28
                                                                 5074

                                                                   1
Figure AX3-31  (cont'd).
August 2005
                                      AX3-47          DRAFT-DO NOT QUOTE OR CITE

-------
                                      Phoenix - Mesa, AZ MSA
                              50     0
                                                            2QO
                                          Kilometers
                             c.
                              Site  A   B
                                             H    I
                              mean
                              (ppm)
             0.021 0.025  0.028 0.046  0.021 0.024  0.024 0.028  0.038 0.041 0.024
                                    1   0.87  0.91  0.59  0.93   0.89  0.91  0.90  0.78  078  0.95
                                       0.017 0.018 0.046 0.012 0.016 0.014 0.017 0034 0.037 0.011
                                       0.38  0.35  0.59  0.31   0.35  0.32  0.34  0.54  055  0.27
                                       7580  8031  7991 7925  7593  7901  6651 8020 8081  7920
                                        1    0.90  0.69  0.89   0.92  0.89  0.93  0.83  085  0.90
                                           0.018 0.038 0.016 0.013 0.015 0.014 0027 0.029 0.016
                                            0.25  0.46  0.41   0.33  0.30  0.22  0.38  041  0.32
                                            7962  7912 7872  7502  7826 6768 7895 8011  7858

                                             1   0.62  0.92   0.92  0.91  0.92  0.82  082  0.95
                                                0.040 0.019 0.016 0.017 0.016 0027 0.030 0.013
                                                0.47  0.40   0.34  0.28  0.24  0.39  041  0.28
                                                8377 8326  7965  8273 7022 8367 8468  8319

                                                 1    058   063  058  061  067  079  058
                                                    0.047 0.043 0.042 0.043 0025 0.018 0.045
                                                     0.61   0.56  0.51  0.49  0.23  016  0.56
                                                    8286  7925  8231  6992 8332 8429  8281

                                                      1    0.92  0.94  0.93  0.82  080  0.95
                                                         0.014 0.012 0.015 0033 0.037 0.012
                                                          035  035  037  057  058  031
                                                         7894  8180 6938 8279 8369  8238
                                                           1   0.90  0.94  0.82  084  0.93
                                                              0.015 0.012 0030 0.033 0.013
                                                              0.35  0.30  0.50  052  0.31
                                                              7863 6599 7913 8009  7846

                                                               1   0.93  0.81  080  0.93
                                                                  0.014 0029 0.033 0.013
                                                                  0.25  0.44  046  0.29
                                                                  6882 8225 8322  8166

                                                                   1   0.82  083  0.93
                                                                      0029 0032 0013
                                                                       0.41  043  0.25
                                                                      6963 7078  6912

                                                                        1   086  0.81
                                                                           0.016 0.031
                                                                           017  0.50
                                                                           8411  8265

                                                                            1    0.80
                                                                               0.035
                                                                                0.51
                                                                                8367
                                                                                 1
                        Key
              Pearson correlation coefficient
              90th %-ile difference in concentration
              Coefficient of divergence
              Number of paired observations
Figure AX3-32.
Locations of O3 sampling sites (a) by AQS ID# (b) and intersite
correlation statistics (c) for the Phoenix-Mesa, AZ MSA. The mean
observed O3 concentration at each site is given above its letter code.
For each data pair, the Pearson correlation coefficient, 90th percentile
difference in absolute concentrations, the coefficient of divergence,
and number of observations  are given.
August 2005
                              AX3-48
DRAFT-DO NOT QUOTE OR CITE

-------
                                         Fresno, CA MSA
                                                         c.
Figure AX3-33.
Site
mean
(ppm)
A



B



A B
0.031 0.031
1 0.97
0.010
0.19
7804
1



C
0.047
0.94
0.027
0.34
8124
0.95
0.026
0.40
7777
D
0.040
0.93
0.022
0.25
7950
0.92
0.022
0.33
7624
E
0.036
0.95
0.016
0.19
8066
0.97
0.013
0.23
7726
F
0.030
0.92
0.015
0.25
8055
0.92
0.017
0.25
7710
                                                                      1   0.92  0.93  0.94
                                                                         0.019 0.022  0.027
                                                                         0.20  0.29  0.40
                                                                         7939 8055  8036
50 0 Kilometers 200 D
i i i i i i i i i i
b.

AIRS Site ID
Site A 06-019-0007
SiteB 06-019-0008
SiteC 06-019-0242
SiteD 06-019-4001
SiteE 06-019-5001
Site F 06-039-0004
E
F
1 0.92 0.9;
0.017 0.02
0.22 0.3^
7876 787
1 0.9C
0.02
0.2£
799
1
Key
Pearson correlation coefficient
90th %-ile difference in concentration
Coefficient of divergence
Number of paired observations

                       Locations of O3 sampling sites (a) by AQS ID# (b) and intersite
                       correlation statistics (c) for the Fresno, CA MSA. The mean observed O3
                       concentration at each site is given above its letter code. For each data
                       pair, the Pearson correlation coefficient, 90th percentile difference in
                       absolute concentrations, the coefficient of divergence, and number of
                       observations are given.
1      cases where sites in an urban area may be moderately to highly correlated but showed substantial
2      differences in absolute concentrations. In many cases, values for P90 equaled or exceeded
3      seasonal mean O3 concentrations.  This was reflected in both values for P90 and for the COD.
4           It is instructive to compare the metrics for spatial variability shown in Table AX3-5 to
5      those calculated for PM2 5 and PM10_2 5 in the PM AQCD (U.S. Environmental Protection
6      Agency, 2004a). The values for concentrations and concentration differences are unique to the
7      individual species, but the intersite correlation coefficients and the COD values can be directly
8      compared.
      August 2005
                                        AX3-49
DRAFT-DO NOT QUOTE OR CITE

-------
                                        Bakersfield, CA MSA
             a.
             b.
                                         ®
                           Kilometers
                         AIRS Site ID
                     Site A 06-029-0007
                     Site B 06-029-0008
                     SiteC 06-029-0010
                     SiteD 06-029-0011
                     SiteE 06-029-0014
                     Site F 06-029-0232
                     SiteG 06-029-5001
                     Site H 06-029-6001
                             C.
                                                200
Site
mean
(ppm)
A


B


A B
0.039 0.047
1 0.87
0.026
0.28
8299
1


C
0.029
0.82
0.030
0.40
7835
0.76
0.042
0.47
7779
D
0.047
0.62
0.038
0.35
8127
0.68
0.028
0.28
8079
E
0.030
0.80
0.032
0.45
8307
0.75
0.044
0.51
8251
F
0.038
0.88
0.019
0.24
7915
0.87
0.025
0.29
7880
G
0.043
0.87
0.022
0.24
8283
0.89
0.021
0.20
8244
H
0.030
0.80
0.031
0.40
8061
0.75
0.042
0.47
8006
                              D
                                               0.62  0.95  0.88  0.84  0.94
                                              0.047 0.014 0.026 0.032 0.015
                                               0.48  0.30  0.37  0.43  0.29
                                              7612  7791  7469  7772  7549
                                                                        1
             0.62  0.69  0.68  0.61
             0.047 0.033 0.032 0.047
             0.49  0.34  0.30  0.46
             8096  7695  8066 7903
                                                    1
                                             Key
                                  Pearson correlation coefficient
                                  90th %-ile difference in concentration
                                  Coefficient of divergence
                                  Number of paired observations
                  0.86  0.84 0.96
                 0.028 0.033 0.013
                  0.41  0.46 0.28
                  7862  8244 8024

                   1   0.89 0.85
                      0.021 0.028
                      0.25 0.37
                      7843 7626

                       1   0.84
                          0.033
                          0.42
                          8005
                                                                                          1
       Figure AX3-34.
Locations of O3 sampling sites (a) by AQS ID# (b) and intersite
correlation statistics (c) for the Bakersfield, CA MSA. The mean
observed O3 concentration at each site is given above its letter code.
For each data pair, the Pearson correlation coefficient, 90th percentile
difference in absolute concentrations, the coefficient of divergence, and
number of observations are given.
1      In general, the variability in O3 concentrations is larger than for PM2 5 concentrations and

2      comparable to that obtained for PM10_2 5. Intersite correlation coefficients in some areas (e.g.,

3      Philadelphia, PA; Atlanta, GA; Portland, OR) can be very similar for both PM2 5 and for O3.

4      However, there is much greater variability in the concentration fields of O3 as evidenced by the

5      much higher COD values. Indeed, COD values are higher for O3 than for PM25 in each of the

6      urban areas examined.  In all of the urban areas  examined for O3 some site pairs are always very
       August 2005
                        AX3-50
DRAFT-DO NOT QUOTE OR CITE

-------
           a.
      Los Angeles - Orange County, CA CMSA
                                       b.
           50
            I	
       0
        I
 100
	I
                           Kilometers
                                                         AIRS Site ID
Site A
SiteB
SiteC
SiteD
SiteE
SiteF
SiteG
SiteH
Site I
Sited
SiteK
SiteL
SiteM
SiteN
06-037-
06-037-
06-037-
06-037-
06-037-
06-037-
06-037-
06-037-
06-037-
06-037-
06-037-
06-037-
06-037-
06-037-
0002
0016
•0030
0113
1002
1103
1201
1301
•1601
•1701
2005
4002
5001
9002
Figure AX3-35.
Locations of O3 sampling sites (a) by AQS ID# (b) and intersite
correlation statistics (c) for the Los Angeles-Orange County, CA CMSA.
The mean observed O3 concentration at each site is given above its letter
code.  For each data pair, the Pearson correlation coefficient, 90th
percentile difference in absolute concentrations, the coefficient of
divergence, and number of observations are given.
August 2005
                      AX3-51
     DRAFT-DO NOT QUOTE OR CITE

-------
                         Los Angeles  - Orange County,  CA CMSA
                 c.
Site
mean
(ppm)
A





P



A B
0.021 0.029

1 095
0017
032
8378
1






C
0.021

087
0020
033
6723
082
0029
036
6722
1



D
0.020

064
0028
041
8387
060
0034
042
8386
0.75
0.021
0.32
6731
E
0.018

088
0018
038
8202
086
0028
044
8201
0.88
0.018
0.35
6546
F
0.019

085
0020
036
8292
080
0029
045
8291
0.95
0.010
0.22
6636
G H
0.021 0.013

0 82 0 72
0021 0029
0 37 0 44
8364 8362
0 80 0 67
0027 0038
0 42 0 53
8363 8361
0.84 0.83
0.019 0.020
0.29 0.37
6708 6729
I
0.016

089
0019
0 41
8381
084
0030
052
8380
0.91
0.013
0.37
6725
J
0.013

091
0022
046
8386
088
0032
059
8385
0.85
0.020
0.49
6730
K
0.021

091
0015
033
8384
087
0024
039
8383
0.90
0.017
0.24
6728
L
0.022

061
0029
044
8379
057
0033
047
8378
0.72
0.022
0.31
6723
M N
0.027 0.042

0 55 0 58
0035 0048
0 45 0 47
8341 6321
0 53 0 56
0034 0042
0 46 0 42
8340 6321
0.68 0.61
0.030 0.048
0.37 0.45
6686 4998
                 H
                 K
                 M
                 N
                                             0.76   0.82  0.73  0.77   0.77  0.69  0.73  0.75  0.81  0.62
                                            0.023 0.018 0.023  0.023 0.022  0.027 0.024 0.021  0.025 0.046
                                             0.31   0.31  0.31  0.36   0.35  0.43  0.34  0.31  0.30  0.40
                                            8210  8300 8372  8370  8389  8394 8392  8387  8349  6321

                                              1    0.90  0.86  0.79   0.90  0.87  0.91  0.67  0.66  0.59
                                                  0.014 0.019  0.023 0.015  0.019 0.014 0.027  0.034 0.050
                                                   0.29  0.31  0.35   0.33  0.38  0.28  0.37  0.37  0.45
                                                  8184 8187  8185  8204  8210 8207  8202  8164  6322
                                                    1
                                  0.85  0.85   0.93  0.86  0.91  0.76  0.74  0.66
                                 0.018  0.019 0.012  0.019 0.015 0.021  0.029 0.047
                                  0.31  0.36   0.33  0.42  0.26  0.35  0.40  0.50
                                 8277  8275  8294  8300 8297  8292  8254  6325

                                   1    0.74   0.86  0.84  0.86  0.67  0.68  0.70
                                       0.026 0.020  0.023 0.017 0.026  0.028 0.042
                                       0.36   0.35  0.45  0.31  0.36  0.36  0.43
                                       8347  8366  8371 8369  8364  8349  6326

                                        1    0.86  0.74  0.80  0.84  0.73  0.61
                                            0.017  0.018 0.025 0.021  0.033 0.053
                                             0.26  0.37  0.38  0.35  0.43  0.51
                                            8364  8369 8367  8362  8324  6299

                                              1    0.89  0.93  0.77  0.71  0.66
                                                  0.016 0.015 0.022  0.032 0.048
                                                  0.31  0.30  0.34  0.39  0.48
                                                  8388 8386  8381  8343  6321

                                                    1    0.90  0.67  0.62  0.63
                                                       0.021  0.028  0.037 0.053
                                                        0.43  0.46  0.50  0.56
                                                       8391  8386  8348  6323
                                                                                1
              Key
Pearson correlation coefficient
90th %-ile difference in concentration
Coefficient of divergence
Number of paired observations
             0.71  0.65   0.65
            0.025  0.031  0.045
             0.37  0.40   0.46
            8384  8346  6321

              1    0.74   0.62
                  0.025 0.043
                  0.33   0.42
                  8341  6321

                   1    0.63
                       0.038
                        0.36
                       6299

                         1
Figure AX3-35 (cont'd).
August 2005
                               AX3-52
DRAFT-DO NOT QUOTE OR CITE

-------
                   Riverside - Orange County, CS CMSA
      a.
                                         b.
           50
            i i
   0
   I
                         Kilometers
 200
	i
AIRS Site ID
Site A
SiteB
SiteC
SiteD
SiteE
SiteF
SiteG
SiteH
Sitel
Site J
SiteK
SiteL
SiteM
SiteN
SiteO
SiteP
SiteQ
SiteR
06-065-0012
06-065-2002
06-065-5001
06-065-6001
06-065-8001
06-065-9001
06-071-0001
06-071-0005
06-071-0012
06-071-0017
06-071-0306
06-071-1004
06-071-1234
06-071-2002
06-071-4001
06-071-4003
06-071-9002
06-071-9004
Figure AX3-36.
Locations of O3 sampling sites (a) by AQS ID# (b) and intersite
correlation statistics (c) for the Riverside-Orange County, CA CMSA.
The mean observed O3 concentration at each site is given above its letter
code.  For each data pair, the Pearson correlation coefficient, 90th
percentile difference in absolute concentrations, the coefficient of
divergence, and number of observations are given.
August 2005
                     AX3-53
    DRAFT-DO NOT QUOTE OR CITE

-------
                           c.
                           Site
                                    Riverside -  Orange  County,  CS CMSA
 ABODE    FGH    I     JKLMNOPQR
                          mean
                          (ppm)
0.037 0.035 0.042 0.033 0.026 0.036 0.030 0.047 0.048 0.044 0.032 0.022 0.038 0.020 0.037 0.032 0.040 0.026
                                      0.71  0.77  0.84  0.79  0.81  0.76  0.70  0.69  0.69  0.75 0.76 0.55 0.75 0.76  0.81  0.53  0.80
                                     0.032 0.030 0.026 0.034 0.027 0.031 0.038 0.037 0.034 0.031 0.037 0.036 0.040 0.029 0.028 0.037 0.033
                                      0.33  0.35  0.34  0.42  0.33  0.38  0.38  0.39  0.37  0.36 0.47 0.38 0.54 0.35  0.35  0.38  0.41
                                      8397  8392 8394 8386  8035 7741  8393  8051  7844  7498 8385 7815 8394 7997 8391 7887 8364
                                       1   0.85  0.69  0.64  0.66  0.75  0.62  0.64  0.81  0.72 0.60 0.65 0.56 0.71  0.61  0.52  0.61
                                          0.023 0.035 0.039 0.036 0.031 0.041  0.038 0.029 0.031 0.043 0.031 0.046 0.031 0.038 0.035 0.040
                                           0.25  0.34  0.42  0.33  0.33  0.33  0.34  0.30  0.34 0.46 0.31 0.51 0.31  0.37  0.33  0.42
                                           8394 8395 8388  8040 7919  8393  8233  8017  7678 8389 7990 8395 8179 8392 8059 8369
                                            1
                            M
                            O
                0.73  0.66  0.69  0.74  0.67  0.72  0.81  0.73  0.61  0.67  0.59 0.75 0.65 0.65 0.65
               0.037 0.044 0.036 0.035 0.036 0.030 0.025 0.034 0.049 0.030 0.052 0.030 0.039 0.031 0.045
                0.38  0.47  0.35  0.39  0.32  0.32  0.28  0.38  0.52  0.30  0.56 0.32 0.40 0.33 0.46
               8393 8385  8034 7752 8391 8059 7852  7506  8384  7823  8393 8005 8390 7896 8362

                 1   0.90  0.91  0.79  0.70  0.68  0.67  0.77  0.87  0.57  0.84 0.75 0.87 0.47 0.89
                     0.023 0.019 0.029 0.041 0.041 0.039 0.031  0.028 0.039 0.032 0.033 0.023 0.040 0.024
                     0.33  0.25  0.34  0.41  0.45  0.42  0.37  0.42  0.41  0.51 0.38 0.33 0.44 0.35
                     8387  8036 7742 8393 8049 7842  7496  8386  7813  8395 7995 8392 7885 8364

                       1    0.88  0.76  0.66  0.63  0.60  0.74  0.92  0.50  0.93 0.69 0.87 0.42 0.94
                          0.026 0.031 0.048 0.046 0.045 0.032 0.017 0.042 0.018 0.037 0.025 0.042 0.015
                           0.36  0.38  0.51  0.54  0.51  0.42  0.34  0.50  0.43 0.46 0.39 0.52 0.36
                           8028 7735 8385 8042 7835  7490  8378  7806  8387 7988 8384 7878 8356

                            1    0.76  0.64  0.62  0.63  0.74  0.83  0.54  0.81 0.70 0.82 0.44 0.85
                               0.031 0.042 0.039 0.037 0.031  0.032 0.037 0.034 0.032 0.026 0.039 0.028
                                0.33  0.37  0.39  0.36  0.37  0.43  0.36  0.53 0.35 0.31 0.38 0.35
                               7402 8034 7709 7487  7171  8027  7610  8036 7640 8033 7545 8010

                                 1   0.68  0.68  0.75  0.85  0.72  0.68  0.68 0.78 0.72 0.46 0.73
                                    0.042 0.041 0.035 0.022 0.035 0.033 0.038 0.029 0.032 0.038 0.033
                                     0.40  0.43  0.40  0.33  0.41  0.39  0.49 0.36 0.37 0.41 0.40
                                     7740 8067 7758  7684  7740  7669  7744 7861 7739 7592 7720

                                      1   0.72  0.65  0.74  0.66  0.59  0.65 0.80 0.71 0.55 0.70
                                         0.030 0.033 0.038 0.051  0.037 0.054 0.030 0.041 0.036 0.047
                                          0.26  0.26  0.40  0.56  0.27  0.59 0.31 0.39 0.28 0.46
                                         8047 7840  7494  8384  7811  8393 7993 8390 7883 8363

                                           1   0.73  0.70  0.63  0.64  0.60 0.78 0.68 0.63 0.65
                                              0.022 0.039 0.049 0.027 0.052 0.030 0.040 0.024 0.046
                                               0.19  0.43  0.59  0.24  0.61 0.34 0.42 0.16 0.48
                                              7983  7836  8047  7940  8051 8059 8046 7899 8027

                                                1   0.72  0.57  0.73  0.54 0.74 0.60 0.57 0.58
                                                    0.035 0.048 0.024 0.051 0.028 0.041 0.025 0.046
                                                    0.41  0.56  0.21  0.59 0.32 0.39 0.21 0.45
                                                    7576  7840  7677  7844 7826 7839 7689 7820

                                                     1   0.70  0.64  0.67 0.84 0.70 0.49 0.72
                                                         0.036 0.032 0.039 0.024 0.033 0.037 0.034
                                                         0.44  0.40  0.51 0.34 0.41 0.42 0.42
                                                         7494  7430  7498 7626 7493 7352 7477

                                                          1   0.46  0.95 0.68 0.88 0.38 0.92
                                                              0.044 0.013 0.041 0.027 0.045 0.018
                                                              0.54  0.40 0.50 0.43 0.55 0.40
                                                              7814  8386 7993 8383 7885 8357

                                                                1   0.42 0.65 0.49 0.50 0.49
                                                                   0.048 0.028 0.039 0.023 0.042
                                                                   0.57 0.31 0.40 0.22 0.45
                                                                   7815 7817 7811 7665 7796

                                                                     1   0.64 0.87 0.38 0.93
                                                                        0.0430.0290.0470.019
                                                                        0.54 0.50 0.58 0.45
                                                                        7996 8392 7887 8364

                                                                          1   0.72 0.58 0.70
                                                                             0.033 0.030 0.037
                                                                             0.37 0.33 0.43
                                                                             7992 7842 7973

                                                                               1   0.46 0.92
                                                                                  0.039 0.020
                                                                                  0.41 0.33
                                                                                  7882 8361

               	.                                         1   0.45
                                                                                       0.041
                                                                                       0.46
                                                                                       7866
                                                     Key
                                   Pearson correlation coefficient
                                   90th %-ile difference in concentration
                                   Coefficient of divergence
                                   Number of paired observations
Figure AX3-36  (cont'd).
August 2005
                                     AX3-54
DRAFT-DO NOT  QUOTE OR CITE

-------
 1     highly correlated with each other (i.e., r > 0.9) as seen for PM2 5. These sites also show less
 2     variability in concentration and are probably influenced most strongly by regional production
 3     mechanisms.
 4
 5     AX3.3.2  Small-scale Horizontal and Spatial Variability in Ozone
 6                Concentrations
 7     Ozone concentrations near roadways
 8          Apart from the larger scale variability in surface O3 concentrations, there is also significant
 9     variability on the micro-scale (< a few hundred meters), especially near roadways and other
10     sources of emissions that react with O3. These sources are not confined to urban areas.  Sources
11     of emissions that react with O3 such as highways and power plants are also found in rural areas.
12     Johnson (1995) described the results of studies examining O3 upwind and downwind of
13     roadways in Cincinnati, OH.  In these studies, O3 upwind of the roadway was about 50 ppb and
14     these values were not found again until distances of about 100 m downwind. The O3 profile
15     varied inversely with that of NO, as might be expected. For peak NO concentrations of 30 ppb,
16     the O3 mixing ratio was about 36 ppb, or about 70% of the upwind value. The magnitude of the
17     downwind depletion of O3 depends on the emissions of NO, the rate of mixing of NO from the
18     roadway and ambient temperature and so depletions of O3 downwind of roadways are expected,
19     but with variable magnitude.  Guidance for the placement of O3 monitors (U.S. Environmental
20     Protection Agency, 1998) states a separation distance that depends on traffic counts.  For
21     example, a minimum separation distance of 100 m from a road with 70,000 vehicles per day is
22     recommended for siting an O3 monitor to avoid interference that would mean a site is no longer
23     representative of the surrounding area. An average rate of about 3,000 vehicles per hour passing
24     by a monitoring site implies a road with rather heavy traffic.  As noted earlier in Section
25     AX3.3.1 for the Lakewood, CA monitoring, O3 levels are lower at sites located near traffic than
26     those located some distance away and the scavenging of O3 by emissions of NO from roadways
27     is a major source of spatial variability in O3 concentrations.  It should also be noted that
28     scavenging of O3 by NO near roadways was more pronounced before the implementation of
29     stringent NOx emissions controls.
30
       August 2005                            AX3-55      DRAFT-DO NOT QUOTE OR CITE

-------
 1      Vertical Variations in Ozone Concentrations
 2           In addition to horizontal variability in O3 concentrations, there are also variations in the
 3      vertical profile of O3 in the lowest layers of the atmosphere to consider. The planetary boundary
 4      layer consists of an outer and an inner portion.  The inner part of the planetary boundary layer
 5      extends from the surface to about one-tenth the height of the planetary boundary layer. Winds
 6      and transported properties, such as O3, are especially susceptible to interactions with obstacles,
 7      such as buildings and trees in the inner boundary layer (atmospheric surface layer) (e.g., Garratt,
 8      1992).  Inlets to ambient monitors (typically at heights of 3 to 5  meters) are located in, and
 9      human and vegetation exposures occur in this part of the boundary layer.
10           Photochemical production and destruction of O3 occurs throughout the planetary boundary
11      layer.  However, O3 is also destroyed on the surfaces of buildings, vegetation, etc.  On most
12      surfaces, O3 is destroyed with every collision. In addition, O3 is scavenged by NO emitted by
13      motor vehicles and soils. These losses imply that the vertical gradient of O3 should always be
14      directed downward.  The magnitude of the gradient is determined by the intensity of turbulent
15      mixing in the  surface layer.
16           Most work characterizing the vertical profile of O3 near the surface has been performed in
17      nonurban areas with the aim of calculating fluxes of O3 and other pollutants through forest
18      canopies and to crops and short vegetation etc.  Corresponding data are sparse for urban areas.
19      However, monitoring sites are often set up in open areas such as parks and playgrounds where
20      surface characteristics may be more similar to those in rural areas than to those in the
21      surrounding urban area. The vertical profile of O3 measured over low vegetation are shown in
22      Figure AX3-37. These measurements were obtained as part of a field campaign to measure the
23      fluxes of several gas and aerosol phase pollutants using the gradient-flux technique in a remote
24      area in Hortobagy National Park in Hungary during late spring of 1994 (Horvath et al., 1995).
25      The labels stable and unstable in the figure refer to atmospheric stability conditions and average
26      represents the overall average.  Ozone concentrations were normalized to their values  at 4 m
27      height. As can be seen from the figure, there was a decrease of about 20% in going from a
28      height of 4 m  down to 0.5 m above the surface during stable conditions, but O3 decreased by
29      only about 7% during unstable conditions. The average decrease was about 10% for all
30      measurements. As might be expected, O3 concentrations at all heights were very highly
31      correlated with one another.  Of course, these values represent averages and there is scatter about

        August 2005                             AX3-56       DRAFT-DO NOT QUOTE OR CITE

-------
                     4-i
               +•»
               .c
               O)
               "i
               I
 1 -
0.5-

 0
                          ^ Average  (n = 1797)
                          + Stable   (n= 937)
                          • Unstable (n = 860)
                                                               x
                          y    4
                                                           Y
                      0.5
                                   0.6          0.7           0.8
                                           Relative O3 Concentration
                                                                         0.9
       Figure AX3-37.   Vertical profile of O3 obtained over low vegetation.  Values shown are
                         relative to concentrations at 4 m above the surface.  Ozone
                         concentrations for unstable and unstable conditions were 41.3 and 24.1
                         ppb, and average O3 concentration weighted by stability class was 33.1
                         ppb at 4 m.
       Source. Horvathetal. (1995).
 1     them, particularly under strong stable conditions. However, these conditions tend to occur
 2     mainly during night and the stability regime during the day in urban areas tends more towards
 3     instability because of the urban heat island effect.  Figure AX3-38 shows the vertical profile
 4     of O3 obtained in a spruce forest in northwestern Hungary in late summer 1991 by the same
 5     group (Horvath et al., 2003). The fall off of O3 in this case is due to uptake by trees, reaction
 6     with ambient NO and with NO emitted by the soil in the forest in addition to deposition on the
 7     surface.
 8
 9     AX3.3.3  Ozone Concentrations at High Elevations
10           The distributions of hourly average concentrations experienced at high-elevation cities are
11     similar to those experienced in low-elevation cities. For example, the distribution of hourly
12     average concentrations for several O3 sites located in Denver were similar to distributions
       August 2005
AX3-57
                                         DRAFT-DO NOT QUOTE OR CITE

-------
                    19-,
                    15-
               +*   10-
               O)
               "i
               I
                    2,5-
                    0.5-
                          ^ Average  (n = 1760)
                          + Stable   (n = 1272)
                          • Unstable  (n =  488)
                       0.5
                                   0.6           0.7           0.8           0.9
                                            Relative O3 Concentration
       Figure AX3-38.   Vertical profile of O3 obtained in a spruce forest. Values shown are
                         relative to concentrations at 19 m above the surface. Mean tree height is
                         14.5 m.  Ozone concentrations for unstable and unstable conditions were
                         36.7 and 33.8 ppb, and the average O3 concentration weighted by stability
                         class was 34.6 ppb at 19 m.
       Source: Horvath et al. (2003).
 1      observed at many low-elevation sites elsewhere in the United States. However, the use of
 2      absolute concentrations (e.g., in units of micrograms per cubic meter) in assessing the possible
 3      impacts of O3 on vegetation at high-elevation sites instead of mixing ratios (e.g., parts per
 4      million) may be an important consideration (see Chapter 9, for further considerations about
 5      exposure and effective dose considerations for vegetation assessments).
 6           Concentrations of O3 vary with altitude and latitude. Although a number of reports contain
 7      data on O3 concentrations at high altitudes (e.g., Coffey et al., 1977; Reiter, 1977; Singh et al.,
 8      1977; Evans  et al., 1985; Lefohn and Jones, 1986), fewer reports present data for different
 9      elevations in the same locality. Monitoring data collected by the Mountain Cloud Chemistry
10      (MCCP) provide useful information for investigating O3 exposure differences at different
11      elevations. When applying different exposure indices to the MCCP data, there appears to be no
12      consistent conclusion concerning the relationship between O3 exposure and elevation.
       August 2005
AX3-58
DRAFT-DO NOT QUOTE OR CITE

-------
 1          Lefohn et al. (1990a) summarized the characterization of gaseous exposures at rural sites in
 2      1986 and 1987 at several MCCP high-elevation sites. Aneja and Li (1992) have summarized
 3      the O3 concentrations for 1986 to 1988. Table AX3-6 summarizes the sites characterized by
 4      Lefohn et al. (1990a).  Table AX3-7 summarizes the concentrations and exposures that occurred
 5      at several of the sites for the period 1987 to 1988. In 1987, the 7- and 12-h seasonal means were
 6      similar at the Whiteface Mountain WF1 and WF3 sites (Figure AX3-39a).  The 7-h mean values
 7      were 0.0449 and 0.0444 ppm, respectively, and the 12-h mean values were 0.0454 and
 8      0.0444 ppm, respectively. Note that, in some cases, the 12-h mean was slightly higher than the
 9      7-h mean value.  This resulted when the 7-h mean period (0900 to 1559 hours) did not capture
10      the period of the day when the highest hourly mean O3 concentrations were experienced.
11      A similar observation was made, using the 1987 data, for the MCCP Shenandoah National Park
12      sites. The 7-h and 12-h seasonal means were similar for the SHI and SH2 sites (Figure
13      AX3-39b). Based on cumulative indices, the Whiteface Mountain summit (1483-m) site (WF1)
14      experienced a higher exposure than the WF3 (1026-m) site (Figure AX3-39c). Both the sum of
15      the were higher at the WF1 site than at the WF3 site.  The site at the base of the mountain (WF4)
16      experienced the lowest exposure of the three O3 sites. Among the MCCP Shenandoah National
17      Park sites, the SH2 site experienced marginally higher O3 exposures, based on the index that
18      sums all of the hourly average concentrations (referred to as "total dose" in Figure AX3-39c) and
19      sigmoidal values, than the SHI high-elevation site (Figure AX3-39d). The reverse was true for
20      concentrations >0.07 ppm (SUM07) and the number of hourly concentrations >0.07 ppm the
21      sums of the concentrations >0.07 ppm and the number of hourly concentrations >0.07 ppm.
22      When the Big Meadows, Dickey Ridge, and Sawmill Run, Shenandoah National Park, data for
23      1983 to 1987 were compared, it again was found that the 7-h and 12-h seasonal means were
24      insensitive to the different O3 exposure patterns.  A better resolution of the differences was
25      observed when the cumulative indices were used (Figure AX3-40).  There was no evidence that
26      the highest elevation site, Big Meadows, consistently had experienced higher O3 exposures than
27      the other sites. In 2  of the 5 years, the Big Meadows site experienced lower exposures than the
28      Dickey Ridge and Sawmill Run sites, based on the sum of all concentration or sigmoidal indices.
29      For 4 of the  5 years, the SUM07 index yielded the same result.
30
       August 2005                            AX3-59      DRAFT-DO NOT QUOTE OR CITE

-------
Table AX3-6. Description of Mountain Cloud Chemistry Program Sites
V
£
to
o
o
X
OJ
ON
O
O
H
6
o
0
H
O
O
H
W
O
O
HH
H
W
Site
Rowland Forest (HF1), ME
Mt. Moosilauke (MSI), NH
Whiteface Mountain (WF1), NY
Shenandoah NP (SHI), VA
Shenandoah NP (SH2), VA
Shenandoah NP (SH3), VA
Whitetop Mountain (WT1), VA
Mt. Mitchell (MM 1),NC
Mt. Mitchell (MM2), NC

Elevation (m) Latitude
65 458° 11'
1000 438° 59' 18"
1483 448° 23' 26"
1015 38° 37' 12"
716 38° 37' 30"
524 38° 37' 45"
1689 36° 38' 20"
2006 35° 44' 15"
1760 35° 45'

Longitude
68° 46'
71° 48' 28"
73° 51' 34"
78° 20' 48"
78° 21' 13"
78° 21' 28"
81° 36' 21"
82° 17' 15"
82° 15'


-------
Table AX3-7.  Seasonal (April-October) Percentiles, SUM06, SUM08, and W126 Values for the MCCP Sites
_-
S-
to
o
o






X
1
l~"1

O
H
6
o
0
H
O
O
H
W
O
O
HH
H
W
Site
Howland Forest, ME
(HF1)
Mt. Moosilauke, NH
(MSI)
Whiteface Mountain,
NY (WF1)
(36-031-0002)
Whiteface Mountain,
NY (WF3)

Whiteface Mountain,
NY (WF4)

Mt. Mitchell, NC
(MM1)
Mt. Mitchell, NC
(MM2)
Shenandoah Park,
VA (SHI)
Shenandoah Park,
VA (SH2)
Year
1987
1988
1987
1988
1987
1988
1987


1987

1987
1988
1989
1992
1987
1988
1987
1988
1987*
1988
Min.
0.000
0.000
0.006
0.010
0.011
0.014
0.010


0.000

0.008
0.011
0.010
0.005
0.017
0.009
0.000
0.006
0.003
0.006
10
0.013
0.012
0.027
0.026
0.029
0.025
0.025


0.011

0.034
0.038
0.038
0.036
0.032
0.029
0.023
0.024
0.027
0.029
30
0.021
0.021
0.036
0.033
0.037
0.033
0.033


0.023

0.044
0.054
0.047
0.043
0.042
0.041
0.036
0.036
0.040
0.042
50
0.028
0.028
0.045
0.043
0.046
0.043
0.039


0.031

0.051
0.065
0.054
0.048
0.049
0.050
0.044
0.047
0.049
0.054
70
(ppm)
0.035
0.036
0.053
0.055
0.053
0.056
0.047


0.041

0.058
0.075
0.059
0.053
0.056
0.060
0.054
0.058
0.059
0.064
90
0.046
0.047
0.065
0.076
0.067
0.078
0.064


0.056

0.067
0.095
0.068
0.063
0.067
0.080
0.069
0.077
0.071
0.083
95
0.052
0.054
0.074
0.087
0.074
0.089
0.075


0.065

0.074
0.106
0.072
0.069
0.073
0.092
0.076
0.087
0.077
0.095
99
0.065
0.076
0.086
0.113
0.087
0.110
0.091


0.081

0.085
0.126
0.081
0.081
0.083
0.110
0.085
0.103
0.086
0.108
Max.
0.076
0.106
0.102
0.127
0.104
0.135
0.117


0.117

0.105
0.145
0.147
0.096
0.096
0.162
0.135
0.140
0.145
0.145
No.
Obs.
4766
4786
4077
2835
4703
4675
4755


4463

3539
2989
2788
3971
3118
2992
3636
3959
2908
4661
SUM06
5.9
10.9
45.0
51.9
63.5
94.4
45.4


23.8

59.4
145.1
54.8
37.8
47.0
68.7
54.2
80.9
55.7
133.8
SUM08
(ppm-h)
0.0
2.9
9.5
21.2
12.2
40.8
14.4


5.1

7.8
69.7
3.5
4.4
5.1
28.1
8.5
29.6
7.8
55.8
W126
7.7
11.6
40.1
43.4
50.5
78.3
40.3


21.3

46.5
116.6
40.7
36.7
37.4
57.7
42.0
67.2
41.8
109.4

-------
            Table AX3-7 (cont'd). Seasonal (April-October) Percentiles, SUM06, SUM08, and W126 Values for the MCCP Sites
_-
S-
to
o
o



Site
Shenandoah Park,
VA (SH3)
Whitetop Mountain,
VA (WT1)
Year
1987
1988
1987
1988
Min.
0.000
0.000
0.01
0.000
10
0.018
0.020
0.038
0.030
30
0.029
0.031
0.051
0.046
50
0.037
0.040
0.059
0.058
70
(ppm)
0.047
0.051
0.066
0.068
90
0.061
0.067
0.078
0.084
95
0.068
0.076
0.085
0.094
99
0.080
0.097
0.096
0.119
Max.
0.108
0.135
0.111
0.163
No.
Obs.
3030
4278
4326
3788
SUM06
23.1
52.3
147.7
133.8
SUM08
(ppm-h)
2.6
15.6
32.4
51.0
W126
19.2
44.2
105.7
102.8
X
OJ

ON
to
H

6
o


o
H

O

O
H
W

O


O
HH
H
W
        *Calculations based on a May-September season.

-------
              0.05,
              0.00
                                                       0.05,  b.
                    WF 1
                             WF3
                                      WF4
                                                       0.00
                                                             SH 1
                                                                      SH2
                                                                               SH3
              200 n C.
                                     • Total Dose
                                     • Sigmoidal
                                     D Sum 20.07 ppm
        200n
                                                     S  100
                                                     
-------
              a. May - September 1983
          300 -i
                                  Total Dose
             Big Meadows Dickey Ridge Sawmill Run
            c. June - September 1985
                                                  300 -i
                                D Sum >0.07 ppm   £•
                                               £ 200 -
                                               Q.
                                               Q.
                                                    0 -b
                              • Total Dose
                              GSigmoidal
                              DSum>0.07 ppm
            Big Meadows Dickey Ridge  Sawmill Run
              b. May - September 1984
          300 -i
                                  Total Dose
                                DSum>0.07 ppm   -r-
             d. May - September 1986
                                                   300 -i
                                                                          Total Dose
             Big Meadows Dickey Ridge Sawmill Run
             Big Meadows Dickey Ridge Sawmill Run
                                  e. May - September 1987
                              300 -i
                                                    • Total Dose
                                                    | Sigmoidal
                                                    DSum>0.07 ppm
                                  Big Meadows Dickey Ridge Sawmill Run
Figure AX3-40a-e.   Integrated exposures for three non-Mountain Cloud Chemistry
                     Program Shenandoah National Park sites, 1983 to 1987.

Source: Lefohnetal. (1990b).
August 2005
AX3-64      DRAFT-DO NOT QUOTE OR CITE

-------
         Table AX3-8. Summary Statistics for 11 Integrated Forest Study Sites3
Site Year
HIGH ELEVATION SITES
Whiteface Mtn, NY



Great Smoky Mtns NP



Coweeta Hydrologic Lab,
NC



LOW ELEVATION SITES
Huntington Forest, NY



Howland, ME



Oak Ridge, TN



Thompson Forest, WA




1987
1987
1988
1988
1987
1987
1988
1988
1987
1987
1988
1988

1987
1987
1988
1988
1987
1987
1988
1988
1987
1987
1988
1988
1987
1987
1988
1988
Quarter

2
3
2
3
2
o
5
2
o
3
2
3
2
3

2
3
2
o
5
2
o
3
2
3
2
3
2
o
3
2
o
5
2
3
24-h

42
45
49
44
54
53
71
59
50
47
61
57

36
24
40
37
34
26
36
24
42
29
40
32
36
30
32
32
12-h

43
44
50
43
52
51
70
57
48
44
59
54

42
32
46
46
39
32
41
30
53
44
57
47
43
36
39
39
7-h

42
43
49
43
49
49
68
55
47
42
59
51

42
33
46
48
39
31
41
30
50
41
58
51
41
34
37
36
1-h Max.

104
114
131
119
99
95
119
120
85
95
104
100

88
76
106
91
69
76
90
71
112
105
104
122
103
94
103
140
SUM06
103 ppb-h

13.2
30.1
33.5
22.6
57.1
34.3
126.3
74.7
32.4
24.1
81.6
63.6

9.8
5.4
19.2
18.6
1.9
3.8
8.1
1.7
39.5
24.3
26.4
19.7
10.7
10.3
8.1
13.5
SUM08
103 ppb-h

2.5
11.8
13.9
10.4
10.9
8.8
61.2
22.2
2.6
2.4
18.5
19.8

0.9
0.2
6.1
2.7
0.0
0.0
2.9
0.0
13.5
9.0
9.8
7.7
3.6
2.1
2.3
6.7
August 2005
AX3-65
DRAFT-DO NOT QUOTE OR CITE

-------
            Table AX3-8 (cont'd). Summary Statistics for 11 Integrated Forest Study Sites.3
Site
LOW ELEVATION SITES
B.F. Grant Forest, GA



Gainesville, FL



Duke Forest, NC



Nordmoen, Norway



Year
(cont'd)
1987
1987
1988
1988
1987
1987
1988
1988
1987
1987
1988
1988
1987
1987
1988
1988
Quarter

2
3
2
3
2
o
5
2
o
3
2
3
2
3
2
o
3
2
o
5
24-h

32
33
47
32
42
29
35
20
38
52
54
38
32
14
22
11
12-h

46
52
63
47
53
44
48
29
48
59
69
51
40
18
28
15
7-h

48
54
64
48
50
41
51
30
52
50
75
54
41
20
29
16
1-h Max.

99
102
127
116
b
b
84
70
100
124
115
141
75
32
53
30
SUM06
103 ppb-h

26.1
31.3
53.1
24.1
b
b
23.4
1.9
29.2
b
b
52.9
2.4
0.0
0.0
0.0
SUM08
103 ppb-h

5.1
10.3
21.9
7.4
b
b
0.5
0.1
7.8
b
b
23.4
0.0
0.0
0.0
0.0
        "Concentration in ppb.
        bData were insufficient to calculate statistic.
        Source: Taylor etal. (1992).
1      In most cases, mixing ratios or mole fractions are used to describe O3 concentrations. Lefohn
2      et al. (1990b) pointed out that the manner in which concentration is reported may be important
3      when assessing the potential impacts of air pollution on high-elevation forests. Given the same
4      part-per-million value experienced at both a high- and low-elevation site, the absolute
5      concentrations (i.e., micrograms per cubic meter) at the two elevations will be different, because
6      both O3 and ambient air are gases, and changes in pressure directly affect their volume.
7      According to Boyle's law, if the temperature of a gas is held constant, the volume occupied by
       August 2005
AX3-66
DRAFT-DO NOT QUOTE OR CITE

-------
 1      the gas varies inversely with the pressure (i.e., as pressure decreases, volume increases). This
 2      pressure effect must be considered when measuring absolute pollutant concentrations. At any
 3      given sampling location, normal atmospheric pressure variations have very little effect on air
 4      pollutant measurements. However, when mass/volume units of concentration are used and
 5      pollutant concentrations measured at significantly different altitudes are compared, pressure
 6      (and, hence, volume) adjustments are necessary.  In practice, the summit site at Whiteface
 7      Mountain had a slightly higher O3 exposure than the two low-elevation sites (Lefohn et al.,
 8      1991). However, at Shenandoah National Park sites, the higher elevation site experienced lower
 9      exposures than lower elevation sites in some years.
10          These exposure considerations are trivial at low-elevation sites.  However, when one
11      compares exposure-effects results obtained at high-elevation sites with those from low-elevation
12      sites, the differences may become significant (Lefohn et al., 1990b). In particular, assuming that
13      the sensitivity of the biological target is identical at both low and high elevations, some
14      adjustment will be necessary when attempting to link experimental data obtained at
15      low-elevation sites with air quality data monitored at the high-elevation stations.  This topic is
16      further discussed in Annex AX9 when considering effective dose considerations for predicting
17      vegetation effects associated with O3.
18
19
20      AX3.4  DIURNAL PATTERNS IN OZONE CONCENTRATION
21      AX3.4.1  Introduction
22          Diurnal variations in O3  at a given location are controlled by a number of factors such as
23      the relative importance of transport versus local photochemical production and loss rates, the
24      timing for entrainment of air from the nocturnal residual boundary layer and the diurnal
25      variability in mixing layer height.
26          The form of an average diurnal pattern provides some information on sources, transport,
27      and chemical formation and destruction effects at various sites (Lefohn, 1992). Atmospheric
28      conditions leading to limited transport from source regions will produce early afternoon peaks.
29      However, long-range transport processes will influence the actual timing of a peak, from
30      afternoon to evening or early morning hours.  Ozone is rapidly depleted near the surface below
31      the nocturnal inversion layer (Berry, 1964). Mountainous sites, which are above the nocturnal

        August 2005                            AX3-67      DRAFT-DO NOT QUOTE OR CITE

-------
 1      inversion layer, do not necessarily experience this depletion (Stasiuk and Coffey, 1974). Taylor
 2      and Hanson (1992) reported similar findings, using data from the Integrated Forest Study.  The
 3      authors reported that intraday variability was most significant for the low-elevation sites due to
 4      the pronounced daily amplitude in O3 concentration between the predawn minimum and
 5      mid-afternoon-to-early evening maximum. The authors reported that interday variation was
 6      more significant in the high-elevation sites.  Ozone trapped below the inversion layer is depleted
 7      by dry deposition and chemical reactions if other reactants are present in sufficient quantities
 8      (Kelly et al., 1984). Above the nocturnal inversion layer, dry deposition does not generally
 9      occur, and the concentration of O3 scavengers is generally lower, so O3 concentrations remain
10      fairly constant (Wolff et al.,  1987). A flat diurnal pattern is usually interpreted as indicating a
11      lack of efficient scavenging of O3 or a lack of photochemical precursors, whereas a strongly
12      varying diurnal pattern is taken to indicate the opposite.
13           An analysis that identified when the highest hourly average concentrations were observed
14      at rural agricultural and forested sites was described in 1996 O3 AQCD.  A review of the hourly
15      average data collected at all rural agricultural and forested sites in Environmental Protection
16      Agency's AQS database for 1990 to 1992 was undertaken to evaluate the percentage of time
17      hourly average concentrations >0.1 ppm occurred during the period of 0900 to 1559 hours in
18      comparison with the 24-h period. It was found  that 70% of the rural-agricultural and forested
19      sites used in the analysis experienced at least 50% of the occurrences >0.1 ppm during the period
20      of 0900 to 1559 hours when compared to the 24-h period.  When O3 monitoring sites in
21      California were eliminated, approximately 73% of the remaining sites experienced at least 50%
22      of the occurrences >0.10 ppm during the daylight 7-h period when compared with the
23      24-h period.
24
25      Diurnal Variations in the Nationwide Data Set
26           Composite urban, diurnal variations in hourly averaged O3 for April through October 2000
27      to 2004 are shown in Figure AX3-41. As can be seen from Figure AX3-41, daily 1-h O3 maxima
28      tend to occur in mid-afternoon and daily 1-h O3 minima tend to occur during the early morning.
29      However, there is also considerable spread in these times.  Therefore, some caution must be
30      exercised in extrapolating results from one city  to another and when attempting to judge the time
31      of day when the daily 1-h maximum occurs.

        August 2005                             AX3-68      DRAFT-DO NOT QUOTE OR CITE

-------
                                                 Urban Sites
              C
              o
                 0.200 -i
              •=•  0.150
              Q.
                 0.100 -
              c.
              03
              o
              o
              o
              o
              O  0.050
                 0.000 -I
                      22 23 00 01  02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 00 01
                                                    hour

       Figure AX3-41. Composite, nationwide diurnal variability in hourly averaged O3 in urban
                       areas. Values shown are averages  from April to October 2000 to 2004.
                       Boxes define the interquartile range and the whiskers the minima
                       and maxima.
       Source: Fitz-Simons et al. (2005).
 1          Corresponding data for 8 hour average O3 data are shown in Figure AX3-42. As can be
 2     seen from Figure AX3-42, daily maximum eight hour O3 concentrations tend to occur from
 3     about 10 a.m. to about 6 p.m. As can be seen from Figure AX3-42, they can also occur at
 4     slightly different times and the variation in the 8-h averages is smoother than for the 1-h
 5     averages. The minima in the 8 h averages tend to occur starting at about midnight.
 6
 7     AX3.3.2 Diurnal Patterns in Urban Areas
 8     Diurnal Variations in EPA's 12 Cities
 9          The diurnal variability of hourly averaged O3 in the twelve urban areas considered for
10     inclusion in EPA's human health exposure assessment-risk assessment for the current review is
11     illustrated in Figure AX3-43a-l for April to October.  Daily maximum 1-h concentrations tend to
       August 2005
AX3-69
DRAFT-DO NOT QUOTE OR CITE

-------
                                                Urban Sites
0.100 -
Q.
0.100 ppm) occurring either late in the evening as in
3     Boston, past midnight as in Los Angeles and Sacramento, or mid-morning as in Houston.
4     Typically, high values such as these are found during the daylight hours in mid to late afternoon.
5     The reasons for the behavior of O3 during the night at the above mentioned locations are not
6     clear. Measurement issues may be involved or there may be physical causes such as transport
7     phenomena, as discussed in Chapter 2.  As discussed in Chapter 2, and in greater detail in
8     Section AX2.3.3, nocturnal low level jets are capable of producing secondary O3 maxima
9     at night.
      August 2005
AX3-70
DRAFT-DO NOT QUOTE OR CITE

-------
         a. Boston-Worcester-Manchester, MA-NH
            b. New York-Newark-Bridgeport, NY-NJ-CT-PA
   O.
   o
  O
      £ O-12^ •
      Q.
      Q.
                                                     O
                                                      o>
                                                      c
                                                      o
         22 23 00 01 02 03 04 OS OS 07 OS 09 10 11 12 13 14 15 16 17 16 19 20 21 22 23 00 01

                             hour
            22 230001 02 03 04 05 06 0? 08 09 10 11 12 13 14151617 10 1920 21 22 23 00 01

                                hour
         c. Phlladelphla-Camden-Vineland, PA-NJ-DE-IWE
            d. Washington-Baltimore-Northern Virginia, DC-MD-VA-WV
  — 0 125 •


  O.
  O
  O
                                                     —. 0.12S
      O
      o
         22 23 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 00 01

                             hour
            22 23 0001 C2 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17181920 21 22 2300 01

                                hour
        e. Atlanta-Sandy Springs-Gainesville, GA-AL
   o
  o
  0  0,025-
            f. Cleveland-Akron-Elyria, OH
                                                     —- 0.125 -
      O
      O
         22 23 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 2C 21 22 230001

                             hour
            22230001 02030405060708091011 12131415161718192021 22230001

                                hour
Figure AX3-43a-f.  Diurnal variability in hourly averaged O3 in selected urban areas.

                       Values shown are averages from April to October 2000 to 2004.

                       Boxes define the interquartile range and the whiskers the minima

                       and maxima.



Source: Fitz-Simons et al. (2005).
August 2005
AX3-71        DRAFT-DO NOT QUOTE OR CITE

-------
        g. Detroit-Warren-Flint, MI
            h. Chicago-Naperville-Michigan City, IL-IN-WI
  o
  o
      o
      o
              1 02030405060708091011 121314151617 1819202122230001
                            hour
            22 23 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23
                                hour
        i. St. Louis-St. Charles-Farmington, MO-IL
            j. Houston-Baytown-Huntsville, TX
                                                     o
                                                     o
        22 23 00 01 02 03 04 05 06 07 08 C
                          i 10 11 12 13 14 15 16 17 18 192021 2223 00 01
                            hour
            22 230001 020304050607 08 09 10 11 12 13 14 15 16 17 13 19 20 21 22 23 OC 01
                                hour
         k. Sacramento-Arden-Arcade-Truckee, CA-NV
  S
  E  0.100
  O
  o
            I. Los Angeles-Long Beach-Riverside, CA
      O
      o
                                                     O  0.050
                                                     N
                                                     O
                                                        o ooo -Li.
        22 23 0301 0203 04 05 0607 0809 10 11 12 13 14 15 16 17 18 1920 21 22 2300 01
                            hour
            2223000102030405060738091011 "21314 15 16 17 13 19 20 21 22230001
                                hour
Figure AX3-43g-l.   Diurnal variability in hourly averaged O3 in selected urban areas.
                       Values shown are averages from April to October 2000 to 2004.
                       Boxes define the interquartile range and the whiskers the minima
                       and maxima.

Source: Fitz-Simons et al. (2005).
August 2005
AX3-72       DRAFT-DO NOT QUOTE OR CITE

-------
 1           The diurnal variability of O3 averaged over 8 hours in the same twelve urban areas is
 2      shown in Figures AX3-44a-l.  The diurnal patterns of O3 are broadly similar between 1-h
 3      averages and 8-h averages. A typical pattern shows the  8-h daily maximum occurring from
 4      about 10 a.m to about 6 p.m., with some deviations from these times. However, as shown in
 5      Figures AX3-44a for Boston and AX3-44k for Sacramento, the highest 8-h daily maximum
 6      values occur starting in mid-afternoon and extending into late evening. These results suggest
 7      that transport processes are playing the dominant role in determining the timing of the highest
 8      daily maxima in these areas.
 9           On days with high 1-h daily maximum concentrations (e.g., >0.12 ppm) the maxima tend
10      to occur in a smaller time window centered in the middle of the afternoon, compared to days in
11      which the maximum is lower.  For example, on the high O3 days the 1-h maximum occurs from
12      about 11 a.m. to about 6 p.m.  However, on days in which the 1-h daily maximum is <0.080
13      ppm, the daily maximum can occur at any time during the day or night, with only a 50%
14      probability that it occurs between 1 and 3 p.m., in each of the 12 cities. Photochemical reactions
15      in combination with diurnal emissions patterns are expected to produce mid-afternoon peaks in
16      urban areas. These results suggest that transport from outside the urban airshed plays the major
17      role for determining the timing of the daily maxima for low peak O3 levels. This pattern is more
18      typical for the Los Angeles-Long Beach-Riverside, CA area even for high O3 days.
19           The same general patterns emerge for the timing of the  1-h daily maximum O3
20      concentration as are found for the daily maximum  8-h average O3 concentration. As mentioned
21      above, the daily maximum 8-h O3 concentrations are generally found between the hours of
22      10 a.m and 6 p.m. However, there are significant fractions of the time when this is not the case
23      for high values, as in Houston, TX and Los Angeles, CA, or in general for lower values at any of
24      the cities examined. Although the 8-h average O3 concentration is highly correlated with the
25      daily maximum 1-h average O3 concentration, there are situations where the daily maximum 8-h
26      average O3 concentration may be driven by very high values in the daily maximum 1-h
27      average O3 concentration as illustrated in Figure 3-43J. In cases such as these, the predicted 8-h
28      average might overestimate the short-term O3 concentration later in the day.
       August 2005                            AX3-73      DRAFT-DO NOT QUOTE OR CITE

-------
        a. Boston-Worcester-Manchester, MA-NH
            b. New York-Newark-Bridgeport, NY-NJ-GT-PA
  £  0.100.
  a.
  Q.
  O
  ••B  0.075 •
  O
  o
  c
  s
                                                     E 0.100
      o
      '•£ O.Q75
      O
      9)
        22 230D01 02 03 0405 06 07 0509 10 11 12 13 1415 16 17 18 19 20 21 2223 00 01

                            hour
            22 230001 02030405060708 09 10 11 12 13 14 15 16 17 18 19 20 21 2223 00 01

                                hour
        c. Philadelphia-Camden-Vineland, PA-NJ-DE-ME
           d. Washington-Baltimore-Northern Virginia, DC-MD-VA-WX
  E  0.100 •
  Q.
  a
                                                     O
                                                     d>
        22 23 00 01 02 030405 OC 070809 10 11 12 13 14 15 1G 17 18 192021 22 23 00 01

                            hour
            2223000102030405060708091011 12-3 14 15 16 17 18 19 20 21 22 23 00 01

                                hour
        e. Atlanta-Sandy Springs-Gainesville, GA-AL
  O.
  O
  O
                                         1
            f, Cleveland-Akron-EIyria, OH
                                                     O
                                                     0)
                                                     c
                                                     o
        22230001 02 03 04 05 06 07 OS 09 10 11 12131415161718192021 22230001

                            hour
            22 23 0001 02030405060708 09 10 11 12 13 14 15 16 17 13 19 20 21 22 2300 01

                                hour
Figure AX3-44a-f.   Diurnal variability in 8 hour averaged O3 in selected urban areas.

                       Values shown are averages from April to October 2000 to 2004.

                       Boxes define the interquartile range and the whiskers the minima

                       and maxima.  The hour refers to the start of the 8-h averaging period.


Source:  Fitz-Simons et al. (2005).
August 2005
AX3-74        DRAFT-DO NOT QUOTE OR CITE

-------
        g. Detroit-Warren-Flint, MI
                                                  h. Chicago-Naperville-Michigan City, IL-IN-WI
  ~  0-125 •
  O
  o
  O  n m
                                                    —  0.125
        22230001 02030405060708091011 121314151617 1819202122230001
                            hour
                                                  22 23 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 00 01
                                                                     hour
        i. St. Louis-St. Charles-Farmington, MO-IL
  —  0.125
                                                  j. Houston-Baytown-Huntsville, TX
                                                    E
                                                    5  "-<"5 •
                                                    O
                                                       0.050 -
                                                    O
                                                    c
                                                    o
                                                    0  0.025
                                                                                           fflSE
22 23 00 01 02 03 04 05 06 070809 10 11 12 13 14 15 10 17 18 192021 2223 00 01
                   hour
                                                          22 230001 020304050607 08 09 10 11 12 13 14 15 16 17 13 19 20 21 22 23 OC 01
                                                                              hour
        k. Sacramento-Arden-Arcade-Truckee, CA-NV
  -—  0.125
  a
  S
  E  0.075
  O
  O
                                                  I. Los Angeles-Long Beach-Riverside, CA
                                                    a
                                            o
                                            O
        22 23 0301 0203 04 05 0607 0809 10 11 12 13 14 15 16 17 18 1920 21 22 2300 01
                            hour
                                                  2223000102030405060738091011 "21314 15 16 17 13 19 20 21 22230001
                                                                      hour
Figure AX3-44g-l.  Diurnal variability in 8 hour averaged O3 in selected urban areas.
                      Values shown are averages from April to October 2000 to 2004.
                      Boxes define the interquartile range and the whiskers the minima
                      and maxima. The hour refers to the start of the 8-h averaging period.

Source: Fitz-Simons et al. (2005).
August 2005
                                      AX3-75        DRAFT-DO NOT QUOTE OR CITE

-------
 1           As an aid to better understanding the nature of the diurnal patterns shown in the figures for
 2     EPA's 12 cities, Figures AX3-45a-d show the hours in which the 1-h daily maximum O3
 3     concentration occurs in four of the cities.  As can be seen from Figures AX3-45a-c for the
 4     Philadelphia, Atlanta, and Houston areas, the maximum tends to occur from about 2 p.m. to
 5     4 p.m. about half of the time, and most values occur between about 12 p.m to 6 p.m at higher
 6     values of the daily maximum 1-h O3 concentration.  Although values at Houston can occur
 7     earlier, these are most likely due to episodic releases from the petrochemical industries.
 8     For lower values of the 1-h daily maximum, most of the daily maxima still occur in the
 9     afternoon, but maxima can also occur at any time of the day or night. In the Los Angeles area,
10     as shown in Figure AX3-45d high values  of daily 1-h O3 maxima can occur at any time during
11     the day or night but with most values occurring during the afternoon.
12           Figures AX3-46a-d show the hours  in which the 8-h daily maximum O3 concentration
13     begins. The mean time is about 10 a.m. at these four cities indicating that the 8-h daily
14     maximum tends to occur on average from about 10 a.m to 6 p.m.  However, there can be
15     deviations from these times.  The same general pattern in which the maxima tend to occur within
16     a narrower time frame at high values than at low values is found in the four cities shown.
17           The patterns of diurnal variability for both 1-h and 8-h averages have remained quite stable
18     over the 15 year period from 1990 to 2004, with times of occurrence of the daily maxima
19     varying by no more than an hour from year to year in each of the  12 cities.
20
21     Weekday/Weekend Differences
22           In addition to varying diurnally, O3  concentrations also vary from weekdays to weekends.
23     Heuss et al. (2003) described the results of a nationwide analysis  of weekday/weekend
24     differences that demonstrated significant variation in these differences across the United States.
25     Weekend 1-h or 8-h maximum O3 varied from 15% lower to 30% higher than weekday levels
26     across the U.S.  The weekend O3 increases were primarily found in and around large coastal
27     cities in California and large cities  in the Midwest and Northeast  Corridor. Many sites that
28     experienced elevated weekday O3 also had higher O3 on weekends even though the  traffic and O3
29     precursor levels were substantially reduced on weekends. The authors reported that detailed
30     studies of this phenomenon indicated that the primary  cause of the higher O3 on weekends was
       August 2005                            AX3-76      DRAFT-DO NOT QUOTE OR CITE

-------
         a. Philadelphia-Camden-Vineland, PA-NJ-DE-MD      b. Atlanta-Sandy Springs-Gainesville, GA-AL
24 •
18 •
12 -
6-
n-

























f



li


24 •
r IT r T = :
Ml °12;
6 -
n .



















J 1 T T T
]E


!!!T


            < 0.07 0.07  0.08  0.09  0.10 0.11 0.12 0.13 0.14 > 0.14
         c. Houston-Baytown-Huntsville, TX
24 •
18 •
12 -
6-
0-









]E



3E



3E



3E



3E



]E


till
3TTT


          < 0.07 0.07 0.08 0.09 0.10 0.11  0.12  0.13  0.14 > 0.14

       d.  Los Angeles-Long Beach-Riverside, CA
       24 -i
                                                        18 -
                                                        12 -
                                                         6 -
                                                         0 -I
            < 0.07 0.07  0.08  0.09  0.10 0.11 0.12 0.13 0.14 > 0.14
                                                           < 0.07 0.07 0.08 0.09 0.10 0.11  0.12 0.13 0.14 > 0.14
       Figure AX3-45a-d.  Time of occurrence of daily maximum 1-h O3 concentration in four
                            cities, averaged from April to October, 2000 to 2004.
       Source: Fitz-Simons et al. (2005).
1      the reduction in oxides of nitrogen emissions on weekends in a volatile organic compound
2      (VOC)-limited (NOx-saturated) chemical regime (cf., Chapter 2).  Heuss et al. (2003)
3      hypothesized that the lower O3 on weekends in other locations may result from NOX reductions
4      in a NOx-limited regime (cf, Chapter 2).
       August 2005
AX3-77      DRAFT-DO NOT QUOTE OR CITE

-------
         a.  Philadelphia-Camden-Vineland, PA-NJ-DE-MD     b. Atlanta-Sandy Springs-Gainesville, GA-AL
24 •
18 •
12 -
6-
0-

E



3^



] E


3
j
--


L

24 •
18 •
4 | S e H
6 -
n .

c



] C



:c



:c

'{{HfB


            <0.07 0.07  0.08  0.09  0.10 0.11  0.12  0.13  0.14
         c.  Houston-Baytown-Huntsville, TX
                                                          < 0.07 0.07 0.08 0.09 0.10  0.11  0.12  0.13  0.14 > 0.14
       d. Los Angeles-Long Beach-Riverside, CA
24 -
18 •
"
12 -
-
6-
0-



:





: E





q c





1 C



T

:TC
.


T

1 Y E
.




3E





3 c


24 -
18 •
3
°12-

6 -
, 	 0 -



C





] L





] L





: c





: c





: c





: c





__


n cp i±i i±i
^

            < 0.07 0.07 0.08 0.09 0.10 0.11  0.12  0.13  0.14 > 0.14
                                                          < 0.07 0.07 0.08 0.09 0.10  0.11  0.12  0.13  0.14 > 0.14
       Figure AX3-46a-d.   Time of occurrence of daily maximum 8-h average O3 concentration
                            in four cities, averaged from April to October, 2000 to 2004.  The
                            hour refers to the start of the 8-h averaging period.
       Source: Fitz-Simons et al. (2005).
1           Pun et al. (2003) described the day-of-week behavior for O3 in Chicago, Philadelphia, and
2      Atlanta.  In Chicago and Philadelphia, maximum  1-h average O3 increases on weekends.  In
3      Atlanta, O3 builds up from Mondays to Fridays and declines during the weekends.  Fujita et al.
4      (2003) pointed out that since the mid-1970s, O3 levels in portions of California's South Coast
5      Air Basin on weekends have been as high as or higher than levels on weekdays, even though
       August 2005
AX3-78
DRAFT-DO NOT QUOTE OR CITE

-------
 1      emissions of O3 precursors are lower on weekends.  Blanchard and Tanebaum (2003) noted that
 2      despite significantly lower O3 precursor levels on weekends, 20 of 28 South Coast Air Basin
 3      sites showed statistically significant higher mean O3 levels on Sundays than on weekdays.
 4      Chinkin et al. (2003) noted that ambient O3 levels in California's South Coast Air Basin can be
 5      as much as 55% higher on weekends than on weekdays under comparable meteorological
 6      conditions.
 7          Figures AX3-47a-h show the contrast in the patterns of hourly averaged O3 in the greater
 8      Philadelphia, Atlanta, Houston and Los Angeles areas from weekdays to weekends from May to
 9      September 2004.  Daily maximum concentrations occur basically at the same time on either
10      weekdays or weekends. Mean O3 concentrations at midday are about the same on weekdays and
11      weekends in Atlanta, Philadelphia, and Houston, but are higher on weekends in the Los Angeles
12      area.  Figures AX3-48a-h show the weekday/weekend differences for the 8-h averages.  As can
13      be seen from the figures, the lowest O3 concentration observed during weekend afternoons tend
14      to be higher than on weekday afternoons. On weekends, traffic volumes are lower and there are
15      fewer diesel vehicles on the road, resulting in a lower rate of scavenging by NO.  The spike in
16      values shown for Houston in mid-morning shown in Figure AX3-47 resulted from the release of
17      highly reactive hydrocarbons from the petrochemical industry (which could occur on any day of
18      the week). Otherwise, the maximum O3 concentrations could be seen to occur during the week
19      as they do in Philadelphia and Atlanta, in contrast to Los Angeles.
20
21      Spatial Variability in Diurnal Patterns
22          Daily maxima in either the 1-h or 8-h averages do not necessarily occur at the same time of
23      day at each site in the 12 cities, and the diurnal pattern observed at individual sites can vary from
24      the composites shown in Figures AX3-41 and 42. Differences in diurnal patterns between sites
25      are related to differences in transport times from sources of precursors, chemical  reactions, in
26      particular, titration of O3 by NO from local sources. Figure AX3-49a shows the diurnal pattern
27      of 1-h average O3 at a site in downtown Detroit, MI (cf.  Site J in Figure AX3-30). This site is
28      affected by nearby traffic emissions.  Figure AX3-49b shows the diurnal pattern at a site well
29      downwind (cf, Site D in Figure AX3-30). The peak 1-h average O3 concentrations tend be
30      higher at the downwind site than at the site in the urban core. Figure AX3-50a shows the diurnal
31      patterns at a site in downtown St. Louis (cf. Site P in Figure AX3-31) and Figure AX3-50b

        August 2005                            AX3-79       DRAFT-DO NOT QUOTE OR CITE

-------
             a.  Philadelphia-Camden-Vineland, PA-NJ-DE-MD
                            (week day)
        E
        a.
        a.
        o
        o
             22 23 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 00 01
                                hour

             c. Atlanta-Sandy Springs-Gainesville, GA-AL
                          (week day)
        B.
        D.
        O
        O
           b.  Philadelphia-Camden-Vineland, PA-NJ-DE-MD
                           (week end)
     ~ 0,125
     Q.
     a
                                                       _o
                                                       I
                                                       £  0.075 •
      o
     O
                                                        o
                                                        8
           22230001 02 03 04 05 06 07 OB 09 10 11 12131415161718192021 22230001
                              hour

           d. Atlanta-Sandy Springs-Gainesville, GA-AL
                         (week end)
     -~ 0.125 •
     Q.
     O.
                                                        O
                                                        I
                                                        S  0.075 •
                                                        
-------
             e.  Houston-Baytown-Huntsville, TX
                       (week day)
        o.
        Q.
        O
        O
                    Biaa
                             II
             22 23 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 00 01
                                hour

             g.  Los Angeles-Long Beach-Riverside, CA
                          (week day)
        O
        o
O  0.050 •
N
O
             2223 00 01 02 0304 05 0607 0809 10 11 12 13 14 15 16 17 18 192021 22 2300 01
                                hour
                                                      f. Houston-Baytown-Huntsville, TX
                                                                (week end)
                                                       _o
                                                o
                                                o
                                                        O  0.050 •
                                                        s
                                                      22230001 02 03 04 05 06 07 OB 09 10 11 12131415161718192021 22230001
                                                                         hour

                                                      h. Los Angeles-Long Beach-Riverside, CA
                                                                  (week end)
                                                       £
                                                       S  0.100
                                                o
                                                o
                                                        O  0.050
                                                        N
                                                        O
                                                      2223 0001 02 03 0405 06 0708 09 10 11 12 13 14 15 16 17 18 1920 21 22 230001
                                                                         hour
       Figure AX3-47e-h.  Diurnal variations in hourly averaged O3 on weekdays and weekends
                            in four cities.  Values shown represent averages from
                            May to September of 2004.
       Source: Fitz-Simons et al. (2005).
1      in a relatively unpopulated area. The diurnal pattern of hourly averaged O3 is much flatter and
2      the 1-h peak concentration is reached about 5 or 6 p.m., on average.  The cause of the rise in
3      concentrations at 2 a.m. is not clear.
4            The diurnal variation in the 8-h averages observed at the two contrasting sites in these three
5      areas are shown in Figures AX3-52a,b, 53a,b, and 54a,b. As might be expected the patterns are
6      somewhat flatter than for the 1-h averages.  This implies that the difference in 8-h averages can
7      be substantial (i.e., over a factor of two) during early morning and afternoon and evening.
       August 2005
                                           AX3-81       DRAFT-DO NOT QUOTE OR CITE

-------
             a. Philadelphia-Camden-Vineland, PA-NJ-DE-MD
                             (week day)
        E 0,100
        a.
        o
        S5 0075
        2
              22 23 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 00 01
                                 hour


              c. Atlanta-Sandy Springs-Gainesville, GA-AL
                          (week day)
        £ 0.100
        B.
        D.
        O
        5 0.075
           b. Philadelphia-Camden-Vineland, PA-NJ-DE-MD
                           (week end)
      E  0.100
      a.
           22230001 02 03 04 05 06 07 OB 09 10 11 12131415161718192021 22230001
                               hour


           d. Atlanta-Sandy Springs-Gainesville, GA-AL
                         (week end)
      £  0.100
      Q.
      O.
                                                        o
                                                        g  0.025
              2223 00 01 02 0304 05 0607 0809 10 11 12 13 14 15 16 17 18 192021 22 2300 01
                                 hour
           2223 0001 02 03 0405 06 0708 09 10 11 12 13 14 15 16 17 18 1920 21 22 230001
                              hour
       Figure AX3-48a-d.  Diurnal variations in 8-h averaged O3 on weekdays and weekends in
                             four cities. Values shown represent averages from May to September
                             of 2004.  The hour refers to the start of the 8-h averaging period.

       Source: Fitz-Simons et al. (2005).
1            The general pattern that emerges from the site to site variability within the urban areas

2      examined is that peaks in 1-h average concentrations are higher and tend to occur later at

3      downwind sites than in the urban cores. To the extent that monitoring site are either near to or

4      remote from sources of precursors in urban/suburban areas, the behavior of O3 will follow these

5      basic patterns. Similar relations are found for the 8-h average O3 concentrations.
       August 2005
AX3-82       DRAFT-DO NOT QUOTE OR CITE

-------
             e. Houston-Baytown-Huntsville, TX
                       (week day)
             22 23 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 00 01
                               hour


             g. Los Angeles-Long Beach-Riverside, CA
                         (week day)
       —  0,125

       B.
       D.
       O
       o
       s
       o
           f. Houston-Baytown-Huntsville, TX
                    (week end)
           22230001 02 03 04 05 06 07 OB 09 10 11 12131415161718192021 22230001
                             hour


           h. Los Angeles-Long Beach-Riverside, CA
                       (week end)
     -~ 0.125 •

     Q.
     O.
                                                      O

                                                      I
                                                      S 0.075 •
                                                      
-------
                  0.125-1
                  0.100-
               Q.
               O.

               C
               ,°  0.075
               C
               an
               c
               O  0.050
               0)
               c
               s
               O
                  0.025 -
                  0.000 -
                          Detroit-Warren-Flint, Ml
                              Site: 261630016
                      22 23 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 00 01
                                                  hour


Figure AX3-49a.  Diurnal variations in hourly averaged O3 at a site in downtown Detroit,
                   MI.


Source: Fitz-Simons et al. (2005).
                  0.125 n
               Q.
               5:
               c
               °  0.075
O
O  0.050

-------
                  0.100-1
                  0.080 -

               Q.
               O.
               c
               .2 0.060
c
an
a
c
O  0.040
0)
c
s
O
   0.020 -
                  0.000 -
                    St. Louis-St. Charles-Farmington, MO-IL
                              Site: 295100072
                       TTTULJ
                                                       ±
                                                                      ±
                                                           XT T U
                       22 23 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 00 01
                                                   hour

Figure AX3-50a.   Diurnal variations in hourly averaged O3 at a site in downtown
                     St. Louis, MO.

Source:  Fitz-Simons et al. (2005).
                  0.100 -i
                  0.080 -
               Q.
c
.2  0.060
I
c
u
O  0.040
i)
C
s
o
   0.020 -
                  0.000 -I
                    St. Louis-St. Charles-Farmington, MO-IL
                              Site: 170831001
                                                             ±
                       22 23 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 00 01
                                                   hour

Figure AX3-50b.  Diurnal variations in hourly averaged O3 at a site downwind of
                    downtown St. Louis.

Source:  Fitz-Simons et al. (2005).
August 2005
                             AX3-85        DRAFT-DO NOT QUOTE OR CITE

-------
                  0.200 i
               •£  0.150-

               a
               a.

               c
               o
               "^
               ra


               g  0.100 H
               u
               c
               o
               o
               01


               I

                  0.050 -
                  0.000 -I
Los Angeles-Long Beach-Riverside, CA

          Site: 060719004
                         + IT
                          nnnn
                      22 23 00 01 02 03 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 00 01


                                                 hour
Figure AX3-51a.   Diurnal variations in hourly averaged O3 at a site in

                    San Bernadino, CA.



Source: Fitz-Simons et al. (2005).
                  0.200 -i
               •=•  0.150

               Q.
               a.

               c
               o
               53  0.100
               u
               c
               O
               o
               
-------
                 0.100 -i
                 0.800 -
               Q.

               Q.
              .2  0.600 -
              *J

              I

              11
              o
              c
              o
              O  0.400

              0)
              c
              O
              N
              O
                 0.000 -
Detroit-Warren-Flint, Ml

   Site: 261630016
                      X XT
                           T
                    In
T
T
                     22 23 00 01 02 03 04 05 06 07
                                              10 11 12 13 14 15 16 17 18 19 20 21 22 23 00 01


                                                hour
Figure AX3-52a.  Diurnal variations in 8-h average O3 at a site in downtown Detroit, ML



Source: Fitz-Simons et al. (2005).
                 0.100 -i
               Q.

               5:

               c

               ,2 0.600 -
               o
               O 0.400 -
               
-------
                ~  0.600


                Q.
                Q.
                u
                c
                o
                O
                o
                c
                o


                   0.200 -
                                   St. Louis-St. Charles-Farmington, MO-IL

                                             Site: 295100072
±
                                                                      ±
                       22 23 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 00 01


                                                  hour
Figure AX3-53a.  Diurnal variations in 8-h average ozone at a site in downtown

                   St. Louis, MO.



Source: Fitz-Simons et al. (2005).
                —  0.600 -


                Q.
                O.
                C
                O
                o
                o>
                c
                O


                   0.200 -
                                   St. Louis-St. Charles-Farmington, MO-IL

                                             Site: 170831001

                                                          ±
                       22 23 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 00 01

                                                  hour




Figure AX3-53b.  Diurnal variations in 8-h average O3 at a site downwind of downtown

                   St. Louis, MO.



Source: Fitz-Simons et al. (2005).
August 2005
  AX3-88       DRAFT-DO NOT QUOTE OR CITE

-------
                  0.150 -i
               0.
               0.
               o
               1
               o
               c
               o
               o
               0)
               g

               8
                  0.100-
                  0.000 -I
  Los Angeles-Long Beach-Riverside, CA
            Site: 060719004
                      BDQ
                                   ±
                                     ±
                                       BOB
                      22 23 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 00 01

                                                 hour




Figure AX3-54a.  Diurnal variations in 8-h average O3 at a site in San Bernadino, CA.



Source: Fitz-Simons et al. (2005).
                                  Los Angeles-Long Beach-Riverside, CA
                                            Site: 060719002
               a.
               o
               O
               c
               O
               O
               o
               N
               O
                  0.075 -
                  0.050 -
                  0.025 -
                  0.000 -I
L
r
                      22 23 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 00 01

                                                 hour



Figure AX3-54b.  Diurnal variations in 8-h average O3 at a site in Riverside County well

                   downwind of sources.


Source: Fitz-Simons et al. (2005).
August 2005
           AX3-89       DRAFT-DO NOT QUOTE OR CITE

-------
                                    Rural (CASTNET) Sites
          0.200 -i
       E
       g.  0.150
       c
       o
       "f
       ra
       ^
       C  0.100
       
-------
 1      observed during any particular hour at night at the CASTNET sites (-0.130 ppm) are
 2      substantially higher than observed in urban areas (<0.100 ppm) and daily 1-h maxima at
 3      CASTNET sites have exceeded 0.150 ppm.  The diurnal variations in 8-h average O3
 4      concentrations are also much smaller at the CASTNET sites than at the urban sites. Note also
 5      that the maxima in 8-h average O3 concentrations are higher at the CASTNET sites than at the
 6      urban sites.
 7           The diurnal variability of O3 in urban/suburban areas or in areas affected by local power
 8      plants and highways is usually much greater than in other more isolated areas.  The diurnal
 9      variability of two sites that are characteristic of these two patterns is shown in Figure AX3-57.
10      The Jefferson County, KY site is characterized as suburban-residential in the AQS database and
11      is near Louisville, KY. High levels of O3 and NO can be found there. The Oliver County,
12      ND site is  characterized as rural-agricultural in the AQS database. This site is fairly isolated
13      from combustion sources of precursors and is not near any large urban area.  As can be seen
14      from Figure AX3-57,  the diurnal variability of O3 is much smaller at the North Dakota site than
15      it is at the Kentucky site. The Kentucky site is influenced strongly by emissions of NO that
16      scavenge O3  during the night and by photochemical reactions that form O3 during the day. These
17      sources are lacking in the vicinity of the North Dakota site, and O3 observed there arrives mainly
18      from transport from distant source regions.
19           Logan (1989) described the diurnal variability of O3 at several rural locations, shown in
20      Figure AX3-58, and noted that on average, daily profiles show a broad maximum from about
21      noon to about 6 p.m. at all the eastern sites, except for the peak of Mt. Washington. Further
22      results that document  the diurnal behavior of O3 in the United States during the past few decades
23      can be found in the previous AQCD for O3.  Figure AX3-59 shows diurnal patterns for several
24      national forest sites in the EPA AQS database for 2002.  Several of the sites analyzed exhibit
25      fairly flat average diurnal patterns.  Such a pattern is based on average concentrations calculated
26      over an extended period and caution is urged in drawing conclusions concerning whether some
27      monitoring sites illustrated in the figure experience higher cumulative O3 exposures than other
28      sites. Variation in O3  concentration occurs from hour to hour on a daily basis, and, in some
29      cases, elevated hourly average concentrations are experienced either during daytime or nighttime
30      periods (Lefohn and Mohnen, 1986; Lefohn and Jones, 1986; Logan, 1989; Lefohn et al., 1990a;
31      Taylor et al., 1992). Because the diurnal patterns represent averaged concentrations calculated

        August 2005                             AX3-91      DRAFT-DO NOT QUOTE OR CITE

-------
                                                            Jefferson Co. KY
                                                            Oliver Co. ND
                          1    3    5    7   9    11    13  15   17   19  21   23
                                               Hour of Day
       Figure AX3-57.  The comparison of the seasonal diurnal patterns for urban-influenced
                        (Jefferson County, KY) and a rural-influenced (Oliver County, ND)
                        monitoring sites using 2002 hourly data for April-October.
 1     over an extended period, the smoothing from the averaging tends to mask the elevated hourly
 2     average concentrations.
 3          Lefohn et al. (1990b) characterized O3 concentrations at high-elevation monitoring sites.
 4     The authors reported that a fairly flat diurnal pattern for the Whiteface Mountain summit site
 5     (WF1) was observed (Figure AX3-60a), with the maximum hourly average concentrations
 6     occurring in the late evening or early morning hours. A similar pattern was observed for the
 7     mid-elevation site at Whiteface Mountain (WF3). The site at the base of Whiteface Mountain
 8     (WF4) showed the typical diurnal pattern expected from sites that experience some degree of O3
 9     scavenging. More variation in the diurnal pattern for the highest Shenandoah National Park sites
10     occurred than for the higher elevation Whiteface Mountain sites, with the typical variation for
11     urban-influenced sites in the diurnal pattern at the lower elevation Shenandoah National Park site
12     (Figure AX3-60b). Aneja and Li (1992), in their analysis of the five high-elevation Mountain
13     Cloud Chemistry Program (MCCP) sites, noted the presence of the flat diurnal pattern typical of
       August 2005
AX3-92
DRAFT-DO NOT QUOTE OR CITE

-------
         .Q
         Q.
         a.
          CO
         o
50

40

30

20

10

60

50

40

30
60

50

40

30

20

10

 0
80

70

60

50

40

30

20

10

 0
_aj
— A7 _-^r 	

~--~^-~^'^'i^^^^ NOR
1 1 1 1
' MT
./'•*
— "••*-.,
__ ~~~--
1
                         WFM, NY
                                                        J;N(R)  PA
                                  I
I
I
                                 8        12       16
                                    Time of Day (h)
                                               20
                           24
Figure AX3-58a-d.  Diurnal behavior of O3 at rural sites in the United States in July.
                   Sites are identified by the state in which they are located.
                   (a) Western National Air Pollution Background Network sites
                   (NAPBN); (b) Whiteface Mountain (WFM) located at 1.5 km above
                   sea level; (c) eastern NAPBN sites; and (d) sites selected from the
                   Electric Power Research Institute's Sulfate Regional Air Quality
                   Study. IN(R) refers to Rockport.

Source: Logan (1989).
August 2005
                        AX3-93
       DRAFT-DO NOT QUOTE OR CITE

-------
                        70
                                                         Acadia NP
                                                         Great Smoky Mtn. NP
                                                         Glacier NP
                                                         Yellowstone NP
                                                         Grand Canyon NP
                                                                        \
                             1    3   5   7   9   11   13  15  17   19  21  23
                                                Time of Day
       Figure AX3-59.  Composite diurnal O3 pattern at selected national forest sites in the
                        United States using 2002 hourly average concentration data.
       Source: U.S. Environmental Protection Agency (2003a).
 1
 2
 3
 4
 5
 9
10
11
high-elevation sites that has been described previously in the literature.  Aneja and Li (1992)
noted that the peak of the diurnal patterns over the period of May to October (1986 to 1988)
occurred between 1800 and 2400 hours for the five sites, whereas the minimum was observed
between 0900 and 1200 hours. However, it is important to note that, as indicated by Lefohn
et al. (1990b), the flat diurnal pattern was not observed for all high-elevation sites.
As mentioned earlier, nonurban areas only marginally affected by nearby sources usually have a
flatter diurnal profile than sites located in urban areas. Nonurban O3 monitoring sites experience
differing types of diurnal patterns, as shown in this section. The difference in diurnal patterns
may influence the potential for O3 exposures to affect vegetation.
       August 2005
                                        AX3-94
DRAFT-DO NOT QUOTE OR CITE

-------
             E
             Q_
             o>
             c
             o
             N
             O
                0.06 -\
                0.05
                0.04 -
0.03
                0.02 -
                0.01 -
                0.00
                0.06
                0.00
                                           WF1 (1483 m)
                                           WF3 (1026m)
                                           WF4 (604 m)
                                     8   10  12  14  16  18   20   22   24
                                            Hour
                                                           SH1 (1015m)
                                                           SH3 (716 m)
                                                           SH3 (524 m)
                                     8   10  12  14  16  18   20   22   24
                                            Hour
Figure AX3-60a,b.  Composite diurnal pattern at (a) Whiteface Mountain, NY and
                   (b) the Mountain Cloud Chemistry Program Shenandoah National
                   Park site for May to September 1987.

Source:  Lefohnetal. (1990a).
August 2005
                      AX3-95      DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX3.5  SEASONAL VARIATIONS IN OZONE CONCENTRATIONS
 2     AX3.5.1  Seasonal Variations in Urban Areas
 3     Seasonal Variability
 4          Figures AX3-61a-h show maximum 1-h O3 concentrations by month for selected urban
 5     sites for 2002.  As can be seen from the figure, maximum 1-h O3 concentrations tend to occur
 6     mainly in July and August, but may also occur in other months. For example, they occurred in
 7     June in Washington, DC and Denver, CO.  The number of months for which data are shown
 8     depends on local preference for the length of monitoring during the year. Due to a number of
 9     factors, the absolute magnitude and the timing of the maximum hourly average concentrations
10     varies from year to year.
11          It should not be assumed that highest O3 levels are confined to the summer. Highest
12     average O3 concentrations generally occur at RRMS during the second quarter (i.e., during April
13     or May) versus the third quarter of the year as for urban sites or for nonurban sites heavily
14     affected by regional pollution sources.
15          The seasonal behavior of O3 varies across the 12 cities and high O3 values are also found at
16     some of the 12 cities outside of summer (e.g., Houston and Los Angeles).  Figures AX3-62a-l
17     show the diurnal variability of hourly average O3 averaged over November through March for
18     EPA's 12 cities. Daily maxima tend to occur between about 1 and 2 p.m. standard time is used
19     across the U.S. accounting for the one hour shift from the warm season. As expected, maximum
20     values tend to be lower than during the warmer months.  The diurnal patterns are not as clear as
21     in the warmer season as there is a greater tendency for highest values to occur throughout the
22     day and not only during early afternoon. In most northern cities, the extreme values of the daily
23     maximum 8-h average O3 concentration are a little more than half of those during the
24     warm season and the ratio of the medians are more similar as can be judged by comparison of
25     Figures AX3-41a-l with Figures AX3-62a-l.  Differences are even smaller for the southern cities.
26     Indeed, some of the highest values are found in the Houston CSA outside of summer.
27          Figures AX3-63a-l show the diurnal variability of 8-h average O3 averaged over November
28     through March for EPA's 12 cities.
29
       August 2005                            AX3-96     DRAFT-DO NOT QUOTE OR CITE

-------
         0.20
         0.18
       •p1 0.16
       0. 0.14
       3 0.12
       0, 0.10
       o °08
       N 0.06
       O 0.04
         0.02
         0.00
            a. Phoenix, AZ
                             2002
                                    0401330019
      I
	Mlh
i
                                   E  S
                                   (U  O
                                   •B.  O
                                            b. Denver, CO
                                                                   2002
                                                                          080310014
0.18
0.16
0.14
0.12
0.10
0.08
0.06
0.04
0.02
0.00




I_
• _
II I
•
• ||
III
III
>^>.^= >^ oj >'"S
roro^Q-^^^3
zs zs ro ,
                  S  q.  ^
                              €
                              O
                                        E  E
                                               0.20
                                               0.18
                                               0.16
                                               0.14
                                               0.12
                                               0.10
                                               0.08
                                               0.06
                                               0.04
                                               0.02
                                               0.00
                                            h. Houston, TX
                                                                   2002
                                                                          482011034
                                   Mill
                                                                         E S
                                                                         CD O
                                                                         Q. O
                                                              E  E
                                                              
-------
         a.  Boston-Worcester-Manchester. MA-NH
                                                          b. New York-Newark-Bridgeport, NY-NJ-CT-PA
    o
   Q
                                      mmY.
      O
     o
         22 2300 01 02 03 3405 C6 07 OS 09 1011 12 13 14 15 16 17 18 '92021 2223 0001
                            hour
           22 23 CO 01 02 0304050607 Oa 09 1011 12 '3 14 15 16 17 18 192021 2
                              hour
         c.  Philadelphia-Camden-Vineland. PA-NJ-DE-ME
           d, Washington-Baltimore-Northern Virginia, DC-MD-VA-WV
              1 02 03 Of 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 00 01
                            hour
           22230001 0203 04050607 0809 10 V 12 13 14 15 16 17 18 1920 2' 22230001
                              hour
         e. Atlanta-Sandy Springs-Gainesville, GA-AL
           f. Cleveland-Akron-Elyria, OH
                                          IT
                                                    o
                                                    O
         22 23 00 01 020304 05 06 07 OS 09 J0 11 12 13 14 15 16 17 18 19 2021 22 230001
                            hour
           22230001 02 03 04 0506 D? OS 09 1011 12 '3 14 15 16 17 18 1920 2
                              hour
Figure AX3-62a-f. Diurnal variability in 1-h average O3 concentrations in EPA's 12 cities.
                     Values shown represent averages from November through March, 2000
                     to 2004. Boxes define the interquartile range and the whiskers, the
                     minima and maxima.

Source: Fitz-Simons et al., (2005)
August 2005
AX3-98       DRAFT-DO NOT QUOTE OR CITE

-------
         g. Detroit-Warren-Flint, MI
           h. Chicago-Naperville-Michigan City, IL-IN-WI
0.060 •
— 0.050 •
E
a.
CL
C 0,040 •
g
1
g 0.030 •
u
o
U 0.020 - H
O
C
o
® 0010 •
0 000 •










n n n 4-j n ri



T I T


T



JLJLJL
nn
X
nil ~
rj _
-Mm









i
Mri

-
TuUUUUUuL

         22 23 CO 01 02 03 04 05 06 07 08 09 10 11 12 13 14151617-819202122230001
                            hour
           222300 01 02030405 06 070809 10 11 12 13 14 15 16 17 1B 192021 2223 0001
                              hour
         i. St. Louis-St. Charles-Farmington, MO-IL
           j. Houston-Baytown-Huntsville, TX
    0.
      i
      g-  0,150
         2223 CO 01 02 03 040506 07 0809 10 11 12 13 14 15 16 17 '8 1920 21 22 230001
                            hour
           22230001 02 030405 0607 CS 09 10 11 12 13 14 15 16 17 13 192021 222300 01
                              hour
         k. Sacramento-Arden-Arcade-Truckee, CA-NV
           g.  Los Angeles-Long Beach-Riverside, CA
C.ObO •
*£ T
g; C.06Q-
e
_o

ta
"c r 040

o r
o nri Finn
o
03
C
O C 020 •
o r
JHUHm
r nnn .
r iiii





X n n J,


JL JH I - ~ -
X n f~


LJ T
- - _ - UT T
WUDW.
T T T








n 1 I I AJhpi



~

iUuyiy^
u.uo •
E 0.130 •
ft.
>2:
c
O
+3 0.075 •
E
"c
&
O
1 o 0,050 •
O
o
C
o
Q 0.025 •
I
nmn J
T

j











Inr
:l












MM
-










_L












n H"
1









•








J
IIH
NHL









n
U
n
v









nr









in
-Jnh
n
J^,


T












n
U
n
U
T

T










11
In
1Lr
Jh







T







JL
HM
in










Iflr












in
lyy

2223 CO 01 0203 04050607 08 09 1011 12 13 14 15 16 17 "0 1920 21 22230001 22 23 00 31 02 030405 06 07 OS 09 10 11 12 13 14 15 16 17 18 19 20 21 222300 0
hour hour
Figure AX3-62g-l. Diurnal variability in 1-h average O3 concentrations in EPA's 12 cities.
                     Values shown represent averages from November through March, 2000
                     to 2004. Boxes define the interquartile range and the whiskers, the
                     minima and maxima.

Source: Fitz-Simons et al., (2005)
August 2005
AX3 -99       DRAFT-DO NOT QUOTE OR CITE

-------
         a. Boston-Worcester-Manchester, MA-NH
            b. New York-Newark-Bridgeport, NY-NJ-CT-PA
0.060 •
1= 0.050 .
a
a
c
o
'•5 0.020 •
eO
C
01

§ 0 020 -
O

-------
         g.  Detroit-Warren-Flint, Ml
            h. Chicago-Naperville-Michigan City, IL-IN-WI
   E  C.040
                                                   —  006C
         2223 DO 01 02 030405 36 07 OS 09 10 11 12 13 14 15 J6 17 18 '923 21 22230001
                            hour

jyui
ill!!!
22 23 0001 02 030405 060706 09 10 11 2 13 14
hour
II f I II T
f"| r"
MULJLIU
ITTTTTTTyT
5 18 17 18 192021 22230001
         i. St. Louis-St. Charles-Farmington, MO-IL
            j. Houston-Baytown-Huntsville, TX
         2223 DO 01 02 030405 36 07 OB 09 10 11 12 13 14 15 "6 17 18 '923 21 22230001
                            hour
            2223 0001 02 030405 060708 09 10 11 12 13 1
                               hour
                                                                                 15 16 17 18 1920 21 22230001
         k. Sacramento-Arden-Arcade-Truckee, CA-NV
            g, Los Angeles-Long Beach-Riverside, CA
    C  C.040
                                                   £  0 050 •
          9 73 0001 0? 03 0405 06070809 1011 1? J3 14 15 16 1718 19 ?0?1 ?? 73 00 01
                            hour
            22 23 00 01 02 03 04 05 OS 07 08 09 10 11 121314 IS 16 17 18 IB 20 21 22230001
                               hour
Figure AX3-63g-l. Diurnal variability in 8-h average O3 concentrations in EPA's 12 cities.
                     Values shown represent averages from November through March,
                     2000 to 2004.  Boxes define the interquartile range and the whiskers,
                     the minima and maxima. The hour refers to the start of the 8-h
                     averaging period.

Source: Fitz-Simons et al. (2005).
August 2005
AX3 -101       DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX3.5.2   Seasonal Variations in Nonurban Areas
 2           It is important to characterize the seasons in which the highest O3 concentrations would be
 3     expected to occur in nonurban areas in assessing the effects of O3 on vegetation.  It should not be
 4     assumed that highest O3 concentrations occur at all locations during the summer. For example,
 5     places where highest average O3 concentrations are observed during the spring (i.e., the months
 6     of April or May) versus the summer (Evans et al., 1983; Singh et al., 1978; Lefohn et al., 2001)
 7     are found at many national parks in the West. Figure AX3-23 shows the hourly average
 8     concentrations for Yellowstone National Park (WY) for the period of January to December 2001.
 9     Note that, at the Yellowstone National Park site, the highest hourly average concentrations tend
10     to occur during April and May. Lefohn et al. (2001) and Monks (2000) noted that this was also
11     observed for other RRMS in North America and northern Europe.
12           Differences in the timing of peak O3 concentrations may be associated with the
13     observations by Logan  (1989) that spring and summer O3 concentrations in rural areas of the
14     eastern United States are severely impacted by anthropogenic, and possibly natural emissions of
15     NOX and hydrocarbons, and that O3 episodes occur when the weather is particularly conducive to
16     photochemical formation of O3.  Taylor et al. (1992) reported that the temporal patterns
17     of O3 during quarterly or annual periods exhibited less definitive patterns at 10 forest sites in
18     North America. Based  on the exposure index selected, different patterns were reported.
19     Meagher et al. (1987) reported that for rural O3 sites in the southeastern United States, the daily
20     maximum 1-h average concentration was found to peak during the summer months. Taylor and
21     Norby (1985)
22     reported that Shenandoah National Park experienced both the highest frequency of episodes and
23     the highest mean duration of exposure events during the month of July.
24           Aneja and Li (1992) reported that the maximum monthly O3 levels at several rural sites
25     occurred in either the spring or the summer  (May to August), and the minimum occurred in the
26     fall (September and October). The timing of the maximum monthly values differed across sites
27     and years.  However, in 1988, an exceptionally high O3 concentration year occurred, and the
28     highest monthly average concentration occurred in June for almost  all of the five sites
29     investigated.  June 1988 was also the month in which the greatest number of O3 episodes
30     occurred in the  eastern United States.
        August 2005                           AX3-102     DRAFT-DO NOT QUOTE OR CITE

-------
 1          Lefohn et al. (1990a) characterized the O3 concentrations for several sites in the United
 2      States exhibiting low maximum hourly average concentrations.  Of the three western national
 3      forest sites evaluated by Lefohn et al. (1990a), Apache National Forset (AZ), Custer National
 4      Forest (MT), and Ochoco National Forest (OR), only at Apache National Forest (AZ) did
 5      maximum monthly mean concentrations occur in the spring. The Apache National Forest site
 6      was above mean nocturnal inversion height, and no decrease of concentrations occurred during
 7      the evening hours. Highest hourly maximum concentration, as well as the highest
 8      W126 O3 exposures were also found at this site. Most of the maximum monthly mean
 9      concentrations occurred in the summer at the other. Maximum monthly mean O3 concentrations
10      were found at  the White River Oil Shale site in Colorado during the spring and summer months.
11          The W126 sigmoidal weighting exposure index was also used to identify the month of
12      highest O3 exposure to vegetation. A somewhat more variable pattern was observed than when
13      the maximum monthly average concentration was used.  In some cases, the  highest W126
14      exposures occurred earlier in the year than was indicated by the maximum monthly
15      concentration.  For example, in 1979, the Custer National Forest site experienced its highest
16      W126 exposure in April, although the maximum monthly mean occurred in August.  In 1980, the
17      reverse occurred.
18          There was no consistent pattern for those sites located in the continental United States.
19      Maximum O3 exposures during the spring and summer at the Theodore Roosevelt NP, Ochoco,
20      and Custer National Forest sites and the White River Oil Shale site. The sites at which highest
21      O3 exposures occurred during the period from fall to spring did not necessarily also have the
22      lowest O3  exposures.
23
24
25      AX3.6   TRENDS IN OZONE CONCENTRATIONS
26      Evidence for Trends in Ozone Concentrations at Rural Sites in the United States
27          Year-to-year variability in the nationwide May to September, mean daily maximum 8-h O3
28      concentrations are shown in Figure AX3-64. Data flagged because of quality control issues was
29      removed with concurrence by the local monitoring agency.  Only days for which there was
30      75% data capture (i.e., 18 of 24 hours) were kept, and a minimum of 115 of 153  days (i.e.,
31      75% data capture) were required in each year.  Missing years were filled in  using simple

        August 2005                          AX3-103      DRAFT-DO NOT QUOTE OR CITE

-------
                                      Nationwide Trends, May to September
                                 Mean of Daily Maximum 8-Hour Values, 1990 - 2004
                     0.12-
                     0.11 -
                  E
                  _£  0.09 -_
                  O  0.08 -
                  £  0.07 -
                  c
                  8  0.06 -
                  c
                  O  0.05 -.
                  0)
                  =  0.04 -
                  w
                  N
                  O  0.03 -.
                     0.02-
                     0.01 -
                          1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
                                                    Year
       Figure AX3-64. Year-to-year variability in nationwide mean daily maximum 8-h O3
                       concentrations.  The whiskers on the box plot represent the 10th and
                       90th percentile concentrations.  The "X"s above and below the whiskers
                       are the values that fall below and  above the 10th and 90th percentile
                       concentrations.  The dots inside the box represent the mean, for the
                       statistic, at all sites.
       Source: Fitz-Simons et al. (2005).
 1     linear interpolation, as done in EPA Trends reports. Year-to-year variability in the
 2     corresponding 95th percentile values of the daily maximum 8-h O3 concentrations are shown in
 3     Figure AX3-65.  Sites considered in this analysis are shown in the map in Figure AX3-3.
 4     Mean O3 concentrations were slightly lower in 2003 and 2004 than in earlier years, and as was
 5     shown in Figures AX3-1 and AX3-2, most sites are located in the East.  The summer of 2003
 6     was slightly cooler than normal in the East (Levinson and Waple, 2004) and the summer of 2004
 7     was much cooler than normal in the East (Levinson, 2005) accounting in part for the dip in O3
 8     during these 2 years. Trends in compliance metrics such as the fourth highest daily maximum
 9     8-h and the second highest 1-h daily maximum can be found in the EPA Trends reports and so
10     are not repeated here.
       August 2005
AX3 -104      DRAFT-DO NOT QUOTE OR CITE

-------
                                      Nationwide Trends, May to September
                             95th Percentile of Daily Maximum 8-Hour Values, 1990 - 2004
                  E
                  Q.
                  Q.
                  C
                  o
                  C
                  0)
                  o
                  o
                  o
                  0)
                  o
                  N
                  o
0.18-
0.17-
0.16-
0.15-
0.14-
0.13-
QA2-
0.11 -
0.10-
0.09-
0.08-
0.07-
0.06-
0.05-
0.04-
0.03-
0.02-
                          1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
                                                    Year

       Figure AX3-65. Year-to-year variability in nationwide 95th percentile value of the daily
                       maximum 8-h O3 concentrations. The whiskers on the box plot represent
                       the 10th and 90th percentile values for the statistic. The "X"s above and
                       below the whiskers are the values that fall below and  above the 10th and
                       90th percentile values. The dots inside the box represent the mean, for the
                       statistic, at all sites.

       Source: Fitz-Simons et al. (2005).
 1          Figures AX3-66a-h show year-to-year variability in mean daily 8-h O3 concentrations
 2     observed at selected national park sites across the United States. Figures AX3-67a-h show year-
 3     to-year variability in the 95th percentile value of daily maximum 8-h O3 concentrations at the
 4     same sites shown in Figures AX3-66a-h. The same criteria used for calculating values in
 5     Figures AX3-64 and AX3-65 were used for calculating the May to September seasonal averages
 6     for the national parks shown in Figures AX3-66a-h and 67a-h.  Trends at these national parks are
 7     shown in Table AX3-9.  However, several monitoring sites were moved during the period from
 8     1990 to 2004. Sites were moved at Acadia NP in 1996, Joshua Tree NP in 1993, Mammoth
 9     Cave NP in 1996, Voyageurs NP in 1996, and Yellowstone NP in 1996 and offsets in O3
10     concentrations have resulted.  As a result, trends are not  shown for these sites.
       August 2005
                          AX3-105
DRAFT-DO NOT QUOTE OR CITE

-------
               May to September Mean of Daily Maximum 8-Hour Values, 1990 - 2004
     0.09 i

     0.08-

     0.07-

     0.06-

     0.05-

     0.04-

     0.03-

     0.02-
a. AcadiaNP
     0.01 V-
        1990
                        1996   1998
                          Year
                                                             1996   1998
                                                                Year
                                                        d. Shenandoah NP
     0.09 :

     0.08

     0.07 -

     0.06 -

     0.05 -

     0.04 -

     0.03 -

     002 -

     0.01 :
e. Great Smoky Mountains NP
     0.09 :

     0.08 -

     0.07 -

     0.06 -

     0.05 -

     0.04 -

     0.03 -

     0.02 -

     0.01 -
g. Cape Roma in NWR
                           Year
                                                                       1996   1998
                                                                         Year
                                                    0.09
                                           0.08 •

                                        £  0.07 :
                                        Q.
                                        ^  0.06 -
                                        _o
                                        £  0.05 -

                                        |  0.04 -

                                        5  0.03 :

                                           0.02 :

                                           0.01 :
                                               h. Mammoth Cave NP
                                                                       1996   1998
                                                                         Year
Figure AX3-66a-h.  Year-to-year variability in mean daily maximum 8-h O3 concentrations
                      at selected national park (NP), national wildlife refuge (NWR), and
                      national monument (NM) sites.

Source: Fitz-Simons et al. (2005).
August 2005
                                   AX3 -106      DRAFT-DO NOT QUOTE OR CITE

-------
              May to September Mean of Daily Maximum 8-Hour Values, 1990 - 2004

0.08-
e 0.07 -
c
a.
°: 0.06 -
c
o
1 0.05 -
I 0.04 -
E
O
0 0.03
0.02-
0 01
i. Voyageurs NP








^__, 	 . 	 . __•—-»-"'*'•---•


. — S


0.08 •
c 0.07 •
c
a.
°: 006 •
c
o
1 0.05 •
§ 0.04 •
C
O
0 0.03
0.02 •
n I"H
j. Theodore Roosevelt NP







*— *—*~_^_-^-_^-*-~V
t^^_^^^.* *-"* ~~*~~* • tr^^ ^^




1980 1992 1994 1996 1998 2000 2002 2004 1990 1992 1994 1996 1998 2000 2002 20&
Year Year
0 00 n nn

0.08 -
E 0.07 •
a.
^ 0.06 -
_o
1 0.05 •
| 0.04 -
o
0 0.03 -
0.02 •
0 01 •
k. Rocky Mountain NP



. //~*\.
^ *~ ^s>Vv«xX^\ /
•— *— \y
•/




0.08 •
E 0.07 •
a.
^ 0.06 •
_o
'| 0.05 •
$ 0.04 •
D
0 0.03 •
0.02 -
n n-i
!. Yellowstone NP





* — • — *-^_--*^* •—+—* V^
. . — r^
IT


1990 1992 1994 1996 1998 2000 2002 2004 1990 1992 1994 1996 1998 2000 2002 200
Year Year
0 09 A nn

0.08
£ 0.07 -
Q.
Q.
£ 0.06 -
o
ys
2 0.05 -
E
| 0.04 -
o
o
0.03 -
0.02 -
0 01 -
m. Glacier NP








_
, , 	 ^v J>L A~~«__ ^-«XX^\,

^s^/ 1" \ /
•V \J


0.08
£ 0.07 •
a.
a.
£ 0.06 •
o

| 0.05 •
c
| 0.04 •
o
o
0.03
0.02 •
n n-i
n. Grand Canyon NP




x*~ — *^^









1990 1992 1994 1996 1998 2000 2002 2004 1990 1992 1994 1996 1998 2000 2002 200
Year Year
0 00 A nn

0.08 •
c 0.07 -
c
a
^ 0.06 -
_o
g 0.05 -
1 0.04 -
c
o
0 0.03
0.02
n n-f
o. Sag u arc NM




^^^_^*—~m 	 /*\

^~~+~~*^







0.08 •
- 0.07 •
c
a
^ 0.06 -
_o
jj 0.05 •
1 0.04 -
c
o
0 0.03 •
0.02 •
n m
p. Chirlcahua NM






• — • — •— -^ ~~-+~~^^+-~^^~i^






            1992   1994   1996  1998  2000  2002  2004
                        Year
         1990  1992   1994   1996   1998   2000   2002  2004
                          Year
Figure AX3-66i-p.   Year-to-year variability in mean daily maximum 8-h O3 concentrations
                    at selected national park (NP), national wildlife refuge (NWR), and
                    national monument (NM) sites.

Source: Fitz-Simons et al. (2005).
August 2005
AX3 -107      DRAFT-DO NOT QUOTE OR CITE

-------
                     May to September Mean of Daily Maximum 8-Hour Values, 1990 - 2004
           0.09

           0.08 :

           0.07

           0.06:

           0.05:

           0.04:

           0.03:

           0.02 :
 . Sequoia/Kings Canyon NP
           0.01 V-
              1990
                   1992
                        1994
                              1996   1998
                                 Year
                                         2000  2002
                              1996   1998
                                 Year
           0.09 :

           0.08 -

           0.07 :

           0.06 :

           0.05 :

           0.04 :

           0.03 :

           0.02 :

           0.01
u. Olympic NP
             1990
                   1992
                        1994
                                    1998
                                         2000  2002
                                                           0.09
                                                    2004
                                                    2004
                                           0.08 •

                                           0.07 •

                                           0.06 :

                                           0.05 :

                                           0.04 -

                                           0.03 •

                                           0.02 :

                                           0.01
                                                r. Pinnacles NM
                                                             1990
                                                                   1992
                                                                        1994
                                                              1996   1998
                                                                 Year
                                                                                         2000   2002
                                                                                                    2004
                                                           0.09
                                                           0.08 •

                                                           0.07 :

                                                           0.06 :

                                                           0.05 :

                                                           0.04 •

                                                           0.0:

                                                           0.02 :

                                                           0.01
                                                               t. Denali NP
                                                              1996   1998
                                                                 Year
                                                           0.09
                                           0.08 •

                                           0.07 •

                                           0.06 :

                                           0.05

                                           0.04 -

                                           0.03 •

                                           0.02 :

                                           0.01
                                                v. Lassen Volcanic NP
                                 Year
                                                                              1996   1998
                                                                                Year
       Figure AX3-66q-v.  Year-to-year variability in mean daily maximum 8-h O3 concentrations
                             at selected national park (NP), national wildlife refuge (NWR), and
                             national monument (NM) sites.

       Source: Fitz-Simons et al. (2005).
1            As noted in The Ozone Report—Measuring Progress through 2003 (U.S. Environmental
2      Protection Agency, 2004b), O3 trends in national parks in the South and the East are similar to
3      nearby urban areas and reflect the regional nature of O3 pollution.  For example, O3 trends in
       August 2005
                                    AX3-108
DRAFT-DO NOT QUOTE OR CITE

-------
          May to September 95th Percentile of Daily Maximum 8-Hour Values, 1990 - 2004
     0.11 -

     0.10:

     0.09:

     0.08:

     0.07:

     0.06:

     0.05:

     0.04:

     0.03-
     0.02^
a. Acadia NP
                                  2000   2002  2004
0.11

0.10

0.09:

0.08-

0.07-

0.06-

0.05-

0.04-

0.03-

0.02
                                                  b. Cape Cod NS
                                                     1990   1992   1994
                                                                     1996
                                                                                2000   2002  2004
                                                                        Year
     0.11
                                                                                          2004
                                                                                          2004
        1990   1992   1994  1996   1998
                          Year
                                                                                     2002  2004
Figure AX3-67a-h.
             Year-to-year variability in 95th percentile of daily maximum 8-h O3
             concentrations at selected national park (NP), national wildlife refuge
             (NWR), and national monument (NM) sites.
Source: Fitz-Simons et al. (2005).
August 2005
                                  AX3-109
       DRAFT-DO NOT QUOTE OR CITE

-------
            May to September 95th Percentile of Daily Maximum 8-Hour Values, 1990 - 2004
       0.11 -
       0.10-
       0.09-
       0.08-
       0.07-
       0.06-
       0.05-
       0.04-
       0.03-
       0.02 JT
  i. Voyageurs NP
                                                   0.10 :
                                                   0.09:
                                                   0.08:
                                                   0.07:
                                                   0.06
                                                   0.05:
                                                   0.04:
                                                   0.03-
                                                            0.02
j. Theodore Roosevelt NP
         1990   1992   18
                                        2000   2002   2004
                                                              1990   1992   1994    1996
                                                                                             2000   2002   2004
       0.11
       0.10;
       0.09-
       0.08-
       0.07-
       0.06-
       0.05-
       0.04-
       0.03-
       0.02
           k. Rocky Mountain NP
                                                            0.11
                                                   0.10:
                                                   0.09:
                                                   0.08:
                                                   0.07:
                                                   0.06:
                                                   0.05
                                                   0.04 :
                                                   0.03:
                                                            0.02
                                                                I. Yellowstone NP
         1990   1992   1994   1996
                                        2000   2002   2004
                                                              1990   1992   1994
                                                                                 1996    1998
                                                                                   Year
                                                                                             2000   2002   2004
       0.10-
       0.09-
       0.08-
       0.07-
       0.06-
       0.05
       0.04:
       0.03-
       0.02
           m. Glacier NP
                                                   0.10:
                                                   0.09:
                                                   0.08-
                                                   0.07
                                                   0.06:
                                                   0.05-
                                                   0.04:
                                                   0.03-
                                                            0.02
                                                                n. Grand Canyon NP
         1990   1992   1994
                           1996    1998
                              Year
                                              2002   2004
                                                              1990   1992   1994
                                                                        1996    1998
                                                                          Year
                                                                                             2000   2002   2004
    E
    £  0.08:
       0.04-
       0.03-
           o. Saguaro NM
                                                            0.11
                                                   0.10:
                                                   0.09:
                                                   0.08:
                                                   0.07:
                                                   0.06:
                                                   0.05:
                                                   0.04:
                                                   0.03 :
                                                            0.02
                                                       p. Chiricahua NM
1990   1992   1994   1996    1998    2000
                     Year
                                              2002   2004
                                                              1990   1992   1994
                                                                                 1996    1998
                                                                                   Year
                                                                                             2000   2002   2004
Figure AX3-67i-p.   Year-to-year variability in 95th percentile of daily maximum 8-h O3
                         concentrations at selected national park (NP), national wildlife refuge
                         (NWR), and national monument (NM) sites.
Source: Fitz-Simons et al. (2005).
August 2005
                                         AX3 -110       DRAFT-DO NOT QUOTE OR CITE

-------
                 May to September 95th Percent!le of Daily Maximum 8-Hour Values, 1990 - 2004
           0.10-
           0.09-
           0.08-
           0.07-
           0.06-
           0.05-
           0.04-
           0.03-
           0.02
q. Sequoia/Kings Canyon NP
                                                           0.11
                              1996  1998
                                 Year
              1990
                   1992
                         1994
                              1996
                                         2000   2002
           0.11
           0.10-
           0.09-
           0.08-
           0.07-
           0.06-
           0.05-
           0.04-
           0.03-
u. Olympic NP
           0.02V—
              1990
                              1996  1998
                                 Year
0.10-
0.09-
0.08
0.07-
0.06-
0.05-
0.04-
0.03-
                                                           0.02
                                                           0.11
                                                    2004
                                                               r. Pinnacles NM
                                                              1996   1998
                                                                 Year
                                                           0.10-
                                                           0.09-
                                                           0.08-
                                                           0.07-
                                                           0.06-
                                                           0.05-
                                                           0.04
                                                           0.03-
                                                           0.02
                                                              t. Denali NP
                                                             1990
                                                                  1992
                                                                        1994
                                                                             1996
                                                                                        2000   2002
                                                                                                   2004
                                                           0.11
                                           0.10-
                                           0.09-
                                           0.08-
                                           0.07
                                           0.06-
                                           0.05-
                                           0.04-
                                           0.03-
                                                           0.02
                                               v. Lassen Volcanic NP
                                                              1996   1998
                                                                 Year
       Figure AX3-67q-v. Year-to-year variability in 95th percentile of daily maximum 8-h O3
                            concentrations at selected national park (NP), national wildlife refuge
                            (NWR), and national monument (NM) sites.
       Source: Fitz-Simons et al. (2005).
1      Charleston, SC and Charlotte, NC track those in nearby Cowpens NP and Cape Romaine NP in
2      South Carolina; O3 in Knoxville and Nashville, TN tracks O3 in Great Smoky NP; O3 in
3      Philadelphia, PA and Baltimore, MD tracks Brigantine NP in New Jersey; and New York, NY
       August 2005
                                   AX3-111
       DRAFT-DO NOT QUOTE OR CITE

-------
    Table AX3-9.  Trends in Warm Season (May to September) Daily Maximum 8-h O3 Concentrations at National Parks
                               in the United States (1990 to 2004).  Trends are given as ppb yr'1.
to
o
o
^ Site
AcadiaNP(ME)
Brigantine NWR (NJ)
Cape Cod NS (MA)
Cape Remain NWR (SC)
ChiricahuaNM(AZ)
Cowpens NB (SC)
Denali NP (AK)
Glacier NP (MT)
15 Grand Canyon NP (AZ)
rN
V° Great Smoky Mountains NP (NC-TN)
Mean
trend
0.0401
-0.802
0
0.71
0.221
0
0.17
0
0.25
0.29
p-value
0.037
0.014
0.423
0.046
0.046
0.423
0.12
0.5
0.07
0.248
P95
trend
l.O1
-1.72
0
l.O1
0.2
0.1
0.61
0.1
0
0.9
p-value
0.037
0.004
0.349
0.01
0.084
0.349
0.01
0.349
0.5
0.19
P98
trend
-0.2
-1.92
-0.5
1.0
0.15
0.4
0.61
0.27
0.13
0.4
p-value
0.349
0.003
0.19
0.07
0.218
0.349
0.002
0.19
0.218
0.423
Joshua Tree NP (CA)3

Lassen Volcanic NP (CA)

Mammoth Cave NP (KY)3
0.25
0.141
0.2
2 Downward trend, significant at p = 0.05 level.
3 Site moved. See text for details.
0.19
0.5
fe
~
H
I
o
o
0
H
O
o
H
W
O
O
HH
W
Olympic NP (WA)
Pinnacles NM (CA)

Rocky Mountain NP (CO)
Saguaro NM (AZ)
Sequoia/Kings Cany on NP (CA)
ShenandoahNP(VA)
Theodore Roosevelt NP (ND)
Voyageurs NP (MN)3
Yellowstone NP (WY)3
1 ' T
0.14
-0.1

0.911
-0.2
0.38
0
0.381
—
—
:,al
0.141
0.218

0.004
0.279
0.218
0.461
0.023
—
—

0.31
-0.5

l.O1
-0.3
0
-0.2
0.2
—
—

0.037
0.07

0.014
0.19
0.461
0.385
0.19
—
—

0.2
-0.56

0.88
-0.38
0
0.33
0.2
—
—

0.19
0.057

0.07
0.141
0.539
0.279
0.141
—
—


-------
 1      and Hartford, CT track O3 in Cape Cod NS. The situation is not as clear in the West, where
 2      national parks are affected differently by pollution sources that are located at varying distances
 3      away (e.g., Lassen Volcanic National Park and Yosemite National Park, CA).  However, data
 4      obtained at these sites still provide valuable information about the variability in regional
 5      background concentrations, especially since the West has not been broken down into regions as
 6      has been done by Lehman et al. (2004) and shown in Figure AX3-7.
 7           Caution should be exercised in using trends calculated at national parks to infer
 8      contributions from distant sources either inside or outside of North America, because of the
 9      influence of local and regional pollution.  For example, using a 15-year record of O3 from Lassen
10      Volcanic National Park, a rural elevated  site in northern California; data from two aircraft
11      campaigns;  and observations spanning 18 years from five U.S. West Coast, marine boundary
12      layer sites, Jaffe et al. (2003) reported that O3 in air arriving from the eastern Pacific in spring
13      has increased by approximately 10 ppb from the mid-1980s. They concluded that this positive
14      trend is due to increases of emissions of O3 precursors in Asia. They found positive trends in O3
15      in all seasons.  They also noted that diurnal variations during summer were about 21 ppb, but
16      only about 6 ppb during spring. Although Lassen Volcanic National Park site is not close to any
17      major emission sources or urban centers, the site experiences maximum hourly average O3
18      concentrations  above 0.080 ppm during April to May and above 0.100 ppm during the summer
19      (U.S. Environmental Protection Agency, 2003a), suggesting local  photochemical production, at
20      least during summer.  However, local springtime photochemical production cannot be ruled out.
21      The authors suggested that the likely cause for the spring increases is transport from Asia,
22      because emissions of precursors have decreased in California over the monitoring period. The
23      springtime increases appears to be inconsistent with the summer increases, when there is
24      evidence for the occurrence of more localized photochemical activity. Although emissions of O3
25      precursors may have decreased in California as a whole over the monitoring period, there still
26      may be regional increases in areas that could  affect air quality in Lassen.
27
28
29
        August 2005                            AX3-113      DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX3.7   RELATIONS BETWEEN OZONE, OTHER OXIDANTS, AND
 2               OXIDATION PRODUCTS
 3           Tables of measurements of PAN and peroxypropionyl nitrate (PPN, CH3CH2C(O)OONO2)
 4      concentrations were given in the 1996 O3 AQCD (U.S. Environmental Protection Agency,
 5      1996a). Measurements were summarized for rural and urban areas in the United States, Canada,
 6      France, Greece, and Brazil.  The use of measurements from aboard serve to illustrate or support
 7      certain U.S. results as well as to demonstrate the widespread presence of PANs in the
 8      atmosphere. Additional data for H2O2 were also presented in the 1996 O3 AQCD.  Data for these
 9      species are obtained as part of specialized field studies and not as part of routine monitoring
10      operations and thus are highly limited in their ranges of applicability. As a result, it is difficult
11      to relate the concentrations of O3, other oxidants, and oxidation products on the basis of rather
12      sparse data sets.  This information is simply not available for a large number of environments.
13      Instead, it might be more instructive to examine the relations between O3 and other products of
14      atmospheric reactions on the basis of current understanding of atmospheric photochemical
15      processes.
16           In order to understand co-occurrence between atmospheric species, an important
17      distinction must be made between primary (directly emitted)  species and secondary
18      (photochemically produced) species.  In general, it is more likely that primary species will be
19      more highly correlated with each other, and that secondary species will be more highly
20      correlated with each other than will species from mixed classes.  By contrast, primary and
21      secondary species are less likely to be correlated with each other. Secondary reaction products
22      tend to correlate with each other, but there is considerable variation. Some species (e.g., O3 and
23      organic nitrates) are closely related photochemically and correlate with each other strongly.
24      Others (e.g., O3 and H2O2) show a more complex correlation pattern.
25           Although NO2 is produced mainly by the reaction of directly emitted NO with O3 with
26      some contributions from direct emissions, in practice, it behaves like a primary species. The
27      timescale for conversion of NO to NO2 is fast (5 min or less), so NO and NO2 ambient
28      concentrations rapidly approach values determined by the photochemical steady state.  The sum
29      NO + NO2 (NOX)  behaves like a typical primary species,  while NO and NO2 reflect some
30      additional complexity based on photochemical interconversion. As a primary species, NO2
31      generally does not correlate with O3  in urban environments. In addition, chemical  interactions

        August 2005                             AX3-114     DRAFT-DO NOT QUOTE OR CITE

-------
 1      among O3, NO and NO2 have the effect of converting O3 to NO2 and vice versa, which can result
 2      in a significant anti-correlation between O3 and NO2.
 3           Organic nitrates consist of PAN, a number of higher-order species with photochemistry
 4      similar to PAN (e.g., PPN), and species such as alkyl nitrates with somewhat different
 5      photochemistry.  These species are produced by a photochemical process very similar to that
 6      of O3. Photochemical production is initiated by the reaction of primary and secondary VOCs
 7      with OH radicals, the resulting organic radicals subsequently react with NO2 (producing PAN
 8      and analogous species) or with NO (producing alkyl nitrates). The same sequence (with organic
 9      radicals reacting with NO) leads to the formation of O3.
10           In addition, at warm temperatures, the concentration of PAN forms a photochemical steady
11      state with its radical precursors on a timescale of roughly 30 minutes.  This steady state  value
12      increases with the ambient concentration of O3 (Sillman et al., 1990).  Ozone and PAN may
13      show different seasonal cycles, because they are affected differently by temperature.  Ambient
14      O3  increases with temperature, driven in part by the photochemistry of PAN (see description
15      above). By contrast, the photochemical lifetime of PAN decreases rapidly with increasing
16      temperature. The ratio O3/PAN should show seasonal changes, with highest ratios in summer,
17      although there is no evidence from measurements. Measured ambient concentrations
18      (Figures AX3-68a-d) show a strong correlation between  O3 and PAN, and between O3 and other
19      organic nitrates (Pippin et al., 2001; Roberts et al., 1998).
20           Individual primary VOCs are generally highly correlated with each other and with NOX
21      (Figure AX3-69).  A summary of the results of a number of field studies of the concentrations of
22      precursors including NOX and nonmethane organic compounds (NMOCs) are summarized in the
23      1996O3AQCD.
24           Formation of H2O2 takes place by self-reaction of photochemically generated HO2 radicals,
25      so that there is large seasonal variation of H2O2 concentrations and values in excess of 1 ppb are
26      mainly limited to the summer months (Kleinman,  1991).  Although H2O2 is produced from
27      photochemistry that is closely related to O3, it does not show a consistent pattern of correlation
28      with O3. Hydrogen peroxide is produced in abundance along with O3 only when O3 is produced
29      under NOx-limited conditions.  When the photochemistry is NOx-saturated much less H2O2 is
30      produced.  In addition, increasing NOX tends to slow the  formation of H2O2 under NOx-limited
31      conditions.

        August 2005                            AX3-115     DRAFT-DO NOT QUOTE  OR CITE

-------
                 140
                    0   1000 2000 3000  4000   0    1000 2000  3000  4000  5000
                                            PAN (pptv)
      Figure AX3-68a-d.  Measured O3 (ppbv) versus PAN (pptv) in Tennessee, including
                         (a) aircraft measurements, and (b, c, and d) suburban sites near
                         Nashville.
      Source:  Roberts et al. (1998).
1          Measurements of gas phase peroxides in the atmosphere were reviewed by Lee et al.
2     (2000). Ground level measurements of H2O2 taken during the 1970s indicated values of 180 ppb
3     in Riverside, CA and 10 to 20 ppb during smog episodes in Claremont and Riverside, with
4     values approaching 100 ppb in forest fire plumes. However, later surface measurements always
5     found much lower values. For example, in measurements made in Los Angeles and nearby areas
6     in the 1980s, peak values were always less than about 2 ppb and in a methods intercomparison
      August 2005
AX3 -116     DRAFT-DO NOT QUOTE OR CITE

-------
                                1.5-
• Sp2< 10 ppbvi
o S(D2>"10"ppbvi"
                                0,0
   20
                                              40    60    80
                                                NOv (ppbv)
100
120
       Figure AX3-69.  Measured correlation between benzene and NOy at a measurement site in
                        Boulder, CO. Instances with SO2 >10 ppb are identified separately (open
                        circles), because these may reflect different emission sources.
       Source: Goldan et al. (1995).
 1     study in Research Triangle Park, NC in June 1986, concentrations were <2.5 ppbv. Higher
 2     values ranging up to 5 ppb were found in a few other studies in Kinterbish, Alabama and
 3     Meadview, Arizona. Several of these studies found strong diurnal variations (typically about a
 4     factor three) with maximum values in the mid-afternoon and minimum values in the early
 5     morning. Mean concentrations of organic hydroperoxides at the surface at Niwot Ridge, CO in
 6     the summer of 1988 and State Park, GA during the summer of 1991 were all less than a few ppb.
 7           Early aircraft measurements of H2O2 over the eastern United States were reported by
 8     Heikes et al. (1987). More recent aircraft measurements of hydroperoxide (H2O2, CH3OOH and
 9     HOCH2OOH) concentrations were made as part of the Southern Oxidants  Study intensive
10     campaign in Nashville, TN in July 1995 (Weinstein-Lloyd et al., 1998). The median
11     concentration of total hydroperoxides in the boundary layer between 1100 and 1400 CDT was
12     about 5 ppbv, with more than 50% contribution from organic hydroperoxides. Median O3 was
       August 2005
        AX3 -117     DRAFT-DO NOT QUOTE OR CITE

-------
 1      about 70 ppbv at the same time. The concentrations of the hydroperoxides depended strongly on
 2      wind direction. For example, values were about 40% lower when winds originated from the
 3      N/NW as opposed to the S/SW.
 4          Elevated O3 is generally accompanied by elevated HNO3, although the correlation is not as
 5      strong as between O3 and organic nitrates. Ozone often correlates with HNO3, because they have
 6      the same precursor (NOX). However, HNO3 can be produced in significant quantities in winter,
 7      even when O3 is low.  The ratio between O3 and HNO3 also shows great variation in air pollution
 8      events, with NOx-saturated environments having much lower ratios of O3 to HNO3 (Ryerson
 9      et al., 2001).  Aerosol  nitrate is formed primarily by the combination of nitrate (supplied
10      by HNO3) with ammonia, and may be limited by the availability of either nitrate or ammonia.
11      Nitrate is expected to correlate loosely with O3 (see above), whereas ammonia is not expected to
12      correlate with O3.
13          In addition to nitrate, other oxidants are present in airborne cloud droplets, rain drops and
14      particulate matter.  Measurements of hydroperoxides, summarized by Reeves and Penkett
15      (2003), are available mainly for hydrometeors, but are sparse for ambient particles.
16      Venkatachari et al. (2005a) sampled the concentrations of total reactive oxygen species (ROS) in
17      particles using a cascade impactor in Rubidoux, CA during July 2003. Although the species
18      constituting ROS were not identified, the results were reported in terms of equivalent H2O2
19      concentrations. Unlike O3 and gas phase H2O2 which show strong diurnal variability (i.e., about
20      a factor of three variation between afternoon maximum and early morning minimum), the
21      diurnal variation of particle phase ROS was found to be much weaker (i.e., less than about 20%)
22      at least for the time between 8 a.m. and midnight.  Because the ROS were measured in the fine
23      aerosol size fraction, which has a lifetime with respect to deposition of much greater than a day,
24      little loss is expected but their concentrations might also be expected to increase because of
25      nighttime chemistry, perhaps involving NO3 radicals. The concentration of ROS, expressed as
26      equivalent H2O2 (5.2 to 6.1 x 10'7 M/m3, ranged from 20% to 100% that of O3 (diurnal average:
27      30%), with highest values at night.  The ratio was likely higher at the early morning minimum
28      for O3. In a companion study conducted in Queens, NY during January and early February 2004,
29      Venkatachari et al. (2005b) found much lower concentrations of ROS  of about 1 x 10"7 M/m3.
30      However, O3 levels were also substantially lower leading to ROS concentrations about 20%
31      those of O3. It is of interest to note that gas phase OH concentrations measured at the same time

        August 2005                            AX3-118     DRAFT-DO NOT QUOTE OR CITE

-------
 1     ranged from about 7.5 x lOVcm3 to about 1.8 x lOVcm3, implying the presence of significant
 2     photochemical activity even in New York City during winter.
 3          Peroxyacetylnitrate (PAN) is produced during the photochemical oxidation of a wide range
 4     of VOCs in the presence of NOX. It is removed by thermal decomposition and also by uptake to
 5     vegetation (Sparks et al., 2003; Teklemariam and Sparks, 2004). PAN is the dominant member
 6     of the broader family of peroxyacylnitrates (PANs), which includes as other significant
 7     atmospheric components peroxypropionyl nitrate (PPN) of anthropogenic origin and
 8     peroxymethacrylic nitrate  (MPAN) produced from  oxidation of isoprene.  Measurements and
 9     models show that PAN in  the United States includes major contributions from both
10     anthropogenic and biogenic VOC precursors (Horowitz et al., 1998; Roberts et al.,  1998).
11     Measurements in Nashville during the 1999 summertime Southern Oxidants Study  (SOS)
12     showed PPN and MPAN amounting to 14% and 25% of PAN, respectively (Roberts et al.,
13     2002). Measurements during the TexAQS 2000 study in Houston indicated PAN concentrations
14     of up to 6.5 ppbv (Roberts et al., 2003). PAN measurements in  southern California during the
15     SCOS97-NARSTO study  indicated peak concentrations of 5-10 ppbv, which can be contrasted to
16     values of 60-70 ppbv measured back in 1960 (Grosjean, 2003).  Vertical profiles measured from
17     aircraft over the United States and off the Pacific coasts show PAN concentrations  above the
18     boundary layer of only a few hundred pptv, although there are significant enhancements
19     associated with long-range transport of pollution plumes including from Asia (Kotchenruther
20     et al., 2001a; Roberts et al., 2004).  Decomposition of this anthropogenic PAN as it subsides
21     over North America can lead to significant O3 production, enhancing the O3 background
22     (Kotchenruther et al., 200Ib; Hudman et al., 2004).
23          Relations between primary and secondary components discussed above are illustrated by
24     considering data for O3 and PM2 5. Ozone and PM2 5 concentrations observed at a monitoring site
25     in Fort Meade, MD are plotted as binned means in Figure AX3-70.  These data were collected
26     between July 1999 and July 2001. As can be seen from the figure, PM2 5 tends to be anti-
27     correlated with O3 to the left of the inflection point (at about 30  ppbv O3) and PM25 tends to be
28     positively correlated with  O3 to the right of the inflection point.  Data to the left of the minimum
29     in PM25 were collected mainly during the cooler months of the year, while data to the right of
30     the minimum were collected during the warmer months.  This situation  arises because PM2 5
31     contains a large secondary component during the summer and has a larger primary  component

       August 2005                           AX3-119     DRAFT-DO NOT QUOTE OR CITE

-------
    40

    35"

*E  30"
|  25-
i1
°-  20 f
 c
 (0
S  15 t
T3
 0)
 C
.E  10 f
m
                                I.Wit
                                10       20      30       40      50
                                          Binned Mean 03 (ppbv)
                                                         60
70
       Figure AX3-70.  Binned mean PM2 5 concentrations versus binned mean O3 concentrations
                        observed at Fort Meade, MD from July 1999 to July 2001.
       Source: Chen (2002).
 1     during winter. During the winter, O3 comes mainly from the free troposphere, above the
 2     planetary boundary layer and, thus, may be considered a tracer for relatively clean air.
 3     Unfortunately, data for PM2 5 and O3 are collected concurrently at relatively few sites in the
 4     United States throughout an entire year, so these results, while highly instructive are not readily
 5     extrapolated to areas where appreciable photochemical activity occurs throughout the year.  Ito
 6     et al. (2005) showed the relation between PM10 and O3 on a personal basis in several urban areas
 7     (cf, Figure 7-24). Although PM10 contains proportionately more primary material than
 8     does PM2 5, relations similar to those shown in Figure AX3-70 are found, reflecting the dominant
 9     contribution from PM2 5 to PM10.
10
       August 2005
                              AX3 -120      DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX3.8   RELATIONSHIP BETWEEN SURFACE OZONE AND
 2               OTHER POLLUTANTS
 3     AX3.8.1  Introduction
 4          Several attempts have been made to characterize gaseous air pollutant mixtures (Lefohn
 5     and Tingey, 1984; Lefohn et al., 1987).  The characterization of co-occurrence patterns under
 6     ambient conditions is important for relating human health and vegetation effects to controlled
 7     chamber studies and to ambient conditions. Lefohn et al. (1987) discussed the various patterns
 8     of pollutant exposures. Pollutant combinations can occur at or above a threshold concentration
 9     either together or temporally separated from one another. Patterns that show air pollutant pairs
10     appearing at the same hour of the day at concentrations equal to or greater than a minimum
11     hourly mean value were defined as simultaneous-only daily co-occurrences.  When pollutant
12     pairs occurred  at or above a minimum concentration during the 24-h period, without occurring
13     during the same hour, a "sequential-only" co-occurrence was defined. During a 24-h period, if
14     the pollutant pair occurred at or above the minimum level at the same hour of the day and at
15     different hours during the period, the co-occurrence pattern was defined as "complex-
16     sequential."
17          For characterizing the different types of co-occurrence patterns for O3/NO2, O3/SO2,
18     and NO2/SO2, Lefohn and Tingey (1984) used a 0.05 ppm threshold to identify the number of
19     hourly simultaneous-only co-occurrences for the period May through September at a large
20     number of air quality urban monitoring sites along with rural sites. The selection of a 0.05-ppm
21     threshold concentration was based on vegetation effects considerations.  Data used in the
22     analysis included hourly averaged (1) Environmental Protection Agency Storage and Retrieval
23     of Aerometric Data (SAROAD; now AQS) data for 1981, (2) EPRI-SURE and Eastern Regional
24     Air Quality Study (ERAQS) data for 1978 and 1979, and (3) Tennessee Valley Authority (TVA)
25     data from 1979 to 1982. Lefohn and Tingey (1984) concluded, for the pollutant combinations,
26     that (1) the co-occurrence of two-pollutant mixtures lasted only a few hours per episode and (2)
27     the time interval between episodes was generally large (weeks, sometimes months).
28          Lefohn et al. (1987), using a 0.03-ppm threshold, grouped air quality data from rural and
29     RRMS (as characterized in the EPA database) within a 24-h period starting at 0000 hours and
30     ending at 2359 hours. Data were analyzed for the May to September period. Data used in the
       August 2005                            AX3-121     DRAFT-DO NOT QUOTE OR CITE

-------
 1      analysis included hourly averaged (1) Environmental Protection Agency AQS (SAROAD) data
 2      from 1978 to 1982, (2) EPRI-SURE and -ERAQS data for 1978 and 1979, and (3) TVA data
 3      from 1979 to 1982. Patterns that showed air pollutant pairs appearing at the same hour of the
 4      day at concentrations equal to or greater than a minimum hourly mean value were defined as
 5      simultaneous-only daily co-occurrences.  When pollutant pairs occurred at or above a minimum
 6      concentration during the 24-h period,  without occurring during the same hour, a "sequential-
 7      only" co-occurrence was defined.  During a 24-h period, if the pollutant pair occurred at or
 8      above the minimum level at the same  hour of the day and at different hours during the period,
 9      the co-occurrence pattern was defined as "complex-sequential." A co-occurrence was not
10      indicated if one pollutant exceeded the minimum concentration just before midnight and the
11      other pollutant exceeded the minimum concentration just after midnight. As will be discussed
12      below, studies of the joint occurrence of gaseous NO2/O3 and SO2/O3 reached two conclusions:
13      (1) hourly simultaneous and  daily simultaneous-only co-occurrences are fairly rare and (2) when
14      co-occurrences are present, complex-sequential and sequential-only co-occurrence patterns
15      predominate. The authors reported that year-to-year variability was found to be insignificant;
16      most of the monitoring sites  experienced co-occurrences of any type less than  12% of the
17      153 days.
18           Since 1999, monitoring stations across the United States have been routinely measuring the
19      24-h average concentrations  of PM25. Because of the availability of the PM25 data, daily
20      co-occurrence of PM25 and O3 over a  24-h period was characterized. Because PM25 data are
21      mostly summarized as 24-h average concentrations in the AQS data base, a daily co-occurrence
22      of O3 and PM25 was subjectively defined as when an hourly average O3 concentration >0.05 ppm
23      and a PM2 5 24-h concentration >40 |ig/m3 occurred over the same 24-h period.
24           For exploring the co-occurrence of O3 and other pollutants (e.g., acid precipitation and
25      acidic cloudwater), limited data are available. In most cases, routine monitoring data are not
26      available from which to draw general  conclusions.  However, published results are reviewed and
27      summarized for the purpose  of assessing an estimate of the possible importance of co-occurrence
28      patterns of exposure.
29
        August 2005                            AX3-122     DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX3.8.2  Co-Occurrence of Ozone with Nitrogen Oxides
 2          Ozone occurs frequently at concentrations equal to or greater than 0.05 ppm at many rural
 3     and remote monitoring sites in the United States (U.S. Environmental Protection Agency,
 4     1996a). Therefore, for many rural locations in the United States, the co-occurrence patterns
 5     observed by Lefohn and Tingey (1984) for O3 and NO2 were defined by the presence or absence
 6     of NO2. Lefohn and Tingey (1984) reported that most of the sites analyzed experienced fewer
 7     than 10 co-occurrences (when both pollutants were present at an hourly average concentration
 8     >0.05 ppm). Figure AX3-71 summarizes the simultaneous co-occurrence patterns reported by
 9     Lefohn and Tingey (1984).  The authors noted that several urban monitoring sites in the South
10     Coast Air Basin experienced more than 450 co-occurrences.  For more moderate areas of the
11     country, Lefohn et al.  (1987) reported that even with a threshold of 0.03 ppm O3, the number of
12     co-occurrences with NO2 was small.
                                    100

                                     80
                                  «
                                 «5  60
                                  o
                                  J2
                                     40
                                     20
                                             ll
    +++-
I
                                             OCCOOCOO
                                               CM O ^ IO tD I*- 00
                                         Number of Co-Occurrences
       Figure AX3-71.   The co-occurrence pattern for O3 and NO2.
       Source: Lefohn and Tingey (1984).
       August 2005
AX3 -123     DRAFT-DO NOT QUOTE OR CITE

-------
 1          Using 2001 data from the U.S. EPA AQS database, patterns that showed air pollutant pairs
 2     of O3/NO2 appearing at the same hour of the day at concentrations >0.05 ppm were
 3     characterized. The data were not segregated by location settings categories (i.e., rural, suburban,
 4     and urban and center city) or land use types (i.e., agricultural, commercial, desert, forest,
 5     industrial, mobile, or residential). Data capture was not a consideration in the analysis. The data
 6     were characterized over the EPA-defined O3 season (Table AX3-1). In 2001, there were
 7     341 monitoring sites that co-monitored O3  and NO2.  Because of possible missing hourly average
 8     concentration data during periods when co-monitoring may have occurred, no attempt was made
 9     to characterize the number of co-occurrences in the 0 category.  Thus, co-occurrence patterns
10     were identified for those monitoring sites that experienced one or more co-occurrences.
11          Figure AX3-72 illustrates the results  of the analysis. Similar to the analysis summarized
12     by Lefohn and Tingey (1984), most of the  collocated monitoring sites analyzed, using the 2001
13     data, experienced fewer than 10 co-occurrences (when both pollutants were present at an hourly
14     average concentration >0.05 ppm).
15
35-
30-
                in
                o
                w  25-
                !2
-------
1     AX3.8.3  Co-Occurrence of Ozone with Sulfur Dioxide
2          Because elevated SO2 concentrations are mostly associated with industrial activities (U.S.
3     Environmental Protection Agency, 1992), co-occurrence observations are usually associated
4     with monitors located near these types of sources. Lefohn and Tingey (1984) reported that,
5     for the rural and nonrural monitoring sites investigated, most sites experienced fewer than
6     10 co-occurrences of SO2 and O3.  Lefohn et al. (1987) reported that even with a threshold of
7     0.03 ppm O3, the number of co-occurrences with SO2 was small. Figure AX3-73 illustrates the
8     simultaneous co-occurrence results reported by Lefohn and Tingey (1984).
1 UU
80
(A
J75 60
•5
o
o
f 40
3
Z
20
n
-
-
_






















II..
                                              oooooooo
                                         Number of Co-Occurrences

       Figure AX3-73. The co-occurrence pattern for O3 and SO2.
       Source: Lefohn and Tingey (1984).
1          Meagher et al. (1987) reported that several documented O3 episodes at specific rural
2     locations appeared to be associated with elevated SO2 levels. The investigators defined the
3     co-occurrence of O3 and SO2 to be when hourly mean concentrations were >0.10 and 0.01 ppm,
4     respectively.
      August 2005
AX3-125     DRAFT-DO NOT QUOTE OR CITE

-------
 1           The above discussion was based on the co-occurrence patterns associated with the presence
 2      or absence of hourly average concentrations of pollutant pairs.  Taylor et al. (1992) have
 3      discussed the joint occurrence of O3, nitrogen, and sulfur in forested areas using cumulative
 4      exposures of O3 with data on dry deposition of sulfur and nitrogen. The authors concluded in
 5      their study that the forest landscapes with the highest loadings of sulfur and nitrogen via dry
 6      deposition tended to be the same forests with the highest average O3 concentrations and largest
 7      cumulative exposure. Although the authors concluded that the joint occurrences of multiple
 8      pollutants in forest landscapes were important, nothing was mentioned about the hourly
 9      co-occurrences of O3 and SO2 or of O3 and NO2.
10           Using 2001 data from the EPA AQS database, patterns that showed air pollutant pairs
11      of O3/SO2 appearing at the same hour of the day at concentrations >0.05 ppm were characterized.
12      The data were not segregated by location settings categories (i.e., rural, suburban, and urban and
13      center city) or land use types (i.e., agricultural, commercial, desert, forest, industrial, mobile,
14      or residential).  Data capture was not a consideration in the analysis.  In 2001, there were
15      246 monitoring sites that co-monitored O3 and SO2.  As discussed previously, because of
16      possible missing hourly  average concentration data during periods when co-monitoring may
17      have occurred,  no attempt was made to characterize the number of co-occurrences in the 0
18      category.  Thus, co-occurrence patterns were identified for those monitoring sites that
19      experienced one or more co-occurrences. Figure AX3-74 shows the results from this analysis
20      for the simultaneous co-occurrence of O3 and SO2. Similar to the analysis summarized by
21      Lefohn and Tingey (1984), most of the collocated monitoring sites analyzed, using the 2001
22      data, experienced fewer than 10 co-occurrences (when both pollutants were present at an hourly
23      average concentration >0.05  ppm).
24
25      AX3.8.4  Co-Occurrence of Ozone and Daily PM2 5
26           Using 2001 data from the EPA AQS, the daily co-occurrence of PM2 5 and O3 over a 24-h
27      period was characterized.  There were 362 sites where PM25 and O3 monitors were collocated.
28      As described in the introduction selection of this annex, a daily co-occurrence of O3 and PM25 is
29      subjectively defined as an hourly average O3 concentration >0.05 ppm and a PM2 5 24-h
30      concentration >40 |ig/m3 occurring over the same 24-h period.  Figure AX3-75 illustrates the
        August 2005                            AX3-126      DRAFT-DO NOT QUOTE OR CITE

-------
                 60
                 50-|
              1  40 H
              2  30-
              |  20-
              z
                 10-
                  0
                         OOOOOOOOO
                                                          o   o  o  o   o   o
                                                          o   T-  CM  co   ^   LO
                                   Number of Co-Occurrences (Hours)
      Figure AX3-74. The co-occurrence pattern for O3 and SO2 using 2001 data from AQS.
                160
                140-
              $ 120-
              w 100-
              2  80-
              
-------
 1      AX3.8.5   Co-Occurrence of Ozone with Acid Precipitation
 2           Concern has been expressed about the possible effects on vegetation from co-occurring
 3      exposures of O3 and acid precipitation (Prinz et al., 1985; National Acid Precipitation
 4      Assessment Program, 1987; Prinz and Krause, 1989). Little information has been published
 5      concerning the co-occurrence patterns associated with the joint distribution of O3 and acidic
 6      deposition (i.e., H+). Lefohn and Benedict (1983) reviewed the EPA SAROAD monitoring data
 7      for 1977 through 1980 and, using National Atmospheric Deposition Program (NADP) and EPRI
 8      wet deposition data, evaluated the frequency distribution  of pH events for 34 NADP and 8 EPRI
 9      chemistry monitoring sites located across the United States. Unfortunately, there were few sites
10      where O3 and acidic deposition were co-monitored.
11           As a result, Lefohn and Benedict (1983) focused their attention on O3 and acidic deposition
12      monitoring sites that were closest to one another. In some cases, the sites were as far apart as
13      144 km. Using hourly O3 monitoring data and weekly and event acidic deposition data from the
14      NADP and EPRI databases, the authors identified specific locations where the hourly mean O3
15      concentrations were >0.1 ppm and 20% of the wetfall daily or weekly samples were below pH
16      4.0. Elevated levels of O3 were defined as hourly mean concentrations equal to or greater than
17      0.1 ppm. Although for many cases,  experimental research results of acidic deposition on
18      agricultural crops show few effects at pH levels >3.5 (National Acid Precipitation Assessment
19      Program, 1987), it was decided to use  a pH threshold of 4.0 to take into consideration the
20      possibility of synergistic  effects between O3 and acidic deposition.
21           Based on their analysis, Lefohn and Benedict (1983) identified five sites with the potential
22      for agricultural crops to experience additive, less than additive, or synergistic (i.e., greater than
23      additive) effects from elevated O3 and H+ concentrations.  The authors stated that they believed,
24      based on the available data, the greatest potential for interaction between acid rain and O3
25      concentrations in the United States, with possible effects  on crop yields, may be in the most
26      industrialized areas (e.g., Ohio and Pennsylvania). However, they cautioned that, because no
27      documented evidence existed to show that pollutant interaction had occurred under field growth
28      conditions and ambient exposures, their conclusions  should only be used as a guide for further
29      research.
30           In their analysis, Lefohn  and Benedict (1983) found no collocated sites. The authors
31      rationalized that data from non-co-monitoring sites (i.e., O3 and acidic deposition) could be used

        August 2005                            AX3-128     DRAFT-DO NOT QUOTE OR CITE

-------
 1      because O3 exposures are regional in nature. However, work by Lefohn et al. (1988) has shown
 2      that hourly mean O3 concentrations vary from location to location within a region, and that
 3      cumulative indices, such as the percent of hourly mean concentrations >0.07 ppm, do not form a
 4      uniform pattern over a region.  Thus, extrapolating hourly mean O3 concentrations from known
 5      locations to other areas within a region may provide only qualitative indications of actual O3
 6      exposure patterns.
 7           In the late 1970s and the 1980s, both the private sector and the government funded research
 8      efforts to better characterize gaseous air pollutant concentrations and wet deposition. The event-
 9      oriented wet deposition network, EPRI/Utility Acid Precipitation Study Program, and the weekly
10      oriented sampling network, NADP, provided information that can be compared with hourly
11      mean O3 concentrations collected at several co-monitored locations. No  attempt was made to
12      include H+ cloud deposition information.  In some cases, for mountaintop locations (e.g.,
13      Clingman's Peak, Shenandoah, Whiteface Mountain, and Whitetop Mountain), the H+ cloud
14      water deposition is greater than the H+ deposition in precipitation (Mohnen, 1989), and the
15      co-occurrence patterns associated with O3 and cloud deposition will be different from those
16      patterns associated with O3 and deposition by precipitation.
17           Smith and Lefohn (1991) explored the relationship between O3 and H+ in precipitation,
18      using data from sites that monitored both O3 and wet deposition simultaneously and within
19      one-minute latitude and longitude of each other. The authors reported that individual sites
20      experienced years in which both H+ deposition and total O3 exposure were at least moderately
21      high (i.e., annual H+ deposition >0.5 kg ha"1 and an annual O3 cumulative, sigmoidally weighted
22      exposure (W126) value  >50 ppm-h).  With data compiled from all sites, it was found that
23      relatively acidic precipitation (pH <4.31 on a weekly basis or pH <4.23 on a daily basis)
24      occurred together with relatively high  O3 levels (i.e., W126 values >0.66 ppm-h for  the same
25      week or W126 values >0.18 ppm-h immediately before or after a rainfall event) approximately
26      20% of the time, and highly acidic precipitation (i.e, pH <4.10 on a weekly basis or  pH <4.01 on
27      a daily basis) occurred together with a high O3 level (i.e.,  W126 values > 1.46 ppm-h for the
28      same week or W126 values >0.90 ppm-h immediately before or after a rainfall event)
29      approximately 6% of the time.  Whether during the same  week or before, during, or  after a
30      precipitation event, correlations between O3 level and pH (or H+ deposition) were weak to
        August 2005                            AX3-129     DRAFT-DO NOT QUOTE OR CITE

-------
 1     nonexistent. Sites most subject to relatively high levels of both H+ and O3 were located in the
 2     eastern United States, often in mountainous areas.
 3
 4     AX3.8.6  Co-Occurrence of Ozone with Acid Cloudwater
 5          In addition to the co-occurrence of O3 and acid precipitation, results have been reported on
 6     the co-occurrence of O3 and acidic cloudwater in high-elevation forests.  Vong and Guttorp
 7     (1991) characterized the frequent O3-only and pH-only, single-pollutant episodes, as well as the
 8     simultaneous and sequential co-occurrences of O3 and acidic cloudwater. The authors reported
 9     that both simultaneous and sequential co-occurrences were observed a few times each month
10     above the cloud base. Episodes were classified by considering hourly O3 average concentrations
11     >0.07 ppm and cloudwater events with pH <3.2. The authors reported that simultaneous
12     occurrences of O3 and pH episodes occurred two to three times per month at two southern sites
13     (Mitchell, NC and Whitetop, VA) and the two northern sites (Whiteface Mountain, NY and
14     Moosilauke, NH) averaged  one episode per month.  No co-occurrences were observed at the
15     central Appalachian site (Shenandoah, VA), due to a much lower cloud frequency. Vong and
16     Guttorp (1991) reported that the simultaneous occurrences were usually of short duration
17     (mean =1.5 h/episode) and  were followed by an O3-only episode. As would be expected,
18     O3-only episodes were longer than co-occurrences and pH episodes, averaging an 8-h duration.
19
20
21     AX3.9   THE METHODOLOGY FOR DETERMINING POLICY
22               RELEVANT  BACKGROUND OZONE CONCENTRATIONS
23     AX3.9.1  Introduction
24          Background O3 concentrations used for NAAQS-setting purposes are referred to as Policy
25     Relevant Background (PRB) O3  concentrations.  Policy Relevant Background concentrations are
26     those concentrations that would result in the United States in the absence of anthropogenic
27     emissions in continental North America (the United Sates, Canada and Mexico). Policy
28     Relevant Background concentrations include contributions from natural sources everywhere in
29     the world and from anthropogenic sources outside these three countries.  For the purposes of
30     informing decisions about O3 NAAQS, EPA assesses risks to human health and environmental
       August 2005                           AX3-130     DRAFT-DO NOT QUOTE OR CITE

-------
 1      effects to O3 levels in excess of PRB concentrations. Issues concerning the methodology for
 2      estimating PRB O3 concentrations are described in detail in Annex AX3, Section AX3.9.
 3           Contributions to PRB O3 include: photochemical interactions involving natural emissions
 4      of VOCs, NOX, and CO; the long-range transport of O3 and its precursors from outside North
 5      America; and stratospheric-tropospheric exchange (STE).  Processes involved in STE are
 6      described in detail in Annex AX2.3. Natural sources of O3 precursors include biogenic
 7      emissions, wildfires, and lightning. Biogenic emissions from agricultural activities are not
 8      considered in the formation of PRB O3.
 9           Most of the issues concerning the calculation of PRB O3 center on the origin of springtime
10      maxima in surface O3 concentrations observed at monitoring sites in relatively unpolluted areas
11      of the United States and on the capability of the current generation of global-scale, three-
12      dimensional chemistry transport models to correctly simulate their causes.  These issues are
13      related to the causes of the occurrence of high O3 values, especially those averaged over 1-h to
14      8-h observed at O3 monitoring sites during late winter through spring (i.e., February to June).
15      The issues raised do not affect interpretations of the causes of summertime O3 episodes as
16      strongly. Summertime O3 episodes are mainly associated with slow-moving high-pressure
17      systems characterized by limited mixing between the planetary boundary layer and the free
18      troposphere (Section AX2.3).
19           Springtime  maxima are observed at national parks mainly in the western United States that
20      are relatively clean (Section AX3.2.2; Figures AX3-76a,b) and at a number of other relatively
21      unpolluted monitoring sites throughout the Northern Hemisphere. Spring maxima in
22      tropospheric O3 were  originally attributed to transport from the stratosphere by Regener (1941)
23      as cited by Junge (1963). Junge (1963) also cited measurements  of springtime maxima in O3
24      concentrations  at Mauna Loa (elevation 3400 m) and at Arkosa, Germany (an alpine location,
25      elevation 1860  m).  Measurements of radioactive debris transported downward from the
26      stratosphere as  the result of nuclear testing during the 1960s also  show springtime maxima
27      (Ludwig et al.,  1977).  However, more recent studies (Lelieveld and Dentener, 2000; Browell
28      et al., 2003) attribute the springtime maximum in tropospheric O3 concentrations to tropospheric
29      production rather than transport from the stratosphere. It should be noted here that O3 in the free
30      troposphere is subject to chemical loss on time scales much shorter than for decay of most
        August 2005                            AX3-131      DRAFT-DO NOT QUOTE OR CITE

-------
                        Yellowstone National Park
                     Maximum Hourly Concentration
                                 1998-2001
                                    Month
                          I 1998
I 1999   D 2000   D 2001
Figure AX3-76a.  Monthly maximum hourly average O3 concentrations at Yellowstone
                National Park, Wyoming in 1998,1999, 2000, and 2001.
           ooooooooooooooooooooooooooo

                                        Time


Figure AX3-76b.   Hourly average O3 concentrations at Yellowstone National Park,
                 Wyoming for the period January to December 2001.


Source: U.S. Environmental Protection Agency (2003a).
August 2005
 AX3-132    DRAFT-DO NOT QUOTE OR CITE

-------
 1      radio-isotopes produced by nuclear testing that were used as tracers of stratospheric air such
 2      as 14C, 137Cs and 90Sr.
 3           Springtime O3 maxima were observed in low-lying surface measurements during the late
 4      19th century. However, these measurements are quantitatively highly uncertain, and extreme
 5      caution should be exercised in their use. Concentrations of approximately 0.036 ppm for the
 6      daytime average and of 0.030 ppm for the nighttime averages were reported for Zagreb, Croatia
 7      using the Schonbein method during the 1890s (Lisac and Grubisic, 1991). Of the numerous
 8      measurements of tropospheric O3 made in the 19th century, only the iodine catalyzed oxidation
 9      of arsenite has been verified with modern laboratory methods. Kley et al. (1988) reconstructed
10      the apparatus used between 1876 and 1910 in Montsouris, outside Paris, and evaluated it for
11      accuracy and specificity. They concluded that O3 mixing ratios ranged from 5 to 16 ppb with
12      uncertainty of ±2 ppb.  Interferences from SO2 were avoided as the Montsouris data were
13      selected to exclude air from Paris, the only source of high concentrations of SO2 at that time.
14      Uncertainties in the humidity correction to the Schonbein reading will lead to considerable
15      inaccuracies in the seasonal cycle established by this method (Pavelin et al., 1999).  Because of
16      the uncertainties in the earlier methods, it is difficult to quantify the differences between
17      surface O3 concentrations measured in the last half of the 19th century at certain locations in
18      either Europe or North America with those currently monitored at remote locations in the world.
19           Observations of O3 profiles at a large number  of sites indicate a positive gradient in O3
20      mixing ratios with increasing altitude in the troposphere and a springtime maximum in O3
21      concentrations in the upper troposphere (Logan, 1999).  As discussed in Section AX2.3.1, STE
22      affects the middle and upper troposphere more than the lower troposphere. It is, therefore,
23      reasonable to suppose that the main cause of this positive gradient is STE. However, deep
24      convection transports pollutants upward and can result in an increase in the pollutant mixing
25      ratio with altitude downwind  of surface source regions as shown in Figure AX3-77. This effect
26      can be seen in differences in ozonesonde profiles as one moves eastward across the United States
27      (Newchurch et al., 2003). In  addition, O3 formed by lightning-generated NOX also contributes to
28      the vertical O3 gradient.  (Lelieveld and Dentener, 2000). This O3 could be either background or
29      not, depending on the sources of radical precursors.  Another contributing factor is the increase
30      of O3 lifetime with altitude (Wang et al., 1998). Free-tropospheric O3 is not predominantly of
31      stratospheric origin, nor is it all natural; it is mostly controlled by production within the

        August 2005                            AX3-133     DRAFT-DO NOT QUOTE OR CITE

-------
                  a. Carbon Monoxide

            Surface
                 98°Wf    97C
                  Oklahoma City
96°      95°      94°
     Longitude
                  b. Ozone
             g>
             'o
            Surface
                 98°W|    97C
                  Oklahoma City
96°      95°      94°
     Longitude
Figure AX3-77.  (a) Contour plot of CO mixing ratios (ppbv) observed in and near the
                June 15,1985, mesoscale convective complex in eastern Oklahoma.
                Heavy line shows the outline of the cumulonimbus cloud.  Dark shading
                indicates high CO and light shading indicates low CO. Dashed contour
                lines are plotted according to climatology since no direct measurements
                were made in that area, (b) Same as (a) but for O3 (ppbv).

Source: Dickerson et al. (1987).
August 2005
   AX3-134     DRAFT-DO NOT QUOTE OR CITE

-------
 1      troposphere and includes a major anthropogenic enhancement (e.g., Berntsen et al., 1997;
 2      Roelofs et al., 1997; Wild and Akimoto, 2001).
 3           Stohl (2001), Wernli and Borqui (2002), Seo and Bowman (2002), James et al. (2003a,b),
 4      Sprenger and Wernli (2003), and Sprenger et al. (2003) addressed the spatial and temporal
 5      variability in stratosphere to troposphere transport. Both Stohl (2001) and Sprenger et al. (2003)
 6      produced 1-year climatologies  of tropopause folds based on a 1° by 1° gridded meteorological
 7      model data set. They each found that the probability of deep folds (penetrating to the 800 hPa
 8      level) was maximum during winter (December through February) with the highest frequency of
 9      folding extending from Labrador down the east coast of North America. However, these deep
10      folds occurred in <1% of the 6-h intervals for which meteorological data was assimilated for grid
11      points in the continental United States, with  a higher frequency in Canada. They observed a
12      higher frequency of more shallow folds (penetrating to the upper troposphere) and medium folds
13      (penetrating to levels between 500 and 600 hPa) of about 10% and 1 to 2%, respectively. These
14      events occur preferentially across the subtropics and the southern United States. At higher
15      latitudes, other mechanisms such as the erosion of cut-off lows and the breakup of stratospheric
16      streamers are likely to play an important role in STE.  A 15-year model climatology by Sprenger
17      and Wernli (2003) showed the  consistent pattern of STE occurring over the primary storm tracks
18      along the Asian and North American coasts. This climatology, and the one of James et al.
19      (2003a,b) both found that recent stratospheric air associated with deep intrusions are relatively
20      infrequent occurrences in these models. Thus, stratospheric intrusions are most likely to directly
21      affect the middle and upper troposphere, not the planetary boundary layer. However, this O3 can
22      still exchange with the planetary boundary layer through convection or through large-scale
23      subsidence as described later in this subsection and in  Sections AX2.3.2, AX2.3.3, and AX2.3.4.
24      These results are in accord with the observations of Galeni et al. (2003) over Greece and those of
25      Ludwig et al. (1977) over the western United States. It should also be remembered that
26      stratospheric O3 injected into the upper troposphere is  subject to chemical destruction as it is
27      transported downward toward the surface.
28           Ozone concentrations measured at RRMS in the  Northern Hemisphere  have been compiled
29      by Vingarzan (2004) and are reproduced here in Tables AX3-10, AX3-11, and AX3-12.  Data
30      for annual mean/median concentrations show a broad range, as do annual maximum 1-h
31      concentrations.  Generally, concentrations increase with elevation and the highest concentrations

        August 2005                            AX3-135      DRAFT-DO NOT QUOTE OR CITE

-------
   Table AX3-10. Range of Annual (January-December) Hourly Ozone Concentrations
   	(ppb) at Background Sites Around the World (CMDL, 2004)	
 Location
                                     Elevation (in)
                                                     Period of Record
                                Range of Annual Means
 Pt. Barrow, Alaska
 Ny Alesund, Svalbard, Spitsbergen"
 Mauna Loa, Hawaii0
 11
475
3397
1992-2001
1989-1993
1992-2001
23-29
28-33b
37-46d
 "University of Stockholm Meteorological Institute.
 bAnnual medians
 °10:00- 18:00 UTC.
 dHigh elevation site.
 Source: Vingarzan (2004).
   Table AX3-11. Range of annual (January-December) Hourly Median and Maximum
       Ozone Concentrations (ppb) at Background Stations in Protected Areas of the
                               United States (CASTNet, 2004)
Location
Denali NP, Alaska
Glacier NP, Montana
Voyageurs NP, Minnesota
Theodore Roosevelt NP, North Dakota
Yellowstone NP, Wyoming
Rocky Mountain NP, Colorado
Olympic NP, Washington
North Cascades NP, Washington
Mount Rainier NP, Washington
Lassen NP, California
Virgin Islands NP, U.S. Virgin Islands
Elevation (in)
640
976
429
850
2469
2743
125
109
421
1756
80
Period of Record
1998-2001
1989-2001
1997-2001
1983-2001
1996-2001
1994-2001
1998-2001
1996-2001
1995-2001
1995-2001
1998-2001
Range of Annual Medians
29-34
19-27
28-35
29-43
37-45"
40-47"
19-22
14-18
38371
38-43"
19-24
Range of Annual Maxima
49-68
57-77
74-83
61-82
68-79"
68-102"
50-63
48-69
54-98
81-109"
50-64
 "High elevation site.
 Source: Vingarzan (2004).

   Table AX3-12. Range of annual (January-December) Hourly Median and Maximum
     Ozone Concentrations (ppb) at Canadian Background Stations (CAPMoNa, 2003)
Location
Kejimkujik, Nova Scotiab
Montmorency, Quebec
Algoma, Ontario"
Chalk River, Ontario
Egbert, Ontario"
E.L.A., Ontario
Bratt's Lake, Saskatchewan
Esther, Alberta
Saturna Island, British Columbia
Elevation (in)
127
640
411
184
253
369
588
707
178
Period of Record
1989-2001
1989-1996
1988-2001
1988-1996
1989-2001
1989-2001
1999-2001
1995-2001
1992-2001
Range of Annual Medians
25-34
28-32
27-33
25-31
27-32
28-33
26-29
26-31
23-27
Range of Annual Maxima
76-116
73-99
76-108
79-107
90-113
64-87
63-68
63-78
65-82
 "Canadian Air and Precipitation Monitoring Network.
 'Stations affected by long-range transport of anthropogenic emissions.
 Source: Vingarzan (2004).
August 2005
  AX3 -136      DRAFT-DO NOT QUOTE OR CITE

-------
 1      are found during spring.  The overall average of the annual median O3 concentrations at all sites
 2      in the continental United States is about 30 ppb and excluding higher elevation sites it is about
 3      24 ppb. Maximum concentrations may be related to stratospheric intrusions, wildfires, and
 4      intercontinental or regional transport of pollution.  However, it should be noted that all of these
 5      sites are affected by anthropogenic emissions to some extent making an interpretation based on
 6      these data alone problematic.
 7           Daily 1-h maximum O3 concentrations exceeding 50 or 60 ppb are observed during late
 8      winter and spring in southern Canada and at sites in national parks as shown in Tables
 9      AX3-13, AX3-14, and Figure AX3-78. That these high values can occur during late winter
10      when there are low sun angles and cold temperatures may imply a negligible role for
11      photochemistry and a major role for stratospheric intrusions. However, active photochemistry
12      occurs even at high latitudes during late winter. Rapid O3 loss, apparently due to multiphase
13      chemistry involving bromine atoms (see Section 2.2.10) occurs in the Arctic marine boundary
14      layer.  The Arctic throughout much of winter is characterized by low light levels, temperatures,
15      and precipitation, and can act as a reservoir for O3 precursors such as PAN and alkyl nitrates,
16      which build up and can then photolyze when sun angles are high enough during late winter and
17      early spring. Long-range transport of total odd nitrogen species (NOy) (defined in AX2.2.2) and
18      VOCs to Arctic regions can occur from midlatitude-source regions. In addition, O3 can be
19      transported from tropical areas in the upper troposphere followed by its subsidence at mid and
20      high latitudes (Wang et al., 1998).
21           Penkett (1983), and later Penkett and Brice (1986), first observed a spring peak in PAN at
22      high northern latitudes and hypothesized that winter emissions transported into the Arctic would
23      be mixed throughout a large region of the free troposphere and transformed  into O3 as solar
24      radiation returned to the Arctic in the spring. Subsequent observations (Dickerson, 1985)
25      confirmed the presence of strata of high concentrations of reactive nitrogen  compounds at high
26      latitudes in early spring.  Bottenheim et al. (1990, 1993) observed a positive correlation
27      between O3 and NO2 in the Arctic spring. Jaffe et al. (1991) found NOy concentrations
28      approaching 1 ppb in Barrow, Alaska, in the spring and attributed them to long-range transport.
29           Beine et al. (1997) and Honrath et al. (1997) measured O3, PAN, and NOX in Alaska and
30      Svalbard, Norway and concluded that PAN decomposition can lead to photochemical O3
31      production. At Poker Flat, Alaska, O3 production was directly observable.  Herring et al. (1997)

        August 2005                            AX3-137     DRAFT-DO NOT QUOTE OR CITE

-------
Table AX3-13. Number of Hours >0.05 ppm for Selected Rural O3 Monitoring in the United States
                            by Month for the Period 1988 to 2001
to
o
o








>
X
OJ
i
oo


DRAFT-DO I
^
0
H
O
O
H
W
O
O
H
W
Site Name Month
Denali National Park, Alaska February
Denali National Park, Alaska March
Denali National Park, Alaska April
Denali National Park, Alaska May
Denali National Park, Alaska June
Yellowstone National Park, February
Wyoming
Yellowstone National Park, March
Wyoming
Yellowstone National Park, April
Wyoming
Yellowstone National Park, May
Wyoming
Yellowstone National Park, June
Wyoming







1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001
00000000 0000 14
0 0 0 0 0 0 0 0 52 0 122 17 0 24
217 0 2 0 64 10 31 21 12 51 236 119 0 302
26 1 0 24 10 17 1 54 97 35 79 29 0 98
00000000 27 00 22 06
0 0 11 3 0 21 6 0 1 5 252 23 77

194 2 4 95 26 285 14 7 98 150 509 286 307

228 17 16 217 62 311 185 65 163 385 517 242 461
225 2 10 196 47 180 193 212 216 289 458 240 350

58 67 139 33 28 116 81 94 149 78 212 181 172









-------

-------
Table AX3-14. Number of Hours >0.06 ppm for Selected Rural O3 Monitoring Sites in the United States by Month
                                      for the Period of 1988 to 2001
to
o
o









>
X
OJ
£
o


O
£j
H
6
o
0
H
O
o
H
W
O
O
H
W
Site Name
Denali National Park,
Alaska
Denali National Park,
Alaska
Denali National Park,
Alaska
Denali National Park,
Alaska
Denali National Park,
Alaska
Yellowstone National
Park, Wyoming
Yellowstone National
Park, Wyoming
Yellowstone National
Park, Wyoming
Yellowstone National
Park, Wyoming
Yellowstone National
Park, Wyoming
Glacier National Park,
Montana
Glacier National Park,
Montana



Month
February

March

April

May

June

February

March

April

May

June
February
March




1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001
00000000 00000

00000000000000

00000000000000

00000000200009

00000000000002

0 000000000600

37 0 0 0 0 0 0 0 1 0 120 1 4

59 0 0 29 0 20 4 0 0 64 158 11 77

20 0 0 61 3 42 24 38 26 54 169 49 139

8 7 18 2 1 13 0 0 22 4 27 43 18
000000000000
100000000000





-------
u.uo ppm lor aeietieu isurai \J3 ivioimormg aiies 111 me uniieu aiaies uy ivioiiui
for the Period of 1988 of 2001
Month 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001
April 001000002100
May 27 13 00450 16 19 80

June 03 1010 16 00000

February 10000180000000
March 00000 34 052 15 0940
April 9 0 1 0 0 5 8 0 17 2 57 24 41 0
May 77 6 0 0 40 9 40 2 27 46 53 139 43 6
June 30 0 0 28 17 5 113 12 115 0 5 0 32









-------
             54
       Figure AX3-78.  Maximum hourly average O3 concentrations at rural monitoring sites in
                        Canada and the United States in February from 1980 to 1998.
       Source: Lefohnetal. (2001).
 1     tracked springtime O3 maxima in Denali National Park, Alaska, an area one might presume to be
 2     pristine.  They measured NOX and hydrocarbons and concluded that, in the spring, O3 was
 3     produced predominantly by photochemistry at a calculated rate of 1 to 4 ppb/day, implying that
 4     the O3 observed could be produced on timescales ranging from about a week to a month.
 5     Solberg et al. (1997) tracked the major components of NOy in remote Spitsbergen, Norway for
 6     the first half of the year 1994.  They observed high concentrations of PAN (800 ppt) peaking
 7     simultaneously with O3 (45 to 50 ppb) and attributed this to the long-range transport of pollution
 8     and to photochemical smog chemistry.  These investigators concluded, in general, that large
 9     regions of the Arctic store high concentrations of O3 precursors in the winter and substantial
10     quantities of O3 are produced by photochemical reactions in the spring. Although reactions with
       August 2005
AX3 -142     DRAFT-DO NOT QUOTE OR CITE

-------
 1      high-activation-energy barriers may be ineffective, reactions with low- or no activation-energy
 2      barriers (such as radical-radical reactions) or negative temperature dependencies will still
 3      proceed. Indeed, active photochemistry is observed in the coldest regions of the stratosphere and
 4      mesosphere.  While it is expected that photochemical production rates of O3 will increase with
 5      decreasing solar zenith angle as one moves southward from the locations noted above, it should
 6      not be assumed that photochemical production of O3 does not occur during late winter and spring
 7      at mid- and high-latitudes.
 8           Perhaps the most thorough  set of studies investigating causes of springtime maxima in
 9      surface O3 has been performed as part of the AEROCE and NARE studies (cf Sections
10      AX2.3.4a,b) and TOPSE (Browell et al., 2003). These first two studies found that elevated or
11      surface O3 > 40 ppb at Bermuda,  at least, arises from two distinct sources:  the polluted North
12      American continent and the stratosphere. It was also found that these sources mix in the upper
13      troposphere before descending as shown in Figure AX3-79. (In general, air descending behind
14      cold fronts contains contributions from intercontinental transport and the stratosphere.)  These
15      studies also concluded that it is impossible to determine sources of O3 without ancillary data that
16      could be used either as tracers of sources or to calculate photochemical production and loss rates.
17      In addition, subsiding back trajectories do not necessarily imply a free-tropospheric or
18      stratospheric origin for O3 observed at the surface, since the subsiding conditions are also
19      associated with strong inversions and clear skies that promote O3 production within the boundary
20      layer. Thus, it would be highly problematic to use observations alone as estimates of PRB O3
21      concentrations, especially  for sites at or near sea level.
22           The IPCC Third Assessment Report (TAR) (2001) gave a large range of values for terms in
23      the tropospheric O3 budget. Estimates of O3 STE of O3 ranged over a factor of three from 391 to
24      1440 Tg/year in the twelve models included in the intercomparison; many of the models
25      included in that assessment overestimated O3 STE. However, the overestimates likely reflected
26      errors in assimilated winds in the upper troposphere (Douglass et al., 2003;  Schoeberl et al.,
27      2003; Tan et al., 2004; van Noije et al., 2004). The budgets of tropospheric O3 calculated since
28      the IPCC TAR are shown  in Table AX3-15. Simulation of stratospheric intrusions is notoriously
29      difficult in global models,  and O3 STE is generally parameterized in these models. However, as
30      can be seen from inspection of Table AX3-15, improvements in assimilation techniques have
31      improved and narrowed estimates of STE.  A model intercomparison looking at actual STE

        August 2005                            AX3-143      DRAFT-DO NOT QUOTE OR CITE

-------
                 Altitude
                             Stratospheric Air
                                ,*-
                                Cold Front (A)
                Stratospheric Air
                   Cold Front (B)
                   West
          -1500
Distance (km)
East
       Figure AX3-79.  Schematic diagram of a meteorological mechanism involved in high
                        concentrations of O3 found in spring in the lower troposphere off
                        the American East Coast. Subsidence behind the first cold front meets
                        convection ahead of a second cold front such that polluted air and O3
                        from the upper troposphere/ lower stratosphere are transported in close
                        proximity (or mixed) and advected over the North Atlantic Ocean. The
                        vertical scale is about 10 km; the horizontal scale about 1500 km.  (Note
                        that not all cold fronts are associated with squall lines and that mixing
                        occurs even in their absence.)
       Source: Prados (2000).
 1     events found significant variations in model results that depended significantly on the type and
 2     horizontal resolution of the model (Meloen et al., 2003; Cristofanelli et al., 2003). In particular,
 3     it was found that the Lagrangian perspective (as opposed to the Eulerian perspective used in
 4     most global scale CTMs) was necessary to characterize the depths and residence times of
 5     individual events (Sprenger and Wernli, 2003; James et al., 2003a,b). A few studies of the
 6     magnitude of the O3 STE have been made based on chemical observations in the lower
 7     stratosphere or combined chemistry and dynamics (e.g., 450 Tg/year net global [Murphy and
 8     Fahey, 1994]; 510 Tg/year net global extratropics only [Gettelman et al.,  1997]; and
 9     550 ± 140 Tg/year [Olsen et al., 2002]).
10           Even if the magnitude of cross-tropopause O3 fluxes  in global CTMs are calculated
11     correctly in an annual mean sense, it should be noted that stratospheric intrusions occur
12     episodically following the passage of cold fronts at midlatitudes. Of major concern is the ability
       August 2005
 AX3 -144     DRAFT-DO NOT QUOTE OR CITE

-------
                      Table AX3-15.  Global Budgets of Tropospheric Ozone (Tg year *) for the Present-day Atmosphere
S-
to
o
o

Reference
TAR4
Lelieveld and


Model
1 1 models

Stratosphere-
Troposphere
Exchange
(STE)
770 ± 400
570

Chemical
Production 2
3420 ± 770
3310

Chemical
Loss2
3470 ± 520
3170

Dry
Deposition
770 ± 180
710

Burden
(Tg)
300 ± 30
350

Lifetime
(days)3
24 ±2
33
O
H
O

O
H
W
O

O
HH
H
W
         Dentener (2000)




.
X
OJ
[\
L/l

O
5
Beyetal. (200 la)5
Horowitz et al.
(2003)
Von Kuhlmann
et al. (2003)
Shindell et al. (2003)
Park et al. (2004)
Rotman et al. (2004)
Wong et al. (2004)
GEOS-CHEM
MOZART-2

MATCH-MPIC

GISS
UMD-CTM
IMPACT
SUNYA/UiO GCCM
470
340

540

417
480
660
600
4900
5260

4560

NR6
NR
NR
NR
4300
4750

4290

NR
NR
NR
NR
1070
860

820

1470
1290
830
1100
320
360

290

349
340
NR
376
22
23

21

NR
NR
NR
NR
1 From global CTM simulations describing the atmosphere of the last decade of the 20th century.
2 Chemical production and loss rates are calculated for the odd oxygen family, usually defined as Ox = O3 + O + NO2 + 2NO3 + 3N2O5 + HNO4 +
 peroxyacylnitrates (and sometimes HNO3), to avoid accounting for rapid cycling of O3 with short-lived species that have little implication for its budget.
 Chemical production is mainly contributed by reactions of NO with peroxy radicals, while chemical loss is mainly contributed by the O('D) + H2O reaction
 and by the reactions of O3 with HO2', -OH, and alkenes. Several models in this table do not report production and loss separately ("NR" entry in the table),
 reporting instead net production. However, net production is not a useful quantity for budget purposes, because (1) it represents a small residual between
 large production and loss, (2) it represents the balance between STE and dry deposition, both of which are usually parameterized as a flux boundary
 condition.
3 Calculated as the ratio of the burden to the sum of chemical and deposition losses
4 Means and standard deviations from an ensemble of 11 CTM budgets reported in the IPCC TAR. The mean budget does not balance exactly because
 only 9 CTMs reported chemical production and loss statistics.
5 The Martin et al. (2003b) more recent version of GEOS-CHEM gives identical rates and burdens.
6 Not reported.

-------
 1      of global-scale CTMs to simulate individual intrusions and the effects on surface O3
 2      concentrations that may result during these events.  As noted in Section AX2.3.1, these
 3      intrusions occur in "ribbons" ~ 200 to 1000 km long, 100 to 300 km wide, and 1 to 4 km thick.
 4      An example of a stratospheric intrusion occurred in Boulder, CO (EPA AQS Site 080130011;
 5      formally AIRS) on May 6, 1999 (Lefohn et al., 2001).  At 1700 UTC (1000 hours LSI)
 6      an hourly average concentration of 0.060 ppm was recorded and by 2100 UTC (1400 hours
 7      LSI), the maximum hourly average O3 concentration of 0.076 ppm was measured.  At 0200
 8      UTC on May 7, 1999 (1900 hours LST on May 6), the hourly average concentration declined to
 9      0.059 ppm.  Figure AX3-80 shows the O3 vertical profile that was recorded at Boulder, CO on
10      May 6,  1999, at 1802 UTC (1102 hours LST). The ragged vertical profile of O3 at > 4 km
11      reflects stratospheric air that has spiraled downward around an upper-level low and mixed with
12      tropospheric air along the way. Thus, stratospheric air which is normally extremely cold and dry
13      and rich in O3, loses its characteristics as it mixes downward.  This process was described in
14      Section AX2.3.1 and illustrated in Figures AX2-7a,b and c.
15           The dimensions given above imply that individual intrusions are not resolved properly in
16      the current generation of global-scale CTMs (Figure AX3-80).  However, as noted in Section
17      AX2.3.1, penetration of stratospheric air directly to the planetary boundary layer rarely occurs in
18      the continental United States. Rather, intrusions are more likely to affect the middle and upper
19      troposphere, providing  a reservoir for O3 that can exchange with the planetary boundary layer.
20      In this regard, it is important that CTMs be able to spatially and temporally resolve the exchange
21      between the planetary boundary layer and the lower free troposphere properly.
22
23      AX3.9.2  Capability of Global Models to Simulate Tropospheric Ozone
24           The current generation of global CTMs includes detailed representation of tropospheric
25      O3-NOX-VOC chemistry.  Meteorological information is generally provided by global data
26      assimilation centers.  The horizontal resolution is typically a few hundred km, the vertical
27      resolution is 0.1 to 1 km, and the effective temporal resolution is a few hours. These models  can
28      simulate most of the observed variability in O3 and related species, although the coarse
29      resolution precludes simulation of fine-scale structures or localized extreme events.  On the
30      synoptic scale, at least,  all evidence indicates that global models are adequate tools to investigate
31      the factors controlling tropospheric O3.  Stratosphere-troposphere exchange of O3 in global

        August 2005                            AX3-146     DRAFT-DO NOT QUOTE OR CITE

-------
                          -100
                                       Temperature and Frost Point (C)
                                  -80      -60      -40      -20       0
                                                                          20
                                                                                  40
                       0  10  20  30  40  50  60  70  80  90  100  110  120  130  140 150
                                               Ozone (ppbv)

       Figure AX3-80.  Ozone vertical profile at Boulder, Colorado on May 6,1999 at 1802 UTC
                        (1102 LST).
       Source: Lefohnetal. (2001).
 1
 2
 3
 4
 5
 9
10
11
models is generally parameterized. The parameterizations are typically constrained to match the
global mean O3 cross-tropopause flux, which is in turn constrained by a number of observational
proxies (550 ± 140 Tg O3 year1 [Olsen et al., 2002]).  The model simulations are routinely
compared to ozonesonde observations in the middle and upper troposphere to test the simulation
of stratospheric influence on tropospheric O3 (Logan,  1999). Such evaluations show that the
parameterized cross-tropopause O3 flux in global models results in a good simulation of
tropospheric O3, at least in a mean sense; and that the current generation of models can
reproduce the tropospheric ozonesonde climatology to within 5 to 10 ppbv, even at mid- and
high-northern latitudes in winter, with the correct seasonal cycle.
     Fiore et al. (2003a) used the GEOS-CHEM global  tropospheric chemistry model to
quantify PRB O3 concentrations across the United States. A net global O3 flux of 490 Tg O3
       August 2005
                                       AX3 -147     DRAFT-DO NOT QUOTE OR CITE

-------
 1      year1 from the stratosphere to the troposphere is imposed in the GEOS-CHEM model,
 2      consistent with the range constrained by observations (Olsen et al., 2002). Previous applications
 3      of the model have demonstrated that it simulates the tropospheric ozonesonde climatology
 4      (Logan, 1999) generally to within 5 to 10 ppbv, including at mid- and high-latitudes (Bey et al.,
 5      2001a) over Bermuda in spring (Li et al., 2002) and at sites along the Asian Pacific rim (Liu
 6      et al., 2002). The phase of the seasonal cycle is reproduced to within 1 to 2 months (Bey et al.,
 7      2001a; Li et al., 2002; Liu et al., 2002). An analysis of the 210Pb-7Be-O3 relationships observed
 8      in three aircraft missions over the western Pacific indicates that the model does not
 9      underestimate the stratospheric source of O3 (Liu et al., 2004).  These studies and others (Li
10      et al., 2001; Bey et al., 2001b; Fusco and Logan, 2003) demonstrate that the model provides an
11      adequate simulation of O3 in the free troposphere at northern midlatitudes, including the mean
12      influence from the stratosphere. However, it cannot capture the structure and enhancements
13      associated with stratospheric intrusions, leading to mean O3 under-prediction in regions of
14      preferred stratospheric downwelling.
15           Fiore et al. (2002a, 2003b) presented a detailed evaluation of the model simulation for O3
16      and related species in surface air over the United States for the summer of 1995. They showed
17      that the model reproduces important features of observations including the high tail of O3
18      frequency distributions at sites in the eastern United States (although sub-grid-scale local peaks
19      are underestimated), the O3 to (NOy - NOX) relationships,  and that the highest O3 values exhibit
20      the largest response to decreases in U.S. fossil fuel emissions from 1980 to 1995 (Lefohn et al.,
21      1998).  Empirical orthogonal  functions (EOFs) for the observed regional variability of O3 over
22      the eastern United States are also well reproduced, indicating that GEOS-CHEM captures the
23      synoptic-scale transport processes modulating surface O3 concentrations (Fiore et al., 2003b).
24      One model shortcoming relevant for the discussion below is that excessive convective mixing
25      over the Gulf of Mexico and the Caribbean leads to an overestimate of O3 concentrations in
26      southerly flow over the southeastern United States.  Comparison of GEOS-CHEM with the
27      Multiscale Air Quality Simulation Platform (MAQSIP) regional air quality modeling system
28      (Odman and Ingram, 1996) at 36 km2 horizontal resolution showed that the models exhibit
29      similar skill at capturing the observed variance in O3 concentrations with comparable model
30      biases (Fiore et al., 2003b).
31

        August 2005                            AX3-148     DRAFT-DO NOT QUOTE OR CITE

-------
 1      Simulations to Quantify Background Ozone Over the United States
 2           The sources contributing to the O3 background over the United States were quantified by
 3      Fiore et al. (2003a) with three simulations summarized in Table AX3-16:  (1) a standard
 4      simulation, (2) a background simulation in which North  American anthropogenic NOX,
 5      NMVOC, and CO emissions are set to zero, and (3) a natural O3 simulation in which global
 6      anthropogenic NOX, NMVOC and CO emissions are set  to zero and the CH4 concentration is set
 7      to its 700 ppbv pre-industrial value. Anthropogenic emissions of NOX, nonmethane volatile
 8      organic compounds (NMVOCs), and CO include contributions from fuel use, industry, and
 9      fertilizer application.  The difference between the standard and background simulations
10      represents regional pollution, i.e., the  O3 enhancement from North American anthropogenic
11      emissions. The difference between the background and  natural simulations represents
12      hemispheric pollution, i.e., the O3 enhancement from anthropogenic emissions outside North
13      America. Methane and NOX contribute most to hemispheric pollution (Fiore et al., 2002b).
14      A tagged O3 tracer  simulation (Fiore et al., 2002a) was used to isolate the  stratospheric
15      contribution to the background and yielded results that were quantitatively consistent with those
16      from a  simulation in which O3 transport from the stratosphere to the troposphere was suppressed
17      (Fusco  and Logan,  2003). All simulations were initialized in June 2000; results are reported for
18      March through October 2001.
                  Table AX3-16. Description of Simulations Used for Source Attribution
                                           (Fiore et al., 2003a)
         Simulation    Description                                            Horizontal Resolution
         Standard       Present-day emissions as described in the text                          2° x 2.5°
         Background    North American anthropogenic NOX, NMVOC, and CO emissions          2° x 2.5°
                      set to zero
         Natural        Global anthropogenic NOX, NMVOC, and CO emissions set to zero          4° x 5°
                      and CH4 concentration set to its 700 ppbv preindustrial value
         Stratospheric    Tagged O3 tracer originating from the stratosphere in standard             2° x 2.5°
                      simulation
        August 2005                             AX3-149      DRAFT-DO NOT QUOTE OR CITE

-------
 1           The standard and background simulations were conducted at 2° x 2.5° horizontal
 2     resolution, but the natural simulation was conducted at 4° x 5° resolution to save on
 3     computational time. There was no significant bias between 4° x 5° and 2° x 2.5° simulations
 4     (Fiore et al., 2002a), particularly for a natural O3 simulation where surface concentrations were
 5     controlled by large-scale processes.
 6
 7     AX3.9.3  Mean Background Concentrations: Spatial and Seasonal Variation
 8           The analysis of Fiore et al. (2003a) focused on the 2001 observations from the Clean Air
 9     Status and Trends Network (CASTNet) of rural and remote U.S. sites (Lavery et al., 2002)
10     (Figure AX3-81). Figure AX3-82 shows the mean seasonal cycle in afternoon (1300 to 1700
11     hours LT) O3 concentrations averaged over the CASTNet stations in each U.S. quadrant.
12     Measured O3 concentrations (asterisks) are highest in April to May, except in the Northeast
13     where they peak in June. Model results (triangles) are within 3 ppbv and 5 ppbv of the
14     observations for all months in the Northwest and Southwest, respectively.  Model results for the
15     Northeast are too high by 5 to 8 ppbv when sampled at the CASTNet sites; the model is lower
16     when the ensemble of grid squares in the region are sampled (squares). The model is 8 to
17     12 ppbv too high over the Southeast in summer for reasons discussed in Section AX3.9.2.
18           Results from the background simulation (no anthropogenic emissions in North America;
19     see Table AX3-16) are shown as diamonds in Figure AX3-82. Mean afternoon background O3
20     ranges from 20 ppbv in the Northeast in summer to 35 ppbv in the Northwest in spring. It is
21     higher in the West than in the East because of higher elevation, deeper mixed layers, and
22     longer O3 lifetimes due to the arid climate (Fiore et al., 2002a). It is also higher in spring than in
23     summer, in part because of the seasonal maximum of stratospheric influence (Figure AX3-82)
24     and in part because of the longer lifetime of O3 (Wang et al., 1998).
25           Results from the natural O3 simulation (no anthropogenic emissions anywhere; Table
26     AX3-16) are shown as crosses in Figure AX3-82. Natural O3 concentrations are also highest in
27     the West and in spring when the influence of stratospheric O3 on the troposphere peaks (e.g.,
28     Holton et al., 1995). Monthly mean natural O3 concentration ranges are 18 to  23,  18 to 27, 13  to
29     20, and 15 to 21  ppbv in the Northwest, Southwest, Northeast, and Southeast,  respectively.  The
30     stratospheric contribution (X's) ranges from 7 ppbv in spring to 2 ppbv in summer.
       August 2005                           AX3-150      DRAFT-DO NOT QUOTE OR CITE

-------
                                                '   voy-;"v  '''.
                               • - -X"'
                                Y^k
                     x
                                      X
                             X   x:  x
                           '. .- X MEV
                                                           •,cdw
                                                           G'AS  '.''
       Figure AX3-81.   CASTNet stations in the continental United States for 2001. Sites
                        discussed in Section AX3.9.4 are labeled: VOY = Voyageurs NP, MI;
                        COW = Coweeta, NC; YEL = Yellowstone NP, WY; CAD = Caddo
                        Valley, AR; CVL = Coffeeville, MS; GAS = Georgia Station, GA;
                        GB = Great Basin, NV; GRC = Grand Canyon, AZ; CAN = Canyonlands,
                        UT; MEV = Mesa Verde, CO. Crosses denote sites > .5 km altitude.
       Source: Fiore et al. (2003a).
 1
 2
 3
 4
 5
 9
10
11
     The difference between the background and natural simulations in Figure AX3-82
represents the monthly mean hemispheric pollution enhancement. This enhancement ranges
from 5 to 12 ppbv depending on region and season. It peaks in spring due to a longer O3 lifetime
(Wang et al., 1998) and to e efficient ventilation of pollution from the Asian continent (Liu et al.,
2003). In contrast to hemispheric pollution, the regional pollution influence (O3 produced from
North American anthropogenic emissions, shown as the difference between the squares and
diamonds) peaks in summer and is highest in the East.  For the data in Figure AX3-82, it ranges
from 8 ppbv in the northern quadrants in March to over 30 ppbv in the eastern quadrants in
summer.  Monthly mean observed O3 concentrations are influenced by both regional and
hemispheric pollution in all U.S. regions from March through  October.
       August 2005
                                      AX3-151
DRAFT-DO NOT QUOTE OR CITE

-------
             80

             60

             40

          •£> 20
          .0
          Q.
          Q.  0

          C  80
          O
          N
          °  60

             40

             20
                                           NW
                                           sw

                     J  FMAMJ  JASOND  J FMAMJ  JASOND
                                           Month of 2001
Figure AX3-82.
                      Monthly mean afternoon (1300 to 1700 hours LT) concentrations (ppbv)
                      in surface air averaged over the CASTNet stations (Figure AX3-81) in
                      each U.S. quadrant for March to October 2001. Observations (asterisks)
                      are compared with model values from the standard simulation sampled
                      at the CASTNet sites (triangles) and sampled for the entire quadrant
                      (squares). The vertical lines show the standard deviation in the observed
                      and simulated values. Monthly mean model results for the background
                      (diamonds), natural (crosses), and stratospheric (X's) contributions
                      (Table AX3-16) to surface O3 are shown. The U.S. quadrants are
                      centered at 98.75° W and 37° N.
      Source: Fiore et al. (2003a).
1     AX3.9.4  Frequency of High-Ozone Occurrences at Remote Sites

2          Lefohn et al. (2001) pointed out the frequent occurrence of high-O3 events (>50 and
3     60 ppbv) at remote northern U.S. sites in spring.  Fiore et al. (2003a) replicated the analysis of
4     Lefohn et al. (2001) at the four CASTNet sites that they examined: Denali National Park
5     (Alaska), Voyageurs National Park (Minnesota), Glacier National Park (Montana), and
6     Yellowstone National Park (Wyoming).  The number of times that the hourly O3 observations at
      August 2005
                                     AX3-152     DRAFT-DO NOT QUOTE OR CITE

-------
 1      the sites are >50 and 60 ppbv for each month from March to October 2001 were then calculated
 2      (see results in Table AX3-17) and compared with the same statistics for March to June
 3      1988 to 1998 from Lefohn et al. (2001), to place the 2001 statistics in the context of other years.
 4      More incidences of O3 above both thresholds occur at Denali National Park and Yellowstone
 5      National Park in 2001 than in nearly all of the years analyzed by Lefohn et al. (2001). The
 6      statistics at Glacier National Park, Montana indicate that 2001 had fewer than average incidences
 7      of high-O3 events.  At Voyageurs National Park in Minnesota, March and April 2001 had
 8      lower-than-average frequencies of high-O3 events, but May and June were more typical.
 9      Overall, 2001 was  considered to be a suitable year for analysis of high-O3 events. Ozone
10      concentrations >70 and 80 ppbv occurred most often in May through August in 2001 and were
11      found to be associated with regional pollution by Fiore et al. (2003a).
12           Fiore et al. (2003a) focused their analysis on mean O3 concentrations during the afternoon
13      hours (1300 to 1700 LT), as the comparison  of model results with surface observations is most
14      appropriate in the afternoon when the observations are representative of a relatively deep mixed
15      layer (Fiore et al., 2002a). In addition, the GEOS-CHEM model does not provide independent
16      information on an hour-to-hour basis, because it is driven by meteorological fields that are
17      updated every 6-h and then interpolated. Fiore et al. (2003a) tested whether an analysis
18      restricted to these mean 1300 to 1700 LT surface concentrations captures the same frequency
19      of O3 >50 and 60 ppbv that emerges from an analysis of the individual hourly concentrations
20      over 24 hours. Results are reproduced here in Table AX3-17, which shows that the percentage
21      of individual afternoon (1300 to 1700 LT) hours when O3 >50 and 60 ppbv at the CASTNet sites
22      is always greater than the percentage of all hourly occurrences above these thresholds, indicating
23      that elevated O3 concentrations preferentially occur in the afternoon. Furthermore, Table
24      AX3-17 shows that the frequency of observation of high-O3 events is not diminished when 4-h
25      average (1300 to 1700 LT) concentrations are considered, reflecting persistence in the duration
26      of these events.  Model frequencies of high-O3 events from 1300 to 1700 LT  at the CASTNet
27      sites are similar to  observations in spring, as  shown in Table AX3-17, and about 10% higher in
28      the summer, largely because of the positive model bias in the Southeast discussed in
29      Section AX3.9.2.
30
        August 2005                            AX3-153     DRAFT-DO NOT QUOTE OR CITE

-------
            Table AX3-17. Number of Hours with Ozone Above 50 or 60 ppbv
                              at U.S. CASTNet Sites in 2001
Site
Mar
Apr
May
Observations ^
Denali NP, Alaska
(64° N, 149° W, 0.6 km)
Voyageurs NP, Minnesota
(48° N, 93° W, 0.4km)
Glacier NP, Montana
(49° N, 114° W, 1.0km)
Yellowstone NP, Wyoming
(45° N, 110° W, 2.5km)
All CASTNet sites (71)
All sites, 1300-1700 LT
only (hourly data)
All sites, 1300-1700 LT
mean (4-hour average)
All sites, model
1300-1700 LT mean
24
0
4
307
5468
(11%)
1817
(21%)
435
(20%)
254
(12%)
302
0
0
461
15814
(32%)
4684
(56%)
1153
(55%)
1249
(59%)
98
62
23
350
17704
(36%)
5174
(61%)
1295
(61%)
1527
(69%)
Observations ^
Denali NP
Voyageurs NP
Glacier NP
Yellowstone NP
All sites
All sites, 1300-1700 LT
only
All sites,
1300-1700 LT mean
All sites, model
1300-1700 LT mean
0
0
0
4
519
(1%)
235
(3%)
56
(3%)
13
(1%)
0
0
0
77
4729
(10%)
1798
(22%)
428
(20%)
377
(18%)
9
6
0
139
8181
(16%)
2808
(33%)
697
(33%)
834
(38%)
Jun
50 ppbv
6
95
0
172
16150
(33%)
4624
(56%)
1147
(55%)
1505
(71%)
60 ppbv
2
32
0
18
8199
(17%)
2721
(33%)
671
(32%)
964
(45%)
Jul

0
14
6
140
14489
(29%)
4613
(54%)
1161
(54%)
1475
(67%)

0
0
0
6
5705
(11%)
2235
(26%)
550
(26%)
910
(41%)
Aug

0
17
12
261
15989
(32%)
5075
(60%)
1283
(60%)
1500
(68%)

0
0
0
26
7407
(15%)
2758
(33%)
677
(32%)
834
(38%)
Sep

0
33
0
173
9874
(20%)
3343
(40%)
841
(40%)
1080
(51%)

0
15
0
1
3492
(7%)
1416
(17%)
358
(17%)
502
(24%)
Oct

0
0
0
77
5642
(11%)
1945
(23%)
478
(22%)
591
(27%)

0
0
0
2
2073
(4%)
878
(10%)
214
(10%)
204
(9%)
 Data from 71 U.S. CASTNet sites are included in this analysis: those in Figure AX3-105 plus Denali NP.
 Percentages of total occurrences are shown in parentheses.
 NP = National Park; LT = Local Time.
 Reproduced from Fiore et al. (2003a).
August 2005
AX3-154      DRAFT-DO NOT QUOTE OR CITE

-------
 1     NATURAL VERSUS ANTHROPOGENIC CONTRIBUTIONS TO
 2     HIGH-OZONE OCCURRENCES
 3          Figure AX3-83, reproduced from Fiore et al. (2003a), shows probability distributions of
 4     daily mean afternoon (1300 to 1700 LT) O3 concentrations in surface air at the CASTNet sites
 5     for March through October 2001.  Model distributions for background, natural, and stratospheric
 6     O3 (Table AX3-16) are also shown. The background (long-dashed line) ranges from 10 to
 7     50 ppbv with most values in the 20 to 35 ppbv range. The full 10 to 50 ppbv range of
 8     background predicted here encompasses the previous 25 to 45 ppbv estimates shown in
 9     Table 3-8.  However, background estimates from observations tend to be at the higher end of the
10     range (25 to 45 ppbv), while these results, as well as those from prior modeling studies
11     (Table 3-8) indicate that background O3 concentrations in surface air are usually below 40 ppbv.
12     The background O3 concentrations derived from observations may be overestimated if
13     observations at remote and rural sites contain some influence from  regional pollution (as shown
14     below to occur in the model), or if the O3 versus NOy - NOX correlation is affected by different
15     relative removal rates of O3 and NOy (Trainer et al., 1993). Natural O3 concentrations
16     (short-dashed line) are generally in the 10 to 25 ppbv range and never exceed 40 ppbv.  The
17     range of the hemispheric pollution enhancement (the difference between the background and
18     natural O3 concentrations) is typically 4 to 12 ppbv and only rarely exceeds 20 ppbv (<  1% total
19     incidences). The stratospheric contribution (dotted line) is always less than 20 ppbv and usually
20     below 10 ppbv. Time series for specific sites are presented below.
21
22     CASE STUDIES: INFLUENCE OF THE BACKGROUND ON ELEVATED OZONE
23     EVENTS IN SPRING
24          High-O3 events were previously attributed to natural processes by Lefohn et al. (2001) at:
25     Voyageurs National Park, Minnesota in June and Yellowstone National Park, Wyoming in
26     March through May.  Fiore et al. (2003a) used observations from CASTNet stations in
27     conjunction with GEOS-CHEM model simulations to deconstruct the observed concentrations
28     into anthropogenic and natural contributions.
29          At Voyageurs National Park in 2001, O3 concentrations > 60 ppbv occurred frequently in
30     June but rarely later in summer (Table AX3-17).  A similar pattern was observed in 1995 and
31     1997 and was used to argue that photochemical activity was probably not responsible for these
32     events (Lefohn et al. 2001).  Figure AX3-84 from Fiore et al. (2003a) shows that GEOS-CFffiM

       August 2005                            AX3-155     DRAFT-DO NOT QUOTE  OR CITE

-------
                          0.10n
                      .D
                       Q.
                       -F  0.05-
                       ro
                       .a
                       o
                              0    10   20   30   40   50   60   70   80   90
                                                Ozone (ppbv)

       Figure AX3-83.  Probability distributions of daily mean afternoon (1300 to 1700 LT) O3
                        concentrations in surface air for March through October 2001 at U.S.
                        CASTNet sites (Figure AX3-83): observations (thick solid line) are
                        compared with model results (thin solid line). Additional probability
                        distributions are shown for the simulated background (long-dashed line),
                        natural (short-dashed line), and stratospheric (dotted line) contributions
                        to surface O3 (Table AX3-23).
       Source: Fiore et al. (2003a).
 1     captures much of the day-to-day variability in observed concentrations from mid-May through
 2     June, including the occurrence and magnitude of high-O3 events. The simulated background
 3     contribution (diamonds) ranges from 15 to 36 ppbv with a 25 ppbv mean. The natural O3 level
 4     (crosses) is 15 ppbv on average and varies from 9 to 23 ppbv. The stratospheric contribution
 5     (X's) is always < 7 ppbv.  The dominant contribution to the high-O3 events on June 26 and 29 is
 6     from regional pollution (44 and 50 ppbv on June 26 and 29, respectively, calculated as the
 7     difference between the triangles and diamonds in Figure AX3-84).  The background contribution
 8     (diamonds) is < 30 ppbv on both days, and is composed of a 20 ppbv natural contribution (which
 9     includes 2 ppbv of stratospheric origin) and a 5 ppbv enhancement from hemispheric  pollution
10     (the difference between the diamonds and crosses). Beyond these two high-O3 events,
11     Figure AX3-84 shows that regional pollution drives most of the simulated day-to-day variability
12     and explains all events above 50 ppbv.  In 2001, monthly mean observed and simulated O3
13     concentrations are lower in July (37 and 42 ppbv, respectively) and August (35  and 36 ppbv)
14     than in June (44 and 45 ppbv). Fiore et al. (2003a) hypothesized that the lower  mean  O3 and the
       August 2005
AX3-156     DRAFT-DO NOT QUOTE OR CITE

-------
                    JQ
                    Q.
                    a
                    ""BSBS*1
                    0)
                    c
                    o
                    N
                    O
                       80
                       60
40
20
                                Voyageurs NP, Minnesota (93W, 48N)
                         10       20       30        9       19
                                     May                   June
                                               29
       Figure AX3-84.  Daily mean afternoon (13 to 17 LT) O3 concentrations in surface air at
                       Voyageurs National Park (NP), Minnesota in mid-May through June of
                       2001.  Observations (asterisks) are compared with model values from the
                       standard simulation (triangles). The simulated contributions from
                       background (diamonds), natural (crosses), and stratospheric (X's) O3
                       are also shown.
       Source:  Fiore et al. (2003a).
 1     lack of O3 >60 ppbv in July and August reflects a stronger Bermuda high-pressure system
 2     sweeping pollution from southern regions eastward before it could reach Voyageurs National
 3     Park.
 4          Frequently observed concentrations of O3 between 60  to 80 ppbv at Yellowstone NP in
 5     spring (Figures AX3-76a,b) have been attributed by Lefohn et al. (2001) to natural sources,
 6     because they occur before local park traffic starts and back-trajectories do not suggest influence
 7     from long-range transport of anthropogenic sources.  More hours with O3 >60 ppbv occur in
 8     April and May of 2001 (Table AX3-17) than in the years analyzed by Lefohn et al. (2001). Fiore
 9     et al. (2003a) used GEOS-CFIEM to interpret these events; results are shown in Figure AX3-85.
10     The mean background, natural, and stratospheric O3 contributions in March to May are higher at
11     Yellowstone (38, 22, and 8 ppbv, respectively) as compared to 27, 18, and 5 ppbv at the two
12     eastern sites previously discussed. The larger stratospheric  contribution at Yellowstone reflects
13     the high elevation of the site (2.5 km).  Fiore et al. (2003a) argued that the background at
14     Yellowstone National Park should be considered an upper limit for U.S. PRB O3 concentrations,
       August 2005
                      AX3-157     DRAFT-DO NOT QUOTE OR CITE

-------
   80
">  60
.Q
Q.
a
7 40
o
O  20
    0
                                Yellowstone NP, Wyoming (11OW, 45N)
                                                                                   -i
                                  20
                               March
                                           29
                                   April
 19
May
       Figure AX3-85.  Same as Figure AX3-85 but for Yellowstone National Park, Wyoming in
                        March to May 2001. Observations (asterisks) are compared with model
                        values from the standard simulation (triangles). The simulated
                        contributions from background (diamonds), natural (crosses), and
                        stratospheric (X's) O3 are also shown.
       Source: Fiore et al. (2003a).
 1     because of its high elevation. While Yellowstone receives a higher background concentration
 2     than the eastern sites, the model shows that regional pollution from North American
 3     anthropogenic emissions (difference between the triangles and diamonds) contributes an
 4     additional 10 to 20 ppbv to the highest observed concentrations in April and May.  One should
 5     not assume that regional photochemistry is inactive in spring.
 6          Higher-altitude western sites are more frequent recipients of subsidence events that
 1     transport high concentrations of O3 from the free troposphere to the surface.  Cooper and Moody
 8     (2000) cautioned that observations from elevated sites are not generally representative of
 9     lower-altitude sites.  At Yellowstone, the background O3 rarely exceeds 40 ppbv, but it is even
10     lower in the East.  This point is illustrated in Figure AX3-86, from Fiore et al. (2003a), with time
11     series at representative western and southeastern CASTNet sites for the month of March, when
12     the relative contribution of the background should be high. At the western sites, the background
13     is often near 40 ppbv but total surface O3 concentrations are rarely  above 60 ppbv. While
14     variations in the background play a role in governing the observed  total O3 variability at these
15     sites, regional pollution also contributes. Background concentrations are lower (often <30 ppbv)
       August 2005
                               AX3 -15 8     DRAFT-DO NOT QUOTE OR CITE

-------
   &
   Q.
   O
80
60
40
20
 0
  I

80
60
40
20
80
60
40
20
 0
  I
80
60
40
20
 0
            Great Basin NP, Nevada (114W.39N)
         0           10          20
          Grand Canyon NP, Arizona (112W.36N)
                    10          20
            Canyonlands NP, Utah (110W.38N)
         0           10          20
           Mesa Verde NP, Colorado (108W,37N)
                    10
20
              80
              60
              40
              20
               0
                I

              80
              60
              40
              20
               0
                I

              80
              60
              40
              20
               0
                I

              80
              60
              40
              20
               0
                  Caddo Valley NP, Arkansas (93W,34N)
0          10         20
    Coffeeville, Mississippi (90W,34N)
0          10         20
   Georgia Station, Georgia (84W,33N)
0          10         20
   Coweeta, North Carolina (83W,35N)
           10
                                                              20
                                       March 2001
Figure AX3-86.  Same as Figure AX3-86 but for March of 2001 at selected western
                (left column) and southeastern (right column) sites. Observations
                (asterisks) are compared with model values from the standard simulation
                (triangles). The simulated contributions from background (diamonds),
                natural (crosses), and stratospheric (X's) O3 are also shown.
Source:  Fiore et al. (2003a).
August 2005
        AX3-159     DRAFT-DO NOT QUOTE OR CITE

-------
 1     in the southeastern states where regional photochemical production drives much of the observed
 2     variability. Cooper and Moody (2000) have previously shown that the high O3 concentrations at
 3     an elevated, regionally representative site in the eastern United States in spring coincide with
 4     high temperatures and anticyclonic circulation, conditions conducive to photochemical O3
 5     production. Peak O3 concentrations in this region, mainly at lower elevations, are associated
 6     with lower background concentrations because chemical and depositional loss during stagnant
 7     meteorological conditions suppress mixing between the boundary layer and the free troposphere
 8     (Fiore et al., 2002a).  Surface O3 concentrations >80 ppbv could conceivably occur when
 9     stratospheric intrusions reach the surface. However, based on information given in Section
10     AX2.3.2, these events are rare.
11
12
13     AX3.10 OZONE EXPOSURE IN VARIOUS MICROENVIRONMENTS
14     AX3.10.1   Introduction
15          There are many definitions of exposure.  Human exposure to O3 and related photochemical
16     oxidants are based on the measured O3 concentrations in the individual's breathing zone as the
17     individual moves through time and space. Epidemiological studies generally use the ambient
18     concentrations as surrogates for exposure.  Therefore, human exposure data and models provide
19     the best link between ambient concentrations (from measurements at monitoring sites or
20     estimated with atmospheric transport models), lung deposition and clearance, and estimates of
21     air concentration-exposure-dose relationships.
22          This section discusses the current information on the available human exposure data and
23     exposure model development. This includes information on (a) the relationships between O3
24     measured at ambient monitoring sites and personal exposures and (b) factors that affect these
25     relationships. The information presented in this section is intended to provide critical links
26     between ambient monitoring data and O3 dosimetry as well as between the toxicological and
27     epidemiologic studies presented in Annexes AX4, AX5, AX6, and AX7 of this document.
28
29
30
       August 2005                           AX3-160     DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX3.10.2    Summary of the Information Presented in the Exposure
 2                   Discussion in the 1996 Ozone Criteria Document
 3           The 1996 O3 AQCD (U.S. Environmental Protection Agency, 1996a), based on then
 4     currently available information, indicated that less emphasis should be placed on O3
 5     concentrations measured at ambient monitoring stations. Fixed monitoring stations are generally
 6     used for monitoring associated with air quality standards and do not provide a realistic
 7     representation of individual exposures. Indoor/outdoor O3 ratios reported in the literature were
 8     summarized for residences, hospitals, offices, art galleries, and museums.  The differences in
 9     residential I/O were found to be a function of ventilation conditions.  The I/O ratios were less
10     than unity. In most  cases, indoor and in-transit concentrations of O3 were significantly different
11     from  ambient O3 concentrations. Ambient O3 varied from O3 concentrations measured at fixed-
12     site monitors.  Very limited personal exposure measurements were available at the time the
13     1996 O3 AQCD was published, so estimates of O3 exposure or evaluated models were not
14     provided. The two available personal  exposure studies indicated that only 40%  of the variability
15     in personal exposures was explained by the exposure models using time-weighted indoor and
16     outdoor concentrations. The discussion addressing O3 exposure modeling primarily addressed
17     work reported by McCurdy (1994) on  population-based models (PBMs).  Literature published
18     since publication of the 1996 O3 AQCD has also focused on PBMs. A discussion of individual -
19     based models (IBMs) will be included in the description of exposure modeling in this document
20     to improve our mechanistic understanding of O3 source-to-exposure events and to evaluate their
21     usefulness in providing population-based estimates.
22
23     AX3.10.3    Concepts of Human Exposure
24           Human exposure to O3 and related photochemical oxidants occurs when individuals come
25     in contact with the pollutant through "(a) the visible exterior of the person (skin and openings
26     into the body such as mouth and nostrils) or (b) the so-called exchange boundaries where
27     absorption takes place (skin, mouth, nostrils, lung, gastrointestinal tract)" (Federal Register,
28     1986).  Consequently,  exposure to a chemical, in this case O3, is the contact of that chemical
29     with the exchange boundary (U.S. Environmental Protection Agency,  1992). Therefore,
30     inhalation exposure  to O3 is based on measurements of the O3 concentration near the individual's
31     breathing zone that is not affected by exhaled air.

       August 2005                           AX3-161     DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX3.10.4   Quantification of Exposure
 2           Quantification of inhalation exposure to any air pollutant starts with the concept of the
 3     variation in the concentration of the air pollutant in the breathing zone, unperturbed by exhaled
 4     breath, as measured by a personal exposure monitor as a person moves through time and space.
 5     Since the concentrations of O3 and related photochemical oxidants vary with time and location
 6     and since people move among locations and activities, the exposure and dose received changes
 7     during the day. Furthermore, the amount of pollutant delivered to the lung is dependent upon the
 8     person's minute ventilation rate. Thus, the level of exertion is an important consideration in
 9     determining the potential exposure and dose. Inhalation exposure has been defined as the
10     integral of the concentration as a function of time over the time period of interest for each
11     individual (Ott, 1982,  1985; Lioy, 1990):
12
13                                       E=   c(t)dt                                 (AX3-2)
                                             ti

14     where E is inhalation exposure, eft) is the breathing zone concentration as a function of time and
15     tj and t2 the starting and ending time of the exposure, respectively.
16
17     AX3.10.5    Methods to Estimate Personal Exposure
1 8           There are two approaches for measuring personal exposure; direct and indirect methods
19     (Ott,  1982, 1985; Navidi et al.,  1999). Direct approaches measure the contact of the person with
20     the chemical concentration in the exposure media over an identified period of time. For the
21     direct measurement method, a personal exposure monitor (PEM) is worn near the breathing zone
22     for a  specified time to either continually  collect for subsequent analysis or directly measure the
23     concentrations of the pollutant and the exposure levels.  The indirect approach models
24     concentrations of a pollutant in specific microenvironments. Both methods are associated with
25     measurement error.
26
27
28

       August 2005                             AX3-162     DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX3.10.5.1  Direct Measurement Method
 2           The passive monitors commonly used in the direct method provides integrated personal
 3      exposure information. The monitor's sensitivity to wind velocity, badge placement, and
 4      interference with other copollutants may result in measurement error.
 5           Modified passive samplers have been developed for use in determining O3 exposure. The
 6      difficulty in developing a passive O3 monitor is in identifying a chemical or trapping reagent that
 7      can react with O3.  Zhou and Smith (1997) evaluated the effectiveness of sodium nitrite,
 8      3-methyl-2-benzothiazolinone acetone azine (MBTH),/»-acetamidophenol (p-ATP), and indigo
 9      carmine as O3-trapping reagents.  Only sodium nitrite and MBTH gave sensitive, linear
10      responses at environmentally relevant concentrations. However, MBTH overestimated the O3
11      concentrations significantly,  suggesting an interference effect. Sodium nitrite was found to be a
12      valid reagent when an effective diffusion barrier was used. Scheeren and Adema (1996) used an
13      indigo carmine-coated glass-fiber filter to collect spectrophotometrically measured O3. The
14      detection limit was 23 ppb for a 1-h exposure, with no interfering oxidants identified. The
15      reagent was valid for a relative humidity range of 20 to 80%. The uptake rate was wind velocity
16      dependent. However, wind velocity dependencies was compensated for by using a small
17      battery-operated fan that continuously blew air across the face of the monitor at a speed of
18      1.3 m/s. The overall accuracy of the sampler, after correcting for samples collected under low-
19      wind conditions, was 11 ± 9% in  comparison to a continuous UV-photometric monitor. Sample
20      stability was > 25 days in a freezer. Bernard  et al. (1999) employed a passive sampler consisting
21      of a glass-fiber filter coated with  a l,2-di(4-pyridyl)ethylene solution.  The sample was analyzed
22      spectrophotometrically after  color development by the addition of 3-methyl-2-benzothiazolinone
23      hydrazone hydrochloride.  The sampler was used at 48 sites in Montpellier, France. The results
24      from the passive sampler were highly correlated (0.9, p < 0.0001) with the results from the UV
25      absorption analyzer of the regional air quality network.  Detection limits were 17 ppb for 12-h
26      and 8 ppb for 24-h samples with an overall variation coefficient of 5% for field-tested paired
27      samples.  The imprecision was estimated to be 1.0 ppb.
28           A series of studies have been conducted using a passive sampler developed by Koutrakis
29      et al. (1993) at the Harvard School of Public Health. The sampler used sodium nitrate as the
30      trapping reagent and included a small fan to assure sufficient movement of air across the face of
31      the badge when sampling was done indoors.  The passive sampler has been evaluated against the

        August 2005                            AX3-163      DRAFT-DO NOT QUOTE OR CITE

-------
 1      standard UV absorption technique used in studies in southern California (Avol et al., 1998a;
 2      Geyh et al., 1999, 2000; Delfmo et al., 1996), Baltimore, MD (Sarnat et al., 2000), and Canada
 3      (Brauer and Brook, 1997).
 4          Avol et al. (1998a) used nitrite-coated passive samplers to measure O3 air concentrations
 5      indoors and outdoors of 126 homes between February and December 1994 in the Los Angeles
 6      metropolitan area.  The detection limit of the method was near 5 ppb. The inconsistent sampler
 7      response due to changes in wind pattern and changes in personal activity made the sampler
 8      unacceptable for widespread use.  The results of the study are discussed later in this chapter.
 9      Geyh et al. (1997,  1999) compared passive and active personal O3 air samplers based on
10      nitrite-coated glass-fiber filters.  The active sampler was more sensitive allowing for the
11      collection of short-term, 2.6-h samples. Comparison between the two samplers and UV
12      photometric O3 monitors demonstrated generally good agreement (bias for active personal
13      sampler of-6%). The personal sampler also had high precision (4% for duplicate analyses) and
14      good compliance when used by children attending summer day camp in Riverside, CA.
15
16      AX3.10.5.2 Indirect Measurement Method
17          The indirect method determines and measures the concentrations in all of the locations or
18      "microenvironments" a person encounters or determines the exposure levels through the use of
19      models or biomarkers.  The concept of microenvironments is critical for understanding human
20      exposure and aids in the development of procedures for exposure modeling using data from
21      stationary monitors (indoor and outdoor).  Microenvironments were initially defined as
22      individual or aggregate locations (and sometimes even as activities taking place within a
23      location) where a homogeneous concentration of the pollutant is encountered for a specified
24      period of time.  Thus, a microenvironment has often been identified with an "ideal" (i.e.,
25      perfectly mixed) compartment of classical compartmental modeling. More recent and general
26      definitions view the microenvironment as a "control volume," indoors or outdoors, that can be
27      fully characterized by a set of either mechanistic or phenomenological governing equations,
28      when properly parameterized, given appropriate initial and boundary conditions. The boundary
29      conditions typically reflect interactions with the ambient air and with other microenvironments.
30      The parameterizations of the governing equations generally include the information on attributes
31      of sources and sinks within each microenvironment. This type of general definition allows for
32      the concentration within a microenvironment to be nonhomogeneous, provided its spatial profile

        August 2005                            AX3-164      DRAFT-DO NOT QUOTE OR CITE

-------
 1      and mixing properties can be fully predicted or characterized. By adopting this definition, the
 2      number of microenvironments used in a study is kept manageable, while existing variabilities in
 3      concentrations are still taken into account.  The "control volume" variation could result in a
 4      series of microenvironments in the same location. If there are large spatial gradients within a
 5      location for the same time period, the space should be divided into the number of
 6      microenvironments needed to yield constant pollutant concentrations; the alternative offered by
 7      the control volume approach is to provide concentration as a function of location within it,
 8      so that the appropriate value is selected for calculating exposure. Thus, exposure to a person in a
 9      microenvironment is calculated using a formula analogous to equation AX3-3, but as the sum of
10      the discrete products of measured or modeled concentrations (specific to the receptor and/or
1 1      activity of concern) in each microenvironment by the time spent there. The equation is
12      expressed as:
13
                                             «
                                                                                       (AX3-3)
14
15      where / specifies microenvironments from 1 to n, ci is the concentration in the rth
16      microenvironment, and A ti is the duration spent in the rth microenvironment. The total exposure
17      for any time interval for an individual is the sum of the exposures in all microenvironments
18      encountered within that time interval. The concentration and time component in this approach
19      can contribute to measurement error. However, this method should provide an accurate
20      determination of exposure provided that all microenvironments that contribute significantly to
21      the total exposures are included and the concentration assigned to the microenvironment is
22      appropriate for the time period spent in  those environments.  Results from the error analysis
23      models developed by Navidi et al. (1999) indicated that neither the microenvironmental or
24      personal sampling approach gave reliable health effect estimates when measurement errors were
25      uncorrected.  The nondifferential measurement error biased the effect estimates toward zero
26      under the model assumptions.  However, if the measurement error was correlated with the health
27      response, a bias away from the null could result.
28           Microenvironments typically used to determine O3 exposures include indoor residences,
29      other indoor locations, outdoors near roadways, other outdoor locations, and in-vehicles.

        August 2005                            AX3-165      DRAFT-DO NOT QUOTE OR CITE

-------
 1      Outdoor locations near roadways are segregated from other outdoor locations because N2O
 2      emissions from automobiles alter O3 and related photochemical oxidant concentrations compared
 3      to concurrent typical ambient levels. Indoor residences are typically separated from other indoor
 4      locations, because of the time spent there and potential differences between the residential
 5      environment and the work/public environment. A special concern for O3 and related
 6      photochemical oxidants is their diurnal weekly (weekday-weekend) and seasonal variability.
 7      Few indoor O3 sources exist, but include electronic equipment, O3 generators, and copying
 8      machines.  Some secondary reactions of O3 take place indoors that produce related
 9      photochemical oxidants that could extend the exposures to those species above the estimates
10      obtained from O3 alone.  (See discussion on O3 chemistry and indoor sources and concentrations
11      later in this Annex.)
12
13      AX3.10.6  Ozone Exposure Models
14          Measurement efforts to assess population exposures or exposures to large numbers of
15      individuals over long time periods is labor intensive and costly, so exposure modeling is often
16      done for large populations evaluated over time. Predicting (or reconstructing) human exposure
17      to O3 through mechanistic models is complicated by the fact that O3 (and associated
18      photochemical oxidants) is formed in the atmosphere through a series of chemical reactions  that
19      are nonlinear and have a wide range of characteristic reaction timescales. Furthermore, these
20      reactions require the precursors VOCs and NOX that are emitted by  a wide variety of both
21      anthropogenic and natural (biogenic) emission sources. This makes O3  a secondary pollutant
22      with complex nonlinear and multiscale dynamics in time and space. Concentration levels
23      experienced by individuals and populations exposed to O3 are therefore affected by (1) emission
24      levels and spatiotemporal patterns of the gaseous precursors:  VOCs and NOX, that can be due to
25      sources as diverse as a power plant in a different state, automobiles on a highway five miles
26      away, and the gas stove in one's own kitchen; (2) ambient atmospheric as well as indoor
27      microenvironmental transport, removal and mixing processes (convective, advective, dispersive
28      and molecular/diffusional); and (3) chemical transformations that take place over a multitude of
29      spatial scales, ranging from regional/sub-continental (100 to 1000 km), to urban (10 to 100 km),
30      to local (1 to 10 km), to neighborhood (< 1 km), and to microenvironmental/personal.  These
        August 2005                            AX3-166     DRAFT-DO NOT QUOTE OR CITE

-------
 1      transformations depend on the presence of co-occurring pollutants in gas and aerosol phases,
 2      both primary and secondary, and on the nature of surfaces interacting with the pollutants.
 3           Further, the strong temporal variability of O3, both diurnal and seasonal, makes it critical
 4      that definitions of integrated or time-averaged exposure employ appropriate averaging times in
 5      order to produce scientifically defensible analyses for either causes of O3 production or health
 6      effects that result from O3 exposure. An understanding of the effect of temporal profiles of
 7      concentrations and contacts with human receptors is essential.  Short-term integrated metrics,
 8      such as hourly averages, 8-h running averages, etc., are needed to understand the relationship
 9      between O3 exposure and observed health and other effects.
10           Health effects associated with O3 have mostly been considered effects of acute exposures.
11      Peak O3 and related photochemical oxidants concentrations typically occur towards the latter
12      portion of the day during the summer months. Elevated concentrations can last for several
13      hours. Regional O3 episodes often co-occur with high concentrations of airborne fine particles
14      making it difficult to assess O3 dynamics and exposure patterns.  Furthermore, O3 participates in
15      multiphase (gas/aerosol) chemical reactions in various microenvironments. Several recent
16      studies show that O3 reacts indoors with VOCs and NOX in an analogous fashion to that
17      occurring in the ambient atmosphere (Lee and Hogsett, 1999; Wainman et al., 2000; Weschler
18      and Shields, 1997). These reactions produce secondary oxidants and other air toxics that could
19      play a significant role in cumulative human exposure and health-related effects within the
20      microenvironment.
21
22      Terminology
23           Models of human exposure to O3 can be characterized and differentiated based upon a
24      variety of attributes. For example, exposure models can be classified as (1) potential exposure
25      models, typically maximum outdoor concentration versus "actual" exposure, including locally
26      modified microenvironmental outdoor and indoor exposures; (2) population versus "specific
27      individual"-based exposure models; (3) deterministic versus probabilistic models;  and
28      (4) observation versus mechanistic air quality model-driven estimates of spatially and temporally
29      varying O3 concentration fields, etc.
30           Some points should be made regarding terminology and the directions of exposure
31      modeling research (as related specifically to O3 exposure assessments) before proceeding to

        August 2005                            AX3-167     DRAFT-DO NOT QUOTE OR CITE

-------
 1      discuss specific recent activities and developments. First, it must be understood that significant
 2      variation exists in the definitions for much of the terminology used in the published literature.
 3      The science of exposure modeling is an evolving field and the development of a "standard" and
 4      commonly accepted terminology is a process in evolution.  Second, very often procedures/efforts
 5      listed in the scientific literature as "exposure models/exposure estimates," etc., may in fact refer
 6      to only a subset of the steps or components required for a complete exposure assessment. For
 7      example, some efforts focus solely on refining the subregional or local spatiotemporal dynamics
 8      of local O3 concentrations starting from "raw" data representing monitor observations or regional
 9      grid-based model estimates.  Nevertheless, such efforts are included in the discussion of the next
10      subsection, as they can provide improved tools for the individual components that constitute a
11      complete exposure assessment. On the other hand, formulations that are identified as exposure
12      models, but focus only on ambient air quality predictions, are not included in the discussion that
13      follows, as they do not provide true exposure estimates but, rather,  ambient air estimates. These
14      models are reviewed in an earlier section of this annex. It is recognized that ambient air
15      concentrations are used as surrogates for exposure in some epidemiological studies. Third,
16      O3-exposure modeling is very often identified explicitly with population-based modeling, while
17      models describing the specific mechanisms affecting the exposure of an individual to O3, and
18      possibly some of the co-occurring gas and/or aerosol phase pollutants, are usually associated
19      with studies focusing on indoor chemistry modeling. Finally, in recent years, the focus of either
20      individual- or population-based exposure modeling research has shifted from O3 to other
21      pollutants, mostly airborne toxics and particulate matter.  However, many of the modeling
22      components that have been developed in these efforts are directly applicable to O3 exposure
23      modeling and are, therefore, mentioned in the following discussion.
24
25      A General Framework for Assessing Exposure to Ozone
26          Once the individual and relevant activity locations for Individual Based Modeling (IBM),
27      or the population and associated spatial (geographical) domain for Population Based Modeling
28      (PBM) have been defined, along with the temporal framework of the analysis (period,
29      resolution), the comprehensive modeling of individual/population exposure to O3 (and related
30      pollutants) will generally require several steps (or components, as some of them do not have to
31      be performed in sequence).  The steps represent a "composite" outline based on frameworks

        August 2005                            AX3-168     DRAFT-DO NOT QUOTE OR CITE

-------
 1      described in the literature over the last 20 years (Ott, 1982, 1985; Lioy, 1990; Georgopoulos and
 2      Lioy, 1994; U.S. Environmental Protection Agency, 1992, 1997) as well as on the structure of
 3      various existing inhalation exposure models (McCurdy, 1994; Johnson et al., 1992; Nagda et al.,
 4      1987; U.S. Environmental Protection Agency, 1996c; ICF Consulting, 2003; Burke et al., 2001;
 5      McCurdy et al., 2000; Georgopoulos et al., 2002a,b; Freijer et al., 1998; Clench-Aas et al., 1999;
 6      Kiinzli et al., 1997).  The conceptional frameworks of the models are similar. Figures
 7      AX3-87a,b provides a conceptual overview of an exposure model. The steps involved in
 8      defining exposure models include (1) estimation of the background or ambient levels of O3
 9      through geostatistical analysis of fixed monitor data, or emissions-based, photochemical, air
10      quality modeling; (2) estimation of levels and temporal profiles of O3 in various outdoor and
11      indoor microenvironments such as street canyons, residences, offices, restaurants, vehicles, etc.
12      through linear regression of available observational data sets, simple mass balance models,
13      detailed (nonlinear) gas or gas/aerosol chemistry models, or detailed combined chemistry and
14      computational fluid dynamics models; (3) characterization of relevant attributes of individuals or
15      populations under study (age, gender, weight, occupation, etc.); (4) development of activity
16      event (or exposure event) sequences for each member of the sample population or for each
17      cohort for the exposure period; (5) calculation of appropriate inhalation (in general intake) rates
18      for the individuals of concern, or the members of the sample population, reflecting/combining
19      the physiological attributes of the study subjects and the activities pursued during the individual
20      exposure events; (6) combination of intake rates and microenvironmental concentrations for each
21      activity event to assess dose; (7) calculation of event-specific exposure and intake dose
22      distributions  for selected time periods (1-h and 8-h daily maximum, O3 season averages, etc.);
23      and (8) use of PBM to extrapolate population sample (or cohort) exposures and doses to the
24      entire populations of interest. This process should aim to quantify, to the extent possible,
25      variability and uncertainty in the various components, assessing their effects on the estimates of
26      exposure.
27          Implementation of the above components of comprehensive exposure modeling has
28      benefitted significantly from recent advances and expanded availability of computational
29
        August 2005                            AX3-169     DRAFT-DO NOT QUOTE OR CITE

-------
S-
to
o
>
X
             Sector/Location data I
             (latitude/longitude)  1
            ^                 \
Site-specific
air quality and
temperature data
i
r
              Defined Study Area
             (sectors within a city
             radiusand with air and
              temperature data)
                                                                       Step 2
                                                                   Generate N number
                                                                     of simulated
                                                                      individuals   ,

f Population data /
, (age/gender/race) I
^ r
Population within
a study area

Age-specific physiological
distribution data (body
weight, height, etc)

Distribution functions for
profile variables
(e.g, probability of air
conditioner)
^^_


^^


[Com muting flow data/ \
(origin/destination
sectors) \J
^^-^Stoct
r
iastic^~~^^^
^^^^ profile generator^ —

Age group specific
employment
probabilities


Distribution functions
for daily varying profile
variables (e.g., window
status, car speed)



/A simulated individual with \
the following profile:
• Home sector
• Work sector
• Age
• Gender
• Face
• Employment status
• Gas stove
• Pilot gas light
• Air conditioner
• Height
• Weight
• Amount of hemoglobin
• Lung diffusivity
• Endogenous CO production rate
V • Plus 12 more variables J

O
O
2
O
H
O
c
o
H
W
O
V
O
HH
H
W

r r \ - National
(^ ~^) -Simulation <^
step —

- Area-speciTic
J> - Data processor

| -intermediate step
^^^]J - Output data
        Figure AX3-87a.  Detailed diagram illustrating components of an exposure model.

        Source: ICF Consulting and ManTech Environmental Technology, Inc. (2003)

-------
 S-
 to
 o
>
x
 fe
 H
 6
 O

 O
 H
O
                                / Diary evenfs/activlties  /
                                 and personal information
                                I   (e.g., from CHAD)    I
                                  Activity diary pools by day
                                  type/temperature category
                                                                     The selected diary records
                                                                        for each day in the
                                                                         simulation period
                                                                                                                                            Step 5
                                                                                                                                      Determine exposure
                                                                                                                                       for each simulated
                                                                                                                                          individual
                                                                                                       aleulate concentrator
                                                                                                       in microenvironments
                                                                                                        for each simulated
                                                                                                            individual
                                        Identify
                                    concentrations in
                                    microenviroments
                                        visited
                                                                                                   Microerwironiments denned by
                                                                                                  grouping of CHAD location codes
                                                                                                  Select calculation methods for each
                                                                                                  microenvironment:
                                                                                                  » Factors
                                                                                                  * Mass balance
                                                                                                  Hourly ambient air quality data for
                                                                                                     the home and work sectors
                                   Each day in the simulation
                                    period, is assigned with
                                    activity pool # based on
                                  daily max/mean temperature
                                                                       A sequence of events
                                                                      (i.e., microenvironments
                                                                     visited and minutes spent)
                                                                      in the simulation period
Hourly concentrations in ail
microenvironments during
    simulation period
                                                                                                                                    Hourly concentrations and
                                                                                                                                      minutes spent in each
                                                                                                                                    microenviroment visited by
                                                                                                                                     the simulated individual
                                     Maximum/mean daily
                                      temperature data
                                                              J Physiological parameters
                                                                     from profile
Average exposure
concentrations
for simulated person;
* Hourly
* Daily
» Monthly
  Annual
Average does for
simulated person;
* Hourly
* Daily
* Monthly
  Annual
          Figure AX3-87b.   Detailed diagram illustrating components of an exposure model.

          Source:  ICF Consulting and ManTech Environmental Technology, Inc. (2003)
o
HH
H
W

-------
 1     technologies such as Relational Database Management Systems (RDBMS) and Geographic
 2     Information Systems (Purushothaman and Georgopoulos, 1997, 1999a,b).
 3
 4     AX3.10.6.1   Population Exposure Models
 5          Existing comprehensive inhalation exposure models treat human activity patterns as
 6     sequences of exposure events in which each event is defined by a geographic location and
 7     microenvironment.  The U.S. EPA has supported the most comprehensive efforts in this area,
 8     leading to the development of the National Ambient Air Quality Standard Exposure Model and
 9     Probabilistic National Ambient Air Quality Standard Exposure Model (NEM and pNEM)
10     (Johnson, 2003) and the Modeling Environment for Total Risk Studies/Simulation of Human
11     Exposure and Dose System (MENTOR/SHEDS) (McCurdy et al., 2000).  The Total Risk
12     Integrated Methodology Inhalation Exposure (TRIM.Expo) model, also referred to as the Air
13     Pollutants Exposure (APEX) model, was developed by the U.S. EPA as a tool for estimating
14     human population exposure to criteria and air toxic pollutants. TRIM.Expo serves as the human
15     inhalation exposure model within the Total Risk Integrated Methodology (TRIM) framework
16     (ICF Consulting and ManTech Environmental Technology,  Inc. (2003)). TRIM.Expo, a PC-
17     based model derived from the probabilistic NAAQS Exposure Model (pNEM), was used in the
18     last O3 NAAQS review (Johnson et al., 1996a, 1996b). Over the past five years, TRIM.Expo has
19     undergone several significant improvements in the science reflected in the model and in the
20     databases input to the model.
21          Recent European efforts have produced some formulations that have similar general
22     attributes as the above models but generally involve major simplifications in some of their
23     components. Examples  of recent European models addressing O3 exposures include the AirPEx
24     (Air Pollution Exposure) model  (Freijer et al., 1998), which basically replicates the pNEM
25     approach, and the AirQUIS (Air Quality Information System)  model (Clench-Aas et al.,  1999).
26          The NEM/pNEM,  TRIM.Expo, and MENTOR/SHEDS families of models provide
27     exposure estimates, defined by concentration and minute ventilation rate for each individual
28     exposure event, and provide  distributions of exposure  and O3 dose over any averaging  period of
29     concern from 1 h to an entire O3 season.  The above families of models also simulate certain
30     aspects of the variability and uncertainty in the principal factors affecting exposure. pNEM
31     divides the population of interest into representative cohorts based on the combinations of

       August 2005                           AX3-172     DRAFT-DO NOT QUOTE OR CITE

-------
 1      demographic characteristics (age, gender, employment), home/work district, residential cooking
 2      fuel, and then assigns activity diary records (Glen et al., 1997) to each cohort according to
 3      demographic characteristic, season, day-type (weekday/weekend), and temperature. TRIM.Expo
 4      and MENTOR/SHEDS generates a population demographic file containing a user-defined
 5      number of person-records for each census tract of the population based on proportions of
 6      characteristic variables (age, gender, employment, housing) obtained for the population of
 7      interest, and then assigns the matching activity information based on the characteristic variables.
 8      A discussion of databases on time-activity data, and their influence on estimates of long-term
 9      ambient O3 exposure, can be found in Kiinzli et al. (1997), McCurdy (2000), and McCurdy et al.
10      (2000).
11          More recent exposure models are designed (or have been redesigned) to obtain such
12      information from CHAD (Consolidated Human Activities Database; www.epa.gov/chadnetl:
13      see Table AX3-18).  There are now about 22,600 person-days of sequential daily activity pattern
14      data in CHAD. All ages of both genders are represented in CHAD.  The data for each subject
15      consist of one or more days of sequential activities, in which each activity is defined by
16      start time, duration, activity type (140 categories), and microenvironment classification
17      (110 categories). Activities vary from 1 min to 1 h in duration. Activities longer than 1  h are
18      subdivided into clock-hour durations to facilitate exposure modeling.  A distribution of values
19      for the ratio of oxygen uptake rate to body mass (referred to as metabolic equivalents or METs)
20      is provided for each activity type listed. The forms and parameters of these distributions were
21      determined through an extensive review of the exercise and nutrition literature.  The primary
22      source of distributional data was Ainsworth et al. (1993), a compendium developed specifically
23      to "facilitate the coding of physical activities and to promote comparability across  studies."
24      Other  information on activity patterns has been reported by Klepeis et al. (1996, 2001); Avol
25      et al. (1998b); Adams (1993); Shamoo et al.  (1994); Linn et al. (1996); Kunzli et al. (1997).
26          Use of the information in CHAD provides a rational way for incorporating realistic intakes
27      into exposure models by linking inhalation rates to activity information.  As mentioned earlier,
28      an exposure event sequence derived from activity-diary data is assigned to each population unit
29      (cohort for pNEM- or REHEX-type models, or individual for TRIM.Expo or MENTOR/SHEDS-
30      type models).  Each exposure event is typically defined by a start and duration time, a
31      geographic location and microenvironment, and activity level. The most recent pNEM,

        August 2005                             AX3-173     DRAFT-DO NOT QUOTE OR CITE

-------
                         Table AX3-18.  Activity Pattern Studies Included in the Consolidated Human Activity Database (CHAD)
•s
S-
to
o
o





Study Name
Baltimore
CARB: Adolescents and Adults
CARB: Children
Cincinnati (EPRI)
Denver (EPA)
Los Angeles: Elem. School
Calendar Time
Period
of the Study
Jan-Feb 1997
Jul-Aug 1998
Oct 1987-Sept 1988
Apr 1989-Feb 1990
Mar- Apr and Aug
1985
Nov 1982-Feb 1983
Oct 1989
Diary
Age1
65+
12-94
0-11
0-86
18-70
10-12
Days2
391
1762
1200
2614
805
51
Type3
Diary; 15-min
blocks
Retrospective
Retrospective
Diary
Diary
Diary
Time4
24-h Standard
24-h Standard
24-h Standard
24 h; nominal
7 pm-7 am
24 h; nominal
7 pm-7 am
24-h Standard
Rate5
No
No
No
Yes
No
Yes
Documentation or
Reference
Williams et al. (2000a,b)
Robinson et al. (1991)
Wiley etal. (1991a)
Wiley etal. (1991b)
Johnson (1989)
Akland etal. (1985)
Johnson (1984)
Spier etal. (1992)
Notes
Multiple days, varying from 5-15;
part of a PM25 PEM study


3 consecutive days; 186 P-D
removed7
Part of CO PEM6 study; 2 consec.
days; 55 P-D removed7
7 P-D removed7
 >
 X
 O
 o
 2
 o
 H
o
 c
 o
 H
 W
 O
 V
 O
 HH
 H
 W
Children

Los Angeles: High School
Adoles.

National: NHAPS-A8
           National: NHAPS-B8

           University of Michigan:
           Children

           Valdez, AK
Washington, DC (EPA)
                                           Sept-Oct 1990
                                                                 13-17
                                   43     Diary
                                                                                                   24-h Standard
Yes     Spier etal. (1992)
                                                                                                                                                       23 P-D removed7
                               Sept 1992-Oct 1994      0-93      4723    Retrospective    24-h Standard


                               As above               0-93      4663    Retrospective    24-h Standard

                               Feb-Dec 1997          0-13      5616    Retrospective    24-h Standard
Nov 1990-Oct 1991      11-71      401     Retrospective    Varying 24-h
                                                         period

Nov 1982-Feb 1983      18-98      699     Diary           24 h; nominal
                                                          7 pm-7 am
                                                                                                                     No9    Klepeis etal. (1995)         A national random-probability
                                                                                                                            Tsang and Klepeis (1996)     survey
                                                                          No9     As above

                                                                          No     Institute for Social
                                                                                  Research (1997)
                                  As above

                                  2 days of data:  one is a weekend
                                  day
                                                                                                          No     Goldstein et al. (1992)        4 P-D removed7
No    Akland etal. (1985)
       Hartwell et al. (1984)
Part of a CO PEM6 study; 6 P-D
removed7
1 All studies included both genders. The age range depicted is for the subjects actually included; in most cases, there was not an upper limit for the adult studies. Ages are inclusive.  Age 0 =
 babies < 1 year old.
2 The actual number of person-days of data in CHAD after the "flagging" and removal of questionable data. See the text for a discussion of these procedures.
3 Retrospective:  a "what did you do yesterday" type of survey; also known as an ex post survey. Diary: a "real-time" paper diary that a subject carried as he or she went through the day.
4 Standard = midnight to midnight.
5 Was activity-specific breathing rate data collected?
6 PEM = a personal monitoring study. In addition to the diary, a subject carried a small CO or PM2 5 monitor throughout the sampling period.
7 P-D removed = The number of person-days of activity pattern data removed from consolidated CHAD because of missing activity and location information; completeness criteria are listed in
 the text.
8 National Human Activity Pattern Study; A = the air version; B = the water version. The activity data obtained on the two versions are identical.
9 A question was asked regarding which activities (within each 6-h time block in the day) involved "heavy breathing," lifting heavy objects, and running hard.
           Source: U.S. Environmental Protection Agency (2004a)

-------
 1      TRIM.Expo, and MENTOR/SHEDS models have defined activity levels using the activity
 2      classification coding scheme incorporated into CHAD.  A probabilistic module within the
 3      TRIM.Expo and MENTOR/SHEDS-type models converts the activity classification code of each
 4      exposure event to an energy expenditure rate, which in turn is converted into an estimate of
 5      oxygen uptake rate. The oxygen uptake rate is then converted into an estimate of ventilation rate
 6      (Vs), expressed in L/min. Johnson (2001) reviewed the physiological principles incorporated
 7      into the algorithms used in pNEM and TRIM.Expo to convert each activity classification code to
 8      an oxygen uptake rate and describes the additional steps required to convert oxygen uptake
 9      toVfi.
10           McCurdy (1997a,b, 2000) recommended that ventilation rate be estimated as a function of
11      energy expenditure rate. The energy expended by an individual during a particular activity can
12      be expressed as:
13
                                      EE  = (MET)(RMR)                               (AX3 -4)
14
15      where EE is the average energy expenditure rate (kcal/min) during the activity, MET (metabolic
16      equivalent of work) is a ratio specific to the activity and is dimensionless, and RMR is the
17      resting metabolic rate of the individual expressed in terms of number of energy units expended
18      per unit of time (kcal/min).  If RMR is specified for an individual, then the above equation
19      requires only an activity-specific estimate of MET to produce an estimate of the energy
20      expenditure rate for a given activity. McCurdy et al. (2000) developed MET distributions for the
21      activity classifications appearing in the CHAD database.
22           An important source of uncertainty in existing exposure modeling involves the creation of
23      multiday, seasonal, or year long exposure activity sequences based on 1- to 3-day activity data
24      for any given individual from CHAD. Currently, appropriate longitudinal data are not available
25      and the existing models use various rules to derive longer-term activity sequences using 24-h
26      activity data from CHAD.
27           The pNEM family of models used by the EPA has evolved considerably since the
28      introduction of the first NEM model in the 1980s (Biller et al., 1981).  The first such
29      implementations of pNEM/O3 in the 1980s used a reduced form of a mass balance equation to
        August 2005                            AX3-175     DRAFT-DO NOT QUOTE OR CITE

-------
 1      estimate indoor O3 concentrations from outdoor concentrations. The second generation of
 2      pNEM/O3 was developed in 1992 and used a simple mass balance model to estimate indoor O3
 3      concentrations. Subsequent enhancements to pNEM/O3 and its input databases included
 4      revisions to the methods used to estimate equivalent ventilation rates (ventilation rate divided by
 5      body surface), to determine commuting patterns, and to adjust ambient O3 levels to simulate
 6      attainment of proposed NAAQS. During the mid-1990s, the EPA applied updated versions of
 7      pNEM/O3 to three different population groups in nine selected urban areas (Chicago, Denver,
 8      Houston, Los Angeles, Miami, New York, Philadelphia, St. Louis, and Washington): (1) the
 9      general population of urban residents, (2) outdoor workers, and (3) children who tended to spend
10      more time outdoors than the average child. Reports by Johnson et al. (1996a,b,c) describe these
11      versions of pNEM/O3 and summarize the results of the application of the model to the nine urban
12      areas. These versions of pNEM/O3 used a revised probabilistic mass balance model to determine
13      O3 concentrations over 1-h periods in indoor and in-vehicle microenvironments (Johnson, 2003).
14      The model assumed that there are no indoor sources of O3, that the outdoor O3 concentration and
15      AER during the clock hour is constant at a specified value, and that O3 decays  at a rate
16      proportional to the outdoor O3 concentration and the indoor O3 concentration.
17           The new pNEM-derived model, TRIM.Expo,  differs from earlier pNEM  models in that the
18      probabilistic features of the model are incorporated into a Monte Carlo framework. Instead of
19      dividing the population of interest into a set of cohorts, TRIM.Expo generates  individuals as if
20      they were being randomly sampled from the population.  TRIM.Expo provides each generated
21      individual with a demographic profile that specifies values for all parameters required by the
22      model.  The values are selected from distributions and databases that are specific to the age,
23      gender, and other specifications  stated in the demographic profile. The EPA plans to develop
24      future versions of TRIM.Expo applicable to O3 and other criteria pollutants.
25           The latest version of TRIM.Expo allows for finer geographical units such as census tracts
26      and automatically assigns population to the nearest  monitor within a cutoff distance. Exposure
27      district-specific temperatures can be specified and the user can select the variables that affect
28      each parameter (e.g., the AER parameter in certain indoor microenvironments  may depend on air
29      conditioning status or window position).  The mass balance algorithms have been enhanced to
30      allow window position or vehicle speed to also be considered in determining AERs.
        August 2005                            AX3-176     DRAFT-DO NOT QUOTE OR CITE

-------
 1           The TRIM.Expo model simulates individual movement through time and space to provide
 2      an estimate of exposure to a given pollutant in the indoor, outdoor, and in-vehicle
 3      microenvironments.  The model is highly versatile, allowing input data for specific applications.
 4      TRIM.Expo provides a good balance in terms of precision and resource expenditure compare
 5      with the more narrowly focused site-specific model and the broadly applicable national
 6      screening-level models.
 7           A key strength of TRIM.Expo is its ability to estimate hourly exposures and doses for all
 8      simulated individuals in the sampled population.  TRIM.Expo is capable of estimating exposures
 9      of workers in the geographic area where they work, in addition to the geographic area where
10      they live.  TRIM.Expo is able to represent much of the variability in the exposure estimates
11      resulting from the variability of the factors affecting human exposure by incorporating stochastic
12      processes representing the natural variability of personal profile characteristics, activity patterns,
13      and microenvironment parameters .
14           A limitation of TRIM.Expo is that uncertainty in the predicted distributions has not been
15      addressed. Certain aspects of the personal profiles are held constant (e.g., age) which could be
16      an issue for simulations with long timeframes. The combined data set for activity patterns
17      (CHAD) are from a number of different studies and may not constitute a representative sample.
18      However, the largest portion of CHAD (about 40 percent) is from a study of national scope and
19      research has shown that activity patterns are generally similar once you take into account age,
20      gender, day of week, and season/temperature. The commuting data addresses only home-to-
21      work travel and may not accurately reflect current commuting patterns. The population not
22      employed outside the home is assumed to always remain in the residential census tract.
23      Although several of the TRIM.Expo microenvironments account for time spent in travel, the
24      travel is assumed to always occur in basically a composite of the home and work tract.  Seasonal
25      or year long sequences for a simulated individual are created by sampling human activity data
26      from more than one subject, possibly causing an underestimation of the variability from person
27      to person and an overestimation of the day to day variability for any given individual. The
28      model does not capture certain  correlations among human activities that can impact
29      microenvironmental concentrations (e.g., cigarette smoking leading to an individual opening a
30      window, which in turn affects the amount of outdoor air penetrating the residence).
        August 2005                            AX3-177     DRAFT-DO NOT QUOTE OR CITE

-------
 1          MENTOR/SHEDS estimates the population distribution of pollutant exposure by randomly
 2      sampling from various input distributions.  MENTOR/SHEDS is capable of simulating
 3      individuals exposures in eight microenvironments (outdoors, residence, office, school, store,
 4      restaurant, bar, and vehicles) using spatial concentration data for each census tract for outdoor
 5      pollutant concentrations. The indoor and in-vehicle pollutant concentrations are calculated using
 6      specific equations for the microenvironment and ambient pollutant concentration relationship.
 7      Model simulations use demographic data at the census tract level. Randomly selected
 8      characteristics for a fixed number of individual are selected to match demographics within the
 9      census tract for age, gender, employment status, and housing type. Smoking prevalence
10      statistics by gender and age is randomly selected for each individual in the simulation.  Diaries
11      for activity patterns are matched for the simulated individual by demographic characteristics.
12      The essential attributes of some of the O3 exposure models and approaches are summarized in
13      Table AX3-19.
14          Rifai et al. (2000) compared applications of an updated version of REHEX, REHEX-II.
15      The applications used NHAPS data for the southern states and the 48-state NHAPS or the
16      Houston-specific time-activity pattern data. The results indicated a sensitivity to the specificity
17      of the activity data:  using Houston-specific data resulted in higher estimates of human exposure
18      in some of the scenarios. For example, using NHAPS data lead to an estimated 275 thousand-
19      exposure-hours between 120 to 130 ppb, while use of the Houston-specific activity data lead to
20      an estimated 297 thousand-exposure-hours between 120 and 130 ppb (8% higher). Using the
21      Houston-specific activity data in the model resulted in about 2,400 person-exposure-hours above
22      190 to 200 ppb O3 while no exposure above this threshold was estimated when the NHAPS
23      activity were used in the model.
24          Of the above families of models only NEM/pNEM implementations have been extensively
25      applied to O3 studies. However, it is anticipated that TRIM.Expo will be useful as an exposure
26      modeling tool for assessing both criteria and hazardous air pollutants in the future. The 1996 O3
27      AQCD (U.S. Environmental Protection Agency, 1996a) focused on the pNEM/O3 family of
28      models, referring to the review by McCurdy (1994) for the fundamental principles underlying its
29      formulation and listing, in addition to the "standard" version, three pNEM/O3-derived models
30      (the Systems Applications International NEM [SAI/NEM]; the Regional Human Exposure
31      Model [REHEX]; and the Event Probability Exposure Model [EPEM]).

        August 2005                             AX3-178      DRAFT-DO NOT QUOTE OR CITE

-------
                                        Table AX3-19. Personal and Population Exposure Models for Ozone
S-
to
o
         Model Name
Model Type
Microenvironments or Predictors
Notes
Reference
>
X
         pNEM
         TRIM.Expo
         Mentor/SHEDS
         REHEX
Probabilistic      General population, outdoor workers,
                 outdoor children
PC-based         Outdoors, indoor residence,
                 in-vehicle
Probabilistic      General population, outdoors, indoors,
                 in-vehicle
                 General population
                                     Provides estimates of exposure within a defined    Johnson et al.
                                     population for a specified period of time.          (1996a,b,c)
                                     Uses activity records from CHAD and CADS
                                     Simulates movement through time and space.
                                     Estimates hourly exposures and doses.
                                     Uncertainties in predicted distributions have not
                                     been addressed.
                                     Employees detailed person oriented exposure
                                     approach that includes personal activity data,
                                     physiology, and microenvironmental conditions.
                                     Allows calculation of exposure and dose for
                                     each activity event.

                                     Provides estimates of exposure for 1-day to
                                     3 yrs. Can represent variability in activities of
                                     the population to capture extremes in exposure
                                     distributions.
                                             ICF Consulting and
                                             ManTech
                                             Environmental
                                             Technology, Inc.
                                             (2003)

                                             Georgopoulos et al.
                                             (2005)
                                             Lurmann and Colome
                                             (1991)
O
O
2
o
H
O
c
o
H
W
O
V
O
HH
H
W

-------
 1      AX3.10.6.2  Ambient Concentrations Models
 2           As mentioned earlier, background and regional outdoor concentrations of pollutants over a
 3      study domain may be calculated either through emissions-based mechanistic modeling or
 4      through ambient-data-based modeling.  Emissions-based models calculate the spatiotemporal
 5      fields of the pollutant concentrations using precursor emissions and meteorological conditions as
 6      inputs.  The ambient-data-based models typically calculate spatial or spatiotemporal
 7      distributions of the pollutant through the use of interpolation schemes, based on either
 8      deterministic or stochastic models for allocating monitor station observations to the nodes of a
 9      virtual regular grid covering the region of interest. (See later discussion on population exposure
10      models).  Kriging, a geostatistical technique, provides standard procedures for generating an
11      interpolated O3 spatial distribution for a given time period, using data from a set of observation
12      points.  The kriging approach, with parameters calculated  specifically for each hour of the period
13      of concern, was compared to the Urban Airshed Model (UAM-IV), a comprehensive
14      photochemical grid-based model for deriving concentration fields.  The concentration fields
15      were then linked with corresponding population data to calculate potential outdoor population
16      exposure.  Higher exposure estimates were obtained with the photochemical grid-based model
17      when O3 concentrations were <120 ppb, however, the situation was reversed when O3
18      concentrations exceeded 120 ppb. The authors concluded that kriging O3 values at the locations
19      studied can reconstruct aspects of population exposure distributions (Georgopoulos et al.,
20      1997a,b).
21           Carroll et al. (1997a,b) developed a spatial-temporal model, with a deterministic trend
22      component, to model hourly O3 levels with the capacity to predict O3 concentrations at any
23      location in Harris County, Texas during the time period between 1980 and  1993.  A fast model-
24      fitting method was developed to handle the large amount of available data and the substantial
25      amount of missing data.  Ozone concentration data used consisted of hourly measurements from
26      9 to 12  monitoring stations for the years 1980 to 1993. Using information from the census tract,
27      the authors estimated that exposure of young children to O3 declined by approximately 20%  over
28      the analysis period. The authors also suggested that the O3 monitors are not sited in locations to
29      adequately measure population exposures.  Several researchers have questioned the suitability of
30      the model for addressing spatial variations in O3 (Guttorp  et al., 1997; Cressie, 1997; Stein and
31      Fang, 1997).

        August 2005                            AX3-180     DRAFT-DO NOT QUOTE OR CITE

-------
 1           Spatiotemporal distributions of O3 concentrations have alternatively been obtained using
 2     methods of the "Spatio-Temporal Random Field" (STRF) theory (Christakos and Vyas,
 3     1998a,b). The STRF approach interpolates monitoring data in both space and time
 4     simultaneously.  This method can analyze information on "temporal trends," which cannot be
 5     incorporated directly in purely spatial interpolation methods such as standard kriging. Further,
 6     the STRF method can optimize the use of data that are not uniformly sampled in either space or
 7     time. The STRF theory was further extended in the Bayesian Maximum Entropy (BME)
 8     framework and applied to O3 interpolation studies (Christakos and Hristopulos, 1998; Christakos
 9     and Kolovos, 1999; Christakos, 2000).  The BME framework can use prior information in the
10     form of "hard data" (measurements), probability law descriptors (type of distribution, mean and
11     variance), interval estimation (maximum and minimum values) and even constraint from
12     physical laws.  According to these researchers, both STRF and BME were found to successfully
13     reproduce O3 fields when adequate monitor data are available.
14
15     AX3.10.6.3   Microenvironmental Concentration Models
16           Once specific ambient/local spatiotemporal O3 concentration patterns have been derived,
17     microenvironments that can represent either outdoor or indoor settings must be characterized.
18     This process can involve modeling of various local sources and sinks as well as
19     interrelationships between ambient/local and microenvironmental concentration levels. Three
20     approaches have been used in the past to model microenvironmental concentrations: empirical,
21     mass balance, and detailed computational fluid dynamics (CFD).
22           The empirical fitting approach has been used to summarize the findings of recent field
23     studies (Liu et al., 1995, 1997; Avol et al., 1998a). These empirical relationships could provide
24     the basis for future, "prognostic" population exposure models.
25           Mass balance modeling has ranged from very simple formulations, assuming ideal
26     (homogeneous) mixing and only linear physicochemical transformations with sources and sinks,
27     to models that take into account complex multiphase chemical and physical interactions and
28     nonidealities in mixing. Mass balance modeling is the most common approach used to model
29     pollutant concentrations in enclosed microenvironments.  As discussed earlier, the simplest
30     microenvironmental setting is a homogeneously mixed compartment in contact with possibly
31     both outdoor/local environments as well as with other microenvironments. The air quality of

       August 2005                           AX3-181     DRAFT-DO NOT QUOTE OR CITE

-------
 1     this idealized microenvironment is affected primarily by transport processes (including
 2     infiltration of outdoor air into indoor air compartments, advection between microenvironments,
 3     and convective transport); sources and sinks (local outdoor emissions, indoor emissions, surface
 4     deposition); and local outdoor and indoor gas and aerosol phase chemistry transformation
 5     processes (such as the formation of secondary organic and inorganic aerosols).
 6          Numerous indoor air quality modeling studies have been reported in the literature;
 7     however, depending on the modeling scenario, only a limited number address physical and
 8     chemical processes that affect O3 concentrations indoors (Nazaroff and  Cass, 1986; Hayes, 1989,
 9     1991). An example of a mass balance indoor air model for O3 and benzene can be found in the
10     work of Freijer and Bloemen (2000). They used outdoor O3 measurements to parameterize a
1 1     simplified linearized formulation of transport, transformation, and sources and sinks in the
12     indoor microenvironment.
13          The pNEM/O3 model includes a sophisticated mass balance model for enclosed (indoor and
14     vehicle)  microenvironments. The general form of this mass balance model is a differential
15     equation that accounts for outdoor concentration, AER, penetration rate, decay rate, and indoor
16     sources.  Each of these parameters is represented by a probability distribution or by a dynamic
17     relationship to other parameters that may change according to time of day, temperature, air
18     conditioning status, window status, or other factors (Johnson, 2003).  The simplest form of the
19     model is represented by the following differential equation for a perfectly mixed
20     microenvironment without an air cleaner:
21
                                  d(~* IN    s-<       o     ,~<
                                  —L = vCOUT +  —- vCm                          (AX3-5)
22
23     where dCIN is the indoor pollutant concentration (mass/volume), t is time in hours, v is the air
24     exchange rate, COUT is the outdoor pollutant concentration (mass/volume), Fis the volume of the
25     microenvironment, and S is the indoor source emission rate.
26          Nazaroff and Cass (1986) extended the mass balance model to include multiple
27     compartments and interactions between different compounds.  The extended model takes into
28     account the effects of ventilation, filtration, heterogeneous removal, direct emission, and
        August 2005                            AX3-182     DRAFT-DO NOT QUOTE OR CITE

-------
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
photolytic and thermal and chemical reactions.  A more in-depth discussion of the mass balance
model may be found in Shair and Heitner (1974) and in Nazaroff and Cass (1986).
     Freijer and Bloemen (2000) used the one-compartment mass balance model to examine the
relationship between O3 I/O ratios as influenced by time patterns in outdoor concentrations,
ventilation rate, and indoor emissions. The microenvironment was 250 m3. Three different
ventilation patterns with the same long-term average AER (0.64 IT1) were used. A source
pattern (direct emissions) of zero was used, because O3 sources are not common.  The time series
for outdoor O3 concentrations consisted of 100-day periods during the summer, with hourly
measured concentrations.  The following assumptions were made:  (1) O3 concentration in the air
that enters the microenvironment is equal to the concentration of the outside air minus the
fraction removed by filtration, (2) the O3 concentration that leaves the microenvironment equals
the O3 concentration in the microenvironment, (3) the decay processes in the microenvironment
are proportional to the mass of the pollutant, and (4) addition or removal of O3 in the
microenvironment also may occur through independent sources and sinks.
     Figure  AX3-88 represents the measured outdoor O3 concentrations and modeled indoor O3
                                63
                                51 -
                            O. 37.6 -
                            0)
                            i   25H
                            o
                              12.5-
                                0
                                  100
                                  101
102
103
104
105
                                                 Time (day)
Figure AX3-88.   Measured outdoor O3 concentrations (thin line) and modeled indoor
                 concentrations (bold line).
Source:  Adapted from Freijer and Bloemen (2000).
concentrations. The indoor modeled O3 concentrations were found to equal approximately 33%
of the outdoor monitored concentrations.
       August 2005
                                       AX3-183
           DRAFT-DO NOT QUOTE OR CITE

-------
 1          Few indoor air models have considered detailed nonlinear chemistry, which, however, can
 2      have a significant effect on the indoor air quality, especially in the presence of strong indoor
 3      sources.  Indeed, the need for more comprehensive models that can take into account the
 4      complex, multiphase processes that affect indoor concentrations of interacting gas phase
 5      pollutants and particulate matter has been recognized and a number of formulations have
 6      appeared in recent years. For example, the Exposure and Dose Modeling and Analysis System
 7      (EDMAS) (Georgopoulos et al., 1997a) included an indoor model with detailed gas-phase
 8      atmospheric chemistry. This indoor model accounts for interactions of O3 with indoor sinks and
 9      sources (surfaces, gas releases) and with entrained gas.  The indoor model was dynamically
10      coupled with the outdoor photochemical air quality models, UAM-IV and UAM-V (Urban
11      Airshed Models), which provided the gas-phase composition of entrained air, and with a
12      physiologically based O3 uptake and dosimetry model.  Subsequent work (Isukapalli and
13      Georgopoulos, 2000; Isukapalli et al.,  1999) expanded the framework and features of the
14      EDMAS model to incorporate alternative representations of gas-phase chemistry as well as
15      multiphase O3 chemistry and gas/aerosol interactions. The new model is a component of the
16      integrated Modeling Environment for Total Risk studies (MENTOR).
17          Sarwar et al. (2001, 2002) modeled estimates of indoor hydroxyl radical concentrations
18      using a new indoor air quality  model, Indoor Chemistry and Exposure Model (ICEM). The
19      ICEM uses a modified SAPRC-99 atmospheric chemistry mechanism to simulate indoor
20      hydroxyl radical production and consumption from reactions of alkenes with O3. It also allows
21      for the simulation of transport processes between indoor and outdoor environments, indoor
22      emissions, chemical reactions, and deposition. Indoor hydroxyl radicals, produced from O3 that
23      penetrates indoors, can adversely impact indoor air quality through dark chemistry to produce
24      photochemical oxidants.
25          S0rensen and Weschler (2002) used CFD modeling to examine the production of a
26      hypothetical product from an O3-terpene reaction under two different ventilation scenarios.  The
27      computational grid used in the model was nonuniform.  There were significant variations in the
28      concentrations of reactants between locations in the room, resulting in varying reaction rates.
29      Because the "age of the air" differed at different locations in the room, the time available for
30      reactions to occur also differed between locations.
        August 2005                           AX3-184     DRAFT-DO NOT QUOTE OR CITE

-------
 1          Very few studies have focused on mechanistic modeling of outdoor microenvironments.
 2     Fraigneau et al. (1995) developed a simple model to account for fast NO-O3 reaction/dispersion
 3     in the vicinity of a motorway.  Proyou et al. (1998) applied a simple three-layer photochemical
 4     box model to an Athens street canyon. However, the adjustments of O3 levels for sources, sinks,
 5     and mixing in outdoor microenvironments are done in a phenomenological manner in existing
 6     exposure models, driven by limited available observations.  On-going research is evaluating
 7     approaches for quantifying local effects on outdoor O3 chemistry in specific settings.
 8          Finally, one issue that should be mentioned is that of evaluating comprehensive prognostic
 9     exposure modeling studies, for either individuals or populations, with field data. Attempts had
10     been made to evaluate pNEM/O3-type models using personal exposure measurements (Johnson
11     et al., 1990).  Although databases that would be adequate for performing a comprehensive
12     evaluation are not expected to be available any time soon, a number of studies are building the
13     necessary information base, as discussed previously.  Some of these studies report field
14     observations of personal, indoor, and outdoor O3 concentrations and describe simple
15     semiempirical personal exposure models that are parameterized using observational data and
16     regression techniques.
17
18     AX3.10.7   Measured Exposures and Monitored  Concentrations
19     AX3.10.7.1   Personal Exposure Measurements
20          Passive O3 monitors have been used in several field studies to determine average daily O3
21     exposure as well as in scripted studies to evaluate O3 exposures over one to  several-hour time
22     periods.  Table AX3-20 list the results of O3 exposure studies.  Delfmo et al. (1996) measured
23     12-h personal daytime O3 exposures in asthmatic patients in San Diego from September through
24     October 1993. They found that the mean personal exposures were 27% of the mean outdoor
25     concentrations. Individual exposure levels among the 12 asthmatic subjects aged 9 to 16 years
26     varied greatly. Mean personal O3 exposure levels were lower  on Friday, Saturday, and Sunday
27     than on other days of the week (10 versus 13 ppb), while the ambient air concentrations were
28     higher Friday through Sunday. The authors suggested that the differences were due to higher
29     weekday NO emissions from local traffic which titrated the  ambient O3 levels. The lower
30     personal  exposure levels on Friday, Saturday, and Sunday may have been an artifact of greater
31     noncompliance, with the badges remaining indoors and, therefore, being exposed to lower O3

       August 2005                             AX3-185     DRAFT-DO NOT QUOTE OR CITE

-------
                            Table AX3-20. Personal Exposure Measurements
        Location, Population, Sample Duration
       Personal Exposure Mean3
            (range) (ppb)
                       Reference
        San Diego, CA, Asthmatics ages 9-18 years,      12
        12 hour
        Vancouver, Canada, Adult Workers, Daily      585
          High indoor time
          Moderate indoor time
          Only outdoor

        Southern California, Subjects 10-38 years       24
          Spring
          Fall

        Montpellier, France, Adults, Hourly           16
          Winter
          Summer

        Souther California, Children 6-12 years,        169
        > 6 days
          Upland   - winter
                   - summer
          Mountain - winter
                   - summer

        Baltimore, MD, Technician, Hourlyb            1
          Winter
          Summer

        Baltimore, MD, Adults 75 ± 7 years, Daily      20
          Winter
          Summer
            12 ± 12 (0-84)
             10 weekend
             12 weekday
               (ND-9)
              (ND-12)
               (2-44)
 13.6 ±2.5 (-to 80)
 10.5 ±2.5 (-to 50)

34.3 ± 17.6 (6.5-88)
 15.4 ±7.7 (6.5-40)
 44.1 ± 18.2(11-88)
           6.2 ±4.7 (0.5-41)
           19 ±18 (0.5-63)
           5.7 ±4.2 (0.5-31)
           25 ± 24 (0.5-72)
           3.5 ± 7.5 (ND-49)
           15 ± 18 (ND-76)
          3.5±3.0(ND-9.9)
           0. ± 1.8 (ND-2.8)
                       Delfinoetal. (1996)
                       Brauer and Brook
                       (1997)
                                 Liu etal. (1997)
                                 Bernard etal. (1999)
                                 Geyh et al. (2000)
                                 Chang et al. (2000)
                                 Sarnat et al. (2000)
        aND = not detected.
        bMeasurements made following scripted activities for 15 days.
1      concentrations.  The overall correlation between the personal exposure concentrations between

2      any two individuals and with the outdoor stationary site was only moderate (r = 0.45; range:

3      0.36 to 0.69). The O3 concentrations at the stationary site exceeded the personal levels by an

4      average of 31 ppb. Avol et al. (1998b) observed a poor correlation between personal exposure

5      and fixed-site monitoring concentrations (r = 0.28, n = 1336 pairs) for a cohort of children

6      (healthy, wheezy, and asthmatic). Personal exposure measurements were generally lower than

7      integrated hourly data.  Sarnat et al. (2000) measured personal O3 exposures during a 12-day

8      longitudinal study of 20 older adults (>64 years) in Baltimore, MD.  The subjects spent >94% of
       August 2005
AX3-186      DRAFT-DO NOT QUOTE OR CITE

-------
 1      the time indoors.  Ten subjects participated in the summer and winter exposures and the
 2      remaining 10 participated in either the summer or winter exposure. No statistically significant
 3      overall correlations were identified between the personal and the ambient O3 concentrations
 4      during either the winter or summer. Only a single individual (n = 14 summer and 13 winter) had
 5      a significant correlation with outdoor concentrations. Geyh et al. (2000) measured indoor and
 6      outdoor concentration and personal O3 exposures in 169 elementary school children from 116
 7      homes during a year-long sampling protocol in 2 communities in southern California (Upland
 8      and Mountain communities).  Samples were collected for 1 week per month. Boys had higher
 9      O3 exposure than girls, probably reflecting the greater amount of time boys spent outdoors
10      compared to girls (3.8 versus 3.2 h for the spring/summer and 2.9 versus 2.2 h for the
11      fall/winter). The average personal O3 exposures were lower than the levels measured at the
12      nearest monitor stations retrieved from the AIRS. There was no significant difference in the O3
13      exposure for both groups during the non-O3 season (6.2 and 5.7 ppb for Upland and the
14      mountain communities, respectively), however, children in the mountainous region were
15      exposed to 35% more O3 than children in Upland during the O3 season (18.8 and 25.4 ppb for
16      Upland and the mountain communities)  (two-tailed t-test,p < 0.01). During the O3 season,
17      differences were found in indoor concentrations and personal O3 exposures between the two
18      communities participating in the study based on ambient air concentrations and differences in air
19      exchange rates in the homes.
20           Brauer and Brook (1997) conducted personal exposure evaluations in three groups in
21      Frazer Valley, Vancouver, Canada. The groups were divided by the amount of time spent
22      outdoors:  (1) the majority of the workday was spent indoors or commuting (25 medical clinic
23      workers), (2) an intermediate amount of time was spent outdoors (25 overnight camp staff
24      members), and (3) the entire exposure monitoring period was outdoors (15 adult farm workers).
25      Time-activity data were collected for the first two groups to assess the proportion of time spent
26      outdoors.  For groups 1 and 2, the lowest quartile of participants based on the fraction of time
27      spent outdoors (0 to 25% and 7.5 to 45%, respectively) had significantly lower O3 exposure
28      (mean personal exposure to outdoor concentration ratio = 0.18 and 0.35, respectively) compared
29      to those in the upper quartile (mean ratio =  0.51 and 0.58, respectively; p < 0.05; Bonferroni
30      multiple range test).  The mean ratio was 0.96 with values ranging from near 0 to 2 for group 3,
31      the group that spent the entire exposure-monitoring period  outdoors. The authors attributed the

        August 2005                             AX3-187      DRAFT-DO NOT QUOTE OR CITE

-------
 1      extreme low ratios to random measurement error at low O3 air concentrations (estimated at
 2      35%), local variability in O3 concentrations, and to differences between ground-level
 3      concentrations (where the personal samples were collected) 3-m above ground level (where the
 4      continuous monitors were located). The highest ratios may be due to either locale variability in
 5      O3 concentrations or to an interference affecting the personal  O3 samplers, particularly at the
 6      lower concentration range, leading to a small positive error. Temporal plots of O3 for the mean
 7      daily personal exposures and ambient concentrations showed the same general trend with
 8      general agreement between the personal exposures and ambient air concentrations for group 3.
 9      However, for groups 1 and 2, the day-to-day variability of the personal exposures and ambient
10      O3 concentrations had consistent patterns, suggesting that the ambient air was the primary source
11      for O3 exposure. The day-to-day variability in personal and continuous measurements was 0.60,
12      0.42, and 0.64 for groups 1 through 3, respectively. The actual O3 concentrations measured in
13      the personal air space were always considerably lower than the ambient concentrations.  Bernard
14      et al. (1999) assessed O3 personal exposure and  in the home and outdoor O concentration for up
15      to 110 subjects.  Measurements were conducted over 5-day periods between June 1995 and
16      October 1996. As anticipated,  O3 concentrations were higher during the warmer months. Mean
17      O3 concentrations for 70 subjects were 22, 35, 17.4, 40.5, and 18 ppb for personal, outdoor
18      home, indoor home, outdoor work, and indoor work, respectively for measurements made during
19      the warmer months.
20           In a study by Liard et al. (1999), 55 mild to moderate asthmatic adults (18 to 65 y old) and
21      39 children (7 to 15 y old) were monitored for O3 exposure. Subjects were monitored during
22      three periods, 4 days per monitoring period and  asked to keep a diary of time spent outdoors and
23      in a car.  Many of the study subjects O3 exposures were often below the level of detection for the
24      method used. Ozone exposure levels correlated with the hours spent outdoors.  Ozone personal
25      exposure correlated poorly with the ambient monitoring measurements, however, the mean
26      values for all subjects correlated with those measurements from the ambient monitoring site
27      (r = 0.83, p  < 0.05). Linn et al. (1996) estimated short-term O3 exposures in 269 children from
28      three communities in the Los Angeles Basin by  monitoring air at head level in school class
29      rooms, on the roof of one level school buildings, and in personal microenvironments of selected
30      individuals. Monitoring was carried out for six  weeks in the fall, winter, and spring, two
31      successive weeks per season at each of three schools. Each subject was monitored for one week

        August 2005                            AX3-188      DRAFT-DO NOT QUOTE OR CITE

-------
 1      in each season over a two year period. According to the authors there were meaningful
 2      associations between personal exposures and central monitoring site O3 measurements (r = 0.61).
 3      Based on information reported in the questionnaires, outdoor activity increased slightly in
 4      communities/seasons with higher pollution.
 5           Lee et al. (2004) found that personal O3 exposure was positively correlated with time spent
 6      outdoors (r = 0.19, p < 0.01) and negatively correlated with time spent indoors (r = -0.17,
 7      p <0.01) in elementary school children. Thirty-three elementary school children from two
 8      Nashville, TN area school districts participated in a six week long O3 monitoring study during
 9      the summer vacation. The study participants maintained a dairy of daily activities during the
10      study period.  An additional 62 children from the same school completed a telephone interview
11      on time/activity at least eight times during the study period.  Study participants wore a passive
12      sampler during their non-sleep time and the sampler was placed near their bed at night.
13      A passive monitor also was placed outside and inside of the home. Personal exposure correlated
14      with the amount of time spent outdoors. Exposures ranged from 0.0013 to 0.0064 ppm for
15      indoor concentrations compared to ambient concentrations of 0.011 to 0.042 ppm O3.
16
17      AX3.10.7.2   Monitored Ambient Concentrations
18           Ozone has been measured more extensively than the other photochemical oxidants.
19      Ambient monitors have been  established in most areas of the country, with extensive monitoring
20      in regions that have been in noncompliance with the previous 1-h daily NAAQS. Monitoring
21      station-measured hourly concentrations also have been used as surrogates of exposure in
22      epidemiological studies and in evaluating exposure-related health  effects. According to the
23      Guideline on Ozone Monitoring Site Selection (U.S. Environmental  Protection Agency, 1998),
24      when designing an O3 monitoring network, consideration should be given to (1) proximity to
25      combustion emission sources, (2) distance from primary emission sources, and (3) the general
26      wind direction and speed to determine the primary transport pathways of O3 and its precursors.
27      Finally, the 1-h daily maximum and 8-h average O3 concentrations can have different spatial
28      patterns with elevated daily 8-h O3 concentrations typically being  over a wider spatial area.
29      Therefore, O3 monitoring networks should determine the highest concentrations expected to
30      occur in the area, representative concentrations for high population density areas, the impact of
        August 2005                            AX3-189     DRAFT-DO NOT QUOTE OR CITE

-------
 1      sources or source categories on air pollution levels, and general background concentration levels
 2      (U.S. Environmental Protection Agency, 1998).
 3           The guideline also states that the monitor's O3 inlet probe should be placed at a height and
 4      location that best approximates where people are usually found.  However, complicating factors
 5      (e.g., security considerations,  availability) sometimes result in the probe placement being
 6      elevated 3 to 15 m above ground level, a different location than the breathing zone (1 to 2 m) of
 7      the populace. Although there are some commonalities in the considerations for the sampling
 8      design for monitoring and for determining population exposures, differences also exist.  These
 9      differences between the location and height of the monitor compared to the locations and
10      breathing zone  heights of people can result in different O3 concentrations between what is
11      measured at the monitor and exposure and, therefore, should be considered when using ambient
12      air monitoring data as a surrogate for exposure in epidemiological  studies and risk assessments.
13      Further, since most people spend the majority of their time indoors, where O3 levels tend to be
14      much lower than outdoor ambient levels, the use of ambient monitoring data to determine
15      exposure generally overestimates true personal O3 exposure, resulting in exposure estimates
16      biased towards  the null. Information on monitored ambient concentrations of O3 and other
17      photochemical  oxidants appears earlier in this chapter.
18
19      AX3.10.7.3   Ozone Concentrations in Microenvironments
20           The 1996 O3 AQCD for Ozone (U.S. Environmental Protection Agency, 1996a) reported
21      O3 I/O ratios for a variety of indoor environments including homes, office/laboratories, a
22      hospital, museums, a school room, and automobiles and other vehicles. Ozone I/O  ratios ranged
23      from 0.09 to 1.0 for residences, 0.19 to 0.8 for offices/laboratories, hospital and school rooms,
24      and 0.03 to 0.87 for museums and art galleries.  The higher O3 ratios were generally noted in
25      indoor environments with high AERs or  100% outside air intake. Studies published since
26      completion of the 1996 O3 AQCD are discussed in this section. The findings of the more recent
27      studies on O3 I/O ratios are included in Table AX3-21.
28           Northeast States for Coordinated Air Use Management (NESCAUM, 2002) monitored
29      levels of O3 inside and outside of nine schools located in the New England states. The schools
30      represented a variety of environmental conditions in terms of ambient O3 concentration, sources,
31      geographic location, population density, traffic patterns, and building types.  Schools were

        August 2005                            AX3-190      DRAFT-DO NOT QUOTE OR CITE

-------
                                                     Table AX3-21.  Indoor/Outdoor Ozone Ratios
S-
to
o
         Location and Ventilation
         Conditions
                                    Indoor/Outdoor Ratio
                                    Mean (range)
                          Comments
                                                                   Reference
>
X
O
O
2
O
H
O
c
o
H
W
O
V
O
HH
H
W
         Toronto, Canada, Homes
             Winter - weekly (68)
             Summer - weekly (38)
             Summer-12 h/day (128)
                    - 12 h/night (36)
Boston, MA, Homes (26)
    Winter - continuously 24 h
    Summer - continuously 24 h
Mexico City, School
    Windows/Doors Open (27)
    Windows/Doors Closed,
    cleaner off (41)
    Windows/Doors Closed,
    cleaner on (47)

Los Angeles, CA
    Homes (95)
    Other locations (57)
                                    0.07 ±0.10 (ND- 0.63)
                                    0.40 ± 0.29 (ND-1.15)
                                    0.30 ± 0.32 (ND - 1.42)
                                    0.43 ± 0.54 (ND - 2.89)
0.30 ± 0.42 (ND-1.31)
0.22 ± 0.25 (ND - 0.88)
                                             0.73 ± 0.04
                                             0.17 ±0.02
                                             0.15 ±0.02
0.28
0.18
Electrostatic air cleaners were present in about 50% of the homes.         Liu et al.
Air conditioners were present in about 80% of the homes, most were       (1995)
central units that used recycled air. Air conditioners used in only 13 of
the 40 homes on a daily basis. Measurements were made both inside
and outside of the homes for 5 consecutive 24-h periods. Homes with
electrostatic air cleaners had higher I/O ratios during the winter months.
The mean average weekly AER for all homes during the winter months
was 0.69 ± 0.88 h"1 with 50% of the homes with an AER of less than
0.41 h~'.  For the summer months, the mean average AER was 1.04 ±
1.28 h"1  with 50% of the homes with an AER of less than 0.52 h.

Study examined the potential for O3 to react with VOCs to form acid       Reiss et al.
aerosols. Carbonyls were formed. No clear trend of O3 with AERs.       (1995)
The average AER was 0.9 h"1 during the winter and 2.6 h"1 during the
summer.  Four residences in winter and nine in summer with 24 h
average concentrations.  Air concentrations varied from 0-34.2 ppb
indoors and 4.4-40.5 ppb outdoors.

Study conducted over 4-day period during winter months. Two-min       Gold et al.
averaged measurements were taken both inside and outside of the         (1996)
school every 30 min from 10 a.m. to 4 p.m. Estimated air exchange rates
were 1.1, 2.1, and 2.5 h ' for low, medium, and high flow rates.  Ozone
concentrations decreased with increasing relative humidity.
Study conducted in September. Monitored O3 concentrations consisted     Johnson
of twenty-one 24-h periods beginning at 7 p.m. and ending at 7 p.m. on     (1997)
the following day. Ozone concentrations were higher at the fixed
monitoring sites during the afternoon. The weather was sunny and the
temperature was high. I/O ratio was lower when windows were closed.
The effect of air conditioning on the I/O was varied.

-------
Table AX3-21 (cont'd). Indoor/Outdoor Ozone Ratios
u£
S-
to
o
o
u\











^
X
L>j
VO
to



h-j

o

0
H
O
O
H
W
O
O
HH
H
W

Location and Ventilation
Conditions

Mexico City
Homes (237)
Schools (59)






Los Angeles, Homes (239)
Summer
Winter




California
Homes
no AC, window opened (20)
AC, windows closed (3)
Munich Germany
Office
Gymnasium
Classroom
Residence
Bedroom
Livingroom
Hotel room
Car

La Rochelle, France




Indoor/Outdoor Ratio
Mean (range)

0.20 ±0.18 (0.04 -0.99)
0.1-0.3
0.3-0.4






0.37 ±0.25 (0.06 -1.5)
0.43 ±0.29
0.32 ±0.21






0.68 (n = 20)
0.09 (n= 3)

0.4-0.9
0.49-0.92
0.54-0.77

0.47 - 1.0
0.74 - 1.0
0.02
0.4-0.6

0-0.45




Comments

Ozone monitoring occurred between September and July. Study
included 3 schools and 145 homes. Most of the homes were large and
did not have air conditioning. Ninety-two percent of the homes had
carpeting, 13% used air filters, and 84% used humidifiers. Thirty -five
percent opened windows frequently, 43% sometimes, and 22% never
between 10 a.m. and 4 p.m. Ozone was monitored at the schools sites
from 8 a.m. to 1 p.m. daily for 14 consecutive days. Homes
were monitored for continuous 24-h periods for 14 consecutive days.
I/O based on 1-h average concentrations.
Four hundred and eighty -one samples collected inside and immediately
outside of home from February to December. Ratios based on 24-h
average O3 concentrations indoors and outdoors. Low outdoor
concentrations resulted in low indoor concentrations. However, high
outdoor concentrations resulted in a range of indoor concentrations
and ratios. I/O ratios were highest during the summer months.

I/O ratio was determined for 20 homes. Only 3 of the homes operated
the air conditioning. I/O ratios based on 24-h continuous ambient
concentrations and 0.5-1 h average indoor concentrations.

Indoor concentrations were dependent on the type of ventilation.









I/O ratio determined for 8 schools. Monitoring conducted for a 2-wk
period. Schools located in various areas with different ambient O3
concentrations, types of ventilation systems, and the presumed
air-tightness of building envelop. Building air-tightness and ambient O3
concentration influenced indoor O3 I/O.
Reference

Romieu
etal. (1998)







Avol et al.
(1998a)





Lee et al.
(1999)


Jakobi and
Fabian
(1997)







Blondeau
et al. (2005)




-------
1
OJ
to
o










Location and Ventilation
Conditions
Montpellier, France, Homes (110)


Southern CA, Homes
Upland Mountains




Table AX3-21 (cont
Indoor/Outdoor Ratio
Mean (range)
0.41


0.68 ±0.18
(windows open)
0.07-0.11
(AC used)


'd). Indoor/Outdoor Ozone Ratios
Comments
Ozone measurements were made over 5 -day periods in and outside of 21
homes during the summer and winter months. The winter I/O ratio was
0.3 1 compared to 0.46 during the summer months.
Ozone measurements were taken at 119 homes (57 in Upland and 62 in
towns located in the mountains) during April and May. I/O ratios were
based on average monthly outdoor concentrations and average weekly
indoor concentrations. I/O ratio was associated with the home location,
number of bedrooms, and the presence of an air conditioner. I/O ratios
based on subset of the homes.

Reference
Bernard
etal. (1999)

Geyh et al.
(2000);
Lee et al.
(2002)


>
X
VO
O
o
o
H
O
o
H
W
         Krakow, Poland
             Museums (5)
                                              0.13 -0.42
         Buildings, Greece
             Thessaloniki
             Athens
                                              0.24 ±0.18 (0.01 to 0.75)
         Southern California, Museum
                                              0 . 1 9 ± 0 . 05
Ozone continuously monitored at five museums and cultural centers.
Monitoring conducted during the summer months for 21-46 h or 28-33
days at each of the sites. The I/O was found to be dependent on the
ventilation rate, i.e., when the ventilation rate was high, the I/O ratio
approached unity, while in rooms sequestered from the outdoor air or
where air was predominantly recycled through charcoal filters, the O3
levels indoors were greatly reduced resulting in low I/O ratios.

There was no heating/air conditioning system in the building at
Thessaloniki. Windows were kept closed during the entire monitoring
period.  Complete air exchange took place every 3 h. I/O ratio ranged
from 0.5-0.90, due to low deposition velocity on indoor surfaces.  The air
conditioning system in continuous use at the Athens site recirculated
the air.  Complete air exchange was estimated to be  1 h.  Monitoring
lasted for 30 days at each site, but only the 7 most representative days
were used for calculation of the I/O ratio.

Measurements made over a 2-wk period (24-h avg). Ratio  for
concentrations at the buffer zone with the roll-up door closed.
                                                                                                                                              Salmon et al.
                                                                                                                                              (2000)
                                                                                                                                              Drakou et al.
                                                                                                                                              (1998)
Hisham and
Grosjean
(1991)
O
HH
H
W
         ND = not detectable.

-------
 1     monitored during the summer months to establish baseline O3 concentrations and again during
 2     the fall after classes started. A monitor was placed directly outside of the school entrance and
 3     50 feet away from the entrance in the hall.  Where available, monitors were placed at locations
 4     identified as "problem" classrooms, classrooms with carpeting, or in rooms close to outdoor
 5     sources of O3.  As expected, outdoor concentrations of O3 were higher than those found indoors.
 6     Averaged O3 concentrations were low during the early morning hours (7:30 a.m.) but peaked to
 7     approximately 25 and 40 ppb around 1:30 p.m. indoors and outdoors, respectively.
 8          Gold et al. (1996) compared indoor and outdoor O3 concentrations in classrooms in Mexico
 9     City under three different ventilation conditions: windows/doors open,  air cleaner off;
10     windows/doors closed, air cleaner off; and windows/doors closed, air cleaner on.  Two-minute
11     averaged outdoor O3 levels varied between 64 and 361 ppb, while indoor O3 concentrations
12     ranged from 0  to 247 ppb.  The highest  indoor O3 concentrations were noted when the
13     windows/doors were open. The AERs were estimated to be 1.1,  2.1, and 2.5 IT1 for low,
14     medium, and high air flow rates, respectively. The authors indicated that the indoor levels, and
15     therefore O3 exposure to students in schools, can be reduced to < 80 ppb by closing windows and
16     doors even when ambient O3 levels reach 300 ppb.
17          In a second Mexico City study, Romieu et al. (1998) measured O3 concentrations inside
18     and outside of 145 homes and three schools.  Measurements were made between November and
19     June. Most of the homes were large and did not have air conditioning.  Ninety-five percent of
20     the homes had carpeting, 13% used air filters, and 84% used humidifiers.  Thirty-five percent of
21     the homeowners reported that they opened windows frequently between the hours of 10 a.m.  and
22     4 p.m., while 43% opened windows sometimes and 22% reported that they never opened
23     windows during that time period. Homes were monitored for continuous 24-h periods for
24     14 consecutive days. Schools were monitored from 8 a.m. to 1 p.m. or continuously for 24 h.
25     During the school monitoring periods the windows were frequently left open and the doors were
26     constantly being opened and closed. The mean indoor and outdoor O3 concentrations during  5-h
27     measurements  at the schools were 22 ppb and 56 to 73 ppb, respectively.  The mean indoor and
28     outdoor O3 concentrations for measurements made over a 7- and 14-day period at the test homes
29     were 5 and 27  ppb and 7 and 37 ppb, respectively. Ozone concentrations inside of homes were
30     dependent on the presence  of carpeting, the use of air filters, and whether the windows were
31     open or closed. Air exchange rates were not reported.

       August 2005                           AX3-194     DRAFT-DO NOT QUOTE OR CITE

-------
 1           Reiss et al. (1995) compared indoor and outdoor O3 concentrations for residences in the
 2     Boston, MA area. Four residences were monitored during the winter months and nine residences
 3     during the summer months. Outside monitors were placed ~1 m away from the house.
 4     Monitoring was conducted over a continuous 24-h period. There were no indoor sources of O3.
 5     Indoor O3 concentrations were higher during the summer months, with concentrations reaching
 6     34.2 ppb. Indoor O3 concentrations reached as high as 3.3 ppb during the winter monitoring
 7     period.  In one instance, O3 concentrations were higher indoors than outdoors. The authors
 8     attributed that finding to analytical difficulties. Outdoor O3 concentrations were generally higher
 9     during the summer monitoring period, with concentrations reaching 51.8 ppb. Indoor
10     concentrations were dependent on the outdoor O3 concentrations and AER. Indoor and
11     outdoor O3 concentrations, including the AERs at the times of the monitoring are included in
12     Table AX3-22.
13           Avol et al. (1998a) monitored 126 home in the Los Angeles metropolitan area during
14     February and December. Uniformly low ambient O3 concentrations were present during the
15     non-O3 seasons. Indoor O3 concentrations were always below outdoor O3 concentrations. The
16     mean indoor and outdoor O3 concentrations over the sampling period were 13 ± 12 ppb and 37 ±
17     19 ppb, respectively. There was a correlation between indoor O3 concentration and ambient
18     temperatures. The effect was  magnified when the windows were open. When a central
19     refrigerant recirculating air conditioner was used, indoor O3 concentrations declined.  The
20     authors were able to predict indoor O3 levels with nearly the same accuracy using measurements
21     made at regional stations coupled with window conditions as with measurements made directly
22     outside the homes coupled with window conditions, suggesting that monitoring station data may
23     be useful in helping to reduce errors associated with exposure misclassification.  The authors
24     cautioned that varying results  may occur at different locations with different housing stock or at
25     different times of the year.
26           Lee et al.  (2002) reported indoor and outdoor O3 concentrations in 119 homes of school
27     children in two communities in southern California: Upland and San Bernadino county.
28     Measurements were made over one year.  Outdoor and indoor O3 concentrations were based on
29     cautioned that varying results  may occur at different locations with different housing stock or at
30     different times of the year.
31

       August 2005                             AX3-195     DRAFT-DO NOT QUOTE OR CITE

-------
          Table AX3-22. Indoor and Outdoor O3 Concentrations in Boston, MA
Indoor Data
Residence
Indoor
Ozone
(Ppb)
Outdoor
Ozone
(Ppb)
AER
(IT1)
Relative
Humidity
(%)
Temp. (°F)
Outdoor Data
Relative
Humidity
(%)
Temp. (°F)
Winter
I
I
2
2
3
3
4
4
-Day 1
-Day 2
-Day 1
-Day 2
-Day 1
-Day 2
-Day 1
-Day 2
3.3
0
2.6
1.7
20.4
3.1
2.2
0.3
11.2
15
24.4
4.4
15.6
24.5
11.4
20.7
1
0.8
1
1
1
0.9
0.7
0.7
25-45
22-40
3-19
3-8
8-24
13-19
26-37
29-38
67-75
65-76
67-71
55-70
62-69
64-70
60-72
61-70
88
62
44
40
33
51
57
77
25
27
17
17
36
38
39
36
Summer
1
1
2
2
3
3
4
4
5
5
6
6
7
7
8
8
9
9
-Day 1
-Day 2
-Day 1
-Day 2
-Day 1
-Day 2
-Day 1
-Day 2
-Day 1
-Day 2
-Day 1
-Day 2
-Day 1
-Day 2
-Day 1
-Day 2
-Day 1
-Day 2
5.6
0.6
0.8
5
34.2
6.9
4.3
4.9
1.4
1.9
0.8
1.7
3.9
0
22.9
23.5
1.6
1.2
32.4
13.4
14.3
24.1
38.9
14
30
40.5
17.5
19
8.2
18.6
40.1
33.9
51.8
31.6
20.9
25
3
2.3
2.4
2.1
4.6
3.1
1.4
1.8
5.1
3.5
0.5
0.7
1.1
1.1
3.2
6.3
2.1
1.7
28-44
37-44
48-54
46-60
48-63
45-53
37-60
38-68
30-50
39-63
59-73
43-66
57-70
58-73
66-81
43-67
N/A
33-52
71-74
70-74
73-79
72-78
64-80
65-69
66-75
67-79
69-74
N/A
74-77
76-78
70-77
72-75
71-77
66-79
N/A
75-79
44
59
54
64
52
52
51
64
54
40
76
47
51
64
75
48
37
70
65
60
70
73
71
62
67
67
58
64
72
76
75
72
75
72
70
66
 N/A = not available
 Adapted from: Reiss et al. (1995)
August 2005
AX3 -196     DRAFT-DO NOT QUOTE OR CITE

-------
 1           Lee et al. (2002) reported indoor and outdoor O3 concentrations in 119 homes of school
 2      children in two communities in southern California: Upland and San Bernadino county.
 3      Measurements were made over one year. Outdoor and indoor O3 concentrations were based on
 4      monthly and weekly averages, respectively.  Housing characteristics were not found to affect
 5      indoor O3 concentrations. Indoor O3 concentrations were significantly lower than outdoor O3
 6      concentrations. Average indoor and outdoor O3 concentrations were 14.9 and 56.5 ppb. Homes
 7      with air conditioning had lower O3 concentrations, suggesting decreased ventilation or greater
 8      O3 removal.
 9           Chao (2001) evaluated the relationship between indoor and outdoor levels of various air
10      pollutants, including O3, in 10 apartments in Hong Kong during May to June. Air monitoring
11      was conducted over a 48-h period.  All participants had the habit of closing the windows during
12      the evenings and using the air conditioner during the sleep hours.  Windows were partially open
13      during the morning. The air conditioners were off during the working hours. Indoor O3
14      concentrations were low in all of the monitored apartments, ranging from 0 to 4.9 ppb with an
15      average of 2.65 ppb.  Outdoor O3 concentrations ranged from 1.96 to 15.68 ppb. Table AX3-23
16      provides information on the indoor and outdoor O3 concentrations and the apartment
17      characteristics.
18           Drakou et al. (1995) demonstrated the complexity of the indoor environment.
19      Measurements of several pollutants, including O3, were made inside and outside of two
20      nonresidential buildings in Thessaloniki and Athens, Greece.  The building in Thessaloniki was a
21      58-m3 metal structure. The ceiling and walls were covered with colored corrugated plastic
22      sheeting, and the flooring was plastic tile.  There was no heating/air conditioning system and the
23      building was closed during the monitoring period. The  AER ranged from 0.3 to 0.35 h"1. The
24      building in Athens was a 51-m3 concrete structure. The air conditioning system (recirculated air)
25      worked continuously during  the monitoring period. A window was opened slightly to
26      accommodate the monitors'sampling lines.  The AER was approximately  1 h"1. Monitoring
27      lasted for 30 days at both locations, however, only data  from the 7 most representative days were
28      reported.  Indoor O3 concentrations closely followed the outdoor concentrations at the
29      Thessaloniki building.  The averaged 7-day indoor and outdoor O3 concentrations were 9.39 and
30      15.48 ppb, respectively. The indoor O3 concentrations at the Athens location were highly
31      variable compared to the outdoor concentration.  The authors suggested that a high hydrocarbon

        August 2005                            AX3-197     DRAFT-DO NOT QUOTE OR CITE

-------
                 Table AX3-23. Indoor and Outdoor O3 Concentrations in Hong Kong
Apartment
No.
1
2
o
5
4
5
6
7
8
9
10
Floor
Area
40
47
140
67
86
43
47
30
26
20
Floor
No.
14
13
2
5
11
32
9
6
35
15
Window Opening
Frequency
Seldom
Sometimes
Sometimes
Seldom
Sometimes
Sometimes
Always
Seldom
Sometimes
Seldom
AER (IT1)
1.44
1.97
0.83
5.27
1.64
15.83
15.91
3.25
2.1
5.5
Indoor
O3 Cone.
0
4.96*
1.0*
4.96*
3.0*
4.01*
0
2.05*
4.9
4.01*
Outdoor
O3 Cone.
1.96
6.01*
6.96*
8.76*
7.80*
8.76*
3.0*
3.0*
15.68
4.96*
         *Estimated Concentration.
         Adapted from Chao (2001).
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
intrusion may be the reason for the variability in O3 concentrations noted at this site. The
averaged 7-day indoor and outdoor O3 concentrations were 8.14 and 21.66 ppb, respectively.
     Weschler et al. (1994) reported that indoor O3 concentrations closely tracked outdoor O3
concentrations at a telephone switching station in Burbank, CA. The switching building was
flat-roofed, two-story (first floor and basement), uncarpeted, with unpainted brick walls. Each
floor was 930 m2 with a volume of 5095 m3.  Indoor O3 concentrations were measured on the
first floor using a perfluorocarbon tracer or an UV photometric analyzer. The AER were
obtained by dividing the volumetric flow to the first floor by the volume of the first floor space
or by perfluorocarbon tracer techniques. The AER ranged from 1.0 to 1.9 h"1.
     The major source of O3 at the switching station was transport from outdoors. Indoor O3
concentrations closely tracked outdoor concentrations, measuring from 30 to 70% of the outdoor
concentrations. Indoor O3 concentrations frequently exceed 50 ppb during the summer months,
but seldom exceeded 25 ppb during the winter.  During the early spring  through late fall, indoor
O3 concentrations peaked during the early afternoon and approach zero after sunset.  Ozone sinks
       August 2005
                                       AX3-198      DRAFT-DO NOT QUOTE OR CITE

-------
1
2
3
4
included a surface removal rate between 0.8 and 1.0 IT1 and reactions with NOX. Figures
AX3-89 and AX3-90 compare the outdoor and indoor O3 concentrations, including the AER, for
two 1-week periods during the study.
                                              July, 1992
                     I
                     o
                      0100
                      o
      Figure AX3-89.  Air exchange rates and outdoor and indoor O3 concentrations during the
                      summer at telephone switching station in Burbank, CA.
      Source:  Weschleret al. (1994).
           The relationships between indoor and outdoor O3 concentrations in five museums and
      cultural centers (Wawel Castle, Matejko Museum, National Museum, Collegium Maius, and
      Cloth Hall) in Krakow, Poland were reported by Salmon et al. (2000). Measurements were made
      for up to 46 h and up to 33 days. Air exchange measurements were only made for the Matejko
      Museum and Wawel Castle. Both were naturally ventilated. However, the summertime AER
      for the Matejko Museum was more than twice that of the Wawel Castle site, 1.26 to 1.44 h"1
      compared to 0.56 to 0.66 h"1.  The highest indoor O3 concentrations were noted at the Matejko
      Museum during the summer. The findings are included in Table AX3-24 for those locations
      with reported AERs.
      August 2005
                                      AX3 -199     DRAFT-DO NOT QUOTE OR CITE

-------
                               W
                                              September, 1992
                       °100
                       o
       Figure AX3-90.  Air exchange rates and outdoor and indoor O3 concentrations during the
                        fall at a telephone switching station in Burbank, CA.
       Source:  Weschleretal. (1994).
 1          Figure AX3-91 shows O3 and PAN concentrations in a private residence in Germany.  All
 2     measurements were made in naturally ventilated rooms.
 3          Johnson (1997) conducted a scripted study using four trained technicians to measure
 4     hourly average O3 concentrations between 07:00 and 19:00 h in Los Angeles, CA during
 5     September and October 1995. The ratio of the microenvironmental concentrations to the fixed
 6     site monitor on days when the O3 levels > 20 ppb were as follows: indoor residence, 0.28; other
 7     locations indoors, 0.18; outdoor near roadways, 0.58;  other locations outdoors, 0.59; and in-
 8     vehicle, 0.21.  The concentrations indoors and within vehicles varied depending on whether the
 9     windows were opened (higher) or closed (lower) and the use of air conditioning.  The lower
10     outdoor concentrations, particularly near roadways, probably reflect the reaction  of O3 with NO
11     emitted by automobiles.
12          A study of the effect of elevation on O3 concentrations  found that concentrations increased
13     with increasing elevation. The ratio of O3 concentrations at street level (3 m) compared to the
14     rooftop (25 m) was between 0.12 and 0.16, though the actual concentrations were highly
       August 2005
AX3-200
DRAFT-DO NOT QUOTE OR CITE

-------
                       Table AX3-24. Indoor and Outdoor Ozone Concentrations
Location
Matejko Museum
Outdoors (Town Hall Tower)
Indoors (third floor, west)
Wawel Castle
Outdoors (Loggia)
Indoors (Senator's Hall)
Wawel Castle, outdoors
Wawel Castle, Room 15
Wawel Castle, Senator's Hall
Matejko Museum, outdoors
Matejko Museum, Indoor Gallery
Duration
(hours) AER
26

1.26
43

0.63
31.8
31
31.8
26.9
26.9
Average O3 (ppb)
1988-2000
20
8.5

14.7
2.5
42a
8
7
21b
9
        aOn Loggia of Piano Nobile Level, high above the street.
        bAt street level.
        Adapted from Salmon et al. (2000).
 1     correlated (r = 0.63) (Vakeva et al., 1999). Differential O3 exposures may, therefore, exist in
 2     apartments that are on different floors. Differences in elevation between the monitoring sites in
 3     Los Angeles and street level samples may have contributed to the lower levels measured by
 4     Johnson (1997). Furthermore,  since O3 monitors are frequently located on rooftops in urban
 5     settings, the concentrations measured there may overestimate the exposure to individuals
 6     outdoors in streets and parks, locations where people exercise and maximum O3 exposure is
 7     likely to occur.
 8          Chang et al. (2000) conducted a scripted exposure study in Baltimore during the summer of
 9     1998 and winter of 1999, during which 1-h O3 samples were collected by a technician who also
10     changed his or her activity every hour. The activities chosen were selected to simulate older
11     (> 65 years) adults, based on those activities reported in the National Human Activity Pattern
12     Survey (NHAPS). The scripted activities took place in five different microenvironments:
       August 2005                            AX3-201     DRAFT-DO NOT QUOTE OR CITE

-------
                     70 ^
                      0
                      00:00
                                                                              0.0
                                                                            00:00
       Figure AX3-91.   Diurnal variation of indoor and outdoor O3 and PAN concentrations
                         measured in a private residence, Freising, Germany, August 11-12,1995.
       Source: Jaboki and Fabian (1997).
 1     indoor residential (apartment in 24-h air conditioned high rise building), indoor other (restaurant,
 2     post office, hospital, shopping mall, and bingo parlor), outdoor near roadway, outdoor away
 3     from roadway, and inside motor vehicle. Personal O3 exposures were significantly lower in
 4     indoor than the outdoor microenvironments, because more time was spent indoors. Mean
 5     summer concentrations were 15.0 ± 18.3 ppb, with a maximum of 76.3 ppb.  Significant
 6     correlation was noted for the indoor nonresidential microenvironments and ambient O3 (r = 0.34
 7     in summer, r = 0.46 in winter), however, the authors indicated that this finding was unclear due
 8     to the low personal/ambient ratios. The personal O3 exposure levels within the outdoor and in-
 9     vehicle microenvironments were significantly correlated with ambient concentration, although
10     the ratio of personal exposure to ambient levels was less than one, with only the top 5% of the
11     ratios exceeding  one, indicating that the ambient measurements lead to the maximum
12     concentrations and exposures. The indoor concentrations did not correlate with outdoor
13     measurements (r =  0.09 and r = 0.05 for summer and winter, respectively). The correlation for
14     outdoor near roadway and outdoor away from roadway was moderate to high (0.68 
-------
 1          The scripted exposure studies show that the O3 concentrations in the various
 2     microenvironments were typically lower than the ambient air concentrations measured at
 3     monitoring stations. Exposure models are useful for accounting for the reduced concentrations
 4     usually encountered in various microenvironments compared to ambient monitoring station
 5     concentrations (see discussion on exposure models earlier in this annex).
 6          Riediker et al. (2003) measured mobile source pollutants inside highway patrol vehicles in
 7     Wake County, NC.  Measurements were made during the 3 p.m. to midnight shift between
 8     August 13 and October 11, 2001 in two patrol cars each day for a total of 50 shifts. All areas of
 9     rural and urban Wake County were patrolled.  The prominent areas patrolled were major
10     highways and interstates.  Ozone concentrations inside the cars were compared with the O3
11     measurements at the fixed station in northern Raleigh. The average O3 concentration inside the
12     cars was 11.7 ppb, approximately one-third of the ambient air concentration.  Jakobi and Fabian
13     (1997) found that O3 concentrations in a moving car were  independent of the type of ventilation
14     (windows closed and ventilator operation/ventilator off and windows open). Ozone
15     concentrations inside the car were found to closely follow the outdoor concentrations.  When the
16     car was parked, O3 concentrations outdoors greatly exceeded concentrations inside the car (see
17     Figure AX3-92).
                  100 -,
                   90 -
                   80 -
                   70 -
               5 60-
               3 50 -
               o  4°
                   30 -
                   20 -
                   10 -
                   Country Road
Country Road
            City
Parking
                                Autobahn
rioo
-90
-80
-70
  0  °
-50  =5"
-40  ^
-30
-20
-10
                     11:30       11:45      12:00      12:15
                                                Time
                                     12:30      12:45
                                         August 23, 1995
       Figure AX3-92. Indoor and outdoor O3 concentration in moving cars.
       Source: Jaboki and Fabian (1997).
       August 2005
                    AX3 -203      DRAFT-DO NOT QUOTE OR CITE

-------
 1           Few studies have been conducted in indoor environments containing O3 sources. Black
 2      et al. (2000) measured O3 concentrations in a photocopy room at the University of California
 3      during one business day.  The room volume was 40 m3.  Ozone concentrations were generally
 4      below 20 ppb, but increased proportionately with increasing photocopier use. Ozone
 5      concentrations reached 40 ppb when the hourly average number of copies reached 45.  Helaleh
 6      et al. (2002) reported daily average O3 concentrations from 0.8 to 1.3 ppb and 0.9 to 1.0 ppb in a
 7      laboratory and photocopy room at a university in Japan. Outdoor O3 concentrations ranged from
 8      6 to 11 ppb. Because only limited information was available on the sampling system used in the
 9      study, the adequacy of the sampling system cannot be determined. Jakobi and Fabian (1997)
10      measured O3 concentrations in an office associated with the use of a photocopier and a laser
11      printer. They  noted a 3.0 ppb increase in O3 from the use of a 3-year old printer run for 20 min
12      at a copy rate of 8 pages/min.  There was no detectable change in O3 concentrations from the use
13      of the laser printer.
14           The U.S. Environmental Protection Agency (Steiber,  1995) measured O3 concentrations
15      from the use of three home/office O3 generators. The O3 generators were placed in a 27 m3room
16      with doors  and windows closed and the heating, ventilating and air conditioning system off; the
17      AER was 0.3 h"1.  The units were operated for 90 min.  Ozone concentrations at the low output
18      setting ranged from nondetectable to 14 ppb (averaged output). At the high output setting,
19      averaged output O3 concentrations exceeded 200 ppb in several cases and had spike
20      concentrations as  high as 480 ppb.
21           Figure AX3-93 includes PAN indoor/outdoor ratios for 10 museums.  Four of the museums
22      were equipped with HVAC and chemical infiltration systems.
23
24      AX3.7.4  Factors Affecting Ozone Concentrations Indoors
25           In the absence of an indoor source, O3 concentrations in indoor environments will depend
26      on the outdoor concentration, the air exchange rate (AER) or outdoor infiltration, indoor
27      circulation rate, removal by indoor surfaces, reactions with other indoor pollutants, and
28      temperature and humidity. Indoor concentrations generally closely track outdoor O3
29      concentrations. Limited information on PAN concentrations indoors also indicate that indoor
30      PAN concentrations tract outdoor concentrations (Jakobi and Fabian, 1997; Hisham and
31      Grosjean, 1991).  Since outdoor concentrations of photochemical oxidants are generally higher

        August 2005                           AX3-204     DRAFT-DO NOT  QUOTE OR CITE

-------
      Figure AX3-93.
                               1.0-
                               0.6-
                               0.4-
                               0.2-
                               0.0
                                        3456
                                                        9 10 11
                  Indoor/outdoor concentration ratios for PAN at 10 southern California
                  museums. Code to locations: 1 El Pueblo Historical Park, Los Angeles;
                  2 Olivas Adobe House, Ventura; 3 Southwest Museum, Los Angeles;
                  4 Page Museum, Los Angeles; 5 Museum of Natural History,
                  Los Angeles; 6 Research Library, University of California, Los Angeles;
                  7 Scott Gallery, Huntington Museum, Pasadena; 8 The J. Paul Getty
                  Museum, Malibu; 9 Los Angeles County Museum of Art, Los Angeles;
                  10 Gene Autry Western Heritage Museum, Los Angeles (buffer zone);
                  11 Gene Autry Museum (Conservation Room).
      Source: Hisham and Grosjean (1991).
1

2

3

4

5
during the warmer months, indoor concentrations will also be highest during that time period.

(See discussion on ambient concentrations of O3 earlier in this chapter.)


Air Exchange Rates

     Indoor O3 increases with increasing air exchange rate (Gold et al., 1996; Lee et al., 1999;

Jakobi and Fabian, 1997). The AER is the balance of the flow of air in and out of a

microenvironment.  Infiltration through unintentional openings in the building envelope is the

dominant mechanism for residential air exchange. Duct systems account for 30% of the total

leakage area of a house. Natural ventilation, airflow through opened windows and doors, also
      August 2005
                                      AX3-205
DRAFT-DO NOT QUOTE OR CITE

-------
 1      influences air exchange in residential buildings. Forced or mechanical ventilation is the
 2      dominant mechanism for air exchange in nonresidential buildings.
 3           Air exchange rates vary depending on the temperature differences, wind effects,
 4      geographical region, type of heating/mechanical ventilation system, and building type (U.S.
 5      Environmental Protection Agency, 1997; Weschler and Shields, 2000; Colome et al., 1994;
 6      Johnson et al., 2004). Air exchange rates are generally higher during the summer and lower
 7      during the winter months (Wilson et al., 1996;  Murray and Burmaster, 1995; Colome et al.,
 8      1994; Research Triangle Institute, 1990). The  Gas Research Institute, Pacific Gas and Electric
 9      Company, San Diego Gas and Electric Company, and Southern California Gas Company
10      measured the air exchange rates in a subset of 293 randomly selected homes in California as part
11      of an air pollution monitoring study. The average AER varied by type of heating system (wall
12      furnaces >forced-air > electric) and building type (multifamily units > single-family units)
13      (Billick et al., 1984,  1996; Colome et al., 1994).
14           Howard-Reed et al. (2002) determined that opening a window or exterior door causes the
15      greatest increase in AERs with differences between the indoor and outdoor temperature being
16      important when the windows were closed. Johnson and Long (2004) conducted a pilot study to
17      evaluate the frequency that windows were left  open in a community. They found that a visual
18      2-h survey could be used to estimate the frequency that windows are left open. The occupancy,
19      season, housing density, absence of air conditioning, and wind speed were factors in whether the
20      windows were open.
21           Johnson et al. (2004) conducted a study using scripted ventilation conditions to identify
22      those factors that affected air exchange inside a test house in Columbus, OH.  The test house was
23      a wood-framed, split-level structure with aluminum siding  covering the wood outer walls.  The
24      house had one exterior door located in the front and another at the rear of the house, single-pane
25      glazed windows, central gas heat, a window  air-conditioning unit, and ceiling fans in three
26      rooms.  Eighteen scenarios with unique air flow conditions were evaluated to determine the
27      effect on the AER. The elements of the scenarios included: exteriors doors open/closed, interior
28      doors open/closed, heater on/off, air conditioner on/off, a ceiling fans on/off.  The lower level
29      was sealed off during the study. The various scenarios were evaluated during the winter season.
30      The average AER for all scenarios ranged from 0.36 to 15.8 h"1.  When all windows and doors
31      were closed, the AER ranged from 0.36 to 2.29 h"1 (0.77 h"1 geometric mean).  When at least one

        August 2005                           AX3-206      DRAFT-DO NOT QUOTE OR CITE

-------
 1      exterior door or window was open the AER ranged from 0.5 to 15.8 IT1 (1.98 IT1 geometric
 2      mean).
 3          Williams et al. (2003a, 2003b) reported air exchange rates ranging from 0.001 IT1 to
 4      4.87 tr1 (overall arithmetic mean of 0.72 IT1) in houses in the Research Triangle Park area in
 5      North Carolina.  Air exchange measurements were made in 37 homes as part of a year long study
 6      PM panel study.
 7          Air exchange rates for homes in Houston, TX, Los Angeles County, CA, and Elizabeth, NJ
 8      were reported by Meng et al. (2005e) as part of the Relationship of Indoor, Outdoor and Personal
 9      Air (RIOPA) study. The RIOPA study was designed to determine indoor (residual), outdoor,
10      and personal exposure to several classes of pollutants.  Approximately 100 homes from each of
11      the areas were sampled across all four seasons.  The mean air exchange rate for the Los Angeles
12      County homes was 1.22 h'1, similar to the air exchange rate (1.51 h'1 ) homes in Los Angeles
13      previously reported by Wilson et al. (1996). The mean air exchange rate for Houston and
14      Elizabeth was 0.71 h'1 and 1.22 h'1, respectively.  Air exchange rates for New Jersey were
15      higher than other reported values in the northeast region.  The authors attributed these
16      differences to differences in the age of housing stock in the various areas.
17          Chan et al. (2005) compared air leakage measurements for more than 70,000 houses across
18      the United States, classified as low-income households, energy program houses, and convention
19      houses, to the building size, construction date and various construction characteristics, and
20      geographical location. The construction date and building size were the two most significant
21      predictors of leakage areas. Older and smaller houses had higher normalized leakage areas than
22      the newer and larger houses.  Based on their evaluation of new and older residential
23      constructions, Sherman and Matson (1997) found that existing home stock (pre-1980) was quite
24      leaky with an AER of approximately 20.0 h"1. Newer constructions were considerably more
25      airtight.
26          Murray and Burmaster (1995) conducted an analysis of data compiled by Brookhaven
27      National Laboratories on AERs for 2,844 residential structures in four climatic regions based on
28      heating degree days.  (Region 1: IN, MN, MT, NH, NY1, VT, WI; Region 2: CO, CT, IL, NJ,
29      NY2, OH, PA, WA; Region 3:  CA3, MD, OR, WA; Region 4: AZ, CA4, FL, TX). Data were
30      also separated by seasons (winter, spring, summer, and fall).  The highest overall AERs occurred
31      during the spring and fall season. However, air exchange rates were variable within and between

        August 2005                            AX3-207      DRAFT-DO NOT QUOTE OR CITE

-------
 1      seasons and between regions. Data from the warmest region during the summer months should
 2      be reviewed with caution because many of the measurements were made in southern California
 3      where windows were more likely to be open than in other areas of the country where air-
 4      conditioning is used.
 5           Air exchange rates for 49 nonresidential buildings (14 schools, 22 offices, and 13 retail
 6      establishments) in California were reported by Lagus Applied Technology, Inc. (1995). Average
 7      mean (median) AERs were 2.45 (2.24), 1.35 (1.09), and 2.22 (1.79) IT1 for schools, offices, and
 8      retail establishments, respectively.  Air infiltration rates for 40 of the 49 buildings were 0.32,
 9      0.31, and 1.12 IT1 for schools, offices, and retail establishments, respectively. Air exchange
10      rates for 40 nonresidential buildings in Oregon and Washington (Turk et al., 1989) averaged
11      1.5(1.3) h'1 (mean median).  The geometric mean of the AERs for six garages was 1.6 h'1 (Marr
12      et al., 1998). Park et al. (1998) reported AERs for three stationary cars (cars varied by age)
13      under different ventilation conditions. Air exchange rates ranged from 1.0 to 3.0 h'1 for
14      windows closed and fan off,  13.3 to 23.5 h'1 for windows opened and fan off, 1.8 to 3.7 h'1 for
15      windows closed and fan on recirculation (two cars tested), and 36.2 to 47.5 h'1 for windows
16      closed and fan on fresh air (one car tested). An average AER of 13.1 h'1 was reported by Ott
17      et al. (1992) for a station wagon moving at 20 mph with the windows closed.
18
19      Ozone Removal Processes
20           The most important removal process for O3 in the indoor environment is deposition on and
21      reaction with indoor surfaces. The rate of deposition is material specific.  The removal rate
22      indoors will depend on the indoor dimensions, surface coverings, and furnishings. Smaller
23      rooms generally have larger surface-to-volume ratios (A/V) and remove O3 faster. Fleecy
24      materials, such as carpets, have larger A/V ratios and remove O3 faster than smooth surfaces
25      (Weschler, 2000). O3 can react with carpet, decreasing O3 concentrations and increasing
26      emissions of formaldehyde, acetaldehyde, and other C5-C10 aldehydes. Off-gassing of
27      4-phenylcyclohexene, 4-vinylcyclohexene, and styrene was reduced (Weschler et al., 1992). The
28      rate of O3 reaction with carpet diminishes with cumulative O3 exposure (Morrison and Nazaroff,
29      2000, 2002). Reiss et al.  (1995) reported significant quantities of acetic acid and smaller
30      quantities of formic acid off-gassing from O3 reactions with latex paint. The emission rate also
31      was relative humidity-dependent, increasing with higher relative humidity. Klen0 et al. (2001)

        August 2005                            AX3-208      DRAFT-DO NOT QUOTE OR CITE

-------
 1      evaluated O3 removal by several aged (1- to 120-month) but not used building materials (nylon
 2      carpet, linoleum, painted gypsum board, hand polished stainless steel, oiled beech parquet,
 3      melamine-coated particle board, and glass plate). Initially, O3 removal was high for all
 4      specimens tested with the exception of the glass plate.  Ozone removal for one carpet specimen
 5      and the painted gypsum board remained high throughout the study. For the oiled beech parquet
 6      and melamine-coated particle board, O3 removal leveled off to a moderate rate.  Morrison et al.
 7      (1998) reported that small amounts of O3 (-9%) are removed by lined ductwork of ventilation
 8      systems. The removal efficiency decreases with continued exposure to O3. Unlined ductwork is
 9      less efficient in removing O3.  Ozone is scavenged by fiberglass insulation (Liu and Nazaroff,
10      2001). More O3 was scavenged (60 to 90%) by fiberglass that had not been previously exposed.
11      Table AX3-25 lists the removal rates for O3 in different indoor environments.
12          Jaboki and Fabian (1997) found the O3 decays exponentially.  PAN decay is dependent on
13      room temperature and possibly the properties  and structure of the materials it comes in contact
14      with. They examined O3 and PAN decay in several closed rooms and in a car after a period of
15      intensive ventilation. Figure AX3-94 plots the O3 and PAN decay rates in these environments.
16          Several studies have examined factors within homes that may scavenge O3 and lead to
17      decreased O3 concentrations (Lee et al.,  1999; Wainman et al., 2000; Weschler and Shields,
18      1997). These reactions produce related oxidant species while reducing indoor O3 levels.  Lee
19      et al. (1999) studied 43 homes in Upland, CA in the Los Angeles metropolitan region and
20      reported that O3 declined faster in homes with more bedrooms, greater amounts of carpeting, and
21      lower ceilings (all of which alter the A/V ratio) and with the use of air conditioning.  Homes
22      with air conditioning had indoor/outdoor (I/O) ratios of 0.07, 0.09, and 0.11.  Homes without air
23      conditioning had an I/O ratio of 0.68 ±0.18.  Closed windows and doors combined with the use
24      of an air cleaner resulted in an I/O ratio of 0.15. The O3 I/O ratio was < 0.2 in homes with gas
25      stoves.
26
27      Ozone Removal Through Chemical Processes
28          Ozone chemical reactions in the indoor environment are analogous to those reactions
29      occurring in the ambient air (See Annex AX2).  Ozone reactions with unsaturated VOCs in the
30      indoor environment are dependent on the O3 indoor concentration, the indoor temperature and, in
31      most cases, the air exchange rate/ventilation rate and mixing factor.  The air exchange rate

        August 2005                           AX3-209     DRAFT-DO NOT QUOTE OR CITE

-------
         Table AX3-25.  Rate Constants (h l) for the Removal of Ozone by Surfaces in Different
                                        Indoor Environments
Indoor environment
Aluminum Room (11.9m3)
Stainless Steel Room (14.9 m3)
Bedroom (40. 8m3)
Office (5 5. 2m3)
Home (no forced air)
Home (forced air)
Department Store
Office (24. 1m3)
Office (20.7 m3)
Office/Lab
Office/Lab
Office/Lab
13 Buildings, 24 ventilation systems
Museum
Museum
Office/Lab
Office/Lab
Office
Lab
Cleanroom
Telephone Office
43 Homes
Surface Removal Rate,
k,, (A/V), h-1
3.2
1.4
7.2
4.0
2.9
5.4
4.3
4.0
4.3
4.3
3.2
3.6
3.6"
4.3
4.3
4
3.2
2.5
2.5
7.6
0.8- 1.0"
2.8±1.3
Reference
Mueller et at. (1973)
Sabersky et al. (1973)
Thompson et al. (1973)
Allen etal. (1978)
Shair and Heitner( 1974)
Shair(1981)
Nazaroff and Cass ( 1986)

Weschler etal. (1989)



Weschler et at. (1994)b
Lee etal. (1999)
        a A/V = assumes surface area to volume ratio = 2.8 m '
        b Large office, small A/V
        Source: Weschler (2000).
1     determines the amount of time available for chemical reactions to take place. At low air
2     exchange rate, the residence time for the pollutants is longer, the reaction time is greater, and the
3     concentration of the products produced by O3 chemistry is greater (Weschler and Shields, 2000,
4     2003). Since O3 is primarily an outdoor pollutant, the air exchange rate will influence the
5     amount of O3 occurring indoors.
      August 2005
AX3 -210      DRAFT-DO NOT QUOTE OR CITE

-------
                                                                    PAN - residence, parquet
                     10
20
30    40    50     60    70    80
       Elapsed Time (minutes)
90    100   110   120
        Figure AX3-94.  Ozone decay processes versus time measured for several indoor rooms.
                        Room temperature in all cases was -27 °C and 29 °C in the car.
        Adapted from: Jaboki and Fabian (1997).
 1          Most unsaturated VOCs in the indoor environment are terpenes or terpene-related
 2     compounds from cleaning products, air fresheners, and wood products. Some of the reaction
 3     products may more negatively impact human health and artifacts in the indoor environment than
 4     their precursors (Wolkoff et al., 1999; Wilkins et al., 2001; Weschler et al., 1992; Weschler and
 5     Shields, 1997; Rohr et al., 2002; N0jgaard et al., 2005). The reactions products of O3 and
 6     terpenes are Criegee biradicals,  nitrate radicals, and peroxyacetyl radicals. Secondary reaction
 7     of the pollutants may form hydroxy, alkyl, alkylperoxy, hydroperoxy, and alkoxy radicals. The
 8     reaction of O3 with alkenes can  produce aldehydes, ketones, and organic acids (Weschler and
 9     Shields, 2000; Weschler et al., 1992).
10
       August 2005
                   AX3-211
                          DRAFT-DO NOT QUOTE OR CITE

-------
 1          The indoor chemistry of O3 is described by the following equations. The initial reaction
 2     produces an ozonide which rapidly decomposes into one of the two possible combinations.
 3
 4                         O3 +  R,C(O)R2C - CR3R4 -> ozonide                   (AX3-6)
 5
 6                       ozonide->R,C(O)R2 + [R3R4C»OO»]*                  (AX3-7)
 7
 8                                               or
 9
10                       ozonide ->• [R^C « OO •]* + R3C(O)R4                  (AX3-8)
11
12     The biradical (*) then rearranges or reacts to produce the highly reactive intermediate products
13     (hydroxy, hydroperoxy, and alkylperoxy radicals) and stable products (aldehydes, ketones and
14     organic acids).
15          Hydroxy radicals formed from the reaction of O3 with VOCs are lost to further reaction
16     with VOCs. For each molecule of O3 consumed, approximately one molecule of the hydroxy
17     radical is produced.  The hydroxy radical also is formed through the reaction of nitric oxide and
18     hydroperoxy, and other intermediate products formed from the reaction of O3 with unsaturated
19     VOCs (Sarwar et al., 2002; Orzechowska and Paulson, 2002).  The hydroxy radical can react
20     with various nitrogen compounds, sulfur dioxide, carbon monoxide and other compounds to
21     produce significantly more toxic compounds. In studies by Pick et al. (2003, 2004), the
22     formation of norpinonic and pinonic acid in a ventilation system injected with a-pinene in the
23     presence of O3 was reported to be almost exclusively the result of oxidation by hydroxy radicals.
24     Fan et al. (2003) attributed the formation of some secondary organic aerosols from the reaction
25     of O3 with 23 VOCs (as toluene) to reactions with hydroxy radicals. Van den Bergh et al. (2000)
26     found that formaldehyde, acetaldehyde, acetone, campholenealdehyde, and pinonaldehyde are
27     generated from the reaction of a-pinene with hydroxy radicals.  A list of the VOCs occurring in
28     the indoor environment known to react with O3 and OH radicals is found in Weschler (2000) and
29     Nazaroff and Weschler (2004).
       August 2005                           AX3-212     DRAFT-DO NOT QUOTE OR CITE

-------
 1           Wilkins et al. (2001) found that formaldehyde, formic acid, acetic acid, methacrolein, and
 2      methylvinyl ketone were produced within 30 seconds of mixing approximately 4.0 ppm O3,
 3      500 ppm isoprene and 4.0 ppmNO2. Only a small amount of the O3 remained. Similar findings
 4      were reported by Clausen et al. (2001).  A 16 second reaction of a mixture of 4.0 ppm O3 and
 5      48 ppm linomene was found to produce l-methyl-4-acetcyclohexene, 3-isopropenyl-6-
 6      oxoheptanal, formaldehyde and formic acid. Acetone, acrolein and acetic acid also were
 7      detected, however, the authors found the production of these compounds difficult to explain
 8      based on the structure of linomene and suggested that they may be artefacts.  When ^-pinene
 9      was injected into a ventilation system containing 75 ppb O3 norpinic acid, pinic acid, glyoxal,
10      methyl glyoxal, norpinionic acid, pinonic acid, a C4 dicarbonyl (C4H6O2), a C5 dicarbonyl
11      (C5H8O2), norpionaldehyde, and pinonaldehyde were formed (Pick et al., 2003, 2004).
12           The reaction between O3 and terpenes has been shown to increase the concentration of
13      indoor particles (Weschler and Shields, 1999, 2003; Weschler, 2004; Clausen et al., 2001; Fan et
14      al, 2003; Wainman et al., 2000).  Sarwar et al. (2002) suggested that the hydroxy radical reacts
15      with terpenes to produce products with low vapor pressures that contribute to fine particle
16      growth. The acidity of particles was found to enhance the yield of secondary organic aerosols
17      when a-pinene ozonolysis was carried out in the presence of ammonium sulfate and sulfuric
18      acid.  There was almost a 40% increase in organic carbon particles when a-pinene ozonolysis
19      occurred in the presence of sulfuric acid aerosols compared to ammonium sulfate aerosols
20      (linuma et al., 2004). Rohr et al. (2002, 2003) measured particle formation in a plexiglas
21      chamber as part of a mouse bioassay study.  They found an increase in ultrafine particle numbers
22      as the result of O3 and a-pinene reactions. When O3 was introduced into the  test chamber,
23      particle concentrations increased to >107 particles/cm3.  Clausen found that the reaction of
24      limonene vapor with O3 increased the particle concentration to 3 x 10s from a background
25      concentration of <103 particles/cm3.  Poupard et al. (2005) and Blondeau et al. (2005) found a
26      negative correlation between O3 and particle concentration in eight school buildings in France.
27      The researchers assumed that the increased particle concentration and decreased O3
28      concentration was likely the result of homogeneous processes involving O3.  However, the
29      assumption could not be verified because the particles were not analyzed  for chemical
30      composition.
        August 2005                            AX3-213     DRAFT-DO NOT QUOTE OR CITE

-------
 1          Weschler et al. (1992) suggested that the reaction been O3 and NO2 in the indoor
 2      environment may be a significant source of HNO3.  When there are elevated concentrations of
 3      both O3 and NO2 in the indoor environment, the following reaction sequence may occur:
 4
 5                                     O3  + NO2 - NO3  + O2
 6
 7                                       NO3 +  NO2  ^ N2O5
 g                                                                                    (AX3-9)
 9                                      N2O5 + H2O  2HNO3
10
11                                  NO3 + ORG -  HNO3  +  ORG
12
13          PAN and PPN are thermally unstable and will decomposed in the indoor environment to
14      peroxyacetyl radicals and NO2 (See equation AX3-10).  Decomposition and formation of PAN in
15      the indoor environment is influenced by NO2 and NO.
16
17                              CH3C(O)OONO2  *  CH3C(O)OO  + NO2               (AX3-10)
18
19          When the concentration ratio of NO/NO2 is greater than 7, less than 10% of the
20      peroxyaceyl radicals will revert to PAN.  Decomposition of PAN is expected to be a relatively
21      fast process when indoor O3 levels are low and when motor vehicle emissions are large or there
22      is an indoor source of NOX (Weshcler and Shields, 1997).
23
24      Sources and Emissions of Indoor Ozone
25          Ozone enters the indoor environment primarily through infiltration from outdoors through
26      cracks and crevices in the building envelope (unintentional  and uncontrolled ventilation) and
27      through building components such as windows and doors and ventilation systems (natural and
28      controlled ventilation).  Natural ventilation is  driven by the natural forces of wind and
29      temperature. Possible indoor sources of O3 are office equipment (photocopiers, facsimile
30      machines, and laser printers) and air cleaners (electrostatic air filters and precipitators and O3
31      generators). Generally  O3 emissions from photocopiers and laser printers are limited due to
32      installed filtering systems (Black and Worthan, 1999; Leovic et al., 1996; Aldrich et al., 1995).
33      However, emissions increase under improper maintenance conditions (Leovic et al., 1996).
34      Well-maintained photocopiers and laser printers usually emit low levels of O3 by catalytically

        August 2005                           AX3-214     DRAFT-DO NOT  QUOTE OR CITE

-------
 1     reducing the O3 to oxygen (Aldrich et al., 1995). Leovic et al. (1996, 1998) measured O3
 2     emissions from four dry-process photocopiers.  Ozone emissions ranged from 1300 to
 3     7900 |ig/h.
 4           Most electrostatic air filters and precipitators are designed to minimize the production of
 5     O3. However, if excessive arcing occurs, the units can emit a significant amount of O3 into the
 6     indoor environment (Weschler, 2000). Niu et al. (2001) measured O3 emissions from 27 air
 7     cleaners that used ionization processes to remove particulates.  The test were conducted in a
 8     2 x 2 x 1.60 m3 stainless steel environmental chamber.  The tests were terminated after 1.5 h
 9     if no increase in O3 concentration was noted. If an increase in O3 was noted, the test was
10     continued, in some cases, for up to 20 h. Most of the evaluated units emitted no O3 or only
11     small amounts. Five units were found to emit O3 ranging from 56 to 2757 |ig/h.
12           There are many brands and models of O3 generators commercially available.  The amount
13     of O3 emitted by each unit depends on the size of the unit. Ozone emission rates have been
14     reported to range from tens to thousands of micrograms per hour (Weschler, 2000; Steiber,
15     1995).  Ozone emissions supposedly  can be regulated using the units control features. However,
16     available information suggests that O3 output may not be proportional to the control setting.
17     Some units are equipped with a sensor that automatically controls O3 output by turning the unit
18     on and off. The effectiveness of the sensor is unknown (U.S. Environmental Protection
19     Agency, 2002).
20           Peroxyacyl nitrates (PAN and PPN) have no known direct emission sources and are
21     formed in the atmosphere from the reaction of NO2 and hydrocarbons (Grosjean et al., 1996).
22     Peroxyacyl nitrates primarily occur in the indoor environment from infiltration through the
23     building envelop and through openings in the building envelopment. Peroxyacyl nitrates also
24     may be formed in the indoor environment through radical chemistry. PAN may be formed from
25     the reaction of the OH- or NO3 with acetaldehyde to form the acetyl radical, CH3CO. The acetyl
26     radical then reacts with oxygen to form and acetylperoxy radical which reacts with NO2 to form
27     PAN.
28
29
30                     OH-(or NO3) + CH3CHO -> CH3CO-                        (AX3-11)
31
32
33                            CH3CO- + O2 -» CH3C(O)OO-                       (AX3-12)

       August 2005                            AX3-215     DRAFT-DO NOT QUOTE OR CITE

-------
34
35                      CH3C(O)OO- + NO2 -> CH3C(O)OO NO2                 (AX3-13)
36
37     PPN is formed from when the reaction of the OH- with propionaldehyde (Weschler and
38     Shields, 1997).
39
40     AX3.10.8  Trends in Concentrations Within Microenvironments
41          There have not been sufficient numbers of measurements of personal exposures or indoor
42     O3 concentrations to directly document trends over time or location. However, since O3
43     concentrations in all microenvironments are primarily derived from ambient sources, trends in
44     ambient air concentrations should reflect trends in personal exposure unless there are differences
45     in activity patterns over time or locations. Overall, a significant downward trend in ambient O3
46     concentrations has occurred from 1980 in most locations in the United States, although the trend
47     in the latter part of the 1990s suggests that continued improvements in air quality may have
48     leveled off. Greater declines in ambient O3 concentrations appear to have occurred in urban
49     centers than in rural regions. The decline in daily and weekly average O3 concentrations from
50     1989 to 1995 in rural regions was 5% and 7%, respectively (Wolff et al., 2001; Lin et al., 2001;
51     Holland et al., 1999). A detailed discussion of O3 trends appears earlier in this annex.
52
53     AX3.10.9  Characterization of Exposure
54     AX3.10.9.1   Use of Ambient Ozone Concentrations
55          The use of ambient air monitoring  stations is still the most common surrogate for assigning
56     exposure in epidemiological studies.  Since the primary source of O3 exposure is the ambient air,
57     monitoring concentration data would provide the exposure outdoors while exercising, a potential
58     important exposure to evaluate in epidemiological studies as well as a relative assignment of
59     exposure with time if the concentration were uniform across the region; the time-activity pattern
60     were the same across the population; and the housing characteristics, such as ventilation rates
61     and the O3 sinks contributing to its indoor decay rates, were constant for the  study area. Since
62     these factors vary by population and location there will be errors in not only  the magnitude of the
63     total exposure, but also in the relative total exposure assignment based solely on ambient
64     monitoring data. As discussed earlier in this section, spatial differences in O3 concentrations
65     within a city and between the height of the monitor and the breath zone (1 to 2 m)  exist,
       August 2005                           AX3-216     DRAFT-DO NOT QUOTE OR CITE

-------
 1      increasing uncertainties. The potential exists to obtain more complete exposure assignments for
 2      both individuals and populations by modeling O3 exposure based on ambient air concentration to
 3      account for spatial variations outdoors and for time spent indoors, provided housing
 4      characteristics and activity patterns can be obtained. For cohort studies, measurement of
 5      personal O3 exposures using passive monitors is also possible.
 6           The potential for error in determining pollutant exposure was expressed by Krzyzanowski
 7      (1997), who indicated that while the typical estimate of exposure in epidemiological studies is
 8      "an average concentration of the pollutant calculated from the data routinely collected in the area
 9      of residence of the studied population. This method certainly lacks precision and, in most cases,
10      the analyses that use it will underestimate the effect of specific concentrations of a pollutant on
11      health." It is further stated that when estimating exposure for epidemiological studies it is
12      important to define:  (1) representativeness of exposure or environmental data for the population
13      at risk, (2) appropriateness of the averaging time for the health outcome being examined, and
14      (3) the relationship between the exposure surrogate and the true exposure relative to the
15      exposure-response function used in the risk assessment. Zeger et al. (2000) suggested that the
16      largest biases will occur because of errors between ambient and average personal exposure
17      measures.  Sheppard et  al. (2005a) further indicated that for non-reactive pollutants with non-
18      ambient sources,  exposure variability will introduce a large exposure error. However, for acute
19      effects, time series studies using ambient concentration measurements are adequate (Sheppard,
20      2005b).
21           Numerous air pollutants can have common ambient air sources resulting in strong
22      correlations among pollutant ambient air concentrations.  As a result, some observed
23      associations between an air pollutant and health effects may be due to confounding by other air
24      pollutants.  Sarnat et al. (2001) found that while ambient air concentrations of some air
25      pollutants were correlated, personal PM2 5 and several gaseous air pollutant (O3, SO2, NO2, CO,
26      and exhaust-related VOCs) exposures were not generally correlated.  The findings were based on
27      the results of a multipollutant exposure study of 56 children and elderly adults in Baltimore, MD
28      conducted during both the summer and winter months. Ambient pollutant concentrations were
29      not associated with corresponding personal exposures, except for PM2 5. The gaseous pollutants
30      were found to be  surrogates of PM25 and were generally not correlated.  The authors concluded
31      that multipollutant models in epidemiologic studies of PM2 5 may not be suitable, and health

        August 2005                            AX3-217     DRAFT-DO NOT QUOTE OR CITE

-------
 1      effects attributed to the gaseous pollutants may be the result of PM25 exposure. It should be
 2      noted that the 95th percentile O3 concentrations in the study was lower than 60 ppb, an O3
 3      concentration at which respiratory effects are noted.  It would be important to examine whether
 4      O3 is a surrogate for personal PM2 5 at high O3 levels when attributing adverse health effects to
 5      O3 or PM2 5.
 6           Kiinzli et al. (1996) assessed potential lifetime  exposure to O3 based on the responses to a
 7      standardized questionnaire completed by 175 college freshmen in California.  Questions
 8      addressed lifetime residential history, schools attended, general and outdoor activity patterns,
 9      driving habits and job history. The purpose was to determine what O3 monitoring data to use for
10      each time period of their lives, the potential correction factor for indoor levels and periods of
11      high activity to account for differential doses to the lung due to physical activity.  The reliability
12      of the responses was checked by having each respondent complete the questionnaire twice, on
13      different days, and the results compared.  A lifetime  O3 exposure history was generated for each
14      participant and a sensitivity analysis performed to evaluate which uncertainties would cause the
15      greatest potential misclassification of exposure.  Assigned lifetime O3 concentrations from the
16      nearest monitor yielded highly reliable cumulative values, although the reliability of residential
17      location decreased with increasing residential locations. Individuals involved in moderate and
18      heavy exercise could be reliably identified.  Such an approach can be used to evaluate health
19      outcomes associated with chronic exposures to O3.
20
21      AX3.10.9.2  Exposure Selection in  Controlled Exposure Studies
22           Ozone exposures in the environment are variable over time due to changes in the ambient
23      concentrations during the day as the photochemical reactions proceed and also because people
24      move between microenvironments that have different concentrations (Johnson, 1997).
25      Exposures are repeated on sequential days since weather conditions that produce O3 can move
26      slowly through or become stagnant within a region.  For simplicity, most controlled-exposure
27      experiments are conducted at a single concentration for a fixed time period, with a limited
28      number of studies being repeated on a single individual. Few studies have been conducted using
29      multipollutants or photochemical agents other than O3, to better represent "real-world" exposures
30      with the exception of NO2. Studies by Hazucha et al. (1992),  Adams (2003), and Adams and
31      Ollison (1997) examined the effect of varying O3 exposure concentrations on pulmonary
32      function.  A description of, and findings in, the studies appears in  Annex AX5 of this document.

        August 2005                            AX3-218     DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX3.10.9.3  Exposure to Related Photochemical Agents
 2           Exposures to other related photochemical agents have not been measured using personal
 3      samplers nor are these agents routinely measured at O3 monitoring stations.  Photochemical
 4      agents produced in the ambient air can penetrate indoors and react with other pollutants to
 5      produce other potentially irritating compounds. Reiss et al. (1995) reported that organic acids,
 6      aldehydes and ketones were produced indoors by reactions of O3 with VOCs.  The produced
 7      compounds included oxidants that can be respiratory irritants.  The indoor concentrations were
 8      dependent upon the O3 concentrations indoors and the AER within the building.  Weschler and
 9      Shields (1997) summarized indoor air chemical reactions that depend directly or indirectly on
10      the presence of O3.  They indicated that O3 concentrations are lower indoors than outdoors partly
11      because of gas-phase reactions that produce other oxidants in an  analogous fashion to
12      photochemical smog in ambient air.  The production of these species indoors is a function of the
13      indoor O3 concentration and the presence of the other necessary precursors, VOCs, and NO2,
14      along with an optimal AER.  A variety of the photochemical oxidants related to O3 that are
15      produced outdoors, such as PAN and PPN, can penetrate indoors. These oxidants are thermally
16      unstable and can decompose indoors to peroxacetyl radicals  and NO2 through thermal decay.
17      PAN removal increases with increasing temperature, and at a given temperature, with increasing
18      NO/NO2 concentration ratio (Grosjean et al., 2001).  Other free radicals that can form indoors
19      include HO» and HO2'.  These free radicals can produce compounds that are known or suspected
20      to be irritating.  Little is known about exposure to some of these agents, as not all have been
21      identified and collection and analytical methodologies have not be developed for their routine
22      determination.  Lee et al.  (1999) reported that homogeneous (gas phase) and heterogenous (gas
23      phase/solid surface) reactions occur between O3 and common indoor air pollutants such as NO
24      and VOCs  to produce secondary products whose  production rate  depends on the AER and
25      surface area within the home. Wainman et al. (2000) found that O3 reacts indoors with
26      J-limonene, emitted from air fresheners, to form fine particles in  the range of 0.1 to 0.2 |im and
27      0.2 to 0.3 |im. The indoor process also produces  compounds that have been identified in the
28      ambient atmosphere. These species, plus others that may form indoors from other terpenes or
29      unsaturated compounds can present an additional exposure to oxidants, other than O3, at higher
30      concentrations than present in ambient air, even as the O3 concentration is being reduced
31      indoors.

        August 2005                             AX3-219     DRAFT-DO NOT QUOTE OR CITE

-------
 1           The announcement of smog alerts or air quality indices may influence individuals to alter
 2      behaviors (avoidance behavior). Neidell (2004), in his evaluation of the effect of pollution on
 3      childhood asthma, examined the relationship between the issuance of smog alerts or air quality
 4      indices for several counties in California and hospital admissions for asthma in children under
 5      age 18 years (not including newborns).  Smog alerts are issued in California on days when O3
 6      concentrations exceed 200 ppb. There was a significant reduction in the number of asthma-
 7      related hospital admissions in children ages 1 to 12 years on smog alert days, indicating that
 8      avoidance behavior might be present on days of high O3 concentrations.
 9           Ozone exposure modeling has been conducted for the general population and sensitive
10      subgroups.  The pNEM/O3 model takes into consideration the temporal and spatial distribution
11      of people and  O3 throughout the area of consideration, variations in O3 concentrations within
12      microenvironments, and the effects of exertion/exercise (increased ventilation) on O3 uptake.
13      The pNEM/O3 model consists of two principal parts: the cohort exposure program and the
14      exposure extrapolation program.  The methodology incorporated much of the general framework
15      described earlier in this section on assessing O3 exposure and consists of five steps: (1) define
16      the study area, population of interest, subdivisions of the study area, and exposure period;
17      (2) divide population of interest into a set of cohorts; (3) develop exposure event sequence for
18      each cohort for the exposure period; (4) estimate pollutant concentration and ventilation rate for
19      each exposure event; and (5) extrapolate cohort exposures to population of interest (U.S.
20      Environmental Protection Agency,  1996b).
21           There are three versions of the pNEM/O3 model: general population (Johnson et al.,
22      1996a), outdoor workers (Johnson et al., 1996b), and outdoor children (Johnson et al.,  1996c,
23      1997). These  three versions of the model have been applied to nine urban areas.  The model also
24      has been applied to a single summer camp (Johnson et al.,  1996c).  The general population
25      version of the  model uses activity data from the Cincinnati Activity Diary Study (CADS;
26      Johnson, 1989). Time-activity studies (Wiley et al., 1991a; Johnson, 1984; Linn et al., 1993;
27      Shamoo et al., 1991; Goldstein et al., 1992; Hartwell et al., 1984) were combined with the CADS
28      data for the outdoor worker version of the model.  Additional time-activity data (Goldstein et al.,
29      1992; Hartwell et al., 1984; Wiley et al., 1991a,b; Linn et al., 1992; Spier et al., 1992) were also
30      added to CADS for the outdoor children of the model (U.S. Environmental Protection Agency,
31      1996b).  Home-work commuting patterns are based on information gathered by the U.S. Census

        August 2005                            AX3-220      DRAFT-DO NOT QUOTE OR CITE

-------
 1      Bureau.  Ozone ambient air concentration data from monitoring stations were used to estimate
 2      the outdoor exposure concentrations associated with each exposure event. Indoor O3 decay rate
 3      is assumed to be proportional to the indoor O3 concentration.  An algorithm assigns the
 4      equivalent ventilation rate (EVR) associated with each exposure event. The outdoor children
 5      model uses an EVR-generator module to generate an EVR value for each exposure event based
 6      on data on heart rate by Spier et al. (1992) and Linn et al. (1992). The models produce exposure
 7      estimates for a range of O3 concentrations at specified exertion levels. The models were used to
 8      estimate exposure for nine air quality scenarios (U.S. Environmental Protection Agency, 1996b).
 9           Korc (1996) used the REHEX-II model, a general purpose air pollution exposure model
10      based on a microenvironmental approach modified to  account for the influence of physical
11      activity along and spatial and the temporal variability  of outdoor air pollution.  Ozone  exposure
12      was estimated by demographic groups across 126 geographic subregions for 1980 to 1982, and
13      for 142 geographic subregions for 1990 to 1992. Simulation results were determined for
14      population race, ethnicity, and per capita income and included indoor, in-transit, and outdoor
15      microenvironments.  Exposure modeling was stratified by age because of differences in time-
16      activity patterns.  Exposure distributions by regional activity pattern data were not considered,
17      rather it was assumed that all individuals within a county had the same exposure distribution by
18      race, ethnicity, and socioeconomic status. Model results for southern California indicated that
19      the segment of the population with the highest exposures were children 6 to  11 years old.
20      Individuals living in low income districts may have greater per capita hours of exposure  to O3
21      above the NAAQS than those living in higher income districts. The author indicated that O3
22      exposure differences by race and ethnicity have declined over time. The noninclusion of details
23      on activity patterns for different populations in the model limit the extrapolations that can be
24      made from the model results.
25           Children appear to have higher exposures than adults and the elderly. Asthmatics appear to
26      ventilate more than healthy individuals, but tend to protect themselves by decreasing their
27      outdoor exercise (Linn et al., 1992). Additional data are still needed to identify and better define
28      exposures to potentially susceptible populations and improve exposure models for the  general
29      population and subpopulation of concern.
30
        August 2005                            AX3-221      DRAFT-DO NOT QUOTE OR CITE

-------
  1       REFERENCES

  2       Adams, W. C. (1993) Measurement of breathing rate and volume in routinely performed daily activities [final
  3             report]. Sacramento, CA: California Environmental Protection Agency, Air Resources Board; contract no.
  4             A033-205.
  5       Adams, W. C. (2003) Comparison of chamber and face mask 6.6-hour exposure to 0.08 ppm ozone via square-wave
  6             and triangular profiles on pulmonary responses. Inhalation Toxicol. 15: 265-281.
  7       Adams, W. C.; Ollison, W. M. (1997) Effects of prolonged simulated ambient ozone dosing patterns on human
  8             pulmonary function and symptomatology. Presented at: 90th annual meeting of the Air & Waste Management
  9             Association; June; Toronto, Ontario, Canada. Pittsburgh, PA: Air & Waste Management Association; paper
10             no. 97-MP9.02.
11       Adams, R. M.; Hamilton, S. A.; McCarl, B. A. (1985) An assessment of the economic effects of ozone on U.S.
12             agriculture. J. AirPollut. Control Assoc. 35: 938-943.
13       Adams, R. M.; Glyer, J. D.; Johnson, S. L.; McCarl, B. A. (1989) A reassessment of the economic effects of ozone
14             on U.S. agriculture. JAPCA 39: 960-968.
15       Ainsworth; Haskell, W. L.; Leon, A. S.; Jacobs, D. R., Jr.; Montoye, H. J.; Sallis, J. F.; Paffenbarger, R. S., Jr.
16             (1993) Compendium of physical activities: classification of energy costs of human physical activities. Med.
17             Sci. Sports Exer. 25: 71 -80.
18       Aldrich, F. D.; Brooks, B. O.; Davis, W. F.; DeBroy, J. A.; Li, S. T.; Schimke, R.; Utter, G. M.  (1995) Health and
19             productivity impact of organic volatiles from electronic office equipment.  Presented at: Pacific Rim
20             conference on occupational and environmental health; October;  Sydney, Australia.
21       Allen, R. J.; Wadden, R. A.; Ross, E. D. (1978) Characterization of potential indoor sources of ozone. Am. Ind. Hyg.
22             Assoc. J. 39: 466-471.
23       Aneja, V. P.; Li, Z. (1992) Characterization of ozone at high elevation in the eastern United States: trends, seasonal
24             variations, and exposure. J. Geophys. Res. [Atmos.]  97: 9873-9888.
25       Avol, E. L.; Navidi, W. C.; Colome, S. D. (1998a) Modeling ozone levels in and around southern California homes.
26             Environ. Sci. Technol.  32: 463-468.
27       Avol, E. L.; Navidi, W. C.; Rappaport, E. B.; Peters, J. M. (1998b) Acute effects of ambient ozone on asthmatic,
28             wheezy, and healthy children. Cambridge, MA: Health Effects Institute; research report no. 82.
29       Beine, H. J.; Jaffe, D. A.; Herring, J. A.; Kelley, J. A.; Krognes, T.; Stordal, F. (1997) High-latitude springtime
30             photochemistry. Part I: NOX, PAN and ozone relationships. J. Atmos. Chem. 27:  127-153.
31       Bernard, N. L.; Gerber, M. J.; Astre, C. M.; Saintot, M. J. (1999) Ozone measurement with passive samplers:
32             validation and use for ozone pollution assessment in Montpellier, France. Environ. Sci. Technol. 33: 217-222.
33       Berntsen, T. K.; Isaksen, I. S.  A.; Myhre, G.; Fuglestvedt, J. S.; Stordal, F.; Larsen, T. A.; Freckleton, R. S.; Shine,
34             K. P. (1997) Effects of anthropogenic emissions on tropospheric ozone and its radiative forcing. J. Geophys.
35             Res. [Atmos.] 102: 28,101-28,126.
36       Berry, C. R. (1964) Differences in concentrations of surface oxidant between valley and mountaintop conditions in
37             the southern Appalachians. J. AirPollut. Control Assoc.  14: 238-239.
38       Bey, I.; Jacob, D. J.; Yantosca, R. M.; Logan, J. A.; Field, B.; Fiore, A. M.; Li, Q.; Liu, H.; Mickley, L. J.; Schultz,
39             M. G. (200 la) Global modeling of tropospheric chemistry with assimilated meteorology: model description
40             and evaluation. J. Geophys. Res.  (Atmos.) 106: 23,073-23,095.
41       Bey, I.; Jacob, D. J.; Logan, J. A.; Yantosca, R. M. (200Ib) Asian chemical outflow to the Pacific in spring: origins,
42             pathways, and budgets. J. Geophys. Res.  [Atmos.] 106: 23,097-23,113.
43       Biller, W. F.; Feagans, T. B.; Johnson, T. R.; Duggan, G. M.; Paul, R. A.; McCurdy, T.; Thomas, H. C. (1981)
44             A general model for estimating exposure associated with alternative NAAQS. Presented  at: 74th annual
45             meeting of the Air Pollution Control Association; June; Philadelphia, PA.  Pittsburgh, PA: Air Pollution
46             Control Association; paper no. 81-18.4.
47       Billick, I.; Johnson, D.; Moschandreas, D.; Relwani, S. (1984) An investigation of operational factors that influence
48             emission rates from gas appliances. In: Berglund, B.; Lindvall, T.; Sundell, J., eds. Indoor air. Volume 4:
49             chemical characterization and personal exposure. Proceedings of the 3rd international conference on indoor
50             air quality and climate; August; Stockholm, Sweden. Stockholm, Sweden: Swedish Council for Building
51             Research; pp. 181-187.
52       Billick, I. H.; Behrens, D. W.; Becker, E. W.; Garrison, C. A.; Wilson, A. L.; Colome, S. D.; Tian, Y. (1996)
53             California residential indoor air quality study: an overview. In: Proceedings of the 1995 international gas
54             research conference; November 1995; Cannes, France. Rockville, MD: Government Institutes, Inc.;
55             pp. 1707-1716.
         August 2005                                  AX3-222      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Black, M. S.; Worthan, A. W. (1999) Emissions from office equipment. In: Raw, G.; Aizlewood, C.; Warren, P.,
  2             eds. Indoor Air '99, v. 2. London, England: Construction Research Communications Ltd.; pp. 454-459.
  3       Black, D. R.; Harley, R. A.; Hering, S. V.; Stolzenburg, M. R. (2000) A new, portable, real-time monitor. Environ.
  4             Sci. Technol. 34:3031-3040.
  5       Blanchard, C. L.; Tanenbaum, S. J. (2003) Differences between weekday and weekend air pollutant levels in
  6             southern California. J. Air Waste Mange. Assoc. 53: 816-828.
  7       Blondeau, P.; lordache, V.; Poupard, O.; Genin, D.; Allard, F. (2005) Relationship between outdoor and indoor air
  8             quality in eight French schools. Indoor Air 15: 2-12.
  9       Bohm, M.; McCune, B.; Vandetta, T. (1991) Diurnal curves of troposheric ozone in the western United States.
10             Atmos. Environ. Part A 25: 1577-1590.
11       Bottenheim,  J. W.; Barrie, L. A.; Atlas, E.; Heidt, L. E.; Niki, H.; Rasmussen, R. A.; Shepson, P. B. (1990)
12             Depletion of lower tropospheric ozone during Arctic spring: The Polar Sunrise Experiment 1988. J. Geophys.
13             Res. (Atmos.) 95:  18,555-18,568.
14       Bottenheim,  J. W.; Barrie, L. A.; Atlas, E. (1993) The partitioning of nitrogen oxides in the lower Arctic troposphere
15             during Spring 1988. J. Atmos. Chem. 17: 15-27.
16       Brauer, M.; Brook, J. R. (1997) Ozone personal exposures and health effects for selected groups residing in the
17             Fraser Valley. In: Steyn, D. G.; Bottenheim, J. W., eds. The Lower Fraser Valley Oxidants/Pacific  '93 Field
18             Study. Atmos. Environ. 31: 2113-2121.
19       Browell, E. V; Hair, J. W.; Butler, C. F.; Grant, W. B.; De Young, R. J.; Fenn, M. A.; Brackett, V G.; Clayton,
20             M. B.; Brasseur, L. A.; Harper, D. B.; Ridley, B. A.; Klonecki, A. A.; Hess, P. G.; Emmons, L. K.; Tie, X.;
21             Atlas, E. L.; Cantrell, C. A.; Wimmers, A. J.; Blake, D. R.; Coffey, M. T.; Hannigan, J. W.; Dibb, J. E.;
22             Talbot, R. W.; Flocke, F.; Weinheimer, A. J.; Fried, A.; Wert, B.;  Snow, J.  A.; Lefer, B. L. (2003) Ozone,
23             aerosol, potential vorticity, and trace gas trends observed at high-latitudes over North America from February
24             to May 2000. J. Geophys. Res. [Atmos.] 108(D4): 10.1029/2001JD001390.
25       Burke, J. M.; Zufall, M. J.; Ozkaynak, H. (2001) A population exposure model for paniculate matter: case study
26             results for PM2 5 in Philadelphia, PA. J. Exposure Anal. Environ. Epidemiol. 11: 470-489.
27       Carroll, R. J.; Chen, R.; George, E. L; Li, T. H.; Newton, H. J.; Schmiediche, H.; Wang, N. (1997a) Ozone exposure
28             and population density in Harris County, Texas. J. Am. Stat. Assoc. 92: 392-404.
29       Carroll, R. J.; Chen, R.; George, E. L; Li, T. H.; Newton, H. J.; Schmiediche, H.; Wang, N. (1997b) Ozone exposure
30             and population density in Harris County, Texas (rejoinder). J. Am. Stat. Assoc. 92: 414-415.
31       Chan, C.-C.; Wu, T.-H. (2005) Effects of ambient ozone exposure on mail carriers' peak expiratory flow rates.
32             Environ. Health Perspect.  113: 735-738.
33       Chang, L.-T.; Koutrakis, P.; Catalano, P. J.; Sun, H. H. (2000) Hourly personal exposures to fine particles and
34             gaseous pollutants—results from Baltimore, Maryland. J. Air Waste Manage. Assoc. 50: 1223-1235.
3 5       Chao, C. Y. (2001) Comparison between indoor and outdoor air contaminant levels in residential buildings from
36             passive sampler study. Build. Environ. 36: 999-1007.
37       Chen, L.-W.  A. (2002) Urban fine paniculate matter: chemical composition and possible origins (dissertation).
38             College Park, MD: University of Maryland, Department of Chemical Physics. Available from: University
39             Microfilms, Ann Arbor, MI; AADAA-I3078297.
40       Chinkin, L. R.; Coe, D. L.; Funk, T. H.; Hafner, H. R.; Roberts, P. T.; Ryan, P. A. (2003) Weekday versus weekend
41             activity patterns for ozone precursor emissions in California's South Coast Air Basin. J. Air Waste Manage.
42             Assoc. 53:829-843.
43       Christakos, G. (2000) Modern spatiotemporal geostatistics. New York, NY: Oxford University Press. (International
44             Association for Mathematical Geology; studies in mathematical geology: 6).
45       Christakos, G.; Hristopulos, D. T. (1998) Spatiotemporal environmental health modelling. A tractatus stochasticus.
46             Boston, MA: Kluwer Academic Publishers.
47       Christakos, G.; Kolovos, A. (1999) A study of the spatiotemporal health impacts of ozone exposure. J. Exposure
48             Anal.  Environ. Epidemiol. 9: 322-335.
49       Christakos, G.; Vyas, V. M. (1998a) A composite space/time  approach to studying ozone distribution over eastern
50             United States. Atmos. Environ. 32: 2845-2857.
51       Christakos, G.; Vyas, V. M. (1998b) A novel method for studying population health impacts of spatiotemporal
52             ozone distribution. Soc. Sci. Med. 47: 1051-1066.
53       Clausen, P. A.; Wilkins, C. K.; Wolkoff, P.; Nielsen, G. D. (2001) Chemical and biological evaluation of a reaction
54             mixture of R-(+)-limonene/ozone: formation of strong  airway irritants. Environ. Int. 26: 511-522.
55       Clench-Aas,  J.; Bartonova, A.; Behler, T.; Grenskei, K. E.; Sivertsen, B.; Larssen, S. (1999) Air pollution exposure
56             monitoring and estimating. Part I. Integrated air quality monitoring system. J. Environ. Monit.  1:313-319.
         August 2005                                  AX3-223       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Code of Federal Regulations. (2000) Appendix D to part 58—Network design for state and local air monitoring
  2             stations (SLAMS), national air monitoring stations (NAMS), and photochemical assessment monitoring
  3             stations (PAMS). C. F. R. 40: pt. 58, app. D.
  4       Coffey, P.; Stasiuk, W.; Mohnen, V. (1977) Ozone in rural and urban areas of New York State: part I. In:
  5             Dimitriades, B., ed. International conference on photochemical oxidant pollution and its control -
  6             proceedings: volume I; September 1976; Raleigh, NC. Research Triangle Park, NC: U.S. Environmental
  7             Protection Agency, Environmental Sciences Research Laboratory; pp. 89-96; report no. EPA-600/3-77-001a.
  8             Available from: NTIS, Springfield, VA; PB-264232.
  9       Colome, S. D.; Wilson, A. L.; Tian, Y. (1994) California residential indoor air quality study. Volume 2. Carbon
10             monoxide and air exchange rate: an univariate and multivariate analysis.  Chicago, IL: Gas Research Institute;
11             report no. GRI-93/0224.3.
12       Cooper, O. R.; Moody, J. L.  (2000) Meteorological controls on ozone at an elevated eastern United States regional
13             background monitoring site. J. Geophys. Res. [Atmos.] 105: 6855-6869.
14       Cressie, N. (1997) Ozone exposure and population density  in Harris County, Texas [comment]. J. Am. Stat. Assoc.
15             92:411-413.
16       Cristofanelli, P.; Bonasoni, P.; Collins, W.; Feichter, J.; Forster, C.; James, P.; Kentarchos, A.; Kubik, P. W.;
17             Land, C.; Meloen, J.;  Roelofs, G. J.; Siegmund, P.; Sprenger, M.; Schnabel, C.; Stohl, A.; Tobler, L.;
18             Tositti, L.; Trickl,  T.;  Zanis, P. (2003)  Stratosphere-troposphere exchange: a model and method evaluation. J.
19             Geophys. Res. 108(D12): 10.1029/2002JD002600.
20       Delfino, R. J.; Coate, B. D.;  Zeiger, R. S.; Seltzer, J. M.; Street, D. H.; Koutrakis, P. (1996)  Daily asthma severity in
21             relation to personal ozone exposure and outdoor fungal spores. Am. J. Respir. Crit. CareMed. 154: 633-641.
22       Dickerson, R. R. (1985) Reactive nitrogen compounds in the Arctic. J. Geophys. Res. 90: 10,739-10,743.
23       Dickerson, R. R.; Huffman, G. J.; Luke, W. T.; Nunnermacker, L. J.; Pickering, K. E.; Leslie, A. C. D.; Lindsey,
24             C. G.; Slinn, W. G. N.; Kelly, T. J.; Damn, P. H.; Delany, A. C.; Greenberg, J. P.; Zimmerman, P. R.;
25             Boatman, J. F.; Ray, J. D.; Stedman, D. H. (1987) Thunderstorms: an important mechanism in the transport of
26             air pollutants. Science (Washington, DC) 235: 460-465.
27       Douglass, A. R.; Schoeberl, M.  R.; Rood, R. B.; Pawson, S. (2003) Evaluation of transport in the lower tropical
28             stratosphere in a global chemistry and transport model. J. Geophys. Res.  [Atmos.] 108(D9):
29             10.1029/2002JD002696.
30       Drakou, G.; Zerefos, C.; Ziomas, I. (1995) A preliminary study on the relationship between  outdoor and indoor air
31             pollution levels. Fresenius' Environ. Bull.  4: 689-694.
32       Drakou, G.; Zerefos, C.; Ziomas, I.; Voyatzaki, M. (1998) Measurements and numerical simulations of indoor O3
33             and NO^ in two different  cases. Atmos. Environ. 32: 595-610.
34       Evans, G.; Finkelstein, P.; Martin, B.; Possiel, N.; Graves, M. (1983) Ozone measurements from a network of
35             remote sites. J. AirPollut. Control Assoc.  33: 291-296.
36       Evans, E. G.; Rhodes, R.  C.; Mitchell, W. J.; Puzak, J. C. (1985) Summary of precision and accuracy assessments
37             for the state and local air monitoring networks: 1982. Research Triangle Park, NC: U.S. Environmental
3 8             Protection Agency, Environmental Monitoring Systems Laboratory; report no. EPA-600/4-85-031. Available
39             from: NTIS, Springfield,  VA; PB85-208171.
40       Faith, R.; Sheshinski, R. (1979)  Misspecification of trend in spatial random-function interpolation with application
41             to oxidant mapping. Philadelphia, PA:  Siam Institute for Mathematics and Society; technical report no. 28.
42       Fan, Z.; Lioy, P.; Weschler,  C.;  Fiedler, N.; Kipen, H.; Zhang, J. (2003) Ozone-initiated reactions with mixtures of
43             volatile organic compounds under simulated indoor conditions. Environ.  Sci.  Technol. 37: 1811-1821.
44       Federal Register. (1986) Guidelines for estimating exposures. F. R. (September 24) 51: 34,042-34,054.
45       Pick, J.; Pommer, L.; Nilsson, C.; Andersson, B.  (2003) Effect of OH radicals, relative humidity, and time on the
46             composition of the products formed in the ozonolysis of a-pinene. Atmos. Environ. 37: 4087-4096.
47       Pick, J.; Nilsson, C.; Andersson, B. (2004) Formation of oxidation products in a ventilation  system. Atmos. Environ.
48             38: 5895-5899.
49       Fiore, A. M.; Jacob, D. J.; Bey,  L; Yantosca, R. M.; Field, B. D.; Fusco, A. C.; Wilkinson, J. G. (2002a) Background
50             ozone over the United States in summer: origin, trend, and contribution to pollution episodes. J. Geophys.
51             Res. (Atmos.) 107(D15):  10.1029/2001JD000982.
52       Fiore, A. M.; Jacob, D. J.; Field, B. D.; Streets, D. G.; Fernandes, S. D.; Jang, C. (2002b) Linking ozone pollution
53             and climate change: the case for controlling methane. Geophys. Res. Lett. 29(19): 10.1029/2002GL015601.
54       Fiore, A.; Jacob, D. J.; Liu, H.; Yantosca, R. M.; Fairlie, T. D.; Li, Q. (2003a) Variability in surface ozone
55             background over the United States: implications for  air quality policy. J.  Geophys. Res. (Atmos.)
56             108(024): 10.1029/2003 JD003855.
         August 2005                                  AX3-224      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Fiore, A. M; Jacob, D. I; Mathur, R.; Martin, R. V. (2003b) Application of empirical orthogonal functions to
  2             evaluate ozone simulations with regional and global models. J. Geophys. Res. (Atmos.) 108(D14):
  3             10.1029/2002JD003151.
  4       Fitz-Simons, T.; McCluney, L.; Rizzo, M. (2005) Analysis of 2004 ozone data for the ozone NAAQS review
  5             [memorandum to Dr. Joseph Pinto]. Research Triangle Park, NC: U.S. Environmental Protection Agency;
  6             August 22.
  7       Fraigneau, Y. C.; Gonzalez, M.; Coppalle, A. (1995) Dispersion and chemical reaction of a pollutant near a
  8             motorway. Sci. Total Environ. 169: 83-91.
  9       Freijer, J. L; Bloemen, H. J. T. (2000) Modeling relationships between indoor and outdoor air quality. J.  Air Waste
10             Manage. Assoc. 50: 292-300.
11       Freijer, J. L; Bloemen, H. J. T.; de Loos, S.; Marra, M.; Rombout, P. J. A.; Steentjes, G. M.; Van Veen, M. P. (1998)
12             Modelling exposure of the Dutch population to air pollution. J. Haz. Mat. 61: 107-114.
13       Fujita, E. M.; Stockwell, W. R.; Campbell, D. E.; Keislar, R. E.; Lawson, D. R. (2003) Evolution of the magnitude
14             and spatial extent of the weekend ozone effect in California's South Coast Air Basin, 1981-2000. J. Air Waste
15             Manage. Assoc. 53:802-815.
16       Fusco, A. C.; Logan, J. A. (2003)  Analysis of 1970-1995 trends in tropospheric ozone at Northern Hemisphere
17             midlatitudes with the GEOS-CHEM model. J. Geophys. Res. (Atmos.)  108: 10.1029/2002JD002742.
18       Garratt, J. R. (1992) The atmospheric boundary layer. Cambridge, United Kingdom: Cambridge University Press.
19             (Houghton, J. T.; Rycroft, M. J.; Dessler, A. J., eds. Cambridge atmospheric and space science series).
20       Georgopoulos, P. G.; Lioy, P. J. (1994) Conceptual and theoretical aspects of human exposure and dose assessment.
21             J. Exposure Anal. Environ. Epidemiol. 4: 253-285.
22       Georgopoulos, P. G.; Arunachalam,  S.;  Wang, S. (1997a) Alternative metrics for assessing the relative effectiveness
23             of NOx and VOC emission reductions in controlling ground-level ozone. J. Air Waste Manage. Assoc.
24             47:838-850.
25       Georgopoulos, P. G.; Walia, A.; Roy, A.; Lioy, P. J. (1997b) Integrated exposure and dose modeling and analysis
26             system. 1. Formulation and testing of microenvironmental and pharmacokinetic components. Environ. Sci.
27             Technol. 31: 17-27.
28       Georgopoulos, P.; Yonone-Lioy, M. J.;  Opeikum, R. E.; Lioy, P.  (2002a) Environmental dynamics and human
29             exposure to copper. Volume I: environmental dynamics and human exposure issues. New York, NY:
30             International Copper Association.
31       Georgopoulos, P.; Wang, S. W.; Vyas, V. M.; Lioy, P. J.; Tan, H. C. (2002b) Environmental dynamics and human
32             exposure to copper. Volume 2: framework and data sources for assessing human exposure to copper in the
33             United States. New York, NY: International Copper Association.
34       Georgopoulos, P. G.; Wang, S.-W.; Vyas, V M.; Sun, Q.; Burke, J.; Vedantham, R.; McCurdy, T.; Ozkaynak, H.
3 5             (2005) A source-to-dose assessment of population exposures to fine PM and ozone in Philadelphia, PA,
36             during a summer 1999 episode. J. Exposure Anal. Environ. Epidemiol.  15: 10.1038/sj.jea.7500422.
37       Gettelman, A.; Holton, J. R.; Rosenlof, K. H. (1997) Mass fluxes of O3,  CH4, N2O, and CF2C12 in the lower
38             stratosphere calculated from observational data. J. Geophys. Res. [Atmos.] 102: 19,149-19,159.
39       Geyh, A. S.; Wolfson, J. M.; Koutrakis, P.; Mulik, J. D.; Avol, E. L. (1997) Development and evaluation of a small
40             active ozone sampler. Environ. Sci. Technol.  31: 2326-2330.
41       Geyh, A. S.; Roberts, P.  T.; Lurmann, F. W.; Schoell, B. M.; Avol, E. L. (1999) Initial field evaluation of the
42             Harvard active ozone sampler for personal ozone monitoring. J. Exposure Anal. Environ. Epidemiol.
43             9: 143-149.
44       Geyh, A. S.; Xue, J.; Ozkaynak, H.;  Spengler, J. D. (2000) The Harvard Southern California chronic  ozone exposure
45             study: assessing ozone exposure  of grade-school-age children in two southern California communities.
46             Environ. Health Perspect. 108: 265-270.
47       Glen, G.; Lakkadi, Y.; Tippett, J. A.; del Valle-Torres, M. (1997) Development of NERL/CHAD: the National
48             Exposure Research Laboratory consolidated human activity database. Research Triangle Park, NC: U.S.
49             Environmental Protection Agency, Office of Research and Development; contract no. 68-D5-0049.
50       Gold, D. R.; Allen, G.; Damokosh, A.; Serrano, P.; Hayes, C.; Castillejos, M. (1996) Comparison of outdoor and
51             classroom ozone exposures for school children in Mexico City. J. Air Waste Manage. Assoc. 46: 335-342.
52       Goldan, P. D.; Trainer, M.; Kuster, W. C.; Parrish, D. D.; Carpenter, J.; Roberts, J. M.; Yee, J. E.; Fehsenfeld, F. C.
53             (1995) Measurements of hydrocarbons, oxygenated hydrocarbons, carbon monoxide, and nitrogen oxides in
54             an urban basin in  Colorado: implications for emission inventories. J.  Geophys. Res. [Atmos.]
55             100:22,771-22,783.
56       Goldstein, B. R.; Tardiff, G.; Hoffnagle, G.; Kester, R. (1992) Valdez health study: summary report. Anchorage,
57             AK: Alyeska Pipeline Service Company.


         August 2005                                 AX3-225      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Grivet, C. D. (1980) Modeling and analysis of air quality data [dissertation]. Palo Alto, CA: Stanford University.
  2             Available from: University Microfilms International; Ann Arbor, MI; publication no. 80-16,824.
  3       Grosjean, D. (2003) Ambient PAN and PPN in southern California from 1960 to the SCOS97-NARSTO. Atmos.
  4             Environ. 37(suppl. 2): S221-S238.
  5       Grosjean, E.; Grosjean, D.; Fraser, M. P.; Cass, G. R. (1996) Air quality model evaluation data for organics: 3.
  6             Peroxyacetyl nitrate and peroxypropionyl nitrate in Los Angeles air Environ. Sci. Technol. 30: 2704-2714.
  7       Grosjean, E.; Grosjean, D.; Woodhouse, L. F. (2001) Peroxyacetyl nitrate and peroxypropionyl nitrate during SCOS
  8             97-NARSTO. Environ. Sci. Technol. 35: 4007-4014.
  9       Guttorp, P.; Meiring, W.; Sampson, P. D. (1997) Ozone exposure and population density in Harris County, Texas
10             [comment]. J. Am. Stat. Assoc. 92: 405-408.
11       Hartwell, T. D.; Clayton, C. A.; Ritchie, R. M.; Whitmore, R. W.; Zelon, H. S.; Jones, S. M.; Whitehurst, D. A.
12             (1984) Study of carbon monoxide exposure of residents of Washington, DC and Denver, Colorado. Research
13             Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and Development,
14             Environmental Monitoring Systems Laboratory; report no. EPA-600/4-84-031.
15       Hayes, S. R. (1989) Estimating the effect of being indoors on total personal exposure to outdoor air pollution.
16             JAPCA 39: 1453-1461.
17       Hayes, S. R. (1991) Use of an indoor air quality model (IAQM) to estimate indoor ozone levels. J. Air Waste
18             Manage. Assoc. 41: 161-170.
19       Hazucha, M. J.; Folinsbee, L. J.; Seal, E., Jr. (1992) Effects of steady-state and variable ozone concentration profiles
20             on pulmonary function. Am. Rev. Respir. Dis. 146: 1487-1493.
21       Heikes, B. G.; Kok, G. L.; Walega, J. G.; Lazrus, A. L. (1987) H2O2, O3 and SO2 measurements in the lower
22             troposphere over the eastern United States during fall. J. Geophys. Res. 92: 915-931.
23       Helaleh, M. I. H.; Ngudiwaluyo, S.; Korenaga, T.; Tanaka, K. (2002) Development of passive sampler technique for
24             ozone monitoring. Estimation of indoor and outdoor ozone concentration. Talanta 58: 649-659.
25       Herring, J. A.; Jaffe, D. A.; Beine, H. J.; Madronich, S.; Blake, D. R. (1997) High-latitude springtime
26             photochemistry. Part II: sensitivity studies of ozone production. J. Atmos. Chem. 27: 155-178.
27       Heuss, J. M.; Kahlbaum, D. F.; Wolff, G. T. (2003) Weekday/weekend ozone differences: what can we learn from
28             them? J. Air Waste Manage. Assoc. 53: 772-788.
29       Hisham, M. W. M.; Grosjean, D. (1991) Sulfur dioxide, hydrogen sulfide, total reduced sulfur, chlorinated
30             hydrocarbons and photochemical oxidants in southern California museums. Atmos. Environ. Part
31             25:1497-1505.
32       Holland, D. M.; Principe, P. P.; Vorburger, L. (1999) Rural ozone: trends and exceedances at CASTNet sites.
33             Environ. Sci. Technol. 33:43-48.
34       Holton, J. R.; Haynes, P. H.; Mclntyre, M. E.; Douglass, A. R.; Rood, R. B.; Pfister, L. (1995)
35             Stratosphere-troposphere exchange. Rev. Geophys. 33: 403-439.Honrath, R. E.; Hamlin, A. J.; Merrill, J. T.
36             (1996) Transport of ozone precursors from the Arctic troposphere to the North Atlantic region. J. Geophys.
37             Res. 101:29,335-29,351.
38       Honrath, R. E.; Hamlin, A. J.; Merrill, J. T. (1996) Transport of ozone precursors from the Arctic troposphere to the
39             North Atlantic region. J. Geophys. Res. 101: 29,335-29,351.
40       Horowitz, L. W.; Liang, J.; Gardner, G. M.; Jacob, D. J . (1998) Export of reactive nitrogen from North America
41             during summertime: sensitivity to hydrocarbon chemistry. J. Geophys. Res. [Atmos.] 103(D11):
42             13451-13476.
43       Horowitz, L. W.; Walters, S.; Mauserall, D. L.; Emmons, L. K.; Rasch, P. J.; Granier,  C.; Tie, X.; Lamarque, J.-F.;
44             Schultz, M. G.; Tyndall, G. S.; Orlando, J. J.; Brasseur, G. P. (2003) A global simulation of tropospheric
45             ozone and related tracers: description and evaluation of MOZART, version 2. J. Geophys. Res. [Atmos.]
46             108(024): 10.1029/2002JD002853.
47       Horvath, L.; Nagy, Z.; Weidinger, T.; Artz, R.; Luke, W. T.; Valigura, R.; Pinto, J. P.; Womack, J. (1995)
48             Measurement effluxes of trace gases (O3, NOX, SO2, CO2, HNO3), paniculate sulfate and nitrate, water
49             vapour over short vegetation by gradient and eddy correlation techniques in Hungary. EGS XX.  General
50             Assembly; April; Hamburg, Germany. Ann. Geophys. 13(suppl.  2): C490.
51       Horvath, L.; Pinto, J.; Weidinger, T. (2003) Estimate of the  dry deposition of atmospheric nitrogen and sulfur
52             species to spruce forest. Idojaras (Q. J. Hung. Meteorol. Serv.) 107: 249-255.
53       Howard-Reed, C.; Wallace, L. A.; Ott, W. R. (2002) The effect of opening windows on air change rates in two
54             homes. J.  Air Waste Manage Assoc. 52:  147-159.
55       Hudman, R. C.; Jacob, D. J.; Cooper, O. R.; Evans, M. J.; Heald, C. L.; Park, R. J.; Fehsenfeld, F.; Flocke, F.;
56             Holloway, J.; Hubler, G.; Kita, K.; Koike, M.; Kondo, Y.; Neuman, A.; Nowak, J.; Oltmans, S.;  Parrish, D.;
         August 2005                                 AX3-226      DRAFT-DO NOT QUOTE OR CITE

-------
  1             Roberts, J. M; Ryerson, T. (2004) Ozone production in transpacific Asian pollution plumes and implications
  2             for ozone air quality in California. J. Geophys. Res. (Atmos.) 109(D23): 10.1029/2004JD004974.
  3       ICF Consulting; ManTech Environmental Technology, Inc. (2003) Total Risk Integrated Methodology
  4             TRIM.Expo(Inhalation) user's document. Volume I: air pollutants exposure model (APEX, version 3) user's
  5             guide. Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Air Quality Planning
  6             and Standards. Available: http://www.epa.gov/ttn/fera/data/apex322/apexusersguidevoli4-24-03.pdf [7
  7             September, 2004].
  8       linuma, Y.; Boge, O.; Gnauk, T.; Herrmann, H. (2004) Aerosol-chamber study of the a-pinene/O3 reaction:
  9             influence of particle acidity on aerosol yields and products. Atmos. Environ. 38: 761-773.
10       Intergovernmental Panel on Climate Change (IPCC). (2001) Climate change 2001: the scientific basis. Contribution
11             of working group I to the third assessment report of the Intergovernmental Panel on Climate Change.
12             Cambridge, United Kingdom: Cambridge University Press.
13       Isukapalli, S. S.; Purushothaman, V.; Georgopoulos, P. G. (1999) Mechanistic modeling of the interrelationships
14             between indoor/outdoor air quality and human exposure in a GIS framework. Presented at: 92nd Air & Waste
15             Management Association Annual Meeting; June; St. Louis, MO. Pittsburgh, PA: Air & Waste Management
16             Association.
17       Ito, K.; De Leon, S. F.; Lippmann, M. (2005) Associations between ozone and daily mortality, analysis and
18             meta-analysis. Epidemiology 16: 446-457.
19       Jaffe, D. A.; Honrath, R. E.; Herring, J. A.; Li, S.-M.; Kahl, J. D. (1991) Measurements of nitrogen oxides at
20             Barrow, Alaska during spring: evidence for regional and northern hemispheric sources of pollution.
21             J. Geophys. Res. [Atmos.] 96: 7395-7405.
22       Jaffe, D.; Price, H.; Parrish, D.; Goldstein, A.; Harris, J. (2003) Increasing background ozone during spring on the
23             west coast of North America. Geophys. Res. Lett. 30: 10.1029/2003GL017024.
24       Jakobi, G.; Fabian, P. (1997) Indoor/outdoor concentrations of ozone and peroxyacetyl nitrate (PAN). Int. J.
25             Biometeorol. 40: 162-165.
26       James, P.; Stohl, A.; Forster, C.; Eckhardt, S.; Seibert, P.; Frank, A. (2003a) A 15-year climatology of
27             stratosphere-troposphere exchange with a Lagrangian particle dispersion model: 1. Methodology and
28             validation. J. Geophys. Res.  108(D12): 10.1029/2002JD002637.
29       James, P.; Stohl, A.; Forster, C.; Eckhardt, S.; Seibert, P.; Frank, A. (2003b) A 15-year climatology of
30             stratosphere-troposphere exchange with a Lagrangian particle dispersion model. 2. Mean climate and seasonal
31             variability. J. Geophys. Res.  (Atmos.) 108(D12): 10.1029/2002JD002639.
32       Johnson,  T. (1984) A study of personal exposure  to carbon monoxide in Denver, Colorado. Research Triangle Park,
33             NC: U.S. Environmental Protection Agency, Environmental Monitoring Systems Laboratory; report no.
34             EPA-600/4-84-014.
3 5       Johnson,  T. A. (1989) Human activity patterns in Cincinnati, Ohio. Palo Alto, CA: Electric Power Research
36             Institute; report no. EPRIEN-6204.
37       Johnson,  T. R. (1995) Recent advances in the estimation of population exposure to mobile source pollutants.
38             J. Exposure Anal. Environ. Epidemiol. 5: 551-571.
39       Johnson,  T. (1997) A pilot study in  Los Angeles to measure personal ozone exposures during scripted activities.
40             Washington, DC: American Petroleum Institute, Health and Environmental Sciences Department;
41             API publication no. DR 218.
42       Johnson,  T. (2001) A guide to selected algorithms, distributions, and databases used in exposure models developed
43             by the Office of Air Quality Planning and Standards (DRAFT). Research Triangle Park, NC: U.S.
44             Environmental Protection Agency, CERM report TR01.
45       Johnson,  T. (2003) A guide to selected algorithms, distributions, and databases used in exposure models developed
46             by the Office of Air Quality Planning and Standards. Research Triangle Park,  NC: U.S. Environmental
47             Protection Agency, Office of Research and Development; EPA grant no. CR827033. Available:
48             http://www.epa.gov/ttn/fera/data/human/report052202.pdf [9 April, 2004].
49       Johnson,  T.; Long, T. (2004) Determining the frequency of open windows in residences: a pilot study in Durham,
50             North Carolina during varying temperature conditions. J. Exposure Anal. Environ. Epidemiol.:
51             10.1038/sj.jea.7500409.
         August 2005                                 AX3-227      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Johnson, T. R.; Paul, R. A.; Capel, J. E.; McCurdy, T. (1990) Estimation of ozone exposure in Houston using a
  2             probabilistic version of NEM. Presented at: 83rd annual meeting and exhibition of the Air & Waste
  3             Management Association; June; Pittsburgh, PA. Pittsburgh, PA: Air & Waste Management Association;
  4             paper no. 90-150.1.
  5       Johnson, T. R.; Wijnberg, L.; Capel, J. E.; Vostal, J. J.  (1992) The use of activity diary data to estimate the
  6             probability of exposure to air pollution. In: Berglund, R. L., ed. Tropospheric ozone and the environment II:
  7             effects, modeling and control [papers from an international specialty conference]; November; Atlanta, GA.
  8             Pittsburgh, PA: Air & Waste Management Association; pp. 713-724. (A&WMA transactions series no.  20).
  9       Johnson, T.; Capel, J.; McCoy, M. (1996a) Estimation of ozone exposures experienced by urban residents using a
10             probabilistic version of NEM and 1990 population data. Research Triangle Park, NC: U.S. Environmental
11             Protection Agency, Office of Air Quality Planning and Standards; contract no. 68-DO-0062.
12       Johnson, T.; Capel, J.; McCoy, M.; Mozier, J. W. (1996b) Estimation of ozone exposures experienced by outdoor
13             workers in nine urban areas using a probabilistic version of NEM. Research Triangle Park, NC: U.S.
14             Environmental Protection Agency, Office of Air Quality Planning and Standards; contract no. 63-D-30094,
15             work assignment nos. 0-1 and 1-4.
16       Johnson, T.; Capel, J.; Mozier, J.; McCoy, M. (1996c)  Estimation of ozone exposures experienced by outdoor
17             children in nine urban areas using a probabilistic version of NEM. Research Triangle Park, NC: U.S.
18             Environmental Protection Agency, Office of Air Quality Planning and Standards; contract no. 63-D-30094.
19       Johnson, T.; Mozier, J.; Capel, J. (1997) Supplement to "Estimation of ozone exposures experienced by outdoor
20             children in nine urban areas using a probabilistic version of NEM (April 1996)". Research Triangle Park, NC:
21             U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards.
22       Johnson, T.; Myers, J.; Kelly, T.; Wisbith, A.; Ollison, W. (2004) A pilot study using scripted ventilation conditions
23             to identify key factors affecting indoor pollutant concentrations and air exchange rate in a residence. J.
24             Exposure Anal. Environ. Epidemiol. 14: 1-22.
25       Junge, C. E. (1963) Air chemistry and radioactivity. New York, NY: Academic Press. (Van Mieghem, J.; Hales,
26             A. L., eds. International geophysics series: v. 4).
27       Kelly, N. A.; Wolff, G. T.; Ferman, M. A. (1984) Sources and sinks of ozone in rural areas. Atmos. Environ.
28             18: 1251-1266.
29       Kleinman, L. I. (1991)  Seasonal dependence of boundary-layer peroxide concentration - the low and high NOX
30             regimes. J. Geophys. Res. (Atmos.) 96(D11): 20721-20733.
31       Klene, J. G.; Clausen, P. A.; Weschler, C. J.; Wolkoff, P. (2001) Determination of ozone removal rates by selected
32             building products using the FLEC emission cell. Environ.  Sci. Technol. 35: 2548-2553.
33       Klepeis, N. E.; Tsang, A. M.; Behar, J. V.  (1996) Analysis of the national human activity pattern survey (NHAPS)
34             respondents from a standpoint of exposure assessment. Washington, DC: U.S. Environmental Protection
3 5             Agency, Office of Research and Development; report no. EPA/600/R-96/074.
36       Klepeis, N. E.; Nelson, W. C.; Ott. W. R.; Robinson, J. P. Tsang, A. M.; Switzer, P.; Behar, J. V.; Hern, S. C.;
37             Engelmann, W. H. (2001) The National Human Activity Pattern Survey (NHAPS): a resource for assessing
3 8             exposure to environmental pollutants. J. Exposure Anal. Environ. Epidemiol.  11: 231-252.
39       Kley, D.; Volz, A.; Mulheims, F. (1988) Ozone measurements in historic perspective. In: Isaksen, I. S. A., ed.
40             Tropospheric ozone - regional and global scale interactions: proceedings of the NATO advanced workshop on
41             regional and global ozone interaction and its environmental consequences; June 1987; Lillehammer, Norway.
42             Dordrecht, The Netherlands: D. Reidel Publishing Company; pp. 63-72. [NATO Advanced Science Institutes
43             (ASI) studies: series C, mathematical and physical sciences v. 227].
44       Knudsen, H. P.; Lefohn, A. S. (1988) The use of geostatistics to characterize regional ozone exposures. In: Heck,
45             W. W.; Taylor, O. C.; Tingey, D. T., eds. Assessment of crop loss from air pollutants: proceedings of an
46             international conference; October 1987; Raleigh, NC. Essex, United Kingdom: Elsevier Science Publishers,
47             Ltd.; pp. 91-105.
48       Korc, M. E. (1996) A socioeconomic assessment of human exposure to ozone in the South Coast Air Basin of
49             California. J. Air Waste Manage. Assoc. 46: 547-557.
50       Kotchenruther, R. A.; Jaffe, D. A.; Beine, H. J.;  Anderson, T. L.; Bottenheim, J. W.; Harris, J. M.; Blake, D. R.;
51             Schmitt, R.  (2001a) Observations of ozone and related species in the northeast Pacific during the PHOBEA
52             campaigns:  2. Airborne observations. J. Geophys. Res. (Atmos.) 106: 7463-7483.
53       Kotchenruther, R. A.; Jaffe, D. A.; Jaegle, L. (200Ib) Ozone photochemistry and the role of peroxyacetyl nitrate in
54             the springtime northeastern Pacific troposphere: results from the Photochemical Ozone Budget of the Eastern
55             North Pacific Atmosphere (PHOBEA) campaign. J. Geophys. Res. (Atmos.) 106: 28,731-28,742.
         August 2005                                  AX3-228       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Koutrakis, P.; Sioutas, C.; Ferguson, S. T.; Wolfson, J. M; Mulik, J. D.; Burton, R. M. (1993) Development and
  2             evaluation of a glass honeycomb denuder/filter pack system to collect atmospheric gases and particles.
  3             Environ. Sci. Technol. 27: 2497-2501.
  4       Krzyzanowski, M. (1997) Methods for assessing the extent of exposure and effects of air pollution. Occup. Environ.
  5             Med. 54: 145-151.
  6       Ktinzli, N.; Lurmann, F.; Segal, M.; Ngo, L.; Balmes, J.; Tager, I. B. (1996) Reliability of lifetime residential history
  7             and activity measures as elements of cumulative ambient ozone exposure assessment. J. Exposure Anal.
  8             Environ. Epidemiol. 6: 289-310.
  9       Ktinzli, N.; Kelly, T.; Balmes, J.; Tager, I. B. (1997) Reproducibility of retrospective assessment of outdoor
10             time-activity patterns as an individual determinant of long-term ambient ozone exposure. Int. J. Epidemiol.
11             26:1258-1271.
12       Lagus Applied Technology, Inc. (1995) Air change rates in non-residential buildings in California. Sacramento, CA:
13             California Energy Commission; contract no. 400-91-034; July.
14       Lavery, T. F.; Rogers,  C. M.; Howell,  H. K.; Burnett, M. C.; Wanta, C. A. (2002)  Clean air status trends network
15             (CASTNet) 2000 annual report. Available: http://www.epa.gov/castnet/library/annualOO.html (9 September
16             2003).
17       Lee, E. H.; Hogsett, W. E. (1999) Role of concentrations and time of day in developing ozone exposure indices for a
18             secondary standard. J. Air Waste Manage. Assoc. 49:  669-681.
19       Lee, K.; Vallarino, J.; Dumyahn, T.  Ozkaynak, H.; Spengler, J. (1999) Ozone decay rates in residences. J. Air Waste
20             Manage. Assoc. 49: 1238-1244.
21       Lee, M.; Heikes, B. G.; O'Sullivan, D. W. (2000) Hydrogen peroxide and organic  hydroperoxide in the troposphere.
22             Atmos. Environ. 34: 3475-3494.
23       Lee, K.; Xue, J.; Geyh, A. S.; Ozkaynak, H.; Leaderer, B. P.; Weschler, C. J.; Spengler, J. D. (2002) Nitrous acid,
24             nitrogen dioxide, and ozone concentrations in residential environments. Environ. Health Perspect.
25             110: 145-150.
26       Lee, K.; Parkhurst, W. J.; Xue, J.; Ozkaynak, H.; Neuberg, D.; Spengler, J. D. (2004) Outdoor/indoor/personal ozone
27             exposures of children in Nashville, Tennessee. J. Air Waste Manage. Assoc. 54: 352-359.
28       Lefohn, A. S. (1992) The characterization of ambient ozone exposures. In: Lefohn, A. S., ed. Surface level ozone
29             exposures and their effects on vegetation. Chelsea, MI: Lewis Publishers, Inc.; pp. 31-92.
30       Lefohn, A. S.; Benedict, H. M. (1983) The potential for the interaction of acidic precipitation and ozone pollutant
31             doses affecting agricultural crops. Presented at: 76th annual meeting of the  Air Pollution Control Association;
32             June; Atlanta, GA. Pittsburgh, PA: Air Pollution Control Association; paper no. 83-2.2.
33       Lefohn, A. S.; Jones, C. K. (1986) The characterization of ozone and sulfur dioxide air quality data for assessing
34             possible vegetation effects. J. AirPollut. Control Assoc. 36: 1123-1129.
3 5       Lefohn, A. S.; Mohnen, V. A.  (1986) The characterization of ozone, sulfur dioxide, and nitrogen dioxide for selected
36             monitoring sites in the Federal Republic of Germany.  J. Air Pollut. Control Assoc. 36: 1329-1337.
37       Lefohn, A. S.; Runeckles, V. C. (1987) Establishing standards to protect vegetation - ozone exposure/dose
38             considerations. Atmos. Environ. 21: 561-568.
39       Lefohn, A. S.; Tingey, D. T. (1984)  The co-occurrence of potentially phytotoxic concentrations of various gaseous
40             air pollutants. Atmos. Environ.  18: 2521-2526.
41       Lefohn, A. S.; Davis, C. E.; Jones, C. K.;  Tingey, D. T.; Hogsett, W. E. (1987) Co-occurrence patterns of gaseous air
42             pollutant pairs at different minimum concentrations in the United States. Atmos. Environ. 21: 2435-2444.
43       Lefohn, A. S.; Knudsen, H. P.; McEvoy, L. R., Jr. (1988) The use of kriging to estimate monthly ozone exposure
44             parameters for the southeastern United States. Environ. Pollut. 53: 27-42.
45       Lefohn, A. S.; Krupa, S. V.; Winstanley, D. (1990a) Surface ozone exposures measured at clean locations around the
46             world. Environ. Pollut. 63: 189-224.
47       Lefohn, A. S.; Shadwick, D. S.; Mohnen,  V. A. (1990b) The characterization of ozone concentrations at a select set
48             of high-elevation sites in the eastern United States. Environ. Pollut. 67: 147-178.
49       Lefohn, A. S.; Benkovitz, C. M.; Tanner,  R. L.; Smith, L. A.; Shadwick, D. S. (1991) Air quality measurements and
50             characterizations for terrestrial effects research. In: Irving, P. M., ed. Acidic deposition: state of science and
51             technology, volume I, emissions, atmospheric processes, and deposition. Washington, DC: The U.S. National
52             Acid Precipitation Assessment Program. (State of science and technology report no. 7).
53       Lefohn, A. S.; Shadwick, D. S.; Ziman, S. D. (1998) The difficult challenge of attaining EPA's new ozone standard.
54             Environ. Sci. Technol. 32: 276A-282A.
55       Lefohn, A. S.; Oltmans, S. J.; Dann, T.; Singh, H. B. (2001) Present-day variability of background ozone in the
56             lower troposphere. J. Geophys.  Res. [Atmos.] 106: 9945-9958.
         August 2005                                 AX3-229       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Lehman, J.; Swinton, K.; Bortnick, S.; Hamilton, C.; Baldridge, E.; Ender, B.; Cox, B. (2004) Spatio-temporal
  2             characterization of tropospheric ozone across the eastern United States. Atmos. Environ. 38: 4357-4369.
  3       Lelieveld, J.; Dentener, F. J. (2000) What controls tropospheric ozone? J. Geophys. Res. [Atmos.] 105: 3531-3551.
  4       Leovic, K. W.; Sheldon, L. S.; Whitaker, D. A.; Hetes, R. G.; Calcagni, J. A.; Baskir, J. N. (1996) Measurement of
  5             indoor air emissions from dry-process photocopy machines. J. Air Waste Manage. Assoc. 46: 821-829.
  6       Leovic, K. W.; Whitaker, D.; Brockmann, C.; Bayer, C. W.;  Sojka, P. E. (1998) Evaluation of low-emitting products
  7             for use in the indoor environment. Presented at: 91st annual meeting of the Air & Waste Management
  8             Association; June; San Diego, CA. Pittsburgh, PA: Air & Waste Management Association; paper no.
  9             98-TP48.02 (A292).
10       Levinson, D. H., ed. (2005) State of the climate in 2004. Bull. Am. Meteorol. Soc. 86: S1-S86.
11       Levinson, D. H.; Waple, A. M., eds. (2004) State of the climate in 2003. Bull. Am. Meteorol.  Soc. 85: S1-S72.
12       Li, Q.; Jacob, D. J.; Logan, J. A.; Bey, I.; Yantosca, R. M.; Liu, H.; Martin, R. V.; Fiore, A. M.; Field, B. D.;
13             Duncan,B.N. (2001) A tropospheric ozone maximum over the Middle East. Geophys. Res. Lett.
14             28: 3235-3238.
15       Li, Q.; Jacob, D. J.; Fairlie, T. D.; Liu, H.; Martin, R. V.; Yantosca, R. M. (2002) Stratospheric versus pollution
16             influences on ozone at Bermuda: reconciling past analyses. J. Geophys. Res. 107(D22):
17             10.1029/2002JD002138.
18       Liard, R.; Zureik, M.; Le Moullec, Y.; Soussan, D.; Glorian,  M.; Grimfeld, A.; Neukirch, F. (1999) Use of personal
19             passive samplers for measurement of NO2, NO, and O3 levels in panel studies. Environ. Res. 81: 339-348.
20       Lin, C.-Y. C.; Jacob, D. J.; Fiore, A. M.  (2001) Trends in exceedances of the ozone air quality standard in the
21             continental United States,  1980-1998. Atmos. Environ. 35: 3217-3228.
22       Linn, W. S.; Shamoo, D. A.; Hackney, J. D. (1992)  Documentation of activity patterns in 'high-risk' groups exposed
23             to ozone in the Los Angeles area. In: Tropospheric ozone and the environment II: effects, modeling and
24             control:  papers from an Air & Waste Management Association international specialty conference; November,
25             1991; Atlanta, GA. Pittsburgh, PA: Air & Waste Management Association; pp. 701-712. (A&WMA
26             publication TR-20).
27       Linn, W. S.; Spier, C. E.; Hackney, J. D. (1993) Activity patterns in ozone-exposed construction workers. J. Occup.
28             Med. Toxicol. 2: 1-14.
29       Linn, W. S.; Shamoo, D. A.; Anderson, K. R.; Peng, R.-C.; Avol, E. L.; Hackney, J. D.; Gong, H., Jr. (1996)
30             Short-term air pollution exposures and responses in Los  Angeles area schoolchildren. J. Exposure Anal.
31             Environ. Epidemiol. 6: 449-472.
32       Lioy, P. J. (1990) Assessing total human exposure to contaminants: a multidisciplinary approach. Environ. Sci.
33             Technol. 24: 938-945.
34       Lisac, I.; Grubisic, V. (1991) An analysis of surface ozone data measured at the end of the  19th century in Zagreb,
35             Yugoslavia. Atmos. Environ. Part A 25:  481-486.
36       Liu, D.-L.; Nazaroff, W. W. (2001) Modeling pollutant penetration across building envelopes. Atmos. Environ.
37             35:4451-4462.
38       Liu, L.-J. S.; Koutrakis, P.; Leech, J.; Broder, I. (1995) Assessment of ozone exposures in the greater metropolitan
39             Toronto area. J. Air Waste Manage. Assoc. 45: 223-234.
40       Liu, L.-J. S.; Delfino, R.; Koutrakis, P. (1997) Ozone exposure assessment in a southern California community.
41             Environ. HealthPerspect. 105: 58-65.
42       Liu, H.; Jacob,  D. J.; Chan, L. Y.; Oltmans, S. J.; Bey, I.; Yantosca, R. M.; Harris, J. M.; Duncan, B. N.; Martin,
43             R. V. (2002) Sources of tropospheric ozone along the  Asian Pacific Rim: an analysis of ozonesonde
44             observations. J. Geophys. Res. [Atmos.]  107(D21):  10.1029/2001JD002005.
45       Liu, H.; Jacob,  D. J.; Bey, I.; Yantosca, R. M.; Duncan, B. N.; Sachse, G. W. (2003) Transport pathways for Asian
46             pollution outflow over the Pacific: interannual and seasonal variations. J. Geophys. Res. (Atmos.)
47             108(020): 10.1029/2002JD003102.
48       Liu, H.; Jacob,  D. J.; Dibb, J. E.; Fiore, A. M.; Yantosca, R. M. (2004) Constraints on the sources of tropospheric
49             ozone from 210Pb-7Be-O3 correlations. J. Geophys. Res. (Atmos.) 109(D7): 10.1029/2003JD003988.
50       Logan, J. A. (1989) Ozone in rural areas of the  United States. J. Geophys. Res.  [Atmos.] 94: 8511-8532.
51       Logan, J. A. (1999) An analysis of ozonesonde data for the troposphere: recommendations for testing 3-D models
52             and development of a gridded climatology for tropospheric ozone. J. Geophys. Res.  104: 16,115-16,149.
53       Ludwig, F. L.;  Reiter, E.; Shelar,  E.; Johnson, W. B. (1977) The relation of oxidant levels to precursor emissions and
54             meteorological features: v. I, analysis and findings. Research Triangle Park, NC: U.S. Environmental
55             Protection Agency, Office of Air Quality Planning and Standards; report no. EPA-450/3-77-022a. Available
56             from: NTIS, Springfield, VA; PB-275 001.
         August 2005                                 AX3-230      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Lurmann, F.; Colome, S. (1991) An assessment of current and future human exposure to ozone in the South Coast
  2             Air Basin. Santa Rosa, CA: Sonoma Technology, Inc.; report no. STI-90070-1022-FR; 1991 AQMP revision,
  3             technical report V-H.
  4       Marr, L. C.; Morrison, G. C.; Nazaroff, W. W.; Harley, R. A. (1998) Reducing the risk of accidental death due to
  5             vehicle-related carbon monoxide poisoning. J. Air Waste Manage. Assoc. 48: 899-906.
  6       Martin, R. V.; Jacob, D. J.; Yantosca, R. M.; Chin, M.; Ginoux, P. (2003) Global and regional decreases in
  7             tropospheric oxidants from photochemical effects of aerosols. J. Geophys. Res. [Atmos.] 108(D3):
  8             10.1029/2002JD002622.
  9       Matheron, G. (1963) Principles of geostatistics. Econ. Geol. 58: 1246-1266.
10       McCurdy, T. R. (1994) Human exposure to ambient ozone. In: McKee, D. J., ed. Tropospheric ozone: human health
11             and agricultural impacts. Boca Raton, FL: Lewis Publishers; pp. 85-127.
12       McCurdy, T. (1997a) Human activities that may lead to high inhaled intake doses in children aged 6-13. Environ.
13             Toxicol. Pharmacol. 4: 251-260.
14       McCurdy, T. (1997b) Modeling the dose profile in human exposure assessments: ozone as an example. Rev.
15             Toxicol. 1: 3-23.
16       McCurdy, T. (2000) Conceptual basis for multi-route intake dose modeling using an energy expenditure approach.
17             J. Exposure Anal. Environ. Epidemiol. 10: 86-97.
18       McCurdy, T.; Glen, G.; Smith, L.; Lakkadi, Y. (2000) The National Exposure Research Laboratory's Consolidated
19             Human Activity Database. J. Exposure Anal. Environ. Epidemiol. 10: 566-578.
20       Meagher, J. F.; Lee, N. T.; Valente, R. J.; Parkhurst, W. J. (1987) Rural ozone in the southeastern United States.
21             Atmos. Environ. 21: 605-615.
22       Meloen, J.; Siegmund, P.; Van Velthoven, P.; Kelder, H.; Sprenger, M.; Wernli, H.; Kentarchos, A.; Roelofs, G.;
23             Feichter, J.; Land, C.; Forster, C.; James, P.; Stohl, A.; Collins, W.; Cristofanelli, P. (2003)
24             Stratosphere-troposphere exchange: a model and method intercomparison. J. Geophys. Res. 108(D12):
25             10.1029/2002JD002274.
26       Meng, Q. Y.; Turpin, B. J.; Korn, L.; Weisel, C. P.; Morandi, M.; Colome, S.; Zhang, J.; Stock, T.; Spektor, D.;
27             Winer, A.; Zhang, L.; Lee, J. H.; Giovanetti, R.; Cui, W.; Kwon, J.; Alimokhtari, S.;  Shendell, D.; Jones, J.;
28             Farrar, C.; Maberti, S. (2005) Influence of ambient (outdoor) sources on residential indoor and personal
29             PM2.5 concentrations: analyses of RIOPA data. J. Exposure Anal. Environ. Epidemiol. 15: 17-28.
30       Mohnen, V. A. (1989) Mountain cloud chemistry project: wet, dry and cloud water deposition. Research Triangle
31             Park, NC: U.S. Environmental Protection Agency, Atmospheric Research and Exposure Assessment
32             Laboratory; report no. EPA/600/3-89/009. Available from: NTIS, Springfield, VA; PB89-148597.
33       Monks, P. S. (2000) A review of the observations and origins of the spring ozone maximum. Atmos. Environ.
34             34: 3545-3561.
35       Morrison, G. C.; Nazaroff, W. W. (2000) The rate of ozone uptake on carpets: experimental studies. Environ. Sci.
36             Technol. 34: 4963-4968.
37       Morrison, G. C.; Nazaroff, W. W. (2002) The rate of ozone uptake on carpet: mathematical modeling. Atmos.
38             Environ. 36: 1749-1756.
39       Morrison, G. C.; Nazaroff, W. W.; Cano-Ruiz, J. A.; Hodgson, A. T.; Modera, M. P. (1998) Indoor air quality
40             impacts  of ventilation ducts: ozone removal and emissions of volatile organic compounds. J. Air Waste
41             Manage. Assoc. 48: 941-952.
42       Mueller, F. X.; Loeb, L.; Mapes, W. H. (1973) Decomposition rates of ozone in living areas. Environ. Sci. Technol.
43             7:342-346.
44       Murphy, D. M.; Fahey, D. W. (1994) An estimate of the flux of stratospheric reactive nitrogen and ozone into the
45             troposphere. J. Geophys. Res. (Atmos.) 99: 5325-5332.
46       Murray, D. M.; Burmaster, D. E. (1995) Residential air exchange rates in the United States: empirical and estimated
47             parametric distributions by season and climatic region. Risk Anal. 15: 459-465.
48       Nagda, N. L.; Rector, H. E.; Koontz, M. D. (1987) Guidelines for monitoring indoor air quality. Washington, DC:
49             Hemisphere Publishing Corporation.
50       National Acid Precipitation Assessment Program. (1987) Interim assessment: the causes and effects of acidic
51             deposition, volume I, executive summary. Washington, DC: Office of the Director of Research.
52       Navidi, W.; Thomas, D.; Langholz, B.; Stram, D. (1999) Statistical methods for epidemiologic studies of the health
53             effects of air pollution. Cambridge, MA: Health Effects Institute; research report no. 86.
54       Nazaroff, W. W.; Cass, G. R. (1986) Mathematical modeling of chemically reactive pollutants in indoor air.
55             Environ. Sci. Technol. 20: 924-934.
56       Nazaroff, W. W.; Weschler, C. J. (2004) Cleaning products and air fresheners: exposure to primary and secondary
57             air pollutants.  Atmos.  Environ. 38: 2841-2865.


         August 2005                                AX3-231      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Neidell, M. J. (2004) Air pollution, health, and socio-economic status: the effect of outdoor air quality on childhood
  2             asthma. J. Health Econ. 23: 1209-1236.
  3       Newchurch, M. J.; Ayoub, M. A.; Oltmans, S.; Johnson, B.; Schmidlin, F. J. (2003) Vertical distribution of ozone at
  4             four sites in the United States. J. Geophys. Res. 108(D1): 10.1029/2002JD002059.
  5       Niu, J.; Tung, T. C. W.; Burnett, J. (2001) Ozone emission rate testing and ranking method using environmental
  6             chamber. Atmos. Environ. 35: 2143-2151.
  7       Nejgaard, J. K.; Christensen, K. B.; Wolkoff, P. (2005) The effect on human eye blink frequency of exposure to
  8             limonene oxidation products and methacrolein. Toxicol. Lett. 156: 241-251.
  9       Northeast States for Coordinated Air Use Management (NESCAUM). (2002) Indoor/outdoor school air monitoring
10             project. Boston, MA. Available: http://www.nescaum.org/pdf/schoolmonitoring.pdf [29 October, 2003].
11       Odman, M. T.;  Ingram, C. I. (1996) Multiscale air quality simulation platform (MAQSIP): source code
12             documentation and validation. Research Triangle Park, NC: MCNC; technical report ENV-96TR002-vl.O.
13             Available:  http://www.ce.gatech.edu/~todman/maqsip.pdf (9 September 2003).
14       Olsen, M. A.; Douglass, A. R.; Schoeberl, M. R. (2002) Estimating downward cross-tropopause flux using column
15             ozone and potential vorticity. J. Geophys. Res. 107(D22):  10.1029/2001JD002041.
16       Oltmans, S. J.; Lefohn, A. S.; Scheel, H. E.; Harris, J. M.; Levy, H., II; Galbally, I. E.; Brunke, E.-G.; Meyer, C. P.;
17             Lathrop, J. A.; Johnson, B. J.; Shadwick, D. S.; Cuevas, E.; Schmidlin, F. J.; Tarasick, D. W.; Claude, H.;
18             Kerr, J. B.; Uchino, O.; Mohnen, V. (1998) Trends of ozone in the troposphere. Geophys. Res. Lett. 25:
19             139-142.
20       O'Neill, M. S.; Ramirez-Aguilar, M.; Meneses-Gonzalez, F.; Hernandez-Avila, M.; Geyh,  A. S.; Sienra-Monge, J.  J.;
21             Romieu, I. (2003) Ozone exposure among Mexico City outdoor workers. J. Air Waste Manage. Assoc.
22             53: 339-346.
23       Orzechowska, G.  E.; Paulson, S. E. (2002) Production of OH radicals from the reactions of C4-C6 internal alkenes
24             and styrenes with ozone in the gas phase. Atmos. Environ. 36: 571-581.
25       Ott, W. R.  (1982) Concepts of human exposure to air pollution. Environ. Int. 7: 179-196.
26       Ott, W. R.  (1985) Total human exposure: an emerging science focuses on humans as receptors of environmental
27             pollution. Environ. Sci. Technol. 19: 880-886.
28       Ott, W.; Langan, L.; Switzer, P. (1992) A time series model for cigarette smoking activity  patterns: model validation
29             for carbon monoxide and respirable particles in a chamber and an automobile. In: Measuring, understanding,
30             and predicting exposures in the 21st century: proceedings of the conference; November 1991; Atlanta, GA.  J.
31             Exposure Anal. Environ. Epidemiol. 2(suppl.  2): 175-200.
32       Park, J.-H.; Spengler, J. D.; Yoon, D.-W.; Dumyahn, T.; Lee, K.; Ozkaynak, H. (1998) Measurement of air
3 3             exchange rate of stationary vehicles and estimation of in-vehicle exposure. J. Exposure Anal. Environ.
34             Epidemiol. 8: 65-78.
35       Park, R. J.; Pickering, K. E.; Allen, D. J.; Stenchikov, G. L.; Fox-Rabinovitz, M. S. (2004) Global simulation of
36             tropospheric ozone using the University of Maryland Chemical Transport Model (UMD-CTM):  1. Model
37             description and evaluation. J. Geophys. Res. [Atmos.] 109(D9): 10.1029/2003JD004266.
38       Pavelin, E. G.; Johnson, C. E.; Rughooputh, S.;  Toumi, R. (1999) Evaluation of preindustrial surface ozone
39             measurements made using Schonbein's method. Atmos. Environ.  33: 919-929.
40       Penkett, S. A. (1983) PAN in the natural atmosphere. Nature (London) 302: 293-294.
41       Penkett, S. A.; Brice, K. A. (1986) The spring maximum in photo-oxidants in the Northern Hemisphere troposphere.
42             Nature (London) 319: 655-657.
43       Pinto, J. P.; Lefohn, A. S.; Shadwick, D. S. (2004) Spatial variability of PM2.5 in urban areas in the United States.
44             J. Air Waste Manage. Assoc. 54: 440-449.
45       Pippin, M. R.; Bertman, S.; Thornberry, T.; Town, M.; Carroll, M. A.; Sillman, S. (2001) Seasonal variation of
46             PAN, PPN, and O3 at the upper Midwest PROPHET site. J. Geophys. Res.  (Atmos.) 106: 24,451-24,463.
47       Poupard, O.; Blondeau, P.; lordache, V.; Allard, F. (2005) Statistical analysis of parameters influencing the
48             relationship between outdoor and indoor air quality in schools. Atmos. Environ. 39: 2071-2080.
49       Prados, A.  I. (2000) The export of ozone and other atmospheric trace species from North America: observations and
50             numerical simulations [dissertation]. College Park, MD: The University of Maryland.
         August 2005                                 AX3-232       DRAFT-DO NOT QUOTE OR CITE

-------
1       Prinz, B.; Krause, G. H. M. (1989) State of scientific discussion about the causes of the novel forest decline in the
2            Federal Republic of Germany and surrounding countries. In: Bucher, J. B.; Bucher-Wallin, I., eds. Air
3            pollution and forest decline, v. I: Proceedings of the 14th international meeting for specialists in air pollution
4            effects on forest ecosystems; October 1988; Interlaken, Switzerland. Birmensdorf, Switzerland:
5            Eidgenossische Anstalt fur das forstliche Versuchswesen (EAFV); pp. 27-34.
        August 2005                                 AX3-233       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Prinz, B.; Krause, G. H. M; Jung, K.-D. (1985) Untersuchungen der LIS Essen zur Problematik der Waldschaden
  2             [Investigations of the LIS Essen into the problem of forest damage]. In: Kortzfleisch, G., ed. Waldschaden:
  3             Theorie und Praxis auf der Suche nach Antworten [Forest damage: theory and practice in the search for
  4             answers]. Mtinchen, Germany: R. Oldenbourg Verlag; pp. 143-194.
  5       Proyou, A. G.; Ziomas, I. C.; Stathopoulos, A. (1998) Application of a three-layer photochemical box model in an
  6             Athens street canyon. J. Air Waste Manage. Assoc. 48: 427-433.
  7       Pun, B. K.; Seigneur, C.; White, W. (2003) Day-of-week behavior of atmospheric ozone in three U.S. cities. J. Air
  8             Waste Manage. Assoc. 53: 789-801.
  9       Purushothaman, V.; Georgopoulos, P. G. (1997) Computational tools to aid the estimation and visualization of
10             potential human exposure to ozone. In: Proceedings of the Air & Waste Management Association Specialty
11             Conference on Computing in Environmental Resource Management; December, 1996; Research Triangle
12             Park, NC. Pittsburgh, PA: A&WMA; VIP-68.
13       Purushothaman, V.; Georgopoulos, P. G. (1999a) Evaluation of regional emissions control strategies for ozone
14             utilizing population exposure metrics: a new GIS-based approach applied to the July 1995 episode over the
15             OTAG domain. Piscataway, NJ: Environmental and Occupational Health Sciences Institute, Ozone Research
16             Center; ORC technical report, ORC-TR99-02.
17       Purushothaman, V.; Georgopoulos, P. G. (1999b) Integrating photochemical modeling, geostatistical techniques and
18             geographical information systems for ozone exposure assessment. Piscataway, NJ: Environmental and
19             Occupational Health Sciences Institute, Ozone Research Center; ORC technical report, ORC-TR99-01.
20       Reagan, J. (1984) Air quality data interpretation. In: Heck, W. W.; Taylor, O. C.;  Adams, R. M.; Bingham, G. E.;
21             Miller, J. E.; Preston, E. M.; Weinstein, L. H., eds. National Crop Loss Assessment Network (NCLAN) 1982
22             annual report. Corvallis, OR: U.S. Environmental Protection Agency, Corvallis Environmental Research
23             Laboratory; pp. 198-219; report no. EPA-600/4-84-049. Available from: NTIS, Springfield, VA;
24             PB84-169358.
25       Reeves, C. E.; Penkett, S. A.  (2003) Measurements of peroxides and what they tell us. Chem. Rev. 103: 5199-5218.
26       Regener, E. (1941) Ozonschicht und atmospharische turbulenz. Ber. Dtsch. Wetterdienstes US-Zone No. 11: 45-57.
27       Reiss, R.; Ryan, P. B.; Koutrakis, P.; Tibbetts, S. J. (1995) Ozone reactive chemistry on interior latex paint. Environ.
28             Sci. Technol. 29: 1906-1912.
29       Reiter, E. R. (1977) The role of stratospheric import on tropospheric ozone concentrations. In: Dimitriades, B., ed.
30             International conference on photochemical oxidant pollution and its control - proceedings: volume I;
31             September 1976; Raleigh, NC. Research Triangle Park, NC: U.S. Environmental Protection Agency,
32             Environmental Sciences Research Laboratory; pp. 393-410; report  no. EPA-600/3-77-001a. Available from:
33             NTIS, Springfield, VA; PB-264232.
34       Research Triangle Institute. (1990) An investigation of infiltration and indoor air quality. Albany, NY: New York
3 5             State Energy Research and Development Authority; report no. NYERDA 90-11.
36       Riediker, M.; Williams, R.; Devlin, R.; Griggs, T.; Bromberg, P. (2003) Exposure to paniculate matter, volatile
37             organic compounds, and other air pollutants inside patrol cars. Environ. Sci. Technol. 37: 2084-2093.
38       Rifai,  H. S.; Hopkins, L.; Mehdizadeh, F. (2000) Developing human exposure estimates for indoor/outdoor air
39             pollution in Houston. In: Environmental Institute of Houston -1999 Annual Report. Houston, TX:
40             Environmental Institute of Houston.
41       Roberts, J. M.; Williams, J.; Baumann, K.; Buhr, M. P.; Goldan, P. D.; Holloway, J.; Htibler, G.; Kuster, W. C.;
42             McKeen, S. A.; Ryerson, T. B.; Trainer, M.; Williams, E. J.; Fehsenfeld, F. C.; Bertman, S. B.; Nouaime, G.;
43             Seaver, C.; Grodzinsky, G.; Rodgers, M.; Young, V.  L. (1998) Measurements of PAN, PPN,  and MPAN
44             made during the 1994  and 1995 Nashville Intensives  of the Southern Oxidant Study: implications for regional
45             ozone production from biogenic hydrocarbons. J. Geophys. Res. [Atmos.] 103: 22,473-22,490.
46       Roelofs, G. J.; Lelieveld, J.; Van Dorland, R. (1997) A three-dimensional chemistry/general circulation model
47             simulation of anthropogenically derived ozone in the  troposphere and its radiative climate forcing. J.
48             Geophys. Res. [Atmos.] 102: 23,389-23,401.
49       Rohr,  A. C.; Wilkins, C. K.; Clausen, P.  A.; Hammer, M.; Nielsen, G.  D.; Wolkoff, P.; Spengler, J. D. (2002) Upper
50             airway and pulmonary effects of oxidation products of (+)-a-pinene, J-limonene, and isoprene in BALB/c
51             mice. Inhalation Toxicol.  14: 663-684.
52       Rohr,  A. C.; Shore, S. A.; Spengler, J. D. (2003) Repeated exposure to isoprene oxidation products causes enhanced
53             respiratory tract effects in multiple murine strains. Inhalation Toxicol.  15: 1191-1207.
54       Romieu, I.; Lugo, M. C.; Colome, S.; Garcia A. M.; Avila, M. H.; Geyh, A.; Velasco, S. R.; Rendon, E. P.  (1998)
5 5             Evaluation of indoor ozone concentration and predictors of indoor-outdoor ratio in Mexico City. J. Air Waste
56             Manage. Assoc. 48: 327-335.
         August 2005                                 AX3-234      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Rotman, D. A.; Atherton, C. S.; Bergmann, D. J.; Cameron-Smith, P. I; Chuang, C. C.; Connell, P. S.; Dignon,
  2             J. E.; Franz, A.; Grant, K. E.; Kinnison, D. E.; Molenkamp, C. R.; Proctor, D. D.; Tannahill, J. R. (2004)
  3             IMPACT, the LLNL 3-D global atmospheric chemical transport model for the combined troposphere and
  4             stratosphere: model description and analysis of ozone and other trace gases. J. Geophys. Res. [Atmos.]
  5             109(D4): 10.1029/2002JD003155.
  6       Ryerson, T. B.; Trainer, M; Holloway, J. S.; Parrish, D. D.; Huey, L. G.; Sueper, D. T.; Frost, G. J.; Donnelly,
  7             S. G.; Schauffler, S.; Atlas, E. L.; Kuster, W. C.; Goldan, P. D.; Htibler, G.; Meagher, J. F.; Fehsenfeld, F. C.
  8             (2001) Observations of ozone formation in power plant plumes and implications for ozone control strategies.
  9             Science (Washington, DC) 292: 719-723.
10       Sabersky, R. H.; Sinema, D. A.; Shair, F. H. (1973) Concentrations, decay rates, and removal of ozone and their
11             relation to establishing clean indoor air. Environ. Sci. Technol. 7: 347-353.
12       Salmon, L. G.; Cass, G. R.; Bruckman, K.; Haber, J. (2000) Ozone exposure inside museums in the historic central
13             district of Krakow, Poland. Atmos. Environ. 34: 3823-3832.
14       Sarnat, J. A.; Koutrakis, P.; Suh, H. H. (2000) Assessing the relationship between personal paniculate and gaseous
15             exposures of senior citizens living in Baltimore, MD. J. Air Waste Manage. Assoc. 50: 1184-1198.
16       Sarnat, J. A.; Schwartz, J.; Catalano, P. J.; Suh, H. H. (2001) Gaseous pollutants in paniculate matter epidemiology:
17             confounders or surrogates? Environ. Health Perspect. 109: 1053-1061.
18       Sarwar, M.; Corsi, R.; Kimura, Y.; Allen, D.; Weschler, C. (2001) Hydroxyl radicals in indoor environments. In:
19             Proceedings of the Air & Waste Management Association's 94th Annual Conference & Exhibition; June;
20             Orlando, FL. Pittsburgh, PA: Air & Waste Management Association.
21       Sarwar, G.; Corsi, R.;  Kumura, Y.; Allen, D.; Weschler, C. J. (2002) Hydroxyl radicals in indoor environments.
22             Atmos. Environ. 36: 3973-3988.
23       Scheeren, B. A.; Adema,  E. H. (1996) Monitoring ambient ozone with a passive measurement technique method,
24             field results and strategy. Water Air Soil Pollut. 91: 335-350.
25       Schoeberl, M. R.; Douglass, A. R.; Zhu, Z.; Pawson, S. (2003) A comparison of the lower stratospheric age spectra
26             derived from a  general circulation model and two data assimilation systems. J. Geophys. Res. (Atmos.)
27             108(D3): 10.1029/2002JD002652.
28       Seo, K.-H.; Bowman,  K. P. (2002) Lagrangian estimate of global stratosphere-troposphere mass exchange.
29             J. Geophys. Res. (Atmos.)  107(D21): 10.1029/2002JD002441.
30       Shair, F. H. (1981) Relating indoor pollutant concentrations of ozone and sulfur dioxide to those outside: economic
31             reduction of indoor ozone through selective filtration of the make-up air. ASHRAE Trans. 87(pt. 1): 116-139.
32       Shair, F. H.; Heitner, K. L. (1974) Theoretical model for relating indoor pollutant concentrations to those outside.
33             Environ. Sci. Technol. 8:444-451.
34       Shamoo, D. A.; Johnson,  T. R.; Trim, S. C.; Little, D. E.; Linn, W. S.; Hackney, J. D. (1991) Activity patterns in a
3 5             panel of outdoor workers exposed to oxidant pollution. J. Exposure Anal. Environ. Epidemiol. 1: 423-438.
36       Shamoo, D. A.; Linn,  W. S.; Peng, R.-C; Solomon, J. C.; Webb, T. L.; Hackney, J. D.; Gong, H., Jr. (1994)
3 7             Time-activity patterns and diurnal variation of respiratory status in a panel of asthmatics: implications for
3 8             short-term air pollution effects. J. Exposure Anal. Environ. Epidemiol. 4: 133-148.
39       Sheppard, L.; Slaughter, J. C.; Schildcrout, J.; Liu, L.-J. S.; Lumley, T. (2005a) Exposure and measurement
40             contributions to estimates of acute air pollution effects. J. Exposure Anal. Environ. Epidemiol. 15: 366-376.
41       Sheppard, L. (2005b) Acute air pollution effects: consequences of exposure distribution and measurements. J.
42             Toxicol. Environ. Health Part A 68: 1127-1135.
43       Sheppard, L.; Slaughter, J. C.; Schildcrout, J.; Liu, L.-J. S.; Lumley, T. (2005) Exposure and measurement
44             contributions to estimates of acute air pollution effects. J. Exposure Anal. Environ. Epidemiol. 15: 366-376.
45       Sherman, M. H.; Matson, N. E. (1997) Residential ventilation and energy characteristics. Presented at: 1997
46             American Society  of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) winter meeting;
47             February; Philadelphia, PA. Atlanta, GA: American Society of Heating, Refrigerating and Air-Conditioning
48             Engineers, Inc.  Report no.  CONF-9702141. ASHRAE Trans. 103: 717-730.
49       Shindell, D. T.; Faluvegi, G.; Bell, N. (2003) Preindustrial-to-present-day radiative forcing by tropospheric ozone
50             from improved simulations with the GISS chemistry-climate GCM. Atmos. Chem. Phys. 3:  1675-1702.
51       Sillman, S.; Logan, J.  A.; Wofsy, S. C. (1990) The sensitivity of ozone to nitrogen oxides and hydrocarbons in
52             regional ozone  episodes. J. Geophys. Res. [Atmos.] 95: 1837-1851.
53       Singh, H. B.; Ludwig, F. L.; Johnson, W. B. (1977) Ozone in clean remote atmospheres: concentrations and
54             variabilities. New York, NY: Coordinating Research Council, Inc.; report no. CRC-APRAC CAPA-15-76.
55             Available from: NTIS, Springfield, VA; PB-272290.
56       Singh, H. B.; Ludwig, F. L.; Johnson, W. B. (1978) Tropospheric ozone: concentrations and variabilities in clean
57             remote atmospheres. Atmos. Environ.  12: 2185-2196.


         August 2005                                 AX3-235      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Smith, L. A.; Lefohn, A. S. (1991) Co-occurence of ozone and wet deposited hydrogen ion in the United States.
  2             Atmos. Environ. Part A 25:  2707-2716.
  3       Solberg, S.; Krognes, T.; Stordal, F.; Hov, 0.; Beine, H. I; Jaffe, D. A.; Clemitshaw, K. C.; Penkett, S. A. (1997)
  4             Reactive nitrogen compounds at Spitsbergen in the Norwegian Arctic. J. Atmos. Chem. 28: 209-225.
  5       Serensen, D. N.; Weschler, C. J. (2002) Modeling-gas phase reactions in indoor environments using computational
  6             fluid dynamics. Atmos. Environ. 36: 9-18.
  7       Spier, C. E.; Little, D. E.; Trim, S.  C.; Johnson, T. R.; Linn, W. S.; Hackney, J. D. (1992) Activity patterns in
  8             elementary and high school students exposed to oxidant pollution. J. Exposure Anal. Environ. Epidemiol.
  9             2:277-293.
10       Sprenger, M; Wernli, H. (2003) A northern hemispheric climatology of cross-tropopause exchange for the ERA
11             15 time period (1979-1993). J. Geophys. Res. 108(D12): 10.1029/2002JD002636.
12       Sprenger, M.; Maspoli, M. C.; Wernli, H. (2003) Tropopause folds and cross-tropopause exchange: a global
13             investigation based upon ECMWF analyses for the time period March 2000 to February 2001. J. Geophys.
14             Res. (Atmos.) 108(D12): 10.1029/2002JD002587.
15       Stasiuk, W. N., Jr.; Coffey, P. E. (1974) Rural and urban ozone relationships in New York State. J. Air Pollut.
16             Control Assoc. 24: 564-568.
17       Steiber, R. S. (1995) Ozone generators in indoor air settings. Research Triangle Park, NC: U.S. Environmental
18             Protection Agency, National Risk Management Research Laboratory; report no. EPA-600/R-95-154.
19             Available from: NTIS, Springfield, VA; PB96-100201.
20       Stein, M. L.; Fang, D. (1997) Ozone exposure and population density in Harris County, Texas [comment]. J. Am.
21             Stat. Assoc. 92:408-411.
22       Stohl, A. (2001) A 1-year Lagrangian "climatology" of airstreams in the Northern Hemisphere troposphere and
23             lowermost stratosphere. J. Geophys. Res. (Atmos.) 106: 7263-7279.
24       Tan, W. W.; Geller, M. A.; Pawson,  S.; Da Silva, A. (2004) A case  study of excessive subtropical transport in the
25             stratosphere of a data assimilation system. J. Geophys. Res. (Atmos.) 109(D11102): 10.1029/2003 JD004057.
26       Taylor, G. E., Jr.; Hanson, P. J. (1992) Forest trees and tropospheric ozone: role of canopy deposition and leaf
27             uptake in developing exposure-response relationships. Agric. Ecosyst. Environ. 42: 255-273.
28       Taylor, G. E., Jr.; Norby, R. J. (1985) The significance of elevated levels of ozone on natural ecosystems  of North
29             America. In: Lee, S. D., ed.  Evaluation of the scientific basis for ozone/oxidants standards: proceedings of an
30             APCA international specialty  conference; November 1984; Houston, TX. Pittsburgh, PA: Air Pollution
31             Control Association; pp. 152-175. (APCA transactions: TR-4).
32       Taylor, G. E., Jr.; Ross-Todd, B. M.; Allen, E.; Conklin, P.; Edmonds, R.; Joranger, E.; Miller, E.; Ragsdale, L.;
33             Shepard, J.; Silsbee, D.; Swank, W. (1992) Patterns of tropospheric  ozone in forested landscapes of the
34             Integrated Forest Study. In:  Johnson, D. W.; Lindberg, S. E., eds. Atmospheric  deposition and forest nutrient
35             cycling: a synthesis of the integrated forest study. New York, NY: Springer-Verlag; pp. 50-71. (Billings, W.
36             D.; Golley, F.; Lange, O. L.; Olson, J. S.; Remmert, H., eds. Ecological studies: analysis and synthesis: v. 91).
37       Thompson, C. R.; Hensel, E. G.; Kats, G. (1973) Outdoor-indoor levels of six air pollutants. J. Air Pollut. Control
38             Assoc. 23:881-886.
39       Trainer, M.; Parrish, D. D.; Buhr, M. P.; Norton, R. B.; Fehsenfeld, F. C.; Anlauf, K. G.; Bottenheim, J. W.; Tang,
40             Y. Z.; Wiebe, H. A.; Roberts,  J. M.; Tanner, R. L.; Newman, L.; Bowersox, V.  C.; Meagher, J. F.;  Olszyna,
41             K. J.; Rodgers, M. O.; Wang,  T.; Berresheim, H.; Demerjian, K. L.; Roychowdhury, U. K. (1993) Correlation
42             of ozone with NOy in photochemically aged air. J. Geophys. Res. [Atmos.] 98: 2917-2925.
43       Turk, B. H.; Grimsrud, D. T.; Brown, J. T.; Geisling-Sobotka, K.  L.; Harrison, J.; Prill, R. J. (1989) Commercial
44             building ventilation rates and particle concentrations. ASHRAE Trans. 95(part  1): 422-433.
45       U.S. Environmental Protection Agency. (1986) Air quality criteria for ozone and other photochemical oxidants.
46             Research Triangle Park, NC: Office of Health and Environmental Assessment, Environmental Criteria and
47             Assessment Office; report nos. EPA-600/8-84-020aF-eF. 5v. Available from: NTIS, Springfield, VA;
48             PB87-142949.
49       U.S. Environmental Protection Agency. (1992) National air quality and emissions trends report, 1991. Research
50             Triangle Park, NC: U.S. Environmental Protection Agency,Office of Air Quality Planning and Standards;
51             report no. EPA/450-R-92-001. Available  from: NTIS, Springfield, VA; PB93-143998.
         August 2005                                 AX3-236      DRAFT-DO NOT QUOTE OR CITE

-------
  1       U.S. Environmental Protection Agency. (1996a) Air quality criteria for ozone and related photochemical oxidants.
  2             Research Triangle Park, NC: Office of Research and Development; report nos. EPA/600/AP-93/004aF-cF. 3v.
  3             Available from: NTIS, Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available:
  4             http://cfpub2.epa.gov/ncea/.
  5       U.S. Environmental Protection Agency. (1996b) Air quality criteria for paniculate matter. Research Triangle Park,
  6             NC: National Center for Environmental Assessment-RTF Office; report nos. EPA/600/P-95/001aF-cF. 3v.
  7       U.S. Environmental Protection Agency. (1996c) Review of national ambient air quality standards for ozone:
  8             assessment of scientific and technical information. OAQPS staff paper. Research Triangle Park, NC: Office
  9             of Air Quality Planning and Standards; report no.  EPA/452/R-96/007. Available from: NTIS, Springfield,
10             VA; PB96-203435. Available: http://www.epa.gov/ttn/naaqs/standards/ozone/s_o3_pr_sp.html
11             (9 September 2003).
12       U.S. Environmental Protection Agency. (1997) Responses to  significant comments on the 1996 proposed rule on the
13             national ambient air quality standards for paniculate matter (December 13, 1996; 61 FR 65638). Research
14             Triangle Park, NC: National Center for Environmental Assessment; docket number A-95-54; July.
15             ECAO-CD-92-0671; STD-081.
16       U.S. Environmental Protection Agency. (1998) Guideline on  ozone monitoring site selection. Research Triangle
17             Park, NC: U.S. Environmental Protection Agency, National Center for Environmental Assessment;
18             EPA-454/R-98-002.
19       U.S. Environmental Protection Agency. (2002) Latest findings on national air quality: 2001 status and trends.
20             Research Triangle Park, NC: Office of Air Quality Planning and Standards; report no. EPA-454/K-02-001.
21             Available: http://www.epa.gov/air/aqtrndOI/ (22 July 2003).
22       U.S. Environmental Protection Agency. (2003) Technology Transfer Network: Air Quality System (AQS).
23             Washington, DC: Office of Air and Radiation. Available: http://www.epa.gov/ttn/airs/airsaqs/ [24 August,
24             2005].
25       U.S. Environmental Protection Agency. (2004a) Air quality criteria for paniculate matter. Research Triangle Park,
26             NC: National Center for Environmental Assessment; report no. EPA/600/P-99/002aF-bF. 2v. Available:
27             http://cfpub.epa.gov/ncea/ [9 November, 2004].
28       U.S. Environmental Protection Agency. (2004b) The ozone report: measuring progress through 2003. Cincinnati,
29             OH: U.S. Environmental Protection Agency; report no. EPA454/K-04-001.
30       Vakeva, M; Ha'meri, K.; Kulmala, M; Lahdes, R.; Ruuskanen, J.; Laitinen, T. (1999) Street level versus rooftop
31             concentrations of submicron aerosol particles and gaseous pollutants in an urban street canyon. Atmos.
32             Environ. 33: 1385-1397.
33       Van den Bergh, V.; Vanhees, I.; De Boer, R.; Compernolle, F.; Vinckier, C. (2000) Identification of the oxidation
34             products of the reaction between a-pinene and hydroxyl radicals by gas and high-performance liquid
35             chromatography with mass spectrometric detection. J.  Chromatogr. A 896: 135-148.
36       Van Noije, T. P. C.; Eskes, H. J.; Van Weele, M.; Van Velthoven, P. F. J. (2004) Implications of enhanced
3 7             Brewer-Dobson circulation in European Centre for medium-range weather forecasts reanalysis ERA-40 for
38             the stratosphere-troposphere exchange of ozone in global chemistry transport models. J. Geophys. Res.
39             (Atmos.) 109(019308): 10.1029/2004 JD004586.
40       Venkatachari, P.; Hopke, P. K.; Grover, B. D.; Eatough,  D. J. (2005a) Measurement of particle-bound reactive
41             oxygen species inRubidoux aerosols. J. Atmos. Chem. 50: 49-58.
42       Venkatachari, P.; Hopke, P .K.; Brune, W. H.; Ren, X.; Lesher, R.; Mao, J.; Mitchell, M. (2005b)  Characterization
43             of reactive oxygen species trends in Flushing, New York. Atmos. Environ.: in press.
44       Vingarzan, R. (2004) A review of surface ozone background  levels and trends. Atmos. Environ. 38: 3431-3442.
45       Vong, R. J.; Guttorp, P. (1991) Co-occurence of ozone and acidic cloudwater in high-elevation forests. Environ. Sci.
46             Technol. 25: 1325-1329.
47       VonKuhlmann, R.; Lawrence, M. G.; Crutzen, P. J.; Rasch, P. J. (2003) A model for studies of tropospheric ozone
48             and nonmethane hydrocarbons: model description and ozone results. J. Geophys. Res. [Atmos.]
49             108(09): 10.1029/2002JD002893.
50       Wainman, T.; Zhang, J.; Weschler, C. J.; Lioy, P. J. (2000) Ozone and limonene in indoor air: a source of submicron
51             particle exposure. Environ. Health Perspect. 108:  1139-1145.
52       Wang, Y.; Logan, J. A.;  Jacob, D. J. (1998) Global simulation of tropospheric O3-NOx-hydrocarbon chemistry 2.
53             Model evaluation and global ozone budget. J. Geophys. Res. (Atmos.) 103: 10,727-10,755.
54       Weinstein-Lloyd, J. B.; Lee, J. H.; Damn, P. H.; Nunnermacker, L. J.; Springston, S. R.; Newman, L. (1998)
55             Measurements of peroxides and related species during the 1995 summer intensive of the Southern Oxidants
56             Study in Nashville, Tennessee.  J. Geophys. Res. (Atmos.) 103: 22361-22373.
         August 2005                                 AX3-237      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Wernli, H.; Bourqui, M. (2002) A Lagrangian "1-year climatology" of (deep) cross-tropopause exchange in the
  2             extratropical Northern Hemisphere. J. Geophys. Res. (Atmos.) 107(D2): 10.1029/2001JD000812.
  3       Weschler, C. J. (2000) Ozone in indoor environments: concentration and chemistry. Indoor Air 10: 269-288.
  4       Weschler, C. J. (2004) Chemical reactions among indoor pollutants: what we've learned in the new millennium.
  5             Indoor Air 14(suppl. 7): 184-194.
  6       Weschler, C. J.; Shields, H. C. (1997) Potential reactions among indoor pollutants. Atmos. Environ. 31: 3487-3495.
  7       Weschler, C. J.; Shields, H. C. (1999) Indoor ozone/terpene reactions as a source of indoor particles. Atmos.
  8             Environ. 33:2301-2312.
  9       Weschler, C. J.; Shields, H. C. (2000) The influence of ventilation on reactions among indoor pollutants: modeling
10             and experimental observations. Indoor Air. 10: 92-100.
11       Weschler, C. J.; Shields, H. C. (2003) Experiments probing the influence of air exchange rates on secondary organic
12             aerosols derived from indoor chemistry. Atmos. Environ. 37: 5621-5631.
13       Weschler, C. J.; Shields, H. C.; Naik, D. V. (1989) Indoor ozone exposures. JAPCA 39: 1562-1568.
14       Weschler, C. J.; Hodgson, A. T.; Wooley, J. D. (1992) Indoor chemistry: ozone, volatile organic compounds, and
15             carpets. Environ. Sci. Technol. 26: 2371-2377.
16       Weschler, C. J.; Shields, H. C.; Naik, D. V. (1994) Indoor chemistry involving O3, NO, and NO2 as evidenced by
17             14 months of measurements at a site in southern California. Environ. Sci. Technol.  28: 2120-2132.
18       Wild, O.; Akimoto, H. (2001) Intercontinental transport of ozone and its precursors in a three-dimensional global
19             CTM. J. Geophys. Res.  [Atmos.] 106: 27,729-27,744.
20       Wiley, J. A.; Robinson, J. P.; Piazza, T.; Garrett, K.; Cirksena, K.; Cheng, Y.-T.; Martin, G. (199la) Activity
21             patterns of California residents. Final report. Sacramento, CA:  California Air Resources Board; report no.
22             ARB/R93/487. Available from: NTIS, Springfield, VA.; PB94-108719.
23       Wiley, J. A.; Robinson, J. P.; Cheng, Y.-T.; Piazza, T.; Stork, L.; Pladsen, K. (1991b) Study of children's activity
24             patterns: final report. Sacramento, CA: California Air Resources Board; report no. ARB-R-93/489.
25       Wilkins, C. K.; Clausen, P. A.; Wolkoff, P.; Larsen, S. T.; Hammer, M.; Larsen, K.; Hansen, V.; Nielsen, G. D.
26             (2001) Formation of strong irritants in mixtures of isoprene/ozone and isoprene/ozone/nitrogen dioxide.
27             Environ. Health Perspect. 109: 937-941.
28       Williams, R.; Suggs, J.; Rea, A.; Leovic, K.; Vette, A.; Croghan, C.; Sheldon, L.; Rodes, C.; Thornburg,, J.; Ejire,
29             A.; Herbst, M.; Sanders, W., Jr. (2003a) The Research Triangle Park paniculate matter panel study: PM mass
30             concentration relationships Atmos. Environ. 37: 5349-5363.
31       Williams, R.; Suggs, J.; Rea, A.; Sheldon, L.; Rodes, C.; Thornburg, J. (2003b) The Research Triangle Park
32             paniculate matter panel study: modeling ambient source contribution to personal and residential PM mass
33             concentrations. Atmos. Environ. 37:  5365-5378.
34       Wilson, A. L.; Colome, S. D.; Tian, Y.; Becker, E. W.; Baker, P. E.; Behrens, D. W.; Billick, I. H.; Garrison, C. A.
3 5             (1996) California residential air exchange rates and residence volumes. J. Exposure Anal. Environ.
36             Epidemiol. 6:311-326.
37       Wolff,  G. T.; Lioy, P. J.; Taylor, R. S. (1987) The diurnal variations of ozone at different altitudes on a rural
3 8             mountain in the eastern United States. JAPCA 37: 45-48.
39       Wolff,  G. T.; Dunker, A. M.; Rao, S. T.; Porter, P. S.; Zurbenko, I. G. (2001) Ozone air quality over North America:
40             part I—a review of reported trends. J. Air Waste Manage. Assoc. 51: 273-282.
41       Wolkoff, P.;  Clausen, P. A.; Wilkins, C. K.; Hougaard, K. S.; Nielsen, G. D. (1999) Formation of strong airway
42             irritants in a model mixture of (+)-a-pinene/ozone. Atmos. Environ.  33: 693-698.
43       Wong, S.; Wang, W.-C.; Isaksen, I. S. A.; Berntsen, T. K.; Sundet, J.  K. (2004) A global climate-chemistry model
44             study of present-day tropospheric chemistry and radiative forcing from changes in tropospheric O3 since the
45             preindustrial period. J. Geophys. Res. [Atmos.] 109(D11): 10.1029/2003JD003998.
46       Zeger,  S. L.;  Thomas, D.; Dominici, F.; Samet, J. M.; Schwartz, J.; Dockery, D.; Cohen, A. (2000) Exposure
47             measurement error in time-series studies of air pollution:  concepts and consequences. Environ. Health
48             Perspect. 108: 419-426.
49       Zhou, J.;  Smith, S. (1997) Measurement of ozone concentrations in ambient air using a badge-type passive  monitor.
50             J. Air Waste Manage. Assoc. 47: 697-703.
         August 2005                                  AX3-238       DRAFT-DO NOT QUOTE OR CITE

-------
 i                ANNEX AX4.  DOSIMETRY OF OZONE IN
 2                           THE RESPIRATORY TRACT
 3
 4
 5     AX4.1  INTRODUCTION
 6           This annex serves to provide supporting material for Chapter 4 - Dosimetry, Species,
 7     Homology, Sensitivity, and Animal-to-Human Extrapolation.  It includes tables that summarize
 8     new literature published since the last O3 criteria documents (U.S. Environmental Protection
 9     Agency, 1996). In addition, it provides descriptions of those new findings, in many cases, with
10     more detail than is provided in the chapter.
11           Dosimetry refers to the measurement or estimation of the quantity of or rate at which a
12     chemical and/or it reaction products are absorbed and retained at target sites within the
13     respiratory tract (RT).  The compound most directly responsible for toxic effects may be the
14     inhaled gas O3 or  one of its chemical reaction products. Complete identification of the actual
15     toxic agents and their integration into dosimetry is a complex  issue that has not been resolved.
16     Thus, most dosimetry investigations are  concerned with the dose of the primary inhaled
17     chemical. In this  context, a further confounding aspect can be the units of dose (e.g., mass
18     retained per breath, mass retained per breath per body weight, mass retained per breath per
19     respiratory tract surface area). That is, when comparing dose  between species, what is the
20     relevant measure of dose?  This question has not been answered; units are often dictated by the
21     type of experiment or by a choice made by the investigators. There is also some lack of
22     agreement as to what constitutes "dose." Dahl's seminal paper (1990) classified O3 as a reactive
23     gas and discussed the characterization of dose measurement by parameters including:
24     (1) inhaled O3 concentration; (2) amount of O3 inhaled as determined by minute volume, vapor
25     concentration, and exposure duration; (3) uptake or the amount of O3 retained (i.e., not exhaled);
26     (4) O3 or its active metabolites delivered to target cells or tissues; (5) O3 or its reactive
27     metabolites delivered to target biomolecules or organelles; and (6) O3 or its metabolites
28     participating in the ultimate toxic reactions - the effective dose.  This characterization goes from
29     least complex to greatest, culminating in measurement of the fraction of the inhaled O3 that
30     participates in the effects of cellular perturbation and/or injury. Understanding dosimetry as it
31     relates to O3-induced injury is complex due to the fact that O3 interacts primarily with the

       August 2005                             AX4-1      DRAFT-DO NOT  QUOTE OR CITE

-------
 1      epithelial lining fluid (ELF) which contains surfactant and antioxidants. In the upper airways
 2      ELF is thick and highly protective against oxidant injury. In lower airways ELF is thinner, has
 3      lower levels of antioxidants, and thus, allows more cellular injury. Adding to the complexity is
 4      the fact that O3 can react with molecules in the ELF to create even more reactive metabolites,
 5      which can then diffuse within the lung or be transported out of the lung to generate systemic
 6      effects.  Section 5.3.1 contains further information on the cellular targets of O3 interactions and
 7      antioxidants.
 8           Experimental dosimetry studies in laboratory animals and humans, and theoretical
 9      (dosimetry modeling) studies, have been used to obtain information on O3 dose.  Since the last
10      ozone criteria document (U.S. Environmental Protection Agency, 1996), all new experiments
11      have been carried out in humans to obtain direct measurements of absorbed O3 in the RT, the
12      upper RT (URT) region proximal to the tracheal entrance, and in the tracheobronchial (TB)
13      region; no uptake experiments have been performed using laboratory animals.  Experimentally
14      obtaining dosimetry data is extremely difficult in smaller regions or locations, such as specific
15      airways or the centriacinar region (CAR; junction of conducting airways and gas exchange
16      region), where lesions caused by O3 occur. Nevertheless, experimentation is important for
17      determining dose, making dose comparisons between subpopulations and between  different
18      species, assessing hypotheses and concepts, and validating mathematical models that can be used
19      to predict dose at specific respiratory tract sites and under more general conditions.
20           Theoretical studies are based on the use of mathematical models developed for the
21      purposes of simulating the uptake and distribution of absorbed gases in the tissues and fluids of
22      the RT.  Because the factors affecting the transport and absorption of gases are applicable to all
23      mammals, a model that uses appropriate species or disease-specific anatomical and ventilatory
24      parameters can be used to describe absorption in the  species and in different-sized, aged, or
25      diseased members of the same species.  More importantly, models also may be used to make
26      interspecies and intraspecies dose comparisons, to compare and reconcile data from different
27      experiments, to predict dose in conditions not possible or feasible experimentally, and to better
28      understand the processes involved in toxicity.
29           A review (Miller, 1995) of the factors influencing RT uptake of O3 stated that structure of
30      the RT region, ventilation, and gas transport mechanisms were important.  Additionally, local
31      dose is the critical link between exposure and response.  A more detailed discussion of

        August 2005                             AX4-2       DRAFT-DO NOT QUOTE OR CITE

-------
 1     experimental and theoretical dosimetry studies is available in the 1996 O3 AQCD (Volume III,
 2     Chapter 8, U.S. Environmental Protection Agency, 1996).
 3
 4
 5     AX4.2  EXPERIMENTAL OZONE DOSIMETRY INVESTIGATIONS
 6          There have been some advances in understanding human O3 dosimetry that better enable
 7     quantitative extrapolation from laboratory animal data. The next two sections review the
 8     available new experimental studies on O3 dosimetry, which involve only human subjects and are
 9     all from the same laboratory. Of the studies considered in the following discussion, five
10     involved the use of the bolus response method as a probe to obtain information about the
11     mechanism of O3 uptake in the URT and TB regions.  Of the remaining two investigations, one
12     focused on total uptake by the RT and the other on uptake by the nasal cavities.  Table AX4-1
13     provides a summary of the newer studies.
14
15     AX4.2.1  Bolus-Response Studies
16          The bolus-response method has been used as a probe to explore the effects of physiological
17     and anatomical differences or changes on the uptake of O3 by human beings.
18          Asplund et al. (1996) studied the effects of continuous O3 inhalation on O3 bolus uptake
19     and Rigas et al. (1997) investigated the potential effects of continuous coexposure to O3,
20     nitrogen dioxide (NO2), or sulfur dioxide (SO2) on O3 absorption. In both of these studies,
21     subjects were exposed "continuously" to a gas for 2 h. Every 30 min, breathing at 250 mL/s,
22     a series of bolus test breaths was performed targeted at the lower conducting airways.
23     Differences in bolus-response absorbed  fraction from an established baseline indicated the
24     degree to which the "continuous" gas exposure affected the absorption of O3.  Depending on the
25     gas and concentration, changes in absorbed fraction ranged from -3 to +7 % (see Table AX4-1).
26     Continuous O3 exposure lowered the uptake of O3, whereas NO2 and SO2 increased the uptake
27     of O3.  The investigators concluded that in the tested airways, NO2 and SO2 increased the
28     capacity to absorb O3 because more of the compounds oxidized by O3 were made available.
29     On the other hand, they conjectured that continuous O3 exposure depleted these compounds,
30     thereby reducing O3 uptake.
31

       August 2005                            AX4-3        DRAFT-DO NOT QUOTE OR CITE

-------
                              Table AX4-1. New Experimental Human Studies on Ozone Dosimetry a
•s
S-
to
o
o
u\













^
X
_k
Purpose/Objective

Determine the effect
of continuous O3
inhalation on O3
uptake





Evaluate the
influence of VD on
intersubject variation
of O3 dose.




Subject
Characteristics

8 male,
3 female,
22-3 1 years old,
166-186 cm,
64-93 kg




10 male,
22-30 years old,
163-186 cm,
64-92 kg;
10 female,
22-35 years old,
149-177 cm,
48-81 kg
Region of
Interest

Central
conducting
airways
(70-120 mL
from lips)




Conducting
airways






Breathing
Patterns/Exposure

2 h of continuous exposure at
rest: 0.0, 0.12, and 0.36 ppm
O3. Spontaneous breathing.
Bolus test breaths every
30 minutes using 250 mL/sec
constant flow rate.



Bolus-response test
(VT = 500ml at 250 mL/sec
constant flow rate). Fowler
single-breath N2 washout
method to determine VD.



Results Reference

Averaged over all subjects and the Asplund et al.
4 measurement intervals, the absorbed fraction (1996)
(AF) changed +0.04, -0.005, and -0.03 for the
0, 0. 12, and 0.36 ppm continuous exposures,
respectively. These changes are
approximately +6, -1, and -4 % based on an
average AF value of 0.7 in the range
70 -120 ml.b Both nonzero exposures were
significantly different than the air exposure.
On average, for the same VP , women had a Bush et al.
larger AF than men; women had a smaller VD ( 1 996a)
than men. However, for the same value of
VP/VD, AF for men and women were
indistinguishable. Further analysis indicated
"that previously reported gender differences
may be due to a failure in properly accounting
for tissue surface within the conducting
                                                                         airways .



O
E>
|!TJ
H
tb
\«l/
o
0
H
O
o
H
W
O
Investigate the effect 6 male, 21-29 years
of continuous old, 165 185 cm,
exposure to O3, 60-92 kg;
nitrogen dioxide and 6 female,
sulfur dioxide on O3 19-33 years old,
absorption. 152-173 cm,
48-61 kg








Lower 2 h of continuous exposure
conducting at rest: O3 (0, 0.36 ppm),
airways SO2 (0, 0.36 ppm), or NO2
(70-120 mL (0, 0.36, 0.75 ppm).
from lips) 5-min Bolus test every
30 minutes: VT = 500 ml;
250 mL/sec constant flow
rate.







Averaging over all subjects or by gender, all Rigas et al.
exposures except O3 resulted in an increase of (1997)
AF. Based on an AF reference valueb,
change in AF ranged from -3 to +7 %.
the
Only
the O3 and the NO2 (0.36 ppm) exposures
were significantly different from the air
exposures.

















O

-------
6-
00
S-
o Purpose/Objective
L/l
Compare the
absorption of
chlorine and O3.
Determine how the
physical-chemical
properties of these
compounds affect
their uptake
distribution in
the RT.
Table AX4-1
Subject
Characteristics
5 male, 2 1-26 years
old, 168-198 cm,
64-95 kg; 5 female,
18-28 years old,
162-178 cm,
55-68 kgc




(cont'd). New Experimental Human Studies on Ozone Dosimetrya
Region of
Interest
Conducting
airways.
Nasal and
oral routes






Breathing
Patterns/Exposure
Bolus-response technique;
VT = 500 ml; 3 constant flow
rates: 150, 250, and
1000 mL/sec.







Results
Ozone dose to the URT was sensitive to the
mode of breathing and to the respiratory rate.
With increased airflow rate, O3 retained by the
upper airways decreased from 95 to 50%.
TB region dose ranged from 0 to 35%.
Mass transfer theory indicated that the
diffusion resistance of the tissue phase is
important for O3. The gas phase resistances
were the same for O3 and C12.


Reference
Nodelman
and Ultman
(1999a)c







X
         To determine O3
         uptake relative to
         inhaled O3 dose.
5 male,
5 female,
18-35 years old,
175 ± 13 (SD) cm,
72 ± 13  (SD) kg
Respiratory     Breath-by-breath calculation
tract; oral      of O3 retention based on data
breathing      from fast response analyzers
               for O3 and airflow rates. Oral
               breathing:  0.2 or 0.4 ppm O3
               at VE of approximately
               20 L/min for 60 min or
               40 L/min for 30 min.
The FA for all breaths was 0.86.
Concentration, minute volume, and time have
small but statistically significant effects on AF
when compared to intersubject variability.
The investigators concluded: for a given
subject, constant O3 exposure, a given exercise
level, and time <2 h, inhaled dose is a
reasonable surrogate for the actual uptake
ofO3.  However, the actual doses may vary
considerably among individuals who are
exposed to similar inhaled doses.
Rigas et al.
(2000)
H
6
o
o
H
O
o
H
W
O
O

-------
                                   Table AX4-1 (cont'd).  New Experimental Human Studies on Ozone Dosimetry!
S-
to
o
         Purpose/Objective
                               Subject
                               Characteristics
Region of
Interest
Breathing
Patterns/Exposure
Results
Reference
X
H
6
o
o
H
/O
o
H
W
O
         Study the effect of
         gas flow rate and O3
         concentration on O3
         uptake in the nose.
         Evaluate intersubject
         variability in O3
         uptake; correlate
         differences in
         breathing pattern and
         lung anatomy with O3
         uptake
                               7 male, 3 female,
                               26 ± years,
                               170 ± 11 (SD)cm,
                               75 ± 20  (SD) kg
                               nonsmokers,
                               32 male, 22.9 ± 0.8
                               years old,178±l cm,
                               80.6 ± 2.5kg;
                               28 female,
                               22.4 ± 0.9 years old,
                               166 ± 1 cm,
                               62.1 ±2.2 kg
Nasal          For a given flow rate and
cavities        exposure concentration, the
               subjects inhaled through one
               nostril and exhale through the
               other. For two 1-h sessions,
               a series of 9-12 measurements
               of AF were carried out for
               10 s each: (1) O3 exposure
               concentration = 0.4 ppm; flow
               rates = 3, 5, 8, and 15 L/min.
               (2) O3 exposure = 0.1, 0.2,
               and 0.4 ppm; flow rate =
               15 L/min. (3) O3 exposure
               = 0.4 ppm, flow rate
               = 15 L/min; AF was measured
               every 5 min for 1 h.

Respiratory     Continuous:  1  h exposure to
tract; oral      0.25 ppm, exercising at
breathing      30L/min.
               Bolus: breath-by-breath
               calculation of O3 retention.
               Timing of bolus varied to
               create penetration volumes of
               10 to 250 ml. Peak inhaled
               bolus of ~1 ppm.
                              (1) With the exposure concentration at
                              0.4 ppm O3, AF decreased from 0.80 to
                              0.33 when the flow rate was increased from
                              3 to 15 L/min.  (2) At a flow rate of 15 L/min,
                              the AF changed from 0.36 to 0.32 when the
                              exposure concentration increased from 0.1 to
                              0.4 ppm O3. (3) Statistical analysis indicated
                              that the AF was not associated with the time at
                              which the measurement was taken.
                              Continuous:  Fractional O3 uptake efficiency
                              ranged from 0.70 to 0.98 (mean 0.89 ± 0.06).
                              Inverse correlation between uptake and
                              breathing frequency. Direct correlation
                              between uptake and tidal volume. Intersubject
                              differences in forced respiratory responses not
                              due to differences in O3 uptake.

                              Bolus:  The penetration volume at which 50%
                              of the bolus was taken up was 90.4 ml in
                              females and 107 ml in males.  Distribution of
                              O3 shifts distally as the size of the airway
                              increases.
                                            Santiago
                                            etal. (2001)
                                            Ultman et al.
                                            (2004)
         a See Appendix A for abbreviations and acronyms.
         b Fig. 4, Hu et al. (1994), for the 250 mL/s curve and penetration volume range of 70 - 120 ml; the average AF is approximately 0.7.
         0 Subject characteristics are from Nodelman and Ultman (1999b).
O

-------
 1           Bush et al. (1996a) investigated the effect of lung anatomy and gender on O3 absorption in
 2      the conducting airways during oral breathing using the bolus-response technique. Absorption
 3      was measured using this technique applied to 10 men and 10 women. Anatomy was defined in
 4      terms of forced vital capacity (FVC), total lung capacity (TLC), and dead space (VD).  The
 5      absorbed fraction data were analyzed in terms of a function of penetration volume, airflow rate,
 6      and an "intrinsic mass transfer parameter (Ka)", which was determined for each subject and
 7      found to be highly correlated with VD, but not with height, weight, age, gender, FVC, or TLC.
 8      That is, in all subjects, whether men or women, dosimetry differences could be explained by
 9      differences in VD.  Based on Hu et al. (1994), where absorbed fraction was determined for
10      several flow rates, Bush et al. (1996a) inferred that Ka was proportional to flow rate/VD. The
11      investigators point out that the applicability of their results may be limited because of their
12      assumptions that Ka was independent of location in the RT and that there was no mucous
13      resistance.  They also suggested that the dependence of Ka on flow rate and VD be restricted to
14      flow rates < 1000 mL/s until  studies at higher rates have been performed.
15           With flow rates of 150, 250, and 1000 mL/s, Nodelman and Ultman (1999b) used the
16      bolus-response technique to  compare the uptake distributions of O3 and chlorine gas (C12), and to
17      investigate how their uptakes were affected by their physical and chemical properties. Ozone
18      dose to the URT was found to be sensitive to the mode of breathing and to the airflow rate.  With
19      increased rate, O3 retained by the upper airways decreased from 95 to 50% and TB region dose
20      increased from 0 to 35%. At the highest flow rate only  10% of the O3 reached the pulmonary
21      region. Mass transfer theory indicated that the diffusion resistance of the tissue phase is
22      important for O3.  The gas phase resistances were found to be the same for O3 and C12 as
23      expected. These resistances were inversely related to the volumes of the oral and nasal cavities
24      during oral and nasal breathing, respectively.
25           Ultman et al. (2004) used both bolus and continuous exposures to test the hypotheses that
26      differences in O3 uptake in lungs are responsible for variation in O3-induced changes in lung
27      function parameters and that differences in O3 uptake are due to variations in breathing patterns
28      and lung anatomy. Thirty-two males and 28 female nonsmokers were exposed to bolus
29      penetration volumes ranging from 10 to 250 ml, which was determined by the timing of the
30      bolus injection. The  subjects controlled their breathing to generate a target respired flow of
31      lOOOml/sec.  At this high minute ventilation, there was very little uptake in the upper airway and

        August 2005                             AX4-7       DRAFT-DO NOT QUOTE OR CITE

-------
 1      most of the O3 reached areas where gas exchange takes place.  To quantify intersubject
 2      differences in O3 bolus uptake, they measured the penetration volume at which 50% of the O3
 3      was taken up. Values for penetration volume ranged from 69 to 134 ml and were directly
 4      correlated with the subjects' values for anatomic dead space volume.  A better correlation was
 5      seen when the volume of the upper airways was subtracted.  The penetration volume at which
 6      50% of the bolus was taken up was 90.4 ml in females and 107 ml in males. This significant
 7      difference in uptake suggests to the authors that in females the smaller airways, and associated
 8      larger surface-to-volume ratio, enhance local O3 uptake and cause reduced penetration of O3 into
 9      the distal lung. Thus, these findings indicate that overall O3 uptake is not related to airway size,
10      but that the distribution of O3 shifts distally as the size of the airway in increased.
11
12      General comment on estimating mass transfer coefficients.  Bush et al. (1996b) and Nodelman
13      and Ultman (1999a) used a simple model to analyze their bolus-response data. This model
14      presented by Hu et al. (1992, 1994) assumed steady-state mass transfer by convection (but no
15      dispersion) and the mass transfer of O3 to the walls of a tube of uniform cross-sectional area.
16      These assumptions led to an analytical solution (for the absorbed fraction) which was a function
17      of an "overall mass transfer coefficient," penetration volume, and airflow rate. As the
18      investigators have shown, the model is very useful for statistical analysis and hypothesis testing.
19      Given the absorbed fraction data, the model overall mass transfer  coefficients were estimated for
20      each flow rate. In those bolus-response studies that used this method to analyze data, there was
21      no discussion of the models' "accuracy" in representing mass transfer in the human respiratory
22      tract with respect to omitting dispersion. In addition, the formulation of the gas phase mass
23      transfer coefficient does not take into account that it has a theoretical lower limit greater than
24      zero as the airflow rate goes to zero (Miller et al., 1985; Bush et al., 2001). As a consequence,
25      there is no way to judge the usefulness of the values of the estimated mass transfer coefficients
26      for dosimetry simulations that are based on convection-dispersion equations, or whether or not
27      the simple model's mass transfer coefficients, as well as other parameters derived using these
28      coefficients, are the same as actual physiological parameters.
29
30
        August 2005                              AX4-8       DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX4.2.2   General Uptake Studies
 2           Rigas et al. (2000) performed an experiment to determine the ratio of O3 uptake to the
 3      quantity of O3 inhaled (fractional absorption, FA). Five men and five women were exposed
 4      orally to 0.2 or 0.4 ppm O3 while exercising at a minute volume of approximately 20 L/min for
 5      60 minutes or 40 L/min for 30 minutes. Ozone retention was calculated from breath-by-breath
 6      data taken  from fast response analyzers of O3 and airflow rates.  The FA was statistically
 7      analyzed in terms of subject, exposure concentration, minute volume, and exposure time.
 8           Fractional absorption ranged from 0.56 to 0.98 with a mean ± SD of 0.85 ± 0.06 for all
 9      2000 recorded breaths. Intersubject differences had the largest influence on FA, resulting in a
10      variation of approximately 10%.  Statistical analysis indicated that concentration,  minute
11      volume, and exposure time had statistically significant effects on FA. However, relatively large
12      changes in these variables were estimated to result in relatively small changes in FA. Note: the
13      quantity of O3 retained by the RT is equal to FA times the quantity of O3 inhaled;  thus, relatively
14      large changes in concentration, minute volume, or exposure time may result in relatively large
15      changes in the amount of O3 retained by the RT or absorbed locally.  Also, according to Overton
16      et al. (1996), difference in PAR dose due to anatomical variability may be considerably larger
17      than corresponding small changes in FA would indicate.
18           Santiago et al. (2001) studied the effects of airflow rate and O3 concentration on O3 uptake
19      in the nasal cavities of three women and seven men.  Air was supplied at a constant flow rate to
20      one nostril and exited from the other nostril while the subject kept the velopharyngeal aperture
21      closed by raising the soft palate.  Thus, a constant unidirectional flow of air plus O3 was
22      restricted to the nasal cavities. The fraction of O3 absorbed was calculated using the inlet and
23      outlet concentrations. Inlet concentration and airflow rate were varied in order to determine their
24      effect on O3 uptake.
25           The mean FA decreased from 0.80 to 0.33 with an increase in flow rate from 3 to
26      15 L/min.  The effect of both flow rate and subject on FA was statistically significant. Further
27      analysis indicated that the overall mass transfer coefficient was highly correlated with the flow
28      rate and that the gas phase resistance contributed from 6.3% (15 L/min) to 23% (3 L/min) of the
29      total resistance to O3 transfer to the nasal cavity surface. Concentration had a small, but
30      statistically significant effect on FA, when the inlet concentration was increased from 0.1  to
31      0.4 ppm O3, FA decreased from 0.36 to 0.32. The investigators observed that differences in FA

        August 2005                             AX4-9        DRAFT-DO NOT QUOTE OR CITE

-------
 1      among subjects were important; generally, subject variability accounted for approximately half
 2      of the total variation in FA.
 3           As mentioned above Ultman et al. (2004) tested hypotheses that differences in O3 uptake in
 4      lungs are responsible for variation in O3-induced changes in lung function parameters and that
 5      differences in O3 uptake are due to variations in breathing patterns and lung anatomy.  Thirty-
 6      two males and 28 female nonsmokers were exposed continuously for 1 h to either clean air or
 7      0.25 ppm ozone while exercising at a target minute ventilation of 30 L/min.  They first
 8      determined the forced expiratory response to clean air, then evaluated O3 uptake measuring dead
 9      space volume, cross-sectional area of peripheral lung (Ap) for CO2 diffusion, FEVb FVC,
10      and FEF250/0.75o/0.  The fractional O3 uptake efficiency ranged from 0.70 to 0.98, with a mean of
11      0.89 ± 0.06.  They found an inverse correlation between uptake and breathing frequency and  a
12      direct correlation between uptake and tidal volume.  There was a small, but statistically
13      significant decrease in uptake efficiency during the four sequential 15 minute intervals of the 1 h
14      exposure (0.906 ± 0.058 vs. 0.873 ± 0.088, first and last interval, respectively), in part due to the
15      increased breathing frequency and decreasing tidal volume occurring over the same  period.
16      Ozone uptake rate correlated with individual %Ap, but did not correlate with individual "/oFEVj.
17      Neither of these parameters correlated with the penetration volume determined in the bolus
18      studies mentioned above.  The authors concluded that the intersubject differences in forced
19      respiratory responses were not due to differences in O3 uptake. However, these data did  partially
20      support the second hypothesis, i.e., that the differences in cross-sectional area available for gas
21      diffusion induce differences in O3 uptake.
22
23
24      AX4.3  DOSIMETRY MODELING
25           When all of the animal and human in vivo O3 uptake efficiency data are compared, there is
26      a good degree of consistency across  data sets (U.S. Environmental Protection Agency, 1996).
27      This agreement raises the level of confidence with which these data sets can be used to support
28      dosimetric model formulations.
29           Recent data indicate that the primary site of acute cell injury occurs in the conducting
30      airways (Postlethwait et al., 2000).  These data must be considered when developing models that
31      attempt to predict site-specific locations of O3-induced injury. The early models computed

        August 2005                             AX4-10      DRAFT-DO NOT QUOTE OR CITE

-------
 1      relationships between delivered regional dose and response with the assumption that O3 was the
 2      active agent responsible for injury.  It is now known that reactive intermediates such as
 3      hydrohydroxyperoxides and aldehydes are important agents mediating the response to O3
 4      (further discussed in Section 5.3.1).  Thus, models must consider O3 reaction/diffusion in the
 5      epithelial lining fluid (ELF) and ELF-derived reactions products.
 6           Table AX4-2 presents a summary of new theoretical studies on the uptake of O3 by the RTs
 7      (or regions) of humans and laboratory animals that have become available since the 1996 review.
 8      They are discussed below.
 9           Overton and Graham (1995) described the development and simulation results of a
10      dosimetry model that was applied to a TB region anatomical model that had branching airways,
11      but which had identical single-path pulmonary units distal to each terminal bronchiole. The
12      anatomical model of the TB region  was based on Raabe et al. (1976), which reported lung cast
13      data for the TB region of a 330 g rat.
14           Rat effects data (from the PAR) are available that are identified with the lobe and the
15      generation in the lobe from which tissue samples were obtained (Pinkerton et al., 1995, 1998).
16      Models, like Overton et al. (1996), can be helpful in understanding the distribution of the
17      magnitude of such effects as well as suggesting sampling sites for future experiments.
18           Using computational fluid dynamics (CFD), Cohen-Hubal  et al. (1996) explored the effect
19      of the mucus layer thickness in the nasal passage of a rat. The nasal lining was composed of
20      mucus and tissue layers in which mass transfer was by molecular diffusion with first order
21      chemical reaction.  Physicochemical parameters for O3 were obtained from the literature.  Three
22      scenarios were considered: 10 jim thick mucus layer, no mucus  layer, and two nasal passage
23      regions each with a different mucus layer thickness.  Predictions of overall uptake were within
24      the range of measured uptake. Predicted regional O3 flux was correlated with measured cell
25      proliferation for the CFD simulation that incorporated two regions, each with a different mucus
26      thickness.
27           The reaction rate constant used by Cohen-Hubal and co-workers may be too low. Using
28      bolus-response data, Hu et al. (1994) and Bush et al. (2001) estimated a reaction rate constant
29      that is more than a 1000 times as large as that used by Cohen-Hubal et al. (1996). A rate
30      constant this large could result in a  conclusion different than those based on the  smaller constant.
31

        August 2005                             AX4-11       DRAFT-DO NOT QUOTE OR CITE

-------
                                                   Table AX4-2. New Ozone Dosimetry Model Investigations"
S-
to
o
          Purpose/Objective
                      Type of mass transport
                      model/Anatomical modelb
                              Species/ RT region
                              of interest/Regional
                              anatomical models
Ventilation and
Exposure
Results
                                                                                                                                                             Reference
to
 H
 6
 o

 o
 H
/O
 o
 H
 W
 O
 O
 HH
 H
 W
          To describe an RT
          dosimetry model that
          uses a branching TB
          region anatomical
          model and to
          illustrate the results
          of its application to a
          rat exposed to O3.
To incorporate into
the CFD model of
Kimbelletal. (1993)
resistance to mass
transfer in the nasal
lining and to
investigate the
effects of this lining
on O3 uptake.


To determine if the
single-path model is
able to simulate
bolus inhalation data
recorded during oral
breathing at quiet
respiratory flow.
                      One-dimensional (along axis
                      of airflow), time-dependent,
                      convection-dispersion
                      equation of mass transport
                      applied to each airway or
                      model segment. URT: single
                      path; TB: asymmetric
                      branching airways. PUL:
                      single path anatomical model
                      distal to each terminal
                      bronchiole.
Three dimensional steady-
state Navier-Stokes equations
for solving air velocity flow
field. Three dimensional
steady-state convection-
diffusion equation for O3
transport.
Three-dimensional CFD
model of the nasal passages
of a rat.

Single-path, one-dimensional
(along axis of airflow),
time-dependent, convection-
dispersion equation of mass
transport.  Single-path
anatomical model.
                              Rat/ RT/URT:
                              Patraetal. (1987).
                              TB: multiple path
                              model of Raabe et al.
                              (1976). PUL:
                              Mercer etal. (1991).
                                                             Rat/nasal passages
                                                             Nasal passages:
                                                             Kimbelletal. (1993).
                                                             Human/ RT/URT
                                                             (oral): Olson et al.
                                                             (1973).
                                                             LRT: Weibel(1963).
f= ISObpm;
VT=1.5,2.0,
2.5 mL.
One constant
concentration.
Steady-state
unidirectional
("inhalation")
flow rate
= 576mLl/min.
One constant
concentration.
VT = 500 mL,
f= 15bpm,
constant flow
rate = 250 mL/s.
Bolus-response
simulations.
(protocol used is
described by Hu
etal., 1992).
(1) For VT = 2.0 mL, f = 150 bpm:  The general shape
of the dose versus generation plot along any path from
the trachea to a sac is independent of path:  generally
the tissue dose decreases with increasing generation
index.  In the  TB region, the coefficient of variation for
dose ranges from 0 to 34 %, depending on generation.
The maximum ratio of the largest to smallest dose in
the same generation is 7; the average ratio being 3.
In the first PUL region model segment, the coefficient
of variation for the dose is 29 %. (2) The average dose
to the first PUL region model segment increases with
increasing VT.

Predictions of overall uptake were within the range of
measured uptake. Results suggest that mucus resistance
is important for describing O3 dosimetry and this
thickness may play a role for determining patterns
of O3-induced lesions in the rat nasal passage.
Simulations are sensitive to conducting airway volume
but are relative insensitive to characteristics of the
respiratory airspace. Although the gas-phase resistance
to lateral diffusion limits O3 absorption during quiet
breathing, diffusion through mucus may become
important at the large respiratory flows that are
normally associated with exercise.  The single-path
convection-diffusion model was a reasonable approach
/to simulate the bolus-response data.
                                                                                                                                                             Overton and
                                                                                                                                                             Graham
                                                                                                                                                             (1995)
                                                                                                                                                             Cohen-
                                                                                                                                                             Hubal et al.
                                                                                                                                                             (1996)
                                                                                                                                                             Bush et al.
                                                                                                                                                             (1996b)

-------
                                           Table AX4-2 (cont'd).  New Ozone Dosimetry Model Investigations a
to
o
o
H

6
o


o
H

O

O
H
W

O


O
HH
H
W
Purpose/Objective
To assess age- and
gender-specific
differences in
regional and
systemic uptake.


To examine the
impact on predictions
due to the value used
for the TB region
volume at FRC and
due to TB region
volume change
during respiration.



To make parameter
modifications so that
a single-path model
would simulate AF
from bolus-response
experiments
involving O3
(and C12).








Type of mass transport
model/Anatomical model b
PBPK, at ages 1,3,6, months
and 1,5, 10, 15,25, 50, and
75 years.




Single-path, one-dimensional
(along axis of airflow),
time-dependent, convection-
dispersion equation of mass
transport. Single-path
anatomical model.





Single-path, one-dimensional
(along axis of airflow), time-
dependent, convection-
dispersion equation of mass
transport. Single-path
anatomical model.










Species/ RT region
of interest/Regional
anatomical models
Human/ET/TB






Human /RT/
URT: Nunnetal.
(1959)
LRT: Weibel(1963)

Rat/RT/
URT: Patraetal.
(1987)
TB: Yehetal. (1979)
PUL: Mercer etal.
(1991)
Human/ RT/URT
(oral): Olson et al.
(1973).
URT (nasal):
Olson etal. (1973)
and Guilmette et al.
(1989)
LRT: Weibel(1963).








Ventilation and
Exposure
Pulmonary
ventilation ranged
from 34 mL/s (in
1 -month-old) to
190 mL/s (in
15-year-old).
VT varied with age
Human: VT = 500,
2250mL;f=15,
30 bpm.

Rat: VT=1.4,
2.4 ml; f= 96,
157 bpm.

One constant
concentration.

Oral & nasal
breathing. Flow
rates = 150,250,
1000 mL/s,
VT = 500 mL.
Bolus-response
simulations









Results
Regional extraction is insensitive to age. Extraction per
unit surface area is 2- to 8-fold higher in infants
compared to adults. PU and ET regions have a large
increase in unit extraction with increasing age. Early
postnatal period is time of largest differences in PK,
due to immaturity of metabolic enzymes.

(1) A better understanding and characterization of the
role of TB region expansion (mainly the rat) and
volume is important for an improved understanding of
respiratory-tract dosimetry modeling of reactive gases.
(2) Extrapolations based on dose in the PAR can differ
significantly from those based on exposure
concentration or total uptake.
(3) Human subjects who appear similar outwardly may
have very different PAR doses and potentially different
responses to the same exposure.(Uptake by the URT
was not considered.)
(Simulation results for O3 only) (1) Using parameter
values from the literature and assuming that absorption
was gas-phase controlled, the simulations of O3 data
were realistic at flow rate = 250 mL/s, but not realistic
at 1000 mL/s. (2) Accurate simulations at 250 mL/s
required modification of mass transfer coefficients
reported in the literature for the conducting airways.
(3) It was necessary to include a diffusion resistance for
the epithelial lining fluid based on an assumed O3
reaction rate constant that was much greater than in in
vitro estimates. (4) Partial validation of the final
parameters (determined at 250 mL/s) was obtained by
simulations of bolus-response data at flow rates of
150 and 1000 mL/s. Validation was obtained also by
simulating internal measurements of O3 in subjects
exposed during quiet breathing.
Reference
Sarangapani
et al. (2003)





Overton
etal. (1996)









Bush et al.
(2001)














         "See Appendix A for abbreviations and acronyms.

         bThe anatomical models used in an investigation generally differ from those described in the references, e.g., dimensions are often scaled to dimensions appropriate to the

          dosimetry investigation; or the original structure may be simplified, keeping or scaling the original dimensions.

-------
 1           With an RT dosimetry model, Overton et al. (1996) investigated the sensitivity of
 2      absorbed fraction (AF), proximal alveolar region (PAR) dose, and PAR dose ratio to TB region
 3      volume (Vra) and TB region expansion in human beings and rats.  The PAR was defined as the
 4      first generation distal to terminal bronchioles and the PAR dose ratio was defined as the ratio of
 5      a rat's predicted PAR dose to a human's predicted PAR dose. This ratio  relates human and rat
 6      exposure concentrations so that both species receive the same PAR dose. In rats the PAR is a
 7      region of major damage from O3. For each species, three literature values of V^ were used:
 8      a mean value and the mean ± twice the SD.  The following predictions were obtained:
 9           (1) The sensitivity of AF and PAR dose to V^ depends on species, ventilation, TB region
10      overall mass transfer coefficient (k^), and expansion. For k^ = 0.26 cm/s and quiet breathing,
11      AF was predicted to vary by less than 3% for the ±2 SD range of VTB.  In contrast, the PAR
12      dose predicted for the smallest VTB is five times larger than the PAR dose predicted with the
13      largest VTB. The effect of V^ is much less during heavy exercise: the ratio of maximum to
14      minimum PAR dose was approximately 1.5.  In any case, the simulations predicted that
15      fractional changes in AF due to different V^ are not, in general, a good predictor of the
16      fractional changes in PAR doses.
17           (2) Relative to no expansion in the TB region, expansion decreases both AF and PAR
18      dose.  The largest effect of including expansion in the human simulations was to decrease the AF
19      by «8%; in rats, the maximum decrease was -45%. The PAR doses decreased relatively more,
20      25 and 65% in human beings and rat, respectively.
21           (3) The authors attempted to obtain an understanding as to uncertainty or variability in
22      estimates of exposure concentrations (that give the same PAR dose in both species) if the
23      literature mean value of VTB was used. For various  values off, VT, kra, and expansion, the PAR
24      dose ratios at upper and lower values of VTB deviated in absolute values from the PAR dose ratio
25      calculated at the mean values of VTB by as little as 10% to as large as 310%.  The smallest
26      deviation occurred at the largest VT and smallest kra for both species; whereas, the largest
27      deviation occurred at the smallest VT and largest kra for both species.
28           Bush et al. (2001) modified the single-path model of Bush et al. (1996b) in order to be able
29      to simulate absorbed fraction data for O3 (and C12, which is not considered) for three airflow
30      rates and for oral and nasal breathing.  By adjusting several parameters a reasonable agreement
31      between predicted and experimental values was obtained.  On the other hand, the O3 plots of the

        August 2005                            AX4-14      DRAFT-DO NOT QUOTE OR CITE

-------
 1      experimental and predicted values of absorbed fraction versus penetration volume (e.g.,
 2      Figures 4 and 5 of Bush et al., 2001) show sequential groups composed of only positive or only
 3      negative residuals, indicating a lack of fit. Possibly adjusting other parameters would eliminate
 4      this. To obtain an independent validation of the model, Bush et al. (2001) simulated
 5      measurements of O3 concentrations made by Gerrity et al. (1995) during both inhalation and
 6      exhalation at four locations between the mouth and the bronchus intermedius of human subjects.
 7      Simulated and experimental values obtained are in close agreement.  Note, however, that Bush
 8      et al. made no quantitative assessment of how well their simulations agreed with the
 9      experimental data; assessments were made on the basis of visual inspection of experimental and
10      simulated values plotted on the same figure.  Thus, evaluation of the model was, or is,
11      subjective.
12          Recently Sarangapani et al. (2003) used physiologically based pharmacokinetic (PBPK)
13      modeling to characterize age- and gender-specific differences in both regional and systemic
14      uptake of O3 in humans. This model indicated that regional extraction of O3 is relatively
15      insensitive to age, but extraction per unit surface area is 2- to 8-fold higher in infants compared
16      to adults, due to the region-specific mass transfer coefficient not varying with age.  The PU and
17      ET regions have a large increase in unit extraction with increasing age because both regions
18      increase in surface area. Males and females in this model have similar trends in regional
19      extraction and regional unit extraction.  In early childhood, dose metrics were as much as
20      12 times higher than adult levels, but these differences leveled out with age, such that inhalation
21      exposures varied little after age 5. These data suggest that the early postnatal period is the time
22      of the largest difference in pharmacokinetics observed, and this difference is primarily due to the
23      immaturity of the metabolic enzymes used to clear O3 from the respiratory tract.
24          Mudway and Kelly (2004) attempted to model O3 dose-inflammatory response using a
25      meta-analysis of 23 exposures in published human chamber studies.  The O3 concentrations
26      ranged from 0.08 to 0.6 ppm and the exposure durations ranged from 60 to 396 minutes. The
27      analysis showed linear relationships between O3 dose and neutrophilia in bronchoalveolar lavage
28      fluid (BALF). Linear relationships were also observed between O3 dose and protein leakage
29      into BALF.
30
31

        August 2005                            AX4-15      DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX4.4  SPECIES HOMOLOGY, SENSITIVITY AND ANIMAL-TO-
 2              HUMAN EXTRAPOLATION
 3          Biochemical differences among species are becoming increasingly apparent and these
 4     differences may factor into a species' susceptibility to the effects of O3 exposure. Lee et al.
 5     (1998) compared SD rats and rhesus monkeys to ascertain species differences in the various
 6     isoforms of CYP moonoxygenases in response to O3 exposure (discussed in more detail in
 7     Section 5.3.1.2).  Differences in activities between rat and monkey were 2- to 10-fold, depending
 8     on the isoform and the specific lung region assayed. This study supports the view that
 9     differential expression of CYPs is a key factor in determining the toxicity of O3. As further
10     characterization of species- and region-specific CYP enzymes occurs, a greater understanding of
11     the differences in response may allow more accurate extrapolation from  animal exposures to
12     human exposures and toxic effects.
13          Arsalane et al. (1995) compared guinea pig and human AM recovered in BALF and
14     subsequently exposed in vitro to 0.1 to 1 ppm for 60 minutes.  Measurement of inflammatory
15     cytokines showed a peak at 0.4 ppm in both species. Guinea pig AM had an increase in IL-6 and
16     TNF-a while human AM had increases in TNF-a, IL-lb, IL-6 and IL-8.  This exposure also
17     caused an increase in mRNA expression for TNF-a, IL-lb, IL-6 and IL-8 in human cells.
18     At 0.1 ppm exposures, only TNF-a secretion was increased.  These data suggest similar cytokine
19     responses in guinea pigs and humans, both qualitatively and quantitatively.
20          Dormans et al. (1999) continuously exposed rats, mice, male guinea pigs to filtered air,
21     0.2, or 0.4 ppm O3 for 3 to 56 days or for 28 days to follow recovery at 3, 7, and 28 days PE.
22     Depending on the endpoint studied, the species varied in sensitivity.  Greater sensitivity was
23     shown in the mouse as determined by biochemical  endpoints, persistence of bronchiolar
24     epithelial hypertrophy, and recovery time.  Guinea pigs were more sensitive in terms of the
25     inflammatory response though all three species had increases in the inflammatory response after
26     three days that did not decrease with exposure. In all species the longest exposure to the highest
27     dose caused increased collagen in ductal septa and large lamellar bodies in Type II cells, but that
28     response also occurred in rats and guinea pigs at 0.2 ppm. No fibrosis was seen at the shorter
29     exposure times and the authors question whether fibrosis occurs in healthy humans after
30     continuous exposure. The authors do not rule out the possibility that some of these differences
       August 2005                            AX4-16      DRAFT-DO NOT QUOTE OR CITE

-------
 1      may be attributable to differences in total inhaled dose or dose actually reaching a target site.
 2      Overall, the authors rated mice as most susceptible, followed by guinea pigs and rats.
 3           Comparisons of airway effects in rats, monkeys and ferrets resulting from exposures of
 4      1.0 ppm O3 for 8 h (Sterner-Kock et al. 2000) demonstrated that monkeys and ferrets had a
 5      similar inflammatory responses and epithelial necrosis. The response of these two species was
 6      more severe than that seen in rats. These data suggest that ferrets are a good animal model
 7      for O3-induced airway effects due to the similarities in pulmonary structure between primates
 8      and ferrets.
 9           The rat is a key species used in O3 toxicological studies, but Watkinson and Gordon,
10      (1993) suggest that, because the rat has both behavioral and physiological mechanisms that can
11      lower core temperature in response to acute exposures, extrapolation of these exposure data to
12      humans may be limited. Another laboratory (Iwasaki et al., 1998) has demonstrated both
13      cardiovascular and thermoregulatory responses to O3 at exposure to 0.1, 0.3, and 0.5 ppm O3
14      8 h/day for 4 consecutive days. A dose-dependent disruption of HR and Tco were seen on the
15      first and second days of exposure, which then recovered to control values. Watkinson et al.
16      (2003) exposed rats to 0.5 ppm O3 and observed this hypothermic response which included
17      lowered HR, lowered Tco, and increased inflammatory components in BALF. The authors
18      suggest that the response is an inherent reflexive pattern that can possibly attenuate O3 toxicity in
19      rodents. They discuss the cascade of effects created by decreases in Tco, which include:
20      (1) lowered metabolic rate,  (2) altered  enzyme kinetics, (3) altered membrane function,
21      (4) decreased oxygen consumption and demand, (5) reductions in minute ventilation, which
22      would act to limit the dose of O3  delivered to the lungs. These effects are concurrent with
23      changes in HR which lead to: (1) decreased CO, (2) lowered BP, (3) decreased tissue perfusion,
24      all of which may lead to functional deficits. The hypothermic response has not been observed in
25      humans except at very high exposures.
26
        August 2005                             AX4-17      DRAFT-DO NOT QUOTE OR CITE

-------
  1       REFERENCES

  2       Arsalane, K.; Cosset, P.; Vanhee, D.; Voisin, C.; Hamid, Q.; Tonnel, A.-B.; Wallaert, B. (1995) Ozone stimulates
  3             synthesis of inflammatory cytokines by alveolar macrophages in vitro. Am. J. Respir. Cell Mol. Biol.
  4             13:60-68.
  5       Asplund, P. T.; Ben-Jebria, A.; Rigas, M. L.; Ultman, J. S. (1996) Longitudinal distribution of ozone absorption in
  6             the lung: effect of continuous inhalation exposure. Arch. Environ. Health 51: 431-438.
  7       Bush, M. L.; Asplund, P. T.; Miles, K. A.; Ben-Jebria, A.; Ultman, J. S. (1996a) Longitudinal distribution of O3
  8             absorption in the lung: gender differences and intersubject variability. J. Appl. Physiol. 81: 1651-1657.
  9       Bush, M. L.; Raybold, T.; Abeles, S.; Hu, S.-C; Ben-Jebria, A.; Ultman, J. S.  (1996b) Longitudinal distribution of
10             ozone absorption in the lung: simulation with a single-path model. Toxicol. Appl. Pharmacol. 140: 219-226.
11       Bush, M. L.; Zhang, W.; Ben-Jebria, A.; Ultman, J. S. (2001) Longitudinal distribution of ozone and chlorine in the
12             human respiratory tract: simulation of nasal and oral breathing with the single-path diffusion model. Toxicol.
13             Appl. Pharmacol. 173: 137-145.
14       Cohen-Hubal, E. A.; Kimbell, J. S.; Fedkiw, P. S. (1996) Incorporation of nasal-lining mass-transfer resistance into a
15             CFD model for prediction of ozone dosimetry in the upper respiratory tract. Inhalation Toxicol. 8: 831-857.
16       Dahl, A. R. (1990) Dose concepts for inhaled vapors and gases. Toxicol. Appl. Pharmacol. 103: 185-197.
17       Dormans, J. A. M. A.; VanBree, L.; Boere, A. J. F.; Marra, M.; Rombout, P. J. A. (1999) Interspecies differences in
18             time course of pulmonary toxicity following repeated exposure to ozone. Inhalation Toxicol. 11: 309-329.
19       Gerrity, T. R.; Biscardi, F.; Strong, A.; Garlington, A. R.; Brown, J. S.; Bromberg, P. A. (1995) Bronchoscopic
20             determination of ozone uptake in humans. J. Appl. Physiol. 79: 852-860.
21       Guilmette, R. A.; Wicks, J. D.; Wolff, R. K. (1989) Morphometry of human nasal airways in vivo using magnetic
22             resonance imaging. J. Aerosol. Med. 2: 365-377.
23       Hu, S. C.; Ben-Jebria, A.; Ultman, J. S. (1992) Longitudinal distribution of ozone absorption in the lung: quiet
24             respiration in healthy subjects. J. Appl. Physiol. 73:  1655-1667.
25       Hu, S.-C.; Ben-Jebria,  A.; Ultman, J. S. (1994) Longitudinal distribution of ozone absorption in the lung: effects  of
26             respiratory flow. J. Appl. Physiol. 77:  574-583.
27       Iwasaki, T.; Takahashi, M.; Saito, H.; Arito, H. (1998) Adaptation of extrapulmonary responses to ozone exposure in
28             conscious rats. Ind. Health 36: 57-60.
29       Kimbell, J. S.; Gross, E. A.; Joyner, D. R.; Godo, M. N.; Morgan, K. T. (1993) Application of computational fluid
3 0             dynamics to regional dosimetry of inhaled chemicals in the upper respiratory tract of the rat. Toxicol. Appl.
31             Pharmacol. 121:253-263.
32       Lee, C.; Watt, K. C.; Chang, A. M.; Plopper, C. G.; Buckpitt, A. R.; Pinkerton, K. E. (1998) Site-selective
3 3             differences in cytochrome P450 isoform activities: comparison of expression in rat and rhesus monkey lung
34             and induction in rats. Drug Metab. Dispos. 26: 396-400.
35       Mercer, R. R.; Anjilvel, S.; Miller, F. J.; Crapo, J. D. (1991) Inhomogeneity of ventilatory unit volume and its effects
36             on reactive gas uptake. J. Appl. Physiol. 70: 2193-2205.
37       Miller, F. J. (1995) Uptake and fate of ozone in the respiratory tract. Toxicol. Lett. 82/83: 277-285.
38       Miller, F. J.; Overton, J. H., Jr.; Jaskot, R. H.; Menzel, D. B. (1985) A model of the regional uptake of gaseous
3 9             pollutants in the lung: I. the sensitivity of the uptake of ozone in the human lung to lower respiratory tract
40             secretions and exercise. Toxicol. Appl. Pharmacol. 79: 11-27.
41       Mudway, I. S.; Kelly, F. J. (2004) An investigation of inhaled ozone dose and the magnitude of airway inflammation
42             in healthy adults. Am. J. Respir. Crit. Care Med. 169: 1089-1095.
43       Nodelman, V.; Ultman, J. S. (1999a) Longitudinal distribution of chlorine absorption in human airways: comparison
44             of nasal and oral quiet breathing. J. Appl. Physiol. 86: 1984-1993.
45       Nodelman, V.; Ultman, J. S. (1999b) Longitudinal distribution of chlorine absorption in human airways:
46             a comparison to ozone absorption. J. Appl. Physiol.  87: 2073-2080.
47       Nunn, J. F.; Campbell, E. J. M.; Peckett, B. W. (1959) Anatomical subdivisions of the volume of respiratory dead
48             space and the effect of position of the jaw. J. Appl. Physiol. 14: 174-176.
49       Olson, D. E.; Sudlow, M. F.; Horsfield, K.; Filley,  G. F. (1973) Convective patterns of flow during inspiration. Arch.
50             Intern. Med. 131:51-57.
51       Overton, J. H.; Graham, R. C. (1995) Simulation of the uptake of a reactive gas in a rat respiratory tract model with
52             an asymmetric tracheobronchial region patterned on complete conducting airway cast data. Comput. Biomed.
53             Res. 28: 171-190.
         August 2005                                  AX4-18        DRAFT-DO NOT QUOTE OR CITE

-------
  1       Overton, J. H.; Graham, R. C.; Menache, M. G.; Mercer, R. R.; Miller, F. J. (1996) Influence of tracheobronchial
  2             region expansion and volume on reactive gas uptake and interspecies dose extrapolations. Inhalation Toxicol.
  3             8:723-745.
  4       Patra, A. L.; Menache, M. G.; Shaka, N. B.; Gooya, A. (1987) A morphometric study of nasal-pharyngeal growthfor
  5             particle deposition in the rat. Am. Ind. Hyg. Assoc. J. 48: 556-562.
  6       Pinkerton, K. E.; Menache, M. G.; Plopper, C. G. (1995) Consequences of prolonged inhalation of ozone on F344/N
  7             rats: collaborative studies. Part IX. Changes in the tracheobronchial epithelium, pulmonary acinus, and lung
  8             antioxidant enzyme activity. Cambridge, MA: Health Effects Institute; pp. 41-98; research report no. 65.
  9             Available from: NTIS, Springfield, VA; PB95-261996.
10       Pinkerton, K. E.; Weller, B. L.; Menache, M. G.; Plopper, C. G. (1998) Consequences of prolonged inhalation of
11             ozone on F344/N rats: collaborative studies. Part XIII. A comparison of changes in the tracheobronchial
12             epithelium and pulmonary acinus in male rats at 3 and 20 months. Cambridge, MA: Health Effects Institute;
13             research report no. 65.
14       Postlethwait, E. M.; Joad, J. P.; Hyde, D. M.; Schelegle, E. S.; Brie, J. M.; Weir, A. J.; Putney, L. F.; Wong, V. J.;
15             Velsor, L. W.; Plopper, C. G. (2000) Three-dimensional mapping of ozone-induced acute cytotoxicity in
16             tracheobronchial airways of isolated perfused rat lung. Am. J. Respir. CellMol. Biol. 22: 191-199.
17       Raabe, O.  G.; Yeh, H. C.; Schum, G. M.; Phalen, R. F. (1976) Tracheobronchial geometry: human, dog, rat,
18             hamster. Albuquerque, NM: Lovelace Foundation; report no. LF-53.
19       Rigas, M. L.; Ben-Jebria, A.; Ultman, J. S. (1997) Longitudinal distribution of ozone absorption in the lung: effects
20             of nitrogen dioxide, sulfur dioxide, and ozone exposures. Arch. Environ. Health 52: 173-178.
21       Rigas, M. L.; Catlin, S. N.; Ben-Jebria,  A.; Ultman, J. S. (2000) Ozone uptake in the intact human respiratory tract:
22             relationship between inhaled dose and actual dose. J. Appl. Physiol. 88: 2015-2022.
23       Santiago, L.  Y.; Harm, M. C.; Ben-Jebria, A.; Ultman, J. S. (2001) Ozone adsorption in the human nose during
24             unidirectional airflow. J. Appl. Physiol.  91: 725-732.
25       Sarangapani, R.; Gentry, P. R.; Covington, T. R.; Teeguarden, J. G.; Clewell, H. J., III. (2003) Evaluation of the
26             potential impact of age- and gender-specific lung morphology and ventilation rate on the dosimetry of vapors.
27             Inhalation Toxicol. 15: 987-1016.
28       Sterner-Kock, A.;  Kock, M.; Braun, R.; Hyde, D. M. (2000) Ozone-induced epithelial injury in the ferret is  similar
29             to nonhuman primates. Am. J. Respir. Crit. Care Med. 162: 1152-1156.
30       Ultman, J. S.; Ben-Jebria, A.; Arnold, S. F. (2004) Uptake distribution of ozone in human lungs: intersubject
31             variability in physiologic response. Boston, MA: Health Effects Institute; research report no. 125. Available:
32             http://www.healtheffects.org/Pubs/Ultman.pdf [29 July, 2005].
33       U.S. Environmental Protection Agency. (1996) Air quality criteria for ozone and related photochemical oxidants.
34             Research Triangle Park, NC: Office of Research and Development; report nos. EPA/600/AP-93/004aF-cF. 3v.
3 5             Available from: NTIS, Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available online at:
36             www.epa.gov/ncea/ozone.htm.
37       Watkinson, W. P.; Gordon, C. J. (1993) Caveats regarding the use of the laboratory rat as a model for acute
3 8             lexicological studies: modulation of the toxic response via physiological and behavioral mechanisms.
39             Toxicology 81: 15-31.
40       Watkinson, W. P.; Campen, M. J.; Wichers, L. B.; Nolan, J. P.; Costa, D. L. (2003) Cardiac and thermoregulatory
41             responses to inhaled pollutants in healthy and compromised rodents: modulation via interaction with
42             environmental factors. Environ. Res. 92: 35-47.
43       Weibel, E. R. (1963) Morphometry of the human lung. New York, NY: Academic Press Inc.
44       Yeh, H. C.; Schum, G. M.; Duggan, M.  T. (1979) Anatomic models of the tracheobronchial and pulmonary regions
45             of the rat. Anat. Rec. 195: 483-492.
46
47
         August 2005                                   AX4-19        DRAFT-DO NOT QUOTE OR CITE

-------
    AX5. ANNEX TO CHAPTER 5 OF OZONE AQCD
August 2005                  AX5-1    DRAFT-DO NOT QUOTE OR CITE

-------
                                              Table AX5-1.  Cellular Targets of Ozone Interaction
_-
S-
to
o
o
Concentration
ppm Duration
0.22 4 h with
exercise
0.5 - 10 both 1 h
with and
without 5% CO2
0.25 - 1.0 30-60 min
Species
Rat, male, SD, 90 days
old, 300-330 g,
n = 6/group
Rat, male, SD, 90 days
old, 300-330 g,
n = 6-14/group
Rat, male, SD,
250-275 g,
Effects
Demonstrated the ozonation of PUFA to form nonanal and hexanal in rat B ALF.
Increases in nonanal not accompanied by significant changes in lung function,
epithelial permeability, or airway inflammation. Hexanal levels did not increase
significantly. Levels of both aldehydes returned to baseline by 18 h PE.
O3 plus CO2 increased the VT and the yield of aldehydes with a maximal
aldehyde yield at 2.5 ppm for 1 h. 0.5 ppm O3 with 5% CO2, levels of hexanal
and nonanal increased at 30 minutes, decreased slightly from that level at 60
minutes, was maximal at 90 minutes and then dropped to 60 minutes levels at
120 minutes. Levels of heptanal did not change appreciably during this time
course. Suggests that levels of these aldehydes were dependent on a dynamic
relationship between their production and the disappearance from the ELF.
PUFAs directly react with O3. The amount of bioactive lipids produced is
inversely related to AA availability.
Reference
Frampton
et al. (1999)
Pryor et al.
(1996)
Postlethwait
et al. (1998)
X

H
6
o
2
0
H 0.4
O
c
o
H
W
O

O
H
W
Chemical
systems:
15-10 min.
Cell: 1-2 h




1 h with
exercise







Interfacial films
(dipalmitoylgly cero-3 -
phosphocholine
(DPPC));Rat, SD
BALF; human lung
fibroblast cell line


Human AM








DPPC films: reduced O3 reactive absorption by antioxidants. Human lung
fibroblast cell line: AA produced cell injury; high levels of O3 and AA were
needed to induce cell injury; the DPPC films reduced the amount of cell injury.
Suggests that O3 reactions with ELF substrates cause cell injury; that films of
active, saturated phospholipids reduce the local dose of O3-derived reaction
products; and that these interfacial phospholipids modulate the distribution of
inhaled O3 and the extent of site-specific cell injury.

Ozone exposure caused apoptosis, an increase in a 32-kDa protein adduct, and
an increase in ferritin and a 72-kDa heat shock protein. Exposure of AM to
HNE replicated these effects suggesting that creation of protein adducts and
apoptotic cell death are cellular toxic effects of acute O3 exposure and that they
are mediated, at least in part by HNE.




Connor
et al. (2004)






Hamilton
et al. (1998)








-------
                                             Table AX5-2. Effects of Ozone on Lung Monooxygenases
to
o
o
>
X
Concentration
ppm            Duration     Species
                          Effects
                                                                    Reference
1.0

1.0

8h

90 days

Rat, male, SD,
350-600 g,
n= 3-6/group

Increases CYP2E1 activity in lobar-bronchi and major daughter airway
with 8 h exposure. Decreased CYP2E1 activities in both major and minor
daughter airways with 90 day exposure. O3 does not result in consistent
dramatic alterations in CYP2E1 activities.
Watt et al.
(1998)


                         8 h/day for
                         90 days

                         2h
        NA
Rat, male, SD,
275-300 g

Mice, male, 2-3 months
old, Clara Cell Secretory
Protein deficient,
WT strain 129
n = 3/group

Rat, male, SD, adult
monkey, Rhesus, 0.75 to
9.7 years old
CYP2B activity increased. Linked to Clara cells in distal lung only—not in   Paige et al.
trachea or proximal airway.                                            (2000a)

CCSP-<-mice had increases in IL-6 and MT mRNA that preceded           Mango et al.
decreases in Clara cell CYP2F2 mRNA.  WT mice had levels change,        (1998)
but to a lesser degree.
Microdissection for regiospecific and species-specific differences in         Lee et al.
isoforms of CYPs. Rat parenchyma:  both CYP1A1 and CYP2B            (1998)
were highest.  Rat airways:  CYP2E1 was highest. Rat airways and
parenchyma: P450 reductase activities were high, and conversely, low in
trachea. Monkeys: did not exhibit site-selective differences in CYP2B1,
CYP1A1, or P450 reductase; however, they had high CYP2E1 activity
in parenchyma and distal bronchioles.
         CYP = Cytochrome P-450
         WT = wild-type
         MT = metallothionein
         CCSP = Clara Cell Secretory Protein

-------
                Table AX5-3.  Antioxidants, Antioxidant Metabolism, and Mitochondrial Oxygen Consumption
S-
to
o
Concentration
ppm             Duration       Species
                                          Effects
                                                                                               Reference
X
-k
1.0
6 h/day,
5 days/week
for 2 or 3 mo.
Rat, male Fischer F344,
30-32 days old, n = 4
0.2
6h
Dog, male, mongrel,
~/9 kg, n = 6/group
0,0.1,0.25,
0.5, 1.0 or 1.5
30-720 min
Model of continually
mixed, interfacial exposure
                                                                   Immunohistochemistry and immunogold labeling studies. In epithelial      Weller et al.
                                                                   cells in airways and parenchyma: reduced Cu-Zn SOD labeling with        (1997)
                                                                   O3-exposure. In the CAR regions, in both AMs and Type II epithelial
                                                                   cells:  significantly increased levels of Mn SOD. Mn SOD levels were not
                                                                   increased in Type I epithelial cells, fibroblasts, or Clara cells.  Suggests
                                                                   that the increased levels of Mn SOD in Type II cells in the proximal
                                                                   alveolar duct confer tolerance and protection from further O3-induced
                                                                   injury.

                                                                   Blocking antioxidant transport with probenecid caused heterogeneously     Freed et al.
                                                                   distributed increases in peripheral airway resistance and reactivity.          (1999)
                                                                   Probenecid inhibited O3-induced neutrophilic inflammation, which is
                                                                   evidence of a dissociation between airway function and inflammation.
                                                                   Probenecid caused a 50-60% decrease in plasma urate, a decrease in
                                                                   ascorbate, and a decrease in B ALF protein.  Suggests probenecid has
                                                                   either a direct or indirect effect on either cytokine or leukotriene transport.

                                                                   Modeled the  interactions of O3 with three ELF antioxidants, AA, UA and    Mudway and
                                                                   GSH.  Ranking of reactivity with O3 was UA>AA>GSH. Antioxidants   Kelly (1998)
                                                                   caused no changes in sample pH and protein carbonyl formations.
                                                                   Consumption of the antioxidants occurred in a linear fashion with time
                                                                   and a positive relationship to O3 concentration. Suggests that GSH is
                                                                   not an important substrate for O3; UA appeared to be the most important
                                                                   substrate.
Cu = Copper
Zn = Zinc
Mn = Manganese
SOD = Superoxide Dismutase
AM = Alveolar Macrophage
BALF = Bronchoalveolar Lavage Fluid
AA = Ascorbic Acid
UA = Uric Acid
GSH = Reduced Glutathione

-------
                                              Table AX5-4.  Lipid Metabolism and Content of the Lung
S-
to
o
         Concentration
         ppm
Duration
Species/Cell Line
Effects
Reference
>
X
        NA
        NA
60 min
Cultured human epithelial
cells (BEAS-2B)
               Cultured human bronchial
               epithelial cells (NHBE)
               and BEAS-2B cells
         0.06, 0.125, and
         0.25 ppm
2 to 48 h
Lung (calf surfactant)
                          30 min
               Human red blood cell
               (RBC) model
Incubation withlO uM lipid oxidation products caused significant           Pryor et al.
release of arachidonic acid. Suggests that lipid oxidation products cause     (1995)
activation of specific lipases, which then trigger the activation of second
messenger pathways (e.g., phospholipase A2 or phospholipase C).

Incubation with lipid ozonation product                                  Kafoury et al.
1 -palmitoyl-2-(9-oxononanoyl)-sn-glycero-3 -phosphocholine (PC-ALF)     (1999)
and 1-hydroxy-l-hydroperoxynonane (HHP-C9).  PC-ALF elicited
release of (PAF) and prostaglandin E2, but not IL-6. HHP-C9 caused
release of PAF and IL-6 in these cells, but not prostaglandin E2.
Suggests that O3-induced production of lipid ozonation products
causes release of proinflarnmatory mediators that then generate an
early inflammatory response.

Dose- and time-dependent increase in the formation of                     Uhlson et al.
1 -palmitoyl-2-(9'-oxo-nonanoyl)-glycerophosphocholine                   (2002)
(16:Oa/9-al-GPCho), an oxidized phospholipid, which possessed
biological activity in three assays. The 16:Oa/9-al-GPCho:
(1) decreased macrophage viability by necrosis at 6 uM, (2) induced
apoptosis in pulmonary epithelial-like A549 cells at 100-200 uM,
and (3) elicited release of IL-8 from A549 cells at 50-100 uM.

Human RBCs intermittently covered by an aqueous film consisting of       Ballinger et al.
rat BALF or BALF plus added reagents that included AA, UA, GSH,        (2005)
Trolox (a vitamin E analog), SOD, catalase, desferrioxamine,
deithylenetriaminepentaacectic acid, mannitol or BSA.  Oxidation of
lipids in BALF or on membranes was assayed by measuring TEARS
accumulation and loss of acetylcholinesterase activity. AA and GSH
induced dose-dependent oxidative damage to the cell membrane proteins
and lipids via secondary oxidant formation.  Conclusion: early in O3
exposure, ELF antioxidants are high enough to drive reactive absorption
of O3 into the ELF and to concurrently quench secondary reaction
products, thus limiting cell injury. With continued exposure, O3 flux
would decrease, cellular injury would increase due to depleted
antioxidants levels. They hypothesize that especially in areas where
O3 deposition is high, unreacted O3 and cytotoxic products can diffuse
to the cell membranes, causing injury.

-------
                                         Table AX5-4 (cont'd). Lipid Metabolism and Content of the Lung
S-
to
o
         Concentration
         ppm            Duration
               Species/Cell Line
                           Effects
                                                                     Reference
X
         2.0
4h
         0.2, 0.5, or 1.0     1 h
Rat BALF exposed to O3
in vitro.
               Cultured human bronchial
               epithelial cells (16-HBE)
         0.5, 1.0,2.0,
         or 3.0
               Mice, C57BL/6J,
               8-12 week old female.
               n=4
The major reaction products identified were                              Pulfer and
5-hydroperoxy-B-homo-6-oxa-cholestan-3,7a-diol,                        Murphy
5P,6p-epoxycholesterol, and 3p-hydroxy-5oxo-5,6-seco-cholestan-6-al.      (2004)

Extracted lipid reaction products either immediately or 24 h PE.
Higher levels of 5p,6p-epoxycholesterol were recovered in the extract
immediately PE.  Both 5 P,6 P-epoxy cholesterol and its most abundant
metabolite, cholestan-6-oxo-3p,5oc-diol, were cytotoxic to 16-HBE cells
at physiologically relevant concentrations.  Both reaction products were
also shown to be potent inhibitors of cholesterol synthesis.

In BALF, lavaged cells, and whole lung homogenate there was a            Pulfer et al.
dose-dependent increase in formation cholestan-6-oxo-3 P,5oc-diol at         (2005)
exposure levels of 0.5, 1.0 and 2.0 ppm O3, respectively. In BALF
and lavaged cells O3 induced increases in 5p,6p-epoxycholesterol at
exposure levels of 0.5 and 1.0, respectively.
        PAF = Platelet-Activating Factor
        RBC= Red Blood Cell
        AA = Ascorbic Acid
        GSH = Reduced Glutathione
        IL-6 = Interleukin-6
        IL-8 = Interleukin-8
        ELF = Epithelial Lining Fluid
        BALF = Bronchoalveolar Lavage Fluid
        BSA - Bovine Serum Albumin
        TEARS = thiobarbituric acid
        UA = Uric Acid

-------
                                                  Table AX5-5. Effects of Ozone on Protein Synthesis
S-
to
o
Concentration
ppm
Duration
                                         Species
Effects
                                                                                                 Reference
         0.4
1,3,7,28, or
56 days
Rat, male, Wistar,
200 g,
n = 5/group
                                                            Centriacinar thickening of septa after 7 days of exposure, which
                                                            progressed at 28 and 56 days of exposure.  After 28 days of O3, the
                                                            increase in collagen content in ductular septa was apparent and it
                                                            increased progressively until the 56 daytime point.  Collagen content
                                                            decreased with PE recovery, the structural fibrotic changes in ductular
                                                            septa did not return to control levels.  Respiratory bronchioles were
                                                            present at an increasing degree, which persisted after an 80-day recovery
                                                            period. Suggests that subchronic O3 exposures in rats creates a
                                                            progression of structural lung injury that can evolve into a more chronic
                                                            form, which included fibrosis.
                                                                      van Bree et al.
                                                                      (2001)
X

-------
                                           Table AX5-6. Effects of Ozone on Differential Gene Expression
S-
to
o
         Concentration
         ppm
Duration
Species
Effects
Reference
         1
8 hours for
three
consecutive
nights
Mice, C57BL/6,
20-25 g,
n = 5-6/group
Ozone exposure induced changes in expression of 260 genes ( 80%
repressed and 20% induced).  Cell cycle genes upregulated: S-adenosyl
methionine decarboxylase 3, ribonucleotide reductase, and clusterin.
NF-KB-induced genes upregulated: serum amyloid protein, topoisomerase
Hoc, monocyte chemoattractant protein, platelet-derived growth factor,
and inhibitor of apoptosis. Downregulation of transcripts for isoforms
of myosins and actins.  CYP family genes downregulated:2a4,and 2el,
and 2f2, as were aryl-hydrocarbon receptor and several glutathione
transferases.  Metallothionein 1 and 2 and lactotransferrin upregulated.
Major histocompatibility complex genes and lymphocyte specific
proteins downregulated.
Gohil et al.
(2003)
X

-------
                                                 Table AX5-7. Effects of Ozone on Lung Host Defenses
S-
to
o
         Concentration
         ppm
                               Duration
Species
Effects
Reference
         Microbiologic Endpoints
         0.1,0.3
                               4 h/day,
                               5 days/week,
                               1 or 3 weeks
Rat, male, Fischer F344,
200-250 g
                               3h
                                              Rat, male, Fischer F344,
                                              8 weeks old,
                                              n = 4-10/group
No effect on cumulative mortality from subsequent lung infection         Cohen et al.
with 4-8 x 106 Listeria monocytogenes, but concentration-related          (2001, 2002)
effects on morbidity onset and persistence. One-week exposed rats:
listeric burdens trended higher than in controls; 0.3 ppm rats displayed
continual burden increases and no onset of resolution; in situ IL-loc,
TNFcc, and IFNy levels 48 and 96 h post-infection (4 x 106 level) higher
than controls. Three-week exposed rats: no O3-related change
in bacterial clearance; IL-loc, TNFoc, and IFNy levels higher than
control only at 48 h post-infection (4 x  106) and only with 0.3 ppm rats.

Single exposure to S. zooepidemicus led to differential clearance           Dong et al.
patterns in exposed rats maintained on ad libitum or O3-mitigating         (1998)
calorie-restricted diets.
X
H
6
o

o
H
/O
o
H
W
O
O
HH
H
W
         Clearance Endpoints (Non-Microbial)

         0.01-1.0              lOmin         Rat, SD
         0.4
                               6h
Dog, male, mixed breed,
1-3 years old,
n = 7/group
Single 10 min exposure of trachea! explants, followed by 1 h incubation    Churg et al.
with particles, led to dose-related increases in uptake of amosite           (1996)
asbestos and titanium dioxide particles. Effect inhibited by added
catalase or desferoxamine, but not by superoxide dismutase.

Increased trachea! permeability to 99mTc-DTPA after direct sublobar        Foster and
exposure to O3.  Clearance halftimes remained significantly lower for      Freed (1999)
1-7 d PE, but recovered by 14 days PE.
         Alveolar Macrophage Endpoints (General)
0.8
0.8
1.0
3h
3h
4h
Rat, male, SD,
250-275 g,
n = 5-6/group
Rat, male, SD,
250-300 g,
n = 5/group
Mice cell line
(WEH1-3)
Increased ex vivo AM adherence to epithelial cultures mitigated by
cell pretreatment with anti-CD 1 Ib or anti-ICAM-1 antibodies.
Increased ex vivo AM adherence to epithelial cell cultures mitigated
by cell pretreatment with anti-TNFoc/IL-la antibodies.
Increased intracellular calcium resting levels in WEHI-3 cells.
Decreased rates of calcium influx due to digitonin.
Bhalla (1996)
Pearson and
Bhalla (1997)
Cohen et al.
(1996)

-------
                                           Table AX5-7 (cont'd).  Effects of Ozone on Lung Host Defenses
S-
to
o
         Concentration
         ppm
Duration
Species
                        Effects
Reference
X
         Alveolar Macrophage Endpoints (Functional)
         0.1,0.3
         0.1,0.3
         0.3
4 h/day,
5 days/week,
1 or 3 weeks
4 h/day,
5 days/week,
1 or 3 weeks
5 h/day,
5 days/week,
4 weeks
                              3h
Rat, male, Fischer F344,
200-250 g
Rat, male, Fischer F344,
200-250 g
                        Superoxide anion: increased AM production (1 week; 0.1,0.3 ppm);
                        no intergroup differences noted after IFNy stimulation.
                        H2O2: reduced production (1 week; 0.1, 0.3 ppm); further reduced
                        production after treatment with IFNy (0.1, 0.3 ppm, 1 and 3 weeks).

                        Increased AM superoxide anion production (1 week; 0.1, 0.3 ppm),
                        Lower H2O2 production (1 week; 0.1,0.3 ppm). Reduced production
                        after treatment with IFNy - superoxide (0.3 ppm, 1 week) and H2O2
                        (0.1 ppm, 1 week) - relative to cells without IFNy treatment.
                        No effects from 3-week exposures.

Rat, male, Fischer F344,   No effect on AM endotoxin-stimulated IL-lcc, IL-6, or
200-250 g,               TNFcc production. Decrease in stimulated, but not spontaneous,
n = 10/group             superoxide formation; variable effects on H2O2 formation. No effect
                        on AM spontaneous, endotoxin-, or IFNy-stimulated, NO formation.
               Rat, male, SD, 250-
               275 g, n = 5-6/group
                        Increased AM motility in response to chemotaxin; effect mitigated
                        by cell pretreatment with anti-CD 1 Ib or anti-ICAM-1 antibodies.
Cohen et al.
(2001)
Cohen et al.
(2002)
                                                                                          Cohen et al.
                                                                                          (1998)
                                                                                          Bhalla (1996)
                                             Rat, male, Fischer F344,
                                             8 weeks old,
                                             n = 4-10/group
                                       Decrease in AM phagocytic activity.
                                                                                          Dong et al.
                                                                                          (1998)
H 0.8
O
o
21
o
H
O
C i.o
O
H
O
O
H
W
3 h Mice, female,
(B6J129SV)
(C57/BL6X 129 NOS"'"),
8-16 weeks old,
n= 3-12/group

24 h/day, Rat, male, Wistar,
3 days 8-12 weeks,




Increased AM spontaneous and IFNy + LPS-induced NOS expression
and NO production and PGE2 release. Initial decrease in ROI
production, with eventual rebound. Knockout (NOS"'") mice AM
incapable of similar response to O3 - no inducible NO or PGE2
above control levels and consistent decreased ROI production.

BALF from exposed rats subsequently inhibited IFNy -induced
AM NO production.




Fakhrzadeh
et al. (2002)




Koike et al.
(1998, 1999)





-------
                                        Table AX5-7 (cont'd). Effects of Ozone on Lung Host Defenses
to
o
o
X
Concentration
ppm
Duration
Cytokines, Chemokines: Production,
0.1,0.3


0.1,0.3







0.3


0.3
1.0
2.5
4 h/day,
5 days/week,
1 or 3 weeks
4 h/day,
5 days/week,
1 or 3 weeks





5 h/day,
5 days/week,
4 weeks
24 or 96 h
1, 2, or4h,
or24h
Species
Effects
Reference
Binding, and Inducible Endpoints
Rat, male, Fischer F344,
200-250 g,

Rat, male, Fischer F344,
200-250 g,
n= 3-5/group





Rat, male, Fischer F344,
200-250 g,
n = 10/group
Mice, male, C57B1/6J,
adult, n = 3/group

Superoxide anion: no intergroup differences noted after IFNy
stimulation. H2O2: reduced production after treatment with IFNy.

Decreased expression of CDS among lung lymphocytes (0. 1 ppm only;
3 weeks); effect exacerbated by stimulation with IFNcc (but not with
IL-loc). Decreased expression of CD25 (IL-2R) on CD3+ lymphocytes
(0.3 ppm only; 3 weeks); effect worsened by treatment with IL-loc
(0.1, 0.3 ppm; 3 weeks). No effects on IL-2-inducible lympho-
proliferation. Reduced AM production of ROIs after treatment with
IFNy; superoxide anion (0.3 ppm, 1 week) and H2O2 (0.1 ppm, 1 week)
- relative to untreated cells.
No effect on AM endotoxin-stimulated IL-lcc, IL-6, or
TNFoc production.

0.3 ppm: Increased lung: MIP-2, MCP-1, and eotaxin mRNA
expression.
1 f\ rvrvm • A -ftf^r A V\ in/^rv^ooi^/l Inner* A^TT5_0 A^f^T5_1 i^/~\tovin on/1
Cohen et al.
(2001)

Cohen et al.
(2002)






Cohen et al.
(1998)

Johnston et al.
(1999a)

                                                                 IL-6 mRNA expression.


                                                                 2.5 ppm: After 2 h, increased lung: MIP-2, MCP-1, eotaxin, and

                                                                 IL-6 mRNA expression.


                                                                 No exposure-related increases in lung IL-loc, IL-lp, IL-lRoc, IL-10,
O
o
o L0
H
O
^H
C
O
H
w i.o
O
w
r*
n
HH
H
W

4 h Mice cell line
(WEH1-3)



6 h Rat, male, SD,
200-250 g,
n= 3-6/group


IL-12, or IFNy mRNA expression.
Decreased binding of IFNy by WEHI-3 cells. Decreased superoxide
anion production by IFNy -treated cells; no similar effect on H2O2
production. Decreased IFNy -stimulated phagocytic activity. No effect
on IFNy-inducible la (MHC Class II) antigen expression.

Increased AM MIP-lcc, CINC, TNFcc, and IL-lp mRNA expression.
Induced increase in MIP-loc and CINC mRNA temporally inhibited
by cell treatment with anti-TNFa/TL-lp antibodies.



Cohen et al.
(1996)



Ishii et al.
(1997)




-------
Table AX5-7 (cont'd). Effects of Ozone on Lung Host Defenses
u£
S-
to
o
o
Concentration
ppm Duration Species


Effects


Reference

'~f> Cytokines, Chemokines: Production, Binding, and Inducible Endpoints (cont'd)











X
Y1
to



O
^
i-rj
H
6
o
n
x^*
H
O

O
H
W
O
o
1 — I
H
W
1.0 24h/day, Rat, male, Wistar,
3 days 8-12 weeks old


1.0 24 h Mice, male, C57B1/6J,
8 weeks old, n = 3/group
1.0 24 h Mice, male, C57B1/6J,
8 weeks old, n = 3/group



1.0 8 h/day, Mice (C57B1/6)
3 days (C57Bl/6Ai" NOS"'")
n= 3-10/group
1.0, 2.5 4 or 24 h Mice, male, C57B1/6J,
adult, n = 3/group


0.6, 2.0 3 h Mice, C57BL/6, Rat,
Wistar, 14-16 weeks
old, n = 4-6/group


0.8 3 h Rat, male, SD,
250-300 g, n = 5/group

0.8 3 h Mice, female,
B6J129SV,
C57B1/6X 129 NOS"',
8-16 weeks,
n= 3-12/group


BALF from exposed rats subsequently inhibited: ConA-stimulated
lymphocyte IFNy production, but had no effect on IL-2 production;
IL-2 -induced lymphoproliferation; and, IFNy -induced AM NO
production.
Increased lung: MIP-2 (4 h PE) and MCP-1 (4 and 24 h PE) mRNA
expression.
Increased lung MIP-2 and MCP-1 mRNA expression (4 and 24 h PE);
no effects on mRNA levels of IL-lcc, IL-lp, IL-lRoc, IL-6, MIF,
MTP-la, MIP-lp, eotaxin, or RANTES at either time point in recovery
period. Enhanced expressions of some cytokines/chemokines were
maintained longer than normal by coexposure to endotoxin.
Knockout (NOS"'") mice have more lavageable MIP-2 after exposure
than wild-type; both greater than control.

Dose-related increases in cytokine/chemokine induction. Increased
lung MIP-lcc, MIP-2, eotaxin (4 and 24 h), IL-6 (4 h only), and iNOS
mRNA expression.

Increased lung MIP-2 (4 h PE) and MCP-1 mRNA expression
(24 h PE); PMN and monocyte increased accumulation in lungs
consistent with sequential expression of the chemokines.
NF-kB activation also increased 20-24 h PE.

Increased ex vivo AM adherence to epithelial cells mitigated by cell
treatment with anti-TNFa or IL-loc antibodies.

Increased AM IFNy + LPS-induced NOS expression and NO
production, as well as induced PGE2 release. Knockout (NOS"'")
mice AM incapable of similar response to O3 - no inducible NO or
PGE2 above control levels.



Koike et al.
(1998, 1999)


Johnston et al.
(2001)
Johnston et al.
(2002)



Kenyon et al.
(2002)

Johnston et al.
(2000a)


Zhao et al.
(1998)



Pearson and
Bhalla (1997)

Fakhrzadeh
et al. (2002)






-------
                                   Table AX5-7 (cont'd). Effects of Ozone on Lung Host Defenses
to
o
o
X
H

6
o


o
H

O

O
H
W

O


O
HH
H
W
Concentration
ppm
Duration
Cytokines, Chemokines: Production,
0.8


2.0
0.08-0.25
OVA
Alveolar
0.3
0.8

0.8
3h


3h
and 1% 4 h,
3 times/week,
4 weeks
Macrophage/Lung NO- and
5 h/day,
5d/week,
4 weeks
3h

3h
Species
Effects
Reference
Binding, and Inducible Endpoints (cont'd)
Mice, female,
B6J129SV,
C57B1/6X 129 NOS ' ,
8-16 weeks,
n= 3-4/group
Rat, female, SD,
200-225 g,
n = 4-6/group
Mice, female, BALB/c,
C57BL/6,
6-8 weeks old,
n = 4-12/group
iNOS-Related Endpoints
Rat, male, Fischer F344,
200-250 g,
n = 10/group
Mice, female,
B6J129SV,
C57B1/6X 129 NOS ' ,
8-16 weeks,
n= 3-12/group
Mice, female,
B6J129SV,
C57B1/6X 129 NOS ' ,
8-16 weeks,
n= 3-4/group
Increased AM IFNy + LPS-induced NOS expression and NO
production.


Increased AM spontaneous and IFNy + LPS-induced NOS expression
and NO production. AM from exposed rats showed rapid
onset/prolonged activation of NF-KB.
O3 - dose-dependent increases in IgE, IL-4, IL-5; recruitment of
eosinophils and lymphocytes in BALB/c; O3 + OVA - increased IgG,
antibody liters, leukotrienes, airway responsiveness, immediate
cutaneous hypersensitivity reactions in BALB/c. In C57BL/6 only
O3 + OVA caused cutaneous hypersensitivity and altered IgG responses.

No effect on AM spontaneous, endotoxin-, or IFNy -stimulated,
NO formation.
Increased AM IFNy + LPS-induced NOS expression and NO
production and PGE2 release. Knockout (NOS"'") mice AM incapable of
similar response to O3 - no inducible NO or PGE2 above control levels.

Increased AM spontaneous and IFNy + LPS-induced NOS expression
and NO production. AM from exposed mice showed rapid and
prolonged activation of NF-KB, STAT-1 (expression, activity),
phosphoinositide 3-kinase, and protein kinase B.
Laskin et al.
(2002)


Laskin et al.
(1998a)
Neuhaus-
Steinmetz
et al. (2000)

Cohen et al.
(1998)
Fakhrzadeh
et al. (2002)

Laskin et al.
(2002)

-------
6-
(TO
c
S-
to
o
o
^yi










>
X
1



o
£
H
6
o
o
H
O

O
H
W
O
^
O
1 — I
H
W
Table AX5-7 (cont'd). Effects of Ozone on Lung Host Defenses

Concentration
ppm Duration

Alveolar Macrophage/Lung NO- and
1.0 8 h/day,
3 days

1.0 24 h/day,
3 days
1.0,2.5 4 or 24 h

2.0 3h


2.0 3h


3.0 6h


0.12, 0.5, or 2 3h

Surface Marker-Related Endpoints

0.8 3h



1.0 4h


1.0 2h






Species

iNOS-Related Endpoints
Mice, C57B1/6,
C57Bl/6Ar NOS"'-,
n= 3-10/group
Rat, male, Wistar,
8-12 weeks, n = 2/group
Mice, male, C57B1/6J,
adult, n = 3/group
Rat, female, SD,
200-250 g,
n= 1-3 /group
Rat, female, SD,
200-225 g,
n= 3-6/group
Rat, female, Brown
Norway, 250-300 g,
n = 4-8/group
Mice, female, BALB/c,
5-6 weeks


Rat, male, SD,
250-275 g,
n = 5-6/group

Mice cell line
(WEH1-3)

Rat, female, SD,
170-210 g,
n = 8-12/group




Effects

(cont'd)
Knockout (NOS"'") mice have more lavageable PMN, MTP-2,
and protein in lungs after exposure than wild-type.

BALF from exposed rats subsequently inhibited IFNy -induced
AM NO production.
Dose-related increase in lung iNOS mRNA expression.

Increased AM spontaneous, IFNy, and LPS-induced NO production,
as well as spontaneous and LPS-induced NOS expression. Effect
somewhat ameliorated by pretreatment with bacterial endotoxin.
Increased AM spontaneous and IFNy + LPS-induced NOS
expression and NO production. AM from exposed rats showed
rapid onset/prolonged activation of NF-icB.
Increased lung iNOS mRNA expression. Effect blocked by
pretreatment with dexamethasone.

Dose-dependent increases in nitrate and Penh; increases in nNOS, but
not iNOS or eNOS.


Increased expression of AM CD 1 Ib, but no effect on ICAM- 1 .



No effect on IFNy-inducible la (MHC Class II) antigen expression
on WEHI-3 cells.

Decreased expression of integrins CD 1 8 on AM and CD 1 Ib on PMN.
No effect on PMN CD62L selection.





Reference


Kenyon et al.
(2002)

Koike et al.
(1998, 1999)
Johnston et al.
(2000a)
Pendino et al.
(1996)

Laskin et al.
(1998b)

Haddad et al.
(1995)

Jang et al.
(2002)


Bhalla (1996)



Cohen et al.
(1996)

Hoffer et al.
(1999)




-------
                                           Table AX5-7 (cont'd). Effects of Ozone on Lung Host Defenses
to
o
o
Concentration
ppm
                              Duration
               Species
Effects
Reference
         Surface Marker-Related Endpoints (cont'd)
         1.0
         0.1,0.3
                              3 days
                              4 h/day,
                              5 days/week,
                              1 or 3 weeks
               Rat, male, Wistar and
               Fischer F344,
               8-10 weeks, n = 3/group

               Rat, male, Fischer F344,
               200-250 g
Increased expression of surface markers associated with antigen          Koike et al.
presentation: la (MHC Class II) antigen, B7.1, B7.2, and CDllb/c        (2001)
on BAL cells. Effect attributed to influx of monocytes.

Decreased expression of CDS among lung lymphocytes (0.1 ppm only;     Cohen et al.
3 weeks); effect exacerbated by stimulation of cells with IFNcc (but not     (2002)
with IL-lcc). Decreased expression of CD25 (IL-2R) on CD3+
lymphocytes (0.3 ppm only; 3 weeks); effect worsened by treatment
of cells with IL-lcc (0.1 and 0.3 ppm; 3 weeks).
X
H
6
o

o
H
O
o
H
W
O
O
HH
H
W
         NK- and Lymphocyte-Related Endpoints
         0.1,0.3
                              4 h/day,
                              5 days/week,
                              1 or 3 weeks
               Rat, male, Fischer F344,
               200-250 g
         0.4,0.8, 1.6
         1.0
12 h           Mice, male, BALB/c,
               6-8 weeks old,
               n = 5-8/group

24 h/day,       Rat, male, Wistar,
3 days         8-12 weeks old
Decreased expression of CDS among lung lymphocytes (0.1 ppm only;     Cohen et al.
3 weeks); effect exacerbated by stimulation of cells with IFNcc            (2002)
(but not with IL-lcc). Decreased expression of CD25 (IL-2R) on
CD3+ lymphocytes (0.3 ppm only; 3 weeks); effect worsened by
treatment of cells with IL-lcc (0.1 and 0.3 ppm; 3 weeks).

Lymphoproliferation: no effect on spontaneous or IL-2-inducible
forms; 0.1 ppm increased response to ConA mitogen (1 week only);
0.3 ppm - decreased response to ConA (1 week only).

Decreased pulmonary delayed-type hypersensitivity reactions to low       Garssen et al.
MW agents, likely via activation of TH2-dependent pathways.             (1997)
Lavage fluid from exposed rats subsequently inhibited ConA-stimulated   Koike et al.
lymphocyte IFNy production, but had no effect on IL-2 production;       (1999)
material also inhibited IL-2-induced lymphoproliferation.

-------
                                            Table AX5-7 (cont'd).  Effects of Ozone on Lung Host Defenses
S-
to
o
         Concentration
         ppm
Duration
Species
Effects
Reference
X
H
6
o

o
H
/O
o
H
W
O
O
HH
H
W
         Susceptibility Factors

         0.3                   24 to 72 h
                               4h
                               24 - 72 h
         0.1
2 h
               Mice
               C57BL/6J
               C3H/HeJ
               C3H/HeOuJ, 6-8 weeks
               old, n = 4-8/group

               CHO-K1 cell line SP-A
               Mice, male, C57BL/6J,
               C3H/HeJ, C3H/HeOuJ,
               6-8 weeks old,
               n = 5-16/group
Mice, male, C57BL/6,
6-8 weeks old
                         Lavageable protein concentration lowered by inhibition of iNOS and by   Kleeberger
                         targeted disruption o£Nos2; reduced Nos2 and Tlr4 mRNA levels in the   et al. (200 Ib)
                         O3-resistant C3H/HeJ mice.
Differences exist biochemically and functionally in SP-A variants.         Wang et al.
O3 exposure affects the ability of variants to stimulate TNFa and IL-8.     (2002)

Identified a candidate gene on chromosome 4, Toll-like receptor 4         Kleeberger
(Tlr4), a gene implicated in endotoxin susceptibility and innate            et al. (2000)
immunity. O3-resistant strain C3H/HeJ and C3H/HeOuJ (differing from
the O3-resistant strain by a polymorphism in the coding region of Tlr4)
were exposed, greater protein concentrations were demonstrated in the
OuJ strain.  Differential expression of Tlr4 mRNA with O3 exposure.
Suggests quantitative trait locus on chromosome 4 is responsible for a
significant portion of the genetic variance in O3 -induced lung
hyperpermeability; potential interaction between the innate and acquired
immune system.

Sensitized the mice to OVA by intratracheal instillation of OVA-pulsed    Depuydt et al.
dendritic cells (the principal antigen-presenting cells in airways).          (2002)
Created Th2 lymphocyte-dependent eosinophilic airway inflammation.
Groups of mice exposed to O3 during sensitization by OVA-pulsed
dendritic cells showed no modification of the allergic sensitization
process, whereas, previously sensitized mice exposed to O3,
demonstrated increases in allergen-induced airway inflammation.
Suggests that dendritic cells are an important component of O3-induced
eosinophilic airway inflammation.

-------
                                           Table AX5-7 (cont'd). Effects of Ozone on Lung Host Defenses
S-
to
o
         Concentration
         ppm
Duration
Species
Effects
Reference
X
         Susceptibility Factors (cont'd)

         0.1,0.5, and 1.0        2h
               Human lymphocytes
         NA
               RatBALF;
               Murine macrophage
               cell line (RAW 264.7)
                         Subsequent to O3 exposure, when lymphocytes were stimulated          Becker et al.
                        with pokeweed mitogen (PWM, a T-cell-dependent stimulus)            (1991)
                        or Staphylococcus aureus Cowan 1 strain (SAC, a T-cell-independent
                        stimulus), both B and T cells were found to be affected by O3
                        preexposure.  T cells also demonstrated an increase in IL-6 and
                        a decrease in IL-2, suggesting that O3 may have direct effects on
                        IgG-producing cells and concurrently an effect that is mediated
                        by altered production of T cell immunoregulatory molecules.

                        Both SP-A and SP-D  directly protected surfactant phospholipids          Bridges et al.
                        and macrophages from oxidative damage.  Both proteins blocked         (2000)
                        accumulation of TEARS and conjugated dienes generated during
                        oxidation of surfactant lipids or low density lipoprotein particles by
                        a mechanism that does not involve metal chelation or oxidative
                        modification of the proteins.
H
6
o

o
H
/O
o
H
W
O
O
HH
H
W
          AM = Alveolar macrophage; PE = Postexposure (i.e., time after O3 exposure ceased); MIP = macrophage inflammatory protein;
          PMN = Polymorphonuclear leukocyte; MLN = Mediastinal lymph node; CINC = cytokine-induce neutrophil chemoattractant;
          BAL = Bronchoalveolar lavage; DTPA = diethylenetriaminepentaacetic acid; ROI = reactive oxygen intermediate/superoxide anion;
          IFN = Interferon; BALT = Bronchus-associated lymphoid tissue; MCP = monocyte chemoattractant protein; CON A = Concanavalin A;
          OVA = Ovalbumin; SP-A = Surfactant Protein A; SP-D = Surfactant Protein D

-------
Table AX5-8. Effects of Ozone on Lung Permeability and Inflammation
•s
S-
to
o
o
u\










;>
X
Y1
oo


Concentration
ppm Duration

0.1 0. 5 h, in vitro
0.2
0.5
0.1 1 h, in vitro
0.2
0.4
1.0

0.2 23 h/day for 1 week
0.4
0.8


0.26 8 h/day, 5 days/week
for 1-90 days


Species

Rat, SD, primary alveolar
Type II cells

Guinea pig, male and
female, Hartley, and
human alveolar
macrophages

Guinea pigs, female
(Hartley), 260-330 g
n= 4-10/group


Mice, male (mast
cell-deficient and
-sufficient), 6-8 weeks old
n = 4-8/group
Effects

Decreased transepithelial resistance (Rt) after 0.5 ppm from 2 to
24 h PE and at 48 h in monolayers subjected to PMNs.
Significantly lower Rt after PMN treatment at 0.2 and 0.5 ppm.
Exposure of guinea pig alveolar macrophages to 0.4 ppm for
60 minutes produced a significant increase in IL-6 and TNFcc,
and an exposure of human alveolar macrophages to identical O3
concentration increased TNFcc, IL-lp, IL-6 and IL-8 protein
and mRNA expression.
Increase in B ALF protein and albumin immediately after
0.8 ppm exposure, with no effect of ascorbate deficiency in diet.
O3 -induced increase in B ALF PMN number was only slightly
augmented by ascorbate deficiency.

Greater increases in lavageable macrophages, epithelial cells and
PMNs in mast cell -sufficient and mast cell-deficient mice repleted
of mast cells than in mast cell-deficient mice. O3 -induced
permeability increase was not different in genotypic groups.
Reference

Cheek et al.
(1995)

Arsalane et al.
(1995)



Kodavanti
etal. (1995)



Kleeberger
etal. (200 Ib)


?P
Prj
H
6
2!
-^H
O
H
0
o
^^
H
W
O
o
1 — I
H
W
0.3






0.1
0.3
1.0

0.3
2.0


48 hand 72 h.
Exposures repeated
after 14 days




60 min



72 h
3h


Mice, male, C57BL/6J and
C3H/HeJ, 6-8 weeks old





Rat basophilic leukemia
cell line (RBL-2H3)


Mice, male and female,
C57BL/6J and C3H/HeJ


Greater B ALF protein, inflammatory cell and LDH response in
C57BL/6J than in C3H/HeJ after initial exposure. Repeated
exposure caused a smaller increase in B ALF protein and number
of macrophages, lymphocytes and epithelial cells in both strains,
but PMN number was greater in both strains of mice compared to
initial exposure.

O3 inhibited IgE- and A23 187 - indued degranulation.
Spontaneous release of serotonin and modest generation of
PGD2 occurred only under conditions that caused cytotoxicity.

Greater PMN response in C57BL/6J than in C3H/HeJ after acute
and subacute exposures. Responses of recombinant mice were
discordant and suggested two distinct genes controlling acute
and subacute responses. Genes termed Inf-1 and Inf-2.
Paquette et al.
(1994)





Peden and
Dailey (1995)


Tankersley and
Kleeberger,
(1994)


-------

 X
 H

 6
 o


 o
 H

O


 O
 H
 W

 O


 O
 HH
 H
 W
Concentration
ppm
0.3
0.3
1.0
2.5
Duration
48 h
24 or 48 h
I,2or4h
2, 4 or 24 h
Species
Mice, C57BL/6J and
C3H/HeJ, 6-8 weeks old
Mice, male, C57BL/6J,
8 weeks old
n = 3/group

Effects
Susceptibility to O3 is linked to a quantitative trait locus,
and TNFa is identified as a candidate gene.
0.3 ppm for 24 h caused increase in mRNA for eotaxin, MlP-la
and MIP-2.
1 ppm for 4 h caused increase in mRNA for eotaxin, MlP-la,
MIP-2, and IL-6.
2.5 ppm for 2 and 4 h caused increase in mRNA for MlP-la,
Reference
Kleeberger
etal. (1997)
Johnston et al.
(1999a)

                                                                  MIP-2 and IL-6 and metallothionein. Greater increases and

                                                                  lethality after 24 h.
0.3




0.4



0.15,0.3,
or 0.5

0.5





72 h




5 weeks



3h


4 h, 12-4 PM for
daytime and 7- 11
PM for nighttime
exposures. Exposures
repeated 16 h later.

Mice, male {HeJ, OuJ,
Nos2 (+/+) [C57BL/6J-
Nos2 (+/+)], and Nos2
(-/-) [C57BL/6J-Nos2
(-/-)]}, 6-8 weeks old
Guinea pigs, male,
(Hartley), 5 weeks old
(350-450 g)
n = 7-8/group
Rat, male, SD
6-8 weeks old,
n = 2-6/group
Rat, male, Wistar,
60-90 days old
n= 5-15/group



O3 induced permeability was decreased by pretreatment with a
nitric oxide synthase inhibitor and in animals with iNOS gene
knocked out.


Ovalbumin instillation in the nose caused an increase in
O3 -induced infiltration of eosinophils in nasal epithelium.


Time-related increase in permeability and inflammation, with a
peak at 8 h PE, after 0.5 ppm. No change following exposure to
0.15 or 3 ppm.
Significantly greater increase in IL-6, but not inflammation,
following a nighttime exposure compared to daytime exposure.
An initial nighttime exposure resulted in lesser inflammation
following a subsequent exposure. Pretreatment with IL-6 receptor
antibody abolished cellular adaptive response without affecting
inflammatory response induced by initial nighttime exposure.
Kleeberger
etal. (2001a)



lijima et al.
(2001)


Bhalla and
Hoffman
(1997)
McKinney
etal. (1998)





-------
                                Table AX5-8 (cont'd). Effects of Ozone on Lung Permeability and Inflammation
£
S-
to
o
o
Concentration
ppm Duration
0.5
1.0
2.0
Species
Rat, male, Fisher,
90 days old
n= 6-12/group
Effects
Increase in BALF protein and albumin occurred immediately
after 2 ppm exposure, and at 18 h after 1 ppm. No increase after
0.5 ppm. The movement of water and protein into airspace were
not coupled.
Reference
Cheng et al.
(1995)
X
to
o
         1.0-2.0
        0.5
        0.5
24 h following a 3-day
(6 h/day) exposure to
cigarette smoke
8 h during nighttime
Mice, C57BL/6, 6-8 weeks
old and rats, Wistar,
14-16 weeks old
n = 4-6/group


Mice, male, B6C3F1,
25 ± 2 g, 10 weeks old,
n = 6/group
Rat, male, Wistar, SD and
Fischer F344, 90 days old
n= 3-8/group
Steady state MCP-1 mRNA increase after 0.6 ppm, with maximal    Zhao et al.
increase after 2 ppm.  After 2 ppm, MIP-2 mRNA peaked at 4 h PE   (1998)
and MCP-1 mRNA peaked at 24 h PE. BALF neutrophils and
monocytes peaked at 24 and 72 h PE, respectively. BALF MCP-1
activity induced by O3 was inhibited by an anti-MCP-1 antibody.

BALF protein, neutrophils and lymphocytes were increased in       Yu et al.
animals exposed to smoke and then to O3. Macrophages from        (2002)
this group also responded with greater release of TNFa upon
LPS stimulation.

Exposure resulted in a significantly greater injury, inflammation     Dye et al.
and BALF levels of IL-6 in Wistar than in SD or F344 rats.          (1999)
£ 0.8
£
H
6
o

o
H
0
0
H
w 08
O
o
HH
H
m
2h and 6 h Rats, male, Fisher F344,
Juvenile (2 months;
180-250 g),
Adult (9 months;
370-420 g),
Old (18 months;
375-425 g), Senescent
(24 months; 400-450 g)
n = 2/group

3 h Rat, male, SD,
6-8 weeks old,
n = 5/group


Comparable effect on the leakage of alveolar protein in rats
of different age groups, but a greater increase occurred in
interleukin-6 and N-acetyl-beta-D-glucosaminidase in senescent
animals than in juvenile and adult rats.






Increased adhesion of macrophages from exposed animals to
rat alveolar type II epithelial cells in culture. Treatment with
anti-TNFcc + anti-IL-lcc antibody decreased adhesion in vitro,
but not permeability in vivo.

Vincent et al.
(1996).








Pearson and
Bhalla (1997)




-------
 X
 
-------
 X
 to
 to
 H

 6
 o


 o
 H

O


 O
 H
 W

 O


 O
 HH
 H
 W
Concentration
ppm Duration
1.0 8 h, assayed 1 and
2hPE
0.2
0.5 In vitro at liquid/air
1.0 interface
1.0 3h
1 2h
1 3h

1 3h

1 6h
0.5 4h
1.0
2.5
Species
Monkeys (Rhesus)
Primary TBE, BEAS-2b S
and HBE1
Rat, male, SD,
6-8 weeks old,
n = 4-5/group
Rats, female, SD,
170-210 g
n= 8-12/group
Rat, male, SD,
6-8 weeks old,
n = 5/group
Rat, male, SD,
250-275 g,
n = 6/group
Rat, male, SD,
200-250 g,
n= 3 -6/group
Mice, male
(129 wild-type or Clara
Cell Secretory Protein -/-),
2-5 months old,
n = 3/group
Effects
Increase in steady state IL-8 mRNA in airway epithelium.
Increase in IL-8 protein staining declined at 24 h after exposure.
Dose related increase in IL-8 release in the conditioned media.
Ozone produced greater toxicity in cell lines than in primary
cultures.
Time-related increase in B ALF protein, fibronectin (Fn), and
alkaline phosphatase (AP) activity. Fn mRNA detected in
macrophages, and AP in Type II cells and in B ALF PMNs from
exposed animals only.
The expression of CD 18 on alveolar macrophages and GDI Ib on
blood PMNs was lowered by exposure, but CD62L expression on
blood PMNs was not affected.
Time-related increase in B ALF albumin, PMNs, MTP-2 and
ICAM-1, and increase in MIP-2 mRNA only at early time point
in B ALF macrophages. MIP-2 mRNA not detected in lung tissue.
Ozone induced increase in BALF albumin, fibronectin and PMN
number was associated with an increase in expression of TNFcc,
IL-lcc, IL-6 and IL-10 mRNA. Pretreatment with anti-TNFcc
antibody caused downregulation of gene expression and reduction
of BALF albumin and PMN number, but not fibronectin.
Increase in number of macrophages with mRNA transcripts and
immunocytochemical staining of IL-1, TNFoc, MIP-2 and cytokine-
induced neutrophil chemoattractact (CINC). Chemokine activities
were reduced by treatment of macrophages with anti-IL-lp and
anti-TNFa antibodies.
Increases in IL-6 and metallothionein mRNA by 2 h after exposure
to 1 ppm. mRNA increases were further enhanced in CCSP -/-
mice.
Reference
Chang et al.
(1998)


Bhalla et al.
(1999)
Hoffer et al.
(1999)
Bhalla and
Gupta (2000)
Bhalla et al.
(2002)
Ishii et al.
(1997)
Mango et al.
(1998)

-------
                                  Table AX5-8 (cont'd).  Effects of Ozone on Lung Permeability and Inflammation
S-
to
o
         Concentration
         ppm            Duration
                       Species
                           Effects
                                                               Reference
X

o
H
W
O
O
HH
H
W
         1.0
         1.0
         1.2
         2.0
         2.0
         1.1
         0.32
8 h/night for three
nights
4h
6h
8h
48 h
(subacute)
3h
(acute)
Mice, (C57B1/6 wild-type
and iNOS knockout)
n= 3-10/group
Mice, male (129 strain,
wild-type and Clara Cell
Secretory Protein-
deficient), 2-3 mo old,
n = 3/group

Rat, male, Brown Norway,
200-250 g,
n = 4/group
                       Mice, male, C57BL/6J,
                       6-8 weeks old,
                       n= 5-10/group
                       Rat, female, SD,
                       6-8 weeks old
Rat, Wistar, -depleted of
neutrophils, 43 days old,
n= 9-10/group


Mice, male, C57BL/6J,
WT
TNRF1KO
TNRF2KO
6-8 weeks old,
n= 3-12/group
O3 exposure produced greater injury, as determined by
measurement of MIP-2, matrix metalloproteinases, total protein,
cell content and tyrosine nitration of whole lung protein, in iNOS
knockout mice than in wild-type mice.
                                                                                                                                          Kenyon et al.
                                                                                                                                          (2002)
Increases in abundance of mRNAs encoding eotaxin, MlP-la        Johnston et al.
and MIP-2 in CCSP-/-, but no change in wild-type mice.             (1999b)
Eotaxin mRNA expression in the lungs increased 1.6-fold            Ishii et al.
immediately after and 4-fold at 20 h.  Number of lavageable          (1998)
eosinophils increased 3- and 15-fold respectively at these time
points. Alveolar macrophages and bronchial epithelial cells
stained positively for eotaxin.

O3-induced increase in protein and PMNs in B ALF, and pulmonary   Longphre et al.
epithelial cell proliferation were significantly reduced in animals      (1999)
pre-treated with UK-74505, a platelet activating factor-receptor
antagonist.

BALF cells from exposed animals released 2 to 3 times greater       Pendino et al.
IL-1 and TNFoc, and greater fibronectin.  Immunocytochemistry       (1994)
showed greater staining of these mediators in lung tissue from
exposed rats.

Epithelial necrosis in the nasal cavity, bronchi, and distal airways.    Vesely et al.
Proliferation of terminal bronchiolar epithelial cells also decreased    (1999)
by O3 exposure, suggesting a role for neutrophils in the repair
process.

TNFR1 and TNFR2 KOs less sensitive to subacute O3 exposure       Cho et al.
than WT.  With acute exposures, airway  hyperreactivity was         (2001)
diminished in KO mice compared to WT mice, but lung
inflammation and permeability were increased.

-------
                                Table AX5-8 (cont'd). Effects of Ozone on Lung Permeability and Inflammation
•s
S-
to
o
o





Concentration
ppm Duration
0.3 24 to 72 h





Species
Mice
C57BL/6J
C3H/HeJ
C3H/HeOuJ
6-8 week old,
n= 5-16/group
Effects
Differential expression of Tlr4 mRNA.





Reference
Kleeberger
et al. (2000)




X
 0.4
H
H
6
0
2;
o
H
O

0
H
W
O 2
^
O
HH
W
1 or 5 days/ 12 h/day,
recovery period in
fresh
air of 5, 10, 15, or 20
days after the 5-day
preexposure




2 h, examined 2, 12,
and48hPE


Rat, Wistar, male,
7 weeks old,
n = 5/group







Rat, female, SD,
200-250 g
n = 4-7/group

Exposure for 5 days caused lower B ALF proteins, fibronectin,
IL-6, and inflammatory cells than animals exposed for 1 day.
Postexposure challenge with single O3 exposures at different time
points showed that a recovery of susceptibility to O3 (as measured
by B ALF levels of albumin, IL-6, and the number of macrophages
and neutrophils) occurred at -15-20 days, but total protein and
fibronectin levels remained attenuated even at 20 days post-5-day
exposure. The recovery with regards to BrdU labeling occurred
in 5-10 days after the 5 day exposure.

Adherence of neutrophils to pulmonary vascular endothelium was
maximal within 2 h after exposure and returned to control levels
by 12hPE.

Van Bree
et al. (2002)








Lavnikova
etal. (1998)



-------
                                 Table AX5-8 (cont'd).  Effects of Ozone on Lung Permeability and Inflammation
S-
to
o
         Concentration
         ppm            Duration
                       Species
                          Effects
                                                              Reference
X

o
H
W
O
O
HH
H
W
         1.0 or 2.5
         0.11
         0.1-2
4,20, or 24 h,
examined immediately
PE
10 min
                         3 days, continuous
3h
24/h day for up to
3 days, assays
immediately or at 16 h
PE
3 h, assayed 6 h PE
(exposure-response)
3 h, assayed 0, 2, 6, or
24 h PE (time-course)
Mice, C57B1/6J, 36 hand
8 week old
Endotoxin(10 ng)
n = 3/group
                       Rat, male, SD,
                       30 days old,
                       n = 6/group
Mice, female, C57BL/6J
CBA C3H/HeJ AKR/J
SJL/J, 6-8 weeks old,
n = 4-7/group
Mice,
C57BL/6,
n = 2-4/group
RPA for IL-12, IL-10, IL-lcc, IL-lp, IL-IRa, MIF, IFNy, MIP-lcc,    Johnston et al.
MIP-2,IL-6, and Mt. Newborn mice: increased Mt mRNA only.      (2000b)
8-week-old mice:  increased MIP-lcc, MIP-2, IL-6, and Mt mRNA.

Both age groups had similar cytokine/chemokine profiles with
endotoxin exposure, suggesting that the responses to endotoxin,
which does not cause epithelial injury, and the responses to O3,
which does, demonstrate that differences in inflammatory control
between newborn and adult mice is secondary to epithelial injury.

General dietary restriction to 20% of the freely-fed diet for           Elsayed (2001)
60 days caused an extreme reduction in body weight and higher
survivability.  Levels of antioxidants and detoxifying enzymes
increased less than in freely fed animals.

 Both exposure levels caused a transient increase in CC16 in         Broeckaert
serum that correlated with BALF changes in protein, LDH, and      et al. (2003)
inflammatory cells. Inverse relationship between preexposure
levels of CC16 in BALF and epithelial damage based on serum
CC16 levels and BALF markers of inflammation. Inverse
relationship between preexposure levels of albumin in BALF
and lung epithelium damage.  Suggests that a major determinant
of susceptibility to O3 is basal lung epithelial permeability.
C57BL/6J mice had lower levels of CC16a (the more acidic form)
than C3H/HeJ. Both the strains had similar levels of CC16b.
Suggests that basal lung epithelial permeability is a major
determinant of susceptibility; greater epithelial permeability
observed in C57BL/6J may be due to difference in the expression
of CC16a and possibly other antioxidant/inflammatory proteins.

O3 > Ippm increased MIP-2 mRNA and recruitment of neutrophils.    Driscoll et al.
MIP-2 increase was immediate and decreased to control by           (1993)
24hPE.

-------
                                 Table AX5-8 (cont'd).  Effects of Ozone on Lung Permeability and Inflammation
to
o
o
X

-------
                                  Table AX5-8 (cont'd).  Effects of Ozone on Lung Permeability and Inflammation
S-
to
o
         Concentration
         ppm             Duration
Species
Effects
Reference
                          3h
Mice, male and female,
C57BL/6J, assayed
immediately or 3, 6, 9,
or21hPE
Increase in tissue expression of 1C AM-1 3 -9 hPE, remaining
until 21 h PE. Bronchioles and terminal bronchiole/alveolar duct
regions: Enhanced ICAM-1-IR 0 - 3 h PE, returning to baseline
by 21 and 9 h, respectively. Lung parenchyma: maximal
ICAM-1 expression and PMN influx concurrent 3 h, followed by
transepithelial migration of PMNs to the airway lumen. Suggests
regional variations in airway inflammatory activity; upregulation
of ICAM-1 may play a role in local regulation of PMN influx to
the airways after acute exposure.
Takahashi
etal. (1995a)
                          4h
X

o
H
W
O
O
HH
H
W
                          2h
Rat, SD, male, 225-250 g,
treated with 10 mg/kg
ebselen every 12 h from
1 h before O3 exposure,
n = 4/group
Human transformed
bronchial epithelial cells
(16-HBE)
Guinea pigs, Hartley,
male,
450-550 g,
n= 3-5/group
Ebselen significantly decreased pulmonary inflammation (albumin
and PMN in BALF) 18 h PE without altering AM expression of
iNOS.  Ebselen inhibited the nitration reaction of tyrosine residues
and enhanced expression of Cu-, Zn-,and Mn SOD.  Suggests that
ebselen scavenges peroxynitrite during Cyinduced inflammation
and may protect against acute lung injuries by modulating the
oxidant-related inflammatory process.

NO donors increased IL-8 production dose-dependency.
TNFoc plus IL-lp plus INFy increased IL-8 in culture
supernatant of epithelial cells.  NOS inhibitors
(aminoguanidine plus NG-nitro-L-arginine methyl ester)
attenuated the cytokine-induced IL-8 production.

O3 induced AHR to acetylcholine and increased PMN in BALF.
Persisting for 5 h.  Pretreatment with NOS inhibitors did not affect
AHR or PMN accumulation 0 h PE, but, inhibited at 5 h PE.

Suggests that endogenous NO, through upregulation of IL-8,
modulates O3-induced airway inflammation and AHR.
Ishii et al.
(2000a)
Inoue et al.
(2000)

-------
fa                               Table AX5-8 (cont'd).  Effects of Ozone on Lung Permeability and Inflammation
«j
^       Concentration
g       ppm            Duration              Species                   Effects                                                      Reference
i^f\      	
         0.12,            3h                    Mice, female, BALB/c,      O3 exposure caused dose-dependent increases in nitrate (indicative    Jang et al.
         0.5,                                   5-6 weeks old             of in vivo NO generation).  Increases in enhanced pause (Penh)        (2002)
         1, or                                                           were also dose-dependent.  Increases in NOS-1, but not in
         2                                                              NOS-3 or iNOS isoforms.  Suggest that NOS-1 may induce
                                                                        airway responsiveness by neutrophilic airway inflammation.


         PMN = Polymorphonuclear leukocyte
         PE = Postexposure (time after O3 exposure ceased)
         BAL = Bronchoalveolar lavage
         BALF = Bronchoalveolar lavage fluid


>
X

-------
                               Table AX5-9. Effects of Ozone on Lung Structure:  Acute and Subchronic Exposures
_-
OJ
to
O


Concentration
ppm
0.1
0.5
1.0
0.2
0.4
Duration
8 h/ day x 1 day
8 h/day x 1 day
8 h/day x 1, 10,
75, and 90 days
3, 7, 28, and
56 days; 3-, 7-,
Species
Rat; male, SD,
350-600 g,
n = 3-6/group
Mice, male, NIH,
Rat, male, Wistar
Effects
No dose-related response on CYP2E1, one of six P450 enzymes identified
in respiratory tissue. CYP2E1 activity was elevated (250% and 280%)
in the lobar bronchi/major daughters airways immediately after 1.0 ppm O3
exposure for 1 day and 10 days, respectively, but not in the trachea or distal
bronchioles; CYPE1 activity was unchanged and decreased after 1.0 ppm
O3 exposure for 75 and 90 days, respectively.
Concentration-related centriacinar inflammation, with a maximum after
3 days of exposure; number of alveolar macrophages and pulmonary cell
Reference
Watt et al.
(1998)
Dormans
et al. (1999)
X
to
VO
O
HH
H
W
0.2
0.4
0.8
                         and 28-day
                         recovery from
                         28 days of
                         exposure
23 h/day for
7 days
                                   RIV:Tox
                                   Guinea pig,
                                   male, Hartley
                                   Crl:(HA)BR,
                                   7 weeks old,
                                   n= 3-9/group
Guinea pig, female,
Hartley; ±AH2 diet
density increased progressively until 56 days of exposure, with the guinea
pig the most sensitive species.  Concentration and exposure-time dependent
hypertrophy of bronchiolar epithelium in mouse only. Exposure to 0.2 ppm
for 3 and 7 days caused significant histological and morphometric changes
in all 3 species; exposure for 56 days caused alveolar duct fibrosis in rat
and guinea pigs.  Total recovery in rats after 28-day exposure, but not in
guinea pigs or mice.

Treatment-related lesions were observed after exposure to 0.4 and 0.8 ppm
O3; lesions were primarily seen in the terminal bronchioles and consisted
of mononuclear cell and neutrophilic infiltrate and thickening of the
peribronchiolar interstitium.  Effects were only marginally exacerbated
by the AH2 (ascorbic acid) deficient diet and lesions were resolved after
1 week in FA.
Kodavanti
et al. (1995)
> 0.4
-ij
H
6
o
!2
O
H
O
O
H
W
O
12 h/day;
1- or 7-day
exposure








Rat, Wistar
RiV:TOX,
male and female,
1, 3, 9, & 18 months
of age,
n = 5-6/group





Centriacinar inflammation (increased alveolar macrophages and PMNs;
increased proximal and ductular septal density) was greatest in young rats
(1 month and 3 months for 1-and 7-day exposures, respectively) and
decreased with age. No major gender differences were noted.







Dormans
et al. (1996)










-------
                         Table AX5-9 (cont'd).  Effects of Ozone on Lung Structure:  Acute and Subchronic Exposures
to
o
o
Concentration
ppm           Duration
                  Species
                     Effects
Reference
>
X
(Si
        0.4
        1.0
2h
Monkey;
adult male Rhesus
                                                       Reduced glutathione (GSH) increased in the proximal intrapulmonary
                                                       bronchus after 0.4 ppm O3 and in the respiratory bronchiole after 1.0 ppm
                                                       O3.  Local O3 dose (measured as excess 18O) varied by as much as a factor
                                                       of three in different airways of monkeys exposed to 1.0 ppm, with
                                                       respiratory bronchioles having the highest concentration and the parenchyma
                                                       the lowest concentration. After exposure to 0.4 ppm, the O3 dose was 60%
                                                       to 70% less and epithelial injury was minimal, except in the respiratory
                                                       bronchiole, where cell loss and necrosis occurred, but was 50% less than
                                                       found at 1.0 ppm.
Plopper et al.
(1998)
0.5 8h
+ BrdU to label
epithelial cells


0.5 8 h/day,
3 or 5 days;
+ fluticasone
propionate (FP)
intranasally



0.5 8 h/day,
3 days
+ endotoxin
(100 ug/mL)
intranasally
Rat, male, Fischer
F344, 13 weeks old,
n = 6/group


Rat, male, Fischer
F344, 203-232 g,
n = 6/group





Rat, male, Fischer
F344/N Hsd, 12
weeks old,
n = 6/group

O3 exposure induced a transient influx of neutrophils and a significant (17%)
loss of NTE cells 2-4 h after exposure. Increased epithelial DNA synthesis
was first detected 12 h PE. LI and ULLI indices of epithelial cell DNA
synthesis were greatest 20-24 h and still elevated 36 h PE; numeric density
of NTE cells returned to control levels 20-24 h PE.
No significant difference of fluticasone propionate on morphometry of the
maxilloturbinates; O3 exposure caused neutrophilic rhinitis with 3.3- and
1.6-fold more intraepithelial neutrophils (3-day and 5-day exposure,
respectively) and marked mucous cell metaplasia (5-day exposure only)
with numerous mucous cells and approximately 60 times more
intraepithelial mucosubstances in the nasal transitional epithelium; FP-
treated rats exposed to O3 had minimal nasal inflammation and mucous
cell metaplasia.
Endotoxin-induced neutrophilia in nasal mucosa with NTE; mucous cell
metaplasia was not detected in air/endotoxin-exposed rats, was observed
in O3/saline-exposed rats, and was most severe in O3/endotoxin-exposed
rats.

Hotchkiss
et al. (1997)



Hotchkiss
et al. (1998)






Fanucchi
et al. (1998)




-------
                         Table AX5-9 (cont'd).  Effects of Ozone on Lung Structure:  Acute and Subchronic Exposures
•s
S-
to
o
o
^






Concentration
ppm Duration

0.5 8 h/day,
1, 2, or 3 days
+ BrdU to label
epithelial cells
+ antirat
neutrophil
antiserum
Species

Rat, male, Fischer
F344/N, 10-12 weeks
old, n = 6-8/group




Effects

Acute O3 exposure induced a rapid increase in rMuc-5AC mRNA levels
prior to the onset of mucous cell metaplasia; neutrophilic inflammation
coincided with epithelial DNA synthesis and upregulation, but was
resolved when mucous cell metaplasia first appeared in the NTE.

Maxilloturbinates lined with NTE determined the epithelial labeling
index, numeric densities of neutrophils, total epithelial and mucous secretory
Reference

Cho et al.
(19993, 2000)





>
X
        0.5
8 h/day,
3 days
+ endotoxin
Rat, male, Fischer
F-344, 10-12 weeks
old, n = 6/group
cells, amount of stored intraepithelial mucosubstances, and steady-state
ratMUC-5AC (mucin) mRNA levels.  Four days after a 3-d exposure,
antiserum-treated, O3-exposed rats had 66% less stored intraepithelial
mucosubstances and 58% fewer mucous cells in their NTE than did controls.
Antiserum treatment had no effects on O3-induced epithelial cell
proliferation or mucin mRNA upregulation.

Enhanced epithelial lesions in the NTE and respiratory epithelium of the
nose and conducting airways by endotoxin and O3 exposures, respectively;
synergistic effects of coexposure mediated by neutrophils.  Endotoxin
increased rMuc-5 AC mRNA levels in the NTE of O3-exposed rats;
neutrophil depletion, however, had no effect on endotoxin-induced
upregulation of mucin gene mRNA levels. Endotoxin enhanced the
O3-induced increase in stored mucosubstances (4-fold increase),
but only in neutrophil-sufficient rats.
Wagner et al.
(2001a,b)
O 0.5
!>
trj
H
6
o
§ '
^^
H
O
c|
o
H
W
O
O
H
W
8 h/day,
1 and 3 days
+ OVA
(1%, 50 uL/nasal
passage)
8h









Rat, male, Brown
Norway, 10-12 weeks
old, n = 6/group


Rat, male, SD,
10 weeks old,
Ferret, male, 18
months old,
Monkey, male,
Rhesus, 4 years old,
n = 4-8/group



O3 enhanced the appearance of eosinophils in the maxilloturbinates of
OVA-challenged rats but did not increase inflammation in other nasal
tissues; O3/OVA coexposures for 3 days increased the number of epithelial
cells as well as the appearance of mucus-containing cells in the NTE lining
the maxilloturbinates.
Severe, acute infiltration of neutrophils along with necrotic bronchiolar
epithelium in all lung regions, especially in the centriacinar region; necrosis
and inflammation was more severe in ferrets and monkeys than in rats.







Wagner et al.
(2002)



Sterner-Kock
et al. (2000)









-------
                          Table AX5-9 (cont'd).  Effects of Ozone on Lung Structure:  Acute and Subchronic Exposures
to
o
o
>
X
(Si
I
to
Concentration
ppm            Duration
Species
Effects
Reference
0.5 8 h/day for 3 days,
assayed 2h or
4 days PE





Rat, male, Fischer
F344/N, with or
without prior
exposure to
100 ug/day
endotoxin,
10-12 weeks old,
n = 8/group
2 h PE: Endotoxin/O3 rats had 48 and 3 times more PMNs in the NTE than
did saline/air- and saline/O3-exposed rats, respectively at 2 h PE. O3-only
rats had 35% more NTE cells and 2-fold more mucin mRNA than did
saline/air-exposed rats.

4 days PE: Endotoxin/O3 rats had 5 and 2 times more IM and mucous cells,
respectively, than did saline/air- and saline/O3- rats.

Cho et al.
(1999b)






                        8 h/day for
                        90 days
                                   Rats, SD, male,
                                   275-300g, treated
                                   i.p. with
                                   1 -nitronaphthalene
                                   (0, 50, or 100 mg/kg)
                      No mucous cell metaplasia was present in those rats killed at 4 days
                      postexposure. Suggests that pre-existing rhinitis augments O3-induced
                      mucous cell metaplasia.

                      1-nitronaphthalene (a pulmonary toxicant requiring metabolic activation)-
                      treated rats exposed to O3 showed greater histopathologic and morphometric
                      effects in the centriacinar region of the lung. Caused denudation of the
                      basement membrane and necrosis of remaining epithelial cells. Increased
                      severity of ciliated cell toxicity in O3-exposed rats.  No differences in the
                      intrapulmonary airways or trachea in sensitivity to 1-nitronaphthalene,
                      suggesting a site-selective synergy between O3 and 1-nitronaphthalene.
                                                                      Paige et al.
                                                                      (2000b)
         AM = Alveolar macrophage
         PE = Postexposure (i.e., time after O3 exposure ceased)
         LM = Light microscopy
         EM = Electron microscopy
                                                         RB = Respiratory bronchiole
                                                         TB = Terminal bronchiole
                                                         IAS = Interalveolar septum
                                                         PMN = Polymorphonuclear leukocyte

-------
6-
fin
c
S-
to
o
Ul













;>
X
OJ
OJ



o
£>
l-rj
H
6
o
21
o
H
O
O
H
W
O
Table AX5-10. Effects of Ozone on Lung Structure: Subchronic and Chronic Exposures





Concentration
ppm

Mexico City
Ambient:
0.018
(>0.12for
18 1-h
intervals)
0.12
0.5
1.0









0.12
0.50
1.0





0.12
1.0






Duration

23 h/day for
7 weeks




6 h/day,
5 days/week
for 20 months









6 h/day,
5 days/week
for 24 and
30 months




6 h/day,
5 days/week,
for 2 or
3 months




Species

Rat, male and
female, Fischer
F344,
8 weeks old


Rat, male,
Fischer F344,
6-8 weeks old









Mice, male and
female, B6C3F1,
6-7 weeks old,
n = 50/group




Rat, male,
Fischer F344/N,
4-5 weeks old,
n = 4/group




Effects

No inflammatory or epithelial lesions in nasal airways or respiratory tract.





LM morphometry of CAR remodeling. Thickened tips of alveolar septa
lining ADs (alveolar entrance rings) 0.2 mm from TB in rats exposed to 0. 12 ppm
and to 0.6 mm in rats exposed to 1.0 ppm. At 0.5 and 1.0 ppm, atrophy of nasal
turbinates, mucous cell metaplasia in NTE, increased volume of interstitium and
epithelium along ADs due to epithelial metaplasia, and bronchiolar epithelial
hyperplasia. At 1.0 ppm, increased AMs and mild fibrotic response (increase in
interstitial matrix and cellular interstitium; the latter due to increase in volume
in interstitial fibroblasts). More effects in PAR than in terminal bronchioles.
Effects not influenced by gender or by aging. Effects similar to, or model of,
early fibrotic human disease (e.g., idiopathic pulmonary fibrosis).


Effects in the nose and centriacinar region of the lung at 0.5 and 1.0 ppm.
Nasal lesions were mild: hyaline degeneration, hyperplasia, squamous
metaplasia, fibrosis, suppurative inflammation of transitional and respiratory
epithelium; and atrophy of olfactory epithelium. Lung lesions:
alveolar/bronchiolar epithelial metaplasia and histiocytosis in terminal
bronchioles, alveolar ducts, and proximal alveoli. Severity was greatest in
mice exposed to 1.0 ppm O3, but there was minimal interstitial fibrosis.

Morphometric changes (epithelial thickening, bronchiolarization) occurred after
2 or 3 months exposure to 1.0 ppm O3; effects were similar to those found with
20 months exposure (see Pinkerton et al., 1995)





Reference

Moss et al.
(2001)




Catalano et al.
(1995a,b);
Chang et al.
(1995);
Harkema et al.
(1994, 1997a,b)
Pinkerton et al.
(1995);
Plopper et al.
(1994);
Stockstill et al.
(1995)
Herbert et al.
(1996)






Pinkerton et al.
(1998)






O
HH
H
W

-------
6-

Table AX5-10 (cont'd). Effects of Ozone on Lung Structure: Subchronic and Chronic Exposures




w Concentration
to ppm
o
^ 0.011

0.25
0.5



0.4





^
X
1
OJ
0.5


o

rrt
H
6
o
0
H
O
o
H
W
O
Duration

6 months

8 h/day,
7 days/week
for 13 weeks


23.5 h/day
for 1, 3, 7, 28,
or 56 days






8 h/day for
1, 3, and
6 months
8 h/day for
5 days,
every 5 days
for a total of
1 1 episodes








Species

Rat, male,
Wistar, 2 months
Rat, male,
Fischer F344/N
HSD,
10-14 weeks old
n = 5-6/group
Rat, Wistar,
7 weeks old,
n = 5/group






Rat, male,
Fischer F344/N

Monkey;
Rhesus,
30-day-olds,
n = 6/group









Effects

CO = 1.25 ppm; PM = 35.18 ug/m3; SO2 = 29.05 ug/m3. Induction of secretory
hypertrophy, acidic mucous secretion, and ciliary damage.
Mucous cell hyperplasia in nasal epithelium after exposure to 0.25 and 0.5 ppm
O3; still evident after 13 weeks recovery from 0.5 ppm O3 exposure. Mucous cell
metaplasia found only after 0.5 ppm O3, but still detectable 13 weeks PE.


Acute inflammatory response (increased PMNs and plasma protein in B ALF)
reached a maximum at day 1 and resolved within 6 days during exposure;
AMs in B ALF increased progressively up to day 56, and slowly returned to near
control levels with PE recovery. Histological examination and morphometry
of the lungs revealed CAR inflammatory responses throughout O3 exposure;
thickening of septa was observed at day 7. Ductular septa thickened progressively
at days 7, 28, and 56 of exposure; showed increased collagen at day 28, which
was further enhanced at day 56. Increased RBs with continuous exposure.
Collagen and bronchiolization remained present after a recovery period.
Increased Bcl-2, a regulator of apoptosis, after 1 month, decreasing somewhat
thereafter, returning to baseline by 13 weeks PE; increased number of metaplastic
mucous cells in NTE after 3 and 6 months.
Increased density and distribution of goblet cells in RB whole mounts stained
with AB/PAS; extensive remodeling of distal airway with O3 and O3 + HDMA
challenge; increased airways resistance and reactivity, and respiratory motor
adaptation also occurred. Authors conclude that periodic cycles of acute injury
and repair associated with the episodic nature of environmental patterns of O3
exposure alters postnatal morphogenesis and epithelial differentiation in the
distal lung of infant primates.






Reference

Lemos et al.
(1994)
Harkema et al.
(1999)



Van Bree et al.
(2002)







Tesfaigzi et al.
(1998)

Schelegle et al.
(2003a);
Chen et al.
(2003);
Plopper and
Fanucchi (2000)







O
HH
H
W

-------
                       Table AX5-10 (cont'd). Effects of Ozone on Lung Structure: Subchronic and Chronic Exposures
•s
S-
to
o
o
01












|>
X
&
Concentration
ppm Duration

0.8 8 h/day for
90 days
+ 1-NN
(100 mg/kg)
0.5 11 episodes of
5 days each,
8 h/day
followed by
9 days of
recovery
0.5 11 episodes of
5 days each,
8 h/day
followed by
9 days of
recovery
Species

Rat, male, SD,
275-301 g


Monkey,
Macaca mulatta,
30 days old



Monkey,
Rhesus,
30 days old,
n = 6/group


Effects

Increased O3 -induced centriacinar toxicity (histopathology, TEM, morphometry)
of 1-Nitronaphthalene (1-NN), a pulmonary cytotoxicant requiring metabolic
activation, especially to ciliated cells.

In small conducting airways O3 caused decrements in density of airway epithelial
nerves. Reduction greater with HDMA + O3. O3 or HDMA + O3 caused increase
in number of PGP 9.5 (pan-neuronal marker) in airway. CGRP-IR nerves were
in close contact with the PGP9.5 positive cells. Appearance of clusters of
PGP9.5+/CGRP cells. Suggests episodic O3 alters developmental pattern of
neural innervation of epithelial compartment.
Abnormalities in the BMZ included: (1) irregular and thin collagen throughout
the BMZ; (2) perclecan depeleted or severely reduced; (3) FGFR-1
immunoreactivity was reduced; (4) FGF-2 immunoreactivity was absent in
perlecan-deficient BMZ, but was present in the lateral intercelluar space (LIS),
in basal cells, and in attenuated fibroblasts; (5) syndecan-4 immunoreactivity
was increased in basal cells.
Reference

Paige et al.
(2000b)


Larson et al.
(2004)




Evans et al.
(2003)




H
6
o

o
H
O
O
H
W
O
O
HH
H
W
          TB = Terminal bronchiole
          PE = Postexposure (i.e., time after O3 exposure ceased)
          AM = Alveolar macrophage
          LM = Light microscopy
          TEM = Transmission Electron Microscopy
          BMZ = Basement Membrane Zone
EM = Electron microscopy
RB = Respiratory bronchiole
IAS = Interalveolar septum
C x T = Product of concentration and time

-------
                                                Table AX5-11. Effects of Ozone on Pulmonary Function
S-
to
o
         Concentration
         ppm
Duration
Species
Effects
                                                                        Reference
>
X
H
6
o

o
H
/O
o
H
W
O
^
O
HH
H
W
         0.5
         0.3
6 or 23 h/day
over 5 days
                         3 h, assayed 1
                         and24hPE
Rats, male,
Fischer 344,
90 days old,
n = 28-36/group,
ambient temperature
10, 22, or 34 °C
                   Mice, male,
                   C57BL/6J C3H/HeJ,
                   53 days old,
                   n = 5-7/group
                   challenged by CO2
                   (5 or 8%)
48 and 72 h, with    Mice,
re-exposure after    C57BL/6J,
14 days of          C3H/HeJ,
recovery            6-8 weeks old
2-3 h, assayed
OhPE
Mice, male,
C57BL/6J,
C3H/HeJ,
4 or 11-12 weeks old,
n = 4-6/group
Toxicity increased with decreases in temperature.  At 10 °C: decreased body
weight, total lung capacity, BALF protein, alkaline phosphatase activity,
% PMN, and lysozyme. Ozone-induced changes in lung volume were
attenuated during the 5 exposure days and returned to control levels after
7 days recovery.  The responses to repeated O3 exposure in rats were
exacerbated by reduced ambient temperature, presumably as a result of
increased metabolic activity.

C57BL/6J mice:  CO2-induced changes in VE were attenuated 1 h after O3
exposure; VT was reduced 1 h after O3 exposure; the diminished VT 1 h
after O3 was coincident with reduced/ mean inspiratory flow, and slope of
VE-to-%CO2 relationship compared with FA.; VE partially reversed 24 h after
O3 relative to FA.

C3H/HeJ:  VT was reduced 1 h after O3 exposure; increased/to sustain the
hypercapnic VE response similar to air exposure.

Suggests that control of ventilation during response to CO2 is governed,
in part, by genetic factors in these two strains of mice, implying differential
O3 susceptibility.

VE and/were measured before and immediately after exposure. Normocapnic
VE was greater following subacute O3 exposure in C57BL/6J mice than in
C3H/HeJ mice, due to increased/and reduced VT, respectively. Ventilatory
responses to both normocapnia and hypercapnia were similar after O3
reexposure in both strains.  Suggests that:  increased VT in C57BL/6J mice
may contribute to the increased susceptibility to lung injury due to a greater
dose of O3 reaching the lower lung; mechanistic separation of airway
inflammation and ventilation.

Using 18O-labeled O3.  C3H/HeJ mice had 46% less 18O in lungs and 61% less
in trachea, than C57BL/6J. C3H/HeJ mice had a greater body temperature
decrease following O3 exposure than C57BL/6J mice. Suggests that the
differences in susceptibility to O3 are due to differences the ability to decrease
body temperature and, consequently decrease the dose of O3 to the lung.
                                                                        Wiester et al.
                                                                        (1996)
                                                                                              Tankersley
                                                                                              etal. (1993)
                                                                                              Paquette
                                                                                              etal. (1994)
                                                                        Slade et al.
                                                                        (1997)

-------
                                          Table AX5-11 (cont'd).  Effects of Ozone on Pulmonary Function
S-
to
o
        Concentration
        ppm            Duration
                                           Species
Effects
                                                                        Reference
X
                        3 h, assayed 6 h
                        after exposure
                                           Mice, male and
                                           female, AKR/J,
                                           C3H/HeJ, CBA/J,
                                           129/J, NJ, C57BL/6J,
                                           C3HeB/FeJ, SJL/J
Measured trachea! transepithelial potential in the six strains and in progeny of
B6 and C3 strain mice. Fl mice and second generation backcrosses with the
resistant parent were O3 resistant. Ratios of 1:1 (resistant: susceptible) were
obtained with second generation backcrosses with the susceptible parent,
suggesting simple autosomal recessive inheritance of susceptibility.
                                                                 Susceptible phenotype:  129/J, A/J, B6,C3HeB/FeJ, and SJL/J.

                                                                 Resistant phenotype: AKR/J, C3, and CBA/J.

                                                                 Different pattern of susceptibility than with inflammation, suggesting that
                                                                 the responses are controlled by disparate genetic factors.

                                           Mice, C3H/HeJ, A/J,    Used whole body plethysmography and enhanced pause index (Penh)
                                           C57BL/6J, 129/SvIm,   evaluations.
                                           CAST/Ei, BTBR,
                                           DBA/2J, FVB/NJ,      C57BL/6J, BALB/cJ, 129/Svlm, BTBR: were highly sensitive to O3;
                                           B ALB/cJ,             exhibited significant increases in Penh to MCh at 6 and 24 h after exposure
                                           n = 6-24/group         to O3.

                                                                 DBA/2J, A/J, FVB/NJ, CAST/Ei, C3H/HeJ: increases in sensitivity to MCh
                                                                 at 6 h after exposure, return to near baseline by 24 h after exposure to O3.
                                                                        Takahashi
                                                                        et al. (1995b)
                                                                                                                                        Savov et al.
                                                                                                                                        (2004)
H
6
o

o
H
/O
o
H
W
O
O
HH
H
W
        VE = Minute ventilation
        VT = Tidal volume
        /= Frequency of breathing
        FA = Filtered air
        MCh = Methacholine

-------
                   Table AX5-12.  Effects of Ozone on Airway Responsiveness
•s
S-
to
o
o














X

oo



Concentration
ppm

0.1
0.3





0.15
0.30
0.60
1.2
0.3




0.5



Exposure
Duration

4 h/day,
4 days/week
for 24 weeks




4h



4 h/day for 1,
3, 6, 12, 24,
or 3 8 days


8 h/day for 5 days,
repeated every
14 days for
6 months
Species, Sex, Strain,
and Age

Guinea pig, male and
female, Hartley,
200-250 g,
n = 10-20/group



Guinea pig, male
Hartley,
500-600g,
n = 5-8/group
Guinea pig, male
Hartley,
500-600 g,
n = 6-7/group

Rhesus monkey,
male, 30 days old,
n = 6/group

Observed Effect(s)

O3 exposure did not produce airway hyperresponsiveness to ACh in
nonsensitized animals; in OVA-sensitized animals, there was increased
responsiveness to both nonspecific (ACh) and specific (OVA) airway
challenge that persisted for 4 weeks after exposure 0. 1 and 0.3 ppm O3.
Effects were not gender specific and were not associated with BALF
inflammatory indicators, but were associated with antigen-specific
antibodies in blood.
Increased airway responsiveness to Hist, but not ACh, 16-18 h after
1.2 ppm O3 exposure only. Increased responsiveness to SP occurred after
exposure to > 0.3 ppm O3.

Increased airway responsiveness to SP occurred 16-18 h after exposure
to 0.3 ppm O3 for 1, 3, 6, 12, and 24 days; but not after 48 days. Highly
significant correlation between airway responsiveness and BALF total
cells, AMs, neutrophils, and eosinophils, suggesting that airway
inflammation is involved.
Increased airway responsiveness to Hist after 10 episodes of exposure to
O3 + HDMA in sensitized infant monkeys.


Reference

Schlesinger
etal.
(2002a,b)




Segura et al.
(1997)


Vargas et al.
(1998)



Schelegle
et al. (2003a)


Ih
Guinea pig, male,
Dunkin-Hartley,
250-300 g,
Increased bronchial responsiveness at 3 h, but not 24 h after O3; OVA had    Sun et al.
no effect on baseline, but enhanced airway responsiveness 24 h after O3.      (1997)
H
6
0 1 Ih
o
H
O
0 2 2h
W
O
s
HH
H
W
n = 6-7/group
Mice, male, Ozone caused increased Cdyn and VE, and decreased PaO2 in OVA-
C57BL/6, sensitized mice.
6 weeks old,
n= 10-3 I/group
Rat, male, Fischer Increased airway responsiveness to MCh 2 h PE.
F344,
14 months old,
n = 6/group



Yamauchi
et al. (2002)


Dye et al.
(1999)





-------
                                        Table AX5-12 (cont'd). Effects of Ozone on Airway Responsiveness
S-
to
o
         Concentration   Exposure
         ppm            Duration
                    Species, Sex, Strain,
                    and Age
                     Observed Effect(s)
                                                                     Reference
>
X
^1
i
\D
        0.3-3.0
3h
        0.3
5h
                         3h
Mice, AJ, male and
female,
aged 2, 4, 8, or
12 weeks,
n = 42-50/group
Mice, BALB/c,
3 weeks old,
n = 5/group

Mice, C57B1/6J, and
ob/ob,
8-12 weeks old,
n = 6-7/group
                                             Rats, male and
                                             female, SD,
                                             2, 4, 6, 8, or 12
                                             weeks,
                                             n= 4-19/group
Nose-only exposure plethysmographs. VE decreased with increasing age.
O3 caused concentration-related decrease in VE at all ages, but with less
response in the 2-week old. Younger mice with less decrease in
O3-induced VE demonstrated 3- to 4-times greater inhaled dose when
normalized for body weight. The 2- to 4- week mice showed no AHR at
any dose, while the 8- and 12 week old mice demonstrated dose-related
increases in AHR. Older mice demonstrated increased levels of IL-6 and
MIP-2.  Suggests that young mice are less sensitive to O3 for the endpoints
of two cytokine and AHR.

Mice OVA-sensitized days 7-14; exposed days 21-23, assayed day 24 or
25 Decrease in Penh in rats exposed to O3,  as a function of MCh
concentration.

Compared C57BL/6J and ob/ob mice ( strain obese due to defect in gene
coding for satiety hormone leptin).  Intravenous MCh challenge induced
AHR and inflammation in both groups, but was greater in obese mice.
Dose per gram of lung tissue was greater in obese mice.  Suggests obese
mice get greater dose of O3.

Nose-only-exposure plethysmographs exposure.

8,12 week rats: O3-induced 40-50% decreases in VE (primarily due
to decrease in VT).

6 week rats: Cyinduced changes in VE were significantly less.
                                                                                                                                       Shore et al.
                                                                                                                                       (2002)
                                                                                                                                       Goldsmith
                                                                                                                                       et al. (2002)


                                                                                                                                       Shore et al.
                                                                                                                                       (2003)
                                                                                                              Shore et al.
                                                                                                              (2000)
H
1
O
O
0
H
/O 2 2h
C|
O
H
W
O
w
o
H
W
z, 4-weeKrais: no u3-i:

naucea cnanges in VE.

BALF protein and PGE2 were greater than in older rats. Suggests higher
delivered dose to younj
greater lung injury.

Guinea pigs, male, AHR to MCh peaked 2
Hartley, 400-600 g, Tazanolast (a mast cell
n = 6/group administered before O3
dependency. Suggests
development of AHR.



*er rats, decreased ventilatory response, and


h PE; PMN in BALF increased until 6 h PE. Igarashi et al.
stabilizing drug, doses 30, 100, or 300 mg/kg) (1998)
exposure inhibited O3-induced AHR dose-
that mast cells may play in role in the





-------
                                        Table AX5-12 (cont'd). Effects of Ozone on Airway Responsiveness
•s
S-
to
o
o
01













X
i
o



Concentration Exposure Species, Sex, Strain,
ppm Duration and Age

1 or 3 4 h, assayed 4 to Mice,
72 PE normal WBB6F1
(1/1) and mast cell-
deficient WBB6F1-
kit W /kit W-v
(kit W /kit W-v)
8-12 weeks old,
n= 3-1 I/group
2 2 h Cat, 2-3 kg,
n = 5/group



0.75 4 h, MCh challenge Mice
6hPE FVB/N, and
FVB/N with
P2-AR transgene,
10-14 weeks old,
n = 10/group

Observed Effect(s)

Demonstrated O3-induced cutaneous, as well as bronchial, mast cell
degranulation. PMN influx observed at 1 ppm only in normal mice.
AHR in response to MCh with 1 ppm observed in both. Suggests that
mast cells are involved in O3 -induced PMN influx but not AHR.




Cats anesthetized and mechanically ventilated, challenged with ACh.
Pretreated with polyethylene glycol-superoxide dismutase (PEG-SOD)
orPEG-catalase (PEG-CAT) 5 min before O3 exposure. PEG-SOD
partially prevented O3-induced AHR, PEG-CAT did not. Suggests
superoxide involvement in O3-induced AHR.
Targeted expression (using CCSP promoter) of p2-adrenergic receptors
(P2-AR) to airway epithelium to mimic agonist activation. Heterozygous
mice from generations 2 to 4 used. MCh challenge dose needed to increase
Penh was greater in CCSP-P2-AR mice. CCSP-P2-AR mice less responsive
to O3. Suggests that p2-ARs regulate airway responsiveness and that
P -agonists induce bronchodilation through activation of receptors on
smooth muscle cells and epithelial cells.
Reference

Noviski et al.
(1999)






Takahashi
etal. (1993)



McGraw et al.
(2000)





H
6
o

o
H
O
O
H
W
O
O
HH
H
W
                         2h, assayed 2 hPE
Rat, male, SD,
2.5-3.5 months old,
n = 5/group
Neonatal rats treated with capsaicin. Challenged with MCh following O3.

Capsaicin-treated rats: O3 had no effect on pulmonary conductance;
decreased dynamic compliance; increase in AHR.

During O3 exposure:  50% decrease in HR and 2.5 °C decrease in core
temperature in both controls and capsaicin rats.

Suggests that C-fibers inhibit O3-induced AHR but do not modulate HR
or core temperature.
Jimba et al.
(1995)

-------
                                         Table AX5-12 (cont'd). Effects of Ozone on Airway Responsiveness
to
o
o
X
-k
H
6
o

o
H
O
O
H
W
O
O
HH
H
W
Concentration
ppm
1
Exposure
Duration
3 h, assayed 4h PE
Species, Sex, Strain,
and Age
Rat, SD, treated
with capsaicin
or tachykinin
antagonists,
n = 6/group
Observed Effect(s)
Rats treated with CP-99994 (neurokinin-1 receptor antagonist) and
SR-48968 (neurokinin-2-receptor antagonist). O3 induced greater
numbers of PMN in BALF of treated rats. The antagonists has no
effects on pulmonary mechanics or airway responsiveness.
Reference
Takebayashi
et al. (1998)
                          3 h, ACh challenge
                          Oand24hPE
Mice, 8 weeks old,
15-25 g.
DBA/2J, AKR/J,
A/J, C3H/HeJ,
C57BL/6J, SJL/H,
129/J,
n = 6-8/group
Suggests that tachykinins are involved in the protective effects of C-fibers
against O3-induced inflammation.

Capsaicin treatment induced increased PMN in BALF. O3 exposure
reduced VE in both vehicle and capsaicin-treated rats, but the capsaicin
treatment caused a greater, more immediate reduction.

Suggests that the increase in BALF PMN is not due to a greater inhaled
dose of O3 reaching the lung.

Differing susceptibility to O3: Hyperreactive-DBA/2J, AKR/J, A/J;
Hyporeactive-C3H/HeJ, C57BL/6J, SJL/H; Intermediate 129/J.
ACh challenge 25 or 50 ug/kg.

ACh 24 h PE:  Airway responses increased in A/J strain at 25 ug/kg and
in C57BL/6J and SJL/J strains at 50 ug/kg.

Ozone did not alter reactivity in other strains. MCh or carbachol challenge
not affected by O3, suggesting that cholinesterase function is affected by
O3. C57BL/6J and A/J mice treated with cyclophosphamide
(an immunosuppressant) or anti-PMN caused decreased Cyinduced
PMN levels but did not alter AHR, suggesting that O3-induced ACh
hyperreactivity correlates with susceptibility and that PMN influx
is independent of AHR.
Zhang et al.
(1995)

-------
                                         Table AX5-12 (cont'd).  Effects of Ozone on Airway Responsiveness
S-
to
o
         Concentration
         ppm
Exposure
Duration
                     Species, Sex, Strain,
                     and Age
                      Observed Effect(s)
                                                                     Reference
X
to
H
6
o

o
H
/O
o
H
W
O
O
HH
H
W
         0.05
4 h, challenged with  Rats, male, Long-
                          iv 5-HT
                    Evans, SD, Fisher
                    344, Brown-Norway,
                    BDII, BDE, DA,
                    Lewis and Wistar,
                    6-8 weeks old,
                    n = 10/group
                      AHR: developed in Lewis, BDII and Long-Evans rats 90 min after O3.
                      Baseline AHR differed among strains; did not correlate with O3-induced
                      AHR.

                      PMN influx: did not occur in any strain.

                      LE rats:  AHR lasted > 12 h PE with no change in BALF PMN, LDH,
                      alkaline phosphatase, or protein.

                      Suggests that O3.induced AHR occurs without airway inflammation and
                      that genetic factors may alter the sensitivity to O3.
         0.2
7 h, assayed 3 h PE   Rabbits, New
         0.4
4 h, assayed by
ACh, SP, or
histamine challenge
0 or 48 h PE
                      O3-induced decrease in tracheal transepithelial potential difference, but no
Zealand White, 5 kg,    change in lung resistance.
n = 5-7/group
                      ACh challenge:  no change in compartmentalized lung resistance;
                      140% increase in Cyinduced lung resistance.

                      Bilateral vagotomy: no change in compartmentalized lung resistance;
                      enhancement of O3-induced peripheral lung reactivity.
Suggests that O3 exposure may affect tracheal epithelial function and
increase central airway reactivity, possibly through vagally -mediated
mechanisms.

Used isolated perfused lung model allowing partitioning of the total
pressure gradient into arterial, pre- and postcapillary, and venous
components. O3 -induced inhibition of pulmonary mechanical reactivity to
ACh, SP, and histamine. No change in baseline pulmonary resistance or
dynamic compliance. At 48 h PE O3 altered vasoreactivity of the vascular
bed. . Ozone-induced modification of the vasoreactivity of the vascular bed
at 48 h PE and elevation of arterial segmental pressure.  Suggests that O3
can directly induce vascular constriction both immediately and two days
following exposure and can inhibit ACh-, SP-, and histamine-induced
changes in lung mechanics.
                    Rabbits, New
                    Zealand White, male
                    and female, 2.5-3 kg,
                    n = 4-6/group
                                                                                                               Depuydt et al.
                                                                                                               (1999)
                                                                                                               Freed et al.
                                                                                                               (1996)
                                                                                           Delaunois
                                                                                           et al. (1998)

-------
                                         Table AX5-12 (cont'd).  Effects of Ozone on Airway Responsiveness
•s
S-
to
o
o
01









Concentration
ppm

0,0.12, 0.5, or
1.0






0.5

Exposure
Duration

6 h/day,
5 days/week for
20 months





8 h/day for 7 days

Species, Sex, Strain,
and Age

Rat, male and
female,
Fischer 344,
6-7 weeks old




Guinea pig, male,
Hartley, 5 weeks old
Observed Effect(s)

Isolated eighth generation airways following O3 exposure. Circumferential
tension development was measured in response to bethanechol,
acetylcholine, and electrical field stimulation and normalized to smooth
muscle area. 0.5 ppm caused an increase in smooth muscle area.
Maximum responses of the small bronchi of male rats were significantly
reduced after exposure to 0. 12 and 0.5 ppm O3. Suggests that O3 -induced
increases in airway responsiveness do not persist with near-lifetime
exposure and that chronic exposure alters smooth muscle cell function.
Repeated exposure increased rapidly adapting receptor activity to
substance P, methacholine, and hyperinflation; no significant effects on
Reference

Szarek et al.
(1995)






Joad et al.
(1998)
X
H
6
o

o
H
O
O
H
W
O
O
HH
H
W
         0.5
8 h/day for 5 days
followed by 9 days
of FA; for
11 episodes
Monkey, Rhesus,
30-day old,
n = 6/group
baseline or substance P- and methacholine-induced changes in lung
compliance and resistance.
Suggest that because agonist-induced changes in receptor activity precede
lung function changes, the responsiveness of rapidly adapting receptors
was enhanced.

Half of the monkeys were sensitized to house dust mite allergen (HDMA)
at 14 and 28 days of age before exposure.  HDMA and histamine aerosol
challenges administered until Raw doubled. Baseline Raw elevated after 10
exposure episodes in the HDMA + O3 group compared to the FA, HDMA,
and O3 exposure groups.  Aerosol challenge with HDMA at the end of the
10th episode did not significantly affect Raw, VT,/ or SaO2. AHR appeared
to develop following episode 6. Aerosol challenge with HDMA at the end
of the 10th episode did not significantly affect Raw,  VT,/ or SaO2.  HDMA
+ O3 group: had increases in serum IgG, histamine and eosinophils; greater
alteration in airway structure and content suggesting that O3 can enhance
the structural remodeling and allergic effects of HDMA sensitization
exposure groups.
Schelegle
et al. (2003a)

-------
                                         Table AX5-12 (cont'd).  Effects of Ozone on Airway Responsiveness
to
o
o
X
H
6
o

o
H
O
O
H
W
O
O
HH
H
W
Concentration
ppm
1
Exposure
Duration
8h
Species, Sex, Strain,
and Age
Rat, Wistar,
some treated
with capsaicin,
n = 6-8/group
Observed Effect(s)
Vehicle-treated: O3-induced rapid shallow breathing pattern; BrdU label
started at the bifurcation of the main stem bronchi and increased distally.
Capsaicin treated: no O3-induced changes in respiratory frequency;
Reference
Schelegle
etal. (2001)
                         2h
                                               Rat, male, Wistar,
                                               100-120 days old,
                                               n = 7-10/group
reduced BrdU labeling density in the terminal bronchioles supplied by
short airway paths. Suggests that O3-induced rapid shallow breathing
is protective of conducting airways and allows distribution of injury to
more distal regions.

Examined the site-specific deposition of 18O at breathing frequencies of 80,
120, 16, or 200 bpm at a VT to produce a constant minute ventilation of
72.8 ml/mm/100 gbody weight.

All frequencies: parenchyma! areas had a lower content of 18O than trachea
and bronchi; right caudal parenchyma! levels did not change.

80 to 160 bpm:  deposition reduced in midlevel trachea and increased in
both mainstream bronchi; increased deposition in parenchyma supplied by
short (cranial) airway paths.

200 bpm:  increased deposition in trachea increased; increased deposition
in right cranial and caudal bronchi regions: decreased content in right
cranial parenchyma!.

Suggests that the effect of rapid, shallow breathing is to create a more
evenly distributed injury pattern with less deposition of O3) in the trachea
and a small effect on deposition in the parenchyma.
Alfaro et al.
(2004)
         MCh = Methylcholine, ACh = Acetylcholine, Hist = Histamine, 5-HT = 5-Hydroxytryptamine, SP = Substance P, FS = Field Stimulation, CCh = Carbachol,
         TX = Thromboxane, KC1 = Potassium Chloride, Pt = Platinum; Route: iv = intravenous, inh = inhalation., sc = subcutaneous, ip = intraperitoneal,
         OVA = Ovalbumin, BALF = bronchoalveolar lavage fluid, HDMA = House Dust Mite Allergen, Cdyn = Dynamic Lung Compliance, VE = Minute Ventilation,
         PaO2 = partial pressure of arterial oxygen, AHR = airway hyperreactivity, VT = Tidal Volume, Raw = airway resistance,/= frequency of breathing,
         SaO2 = oxygen saturation of arterial blood.

-------
I
S-
to
o
<*s\













J>
X
Y1
5




O
§
H
6
o
2|
O
_ 1
O
O
H
W

Table AX5-13. Effects of Ozone on Genotoxicity/Carcinogenicity
Concentration Exposure
ppm Duration
1 0, 12, 24, 48,
72, or 96 h



1 or 2 90 min




2 90 min/day
for 5 days
0.12,0.50, 6h/day,
and 1.0 5 days/week for
up to 9 months



0.5 6 h/day,
5 days/week for
12 weeks


0, 0.12, 0.5,or 1.0 6hr/day,
5 days/wk,
to ppm 2-yr
and lifetime







Species,
Sex, Strain,
and Age
Guinea pigs,
Dunkin-Hartley ,
male,
2 months-old,
n = 4/group
Mice, female,
BALB/c
(20.6 g) or
Muta™ (26.0 g)
n= 11 -21 /group


Mice, female,
A/J,
n = 29-35/group



Mice, male and
female, B6C3Fl5
5-6 weeks old,
n = 20 M and 20
F
Rats, male and
female, Fischer
F-344/N,
n = 5M+5F
(acute)
n = 50 m + 50
F/group




Observed Effect(s)
Following O3 exposure, two main bronchi were removed and tracheobronchial cells were isolated
and assayed for DNA strand breaks using fluormetric analysis. Ozone induced an increase in
BALF protein and in DNA strand breaks, but did not change cell yield or viability. The amount
of DNA in alkali ly sates was decreased at 72 h, which suggests an increase in strand breaks at
that time point.
O3-induced increase in strand breaks up to 200 min following exposure. No effects after 200 min.
O3 did not affect the level of oxidized amino acids in lung or the level of 8-oxo-deoxyguanosine
in nuclear DNA. O3-induced induction of IL-6 mRNA following DNA strand breaks, which does
not support inflammation causing DNA damage.

Mutagenic mice had no O3-induced mutations in ell transgene.

At 5 months, no difference in lung tumor multiplicity or incidence. At 9 months, no differences
in lung tumor multiplicity between control mice and mice exposed to any concentration of O3.
The highest, and only statistically significant lung tumor incidence, was found in the mice exposed
to 0.5 ppm O3. In the O3-exposed mice allowed to recover in filtered air, only the mice exposed to
0.12 ppm O3 had increases in lung tumor incidence and multiplicity. Authors consider the results
to be spurious and of no significance for data interpretation.
Sporadic differences in mean body weight between O3-exposed mice and air-exposed controls,
as well as significant differences in the mean absolute and relative weights of liver, spleen,
kidney, testes, and ovary. No O3-related increased incidence of neoplasms in lung tissue.
Oviductal carcinomas observed upon histopathologic examination in 30% of O3-exposed
female mice.
Cocarcinogenicity study with subcutaneous administration of 0, 0.1, or 1.0 mg/kg body weight
of 4-(N-nitrosomethylamino)-l-(3-pyridyl)-l-butanone (NNK) and inhalation of 0 or 0.5 ppm
O3 to male rats.

O3 caused dose-related increase in inflammation in CAR; increased fibrosis; extension of the
bronchiolar epithelium to the proximal alveoli; but no increase in neoplasms.
NNK (1.0 mg/kg) alone causes increased bronchiolar/alveolar neoplasms.
NNK + O3: no enhancement of neoplasms.

Suggests that O3 is not carcinogenic; does not enhance neoplasms growth; and creates mild
site-specific lesions which persist with continued exposure.
Reference
Femg et al.
(1997)



Bornholdt
et al. (2002)





Witschi et al.
(1999)




Kim et al.
(2001)



Boorman
etal. (1994)









H
W

-------
X
H
6
o

o
H
O
O
H
W
O
O
HH
H
W
Concentration
ppm Duration
NEUROBEHAVIORAL EFFECTS
0.1 4h
0.2
0.5
1.0
0.1 4h
0.4
0.7
1.1
1.5
0.3 30 days
0.6

Species

Rat, male, Wistar,
47-50 days old,
n = 25/group

Rat, male, Wistar,
300-350 g,
n = 10/group


Mice, CD-I
M, F, 28-33 g,
n = 6-7/group
Effects

Rats exposed for 4 h to 0.2, 0.5, and 1 ppm O3 showed long-term memory
deterioration and decreased motor activity, which was reversed 24 h later.
Brain and pulmonary Cu/Zn SOD levels were increased in animals exposed
to 0.1, 0.2, and 0.5 ppm O3, but decreased in animals exposed to 1 ppm O3.
O3 caused memory impairment at >0.7 ppm (one trial passive avoidance test),
decreased motor activity at > 1. 1 ppm, and increased lipid peroxidation at
>0.4 ppm. Lipid perioxidation levels from the frontal cortex, hippocampus,
striatum and cerebellum increased with increasing O3 concentration.

O3 exposure slightly but selectively affected neurobehavioral performance in
male mice assessed with a 5-min open-field test on exposure days 4 and 19
and on day 3 after the end of the exposure. O3 exposure, however, did not
Reference

Rivas-Arancibia
et al. (1998)


Dorado-Martinez
etal. (2001)



Sorace et al.
(2001)

grossly affect neurobehavioral development. Reversal learning in the Morris
water maze test was consistently impaired in both prenatally and adult
exposed mice. In addition, longer latency to step-through in the first trial of
the passive avoidance test and a decrease in wall rearing in the hot-plate test
were recorded in O3 prenatally exposed mice. Except for the first open-field
test, altered responses were observed only in animals exposed to 0.3 ppm O3.
0.35 12 h
0.75
1.5
0.7 4h



0.7 4h



Rat, male, Wistar,
270 g,
n = 10/group
Rat, male, Wistar,
47-50 days old,
n = 6-10/group

Rat, male, Wistar,
27 months old,
n = 3-4/group

O3 exposure decreased paradoxical sleep after 2 h of exposure, and increased
slow wave sleep after 12 h of exposure at all O3 concentrations; 5-HT
concentrations in the pons increased with increasing O3 concentration.
Vitamin E administered before or after O3 exposure blocked memory
deterioration (passive avoidance)and increases in lipid peroxidation levels
in the striatum, hippocampus and frontal cortex that were associated with
oxidative stress.
O3 exposure increased ultrastructural alterations in the hippocampus and
prefrontal cortex in aged rats compared with controls. These areas are related
to learning and memory functions, which are the first degenerative aging
changes observed.
Paz and Huitron-
Resendiz (1996)

Guerrero et al.
(1999)


Nino-Cabrera
et al. (2002)



-------
                                                  Table AX5-14 (cont'd).  Systemic Effects of Ozone
to
o
o
Concentration
ppm             Duration         Species
                    Effects
                                                                       Reference
X
        NEUROBEHAVIORAL EFFECTS (cont'd)
        0.7
        0.8
                 4h
                         12 h/day
                         during dark
                         period

                         4h
Rat, male, Wistar,
47, 540, or 900
days old,
n = 10-30/group
                                  Rat, female, adult,
                                  n = 6/group
                                  Rat, male, Wistar,
                                  n = 24/group
                                          Rat, male, Wistar,
                                          275 g,
                                          n = 10/group
Taurine (43 mg/kg) given before or after O3 exposure improved memory
deterioration in an age-specific manner. Old rats showed peroxidation in all
control groups and an improvement in memory with taurine. When taurine
was applied before O3, peroxidation levels were high in the frontal cortex
of old rats and the hippocampus of young rats; in the striatum, peroxidation
caused by O3  was blocked when taurine was applied either before or after
exposure.

O3 exposure during pregnancy affects the neural regulation of paradoxical
sleep and circadian rhythm of rat pups 30, 60, and 90 days after birth.
                    O3 caused alterations in long-term memory and a significant reduction of
                    dendritic spines. Results provide evidence that deterioration in memory is
                    probably due to the reduction in spine density in the pyramidal neurons of the
                    hippocampus.

                    O3 or its reaction products affect the metabolism of major neurotransmitter
                    systems as rapidly as after 1 h of exposure. There were significant
                    increases in dopamine (DA), and its metabolites noradrenaline (NA)
                    and 3,4 dihydroxyphenylacetic acid (DOPAC), and 5-hydroxyindolacetic
                    acid (5-HIAA) in the midbrain and the striatum.
Rivas-Arancibia
et al. (2000)
Haro and Paz
(1993)


Avila-Costa et al.
(1999)
                                                                                                                             Gonzalez-Pina
                                                                                                                             and Paz (1997)
b> 1.5
•TJ
H
6
O
g 0.5
O

O
— I
w
o
brl
o
H
W
24 h Rat, male, Wistar,
n = 1 I/group


20 h/day Rat, SD,
for 5 days 220-240 g,
n = 10-20/group







Adult rats exposed to O3 spend decreased time in wakefulness and
paradoxical sleep and a significant increase in time in slow-wave sleep.
Neurochemical changes include increased metabolism of serotonin in
the medulla oblongata, pons, and midbrain.
O3 produced marked neural disturbances in structures involved in the
integration of chemosensory inputs, arousal, and motor control.
O3 inhibited tyrosine hydroxylase activity in noradrenergic brainstem cell
groups, including the locus ceruleus (-62%) and the caudal A2 subset (-57%).
Catecholamine turnover was decreased by O3 in the cortex (- 49%) and
striatum (- 18%) but not in the hypothalamus.




Huitron-Resendiz
et al. (1994)


Cottet-Emard
et al. (1997)









-------
                                                   Table AX5-14 (cont'd).  Systemic Effects of Ozone
to
o
o
Concentration
ppm             Duration         Species
                    Effects
                                                                        Reference
X
oo
        NEUROBEHAVIORAL

        0.4, 0.8, or 1.2    24 h
        0.75, 1.5 and 3.0  4h
         1-1.5
                 4h
                                  Cat, male, adult,
                                  n = 5/group
                                  Rat, male, Wistar,
                                  250-300 g,
                                  n= 15/group
Rat, male, Wistar,
-250 g
n = 5/group
Evaluated EEG of sleep-wake organization in cats.  0.4 O3 did not change the    Paz and Bazan-
amount of sleep parameters, did decrease paradoxical sleep during first 8 h       Perkins (1992)
of exposure. At 1.2 ppm paradoxical sleep was reduced during O3 exposure,
followed by a dose-related increase of slow-wave sleep. Suggests O3-induced
changes in sleep patterns.

Recorded evoked potential in visual cortex and lateral geniculate nucleus.       Custodio-
Pl, Nl and P2 components delayed in the visual cortex and lateral geniculate    Ramierez and
nucleus at 3.0 ppm O3. Nl component in the visual cortex affected at 1.5 ppm   Paz (1997)
O3. Suggest O3-induced alterations in synaptic excitability and conduction
mechanisms in the visual pathway.

O3-induced loss of dendritic spines on primary and secondary dendrites of       Colin-Barenque
granule cells;  swelling of Golgi apparatus and mitochondrion; dilation cisterns   et al.
of the rough endoplasmic reticulum; vacuolation of neuronal cytoplasm.         (1999)
Suggests O3-induced oxidative stress creates alterations in the granule layer
of the olfactory bulb and possible modifications of function.
        NEUROENDOCRINE EFFECTS
0.5 to 3.0 3 h
1.0 24 h
Rat, male, SD,
44-47 days old,
n = 4-6/group
Rat, male, SD,
3-4 months old
Hyperthyroid, T4-treated rats (0.1 - 1.0 mg/kg/day for 7 days) had increased
pulmonary injury (BALF LDH, albumin, PMNs) at 18 h PE compared to
control rats.
Hyperthyroid, T3 -treated rats had increased metabolic activity and O3-induced
pulmonary injury, but lipid peroxidation, as assessed by alkane generation,
was not affected.
Huffman et al.
(2001)
Sen etal. (1993)

-------
                                                    Table AX5-14 (cont'd). Systemic Effects of Ozone
to
o
o
Concentration
ppm             Duration         Species
                                                      Effects
                                                                                                              Reference
X
VO
                          AR EFFECTS
         0.1
         0.3
         0.5
                 5h
0.1
0.3
0.5

0.25 to 2.0
8h/day
for 4 days
                          2 h to 5 days
         0.5
                 6h/day
                 23 h/day
                 for 5 days
                 Rat, Wistar
                 young (4-6 month)
                 and old
                 (22-24 month)
                 n = 9-14/group
                                  Rat, male, Wistar,
                                  10 weeks old,
                                  n = 9/group

                                  Rat, Fischer F344,
                                  Mice, C57BL/6J,
                                  C3H/HeJ, Guinea
                                  pig, Hartley,
                                  n = 4-10/group

                                  Rat, male, Fischer
                                  F-344,
                                  100-120 days old,
                                  n = 4-6/group
Transient rapid shallow breathing with slightly increased HR appeared
1-2 min after the start of O3 exposure, possibly due to olfactory sensation;
persistent rapid shallow breathing with a progressive decrease in HR occurred
with a latent period of 1-2 h.  The last 90-min averaged values for relative
minute ventilation tended to decrease with the increase in the level of
exposure to O3 and these values for young rats were significantly lower than
those for old rats. An exposure of young rats to 0.1 ppm O3 for shorter than
5 h significantly decreased the tidal volume and HR and increased breathing
frequency, but no significant changes were observed in old rats.  There were
no differences between young and old rats in non-observable-adverse-effect-
levels (NOAELs) for the O3-induced persistent ventilatory and HR responses,
when the NOAELs were determined by exposure to 0.3 and 0.5 ppm O3.

Circadian rhythms of HR and core body temperature were significantly
decreased on the first and second O3 exposure days in a concentration
dependent manner, and returned to control levels  on the third and fourth days.

Robust and consistent decreases in HR and core body temperature; smaller
decreases in metabolism, minute ventilation, blood pressure, and cardiac
output that vary inversely with ambient temperature and body mass.
                                     Minimal extrapulmonary effects were observed at a core body temperature of
                                     34 °C; O3 exposures at 22 and 10 °C produced significant decreases in heart
                                     rate (160 and 210 beats/min, respectively), core body temperature (2.0 and
                                     3.5 °C, respectively), and body weight (15 and 40 g, respectively).
                                     Decreases in these functional parameters reached their maxima over the first
                                     2 exposure days and returned to control levels after the 3rd day of exposure.
                                                                                                                               Aritoetal. (1997)
Iwasaki et al.
(1998)
                                                                                                              Watkinson et al.
                                                                                                              (2001)
                                                                         Watkinson et al.
                                                                         (1995);
                                                                         Highfill and
                                                                         Watkinson (1996)

-------
                                                Table AX5-14 (cont'd).  Systemic Effects of Ozone
to
o
o
>
X
(Si
I
o
Concentration
ppm
Duration
CARDIOVASCULAR EFFECTS
0.5
0
0.5
8h
24 h/day
6h/day
Species
(cont'd)
Rat, male, Fischer
F-344,
270-330 g,
n = 6/group
Rat, male, Fischer
F-344 kept at one
of three
Effects
O3 exposure increased atrial natriuretic peptides in the heart, lung, and
circulation, suggesting they mediate the decreased BP and pulmonary
edema observed with similar O3 exposures.
0.5 ppm O3 for both 6 h/day and 23 h/day caused decreases in heart rate and
core temperature (termed hypothermic response) and increases in B ALF
inflammatory markers. Exercise in 0 ppm O3 caused increases in heart rate
Reference
Vesely et al.
(1994a,b,c)
Watkinson et al.
(2003)
                                        temperatures:
                                        10 °C, 22 °C,
        0.5             23h/day         34 °C at rest,
                                        moderate or heavy
        0.5                             CO2-stimulated
                                        ventilation,
                                        100-120 days old,
                                        n = 4-8/group

        REPRODUCTIVE AND DEVELOPMENTAL EFFECTS
                                                           and core temperature, 0.5 ppm O3 decreases.  CO2 and O3 induced the
                                                           greatest deficits. Dose, animal mass, and environmental stress are suggested
                                                           to modify the hypothermic response.
0.2
0.4
0.6

0.3
0.6
0.9


Continuous up
to day 17 of
pregnancy

Continuous up
to postnatal day
26


Mice, male and
female, CD-I,
25-27 g

Mice, male and
female, CD-I,
27-30 g,
n= 11-15/group

No significant effects on either reproductive performance, postnatal somatic
and neurobehavioral development (as assessed by a Fox test battery) or adult
motor activity (including within-session habituation); some subtle or
borderline behavioral deficits were noted, however.
O3 caused subtle CNS effects but did not affect the animals' capability to
learn a reflexive response (limb withdrawal); females exposed to 0.6 ppm
O3 showed a reduced preference for the right paw than both their same-sex
controls and 0.6 ppm males. The effect was more robust in the case of an
organized avoidance response (wall-rearing).
Petruzzi et al.
(1995)


Petruzzi et al.
(1999)




-------
                                                  Table AX5-14 (cont'd). Systemic Effects of Ozone
to
o
o
>
X
Concentration
ppm
Duration
REPRODUCTIVE AND DEVEL
0.3
0.6



0.4
0.8
1.2
0.6








Continuous
until gestational
day 17


Continuous
during gestation
days 7-17
Continuous
from birth to
weaning






Species
Effects
Reference
OPMENTAL EFFECTS (cont'd)
Mice, male and
female, CD-I,
28-30 g,
n = 6-9/group

Mice, CD-I


Mice, male and
female, CD-I,
25-27 g,
n= 13-16/group





Exposure to O3 did not grossly affect neurobehavioral development,
as assessed by somatic and sensorimotor development (postnatal day
(PND) 2-20), homing performance (PND 12), motor activity (PND 21),
passive avoidance (PND 22-23), water maze performances (PND 70-74),
and response to a nociceptive stimulus (PND 100).
No effect of O3 on reproductive performance; no significant somatic
developmental effects in O3-exposed pups except for a delay in eye
opening that was not concentration dependent.
Exposure to O3 did not produce any significant impairment of the acquisition
phase during swimming navigation, a sensitive indicator for hippocampal
damage; however, O3 slightly increased the swimming paths during the last
day of the reversal phase. Mice exposed to O3 showed a slightly but
significantly higher swimming speed during all the days, which was unrelated
to differences in body weight and to navigational performances. Moreover,
mice exposed to O3 (with the exception of one animal) had a strong tendency
to make turns to the left while the controls, independent of sex, preferred
clockwise turns.
Sorace et al.
(2001)



Bignami et al.
(1994)

DeH'Onio et al.
(1995a,b)







                         12 h/day for
                         entire gestation,
                         assayed at
                         postnatal day 0,
                         12, and 60
                                 Rat, male and
                                 female, Wistar,
                                 n = 9-10/group
        EFFECTS ON LIVER, SPLEEN, THYMUS

                         3h
1.0
2.0
Rat, female, SD,
200-225 g
                    Histological and planimetric analysis using sagittal sections of the anterior
                    cerebellar lobe.

                    PND 0: O3-induced cerebellar necrosis.

                    PND 12: diminished molecular layer; pale nucleoli and perinucleolar
                    bodies in Purkinjie cells. PND 60: Purkinjie cells with clumps of chromatin
                    around periphery. Suggests that gestational exposure to O3 induces permanent
                    cerebellar damage.
High O3 exposure stimulates hepatocytes to produce increased amounts
of nitric oxide as well as protein, possibly mediated by cytokines such
as TNFa produced by alveolar macrophages. When macrophage function
is blocked, hepatic injury induced by O3 is prevented.
                                                                       Rivas-Manzano
                                                                       and Paz( 1999)
Laskin et al.
(1994,  1996,
1998b);
Laskin and
Laskin (2001)

-------
                                                  Table AX5-14 (cont'd).  Systemic Effects of Ozone
to
o
o
         Concentration
         ppm
                 Duration
Species
Effects
Reference
>
X
(Si
I
to
        EFFECTS ON LIVER, SPLEEN, THYMUS (cont'd)
        2.0
                         2h
                                 Rat, male, Fischer
                                 F-344,
                                 2, 9, or 24 months
                                 old,
                                 n = 2/group
        EFFECTS ON CUTANEOUS TISSUE
        0.5
        0.8
0.8
1.0
10.0

1.0
5.0
10.0
                         2h
                         6h
                         2h
                         2h
        0, 0.8, 1, and 10   2 h
                                 Mice, hairless
                                 female
                                 Mice, SKH-1
                                 hairless
Mice, SKH-1
hairless
                                          Mice
                                          Mice, SKH-1
                                          hairless
                    Utilizing electron paramagnetic resonance (EPR) spectroscopy of chloroform    Vincent et al.
                    extracts of liver homogenates, a significant flux of hydrogen peroxide          (1996)
                    produced from the reaction of O3 with lipids of the extracellular lining could
                    be a source of biologically relevant amounts of hydroxyl radical. EPR signals
                    for carbon-centered alkoxyl and alkyl adducts were detected with C-phenyl
                    N-tert-butyl nitrone (PEN) in the liver of animals exposed to O3.
a tocopherol levels in the stratum corneum (SC) were not affected by O3        Valacchi et al.
exposure (0.5 ppm) alone, but were significantly depleted by combined         (2000)
exposure to UV and O3.

Increased lipid peroxidation in the skin epidermis and dermis activated stress    Valacchi et al.
proteins HSP27 and HO-1, and activated a proteolytic enzyme system          (2003)
(MMP-9) related to matrix injury and repair processes.

High O3 depletes hydrophilic antioxidants in the SC: vit. C decreased to 80%,   Weber et al.
GSH decreased to  41%, and uric acid decreased to 44% of control levels after   (2000)
exposure to > 1.0 ppm O3.

High O3 exerts an oxidizing effect on the outermost layer of the skin (SC);      Weber et al.
depletes low-molecular-weight antioxidants (a tocopherol, vit. C, glutathione,   (2001)
uric acid) in a concentration dependent manner; increases malondialdehyde
levels associated with lipid peroxidation.

1 ppm O3 depleted SC levels of vitamin C ( 80%), GSH (41%), and UA         Weber et al.
(44%). Suggests that hydrophilic antioxidants in the SC modulate the effects    (1999)
of O3-induced oxidative stress.
 RER = Rough endoplasmic reticulum
 PE = Postexposure (i.e., time after O3 exposure ceased)
 TSH = Thyroid stimulating hormone
 T3 = Triiodothyronine
 T4 = Thyroxine
cyt. = Cytochrome
                                                             NADPH = Reduced nicotinamide adenine dinucleotide phosphate
                                                             NADH = Reduced nicotinamide adenine dinucleotide
                                                             B[a]P = Benzo[a]pyrene
                                                             NK = Natural killer
                                                             PHA = Phytohemagglutin
                                                             ConA = Concanavalin A
                                                                                                                          LPS = Lipopolysaccharide
                                                                                                                          SRBC = Sheep red blood cell
                                                                                                                          TEA = Thiobarbituric acid
                                                                                                                          IgE = Immunoglobulin E.
                                                                                                                          PHA = Phytohemagglutin

-------
                                           Table AX5-15.  Interactions of Ozone With Nitrogen Dioxide
to
o
o
>
X
Concentration
ppm
03
NO2 Duration
Species
Endpoints
Interaction

Reference
MORPHOLOGY
0.8
14.4 6 h/day,
7 days/week
for 90 days
Rat, male, SD,
10-12 weeks
old,
Morphometry of lung
parenchyma; DNA probes
for procollagen; in situ
Syngeristic; more
after 7, 78, and 90
peripheral centriacinar lesion, but same
days of exposure.
Farman et al.
(1999)
         0.3       1.2    Continuous
                          for 3 days
BIOCHEMISTRY

 0.8       14.4   6 h/day,
                 7 days/week
                 for 9 weeks
         0.4
            7     90 days
                                       n = 4/group     mRNA hybridization.
                               Rat, male, SD,
                               3 months old,
                               n = 4/group
Rat, male, SD,
10-12 weeks
old
n = 4/group


Rat, male, SD,
200-225 g
               DNA single strand breaks;
               polyADPR synthetase of
               AMs; total cells, protein,
               andLDHinBALF.
Lung hydroxyproline,
hydrooxypyridinium,
DNA, and protein content
of whole lung; morphology
and labeling index.

PMN, pulmonary edema,
fibrosis, MIP-2.
                          None; effect due to O3.
Synergistic; fibrosis after 7-8 weeks of exposure;
50% mortality at -10 weeks.
                                                     Bermudez
                                                     etal. (1999);
                                                     Bermudez
                                                     (2001)
Farman et al.
(1997)
1-3 days: enhanced MIP-2, IL-lp, TNFa, thioredoxin, and   Ishiietal.
IL-6 expression; pulmonary edema and PMN influx, which   (2000b)
reversed by day 8; activation of NFicB.

15-45 days: no tissue responses observed, suggesting
adaptation.

60 and 90 days:  increased lung collagen; increased
expression of transforming growth factor-13 and TNFa,
activation of NFicB.
                                                                                MnSOD and GPX not altered during exposure.

                                                                                Suggests that cytokines play a role in the early responses
                                                                                to combined O3 and NO2 and that pulmonary fibrosis is
                                                                                more dependent on concentration than on cumulative dose.

-------
                                       Table AX5-15 (cont'd). Interactions of Ozone With Nitrogen Dioxide
S-
to
o
         Concentration
              ppm
                   NO2   Duration     Species
                                                     Endpoints
                          Interaction
                                                     Reference
>
X
        BIOCHEMISTRY (cont'd)

          0.8       14.4   1,5, and
                          8 weeks
                                       Rat, male, SD,
                                       10-12 weeks
                                       old,
                                       n = 4/group
Immunohistochemisty and
morphometric analysis of
TNFcc and MnSOD levels
in alveolar ducts.
Triphasic response.  1-3 weeks: initial inflammation;
TNFcc increased in proximal area.

4-5 weeks: partial resolution.

6-8 weeks: rapidly progressive fibrosis; elevated MnSOD;
TNFcc increased in proximal area.

TNFcc increased in interstitial cells at all time points.

Suggests O3-induced increases in MnSOD in areas of
injury and in more protected areas; O3-induced increases
in TNFcc correlate spatiotemporally with injury.
Weller et al.
(2000)
H
6
o

o
H
/O
o
H
W
O
O
HH
H
W
        BAL = Bronchoalveolar lavage.
        PG = Prostaglandin.
        G-6-PD = Glucose-6-phosphate dehydrogenase.
        GOT = GSH-disulfide transhydrogenase.
        GSHPX = GSH peroxidase.
                                                              SOD = Superoxide dismutase.
                                                              DR = Disulfide reductase.
                                                              NADPH-CR = Reduced nicotinamide adenine dinucleotide phosphate-cytochrome c reductase.
                                                              GSH = Glutathione.
                                                              6-PG-D = 6-phosphogluconate dehydrogenase.

-------
                                               Table AX5-16. Interactions of Ozone with Formaldehyde
_-
S-
to
o
o
03
0.4
HCHO
3.6
Duration
8 h/day for
3 days
Species
Rat, male,
Wistar,
n = 20/group
Endpoints
O3 or HCHO: no changes in formaldehyde dehydrogenase,
glutathione S-transferase, glutathione reductase, and glucose-6-
phosphate dehydrogenase activities.
Interaction
Proliferative
response:
No interactive
effects
Reference
Cassee and
Feron(1994)
>
X
                                                          O3 + HCHO:  slightly decreased enzyme activities

                                                          HCHO alone:  rhinitis, necrosis, degeneration, hyperplasia and

                                                          squamous metaplasia of nasal respiratory epithelium.



                                                          O3: PMN influx; disarrangement, flattening and slight basal cell

                                                          hyperplasia of the nonciliated cuboidal epithelium.  Proliferating

                                                          cell nuclear antigen levels elevated.
0.6 10 3h, with
exercise at
2* resting
ventilation
0.2-10 0.4-4 30min
Rat, male,
SD, 7 weeks
Mice,
BALB/c,
n = 4
HCHO did not alter O3 -induced changes in breathing pattern.
Parenchyma! injury attributed to O3 alone.
Continuously measured/ VT, expiratory flow, T;, Te, and
respiratory patterns during acute exposures. HCHO: appeared to
be a pure sensory irritant at lower concentrations. O3: induced
nondose-dependent transient increase in rapid, shallow breathing.
No effect level of HCHO was 0.3 ppm, for O3 1.0 ppm. Suggests
that O3 and HCHO have the same respiratory effects in BALB/c
mice and humans, with similar sensitivities.
Additive effects
in transitional
epithelium
and trachea
Not additive
Mautz (2003)
Nielson et al.
(1999)
H

6
o


o
H

O

o
H
W

O


O
HH
H
W

-------
                                             Table AX5-17. Interactions of Ozone with Tobacco Smoke
_-
OJ
to
o



03
1.5



HCHO
7 ml breaths at
33% challenge


Duration
Ih



Species
Guinea pig,
male, Hartley,
320-460 g,
n = 5 to 9
Endpoints
Pulmonary resistance and dynamic lung compliance were compared
pre- and post-O3 exposure.

Pre-O3: cigarette smoke induced bronchoconstriction after 1 min.
Interaction
Synergistic



Reference
Wu et al.
(1997)


                                                         Post-O3: cigarette smoke induced bronchoconstriction more quickly
                                                         and for a longer period.
>
X
ON
         0.5
ADSS
30 mg/m3
6 h/day
for 3 days
                              24 h
Mice, male,
B6C3F1,
10 weeks old,
n=ll
Selective antagonism of neurokinin 1 and 2 receptors blocked then
enhanced O3-induced bronchoconstriction, suggesting that endogenous
tachykinins modulate O3-induced bronchoconstriction.

Aged and diluted side stream cigarette smoke (ADSS) exposure
followed by O3.

ADSS + O3: greater increase in BALF cells, %PMN, and proteins
compared to alone; 402% increase in proliferating cells in CAR;
LPS-stimulated release of IL-6 decreased; LPS-stimulated release
of TNFcc increased.
                                                                                                                          Synergism
Yu et al.
(2002)
                                                         ADSS alone: no change in number of proliferating cells in CAR;
                                                         LPS-stimulated release of TNFcc increased.

                                                         O3 alone:  280% increase in proliferating cells in CAR;
                                                         LPS-stimulated release of IL-6 decreased.
H
6
o

o
H
O
o
H
W
O
O
HH
H
W
                                                        Suggests that O3-induced lung injury is enhanced by prior ADSS
                                                        exposure.

-------
                                        Table AX5-18. Interactions Of Ozone With Particles
to
o
o
>
X
H

6
o


o
H

O

O
H
W

O


O
HH
H
W
Concentration
03
(ppm)
PM
(mg/m3)
Duration Species
Endpoints
Interaction
Reference
SULFURIC ACID
0.1
0.2
0.1
0.3
0.6
0.1
0.3
0.6
0.6
PARTICI
0.1
0.16
0.30
0.59
0.02-0.15
(0.4 - 0.8 urn)
0.50 (0.3 urn)
0.125 (0.3 urn)
0.50 (0.3 urn)
0.125(0.3 urn)
0.5 (0.06 and
0.3 umMMD)
,E MIXTURES
Diesel PM (MIST
#2975) reacted with
O3 for 48 h
0.05 - 0.22 mg/m3
ammonium bisulfate
0.03 -0.10 mg/m3 C
0.11 -0.39pmNO2
0.02 -0.11 mg/m3
HNO3. (0.3 urn
MMAD)
23.5 h/day or Rat, male, SD,
intermittent 12 h/day 250-275 g,
for up to 90 days n = 6/group
3h Rabbit NZW
male, 3.5-4.5 kg,
n = 5/group
3h Rabbit NZW
male, 3.5-4.5 kg,
n = 5/group
4 h/day for 2 days Rat, male, SD,
250-300 g,
n = 10/group
24 h (IT) Rat, male, SD,
250-300 g,
n = 4-13
4 h/day, 3 days/week Rat, male,
for 4 weeks Fischer F344N,
1 1 weeks old,
n = 8-30
Morphology
Biochemistry
AM intracellular
pH homeostasis and
H+ extrusion
Airway responsiveness
(in vitro bronchial
rings + ACh)
Morphology: volume
percentage of total
parenchyma containing injured
alveolar septae;
bromodeoxyuridine cell
labeling index in the periacinar
region
Inflammation
breathing pattern, morphology,
lavagable protein, and
clearance
No interaction
Antagonism
Antagonism
Synergism: ultrafine + O3, but
not fine
Synergism: fine + O3
Synergism
Complex interactions, but
possible loss of typical
attenuation seen with O3 only
exposure, reflecting
persistence of inflammation.
Last and
Pinkerton
(1997)
Chen et al.
(1995)
El-Fawal
etal. (1995)
Kimmel et al.
(1997)
Madden et al.
(2000)
Mautz et al.
(2001)

-------
                                     Table AX5-18 (cont'd).  Interactions Of Ozone With Particles
to
o
o
>
X
(Si
 I

oo
H

6
o


o
H

O

o
H
W

O


O
HH
H
W
Concentration
O3 PM
(ppm) (mg/m3)
PARTICLE MIXTURES (cont'd)
0.2 0.07and 0.14 mg/m3
ammonium bisulfate
(0.45 urn MMMD);
0.05 and 0.10 mg/m3
carbon
0.2 0.50 mg/m3 ammonium
bisulfate (0.45 um
MMMD) and elemental
carbon)
0.3 0.063 to 1.57 mg/m3
CAPs (Boston) + ip
OVA sensitization

0.4 0.20 and 0.50 mg/m3
fine, H2O2-coated
carbon (0.26 um
MMMD)
0.5 Endotoxin (IN)
100 ug 24 hand
48 h after the 3rd
O3 exposure
0.5 OVA (IN)
50 ul (1%)
Duration
4 h/day, 3 days/week
for 4 weeks
4 h/day, 3 days/week
for 4 weeks; nose only
Sensitized at days 7
and 14. Challenged at
day 2 1,22, and 23.
5 h O3 exposure
4 h/day for 1 or 5 days
8 h/day for 3 days
8 h/day for 1 day or
3 consecutive days
Species
Rat, male,
Fischer F344,
22-24 months
old,
n = 5-10
Rat, Fischer
F344N-NIA,
22-24 months old
Mice, BALB/c,
n = 5 -6/group

Rat, SD,
300 g,
n=10
Rat, Fischer
F344,
10-12 weeks old,
n = 6/group
Rat, Brown,
Norway,
10-12 weeks old,
n = 6/group
Endpoints
BAL protein and albumin;
plasma hydroxylase and
fibronectin
DNA labeling of dividing lung
epithelial and interstitial cells
by 5-bromo-2-deoxyuridine
Airway function

Inflammation
Nasal morphology
Nasal morphology
Interaction
Questionable interaction.
No changes in BALF protein
or prolyl 4-hydroxylase in
blood. Small decease in
plasma fibronectin with
combined exposure.
Synergism. Increased AM
phagocytosis and respiratory
burst. Decreased lung
collagen.
Interaction: increased RL
and airway responsiveness in
normal and OVA-sensitized
mice.
Synergism for effect on day 5.
Greater response at high dose,
contrasts with O3 along where
inflammation was greatest at
0.4 ppm on day 1.
Synergism: increased
intraepithelial mucosubstances
and mucous cell metaplasia
Synergism: increased
intraepithelial mucosubstances
and mucous all metaplasia.
Reference
Bolarin et al.
(1997)
Kleinman
et al. (2000)
Kobzik et al.
(2001)

Kleinman
etal. (1999)
Fanucchi
etal. (1998)
Wagner et al.
(2001a,b)
Wagner et al.
(2002)

-------
                                       Table AX5-18 (cont'd).  Interactions Of Ozone With Particles
to
o
o
>
X
         Concentration
 O3      PM

(ppm)   (mg/m3)\
                                   Duration
Species
Endpoints
Interaction
Reference
PARTICLE MIXTURES (cont'd)
0.8 0.5 mg, 1.5 mg, or 2, 4, and 7 days after
5 mg of PM from IT instillation
Ottawa Canada
(EHC-93)
1 0.11 mg/m3 ultra fine 6 h
carbon (25 nm CMD)
+ endotoxin (IH)







1 Endotoxin (37.5 EU) 4, 20, or 24 h
for 10 minutes


1 Endotoxin (IN) 0, 2, 8 h, repeated after
or 20 ug in 120 uL 24 h





Rat, male, Inflammation
Wistar,
200-250 g,
n = 5/group
Rat, male, Inflammation
Fischer F3 44,
10 weeks and
22 months old,
n = 3/group
Mice, male,
TSK,
14-17 months
old,
n = 6-7/group
Mice, Inflammation
C57BL/6J, 36 h
old, 8 weeks old,
n = 12/group
Rat, male, Lung morphometric analysis
Fischer F344, and inflammation
10-12 weeks old,
n = 6/group



Interaction: increased TNFcc
inBALF.


Interaction: increased
PMNs and ROS release from
BALF cells for old rats and
mice primed with endotoxin;
depressed in young rats.





Synergism: increased BALF
protein and PMNs.


Synergism: increased
BALF PMNs and mucin
glycoprotein; increased
intraepithelial
mucosubstances and
mucingene mRNA.

Ulrich et al.
(2002)


Elder et al.
(2000a,b)








Johnston
et al. (2000b,
2002)

Wagner et al.
(2003)





-------
                                              Table AX5-18 (cont'd).  Interactions Of Ozone With Particles
to
o
O
           Concentration
  03
(ppm)
  PM
(mg/m3)
                                          Duration
                                                        Species
                   Endpoints
Interaction
Reference
X
PARTICLE MIXTURES (cont'd)

  0.8     EHC-93,5 mg/m3 or
         50 mg/m3
                                          4 h exposure, clean air
                                          for 32 h, 3H injection
                                          followed by assay at
                                          90min
           1.8     (EHC93) at 50 mg/m3
                                 4 h, exposure, clean
                                 air for 32 h, 3H
                                 injection followed by
                                 assay at 90 min
Rat, male,          Control rats:  3H labeling
Fischer344,         (indicative of proliferation)
200-250 g,          low in bronchioles and
n = 6/group         parenchyma.

                   ECH-93 alone: no induction
                   of proliferation.

                   O3 alone: increased cell
                   labeling in bronchioles and
                   parenchyma,  suggestive of
                   reparative cell proliferation.

                   ECH-93 + O3: both doses
                   of EHC-93 potentiated
                   proliferation, especially
                   in epithelia of terminal
                   bronchioles and alveolar
                   ducts, but not distal
                   parenchyma.  Suggests that
                   ambient PM can enhance
                   O3-induced proliferations
                   and exacerbate injury.

Rat, male,          O3 +EHC-93: epithelial cell
Fischer344,         injury and proliferation
200-250g,          (3H-labeling) higher than
n = 4/group        single exposures; higher in
                   periductal region than in
                   whole lung counts; greater
                   numbers of AMs and PMNs in
                   lung tissue compartment than
                   in single exposures.  Suggests
                   that exposures to urban PM
                   have few effects alone, but can
                   potentiate O3-induced injury.
                                                                                                         Synergism of proliferation
                                                                                                                              Vincent et al.
                                                                                                                              (1997)
                                                                                               Synergism
                                                                                                                                       Adamson
                                                                                                                                       etal. (1999)

-------
                                              Table AX5-18 (cont'd). Interactions Of Ozone With Particles
&
S-
to
o
o
Concentration
03
(ppm)
PM
(mg/m3)
Duration
Species
Endpoints Interaction
Reference
PARTICLE MIXTURES (cont'd)




0.15



HNO3 50 ug/m3



4 hours/day,
3 days/week
for 40 weeks,
nose-only exposure
Rat, male,
Fischer F344/N, 8
weeks old,
n = 4-5/group
O3 alone: 28% increase in Synergism
lung putrescine.

FDSKD3 alone: 21% decrease in
Sindhu et al.
(1998)


X
O
o
2
o
H
O
c
o
H
W
O
V
O
HH
H
W
           0.5     Carbon Black (CB),
                  0.5 or 1.5 mg/rat
Intratracheal CB
followed by 7 days
or 2 months of O3
Rat, male, Wistar,
7 weeks old,
n = 7-8/group
lung putrescine

O3andO3+HNO3:
56% increase in lung
putrescine

Pulmonary spermidine and
spermine did not change with
any exposure.

Suggests role of putrescine in
regulation of inflammation.

Phagocytotic capacity:
decreased in CB-exposed
group, unchanged in O3 group.
Formation of superoxide anion
radicals and numbers of
ingested particles increased at
2 months in O3 group.

Chemotactic migration:
decreased in CB-treated group.

Suggests that CB impairs
phagocytosis and chemotactic
migration in AMs, whereas O3
stimulates these functions.
Synergism
Creutzenberg
etal. (1995)

-------
                                            Table AX5-18 (cont'd). Interactions Of Ozone With Particles
to
o
o
           Concentration
  O3      PM
(ppm)    (mg/m3)
                                         Duration
                                                      Species
                                                       Endpoints
                                                                     Interaction
                                                                            Reference
        PARTICLE MIXTURES (cont'd)
         0.018
         CH2O
         TSP
         PM10
         PM25
3.3ppb
0.068 mg/m3
0.032 mg/m3
0.016 mg/m3
23 h/day for 7 weeks
Rat, male and
female, Fischer
F344,
9 weeks old,
n = 5/group
Exposure to filtered and
unfiltered Mexico City air.
Histopathology revealed no
nasal lesions in exposed or
control rats; tracheal and lung
tissue from both groups
showed similar levels of
minor abnormalities.
                                                                                                                                  Moss et al.
                                                                                                                                  (2001)
X
(Si
I
ON
to
VMD = Volume median diameter
og = Geometric standard deviation
MMAD = Mass median aerodynamic diameter
CMD = Count median diameter
PM = Particulate Matter
OVA= Ovalbumin
RL = Total Pulmonary Resistance
                                              AED = Aerodynamic diameter
                                              BAL = Bronchoalveolar lavage
                                              AM = Alveolar macrophage
                                              IN = Intranasal
                                              IT = Intratracheal
                                             Unless indicated otherwise, whole-body exposures used

-------
&
^
to
o
o
(^
Table AX5-19. Effects of Other Photochemical Oxidants
X
H

6
o


o
H

O

O
H
W

O


O
HH
H
W
Concentration
ppm Duration
H2O2(10, 20, or 2h,
100 ppb) and/or assayed 0
(NH4)2S04 or24hPE
(429or215Ag/m3;
0.3-0.4 gmmass
median diameter)



PAN
0.13,0.66,1.31 3h

0.077,0.192,0.384 Ih

0, 15, 39, or 78 ppm Ih


3 ppm PAN 1 month




-300 ppb 4, 7, 10, or
12h

78 ppm 24 h



-80 ppm 3H-PAN 1 h



Species
Rat,
female, SD,
n = 4- 1 8/group






PEL from
mice, male,
CD- 1 , 5 weeks
old, n = 25
Mice, male,
B6C3F1,
6 weeks old,
n = 4
Hamster,
Chinese,
n = 4


Salmonella
TA100

Mice, Big
Blue®, male,
4 week old,
n=15




Effects
Exposures alone or in combination.

H2O2 + (NH4)2S04: no effect on BALF cell number, viability, protein or LDH at either
time point; increased PMNs adhered to vascular epithelium; increased TNFa by AMs at
both time points; at 0 h PE, transient increase in superoxide anion by AMs; decreased NO
production; nitrotyrosine detected; heme oxygenase-1 expression upregulated in AMs;
H2O2 only: decreased NO production which persisted for 24 h.
Suggests that PM-induced effects are augmented by H2O2 and that PM-induced tissue
injury may be modulated by cytotoxic mediators produced by AMs.
In vitro: Exposed peripheral blood lympocytes (PEL) to PAN; assayed for chromosome
aberrations, sister chromatid exchanges and DNA damage. DNA damage was observed
at cytotoxic concentrations of PAN. No effects at noncytotoxic exposures.

In vivo: Nose-only exposures. No dose-related effects at any exposure level in any
assay.

Suggests that PAN, both in vitro and in vivo, is not a DNA damaging agent or clastogen.
Measured frequency of thioguanine-resistant lung fibroblasts and frequency of
micronuclei in either the bone marrow or the lungs.

Mutation frequencies not altered from control levels. No chromosome breakage was
observed in bone marrow or lung. Suggests no mutagenicity in vivo.
Gas phase exposure PAN-mduced mutants: 59% GC -> TA; 29% GC -> AT;
2% GC -> CG; 10% multiple mutations — primarily GG -> TT tandem-base
substitutions.

Nose-only exposure. Mutagenic at the lacl gene in the lung; no tandem-base mutations.


Nasal tissue: 3.9% of the radiolabel.
Lung: 0.3 % of radiolabel.
Suggests that PAN is weakly mutagenic in Salmonella (with a signature GG -> TT
transversion) and in mouse lung.
Reference
Morio et al.
(2001)







Kligerman
etal. (1995)






Heddle et al.
(1993)



DeMarini
et al. (2000)










-------
  1       REFERENCES
  2
  3       Adamson, I. Y. R.; Vincent, R.; Bjarnason, S. G. (1999) Cell injury and interstitial inflammation in rat lung after
  4             inhalation of ozone and urban particulates. Am. J. Respir. Cell Mol. Biol. 20:  1067-1072.
  5       Alfaro, M. F.; Putney, L.; Tarkington, B. K.; Hatch, G. E.; Hyde, D. M; Schelegle, E. S. (2004) Effect of rapid
  6             shallow breathing on the distribution of 18O-labeled ozone reaction product in the respiratory tract of the rat.
  7             Inhalation Toxicol. 16: 77-85.
  8       Arito, H.; Takahashi, M.; Iwasaki, T.; Uchiyama, I. (1997) Age-related changes in ventilatory and heart rate
  9             responses to acute ozone exposure in the conscious rat. Ind. Health 35: 78-86.
10       Arsalane, K.; Cosset, P.; Vanhee, D.; Voisin, C.; Hamid, Q.; Tonnel, A.-B.; Wallaert, B. (1995) Ozone stimulates
11             synthesis of inflammatory cytokines by alveolar macrophages in vitro. Am. J. Respir. Cell Mol. Biol.
12             13:60-68.
13       Avila-Costa, M. R.; Colin-Barenque, L.; Fortoul, T. L; Machado-Salas, J. P.; Espinosa-Villanueva, J.;
14             Rugerio-Vargas, C.; Rivas-Arancibia, S. (1999) Memory deterioration in an oxidative stress model and its
15             correlation with cytological changes on rat hippocampus CA1. Neurosci. Lett. 270: 107-109.
16       Ballinger, C. A.; Cueto, R.; Squadrito, G.; Coffin, J. F.; Velsor, L.  W.; Pryor, W. A.; Postlethwait, E. M. (2005)
17             Antioxidant-mediated augmentation of ozone-induced membrane oxidation. Free Radical Biol. Med.
18             38:515-526.
19       Bassett, D.; Elbon-Copp, C.; Otterbein, S.; Barraclough-Mitchell, H.; DeLorme, M.; Yang, H. (2001) Inflammatory
20             cell availability affects ozone-induced lung damage. J. Toxicol. Environ. Health A 64: 547-565.
21       Becker, S.; Quay, J.; Koren, H. S. (1991) Effect of ozone on immunoglobulin production by humanB cells in vitro.
22             J. Toxicol. Environ. Health 34: 353-366.
23       Bermudez, E. (2001) Detection of poly(ADP-ribose) synthetase activity in alveolar macrophages of rats exposed to
24             nitrogen dioxide and ozone. Inhalation Toxicol. 13: 69-84.
25       Bermudez, E.; Ferng, S. F.; Castro, C. E.; Mustafa, M. G. (1999) DNA strand breaks caused by exposure to ozone
26             and nitrogen dioxide. Environ. Res. 81: 72-80.
27       Bhalla, D. K. (1996) Alteration of alveolar macrophage chemotaxis, cell adhesion, and cell adhesion molecules
28             following ozone  exposure of rats. J. Cell. Physiol. 169: 429-438.
29       Bhalla, D. K.; Gupta, S. K. (2000) Lung injury, inflammation, and inflammatory stimuli in rats exposed to ozone.
30             J. Toxicol. Environ. Health 59: 211-228.
31       Bhalla, D. K.; Hoffman, L. (1997) Time course of airway epithelial and inflammatory changes in rats exposed to
32             moderate levels of ozone. Inhalation Toxicol. 9: 829-842.
33       Bhalla, D. K.; Gupta, S. K.; Reinhart, P. G. (1999) Alteration of epithelial integrity, alkaline phosphatase activity,
34             and fibronectin expression in lungs of rats exposed to ozone. J. Toxicol. Environ. Health A 56: 329-343.
35       Bhalla, D. K.; Reinhart, P. G.; Bai, C.; Gupta, S. K. (2002) Amelioration of ozone-induced lung injury by anti-tumor
36             necrosis factor-a. Toxicol. Sci. 69: 400-408.
37       Bignami, G.; Musi, B.; DeH'Omo, G.; Laviola, G.; Alleva, E. (1994) Limited effects  of ozone exposure during
3 8             pregnancy on physical and neurobehavioral development of CD-I mice. Toxicol. Appl. Pharmacol.
39             129:264-271.
40       Bolarin, D. M.; Bhalla, D. K.; Kleinman, M. T. (1997) Effects of repeated exposures of geriatric rats to ozone and
41             particle-containing atmospheres: an analysis of bronchoalveolar lavage and plasma proteins. Inhalation
42             Toxicol. 9: 423-434.
43       Boorman, G. A.; Hailey, R.; Grumbein, S.; Chou, B. J.; Herbert, R. A.; Goehl,  T.; Mellick, P. W.; Roycroft, J. H.;
44             Haseman, J. K.; Sills, R. (1994) Toxicology and carcinogenesis studies of ozone and ozone
45             4-(A^-nitrosomethylamino)-l-(3-pyridyl)-l-butanone in Fischer-344/N rats. Toxicol. Pathol. 22:  545-554.
46       Bornholdt, J.; Dybdahl, M.; Vogel, U.; Hansen, M.; Loft, S.; Wallin, H. (2002) Inhalation of ozone induces DNA
47             strand breaks and inflammation in mice. Mutat. Res.  520: 63-71.
48       Bridges, J. P.; Davis,  H. W.; Demodarasamy, M.; Kuroki, Y.; Howies, G.; Hui, D. Y.; McCormack, F. X. (2000)
49             Pulmonary surfactant proteins A and D are potent endogenous inhibitors of lipid peroxidation and oxidative
50             cellular injury. J. Biol. Chem. 275: 38848-38855.
51       Broeckaert, F.; Clippe, A.; Wattiez, R.; Falmagne, P.; Bernard, A.  (2003) Lung hyperpermeability, Clara-cell
52             secretory potein (CC16), and susceptibility to ozone of five  inbred strains of mice. Inhalation Toxicol.
53             15: 1209-1230.
54       Cassee, F. R.; Feron,  V. J. (1994) Biochemical and histopathological changes in nasal epithelium of rats after 3-day
55             intermittent exposure to formaldehyde and ozone alone or in combination. Toxicol. Lett. 72: 257-268.
         August 2005                                  AX5-64        DRAFT-DO NOT QUOTE OR CITE

-------
  1       Catalano, P. I; Rogus, I; Ryan, L. M. (1995a) Consequences of prolonged inhalation of ozone onF344/Nrats:
  2             collaborative studies. Part X: robust composite scores based on median polish analysis. Cambridge, MA:
  3             Health Effects Institute; research report no. 65.
  4       Catalano, P. I; Chang, L.-Y. L.; Harkema, J. R.; Kaden, D. A.; Last, J. A.; Mellick, P. W.; Parks, W. C.; Pinkerton,
  5             K. E.; Radhakrishnamurthy, B.; Ryan, L. M.; Szarek, J. L. (1995b) Consequences of prolonged inhalation of
  6             ozone on F344/N rats: collaborative studies. Part XI: Integrative summary. Cambridge, MA: Health Effects
  7             Institute; research report no. 65.
  8       Chang, L.-Y.; Stockstill, B. L.; Menache, M. G.; Mercer, R. R.; Crapo, J. D. (1995) Consequences of prolonged
  9             inhalation of ozone on F344/N rats: collaborative studies. Part VIII: morphometric analysis of structural
10             alterations in alveolar regions. Cambridge, MA: Health Effects Institute; pp. 3-39; research report no. 65.
11       Chang, M. M.-J.; Wu, R.; Plopper, C. G.; Hyde, D. M. (1998) IL-8 is one of the major chemokines produced by
12             monkey airway epithelium after  ozone-induced injury. Am. J. Physiol. 275: L524-L532.
13       Cheek, J. M.; McDonald, R. J.; Rapalyea, L.; Tarkington, B. K.; Hyde, D. M. (1995) Neutrophils enhance removal
14             of ozone-injured alveolar epithelial cells in vitro. Am. J. Physiol. 269: L527-L535.
15       Chen, L. C.; Qu, Q.; Amdur, M. O.; Schlesinger, R. B. (1995) Alteration of pulmonary macrophage intracellular pH
16             following inhalation exposure to sulfuric acid/ozone mixtures. Exp. Lung Res. 21: 113-128.
17       Chen, C.-Y.; Bonham, A. C.; Plopper, C. G.; Joad, J. P. (2003) Plasticity in respiratory motor control: selected
18             contribution: neuroplasticity in nucleus tractus solitarius neurons following episodic ozone exposure in infant
19             primates. J. Appl. Physiol.  94: 819-827.
20       Cheng, P.-W.; Boat, T. F.; Shaikh, S.; Wang, O.-L.; Hu, P.-C.; Costa, D. L. (1995) Differential effects of ozone  on
21             lung epithelial lining fluid volume and protein content. Exp. Lung Res. 21: 351-365.
22       Cho, H. Y.; Hotchkiss, J. A.; Harkema, J. R. (1999a) Inflammatory and epithelial responses during the development
23             of ozone-induced mucous cell metaplasia in the nasal epithelium of rats. Toxicol. Sci.  51: 135-145.
24       Cho, H. Y.; Hotchkiss, J. A.; Bennett, C. B.; Harkema, J. R. (1999b) Effects of pre-existing rhinitis on
25             ozone-induced mucous cell metaplasia in rat nasal epithelium. Toxicol. Appl. Pharmacol. 158: 92-102.
26       Cho, H. Y.; Hotchkiss, J. A.; Bennett, C. B.; Harkema, J. R. (2000) Neutrophil-dependent and
27             neutrophil-independent alterations in the nasal epithelium of ozone-exposed rats. Am.  J. Respir. Crit. Care
28             Med. 162: 629-636.
29       Cho, H.-Y.; Zhang, L.-Y.; Kleeberger, S. R. (2001) Ozone-induced lung inflammation and hyperreactivity are
30             mediated via tumor necrosis factor-a receptors. Am. J. Physiol. 280: L537-L546.
31       Churg, A.; Brauer,  M.; Keeling, B. (1996) Ozone enhances the uptake of mineral particles by tracheobronchial
32             epithelial cells in organ culture. Am. J. Respir. Crit. Care Med. 153: 1230-1233.
33       Cohen, M. D.; Zelikoff, J. T.; Qu, Q.; Schlesinger, R. B. (1996) Effects of ozone upon macrophage-interferon
34             interactions. Toxicology 114: 243-252.
35       Cohen, M. D.; Zelikoff, J. T.; Chen, L.-C.; Schlesinger, R. B. (1998) Immunotoxicologic effects of inhaled
36             chromium: role of particle solubility and co-exposure to ozone. Toxicol. Appl. Pharmacol. 152: 30-40.
37       Cohen, M. D.; Sisco, M.; Li, Y.; Zelikoff, J. T.; Schlesinger, R. B. (2001) Ozone-induced modulation of
38             cell-mediated immune responses in the lungs.  Toxicol. Appl. Pharmacol. 171: 71-84.
39       Cohen, M. D.; Sisco, M.; Baker, K.; Li, Y.; Lawrence, D.; VanLoveren, H.; Zelikoff, J. T.; Schlesinger, R. B.
40             (2002) Effects of inhaled ozone on pulmonary immune cells critical to antibacterial responses in situ.
41             Inhalation Toxicol. 14: 599-619.
42       Colin-Barenque, L.; Avila-Costa, M. R.; Fortoul,  T.; Rugerio-Vargas, C.; Machado-Salas, J. P.;
43             Espinosa-Villanueva, J.; Rivas-Arancibia,  S. (1999) Morphologic alteration of the olfactory bulb  after acute
44             ozone exposure in rats. Neurosci. Lett. 274: 1-4.
45       Connor, L. M.; Ballinger, C. A.; Albrecht, T. B.; Postlethwait, E. M. (2004) Interfacial phospholipids inhibit ozone
46             reactive absorption-mediated cytotoxicity in vitro. Am. J. Physiol.: 10.1152/ajplung.00397.2003.
47       Cottet-Emard, J.-M.; Dalmaz, Y.; Pequignot, J.; Peyrin, L.; Pequignot, J.-M. (1997) Long-term exposure to ozone
48             alters peripheral and central catecholamine activity in rats. Pfluegers Arch. 433: 744-749.
49       Creutzenberg, O.; Bellmann, B.; Klingebiel, R.; Heinrich, U.; Muhle, H. (1995) Phagocytosis and chemotaxis of rat
50             alveolar macrophages after a combined or separate exposure to ozone and carbon black. Exp. Toxicol. Pathol.
51             47:202-206.
52       Custodio-Ramirez, V.; Paz, C. (1997) Ozone produces functional deficits in the rat visual pathway.
53             Electroencephalogr. Clin. Neurophysiol. 104:  269-273.
         August 2005                                  AX5-65       DRAFT-DO NOT QUOTE OR CITE

-------
  1       DeMarini, D. M; Shelton, M. L.; Kohan, M. J.; Hudgens, E. E.; Kleindienst, T. E.; Ball, L. M.; Walsh, D.; de Boer,
  2             J. G.; Lewis-Bevan, L.; Rabinowitz, J. R.; Claxton, L. D.; Lewtas, J. (2000) Mutagenicity in lung of Big
  3             Blue(R) mice and induction of tandem-base substitutions in Salmonella by the air pollutant peroxyacetyl
  4             nitrate (PAN): predicted formation of intrastrand cross-links. Mutat. Res. 457: 41-55.
  5       Delaunois, A.; Segura, P.; Montano, L. M.; Vargas, M. H.; Ansay, M.; Gustin, P. (1998) Comparison of
  6             ozone-induced effects on lung mechanics and hemodynamics in the rabbit. Toxicol. Appl. Pharmacol.
  7             150:58-67.
  8       Dell'Omo, G.; Fiore, M.; Petruzzi, S.; Alleva, E.; Bignami, G. (1995a) Neurobehavioral development of CD-I mice
  9             after combined gestational and postnatal exposure to ozone. Arch. Toxicol. 69: 608-616.
10       Dell'Omo, G.; Wolfer, D.; Alleva, E.; Lipp, H.-P. (1995b) Developmental exposure to ozone induces subtle changes
11             in swimming navigation of adult mice. Toxicol. Lett. 81:  91-99.
12       Depuydt, P.; Joos, G. F.; Pauwels, R. A. (1999) Ambient ozone concentrations induce airway hyperresponsiveness
13             in some rat strains. Eur. Respir. J. 14:  125-131.
14       Depuydt, P. O.; Lambrecht, B. N.; Joos, G. F.; Pauwels, R. A. (2002)  Effect of ozone exposure on allergic
15             sensitization and airway inflammation induced by dendritic cells. Clin. Exp. Allergy 32: 391-396.
16       Dong, W.; Selgrade, M. K.; Gilmour,M. L; Lange, R. W.; Park, P.; Luster, M. I.; Kari, F. W. (1998) Altered alveolar
17             macrophage function in calorie-restricted rats. Am. J. Respir. Cell Mol. Biol.  19: 462-469
18       Dorado-Martinez, C.; Parades-Carbajal,  C.; Mascher, D.; Borgonio-Perez, G.; Rivas-Arancibia, S. (2001) Effects of
19             different ozone doses on memory, motor activity and lipid peroxidation levels, in rats. Int. J. Neurosci.
20             108: 149-161.
21       Dormans, J. A. M. A.; Boere, A. J. F.; vanLoveren, H.; Rombout, P. J. A.; Marra, M.;  vanBree, L. (1996)
22             Age-related toxicity in rat lungs following acute and repeated ozone exposure. Inhalation Toxicol. 8: 903-925.
23       Dormans, J. A. M. A.; VanBree, L.; Boere, A. J. F.; Marra, M.; Rombout, P. J. A. (1999) Interspecies differences in
24             time course of pulmonary toxicity following repeated exposure to ozone. Inhalation Toxicol. 11: 309-329.
25       Driscoll, K. E.;  Simpson, L.;  Carter, J.; Hassenbein, D.; Leikauf, G. D. (1993) Ozone inhalation stimulates
26             expression of a neutrophil chemotactic protein, macrophage inflammatory protein 2. Toxicol. Appl.
27             Pharmacol. 119: 306-309.
28       Dye, J. A.; Madden, M. C.; Richards, J. H.; Lehmann, J. R.; Devlin, R. B.; Costa, D. L. (1999) Ozone effects on
29             airway responsiveness, lung injury, and inflammation. Comparative rat strain and in vivo/in vitro
30             investigations. Inhalation Toxicol. 11: 1015-1040.
31       El-Fawal, H. A. N.; McGovern, T.; Schlesinger, R. B. (1995) Nonspecific bronchial responsiveness assessed in vitro
32             following acute inhalation exposure to ozone and ozone/sulfuric acid mixtures. Exp. Lung Res. 21: 129-139.
33       Elder, A. C. P.;  Gelein, R.; Finkelstein J. N.; Cox,  C.; Oberdorster, G. (2000a) Endotoxin priming affects the lung
34             response to ultrafine particles and ozone in young and old rats.  In: Phalen, R. F., ed. Inhalation toxicology:
3 5             proceedings of the third colloquium on paniculate air pollution and human health (first special issue); June,
36             1999; Durham, NC. Inhalation Toxicol. 12(suppl. 1): 85-98.
37       Elder, A. C. P.;  Gelein, R.; Finkelstein, J. N.; Cox, C.; Oberdorster, G. (2000b) Pulmonary inflammatory response to
3 8             inhaled ultrafine particles is modified by age, ozone exposure, and bacterial toxin. In: Grant, L. D., ed.
39             PM2000: paniculate matter and health. Inhalation Toxicol. 12(suppl. 4): 227-246.
40       Elsayed, N. M. (2001) Diet restriction modulates lung response and survivability of rats exposed to ozone.
41             Toxicology 159: 171-182.
42       Evans, M. J.; Fanucchi, M. V; Baker,  G. L.; Van Winkle, L. S.; Pantle, L. M.; Nishio,  S. J.; Schelegle, E. S.;
43             Gershwhin, L. J.; Miller, L. A.; Hyde, D. M.; Sannes, P. L.; Plopper, C. G. (2003) Atypical development of
44             the trachea! basement membrane zone of infant rhesus monkeys exposed to ozone and allergen. Am.  J.
45             Physiol. 285:L931-L939.
46       Fakhrzadeh, L.; Laskin, J. D.; Laskin,  D. L.  (2002) Deficiency in inducible nitric oxide synthase protects mice from
47             ozone-induced lung inflammation and tissue injury. Am. J. Respir. Cell Mol. Biol. 26: 413-419.
48       Fanucchi, M. V.; Hotchkiss, J. A.; Harkema, J. R. (1998) Endotoxin potentiates ozone-induced mucous cell
49             metaplasia in rat nasal epithelium. Toxicol.  Appl. Pharmacol. 152: 1-9.
50       Farman, C. A.; Pinkerton, K. E.; Rajini, P.; Witschi, H.; Last, J.  A. (1997) Evolution of lung lesions in rats exposed
51             to mixtures of ozone and nitrogen dioxide. Inhalation Toxicol.  9: 647-677'.
52       Farman, C. A.; Watkins, K.; Van Hoozen, B.; Last, J. A.; Witschi, H.; Pinkerton, K. E. (1999) Centriacinar
53             remodeling and sustained procollagen gene expression after exposure to ozone and nitrogen dioxide.  Am. J.
54             Respir. Cell Mol. Biol. 20: 303-311.
55       Ferng, S.-F.; Castro, C. E.; Afifi, A. A.; Bermudez, E.; Mustafa, M. G. (1997) Ozone-induced DNA strand breaks in
56             guinea pig tracheobronchial epithelial cells. J. Toxicol. Environ. Health 51: 353-367.
         August 2005                                  AX5-66       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Foster, W. M; Freed, A. N. (1999) Regional clearance of solute from peripheral airway epithelia: recovery after
  2             sublobar exposure to ozone. J. Appl. Physiol. 86: 641-646.
  3       Frampton, M. W.; Pryor, W. A.; Cueto, R.; Cox, C.; Morrow, P. E.; Utell, M. J. (1999) Aldehydes (nonanal and
  4             hexanal) in rat and human bronchoalveolar lavage fluid after ozone exposure. Cambridge, MA: Health Effects
  5             Institute; research report no. 90. Available: www.healtheffects.org/Pubs/Frampton-C.pdf [2000, February 9].
  6       Freed, A. N.; Chou, C. L.; Fuller, S. D.; Croxton, T. L. (1996) Ozone-induced vagal reflex modulates airways
  7             reactivity in rabbits. Respir. Physiol. 105: 95-102.
  8       Freed, A. N.; Cueto, R.; Pryor, W. A. (1999) Antioxidant transport modulates peripheral airway reactivity and
  9             inflammation during ozone exposure. J. Appl. Physiol. 87:  1595-1603.
10       Garssen, J.; VanBree, L.; VanDer Vliet, H.; VanLoveren, H. (1997) Ozone-induced impairment of pulmonary type
11             IV hypersensitivity and airway hyperresponsiveness in mice. Inhalation Toxicol. 9: 581-599.
12       Gohil, K.; Cross, C. E.; Last, J. A. (2003) Ozone-induced disruptions of lung transcriptomes. Biochem. Biophys.
13             Res. Commun. 305: 719-728.
14       Goldsmith, C.-A. W.; Ning, Y.-Y.; Qin, G.; Imrich, A.; Lawrence, J.; Murthy, G. G., K.; Catalano, P. J.; Kobzik, L.
15             (2002) Combined air pollution particle and ozone exposure increases airway responsiveness in mice.
16             Inhalation Toxicol. 14: 325-347.
17       Gonzalez-Pina, R.; Paz, C. (1997) Brain monoamine changes in rats after short periods of ozone exposure.
18             Neurochem. Res. 22: 63-66.
19       Guerrero, A. L.; Dorado-Martinez, C.; Rodriguez, A.; Pedroza-Rios, K.; Borgonio-Perez, G.; Rivas-Arancibia,  S.
20             (1999) Effects of vitamin E on ozone-induced memory deficits and lipid peroxidation in rats. NeuroReport
21             10: 1689-1692.
22       Gupta, S. K.; Reinhart, P.  G.; Bhalla, D. K. (1998) Enhancement of fibronectin expression in rat lung by ozone and
23             an inflammatory stimulus. Am. J. Physiol. 275: L330-L335.
24       Haddad, E.-B.; Liu, S. F.; Salmon, M.; Robichaud, A.; Barnes, P.  J.; Chung, K. F. (1995) Expression of inducible
25             nitric oxide synthase mRNA in Brown Norway rats exposed to ozone: effect of dexamethasone. Eur. J.
26             Pharmacol. Environ. Toxicol. Pharmacol. Sect. 293: 287-290.
27       Hamilton, R. F.; Li, L.; Eschenbacher, W. L.; Szweda, L.; Holian, A. (1998) Potential involvement of
28             4-hydroxynonenal in the response of human lung cells to ozone. Am. J. Physiol. 274: L8-L16.
29       Harkema, J. R.; Morgan, K. T.; Gross, E. A.; Catalano, P. J.; Griffith, W. C. (1994) Consequences of prolonged
30             inhalation of ozone on F344/N rats: collaborative studies. Part VII: effects on the nasal mucociliary apparatus.
31             Cambridge, MA: Health Effects Institute; research report no. 65.
32       Harkema, J. R.; Catalano, P. J.; Hotchkiss, J. A. (1997a) Consequences of prolonged inhalation of ozone on F344/N
33             rats: collaborative studies. Part XII. Atrophy of bone in nasal turbinates. Cambridge, MA: Health Effects
34             Institute; research report no. 65.
35       Harkema, J. R.; Hotchkiss, J. A.; Griffith, W. C. (1997b) Mucous cell metaplasia in rat nasal epithelium after a
36             20-month exposure to ozone: a morphometric study of epithelial differentiation. Am. J. Respir. Cell Mol.
37             Biol. 16: 521-530.
38       Harkema, J. R.; Hotchkiss, J. A.; Barr, E. B.; Bennett, C. B.; Gallup, M.; Lee, J. K.; Basbaum, C.  (1999)
39             Long-lasting effects of chronic ozone exposure on rat nasal epithelium. Am. J. Respir. Cell Mol. Biol.
40             20:517-529.
41       Haro, R.; Paz, C. (1993) Effects of ozone exposure during pregnancy on ontogeny of sleep in rats. Neurosci. Lett.
42             164:67-70.
43       Heddle, J. A.; Shepson, P. B.; Gingerich, J. D.; So, K. W. (1993) Mutagenicity of peroxyacetyl nitrate (PAN)
44             in vivo: tests for somatic mutations and chromosomal aberrations. Environ. Mol. Mutagen. 21: 58-66.
45       Herbert, R. A.; Hailey, J. R.; Grumbein, S.; Chou, B. J.; Sills, R. C.; Haseman, J. K.; Goehl, T.; Miller, R. A.;
46             Roycroft, J. H.; Boorman, G. A. (1996)  Two-year and lifetime toxicity and carcinogenicity studies of ozone in
47             B6C3F1 mice. Toxicol. Pathol. 24: 539-548.
48       Highfill, J. W.; Watkinson, W. P. (1996) Ozone toxicity in the rat. II. Modeling changes due to ambient temperatures
49             and duration. J. Appl. Physiol. 80: 1811-1818.
50       Hoffer, E.; Baum, Y.; Tabak, A.; Frevert, C. (1999) Adhesion molecules of blood polymorphonuclear leukocytes
51             and alveolar macrophages in rats: modulation by exposure to ozone. Hum. Exp. Toxicol. 18: 547-551.
52       Hotchkiss, J. A.; Harkema, J. R.; Johnson, N. F. (1997) Kinetics of nasal epithelial cell loss and proliferation inF344
53             rats following a single exposure to 0.5 ppm ozone. Toxicol. Appl. Pharmacol.  143: 75-82.
54       Hotchkiss, J. A.; Hilaski, R.; Cho, H.; Regan, K.; Spencer, P.;  Slack, K.; Harkema, J. R.  (1998) Fluticasone
55             propionate attenuates ozone-induced rhinitis and mucous cell metaplasia in rat nasal airway epithelium. Am.
56             J. Respir. Cell Mol. Biol. 18: 91-99.
         August 2005                                  AX5-67        DRAFT-DO NOT QUOTE OR CITE

-------
  1       Huffman, L. I; Judy, D. I; Brumbaugh, K.; Frazer, D. G.; Reynolds, J. S.; McKinney, W. G.; Goldsmith, W. T.
  2             (2001) Hyperthyroidism increases the risk of ozone-induced lung toxicity in rats. Toxicol. Appl. Pharmacol.
  3             173: 18-26.
  4       Huitron-Resendiz, S.; Custodio-Ramirez, V.; Escalante-Membrillo, C.; Gonzalez-Pina, R.; Paz, C. (1994) Sleep
  5             alterations and brain regional changes of serotonin and its metabolite in rats exposed to ozone. Neurosci. Lett.
  6             177: 119-122.
  7       Hyde, D. M; Miller, L. A.; McDonald, R. J.; Stovall, M. Y.; Wong, V.; Pinkerton, K. E.; Wegner, C. D.; Rothlein,
  8             R.; Plopper, C. G. (1999) Neutrophils enhance clearance of necrotic epithelial cells in ozone-induced lung
  9             injury in rhesus monkeys. Am. J. Physiol. 277: LI 190-L1198.
10       Igarashi, A.; lijima, H.; Tamura,  G.; Shirato, K. (1998) Tazanolast inhibits ozone-induced airway
11             hyperresponsiveness in guinea pigs. Am. J. Respir. Crit. Care Med. 157: 1531-1535.
12       lijima, M. K.; Kobayashi, T.; Kamada, H.; Shimojo, N. (2001) Exposure to ozone aggravates nasal allergy-like
13             symptoms in guinea pigs.  Toxicol. Lett. 123: 77-85.
14       Inoue, H.; Aizawa, H.; Nakano, H.; Matsumoto, K.; Kuwano, K.; Nadel, J. A.; Kara, N. (2000) Nitric oxide synthase
15             inhibitors attenuate ozone-induced airway inflammation in guinea pigs: possible role of interleukin-8. Am. J.
16             Respir. Crit. Care Med. 161: 249-256.
17       Ishii, Y.; Yang, H.; Sakamoto, T.; Nomura, A.; Hasegawa, S.; Hirata, F.; Bassett, D.  J. P. (1997) Rat alveolar
18             macrophage cytokine production and regulation of neutrophil recruitment following acute ozone exposure.
19             Toxicol. Appl. Pharmacol. 147: 214-223.
20       Ishii, Y.; Shirato, M.; Nomura, A.; Sakamoto, T.; Uchida, Y.; Ohtsuka, M.; Sagai, M.; Hasegawa, S. (1998) Cloning
21             of rat eotaxin: ozone inhalation increases mRNA and protein expression in lungs of brown Norway rats. Am.
22             J. Physiol. 274:  L171-L176.
23       Ishii, Y.; Hashimoto, K.; Hirano, K.; Morishima, Y.; Mochizuki, M.; Masuyama, K.; Nomura, A.; Sakamoto, T.;
24             Uchida, Y.; Sagai, M.; Sekizawa, K. (2000a) Ebselen decreases ozone-induced pulmonary inflammation in
25             rats. Lung 178: 225-234.
26       Ishii, Y.; Hirano, K.; Morishima, Y.; Masuyama, K.; Goto, Y.; Nomura, A.; Sakamoto, T.; Uchida, Y.; Sagai, M.;
27             Sekizawa, K. (2000b) Early molecular and cellular events of oxidant-induced pulmonary fibrosis in rats.
28             Toxicol. Appl. Pharmacol. 167: 173-181.
29       Iwasaki, T.; Takahashi, M.; Saito, H.; Arito, H. (1998) Adaptation of extrapulmonary responses to ozone exposure in
30             conscious rats. Ind. Health 36: 57-60.
31       Jang, A.-S.; Choi, I.-S.; Koh, Y.-L; Park, C.-S.; Lee, J.-S. (2002) The relationship between alveolar epithelial
32             proliferation and airway obstruction after ozone exposure. Allergy  57: 737-740.
33       Jimba, M.; Skornik, W. A.; Killingsworth, C. R.; Long, N. C.; Brain, J. D.; Shore, S. A. (1995) Role of C fibers in
34             physiological responses to ozone in rats. J. Appl. Physiol. 78: 1757-1763.
35       Joad, J. P.; Brie, J. M.; Weir, A. J.; Putney, L.; Hyde, D. M.; Postlewait, E. M.; Plopper, C.  G. (2000) Effect of
36             respiratory pattern on ozone injury to the airways of isolated rat lungs. Toxicol. Appl. Pharmacol. 169: 26-32.
37       Johnston, C. J.; Stripp, B. R.; Reynolds, S. D.; Avissar, N. E.; Reed, C. K.; Finkelstein, J. N. (1999a) Inflammatory
3 8             and antioxidant gene expression in C57BL/6J mice after lethal and sublethal ozone exposures. Exp. Lung
39             Res. 25: 81-97.
40       Johnston, C. J.; Finkelstein, J. N.; Oberdorster, G.; Reynolds, S. D.; Stripp, B. R. (1999b) Clara cell  secretory
41             protein-deficient mice differ from wild-type mice in inflammatory chemokine expression to oxygen and
42             ozone, but not to endotoxin. Exp. Lung Res. 25: 7-21.
43       Johnston, C. J.; Reed, C. K.; Avissar, N. E.; Gelein, R.; Finkelstein, J. N.  (2000a) Antioxidant and inflammatory
44             response after acute nitrogen  dioxide and ozone exposures in C57B1/6 mice. Inhalation Toxicol. 12: 187-203.
45       Johnston, C. J.; Oberdorster, G.;  Gelein, R.; Finkelstein, J. N. (2000b) Newborn mice differ from adult mice in
46             chemokine and cytokine expression to ozone, but not to endotoxin. Inhalation Toxicol. 12: 205-224.
47       Johnston, C. J.; Oberdorster, G.; Finkelstein, J. N. (2001) Recovery from oxidant-mediated lung injury: response of
48             metallothionein, MIP-2, and MCP-1 to nitrogen dioxide, oxygen, and ozone exposures. Inhalation Toxicol.
49             13:689-702.
50       Johnston, C. J.; Oberdorster, G.;  Gelein, R.; Finkelstein, J. N. (2002) Endotoxin potentiates ozone-induced
51             pulmonary chemokine and inflammatory responses. Exp. Lung Res. 28: 419-433.
52       Kafoury, R. M.; Pryor, W. A.; Squadrito, G. L.; Salgo, M. G.; Zou, X.; Friedman, M. (1999) Induction of
53             inflammatory mediators in human airway epithelial cells by lipid ozonation products. Am. J. Respir. Crit.
54             Care Med. 160:  1934-1942.
55       Kenyon, N. J.; Van Der Vliet, A.; Schock, B.C.; Okamoto, T.; McGrew,  G. M.; Last, J. A. (2002) Susceptibility to
56             ozone-induced acute lung injury in iNOS-deficient mice. Am. J.  Physiol. 282: L540-L545.
         August 2005                                   AX5-68        DRAFT-DO NOT QUOTE OR CITE

-------
  1       Kim, M. Y.; Son, J. W.; Cho, M. H.; Choi, C. S.; Chae, C. H.; Lee, M. H. (2001) Oviductural carcinoma in B6C3F1
  2             female mice exposed to 0.5 ppm ozone. Vet. Hum. Toxicol. 43: 370-372.
  3       Kimmel, T. A.; Chen, L. C.; Bosland, M. C.; Nadziejko, C. (1997) Influence of acid aerosol droplet size on
  4             structural changes in the rat lung caused by acute exposure to sulfuric acid and ozone. Toxicol. Appl.
  5             Pharmacol. 144: 348-355.
  6       Kleeberger, S. R.; Levitt, R. C.; Zhang, L.-Y.; Longphre, M.; Harkema, J.; Jedlicka, A.; Eleff, S. M.; DiSilvestre, D.;
  7             Holroyd, K. J. (1997) Linkage analysis of susceptibility to ozone-induced lung inflammation in inbred mice.
  8             Nat. Genet. 17: 475-478.
  9       Kleeberger, S. R.; Reddy, S.;  Zhang, L.-Y.; Jedlicka, A. E. (2000) Genetic susceptibility to ozone-induced lung
10             hyperpermeability: role of toll-like receptor 4. Am. J. Respir. Cell Mol. Biol. 22: 620-627.
11       Kleeberger, S. R.; Ohtsuka, Y.; Ahang, L.-Y.; Longphre, M. (200la) Airway responses to chronic ozone exposure
12             are partially mediated through mast cells. J. Appl. Physiol. 90: 713-723.
13       Kleeberger, S. R.; Reddy, S. P.; Zhang, L.-Y.; Cho, H.-Y.; Jedlicka, A. E. (200Ib) Toll-like receptor 4 mediates
14             ozone-induced murine lung hyperpermeability via inducible nitric oxide synthase.  Am. J. Physiol.
15             280: L326-L333.
16       Kleinman, M. T.; Mautz, W. J.; Bjarnason, S. (1999) Adaptive and non-adaptive responses in rats exposed to ozone,
17             alone and in mixtures,  with acidic aerosols. Inhalation Toxicol. 11: 249-264.
18       Kleinman, M. T.; Bufalino, C.; Rasmussen, R.; Hyde, D.; Bhalla, D. K.; Mautz, W. J. (2000) Toxicity of chemical
19             components of ambient fine paniculate matter (PM2 5) inhaled by aged rats. J. Appl. Toxicol. 20: 357-364.
20       Kligerman, A. D.; Mottus, K.; Erexson, G.  L. (1995) Cytogenetic analyses of the in vitro  and in vivo responses of
21             murine cells to peroxyacetyl nitrate (PAN). Mutat. Res. 341: 199-206.
22       Kobzik, L.; Goldsmith, C.-A. W.; Ning, Y. Y.; Qin, G.; Morgan, B.; Imrich, A.; Lawrence J.;  Murthy, G. G. K.;
23             Catalano, P. J. (2001) Effects of combined ozone and air pollution particle exposure in mice. Boston, MA:
24             Health Effects Institute; research report no. 106.  Available: http://www.healtheffects.org/Pubs/Kobzik.pdf
25             [27 January 2003].
26       Kodavanti, U. P.; Hatch, G. E.; Starcher, B.; Giri, S. N.; Winsett, D.; Costa, D. L. (1995)  Ozone-induced pulmonary
27             functional, pathological, and biochemical changes in normal and vitamin C-deficient guinea pigs. Fundam.
28             Appl. Toxicol. 24:  154-164.
29       Koike, E.; Kobayashi, T.; Nelson, D. J.; Me William, A. S.; Holt, P. G. (1998) Effect of ozone exposure on alveolar
30             macrophage-mediated  immunosuppressive activity in rats. Toxicol. Sci. 41: 217-223.
31       Koike, E.; Kobayashi, T.; Murakami, M.; McWilliam, A. S.; Holt, P. G. (1999) Mechanisms of ozone-induced
32             inhibitory effect of bronchoalveolar lavage fluid on alveolar macrophage-mediated immunosuppressive
33             activity in rats. J. Leukoc. Biol.  66: 75-82.
34       Koike, E.; Kobayashi, T.; Shimojo, N.  (2001) Ozone exposure enhances expression of cell-surface molecules
3 5             associated with antigen-presenting activity on bronchoalveolar lavage cells in rats. Toxicol. Sci. 63: 115-124.
36       Koto, H.; Salmon, M.; Haddad el-B.; Huang, T.-J.; Zagorski, J.;  Chung, K. F. (1997) Role of cytokine-induced
37             neutrophil chemoattractant (CINC) in ozone-induced airway inflammation and hyperresponsiveness. Am.
38             J. Respir. Crit. Care Med. 156: 234-239.
39       Larson, S. D.; Schelegle, E. S.; Walby, W.  F.; Gershwin, L. J.; Fanuccihi, M. V.; Evans, M. J.; Joad, J. P.;
40             Tarkington, B. K.; Hyde, D. M.; Plopper, C. G. (2004) Postnatal remodeling of the neural components of the
41             epithelial-mesenchymal trophic unit in the proximal airways of infant rhesus monkeys exposed to ozone and
42             allergen. Toxicol. Appl. Pharmacol.  194: 211-220.
43       Laskin, D. L.; Laskin, J. D. (2001) Role of macrophages and inflammatory mediators in chemically induced toxicity.
44             Toxicology (Ireland) 160: 111-118.
45       Laskin, D. L.; Pendino, K. J.; Punjabi,  C. J.; del Valle, M. R.; Laskin, J. D. (1994) Pulmonary and hepatic effects of
46             inhaled ozone in rats. Environ. Health Perspect. 102(suppl. 10): 61-64.
47       Laskin, J. D.; Heck, D. E.; Laskin, D. L. (1996) Nitric oxide production in the lung and liver following inhalation of
48             the pulmonary irritant ozone. In: Snyder, R.; Kocsis, J.  J.; Sipes, I. G.; Kalf, G. F.; Jollow, D. J.; Greim, H.;
49             Monks, T. J.;  Witmer,  C. M., eds.  Biological Reactive Intermediates V: Basic Mechanistic Research in
50             Toxicology and Human Risk Assessment: proceedings of the Fifth International Symposium; January 1995;
51             Munich, Germany. Adv. Exp. Med. Biol. 387: 141-146.
52       Laskin, D. L.; Heck, D.  E.; Laskin, J. D. (1998a) Role of inflammatory cytokines and nitric oxide in hepatic and
53             pulmonary toxicity. Toxicol. Lett. 102-103: 289-293.
54       Laskin, D. L.; Sunil, V.; Guo, Y.; Heck, D. E.; Laskin, J. D. (1998b) Increased nitric oxide synthase in the lung after
5 5             ozone inhalation is associated with activation of NF-KB. Environ. Health Perspect. 106(suppl. 5): 1175-1178.
56
         August 2005                                  AX5-69        DRAFT-DO NOT QUOTE OR CITE

-------
  1       Laskin, D. L.; Fakhrzadeh, L.; Heck, D. E.; Gerecke, D.; Laskin, J. D. (2002) Upregulation of phosphoinositide
  2             3-kinase and protein kinase B in alveolar macrophages following ozone inhalation. Role of NF-KB and
  3             STAT-1 in ozone-induced nitric oxide production and toxicity. Mol. Cell. Biochem. 234-235: 91-98.
  4       Last, J. A.; Pinkerton, K. E. (1997) Chronic exposure of rats to ozone and sulfuric acid aerosol: biochemical and
  5             structural responses. Toxicology 116: 133-146.
  6       Lavnikova, N.; Prokhorova, S.; Lakhotia, A. V.; Gordon, R.; Laskin, D. L. (1998) Distinct inflammatory responses
  7             of adherent vascular lung neutrophils to pulmonary irritants. J. Inflammation 48: 56-66
  8       Lee, C.; Watt,  K. C.; Chang, A. M; Plopper, C. G.; Buckpitt, A. R.; Pinkerton, K. E. (1998) Site-selective
  9             differences in cytochrome P450 isoform activities: comparison of expression in rat and rhesus monkey lung
10             and induction in rats. Drug Metab. Dispos. 26: 396-400.
11       Lemos, M.; Lichtenfels, A. J. F.  C.; Amaro, E., Jr.; Macchione, M.; Martins, M. A.; King, M.; Bohm, G. M.;
12             Saldiva, P. H. N. (1994) Quantitative pathology of nasal passages in rats exposed to urban levels of air
13             pollution. Environ. Res. 66: 87-95.
14       Longphre, M.; Zhang, L.-Y.; Harkema, J. R.; Kleeberger, S. R. (1999) Ozone-induced pulmonary inflammation and
15             epithelial proliferation are partially mediated by PAF. J. Appl. Physiol. 86:  341-349.
16       Madden, M. C.; Richards, J. H.;  Dailey, L. A.; Hatch, G. E.; Ohio, A. J. (2000) Effect of ozone on diesel exhaust
17             particle toxicity in rat lung. Toxicol. Appl. Pharmacol. 168: 140-148.
18       Mango, G. W.; Johnston, C. J.; Reynolds, S. D.; Finkelstein, J. N.; Plopper, C. G.; Stripp, B. R. (1998) Clara cell
19             secretory protein deficiency increases oxidant stress response in conducting airways. Am. J. Physiol.
20             275: L348-L356.
21       Mautz, W. J. (2003) Exercising animal models in inhalation toxicology: interactions with ozone and formaldehyde.
22             Environ. Res. 92:  14-26.
23       Mautz, W. J.; Kleinman, M. T.; Bhalla, D. K.; Phalen, R. F. (2001) Respiratory tract responses to repeated inhalation
24             of an oxidant and acid gas-particle air pollutant mixture. Toxicol. Sci. 61: 331-341.
25       McGraw, D. W.; Forbes, S. L.; Mak, J. C. W.; Witte, D. P.; Carrigan, P. E.; Leikauf, G. D.; Liggett, S. B. (2000)
26             Transgenic overexpression of beta(2)-adrenergic receptors in airway epithelial cells decreases
27             bronchoconstriction. Am. J. Physiol. 279: L379-L389.
28       McKinney, W. J.; Jaskot, R. H.;  Richards, J. H.; Costa, D. L.; Dreher, K. L. (1998) Cytokine mediation of
29             ozone-induced pulmonary adaptation. Am. J. Respir. Cell. Mol. Biol. 18: 696-705.
30       Miller, L. A.; Barnett, N. L.; Sheppard, D.; Hyde, D. M. (2001) Expression of the  P6 integrin subunit is associated
31             with sites of neutrophil influx in lung epithelium. J. Histochem. Cytochem.  49: 41-48.
32       Morio, L. A.; Hooper, K. A.; Brittingham, J.; Li, T.-H.; Gordon, R. E.; Turpin, B.  J.; Laskin, D. L. (2001) Tissue
3 3             injury following inhalation of fine  paniculate matter  and hydrogen perioxide is associated with altered
34             production of inflammatory mediators and antioxidants by alveolar macrophages. Toxicol. Appl. Pharmacol.
35             177: 188-199.
36       Moss, O. R.; Gross, E. A.; James, R. A.; Janszen, D. B.; Ross, P. W.; Roberts, K. C.; Howard, A. M.; Harkema,
37             J. R.; Calderon-Garciduenas, L.; Morgan, K. T. (2001) Respiratory tract toxicity in rats exposed to Mexico
3 8             City air. Cambridge, MA: Health Effects Institute; research report no. 100. Available:
39             http://www.healtheffects.org/pubs-research.htm [15 May, 2003].
40       Mudway, I. S.; Kelly, F. J. (1998) Modeling the interactions of ozone with pulmonary epithelial lining fluid
41             antioxidants. Toxicol. Appl. Pharmacol. 148: 91-100.
42       Neuhaus-Steinmetz, U.; Uffhausen, F.; Herz, U.; Renz, H. (2000) Priming of allergic immune responses by repeated
43             ozone exposure in mice. Am. J. Respir. Cell Mol. Biol. 23: 228-233.
44       Nichols, B. G.; Woods, J. S.; Luchtel,  D.  L.; Corral, J.; Koenig, J. Q. (2001) Effects of ozone exposure on nuclear
45             factor-KB activation and tumor  necrosis factor-a expression in human nasal epithelial cells. Toxicol. Sci.
46             60:356-362.
47       Nielsen, G. D.; Hougaard, K.  S.; Larsen, S. T.; Hammer, M.; Wolkoff, P.; Clausen, P. A.; Wilkins, C. K.; Alarie, Y.
48             (1999) Acute airway effects of formaldehyde and ozone inBALB/c mice. Hum. Exp. Toxicol. 18: 400-409.
49       Nino-Cabrera, H. G.; Colin-Barenque, L.; Avila-Costa, M. R.; Espinosa-Villanueva, J.; Fortoul, T. I.;
50             Rivas-Arancibia, S. (2002) Differences between hippocampus and cerebral cortex in aged rats in an oxidative
51             stress model. Int. J. Neurosci. 112: 373-381.
52       Noviski, N.; Brewer, J. P.; Skornik, W. A.; Galli, S. J.; Drazen, J. M.; Martin, T. R. (1999) Mast cell activation is
53             not required for induction of airway hyperresponsiveness by ozone in mice. J. Appl. Physiol.  86: 202-210.
54       Paige, R. C.; Royce, F. H.; Plopper, C. G.; Buckpitt, A. R. (2000a) Long-term exposure to ozone increases acute
55             pulmonary centriacinar injury by 1-nitronaphthalene: I. Region-specific enzyme activity. J. Pharmacol. Exp.
56             Ther. 295: 934-941.
         August 2005                                  AX5-70       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Paige, R. C.; Wong, V.; Plopper, C. G. (2000b) Long-term exposure to ozone increases acute pulmonary centriacinar
  2             injury by 1-nitronaphthalene: II. Quantitative histopathology. J. Pharmacol. Exp. Ther. 295: 942-950.
  3       Paquette, N. C.; Tankersley, C. G.; Zhang, L.-Y.; Kleeberger, S. R. (1994) Repeated subacute ozone exposure of
  4             inbred mice: airway inflammation and ventilation. Exp. Lung Res. 20: 579-594.
  5       Paz, C.; Bazan-Perkins, B. (1992) Sleep-wake disorganization in cats exposed to ozone. Neurosci. Lett.
  6             140:270-272.
  7       Paz, C.; Huitron-Resendiz, S. (1996) The effects of ozone exposure on the sleep-wake cycle and serotonin contents
  8             in the pons of the rat. Neurosci. Lett. 204: 49-52.
  9       Pearson, A. C.; Bhalla, D. K. (1997) Effects of ozone onmacrophage adhesion in vitro and epithelial and
10             inflammatory responses in vivo: the role of cytokines. J. Toxicol. Environ. Health 50: 143-157.
11       Peden, D. B.; Dailey, L. (1995) Modulation of mast cell functions by in vitro ozone exposure. Am. J. Physiol.
12             268: L902-L910.
13       Pendino, K. J.; Shuler, R. L.; Laskin, J. D.; Laskin, D. L. (1994) Enhanced production of interleukin-1, tumor
14             necrosis factor-a, and fibronectin by rat lung phagocytes following inhalation of a pulmonary irritant.
15             Am. J. Respir. Cell Mol. Biol. 11: 279-286.
16       Petruzzi, S.; Fiore, M; Dell'Omo, G.; Bignami, G.; Alleva, E. (1995) Medium and long-term behavioral effects in
17             mice of extended gestational exposure to ozone. Neurotoxicol.  Teratol. 17: 463-470.
18       Petruzzi, S.; De Acetis, L.; Chiarotti, F.; Sorace, A.; Alleva, E. (1999) Limited changes in handedness and morphine
19             reactivity in CD-I mice after pre- and postnatal ozone exposure. Acta Neurobiol. Exp. 59:  115-122.
20       Pinkerton, K. E.; Menache, M. G.; Plopper, C. G. (1995) Consequences of prolonged inhalation of ozone onF344/N
21             rats: collaborative studies. Part IX. Changes in the tracheobronchial epithelium, pulmonary acinus, and lung
22             antioxidant enzyme activity. Cambridge, MA: Health Effects Institute; pp. 41-98; research report no. 65.
23             Available from: NTIS, Springfield, VA; PB95-261996.
24       Pinkerton, K. E.; Weller, B. L.; Menache, M. G.; Plopper, C. G. (1998) Consequences of prolonged inhalation of
25             ozone on F344/N rats: collaborative studies. Part XIII. A comparison of changes in the tracheobronchial
26             epithelium and pulmonary acinus in male rats at 3 and 20 months. Cambridge,  MA: Health Effects Institute;
27             research report no. 65.
28       Plopper, C. G.; Fanucchi, M. V. (2000) Do urban environmental pollutants exacerbate childhood lung diseases?
29             Environ.  Health Perspect. 108: A252-A253.
30       Plopper, C. G.; Chu, F.-P.; Haselton, C. J.; Peake, J.; Wu, J.; Pinkerton, K. E. (1994) Dose-dependent tolerance to
31             ozone: I.  tracheobronchial epithelial reorganization in rats after 20 months' exposure. Am. J. Pathol.
32             144:404-420.
33       Plopper, C. G.; Hatch, G. E.; Wong, V.; Duan, X.; Weir, A. J.; Tarkington, B. K.; Devlin, R. B.; Becker, S.;
34             Buckpitt, A. R. (1998) Relationship of inhaled ozone concentration to acute tracheobronchial epithelial injury,
3 5             site-specific ozone dose and glutathione depletion in rhesus monkeys. Am. J. Respir. Cell Mol. Biol.
36             19:387-399.
37       Postlethwait, E. M.;  Cueto, R.; Velsor, L. W.; Pryor, W. A. (1998) Cyinduced formation of bioactive lipids:
38             estimated surface concentrations and lining layer effects. Am. J. Physiol. 274: L1006-L1016.
39       Pryor, W. A.; Squadrito, G. L.; Friedman, M. (1995) A new mechanism for the toxicity of ozone.  Toxicol. Lett.
40             82/83:287-293.
41       Pryor, W. A.; Bermudez, E.; Cueto, R.; Squadrito, G. L. (1996) Detection of aldehydes in bronchoalveolar lavage of
42             rats exposed to ozone. Fundam. Appl. Toxicol. 34: 148-156.
43       Pulfer, M. K.; Murphy, R. C. (2004) Formation of biologically active  oxysterols during ozonolysis of cholesterol
44             present in lung surfactant. J. Biol. Chem. 279: 26331-26338.
45       Pulfer, M. K.; Taube, C.; Gelfand, E.; Murphy, R. C. (2005) Ozone exposure in vivo and formation of biologically
46             active oxysterols in the lung. J. Pharmacol. Exp. Ther. 312: 256-264.
47       Reinhart, P. G.; Gupta, S. K.; Bhalla, D. K. (1999) Attenuation of ozone-induced lung injury by interleukin-10.
48             Toxicol. Lett. 110:35-42.
49       Rivas-Arancibia, S.; Vazquez-Sandoval, R.; Gonzalez-Kladiano, D.; Schneider-Rivas, S.; Lechuga-Guerrero, A.
50             (1998) Effects of ozone exposure in rats on memory and levels of brain and pulmonary superoxide dismutase.
51             Environ.  Res. 76: 33-39.
52       Rivas-Arancibia, S.; Dorado-Martinez, C.; Borgonio-Perez, G.; Hiriart-Urdanivia, M.; Verdugo-Diaz, L.;
53             Duran-Vazquez, A.; Colin-Baranque, L.; Avila-Costa, M. R. (2000) Effects of taurine on ozone-induced
54             memory deficits and lipid peroxidation levels in brains of young, mature, and old rats. Environ. Res. 82: 7-17.
55       Rivas-Manzano, P.; Paz, C. (1999) Cerebellar morphological alterations in rats induced by prenatal ozone exposure.
56             Neurosci. Lett. 276: 37-40.
         August 2005                                  AX5-71        DRAFT-DO NOT QUOTE OR CITE

-------
  1       Savov, J. D.; Whitehead, G. S.; Wang, J.; Liao, G.; Usuka, J.; Peltz, G.; Foster, W. M; Schwartz, D. A. (2004)
  2             Ozone-induced acute pulmonary injury in inbred mouse strains. Am. J. Respir. Cell Mol. Biol. 31: 69-77'.
  3       Schelegle, E. S.; Alfaro, M. F.; Putney, L.; Stovall, M.; Tyler, N.; Hyde, D. M. (2001) Effect of C-fiber-mediated,
  4             ozone-induced rapid shallow breathing on airway epithelial injury in rats. J. Appl. Physiol. 91: 1611-1618.
  5       Schelegle, E. S.; Miller, L. A.; Gershwin, L. J.; Fanucchi, M. V.; Van Winkle, L. S.; Gerriets, J. E.; Walby, W. F.;
  6             Mitchell, V.; Tarkington, B. K.; Wong, V. J.; Baker, G. L.; Pantle, L. M.; Joad, J. P.; Pinkerton, K. E.;
  7             Wu, R.; Evans, M. J.; Hyde, D. M.; Plopper, C. G. (2003) Repeated episodes of ozone inhalation amplifies
  8             the effects of allergen sensitization and inhalation on airway immune and structural development in Rhesus
  9             monkeys. Toxicol. Appl. Pharmacol. 191: 74-85.
10       Schlesinger, R. B.; Cohen, M. D.; Gordon, T.; Nadziejko, C.; Zelikoff, J. T.; Sisco, M.; Regal, J. F.; Menache, M. G.
11             (2002a) Ozone differentially modulates airway responsiveness in atopic versus nonatopic guinea pigs.
12             Inhalation Toxicol. 14:431-457.
13       Schlesinger, R. B.; Cohen, M.; Gordon, T.; Nadziejko, C.; Zelikoff, J. T.; Sisco, M.; Regal, J. F.; Menache, M. G.
14             (2002b) Ozone-induced modulation of airway hyperresponsiveness in guinea pigs. Boston, MA: Health
15             Effects Institute; research report no. 109.
16       Segura, P.; Montano, L. M.; Bazan-Perkins, B.; Gustin, P.; Vargas, M. H. (1997) Ozone at high-pollution urban
17             levels causes airway hyperresponsiveness to substance P but not to other agonists. Environ. Toxicol.
18             Pharmacol. 3: 91-95.
19       Sen, S.; Dulchavsky, S. A.; Dutta, S. (1993) Effects of triiodothyronine (T3) supplementation upon ozone-induced
20             lung injury. Free Radic. Res. Commun. 18: 299-308.
21       Shore, S. A.; Abraham, J. H.; Schwartzman, I. N.; Murthy, G. G.; Laporte, J. D. (2000) Ventilatory responses to
22             ozone are reduced in immature rats. J. Appl. Physiol. 88: 2023-2030.
23       Shore, S. A.; Johnston, R. A.; Schwartzman, I. N.; Chism, D.; Krishna Murthy, G. G. (2002) Ozone-induced airway
24             hyperresponsiveness is reduced in immature mice. J. Appl. Physiol. 92: 1019-1028.
25       Shore, S. A.; Rivera-Sanchez, Y. M.; Schwartzman, I. N.; Johnston, R. A. (2003) Responses to ozone are increased
26             in obese mice. J. Appl. Physiol. 95: 938-945.
27       Sindhu, R. K.; Mautz, W. J.; Kikkawa, Y. (1998) Chronic exposure to ozone and nitric acid vapor results in
28             increased levels of rat pulmonary putrescine. Arch. Toxicol. 72:  445-449.
29       Slade, R.; Watkinson, W. P.; Hatch, G. E. (1997) Mouse strain differences in ozone dosimetry and body temperature
30             changes. Am. J. Physiol. 272: L73-L77.
31       Sorace, A.; De Acetis, L.; Alleva, E.; Santucci, D. (2001) Prolonged exposure to low doses of ozone: short- and
32             long-term changes to behavioral performance in mice. Environ. Res. 85: 122-134.
33       Spannhake, E. W. (1996) Down-regulation of canine airway mast cell function following exposure to ozone in vivo.
34             Exp. Lung Res. 22:  163-178.
35       Sterner-Kock, A.; Kock, M.; Braun, R.; Hyde, D. M. (2000) Ozone-induced epithelial injury in the ferret is similar
36             to nonhumanprimates. Am. J. Respir. Crit. Care Med. 162: 1152-1156.
37       Stockstill, B. L.; Chang, L.-Y.; Menache, M. G.; Mellick, P. W.; Mercer, R. R.; Crapo, J. D. (1995) Bronchiolarized
3 8             metaplasia and interstitial fibrosis in rat lungs chronically exposed to high ambient levels of ozone. Toxicol.
39             Appl. Pharmacol.  134: 251-263.
40       Sun, J.; Chung, K. F. (1997) Interaction of ozone exposure with airway hyperresponsiveness and inflammation
41             induced by trimellitic anhydride in sensitized guinea pigs. J. Toxicol. Environ. Health 51: 77-87.
42       Sun, J.; Koto, H.; Chung, K. F. (1997) Interaction of ozone and allergen challenges on bronchial responsiveness and
43             inflammation in sensitised guinea pigs. Int. Arch.  Allergy Immunol. 112: 191-195.
44       Szarek, J. L.; Stewart, N. L.; Zhang, J. Z.; Webb, J. A.; Valentovic, M. A.; Catalano, P. (1995) Contractile responses
45             and structure of small bronchi isolated from rats after 20 months' exposure to ozone. Fundam. Appl. Toxicol.
46             28: 199-208.
47       Takahashi, T.; Miura, M.; Katsumata, U.; Ichinose, M.; Kimura, K.; Inoue, H.; Takishima, T.;  Shirato, K.  (1993)
48             Involvement of superoxide in ozone-induced airway hyperresponsiveness in anesthetized cats. Am. Rev.
49             Respir. Dis. 148: 103-106.
50       Takahashi, N.; Yu, X.-Y.; Schofield, B. H.; Kleeberger, S. R.; Scott, A. L.; Hasegawa, S.;  Spannhake, E. W. (1995a)
51             Expression of ICAM-1 in airway epithelium after acute ozone exposure in the mouse. J. Appl. Physiol.
52             79: 1753-1761.
53       Takahashi, M.; Kleeberger, S. R.; Croxton, T. L. (1995b) Genetic control of susceptibility to ozone-induced changes
54             in mouse trachea! electrophysiology. Am. J. Physiol. 269: L6-L10.
55       Takebayashi, T.; Abraham, J.; Murthy, G. G. K.; Lilly, C.; Rodger, I.; Shore, S. A. (1998) Role of tachykinins in
56             airway responses to ozone in rats. J. Appl. Physiol. 85: 442-450.
         August 2005                                  AX5-72        DRAFT-DO NOT QUOTE OR CITE

-------
  1       Tankersley, C. G.; Kleeberger, S. R. (1994) Ozone-induced inflammation and altered ventilation in genetically
  2             susceptible mice: a comparison of acute and subacute exposures. Toxicol Lett. 72: 279-289.
  3       Tankersley, C. G.; Fitzgerald, R. S.; Mitzner, W. A.; Kleeberger, S. R. (1993) Hypercapnic ventilatory responses in
  4             mice differentially susceptible to acute ozone exposure. J. Appl. Physiol. 75: 2613-2619.
  5       Tesfaigzi, J.; Hotchkiss, J. A.; Harkema, J. R. (1998) Expression of the Bcl-2 protein in nasal epithelia of F344/N
  6             rats during mucous cell metaplasia and remodeling. Am. J. Respir. Cell Mol. Biol. 18: 794-799.
  7       Uhlson, C.; Harrison, K.; Allen, C. B.; Ahmad, S.; White, C. W.; Murphy, R. C. (2002) Oxidized phospholipids
  8             derived from ozone-treated lung surfactant extract reduce macrophage and epithelial cell viability. Chem.
  9             Res. Toxicol. 15: 896-906.
10       Ulrich, M. M. W.; Alink, G. M; Kumarathasan, P.; Vincent, R.; Boere, A. J.; Cassee, F. R. (2002) Health effects
11             and time course of paniculate matter on the cardiopulmonary system in rats with lung inflammation.
12             J. Toxicol. Environ. Health Part A 65: 1571-1595.
13       Valacchi, G.; Weber, S. U.; Luu, C.; Cross, C. E.; Packer, L. (2000) Ozone potentiates vitamin E depletion by
14             ultraviolet radiation in the murine stratum corneum. FEES Lett. 466: 165-168.
15       Valacchi, G.; Pagnin, E.; Okamoto, T.; Corbacho, A. M.; Olano, E.; Davis, P. A.; Van der  Vliet, A.; Packer, L.;
16             Cross, C. E.  (2003) Induction of stress proteins and MMP-9 by 0.8 ppm of ozone in murine skin. Biochem.
17             Biophys. Res. Commun. 305: 741-746.
18       Van Bree, L.; Dormans, J. A. M. A.; Boere, A. J. F.; Rombout, P. J. A. (2001) Time study  on development and
19             repair of lung injury following ozone exposure in rats. Inhalation Toxicol. 13: 703-717.
20       Van Bree, L.; Dormans, J. A. M. A.; Koren, H. S.; Devlin, R. B.; Rombout, P. J. A. (2002) Attenuation and recovery
21             of pulmonary injury in rats following short-term, repeated daily exposure to ozone.  Inhalation Toxicol.
22             14:883-900.
23       Vargas, M. H.; Romero, L.; Sommer, B.; Zamudio, P.; Gustin, P.; Montano, L. M. (1998) Chronic exposure to ozone
24             causes tolerance to  airway hyperresponsiveness in guiea pigs: lack of SOD role.  J. Appl. Physiol.
25             84: 1749-1755.
26       Vesely, D. L.; Giordano, A. T.; Raska-Emery, P.; Montgomery, M. R. (1994a) Ozone increases amino- and
27             carboxy-terminal atrial natriuretic factor prohormone peptides in lung, heart, and circulation. J. Biochem.
28             Toxicol. 9: 107-112.
29       Vesely, D. L.; Giordano, A. T.; Raska-Emery, P.; Montgomery, M. R. (1994b) Increase in atrial natriuretic factor in
30             the lungs, heart, and circulatory system owing to ozone. Chest 105: 1551-1554.
31       Vesely, D. L.; Giordano, A. T.; Raska-Emery, P.; Montgomery, M. R. (1994c) Ozone increases atrial natriuretic
32             peptides in heart, lung and circulation of aged vs. adult animals. Gerontology (Basel) 40: 227-236.
33       Vesely, K. R.; Schelegle, E.  S.; Stovall, M. Y.; Harkema, J. R.; Green, J. F.; Hyde, D. M. (1999) Breathing  pattern
34             response and epithelial labeling in ozone-induced airway injury in neutrophil-depleted rats. Am. J. Respir.
35             Cell Mol. Biol. 20:  699-709.
36       Vincent, R.; Janzen, E. G.; Chen, G.; Kumarathasan, P.; Haire, D. L.; Guenette, J.; Chen, J. Z.; Bray, T. M.  (1996)
3 7             Spin trapping study in the lungs and liver of F344 rats after exposure to ozone. Free Radical Res. 25: 475-488.
38       Vincent, R.; Bjarnason, S. G.; Adamson, I. Y. R.; Hedgecock,  C.; Kumarathasan, P.; Guenette, J.; Potvin, M.;
39             Goegan, P.; Bouthillier, L. (1997) Acute pulmonary toxicity of urban paniculate matter and ozone. Am. J.
40             Pathol. 151:  1563-1570.
41       Wagner, J. G.; Hotchkiss,  J.  A.; Harkema, J. R. (200la) Effects of ozone and endotoxin coexposure on rat airway
42             epithelium: potentiation of toxicant-induced alterations. Environ. Health Perspect. 109(suppl. 4): 591-598.
43       Wagner, J. G.; VanDyken, S. J.; Hotchkiss, J. A.; Harkema, J. R. (200Ib) Endotoxin enhancement of ozone-induced
44             mucous cell metaplasia is neutrophil-dependent in rat nasal epithelium. Toxicol.  Sci. 60: 338-347.
45       Wagner, J. G.; Hotchkiss,  J.  A.; Harkema, J. R. (2002) Enhancement of nasal inflammatory and epithelial responses
46             after ozone and allergen coexposure in brown Norway rats. Toxicol. Sci. 67: 284-294.
47       Wagner, J. G.; VanDyken, S. J.; Wierenga, J. R.; Hotchkiss, J. A.; Harkema, J. R. (2003) Ozone exposure enhances
48             endotoxin-induced  mucous cell metaplasia in rat pulmonary airways. Toxicol. Sci. 74: 437-446.
49       Wang, G.; Umstead, T. M.; Phelps, D. S.; Al-Mondhiry, H.; Floras, J. (2002) The effect of ozone exposure  on the
50             ability of human surfactant protein A variants to stimulate cytokine production. Environ. Health Perspect.
51             110:79-84.
52       Watkinson, W. P.; Wiester, M. J.; Highfill, J. W. (1995) Ozone toxicity in the rat. I. Effect of changes in ambient
53             temperature on extrapulmonary physiological parameters. J. Appl. Physiol. 78: 1108-1120.
54       Watkinson, W. P.; Campen,  M. J.; Nolan, J. P.; Costa, D. L. (2001) Cardiovascular and systemic responses  to
5 5             inhaled pollutants in rodents: effects of ozone and paniculate matter. Environ. Health Perspect.
56             109(suppl. 4): 539-546.
         August 2005                                   AX5-73        DRAFT-DO NOT QUOTE OR CITE

-------
  1       Watkinson, W. P.; Campen, M. J.; Wichers, L. B.; Nolan, J. P.; Costa, D. L. (2003) Cardiac and thermoregulatory
  2             responses to inhaled pollutants in healthy and compromised rodents: modulation via interaction with
  3             environmental factors. Environ. Res. 92: 35-47.
  4       Watt, K. C.; Plopper, C. G.; Weir, A. J.; Tarkington, B.; Buckpitt, A. R. (1998) Cytochrome P450 2E1 in rat
  5             tracheobronchial airways: response to ozone exposure. Toxicol. Appl. Pharmacol. 149: 195-202.
  6       Weber, S. U.; Thiele, J. J.; Cross, C. E.; Packer, L. (1999) Vitamin C, uric acid, and glutathione gradients in murine
  7             stratum corneum and their susceptibility to ozone exposure. J. Invest. Dermatol. 113:1128-1132.
  8       Weber, S. U.; Jothi, S.; Thiele, J. J. (2000) High-pressure liquid chromatography analysis of ozone-induced
  9             depletion of hydrophilic and lipophilic antioxidants in murine skin. Methods Enzymol. 319: 536-546.
10       Weber, S. U.; Han, N.; Packer, L. (2001) Ozone: an emerging oxidative stressorto skin. Curr. Probl. in Dermatol.
11             29:52-61.
12       Weller, B. L.; Crapo, J. D.; Slot,  J.; Posthuma, G.; Plopper, C.  G.; Pinkerton, K. E. (1997) Site- and cell-specific
13             alteration of lung copper/zinc and manganese superoxide dismutases by chronic ozone exposure. Am. J.
14             Respir. Cell Mol. Biol. 17: 552-560.
15       Weller, B. L.; Witschi, H.; Pinkerton, K. E. (2000) Quantitation and localization of pulmonary manganese
16             superoxide dismutase and tumor necrosis factor a following exposure to ozone and nitrogen dioxide. Toxicol.
17             Sci. 54: 452-461.
18       Wiester, M. J.; Watkinson, W. P.; Costa, D. L.; Crissman, K. M.; Richards, J. H.; Winsett, D. W.; Highfill, J. W.
19             (1996) Ozone toxicity in the rat. III. Effect of changes in ambient temperature on pulmonary parameters.
20             J. Appl. Physiol. 81: 1691-1700.
21       Witschi, H.; Espiritu, L; Pinkerton, K.  E.; Murphy, K.; Maronpot, R. R. (1999) Ozone carcinogenesis revisited.
22             Toxicol.  Sci. 52: 162-167.
23       Wu, Z.-X.; Morton, R. F.; Lee, L.-Y. (1997) Role of tachykinins in ozone-induced airway hyperresponsiveness to
24             cigarette  smoke in guinea pigs. J. Appl. Physiol.  83: 958-965.
25       Yamauchi, T.; Shima, M.; Kuwaki, T.; Ando, M.; Ohmichi, M.; Fukuda, Y.; Adachi, M. (2002) Acute effects of
26             ozone exposure on lung function in mice sensitized to ovalbumin. Toxicology (Ireland) 172: 69-78.
27       Yu, M.; Pinkerton, K. E.; Witschi, H. (2002) Short-term exposure to aged and diluted sidestream cigarette smoke
28             enhances ozone-induced lung injury inB6C3Fl mice. Toxicol. Sci. 65: 99-106.
29       Zhang, L.-Y.; Levitt, R. C.; Kleeberger, S. R. (1995) Differential susceptibility to ozone-induced airways
30             hyperreactivity in inbred strains of mice. Exp. Lung Res. 21: 503-518.
31       Zhao, Q.; Simpson, L.  G.; Driscoll, K. E.; Leikauf, G. D. (1998) Chemokine regulation of ozone-induced neutrophil
32             and monocyte inflammation. Am. J. Physiol. 274: L39-L46.
33
34
         August 2005                                   AX5-74        DRAFT-DO NOT QUOTE OR CITE

-------
 i         ANNEX AX6.  CONTROLLED HUMAN EXPOSURE
 2                   STUDIES OF OZONE AND RELATED
 3                       PHOTOCHEMICAL OXIDANTS
 4
 5
 6     AX6.1  INTRODUCTION
 7          Results of ozone (O3) studies in laboratory animals and in vitro test systems were presented
 8     in Chapter 5 and Annex 5.  The extrapolation of results from animal studies is one mechanism
 9     by which information on potential adverse human health effects from exposure to O3 is obtained.
10     More direct evidence of human health effects due to O3 exposure can be obtained through
11     controlled human exposure studies of volunteer subjects or through field and epidemiologic
12     studies of populations exposed to ambient O3. Controlled human exposure studies, discussed in
13     this chapter, typically use fixed concentrations of O3 under carefully regulated environmental
14     conditions and subject activity levels.
15          Most of the scientific information selected for review and evaluation in this chapter comes
16     from the literature published since 1996 which, in addition to further study of physiological
17     pulmonary responses and respiratory symptoms, has focused on mechanisms of inflammation
18     and cellular responses to injury induced by O3 inhalation. Older studies are discussed where
19     only limited new data are available and where new and old data are conflicting.  The reader is
20     referred to both the 1986 and 1996 Air Quality Criteria documents (U.S. Environmental
21     Protection Agency, 1986, 1996) for a more extensive discussion of older studies. Summary
22     tables of the relevant O3 literature are included for each of the major subsections.
23     In summarizing the human health effects literature, changes from control are described if
24     statistically significant at a probability (p) value less than 0.05,  otherwise trends are noted
25     as such.
26
27
28
       August 2005                            AX6-1       DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX6.2   PULMONARY FUNCTION EFFECTS OF OZONE EXPOSURE IN
 2               HEALTHY SUBJECTS
 3     AX6.2.1  Introduction
 4           The responses observed in young healthy nonsmoking human adults exposed to ambient O3
 5     concentrations include decreased inspiratory capacity; mild bronchoconstriction; rapid, shallow
 6     breathing pattern during exercise; and symptoms of cough and pain on deep inspiration.
 7     In addition, O3 has been shown to result in airway hyperresponsiveness as demonstrated by an
 8     increased physiological response to a nonspecific bronchoconstrictor, as well as airway injury
 9     and inflammation assessed via bronchoalveolar lavage and biopsy. Reflex inhibition of
10     inspiration and consequent decrease in inspiratory capacity results in a decrease in forced vital
11     capacity (FVC) and total lung capacity (TLC) and, in combination with mild
12     bronchoconstriction, contributes to a decrease in the forced expiratory volume in 1 s (FEVj).
13     Given that both FEVj and FVC are subject to decrease with O3 exposures, changes in the
14     ratio (FEVj/FVC) become difficult to interpret and so are not discussed.
15           The majority of controlled human studies have investigated the effects of exposure to
16     variable O3 concentrations in healthy subjects performing continuous exercise (CE) or
17     intermittent exercise (IE) for variable periods of time. These studies have several important
18     limitations: (1) the ability to study only short-term, acute effects; (2) the inability to link short-
19     term effects with long-term consequences; (3) the use of a small number of volunteers that may
20     not be representative of the general population; and (4) the statistical limitations associated with
21     the small sample size.  Nonetheless, studies reviewed in the 1996 EPA criteria document
22     (U.S. Environmental Protection Agency, 1996) provided a large body of data describing the
23     effects and dose-response characteristics of O3 as function of O3 concentration (C), minute
24     ventilation (VE), and duration or time (T) of exposure.  In most of these studies, subjects were
25     exposed to O3 and to filtered air (FA [reported as 0 ppm O3]) as a control. The most salient
26     observations from these studies were:  (1) healthy subjects exposed to O3 concentrations
27     >0.08 ppm develop  significant reversible, transient decrements in pulmonary function if VE or
28     T are increased sufficiently, (2) there is a large degree of intersubject variability in physiologic
29     and symptomatic responses to O3 and these responses tend to be reproducible within a given
30     individual over a several months period,  and (3) subjects exposed repeated to O3 over several
       August 2005                            AX6-2       DRAFT-DO NOT QUOTE OR CITE

-------
 1      days develop a tolerance to successive exposures, as demonstrated by an attenuation of
 2      responses, which is lost after about a week without exposure.
 3          In this section, the effects of single O3 exposures of 1- to 8-h in duration on pulmonary
 4      function in healthy nonsmoking subjects are examined by reviewing studies that investigate:
 5      (1) the O3 exposure-response relationship; (2) intersubject variability, individual sensitivity, and
 6      the association between responses; and (3) mechanisms of pulmonary function responses and the
 7      relationship between tissue-level events and functional responses.  Discussion will largely be
 8      limited to studies published subsequent to the 1996 EPA criteria document (U.S. Environmental
 9      Protection Agency, 1996)
10
11      AX6.2.2  Acute Ozone Exposures for Up to 2 Hours
12          At-Rest Exposures. Exposure studies investigating the effects of O3 exposures on sedentary
13      subjects were discussed in the 1986 EPA criteria document (U.S. Environmental Protection
14      Agency, 1986). The lowest O3 concentration at which significant reductions in FVC and FEVj
15      were reported was 0.5 ppm (Folinsbee et al., 1978; Horvath et al., 1979). Based on the
16      average O3 responses in these two studies (corrected for FA responses), resting young adults
17      (n = 23, age = 22) exposed to 0.5  ppm O3 have  a -4% reduction in FVC and a -7% reduction
18      FEVj.  At lower O3 concentrations of 0.25 to 0.3 ppm, resting exposures did not significantly
19      affect lung function.
20          Exposures with Exercise. Collectively, the studies reviewed in the 1996 EPA criteria
21      document (U.S. Environmental Protection Agency, 1996) demonstrated that healthy young
22      adults performing moderate to heavy IE or CE of 1 to 2.5 h duration, exposed to 0.12 to
23      0.18 ppm O3 experienced statistically significant decrements in pulmonary function and
24      respiratory symptoms. As an example, 2 hr exposures to 0.12 and 0.18 ppm O3 during heavy IE
25      (exercise VE  = 65 L/min) have resulted in FEVj decrements of 2.0 ± 0.8% (mean ± SE; n = 40)
26      and 9.5 ± 1.1% (n = 89), respectively (McDonnell and Smith,  1994).  Significant decrements in
27      pulmonary function  have been reported in heavily exercising healthy adults exposed for 1 h with
28      CE at O3 concentrations of 0.12 ppm (Gong et al., 1986), 0.16 ppm (Avol et al., 1984), and
29      0.2 ppm (Adams and Schelegle, 1983; Folinsbee et al., 1984).
30          In an attempt to describe O3 dose-response characteristics, many investigators modeled
31      acute responses as a function of total inhaled O3 dose (C * T x VE), which was found to be a

        August 2005                             AX6-3      DRAFT-DO NOT  QUOTE OR CITE

-------
 1      better predictor of response than O3 concentration, VE, or T of exposure, alone. In an analysis of
 2      6 studies with 1 to 2 h exposures to between 0.12 and 0.18 ppm O3 with exercise, Folinsbee et al.
 3      (1988) reported a good correlation (r = 0.81) between total inhaled O3 dose and FEVj
 4      decrements. For a given exposure duration, total inhaled O3 dose can be increased by increases
 5      in C and/or VE . In exposures of fixed duration, results of several studies suggested that O3
 6      concentration was a more important predictor of response or explained more of the variability in
 7      response than VE  (Adams et al., 1981; Folinsbee et al, 1978; Hazucha, 1987). Based on a review
 8      of previously published studies, Hazucha (1987) noted that relative to the FEVj decrement
 9      occurring at a given C and VE, doubling C (e.g., from 0.1 to 0.2 ppm) would increase the FEVj
10      decrement by 400%, whereas doubling the VE (e.g., from an exercise VE of 20 to 40 L/min)
11      which would only increase the FEVj decrement by 190%. Thus, C appears to have a greater
12      affect than VE on FEVj responses even when total inhaled O3 doses are equivalent.
13           New studies (i.e., not reviewed in the 1996 EPA criteria document) that provide
14      spirometric responses for up to 2 h exposures are summarized in Table AX6-1. Most of these
15      newer studies have investigated mechanisms affecting responses, inflammation, and/or effects in
16      diseased groups versus healthy adults, accordingly their findings may be summarized differently
17      in several sections of this chapter.  Rather than being interested in responses due to O3 versus FA
18      exposures, many of the newer studies have tested the effects of a placebo versus treatment in
19      modulating responses to O3 exposure.  Studies appearing in Table 1, but not discussed in  this
20      section, are discussed in other sections of this chapter as indicated within the table.
21           McDonnell  et al. (1997) pooled the results of eight studies entailing 485 healthy male
22      subjects exposed for 2 h on one occasion to one of six O3 concentrations (0.0, 0.12, 0.18,  0.24,
23      0.30, or 0.40 ppm) at rest or one of two levels of IE (VE of 25 and 35 L/min/m2 BSA).  FEVj
24      was measured preexposure, after 1 h of exposure, and immediately postexposure. Decrements
25      in FEVj were modeled by sigmoid-shaped curve as a function of subject age, O3 concentration,
26      VE, and T. The modeled decrements reach a plateau with increasing T and dose rate (C * VE).
27      That is, for a given O3 concentration, exercise VE level, and after a certain length  of exposure,
28      the FEVj response tends not to increase further with increasing duration of exposure. The
29      modeled FEVj responses increased with C * VE and T, decreased with subject age, but were only


        August 2005                             AX6-4      DRAFT-DO NOT QUOTE OR CITE

-------
                         Table AX6-1. Controlled Exposure of Healthy Humans to Ozone for 1 to 2 Hours During Exercisea
C/3
rt Ozone
0 Concentration" Exposure
55 Duration and
ppm Hg/m3 Activity
0.0 0 2hIE
0.4 784 4xl5min
on bicycle,
VE = 30 L/min
Exposure Number and
Conditions Gender of Subject
c Subjects Characteristics
NA 5 M, 4 F Healthy adults
25 ± 2 years old
6 M, 7 F Mild atopic
asthmatics
22 ± 0.7 years old
Observed Effect(s)
O3-induced reductions in FVC (12%, 10%) and FEV;
(13%, 11%) for asthmatic and healthy subjects.
Significant reductions in mid-flows in both asthmatics and
healthy subjects. Indomethacin pretreatment significantly
decreased FVC and FEV[ responses to O3 in healthy but not
asthmatic subjects. See Section AX6. 3. 2 and Tables AX6-3
andAX6-13.
Reference
Alexis et al.
(2000)
X
H
6
o
o
H
O
o
H
W
O
0.0
0.2
0.0
0.2
          0
         392
          0
         392
 0.0
0.33
 0
647
       2hIE
       4x15 min
       at VE = 20
       L/min/m2 BSA
       2hIE
       4x15 min
       at VE = 20
       L/min/m2 BSA
2hIE
4x15 min
on bicycle
ergometer
(600 kpm/min)
                   20 °C
                   50% RH
                   20 °C
                   50% RH
                                                            8 M, 5 F
                                                           10 M, 12 F
Healthy NS
median age 23 years
Healthy NS
mean age 24 years
                                             NA
                                                   9 M
Healthy NS
26.7 ± 7 years old
Median O3-induced decrements of 70 mL, 190 mL, and        Blomberg
400 mL/s in FVC, FEVj, and FEF25.75, respectively.           et al. (1999)
Spirometric responses not predicted of inflammatory
responses. See Sections AX6.2.5.2, AX6.5.6, andAX6.9.3
andTableAX6-12.

Significant O3-induced decrement in FEV[ immediately        Blomberg
postexposure but not significantly different from baseline      et al. (2003)
2 h later.  No correlation between Clara cell protein (CC16)
and FEV; decrement.  CC16 levels, elevated by  O3 exposure,
remained high at 6 h postexposure, but returned to baseline
by 18 h postexposure. See Table AX6-12

O3-induced reductions in FVC (7%).  FRC not altered by O3    Foster et al.
exposure. Post FA, normal gradient in ventilation which       (1993)
increased from apex to the base of the lung.  Post O3,
ventilation shifted away from the lower-lung into middle
and upper-lung regions. The post O3 increase in ventilation
to mid-lung region was correlated with decrease in
midmaximal expiratory flow (r = 0.76, p < 0.05).
0.0
0.35
0
690
2.2 h IE
2 x 30 min
on treadmill
(VE <= 50 L/min)
Final 10 min rest
19-23 °C
48-55%
RH
15 M Healthy NS
25.4 ±2 years old
Pre- to post-O3, mean FVC and FEV; decreased by 12 and
14%, respectively. Following O3 exposure, there was a
pronounced slow phase evident in multibreath nitrogen
washouts which, on average, represented a 24% decrease
in the washout rate relative to pre-O3.
Foster et al.
(1997)
O

-------
                    Table AX6-1 (cont'd). Controlled Exposure of Healthy Humans to Ozone for 1 to 2 Hours during Exercise"
r+ Ozone
O Concentrationb
L/I
ppm
0.0
0.12
0.18
0.24
0.30
0.40

ug/m3
0
235
353
471
589
784
Exposure
Duration and
Activity
2 h rest or IE
(4 x 15 min
at VE = 25 or 35
L/min/m2 BSA)


Exposure Number and
Conditions Gender of
c Subjects
22 °C 485 M (each
40% RH subject
exposed at one
activity level
to one O3
concentration)

Subject
Characteristics
Healthy NS
18 to 36 years old
mean age 24 years





Observed Effect(s)
Statistical analysis of 8 experimental chamber studies
conducted between 1980 and 1993 by the U.S. EPA in
Chapel Hill, NC. Decrement in FEV; described by sigmoid-
shaped curve as a function of subject age, O3 concentration,
VE, and time. Response decreased with age, was minimally
affected by body size corrections, and was not more


Reference
McDonnell
etal. (1997)




                                                                                           sensitive to O3 concentration than VE. Also see Section
                                                                                           AX6.5.
X
Oi
           0.4
784
2hIE
20 min mild-mod.
exercise,
10 min rest
NA
4 M, 5 F     Healthy NS           Subjects previously in Nightingale et al. (2000) study.
            30 ± 3 years old        Placebo-control: Immediately postexposure decrements
                                 in FVC (9%) and FEV1 (14%) relative to pre-exposure
                                 values. FEV; decrement only 9% at 1 hr postexposure.
                                 By 3 h postexposure, recovery in FVC to 97% and FEVj
                                 to 98% of preexposure values. Significant increases in
                                 8-isoprostane at 4 h postexposure. Budesonide for
                                 2 wk prior to exposure did not affect responses.
Montuschi
et al. (2002)
H
6
o
o
H
O
o
H
W
O
0.0 392 2hIE 20 °C
0.2 4x15 min 50% RH
at VE = 20
L/min/m2 BSA


0.4 784 2hIE NA
20 min mild-mod.
exercise,
10 min rest




6 M, 9 F Healthy adults
24 years old

9 M, 6 F Mild asthmatics
29 years old

6 M, 9 F Healthy NS
mean age ~3 1 years






O3-induced FEVj decrement (8%, healthy adults;
3% asthmatics) and PMN increase (20.6%, healthy adults;
15.2% asthmatics). Primary goal was to investigate
relationship between antioxidant defenses and O3
responses in asthmatics and healthy adults.
See Tables AX6-3 andAX6-13.
Placebo-control: O3 caused significant decrements in FEVj
(13.5%) and FVC (10%) immediately following exposure,
a small increase in Mch-reactivity, and increased PMNs and
myeloperoxidase in induced sputum at 4 h postexposure.
FEV; at 96% and FVC at 97% preexposure values at 3 h
postexposure. Budesonide for 2 wk prior to exposure did
not affect spirometric responses. See Section AX6.2.5 and
Table AX6-13.
Mudway
etal. (2001)
Stenfors
et al. (2002)


Nightingale
et al. (2000)






O

-------
                   Table AX6-1 (cont'd).  Controlled Exposure of Healthy Humans to Ozone for 1 to 2 Hours during Exercise"
<-K Ozone
O Concentration" Exposure
J~| Duration and
ppm Hg/m3 Activity
0.0 784 2hIE
0.4 4 x ISminat
VE=18L/min/m2
BSA

2 exposures:
25% subjects
exposed to air-air,
75% to O3-O3
0.0 784 2hIE
0.4 4 x 15min
at VE = 20
L/min/m2 BSA
^
!x!
Exposure Number and
Conditions Gender of Subject
c Subjects Charactenstics
21 °C Weak
40% RH responders Healthy NS
7M, 13F 20 to 59 years old

Strong
responders
21M,21F


20 °C Placebo group
40% RH 1 5 M, 1 F Healthy NS
mean age 27 years
Antioxidant
group
13M,2F
Observed Effect(s)
Significant O3-induced decrements in spirometric lung
function. Young adults (<35 years) were significantly more
responsive than older individuals (>35 years). Sufentanil,
a narcotic analgesic, largely abolished symptom responses
and improved FEV[ in strong responders. Naloxone, an
opioid antagonist, did not affect O3 effects in weak
responders. See Section AX6. 2. 5.1.


Placebo and antioxidant groups had O3-induced decrements
in FEVj (20 and 14%) and FVC (13 and 10%), respectively.
Percent neutrophils and IL-6 levels in BAL fluid obtained
1 h postexposure were not different in the two treatment
groups. See Table AX6-13.

Reference
Passannante
etal. (1998)







Samet et al.
(2001)
Steck-Scott
et al. (2004)


0.0
0.25
                  490
IhCE
VE = 30 L/min
NA
Face mask
exposure
                                                         32 M, 28 F
Healthy NS
22.6 ± 0.6 years old
Mean O3-induced FEV; decrements of 15.9% in males and
9.4% in females (gender differences not significant). FEV[
decrements ranged from -4 to 56%; decrements >15% in
20 subjects and >40% in 4 subjects. Uptake of O3 greater in
males than females, but uptake not correlated with
spirometric responses.
Ultman et al.
(2004)
H
6
o
o
H
O
o
H
W
O
         "See Appendix A for abbreviations and acronyms.
         bListed from lowest to highest O3 concentration.
         'Studies conducted in exposure chamber unless otherwise indicated.
O

-------
1     minimally affected by body size corrections to VE. Fitted and experimental FEVj decrements

2     following a 2 h exposure at three nominal levels of VE are illustrated in Figure AX6-1 as a

3     function of O3 concentration. Their analysis indicated that C was marginally, but not

4     significantly, more important than VE in predicting FEVj response.  Additionally, the McDonnell

5     et al. (1997) analysis revealed that some prior analyses of IE protocols may have over estimated

6     the relative importance of C over VE in predicting FEVj responses by considering only the VE

7     during exercise and ignoring the VE during periods of rest.
                   24-
                   20-
               •£  16-

                E
                «  12-|
                o

               I   .H
               HI
                    4-
                    0-
                   -4-
                       0.0
                             T = 2h
0.1
0.2         0.3
  O3 (ppm)
                                                                       Ve = 40 L/min (n)
                                                                          31 L/min (o)
                                                                       Ve = 10 L/min (A)
0.4
0.5
      Figure AX6-1. FEVt decrements as a function of O3 concentration following a 2 h
                     exposure with incremental exercise (15 min intervals) or rest. Points are
                     experimental data (mean ± SE) and  lines are model predictions for each
                     activity level. Minute ventilation (VE) represent average across intervals
                     of rest and exercise.

      Source: McDonnell et al. (1997).
      August 2005
            AX6-8
             DRAFT-DO NOT QUOTE OR CITE

-------
 1          Ultman et al. (2004) measured O3 uptake and pulmonary responses in 60 young heathy
 2      nonsmoking adults (32 M, 28 F).  A bolus technique was used to quantify the uptake of O3 as a
 3      function of the volume into the lung which the bolus penetrated. From these measurements, the
 4      volumetric depth at which 50% uptake occurred was calculated. This volumetric lung depth was
 5      correlated with conducting airways volume, i.e., a greater fraction of O3 penetrated to deeper into
 6      the lungs of individuals have larger conducting airways volumes. Two weeks after the bolus
 7      measurements, subjects were exposed via a face mask to FA and subsequently two weeks later to
 8      0.25 ppm O3 for 1 h with CE at a target VE of 30 L/min. The breath-by-breath uptake of O3 was
 9      measured.  There was a small but significant reduction in the breath-by-breath uptake of O3 from
10      90.6% on average for the first  15 minutes to 87.3% on average for the last 15  minutes of
11      exposure.  The uptake fraction was significantly greater in males (91.4%) than females (87.1%),
12      which is consistent with the larger fB and smaller VT of the females than males. Uptake was
13      not correlated with spirometric responses.  However, there was tendency for males to have
14      greater O3-induced FEVj decrements than females, 15.9% versus 9.4%, respectively. There was
15      considerable intersubject variability in FEVj decrements which ranged from -4 to 56% with
16      20 subjects having decrements of >15% and 4 subjects with >40% decrements (see Section
17      AX6.4 for additional  discussion regarding intersubject variability).
18          Few studies have measured the effect of ozone on ventilation distribution within the lung.
19      Foster et al. (1993) measured the effect of ozone on the vertical distribution of inspired air in the
20      lung using planar gamma scintigraphy. Nine healthy nonsmoking males (26.7 ± 7 years old)
21      were randomly exposed to FA or 0.33 ppm O3 for 2 h with IE.  After each exposure session,
22      subjects inhaled a 2- to 4-ml bolus of xenon-133 while seated in from of a gamma camera.
23      Images were acquired at the end of the first inspiration and 5-6 breaths later after the xenon had
24      equilibrated between lung regions. Using these images, the distribution of ventilation and
25      volume between upper-, middle-,  and lower-lung regions was quantified. Post-O3 relative to
26      post-FA, there were significant reductions in FVC (FA, 5.23 ±  0.5; O3, 4.88 ± 0.5 liters) and
27      midmaximal expiratory flow (FA, 3.82 ± 0.8; O3, 3.14 ± 0.9 liter/sec). Neither FRC nor the
28      distribution of volume (upper, 26.5%; middle, 42.5%; lower, 31%) between lung compartments
29      were affected by O3 exposure.  After the FA exposure, the  distribution of ventilation per unit
30      volume increased with progression from the apex to the base of the lung, i.e.,  the lower lung
31      regions received the greatest ventilation. Following O3 exposure, there was a significant

        August 2005                             AX6-9       DRAFT-DO NOT QUOTE OR CITE

-------
 1     reduction in the ventilation to the lower-lung and significant increases in ventilation to the
 2     upper- and middle-lung regions relative to the FA values in 7 of the 9 subjects. The post-O3
 3     increase in middle-lung ventilation was correlated with the decrease in midmaximal expiratory
 4     flow (r = 0.76, p < 0.05).
 5          Foster et al. (1997) measured the effect of ozone on ventilation distribution using a
 6     multiple breath nitrogen washout. Fifteen healthy nonsmoking males (25.4 ± 2 years old) were
 7     randomly exposed to FA or 0.35 ppm O3 for 2.2 h with IE.  Subjects alternated between 30 min
 8     periods of rest and treadmill exercise (VE « IQxFVC ~ 50 L/min).  The final exercise period
 9     was followed by 10 min rest period. Multiple breath nitrogen washout and spirometry were
10     measured pre- and immediately postexposure. At 24-h post-O3 exposure, 12 of 15 subjects
11     returned and completed an addition multibreath nitrogen washout maneuver. Pre- to post-O3
12     exposure, the mean FVC and FEVj were significantly decreased by 12 and 14%, respectively.
13     Exposure to FA did not appreciably affect spirometry or the multibreath nitrogen washout.
14     Following O3 exposure, the washout of nitrogen was delayed and resembled a two-compartment
15     washout, whereas pre-O3 exposure the log-linear clearance of nitrogen as a function of expired
16     volume resembled a single-compartment washout.  The clearance rate of the slow compartment
17     was approximated as the slope (Ln[N2] per expired volume) of the nitrogen washout between
18     20% and 9% nitrogen. Post-O3, there was a pronounced slow phase evident in nitrogen washout
19     which, on average, represented a 24% decrease in the washout rate relative to pre-O3.  Data for a
20     single  subject (see Figure 6-1) allowed for the size of the slow compartment to be determined.
21     For this subject, the slow compartment represented 23% of the lung. This  is fairly consistent
22     with Foster et al. (1993) where ventilation to the lower-lung (31% of volume) was reduced
23     post-O3. At 24-h post-O3, 6 of the 12 subjects who completed an additional nitrogen washout
24     maneuver had a delayed washout relative to the pre-O3 maneuver.  This suggests a prolonged O3
25     effect on the small airways and ventilation distribution in some individuals.
26
27     AX6.2.3  Prolonged Ozone Exposures
28          Between 1988 and 1994, a number studies were completed that described the responses of
29     subjects exposed to relatively low (0.08 to 0.16 ppm) O3 concentrations for exposure durations
30     of 4 to 8 h. These studies were discussed in the 1996 criteria document (U.S. Environmental
31     Protection Agency, 1996) and only a select few are briefly discussed here.  Table AX6-2 details

       August 2005                           AX6-10      DRAFT-DO NOT QUOTE OR CITE

-------
                           Table AX6-2. Pulmonary Function Effects after Prolonged Exposures to Ozone"
to
o
o
X
H

6
o


o
H

O

o
H
W

O


O
HH
H
W
Ozone Concentration11 Exposure

ppm
Studies with
0.18


0.0
0.20



0.2



0.0
0.24




Studies with
0.0
0.04
0.08
0.12

0.12








ug/m3 and Activity
4 hr Exposures
353 4hIE
(4 x 50 min)
VE = 35L/min
0 4hIE
392 (4 x 50 min cycle
ergometry or
treadmill running
[VE = 40L/min])
392 4 h IE
(4 x 50 min)
VE = 25 L/min/m2
BSA
0 4hIE
470 (4x15 min)
VE = 20 L/min



>6 hr Exposures
0 6.6 h
78 IE (6 x 50min)
157 VE = 20 L/min/m2
235 BSA

235 3 day-6.6h/day
IE (6 x 50 min)
VE = 17 L/min/m2,
20 L/min/m2
BSA, and 23
L/min/m2 BSA


Number and
Exposure Gender of Subject
Conditions Subjects Characteristics

23 °C 2 M, 2 F Adults NS, 21
50% RH to 33 years old

20 °C FA: 1 1 M, 3 Adult NS, 19 to
50% RH F 41 years old
03: 9M, 3F


20 °C 42 M, 24 F Adults NS,
50% RH 18 to 50 years
old

24 °C 10 M Healthy NS,
40% RH 60 to 69 years
9 M COPD
59 to 71 years



23 °C 1 5 M, 1 5 F Healthy NS,
50% RH 22.4 ± 2.4 yrs
old


23 °C 1 5 M, 1 5 F Healthy NS, 1 8
50% RH to 31 years old








Observed Effect(s)

FVC decreased 19% and FEVj decreased 29% in these four
pre-screened sensitive subjects.

Decrease in FVC, FEVj, VT, and SRaw and increase in fB with
O3 exposure compared with FA; total cell count and LDH
increased in isolated left main bronchus lavage and inflammatory
cell influx occurred with O3 exposure compared to FA exposure.

FEVj decreased by 18.6%; Pre-exposure methacholine
responsiveness was weakly correlated with the functional response
to O3 exposure. Symptoms were also weakly correlated with the
FEVj response (r = -0.31 to -0.37)
Healthy: small, 3.3%, decline in FEVj (p = 0.03 [not reported in
paper], paired-t on O3 versus FA pre-post FEVj). COPD: 8%
decline in FEV; (p = ns, O3 versus FA). Adjusted for exercise,
ozone effects did not differ significantly between COPD patients
and healthy subjects.
See Section AX6. 5.1.

FEVj and total symptoms at 6.6 h exposure to 0.04 ppm not
significantly different from FA. FEVj (-6.4%) and total symptoms
significant at 6.6 h exposure to 0.08 ppm. FEV; (-15.4%) at 6.6 h
not significantly different between chamber and face mask
exposure to 0.12 ppm.
FEVj at 6.6 h decreased significantly by 9.3%, 11.7%, and 13.9%,
respectively at three different exercise VE rates, but were not
significantly different from each other. Total symptoms at the
highest VE protocol were significantly greater than for the lowest
VE protocol beginning at 4.6 h. Largest subjects (2.2 m2 BSA)
had significantly greater average FEVj decrement for the three
protocols, 18.5% compared to the smallest subjects (1.4 m2
BSA), 6.5%.


Reference

Adams
(2000a)

Aris et al.
(1993)



Aris et al.
(1995)


Gong et al.
(1997a)





Adams (2002)




Adams
(2000b)







-------
                               Table AX6-2 (cont'd). Pulmonary Function Effects after Prolonged Exposures to Ozonea
to
o
o
X
ON
to
H
6
o
o
H
O
O
H
W
O
O
HH
H
W
Ozone Concentration11
ppm
(a) 0.08
(b) 0.08
(mean) varied
from 0.03 to
0.15

(a) 0.08



(b) 0.30



(a) 0.12

(b)0.12
(mean) varied
from 0.07 to
0.16
ug/m3
235
235
(mean)



157



588



235

235
(mean)


Exposure
Duration
and Activity
6.6 h
IE (6 x 50 min)
VE = 20 L/min/m2
BSA


6.6 h
IE (6 x 50 min)
VE = 20 L/min/m2
BSA
2h
IE (4 x 15 min)
VE = 35 L/min/m2
BSA
6.6 h IE
(6 x 50 min)
(a,b,c) VE = 20
L/min/m2 BSA
(d)VE=12
L/min/m2 BSA
Number and
Exposure Gender of Subject
Conditions Subjects Characteristics
23 °C 15 M Healthy NS,
50% RH 15 F 18 to 25 years
old



23 °C 15 M Healthy NS,
50% RH 15 F 18 to 25 years
old





23 °C 6 M, 6 F Healthy NS,
50% RH 19 to 25 years
old



Observed Effect(s)
(a) FEVj decreased 6.2% after 6.6 h in square-wave exposures.
Total symptoms significantly increased at 5.6 and 6.6 h.
(b) FEV; decreased 5.6 to 6.2% after 4.6 to 6.6 h, respectively,
in varied exposure; total symptoms significantly increased also
after 4.6 to 6.6 h. No significant difference between face mask
and chamber exposures.
Significantly greater FEV; decrement (12.4%) for 2-h, 0.30 ppm
exposure than for 6.6-h, 0.08 ppm exposure (3.6%).






(a) FEVj decreased 11% at 6.6 h in square-wave exposure.
Total symptoms significant from 4.6 to 6.6 h.

(b) FEV; decreased 13% at 6.6 h; not significantly different from
square-wave exposure. Total symptoms significant from 4.6 to
6.6 h.
Reference
Adams
(2003a)






Adams
(2003b)




Adams and
Ollison(1997)




         (c)0.12          235
         (mean) varied   (mean)
         from 0.11 to
         0.13
         (d)0.12
235
(c) FEVj decreased 10.3% at 6.6 h; not significantly different from
square-wave exposure. Total symptoms significant from 4.6 to
6.6 h.

(d) FEV; decreased 3.6% at 6.6 h; significantly less than for
20 L/min/m2 BSA protocols.
         "See Appendix A for abbreviations and acronyms.
         bListed from lowest to highest O3 concentration.

-------
 1      newer studies of healthy subjects undergoing prolonged exposures at O3 concentrations ranging
 2      from 0.06 to 0.20 ppm. In most of these studies, statistically significant changes in pulmonary
 3      function, symptoms, and airway responsiveness have been observed during and after exposures
 4      to O3 concentrations of 0.08 ppm and higher. As with studies conducted at higher O3
 5      concentrations for shorter periods of time, there is considerable intersubject variability in
 6      response (see Section AX6.4).
 1          Folinsbee et al. (1988) first reported the effects of a 6.6 h exposure to 0.12 ppm O3 in ten
 8      young healthy adults (25 ± 4 yr) with quasi continuous exercise that was intended to simulate a
 9      full workday of heavy physical labor. Except for a 35-min lunch break after 3 h, the subjects
10      exercised at a moderate level ( VE « 40 L/min) for 50 min of each hour.  Ignoring the lunch
11      break during which lung function did not change appreciably, approximately linear decreases
12      were observed in FVC, FEVl3 and FEV25_75 with duration of O3 exposure. Correcting for FA
13      responses, decrements of 8.2, 14.9, and 26.8% in FVC, FEVl3 and FEV25.75 occurred as a result
14      of the O3 exposure.  Using the same 6.6 h protocol, but a lower O3 concentration of 0.08 ppm,
15      Horstman et al. (1990) and McDonnell et al. (1991) observed decrements corrected for FA (and
16      averaged across studies) of 5, 8, and 11% in FVC, FEVl3 and FEV25_75, respectively, in 60 young
17      adults (25 ± 5 years old).  Horvath et al. (1991) observed a 4% (p = 0.03)1 decrement in FEVj
18      using the forementioned protocol (i.e., 6.6 h and 0.08 ppm O3) in 11 healthy adults (37 ± 4 yr).
19      The smaller decrement observed by Horvath et al. (1991) versus Horstman et al. (1990) and
20      McDonnell  et al.  (1991) is consistent with response decreasing as subject age increases (see
21      Section AX6.5.1).
22
23      AX6.2.3.1  Effect of Exercise Ventilation Rate on FEVt Response to 6.6 h Ozone Exposure
24          It is well known that response to O3 exposure is a function of VE in studies of 2 h or less in
25      duration (See Section AX6.2.2).  It is reasonable to expect that response to a prolonged 6.6-h O3
26      exposure is  also function of VE, although quantitative analyses are lacking.
27          In an attempt to quantify this effect, Adams and Ollison (1997) exposed 12 young adults
28      to an average O3 concentration of 0.12 ppm for 6.6 h at varied exercise VE . They observed a
29      mean FEVj  decrements of 10 to 11% in two protocols having a mean exercise VE  of 33 L/min
              'Based on two-tailed paired t-test of data in Table III of Horvath et al. (1991).

        August 2005                            AX6-13      DRAFT-DO NOT QUOTE OR CITE

-------
 1      and a 14% decrement in a protocol with a mean exercise VE  of 36 L/min.  These FEVj
 2      decrements were significantly greater than the average decrement of 3.6% (not significantly
 3      different from FA response) observed at an exercise VE of only 20 L/min.  In a subsequent study
 4      of 30 healthy adults (Adams, 2000b), the effect of smaller exercise VE differences on pulmonary
 5      function and symptoms responses to 6.6 h exposure to 0.12 ppm O3 was examined. FEVj
 6      decrements of 9.3, 11.7, and 13.9% were observed for the exercise VE of 30.2, 35.5, and
 7      40.8 L/min, respectively.  Along with the tendency for FEVj responses to increase with VE, total
 8      symptoms severity was found to be significantly greater at the end of the highest VE protocol
 9      relative to the lowest VE protocol. Although the FEVj responses were not significantly different
10      from each other, the power of the study to detect differences between the three VE  levels was not
11      reported and no analysis was performed using all of the data (e.g., a mixed effects  model). Data
12      from the Adams and Ollison (1997) and Adams (2000b) studies are illustrated in Figure AX6-2
13      with data from three older studies. There is a paucity of data below an exercise  VE of 30 L/min.
14      Existing data for exposure to 0.12 ppm O3 suggest that FEVj responses increase with increasing
15      exercise VE until at least 35 L/min.
16
17      AX6.2.3.2 Exercise Ventilation Rate as  a Function of Body/Lung Size on FEVt Response
18                to 6.6 h Ozone Exposure
19           Typically, with the assumption that the total inhaled O3 dose should be proportional to the
20      lung size of each individual, exercise VE in 6.6 h exposures has been set as a multiple of body
21      surface area (BSA) (McDonnell et al., 1991) or as a product of eight times FVC (Folinsbee et al.,
22      1988; Frank et al., 2001; Horstman et al., 1990). Utilizing previously published data, McDonnell
23      et al. (1997) developed a statistical model analyzing the effects of O3 concentration, VE, duration
24      of exposure, age, and body and lung size on FEVj response.  They concluded that any effect of
25      BSA, height,  or baseline FVC on percent decrement in FEVj in this population of 485 young
26      adults was small if it exists at all. This is consistent with Messineo and Adams (1990), who
27      examined pulmonary function responses in young adult women having small (n =  14) or large
28      (n = 14) lung sizes (mean FVC of 3.74 and 5.11 L, respectively).  Subjects were exposed to
29      0.30 ppm O3 for 1 h with CE (VE = 47 L/min). There was no significant difference between the


        August 2005                             AX6-14      DRAFT-DO NOT QUOTE OR CITE

-------
                         25
~  20-

I   1=
                         10H
                     >
                     LU
                          5-
                              " Adams (2000b)
                              D Adams and Ollison (1997)
                              A Folinsbeeetal. (1988)
                              A Folinsbeeetal. (1994)
                              • Horstmanetal. (1990)
                           15      20       25      30      35
                                         Minute Ventilation (L/min)
                                               40
                    45
      Figure AX6-2.  Average FEVt decrements (±SE) for prolonged 6.6 h exposures to 0.12 ppm
                                              •
                     O3 as a function of exercise VE.  Since age affects response to O3 exposure,
                     selected studies had subjects with mean ages between 22 and 25 years.
                     FEVt decrements were calculated as mean O3 responses minus mean air
                     responses. The SE bars illustrate variability in FEVt responses (pre minus
                     post) on the O3 exposure day in all cases except for Folinsbee et al. (1994),
                     where post O3 exposure variability is illustrated. In one case, the SE for VE
                     of 33 L/Min (10.3% decrement) was taken as the SE of data from protocol
                     with VE of 33 L/min (11% decrement). All studies used a constant
                     0.12 ppm O3 exposure except two (*) which used 0.115 ppm O3 for hours
                     1-2 and 5-6 and 0.13 ppm O3 for hours 3-4 of exposure.
1     group FEVj decrements (22.1 and 25.6% for small and large lung, respectively). In addition,
2     Messineo and Adams (1990) also did a retrospective analysis of 36 young adult males who each

3     had completed similar 1 h exposures to 0.30 ppm O3 with CE (VE  « 70 L/min) and found lung

4     size was not related to FEVj response.
5          Adams (2000b) studied a group of 30 young adult men and women exposed to
6     0.12 ppm O3 for 6.6 h on three occasions while exercising 50 min of each hour at one of three

7     different VE levels (viz., 17, 20, and 23 l/min/m2BSA).  Their postexposure FEVj responses

8     were regressed as a function of BSA (which was directly related to the absolute amount of VE
      August 2005
                       AX6-15
DRAFT-DO NOT QUOTE OR CITE

-------
 1      during exercise and, thus, primarily responsible for individual differences in total inhaled O3
 2      dose). The slope was significantly different from zero (p = 0.01), meaning that the smallest
 3      subjects, who had the lowest exercise VE (« 26 L/min), had a lower FEVj decrement (-5%)
 4      than the largest subjects (-17%), whose exercise VE was -44 L/min. This relationship was not
 5      a gender-based difference, as the mean female's FEVj decrement was -11.2%, which was not
 6      significantly different from the male's -12.2% mean value.  Similarly, when total symptoms
 7      severity response was regressed against BSA, the slope was significantly different than zero
 8      (p = 0.0001), with lower values for smaller subjects than for larger subjects. Results of this
 9      study suggest that for the O3 concentration and exposure duration used, responses are more
10      closely related to VE thanVE normalized to BS A. Further, this observation is in agreement with
11      McDonnell et al. (1997), who observed no evidence that measurements of lung or body size
12      were significantly related to FEVj response in 2 h IE exposures.  These authors state that the
13      absence of an observed relationship between FEVj response and BSA, height, or FVC may be
14      due to the poor correlation between these variables and airway caliber (Collins et al., 1986;
15      Martin et al., 1987). Also, the O3 dosimetry study of Bush et al.  (1996) indicated that
16      normalization of the O3 dose would be more appropriately applied as a function of anatomic
17      dead space.
18
19      AX6.2.3.3 Comparison of 6.6 h Ozone Exposure Pulmonary Responses to Those Observed
20                in 2 h Intermittent Exercise Ozone Exposures
21           It has been shown that greater O3 concentration (Horstman et al., 1990) and higher VE
22      (Adams, 2000b) each elicit greater FEVj response in prolonged,  6.6-h exposures, but data on the
23      relative effect of O3 concentration, VE, and T in prolonged exposures are very limited and have
24      not been systematically compared to data from shorter (<2-h) exposures. In a recent study
25      (Adams, 2003b), the group mean FEVj response for a 2-h IE exposure to 0.30 ppm O3 was
26      -12.4%, while that for a  6.6-h exposure to 0.08 ppm O3 was -3.5%.  The total inhaled O3 dose
27      (as the simple product of C x T x VE ) was 1358 ppm-L for the 2-h exposure and 946 ppm-L for
28      the 6.6-h exposure. Thus, the FEVj decrement was 3.5 times greater and the total inhaled O3
29      dose was 1.44 times greater for the 2-h exposure compared to the 6.6-h exposure. This
30      difference illustrates the limitations of utilizing the concept of total O3 dose for comparisons
31      between studies of vastly different exposure durations.

        August 2005                            AX6-16     DRAFT-DO NOT QUOTE OR CITE

-------
 1           Adams (2003b) also examined whether prolonged 6.6 h exposure to a relatively low O3
 2      concentration (0.08 ppm) and the 2-h IE exposure at a relatively high O3 concentration (0.30
 3      ppm) elicited consistent individual subject effects, i.e, were those most or least affected in one
 4      exposure also similarly affected in the other?  Individual subject O3 exposure reproducibility was
 5      first examined via a regression plot of the postexposure FEVj response to the 6.6-h chamber
 6      exposure as a function of postexposure FEVj response to the 2-h chamber exposure. The R2 of
 7      0.40, although statistically significant, was substantially less than that observed in a comparison
 8      of individual FEVj response to two 2-h IE exposures by chamber and face mask, respectively
 9      (R2 = 0.83). The Spearman rank order correlation for the chamber 6.6-h and chamber 2-h
10      exposure comparison was also substantially less (0.49) than that obtained for the two 2-h
11      exposures (0.85).  The primary reason for the greater variability in the chamber 6.6-h exposure
12      FEVj response as a function of that observed for the two 2-h IE exposures is very likely related
13      to the increased variability in response upon repeated exposure to O3 concentrations lower than
14      0.18 ppm (R = 0.57, compared to a mean R of 0.82 at higher concentrations) reported by
15      McDonnell et al. (1985a). This rationale is supported by the lower R (0.60) observed by
16      Adams (2003b) for the FEVj responses found in 6.6 h chamber and face mask exposures to
17      0.08 ppm O3, compared to an R of 0.91 observed for responses found for the 2 h chamber and
18      face mask exposures to 0.30 ppm O3.
19
20      AX6.2.4  Triangular Ozone Exposures
21           To further explore the factors that determine responsiveness to O3, Hazucha et al. (1992)
22      designed a protocol to examine the effect of varying, rather than constant, O3 concentrations.
23      In this study, subjects were exposed to a constant level of 0.12 ppm O3 for 8 h and to an O3 level
24      that increased linearly from 0 to 0.24 ppm for the first 4 h and then decreased linearly from
25      0.24 to 0 over the second 4 h of the 8 h exposure (triangular concentration profile). Subjects
26      performed  moderate exercise (VE «40 L/min) during the first 30  minutes of each hour.  The total
27      inhaled O3  dose (i.e., C x T x VE ) for the constant versus the triangular concentration profile was
28      almost identical. FEVj responses are illustrated in Figure AX6-3. With exposure to the constant
29      0.12 ppm O3, FEVj declined approximately 5% by the  fifth hour of exposure  and then remained
30      at that level. This observation clearly indicates a response plateau as suggested in other
        August 2005                            AX6-17      DRAFT-DO NOT QUOTE OR CITE

-------
                              4.5

                              4.4
                              4.3
HI
U-
c
(0
                              4.2
                              4.1
                              4.0
                              3.9
                0 ppm
                0.12 ppm
                Constant
                Variable
                                             246
                                          Exposure Duration (h)
       Figure AX6-3.  The forced expiratory volume in 1 s (FEVt) is shown in relation to exposure
                      duration (hours) under three exposure conditions.  Subjects exercised
                      (minute ventilation « 40 L/min) for 30 min during each hour; FEVt was
                      measured at the end of the intervening rest period. Standard error of the
                      mean for these FEVt averages (not shown) ranged from 120 to  150 mL.
       Source: Hazuchaetal. (1992).
 1     prolonged exposure studies (Horstman et al., 1990; McDonnell et al., 1991).  However, with the
 2     triangular O3 concentration profile after a minimal initial response over the first 3 h, Hazucha
 3     et al. (1992) observed a substantial decrease in FEVj corresponding to the higher average O3
 4     concentration that reached a nadir after 6 h (-10.3%). Despite 2 h of continued exposure to a
 5     lower O3 concentration (0.12 to 0.00 ppm, mean = 0.06 ppm), FEVj improved and was only
 6     reduced by 6.3% (relative to the preexposure FEVj) at the end of the 8-h exposure. The authors
 7     concluded that total inhaled O3 dose (C * VE * T) was not a sufficient index of O3 exposure and
 8     that, as observed by others (Adams et al., 1981; Folinsbee et al., 1978; Hazucha, 1987; Larsen
 9     et al.,  1991), O3 concentration appears to be more important in determining exposure effects than
10     is either duration or the volume of air breathed during the exposure. However, it should be noted
       August 2005
                    AX6-18
DRAFT-DO NOT QUOTE OR CITE

-------
 1     that the mean O3 concentration for Hazucha et al.'s triangular exposure profile was 0.12 ppm at
 2     4 h, 0.138 ppm at 5 h, 0.14 ppm at 6 h, and 0.133 ppm at 7 h, before falling to 0.12 ppm at 8 h.
 3     The FEVj responses of the last 4 hours (Figure AX6-3) follow a closely similar pattern as the
 4     total mean O3 concentration over the same time period.
 5          It has become apparent that laboratory simulations of air-pollution risk-assessment need to
 6     employ O3 concentration profiles that more accurately mimic those encountered during summer
 7     daylight ambient air pollution episodes (Adams and Ollison, 1997; Lefohn and Foley, 1993;
 8     Rombout et al., 1986). Neither square-wave O3 exposures or the one 8-h study by Hazucha et al.
 9     (1992) that utilized a triangular shaped varied O3 exposure described above closely resembles
10     the variable diurnal daylight O3 concentration pattern observed in many urban areas experiencing
11     air-pollution episodes (Lefohn and Foley, 1993). Recently, 6.6 h less abrupt triangular O3
12     exposure profiles at lower concentrations more typical of outdoor ambient conditions have been
13     examined (Adams 2003a; Adams and Ollison, 1997).
14          Using a face-mask inhalation system, Adams and Ollison (1997) observed no significant
15     differences in postexposure pulmonary function responses or symptoms between the 6.6-h,
16     0.12 ppm O3 square-wave exposure; and those observed for a triangular O3 profile in which
17     concentration was increased steadily from 0.068 ppm to 0.159 ppm at 3.5 h and then decreased
18     steadily to 0.097 ppm at end exposure. Further, no attenuation in FEVj response during the last
19     2 h was observed in either the 6.6 h square-wave or the triangular exposures.  In a subsequent
20     study (Adams, 2003a), no significant difference was observed in pulmonary function responses
21     or symptoms between face-mask and chamber exposure systems either for a 6.6-h, 0.08 ppm  O3
22     square-wave profile or for the triangular O3 exposure beginning at 0.03 ppm,  increasing steadily
23     to 0.15 ppm in the fourth hour, and decreasing steadily to 0.05 ppm at 6.6 h (mean = 0.08 ppm).
24          For the chamber-exposure comparison, postexposure values for FEVj and symptoms were
25     not significantly different from the responses for the square-wave 0.08 ppm O3 exposure.
26     However, analysis showed that FEVj response for the square-wave protocol did not become
27     statistically significant until the 6.6-h postexposure value, while that for the triangular exposure
28     protocol was significant at 4.6 h (when O3 concentration was 0.15 ppm). Earlier significant
29     FEVj responses for the triangular protocol were accompanied by significant increases in
30     symptoms at 4.6 h, which continued on through the fifth and sixth hours when the mean O3
31     concentration was 0.065 ppm.  Symptoms for the square-wave 0.08 ppm exposure did not

       August 2005                             AX6-19     DRAFT-DO NOT QUOTE OR CITE

-------
 1     become statistically significant until 5.6 h. The FEVj responses during the last two hours of the
 2     triangular exposure by Adams (2003a) did not decrease as dramatically as in the Hazucha et al.
 3     (1992) study (Figure AX6-3). The most probable reason for differences in the triangular O3
 4     profile observations of Hazucha et al. (1992) and those of Adams (2003a) is that the increase
 5     and decrease in Hazucha et al.'s study (i.e., 0 to 0.24 ppm and back to 0) encompassed a much
 6     greater range of O3 concentrations than those used by Adams (2003a), viz., 0.03 ppm to
 7     0.15 ppm from 0 to 3.6 h, then decreasing to 0.05 ppm for the final hour of exposure.
 8     Nonetheless, the greatest FEVj decrement was observed at 6 h of Hazucha et al.'s 8 h triangular
 9     exposure (Figure AX6-3) corresponding to the time when total mean O3 concentration was
10     highest (0.14 ppm), with a very similar response at 7 h when total mean O3 concentration was
11     0.138 ppm.
12          Whereas FEVj decrements during square-wave O3 exposures between 0.08 to 0.12 ppm
13     tend to increase with time of exposure (i.e., with steadily increasing total inhaled dose), FEVj
14     decrements during triangular exposures (Hazucha et al., 1992; Adams, 2003a) occurred 1 to 2 h
15     after the  peak O3 concentration and 1 h to 2 h before the maximal total O3 inhaled dose occurred
16     at the end of exposure.  This  difference, especially because O3 concentration profiles during
17     summer daylight air-pollution episodes rarely mimic a square-wave, implies that triangular O3
18     exposure profiles most frequently observed during summer daylight hours merit further
19     investigation.  These two studies suggest that depending upon the profile of the exposure,
20     the triangular exposure can potentially lead to higher FEVj responses than the square wave
21     exposures at the overall equivalent ozone dose.
22
23     AX6.2.5  Mechanisms of Pulmonary Function Responses
24          Inhalation of O3 for several hours while physically active elicits both subjective respiratory
25     tract symptoms and acute pathophysiologic changes.  The typical symptomatic response
26     consistently reported in studies is that of tracheobronchial airway irritation. This is accompanied
27     by decrements in lung capacities and volumes, bronchoconstriction, airway hyperresponsiveness,
28     airway inflammation, immune system activation, and  epithelial injury.  The severity of
29     symptoms and the magnitude of response depend on inhaled dose, O3 sensitivity of an individual
30     and the extent of tolerance resulting from previous exposures. The development of effects is
        August 2005                            AX6-20      DRAFT-DO NOT QUOTE OR CITE

-------
 1     time dependent during both exposure and recovery periods with considerable overlap of evolving
 2     and receding effects.
 3           Exposure to O3 initiates reflex responses manifested as a decline in spirometric lung
 4     function parameters (1FVC, IFEVl3  1FEF25.75), bronchoconstriction (TSRaw) and altered
 5     breathing pattern (1VT, t fB), which becomes more pronounced as exposure progresses and
 6     symptoms of throat irritation, cough, substernal soreness and pain on deep inspiration develop.
 7     The spirometric lung function decline and the severity of symptoms during a variable (ramp)
 8     exposure profile seem to peak a short time (about 1 to 2 h) following the highest concentration
 9     of O3 (Hazucha et al., 1992; Adams, 2003a). Exposure to a uniform O3 concentration profile
10     elicits the maximum spirometric response at the end of exposure (Hazucha et al., 1992; Adams,
11     2003a). Regardless of exposure concentration profile, as the exposure to O3 progresses, airway
12     inflammation begins to develop and the immune response at both cellular and subcellular level is
13     activated. Airway hyperreactivity develops slower than pulmonary function effects, while
14     neutrophilic inflammation of the airways develops even more slowly and reaches the maximum
15     3 to 6 h postexposure. The  cellular responses (e.g., release of immunoregulatory cytokines)
16     appear to  still be active as late as 20 h postexposure (Torres et al., 2000). Following cessation of
17     exposure, the recovery in terms of breathing pattern, pulmonary function and airway
18     hyperreactivity progresses rapidly and is almost complete within 4 to 6  hours in moderately
19     responsive individuals. Persisting small residual lung function effects are almost completely
20     resolved within 24 hours. Following a 2 h exposure to 0.4 ppm O3 with IE, Nightingale et al.
21     (2000) observed a  13.5% decrement in FEVj. By 3 h postexposure, however, only a 2.7% FEVj
22     decrement persisted.  As illustrated in Figure AX6-4,  a similar postexposure recovery in FVC
23     was observed. In hyperresponsive individuals, the recovery takes longer and as much  as
24     48 hours to return to baseline values.  More slowly developing inflammatory  and cellular
25     changes persist for up to 48 hours. The time sequence, magnitude and the type of responses of
26     this complex series of events, both in terms of development and recovery, indicate that several
27     mechanisms, activated at different times of exposure,  must contribute to the overall lung
28     function response (U.S. Environmental Protection Agency, 1996).
29
        August 2005                            AX6-21      DRAFT-DO NOT QUOTE OR CITE

-------
                          a.
                                4.0-,
                                3.5-
                            LJJ
                          b.
                            o
                                2.5-
                                2.0-
                                    Pre        01234
                                        Time from Exposure (hours)
                                4.5-|
                                4.0-
                                3.5-
                                3.0-
                       24
                                    Pre        0   1    2    3    4    24
                                        Time from Exposure (hours)
      Figure AX6-4a,b.  Recovery of spirometric responses following a 2 h exposure to 0.4 ppm
                         O3 with IE. Immediately postexposure, there were significant
                         decrements (**p < 0.001, ***p < 0.0005) in FVC (10%) and FEVt
                         (13.5%) compared to preexposure values. At 3 h postexposure, FVC
                         and FEVj were at 96 and 97% of preexposure values, respectively.

      Adapted from Nightingale et al. (2000).
1     AX6.2.5.1  Pathophysiologic Mechanisms
2     Breathing pattern changes
3          Human studies consistently report that inhalation of O3 alters the breathing pattern without
4     significantly affecting minute ventilation. A progressive decrease in tidal volume and a
5     "compensatory" increase in frequency of breathing to maintain steady minute ventilation during
      August 2005
AX6-22
DRAFT-DO NOT QUOTE OR CITE

-------
 1      exposure suggests a direct modulation of ventilatory control.  These changes parallel a response
 2      of many animal species exposed to O3 and other lower airway irritants (Tepper et al., 1990).
 3      Although alteration of a breathing pattern could be to some degree voluntary, the presence of the
 4      response in animals and the absence of perception of the pattern change by subjects, even before
 5      appearance of the first subjective symptoms of irritation, suggests an involuntary reflex
 6      mechanism.
 7           Direct recording from single afferent vagal fibers in animals convincingly demonstrated
 8      that bronchial C-fibers and rapidly adapting receptors are the primary vagal afferents responsible
 9      for O3-induced changes in ventilatory rate and depth (Coleridge et al.,  1993; Hazucha and
10      Sant'Ambrogio, 1993).  In spontaneously breathing dogs, an increase in VT/T; (T; decreased more
11      than VT) was attributed  to an increased inspiratory drive due to stimulation of rapidly adapting
12      receptors and bronchial C-fibers by O3 (Schelegle et al., 1993).  Folinsbee and Hazucha (2000)
13      also observed similar changes in VT/T; and other breath-timing parameters in humans exposed
14      to O3 implying activation of the same mechanisms.  They also reported that Pm0A (pressure at
15      mouth at 0.1 sec of inspiration against a transiently occluded mouthpiece which is considered an
16      index of inspiratory drive) increased during controlled hypercapnia without a change in the slope
17      of Pm0 j versus pCO2 relation suggesting that the primary mechanism is an increased inspiratory
18      drive. Since no significant within-individual differences in ventilatory response to CO2 between
19      air exposure and O3 exposure were found, the CO2 chemoreceptors did not modulate the
20      response. Therefore, the principal peripheral mechanism modulating changes in breathing
21      pattern appears to be direct and indirect stimulation of lung receptors and bronchial C-fibers
22      by O3 and/or its oxidative products. The activity of these afferents, centrally integrated with
23      input from other sensory pathways, drives the ventilatory controller, which determines the depth
24      and the frequency of breathing.
25           The potential modulation of breathing pattern by activation of sensory afferents located in
26      extrathoracic airways by O3 has not yet been studied in humans.  Laboratory animal studies have
27      shown that the larynx, pharynx, and nasal mucosa are densely populated by free-ending,
28      unmyelinated sensory afferents resembling nociceptive C-fibers (Spit et al., 1993; Sekizawa and
29      Tsubone, 1994).  They are almost certainly stimulated by O3 and likely contribute to overall
30      ventilatory and symptomatic responses. Nasal only exposure of rats produced O3-induced
31      changes in breathing pattern that are similar to changes found in humans (Kleinman et al., 1999).

        August 2005                             AX6-23      DRAFT-DO NOT QUOTE OR CITE

-------
 1      Symptoms and lung function changes
 2           As already discussed, in addition to changes in ventilatory control, O3 inhalation by
 3      humans will also induce a variety of symptoms, reduce vital capacity (VC) and related functional
 4      measures, and increase airway resistance. Hazucha et al. (1989) postulated that a reduction of
 5      VC by O3 is due to a reflex inhibition of inspiration and not due to a voluntary reduction of
 6      inspiratory effort. Recently, Schelegle et al. (2001) convincingly demonstrated that a reduction
 7      of VC due to O3 is indeed reflex in origin and not a result of subjective discomfort and
 8      consequent premature voluntary termination of inspiration. They reported that inhalation of an
 9      aerosolized topical anesthetic tetracaine substantially reduced if not abolished O3-induced
10      symptoms that are known to be mediated in part by bronchial C-fibers. Yet, such local
11      anesthesia of the upper airway mucosa had a minor and irregular effect on pulmonary function
12      decrements and tachypnea, strongly supporting neural mediation, i.e., stimulation of both
13      bronchial and pulmonary C-fibers, and not voluntary inhibition of inspiration (due to pain) as the
14      key mechanism.
15           The involvement of nociceptive bronchial C-fibers modulated by opioid receptors in
16      limiting maximal inspiration and eliciting subjective symptoms in humans was studied by
17      Passannante et al. (1998). The authors hypothesized that highly variable responses among
18      individuals might reflect the individual's inability or unwillingness to take a full inspiration.
19      Moreover, development of symptoms of pain on deep inspiration, cough and substernal soreness
20      suggested that nociceptive mechanism(s) might be  involved in O3-induced inhibition of maximal
21      inspiration.  If this were so, pain  suppression or inhibition by opioid receptor agonists should
22      partially or fully reverse symptoms and lung functional impairment.  Subjects for this study were
23      pre-screened with exposure to 0.42 ppm O3 and classified either as "weak" (FEVj >95% of
24      preexposure value), "strong" (FEVj < 85% of preexposure value), or "moderate" responders.
25      Sixty two (28 M, 34 F) healthy volunteers (18 to 59 yrs old), known from the previous screening
26      to be "weak" (n = 20) or "strong" (n = 42) O3-responders, participated in this double-blind
27      crossover study. Subjects underwent either two 2 h exposures to air, or two 2 h  exposures to
28      0.42 ppm O3, with 15 min IE at 17.5 1/min/m2 BSA. Immediately following postexposure
29      spirometry the "weak" responders were given (in random order) either the potent opioid receptor
30      antagonist naloxone (0.15 mg/kg) or saline, while "strong" responders received (in random
31      order) either the potent, rapid-acting opioid agonist and analgesic sufentanil (0.2 |ig/kg), or

        August 2005                            AX6-24      DRAFT-DO NOT QUOTE OR CITE

-------
 1      physiologic saline administered through an indwelling catheter. Administration of saline or
 2      naloxone had no significant effect on the relatively small decrements in FEVj observed in
 3      "weak" responders. However, as hypothesized, sufentanil rapidly reversed both the O3-induced
 4      symptomatic effects and spirometric decrements (FEV^ p < 0.0001) in "strong" responders
 5      (Figure AX6-5). All the same, the reversal was not complete and the average post-sufentanil
 6      FEVj remained significantly below (-7.3%) the preexposure value suggesting involvement of
 7      non-opioid receptor modulated mechanisms as well. Uneven suppression of symptoms has
 8      implied involvement of both A-6 and bronchial C-fibers. The plasma p-endorphin (a potent
 9      pain suppressor) levels, though substantially elevated immediately postexposure and post-drug
10      administration, were not related to individuals' O3 responsiveness.  These observations have
11      demonstrated that nociceptive mechanisms play a key role in modulating O3-induced inhibition
12      of inspiration. Moreover, these findings are consistent with and further support the concept that
13      the primary mechanism of O3-induced reduction in inspiratory lung function, is an inhibition of
14      inspiration elicited by stimulation of the C-fibers.  The absence of effect of naloxone in "weak"
15      responders shows that the weak response is not due to excessive endorphin production in those
16      individuals. However, other neurogenic mechanisms not modulated by opioid receptors may
17      have some though limited role in inspiratory inhibition.
18
19      Airway hyperreactivity
20          In addition to limitation of maximal inspiration and its effects on other spirometric
21      endpoints, activation of airway sensory afferents also plays a role in receptor-mediated
22      bronchoconstriction and an increase in airway resistance. Despite this common mechanism,
23      post-O3 pulmonary function changes and either early or late bronchial hyperresponsiveness
24      (BHR) to inhaled aerosolized methacholine or histamine are poorly correlated either in time or
25      magnitude.  Fentanyl and indomethacin, the drugs that have been shown to attenuate O3-induced
26      lung function decrements in humans, did not prevent induction of BHR when administered to
27      guinea pigs prior to O3 exposure (Yeadon et al., 1992). Neither does post-O3 BHR seem to be
28      related to airway baseline reactivity. These findings imply that the mechanisms are either not
29      related or are activated independently in time. Animal studies (with limited support from human
30      studies) have suggested that an early post-O3 BHR is, at least in part, vagally mediated (Freed,
31      1996)  and that stimulation of C-fibers can lead to increased responsiveness of bronchial smooth

        August 2005                             AX6-25      DRAFT-DO NOT QUOTE OR CITE

-------
             LJJ
                                0.5
 1           1.5

Time (hours)
                     2.5
      Figure AX6-5.  Plot of the mean FE Vt (% baseline) vs. time for ozone exposed cohorts.
                      Solid lines represent data for "strong" males (n = 14; solid squares) and
                      females (n = 15; solid circles) that received sufentanil and dotted lines
                      represent data for the same cohorts after receiving saline.  Dashed lines
                      represent data for "weak" males (n = 5; open squares) and females (n = 10;
                      open circles) that received naloxone and dot-dash lines represent data for
                      the same cohorts after receiving saline. The arrow denotes the time of drug
                      administration (-2.1 hrs).  Vertical bars associated with the symbols are
                      one-sided SEM.

      Source: Adapted from Passannante et al. (1998).
1     muscle independently of systemic and inflammatory changes which may be even absent (load

2     et al., 1996). In vitro study of isolated human bronchi have reported that O3-induced airway
3     sensitization involves changes in smooth muscle excitation-contraction coupling (Marthan,

4     1996). Characteristic O3-induced inflammatory airway neutrophilia which at one time was

5     considered a leading BHR mechanism, has been found in a murine model to be only

6     coincidentally associated with BHR, i.e., there was no cause and effect relationship (Zhang et al.,
      August 2005
  AX6-26
DRAFT-DO NOT QUOTE OR CITE

-------
 1      1995). However, this observation does not rule out involvement of other cells such as
 2      eosinophils or T-helper cells in BHR modulation.  There is some evidence that release of
 3      inflammatory mediators by these cells can sustain BHR and bronchoconstriction.  In vitro and
 4      animal studies have also suggested that airway neutral endopeptidase activity can be a strong
 5      modulator of BHR (Marthan et al., 1996; Yeadon et al., 1992). Late BHR observed in some
 6      studies is plausibly due to a sustained damage of airway epithelium and continual release of
 7      inflammatory mediators (Foster et al., 2000).  Thus, O3-induced BHR appears to be a product of
 8      many mechanisms acting at different time periods and levels of the bronchial smooth muscle
 9      signaling pathways.  [The effects ofO3 on BHR are described in Section AX6.8. ]
10
11      AX6.2.5.2 Mechanisms at a Cellular and Molecular Level
12           Stimulation of vagal afferents by O3 and reactive products, the primary mechanism of lung
13      function impairment is enhanced and sustained by what can be considered in this context to be
14      secondary mechanisms activated at a cellular and molecular level. The complexity of these
15      mechanisms is beyond the scope of this section and the reader is directed to Section AX6.9 of
16      this chapter for greater details.  A comprehensive review by Mudway and Kelly (2000) discusses
17      the cellular and molecular mechanisms of O3-induced pulmonary response in great detail.
18
19      Neurogenic airway inflammation
20           Stimulation of bronchial C-fibers by O3 not only inhibits maximal inspiration but, through
21      local axon reflexes, induces neurogenic inflammation. This pathophysiologic process is
22      characterized by release of tachykinins and other proinflammatory neuropeptides. Ozone
23      exposure has been shown to elevate C-fiber-associated tachykinin substance P in human
24      bronchial lavage fluid (Hazbun et al. 1993) and to deplete neuropeptides synthesized and
25      released from C-fibers in human airway  epithelium rich in substance P-immunoreactive axons.
26      Substance P and other transmitters are known to induce granulocyte adhesion and subsequent
27      transposition into the airways, increase vascular permeability and plasma protein extravasation,
28      cause bronchoconstriction, and promote mucus secretion (Solway and Leff,  1991). Although the
29      initial pathways of neurogenic, antigen-induced, and generally immune-mediated inflammation
30      are not the same, they eventually converge leading to further amplification of airway
31      inflammatory processes by subsequent release of cytokines, eicosanoids, and other mediators.

        August 2005                            AX6-27      DRAFT-DO NOT QUOTE OR CITE

-------
 1      Significantly negative correlations between O3-induced leukotriene (LTC4/D4/E4) production and
 2      spirometric decrements (Hazucha et al., 1996), and an increased level of postexposure PGE2, a
 3      mediator known to stimulate bronchial C-fibers, show that these mediators play an important
 4      role in attenuation of lung function due to O3 exposure (Mohammed et al., 1993; Hazucha et al.,
 5      1996). Moreover, because the density of bronchial C-fibers is much lower in the small than
 6      large airways, the reported post-O3 dysfunction of small airways assessed by decrement
 7      in FEF25.75 (Weinman et al., 1995; Frank et al., 2001) may be due in part to inflammation.
 8      Also, because of the relative slowness of inflammatory responses as compared to reflex
 9      effects, O3-triggered inflammatory mechanisms are unlikely to initially contribute to progressive
10      lung function reduction.  It is plausible, however, that when fully activated, they sustain and
11      possibly further aggravate already impaired lung function.  Indeed, a prolonged recovery of
12      residual spirometric  decrements following the initial rapid improvement after exposure
13      termination could be due to slowly  resolving airway inflammation. Bronchial biopsies
14      performed 6 h postexposure have shown that O3 caused a significant decrease in
15      immunoreactivity to substance P in the submucosa (Krishna et al., 1997a). A strong negative
16      correlation with FEVj also suggests that the release of substance P may be a contributing
17      mechanism to persistent post-O3 bronchoconstriction (Krishna et al., 1997a). Persistent
18      spirometry changes observed for up to 48 h postexposure could plausibly be sustained by
19      the inflammatory mediators, many of which have bronchoconstrictive properties (Blomberg
20      etal., 1999).
21
22
23      AX6.3   PULMONARY FUNCTION EFFECTS OF OZONE EXPOSURE IN
24               SUBJECTS WITH PREEXISTING DISEASE
25          This section examines the effects of O3 exposure on pulmonary function in subjects with
26      preexisting disease by reviewing O3 exposure studies that utilized  subjects with (1) chronic
27      obstructive pulmonary disease (COPD), (2) asthma, (3) allergic rhinitis, and (4) ischemic heart
28      disease. Studies of subjects with preexisting disease exposed to O3, published subsequent to or
29      not included in the 1996 Air Quality Criteria Document (U.S. Environmental Protection Agency,
30      1996), are summarized in Table AX6-3.  Studies examining increased airway responsiveness
31      after O3 exposure are discussed in Section AX6.8.

        August 2005                           AX6-28     DRAFT-DO NOT QUOTE OR CITE

-------
                                                Table AX6-3.  Ozone Exposure in Subjects with Preexisting Disease"
to
O
O
X
to
VO
 H
 6
 o
 o
 H
O
 O
 H
 W
 O
 O
 HH
 H
 W
Ozone
Concentration'
ppm
Hg/ni3

Exposure Duration
and Activity

Exposure Number and
Condition Gender of Subjects

Subject Characteristics

Observed Effect(s)

Reference
Subjects with Chronic Obstructive Pulmonary or Heart Disease
0.0
0.24




0.3




0
472




589




4hIE
15 min exercise
15 min rest
VE ~ 20 L/min


3hIE
VE ~ 30 L/min



24 °C 9 M
40% RH 10 M




22 °C 10 M
50% RH

6M

COPD patients
Age-matched healthy NS
All subjects 59-71 years old



Hypertension
42-61 years old

Healthy
41-49 years old
No significant changes in FEVj, FVC, or SRaw due to
ozone in COPD patients. Equivocal SaO2 decrement
during 2nd and 3rd hours of ozone exposure in COPD
patients. Adjusted for exercise, ozone effects did not
differ significantly between COPD patients and healthy
subjects.
O3-induced FEV; decrements of 6.7 and 7.6% in healthy
and hypertensive subjects, respectively. Significant
O3-induced reductions in alveolar-arterial oxygen tension
in both groups. No significant changes in cardiac
enzymes or ECG telemetry.
Gong et al.
(1997a)
Gong and
Tierney (1995)


Gong et al.
(1998)



Subjects with Allergic Rhinitis
0.0
0.2

0
392

IhCE
at VE = 25 L/min/m2
BSA
20 °C 13 M, 1 F
50% RH

Dust mite sensitized asthmatics
mean age 29 ± 5 years

FEVj decrement following O3 of 10% not significantly
different from the 4% decrement following FA. Subjects
received dust mite antigen challenge at 0.5 h FA and O,
Chen et al.
(2004)

0.0 0 1 h CE 20 °C 13 M, 1 F
0.2 392 at VE = 25 L/min/m2 50% RH
BSA





0.125 245 3hIE 27 °C 5 F, 6 M
0.250 490 (10 min rest, 15 min 50 % RH
exercise on bicycle) 6 F, 16 M
VE = 30 L/min

0.125 245 3hIEx4days



Dust mite sensitized asthmatics
mean age 29 ± 5 years






Mild bronchial asthma
20-53 years old
Allergic rhinitis
19-48 years old





FEVj decrement following O3 of 10% not significantly
different from the 4% decrement following FA. Subjects
received dust mite antigen challenge at 0.5 h FA and O3
postexposures and were lavaged 6 h post-challenge.
Amount of allergen producing 1 5% FEVj decrement was
decreased by O3 compared to FA in 9 of 14 subjects.
PMN in proximal airway lavage tended to be greater after
O3 than FA (p = 0.06).
Mean early-phase FEVj response and number of >20%
reductions in FEVj were significantly greater after
0.25ppmO3or4 x 0.125 ppm O3. Most of the >15%
late-phase FEVj responses occurred after 4 days of
exposure to 0.125 ppm O3, as well as significant
inflammatory effects, as indicated by increased sputum
eosinophils (asthma and allergic rhinitis) and increased
sputum lymphocytes, mast cell tryptase, histamine, and
LDH (asthma only).
Chen et al.
(2004)






Holz et al.
(2002)







0.0
0.25



0
490



3hIE,
VE = 30 L/min
15 min ex/10 min
rest/5 min no O3;
every 30 min.
27 °C
54% RH
mouthpiece
exposure

                                                                 13 M, 11 F

                                                                 6M, 6F


                                                                 5M, 5F
Atopic mild asthma

Positive allergen and IgE tests

Healthy NS
O3-induced FEVj decrements of 12.5, 14.1, and 10.2%
in asthmatics, allergic rhinitics and healthy subjects,
respectively (group differences not significant).
Methacholine responsiveness increased in asthmatics.
Allergen responsiveness: increased significantly after O3
exposure in asthmatics (~ 2 dose shift) and a smaller shift
is rhinitics. No change in healthy. Neither allergen or
methacholine response correlated with lung function and
were not correlated with each.
Jorres et al.
(1996)

-------
                              Table AX6-3 (cont'd). Ozone Exposure in Subjects with Preexisting Disease"
to
o
o
X
ON

OJ
O
H

6
o


o
H

O

O
H
W

O


O
HH
H
W
Ozone
Concentration'
ppni ug/ni3
Exposure Duration Exposure
and Activity Conditions
Number and
Gender of Subjects Subject Characteristics
Observed Effect(s)
Reference
Subjects with Asthma
0.4 784







0.0 0
0.4 784





0.12 236

0.0 0
0.2 392




0.4 784



0.0 0
0.12 235

0.4 784






2h IE NA
(15 min rest, 15 min exercise
on bicycle)
VE = 30 L/min




2h IE NA
4x15 min on bicycle,
VE = 40 L/min




Rest 22 °C
40% RH
6 h 22 °C
30 min rest/30 min exercise 50% RH
VE ~ 25 L/min



3h 6x15 min cycle ergometer 31 °C
VE » 32L/min 35% RH
5 consecutive days

1 h rest NA
air-antigen
O3-antigen
2 h IE Head mask
15 min exercise exposure
15 min rest »18°C
VE » 20L/min 60% RH



4 F, 5 M Healthy
(25 ± 2 years old)

7 F, 6 M Mild atopic asthma;
beta agonists only
(22 ± 0.7 years old)


15 Healthy adults
18-40 years old

9 Mild atopic asthmatics
18-40 years old


10 M, 5 F atopic asthma

5 M Healthy NS
5 M Asthmatics, physician
diagnosed,
All 18-45 years


8 M , 2 F Asthmatic NS adults
beta-agonist use only
19-48 years old
ATS criteria for asthma
9 M, 6 F Mild allergic asthma; 18 to
49 years of age.

5 M, 1 F Healthy adults
6 M Atopic asthmatics





Significant reductions in FVC (12%, 10%) and
FEV[ (13%, 1 1%) for asthmatic and healthy
subjects, respectively; attenuated by indomethacin
in healthy subjects only. Significant reductions in
mid-flows which tended to be greater in asthmatics
than healthy subjects. Indomethacin treatment
attenuated mid-flow-reductions somewhat more
in asthmatics than healthy subjects.
Sputum collected 24 h before and 4-6 h post-O3
exposure. Baseline GDI Ib expression positively
correlated with O3-induced PMN. Increased
expression of mCD14 on macrophages following
O3 compared to FA. Asthmatic PMN response
similar to healthy subjects (also see Table AX6-3).
No spirometric data available.
No effect of O3 on airway response to grass
allergen.
Similar spirometric responses in asthmatic and
healthy. However, preexposure FEV1 and FVC
were both —0.4 L lower on O3-day than FA day.
More PMN's in asthmatics. IL-8 and IL-6 higher
in asthmatics exposed to O3. No relationship of
spirometry and symptoms to inflammation.
FEVj decreased 35% on first exposure day.
Methacholine reactivity increased about ten-fold.
Also see Table AX6- 7 for repeated exposure results.

No effect of O3 on airway response to grass or
ragweed allergen.

FEVj responses of healthy and asthmatic similar
(~ 15% decrease). Maximal FEVj response to
methacholine increased similarly in both groups
(12 h postexposure). Larger increase in PC20 in
healthy subjects. Both groups had increased
PMN's in sputum no correlation of PMN's
and lung function.
Alexis et al.
(2000)






Alexis et al.
(2004)





Ball et al.
(1996)
Basha et al.
(1994)




Gong et al.
(1997b)


Hanania
etal. (1998)

Hiltermann
etal. (1995)






-------
                             Table AX6-3 (cont'd).  Ozone Exposure in Subjects with Preexisting Disease"
to
o
o
X
H

6
o


o
H

O

O
H
W

O


O
HH
H
W
Ozone
Concentration'
ppm
Hg/ni3
Number and
Exposure Duration Exposure Gender of
and Activity Conditions Subjects
Subject
Characteristics
Observed Effect(s)
Reference
Adult Subjects with Asthma (cont'd)
0.0
0.15
0.25
0.25



0.0
0.16




0.0
0.25






0.16




0.25
0.40


0.0
0.2



0
294
490
490



0
314




0
490






314




490
784


0
392



3hIE 27±1°C 10M, 11F
15 min rest 56 ± 7 % RH
(VE = 7 L/min)
15 min on bicycle 5M, 10F
(VE = 26 L/min)


7.6 h 18 °C 13 M
25 min treadmill, 40% RH
25 min cycle/10 min rest 7 M, 10 F
per hour.
VE = 27-32 L/min

3hIE, 27 °C 13 M, 11 F
VE = 30 L/min 54% RH
15 min ex/10 min rest/5 mouthpiece 6 M, 6 F
min no O3; every 30 min exposure

5M, 5F


7.6 h 22 °C 4 M, 5 F
25 min treadmill, 25 min 40 % RH
cycle/10 min rest per
hour.
VE = 25 L/min
VE = 25-45 L/min NA 8 M, 4 F
8 M, 10 F
22 M, 16 F

2 h IE 20 °C 6 M, 9 F
4x15 min 50% RH
at VE = 20
L/min/m2 BSA 9 M, 6 F

Healthy NS
28 ± 5 years old

Mild Asthmatic
30 ± 8 years old


Healthy NS,
age 19-32 years.
Moderate Asthmatics,
physician diagnosed,
beta agonist users,
age 19-32 years.
Atopic mild asthma

Positive allergen and
IgE tests

Healthy NS


Mild atopic asthma;
no meds 12 h
pre-exposure
20-35 years old

Asthmatics
Allergic rhinitics
Healthy adults
All <26 years old
Healthy adults
24 years old

Mild asthmatics
29 years old
No significant O3-induced group differences in symptoms or
spirometry. After 0.25 ppm O3, there were significant
decrements in FEV; and FVC that tended to be greater in the
asthmatics than controls. Small but significant neutrophil
increase in asthmatics following 0.15 ppm O3. Significant
neutrophil increases following 0.25 ppm O3 that did not differ
between groups.
FEVj decreased 19% in asthmatics and only 10% in
nonasthmatics. High responders had worse baseline airway
status. More wheeze in asthmatics after O3.



O3-induced FEVj decrements of 12.5, 14.1, and 10.2% in
asthmatics, allergic rhinitics and healthy subjects, respectively
(group differences not significant). Methacholine
responsiveness increased in asthmatics. Allergen responsiveness
increased after O3 exposure in asthmatics (=2 dose shift),
a smaller shift occurred in rhinitics, no change occurred in
healthy subjects. Neither allergen nor methacholine responses
were correlated with each other or with lung function.
Significant FEVj decrease of 9.1 % following O3 exposure;
marked individual variability with responses ranging from 2
% to 26 %.


Healthy 12.2% decrease in FEV1; Rhinitics 10.1%,
asthmatics 12.4%


O3-induced FEVj decrement (8%, healthy adults; 3% asthmatics)
and PMN increase (20.6%, healthy adults; 15.2% asthmatics).
Primary goal was to investigate relationship between antioxidant
defenses and O3 responses in asthmatics and healthy adults
(see Tables AX6-3 andAX6 -13).
Holz et al.
(1999)





Horstman
etal. (1995)




Jorres et al.
(1996)






Kehrl et al.
(1999)



Magnussen
etal. (1994)


Mudway
etal. (2001)
Stenfors
et al. (2002)


-------
                                            Table AX6-3 (cont'd).  Ozone Exposure in Subjects with Preexisting Disease"
55 Ozone
ij Concentration'
u
S ppm
Hg/m3

Exposure Duration Exposure
and Activity Conditions
Number and Subject
Gender of Subjects Characteristics
Observed
Effect(s)
Reference
Adult Subjects with Asthma (cont'd)
0.2
396
2h IE 22 °C
( 1 5 min rest, 1 5 min 40 % RH
5 F, 4 M Mild atopic asthma; no
meds 8 h pre-exposure
Significant decrease in FEVj and a trend toward decreases
in mean inspiratory flow, FEF25, and FEF75 after O3
Newson et al.
(2000)
                            exercise on bicycle)
                            VE = 20 L/min/m2 BSA
                      21-42 years old          exposure.  No significant differences in FEF50, FVC, TLC,
                                             Raw, or sRaw. No correlation between sputum neutrophils
                                             at 6 h postexposure and FEVj immediately after exposure.
            0.4
                     784
                            2 h rest
 X
 ON
 OJ
 to
                                                   21 °C
                                                   40% RH
0.16     314    7.6 h                   18 °C
                25 min treadmill,        40% RH
                25 min cycle/ every
                hour.

0.0       0     4h                     21 °C
0.2      392    50 min exercise,         50% RH
                10 min rest each hour.
                VF » 45-50 L/min
1 1 M , 1 1 F
                                                                       8 M
12 M, 6F
Asthmatics sensitive
to D Farinae,
physician diagnosed,
18 to 35 years

Mild asthmatics,
physician diagnosed,
reactive to dust mite
D. Farinae.

18 adult mild
asthmatics mostly
beta agonist users.
Ozone resulted in nasal inflammation (increased PMN's)      Peden et al.
and caused augmented response to nasal allergen challenge.    (1995)
                                                                                                                     Increased eosinophils and PMN's after O3 exposure more     Peden et al.
                                                                                                                     in initial (bronchial) fraction.  No correlation of eosinophils   (1997)
                                                                                                                     and PMN's, FEVj & FVC decreased 14% and 9%
                                                                                                                     respectively.
FVC, FEVj decreased 17.6% and 25% respectively.
Trend for larger increase in SRaw in asthmatics.  Larger
increase in PMN's and protein in asthmatics indicating
more inflammation. No increase in eosinophils.
Spirometry changes in asthmatics similar to healthy
subjects (Aris et al., 1995; Balmes et al., 1997).
Scannell et al.
(1996)
 H
 6
 o
 o
 H
O
 O
 H
 W
 O
 O
 HH
 H
 W
           "See Appendix A for abbreviations and acronyms.
           'Grouped by rest and exercise; within groups listed from lowest to highest O3 concentration.

-------
 1     AX6.3.1  Subjects with Chronic Obstructive Pulmonary Disease
 2          Five studies of O3-induced responses in COPD patients were available for inclusion in the
 3     1996 criteria document (U.S. Environmental Protection Agency, 1996).  The COPD patients in
 4     these studies were exposed during light IE (4 studies) or at rest (1 study) for 1 to 2 hours to O3
 5     concentrations between 0.1 and 0.3 ppm.  None of theses studies found significant O3-induced
 6     changes in pulmonary function.  Of the four studies examining arterial oxygen saturation, two
 7     reported small but statistically significant O3-induced decreases in the COPD patients. These
 8     limited data suggest COPD patients experience minimal O3-induced effects for 0.3 ppm O3
 9     exposures less than 2 hours in duration. These findings are also consistent decreasing O3 effects
10     with increasing age (see Section AX6.5.7).
11          More recently, Gong et al. (1997a) exposed 9 COPD patients (age range, 59 to 71 years;
12     mean age 66 ± 4 years) and 10 healthy NS (age range, 60 to 69 years; mean age 65 ± 3 years)
13     to 0.24 ppm for 4 h with interment light exercise (-20 L/min). COPD patients had decreases
14     in FEVj following both clean air (-11%, p = 0.06) and O3 (-19%, p < 0.01) exposures.
15     These  FEVj decrements, presumably due to exercise, were primarily attributable to four of
16     the patients who lost greater than 14% of their FEVj following both the air and O3 exposures.
17     Relative to clean air, O3 caused a statistically insignificant FEVj decrement of -8% in COPD
18     patients which was not statistically different from the decrement of -3% in healthy subjects.
19     Ozone-induced symptoms, sRaw, SaO2, and postexposure bronchial activity also exhibited little
20     or no difference between the COPD patients and the healthy subjects.
21
22     AX6.3.2  Subjects with Asthma
23          Based on studies reviewed in the 1996 criteria document (U.S. Environmental Protection
24     Agency, 1996) asthmatics appear to be at least as sensitive to acute effects of O3 as healthy
25     nonasthmatic subjects. At rest, neither adolescent asthmatics nor healthy controls had significant
26     responses as a result of an hour exposure to 0.12 ppm O3. Exposure of adult asthmatics to
27     0.25 ppm O3 for 2 h at rest also caused no significant responses.  Preexposure to between 0.10
28     and 0.25 ppm O3 for 1 hr with light IE does not appear to exacerbate exercise-induced asthma
29     (Fernandes et al., 1994; Weymer et al., 1994). At higher exposures (0.4 ppm O3 with heavy
30     IE for  2 h), Kreit et al. (1989) and Eschenbacher et al.  (1989) demonstrated significantly
       August 2005                            AX6-33      DRAFT-DO NOT QUOTE OR CITE

-------
 1      greater FEVj and FEF25_75 decrements in asthmatics than in healthy controls. With longer
 2      duration exposures to lower O3 levels (0.12 ppm with moderate IE for 6.5 h), asthmatics have
 3      also shown a tendency for greater FEVj decrements than healthy nonasthmatics (Linn et al.,
 4      1994). Newer clinical studies (see Table AX6-3) continue to suggest that asthmatics are at least
 5      as sensitive as healthy controls to O3-induced responses.
 6           Studies of less than 3 h duration have reported similar or tendencies for increased
 7      O3-induced spirometric responses up to O3 concentrations of 0.4 ppm. Similar group decrements
 8      in FEVj and FVC were reported by Hiltermann et al. (1995), who exposed 6 asthmatics and
 9      6 healthy subjects to  0.4 ppm O3 for 2 h with light IE. Alexis et al. (2000) exposed 13 mild
10      atopic asthmatics and 9 healthy subjects for 2 h to 0.4 ppm O3 with IE ( VE = 30 L/min). Similar
11      O3-induced group decrements in FEVj and FVC were also reported by these investigators.
12      A tendency, however, for an increased O3-induced reduction in mid-flows (viz., FEF25, FEF50,
13      FEF60p, FEF75) was reported for the asthmatics relative to the healthy  subjects.  In a larger study,
14      Torres et al. (1996) exposed 24 asthmatics, 12 allergic rhinitis, and 10 healthy subjects to
15      0.25 ppm O3  for 3 h with IE. Statistically significant O3-induced decreases in FEVj occurred in
16      all groups, but tended to be lower in healthy controls (allergic rhinitis, -14.1%; asthmatics,
17      -12.5%; healthy controls, -10.2%).  Holz et al. (1999) exposed 15 asthmatics and 21 healthy
18      controls to 0.15 and 0.25 ppm O3 for 3-h with light IE. After the 0.25 ppm O3 exposure, there
19      were significant decrements in FEVj and FVC that tended to be slightly greater in the asthmatics
20      than controls. One study reported that asthmatics tended to have less of an FEVj response to O3
21      than healthy controls (Mudway et al., 2001). In that study, however, the asthmatics also tended
22      to be older than the healthy subjects which could partially explain their lesser response.
23           Studies between 4 and 8 h duration, with O3 concentrations of 0.2 ppm or less, also suggest
24      a tendency for increased O3-induced pulmonary function responses in asthmatics relative to
25      healthy subjects. Scannell et al. (1996) exposed 18 asthmatics to 0.2 ppm O3 for 4 h with
26      IE (VE -25 L/min/m2 BSA).  Baseline and hourly pulmonary  function measurements of FEVl5
27      FVC, and sRaw were obtained.  Asthmatic responses were compared to 81 healthy subjects who
28      underwent similar experimental protocols (Aris et al., 1995; Balmes et al., 1996).  Asthmatic
29      subjects experienced a significant O3-induced increase in sRaw, FEVj and FVC. The O3-induced
30      increase in sRaw tended to be greater in asthmatics than the healthy subjects, whereas similar
        August 2005                            AX6-34      DRAFT-DO NOT QUOTE OR CITE

-------
 1      group decrements in FEVj and FVC were observed. Basha et al. (1994) also reported similar
 2      spirometric responses between 5 asthmatic and 5 healthy subjects exposed to 0.2 ppm O3 for 6 h
 3      with IE.  However, the mean preexposure FEVj in the asthmatics was about 430 mL less (i.e.,
 4      -12% decreased) on the O3-day relative to the air-day. In a longer exposure duration (7.6 h)
 5      study, Horstman et al. (1995) exposed 17 asthmatics and 13 healthy controls to 0.16 ppm O3 or
 6      FA with alternating periods of exercise (50 min, VE «30 L/min) and rest (10 min).  Both groups
 7      had significant O3-induced decrements in FEVl5 FVC, and FEV25.75. The asthmatic and healthy
 8      subjects  had similar O3-induced reductions in FVC.  The FEVj decrement experienced by the
 9      asthmatics was significantly greater in the healthy controls (19% versus 10%, respectively).
10      There was also tendency for a greater O3-induced decrease in FEF25.75 in asthmatics relative to
11      the healthy subjects (24% versus 15%, respectively).
12           With repeated O3 exposures asthmatics, like healthy subjects (see Section AX6.6)
13      develop tolerance.  Gong et al. (1997b) exposed 10 asthmatics to 0.4 ppm O3, 3 h per day with
14      IE (VE -32 L/min), for 5 consecutive days.  Symptom and spirometric responses were greatest
15      on the first (-35 % FEVj) and second (-34 % FEVj) exposure days, and progressively
16      diminished toward baseline levels (-6 % FEVj) by the fifth exposure day.  Similar to healthy
17      subjects, asthmatics lost their tolerance 4 and 7 days later.
18           Other published studies with similar results (e.g., McBride et al., 1994; Basha et al., 1994;
19      Peden et al., 1995, 1997; Peden, 2001a;  Scannell et al., 1996; Hiltermann et al., 1997, 1999;
20      Michelson et al., 1999; Vagaggini et al., 1999; Newson et al., 2000; Holz et al., 2002) also
21      reported that asthmatics  have a reproducible and somewhat exaggerated inflammatory response
22      to acute O3 exposure (see Section AX6.9}. For instance, Scannell et al. (1996) performed lavages
23      at 18 h post-O3 exposure to assess inflammatory responses in asthmatics.  Asthmatic responses
24      were compared to healthy subjects who underwent a similar experimental protocol (Balmes
25      et al., 1996).  Ozone-induced increases in BAL neutrophils and total protein were significantly
26      greater in asthmatics than healthy subjects.  There was also a  trend for an ozone related increased
27      IL-8 in the asthmatics relative to healthy subjects. Inflammatory responses do not appear to be
28      correlated with lung function responses in either asthmatic or healthy subjects (Balmes et al.,
29      1996, 1997; Holz et al.,  1999). This lack of correlations between inflammatory and spirometric
30      responses may be due to differences in the time kinetics of these responses (Stenfors et al.,
        August 2005                            AX6-3 5      DRAFT-DO NOT QUOTE OR CITE

-------
 1      2002). In addition, airway responsiveness to inhaled allergens is increased by O3 exposure in
 2      subjects with allergic asthma for up to 24 h (see Section AX6.8).
 3           One of the difficulties in comparing O3-induced spirometric responses of healthy subjects
 4      versus asthmatics is the variability in responsiveness of asthmatics.  Most of the asthma studies
 5      were conducted on subjects with a clinical history of mild disease. However, classification of
 6      asthma severity is not only based on functional assessment (e.g., percent predicted FEVj), but
 7      also on clinical symptoms, signs, and medication use (Table AX6-4). Although "mild atopic
 8      asthmatics" are frequently targeted as an experimental group, the criteria for classification has
 9      varied considerably within and across the available published studies. Although the magnitude
10      of group mean changes in spirometry may not be significantly different between healthy and
11      asthmatic subjects, many of the studies have reported clinically significant changes in some
12      individuals.
13           Alexis et al. (2000) explored the possibility that the mechanisms of O3-induced spirometric
14      responses may differ between asthmatics and healthy subjects.  Physician-diagnosed mild
15      atopic asthmatics and healthy subjects were pretreated with 75 mg/day of indomethacin
16      (a COX inhibitor) or placebo and then exposed for 2 h to 0.4 ppm O3 or to FA during mild
17      IE (VE = 30 L/m). The number and severity of O3-induced symptoms were significantly
18      increased in both asthmatics and healthy subjects. These symptom responses were similar
19      between the subject groups  and unaffected by indomethacin pretreatment. Asthmatics and
20      healthy subjects also had similar O3-induced reductions in FVC and FEVj. These restrictive-
21      type responses, occurring due to the  combined effects of bronchoconstriction and reflex
22      inhibition of inspiration (see Section AX6.2.7), were attenuated by indomethacin in the healthy
23      subjects but not the asthmatics.  Thus, in healthy subjects but not asthmatics, COX metabolites
24      may contribute to O3-induced reductions in FVC and FEVj. As assessed by the  magnitude of
25      reductions in mid-flows (viz. FEF25,  FEF50, FEF60p, FEF75), the small airways of the asthmatics
26      tended to be more affected than the healthy subjects.  This suggests asthmatics may be more
27      sensitive to small airway effects of O3, which is consistent with the observed increases in
28      inflammation and airway responsiveness. Indomethacin pretreatment attenuated some of
29      these O3-induced small airways effects (FEF50 in healthy subjects, FEF60p in asthmatics).
30
        August 2005                            AX6-36      DRAFT-DO NOT QUOTE OR CITE

-------
to
o
o
X
                                                  Table AX6-4.  Classification of Asthma Severity1

Classification
Severe
persistent
Moderate
persistent

Days with
Step symptoms
4 Continual
3 Daily

Nights with
symptoms
Frequent
>l/week
Lung Function2
FEVlorPEF PEF
% predicted variability
oral (%)
<60 >30
between >30
60 and 80
Medications3

Daily
High-dose inhaled steroids (ICS)
and long-acting inhaled p2-agonist
If needed, add oral steroids
Low-to-medium-dose ICS and
long-acting p2-agonist (preferred)

Quick relief
Short-acting inhaled
p2-agonist, as needed; oral
steroids may be required
Short-acting inhaled
p2-agonist, as needed;
       Mild
       persistent
>2/week,
but2/week
>80
       Mild
       intermittent
  <2/week
<2/month
^80
                            Or
            Medium-dose ICS (another preferred
            option for children ages <5 years)
                            Or
            Low-to-medium-dose ICS and either
            leukotriene modifier or theophylline

20-30       Low-dose inhaled steroids (preferred)
                            Or
            Cromolyn leukotriene modifier,
            or (except for children aged <5 years)
            nedocromil or sustained release
            theophylline to serum concentration
            of 5-15 ug/mL

 <20        No daily medicine needed
                                                                                                                           oral steroids may be
                                                                                                                           required
Short-acting inhaled
p2-agonist, as needed;
oral steroids may be
required
Short-acting inhaled
p2-agonist, as needed;
oral steroids may be
required
       1 Sources: Centers for Disease Control (2003); National Institutes of Health (1997, 2003).
       2For adults and children aged >5 years who can use a spirometer or peak flow meter.
       3The medications listed here are appropriate for treating asthma at different levels of severity. The preferred treatments, dosage, and type of medication
        recommended vary for adults and children and are detailed in the EPR-Update 2002 stepwise approach to therapy. The stepwise approach emphasizes that
        therapy should be stepped up as necessary and stepped down when possible to identify the least amount of medication required to achieve goals of therapy.
        The stepwise approach to care is intended to assist, not replace, the clinical decision-making required to meet individual patient needs.

-------
 1      AX6.3.3  Subjects with Allergic Rhinitis
 2           Most O3 exposure studies in humans with existing respiratory disease have focused on lung
 3      diseases like COPD and asthma. However, chronic inflammatory disorders of the nasal airway,
 4      especially allergic rhinitis, are very common in the population.  People with allergic rhinitis have
 5      genetic risk factors for the development of atopy that predispose them to increased upper airway
 6      responsiveness to specific allergens as well as nonspecific air pollutants like O3.  Studies
 7      demonstrating the interaction between air pollutants and allergic processes in the human nasal
 8      airways and rhinoconjunctival tissue have been reviewed by Peden (2001b) and Riediker et al.
 9      (2001), respectively.  Ozone exposure of subjects with allergic rhinitis has been shown to induce
10      nasal inflammation and increase airway responsiveness to nonspecific bronchoconstrictors,
11      although to a lesser degree than experienced by asthmatics.
12           McDonnell et al. (1987) exposed nonasthmatic adults with allergic rhinitis to 0.18 ppm O3.
13      The allergic rhinitics were no more responsive to O3 than healthy controls, based on symptoms,
14      spirometry, or airway reactivity to histamine although they had a small but significantly greater
15      increase in SRaw. The data on subjects with allergic rhinitis and asthmatic subjects suggest that
16      both of these groups have a greater rise in Raw to O3 with a relative order of airway
17      responsiveness to O3 being normal < allergic < asthmatic.
18           Bascom et al. (1990) studied the upper respiratory response to acute O3 inhalation, nasal
19      challenge with  antigen, and the combination of O3 plus antigen in subjects with allergic rhinitis.
20      Exposure to O3 caused significant increases in upper and lower airway symptoms, a mixed
21      inflammatory cell influx with a seven-fold increase in nasal lavage PMNs, a 20-fold increase in
22      eosinophils, and a 10-fold increase in mononuclear cells, as well as an apparent sloughing of
23      epithelial cells. McBride et al. (1994) also observed increased nasal PMN's after O3 exposure in
24      atopic asthmatics. Peden et al. (1995), who studied allergic asthmatics exposed to O3 found
25      that O3 causes an increased response to nasal allergen challenge in addition to nasal
26      inflammatory responses. Their data suggested that allergic subjects have an increased immediate
27      response to allergen after O3 exposure. In a follow-up study, Michelson et al.  (1999) reported
28      that 0.4 ppm O3 did not promote early-phase-response mediator release or enhance the response
29      to allergen challenge in the nasal airways of mild,  asymptomatic dust mite-sensitive asthmatic
30      subjects.  Ozone did, however, promote an inflammatory cell influx, which helps induce a more
31      significant late-phase response in this population.

        August 2005                             AX6-3 8     DRAFT-DO NOT QUOTE OR CITE

-------
 1           Torres et al. (1996) found that O3 causes an increased response to bronchial allergen
 2      challenge in subjects with allergic rhinitis. This study also compared responses in subjects
 3      with mild allergic asthma (see Sections AX6.3.2 andAX6.8).  The subjects were exposed to
 4      0.25 ppm O3 for 3 h with IE.  Airway responsiveness to methacholine was determined 1 h before
 5      and after exposure; responsiveness to allergen was determined 3 h after exposure.  Statistically
 6      significant decreases in FEVj occurred in subjects with  allergic rhinitis (13.8%) and allergic
 7      asthma (10.6%), and in healthy controls (7.3%). Methacholine responsiveness was statistically
 8      increased in asthmatics, but not in subjects with allergic rhinitis. Airway responsiveness to an
 9      individual's historical allergen (either grass and birch pollen, house dust mite, or animal dander)
10      was significantly increased after O3 exposure when compared to FA exposure.  In subjects with
11      asthma and allergic rhinitis, a maximum percent fall in FEVj of 27.9 % and 7.8%, respectively,
12      occurred 3 days after O3 exposure when they were challenged with of the highest common dose
13      of allergen. The authors concluded that subjects with allergic rhinitis, but without asthma, could
14      be at risk if a high O3 exposure is followed by a high dose of allergen.
15           Holz et al. (2002) extended the results of Torres et al. (1996) by demonstrating that
16      repeated daily exposure to lower concentrations of O3 (0.125 ppm for 4 days) causes an
17      increased response to bronchial allergen challenge in subjects with preexisting allergic airway
18      disease, with or without asthma. There was no major difference in the pattern of bronchial
19      allergen response between subjects with asthma or rhinitis, except for a 10-fold increase in the
20      dose of allergen required to elicit a similar response (>20% decrease in FEVj) in the asthmatic
21      subjects. Early phase responses were more consistent in subjects with rhinitis and late-phase
22      responses were more pronounced in subjects with asthma. There also was a tendency towards a
23      greater effect of O3 in subjects with greater baseline response to specific allergens chosen on the
24      basis of skin prick test and history (viz., grass, rye, birch, or alder pollen, house dust mite, or
25      animal dander).  These data suggest that the presence of allergic bronchial sensitization, but not a
26      history of asthma, is a key determinant of increased airway allergen responsiveness with O3.
27      [A more complete discussion of airway responsiveness is found in Section AX6.8]
28
29      AX6.3.4   Subjects with Cardiovascular Disease
30           Superko et al. (1984) exposed six middle-aged males with angina-symptom-limited
31      exercise tolerance for 40 min to FA and to 0.2 and 0.3 ppm O3 while they were exercising

        August 2005                             AX6-39     DRAFT-DO NOT QUOTE OR CITE

-------
 1      continuously according to a protocol simulating their angina-symptom-limited exercise training
 2      prescription (meanVE= 35 L/min). No significant pulmonary function impairment or evidence
 3      of cardiovascular strain induced by O3 inhalation was observed.  Gong et al. (1998) exposed
 4      hypertensive (n = 10) and healthy (n = 6) adult males, 41 to 78 years of age, to FA and on the
 5      subsequent day to 0.3 ppm O3 for 3 h with IE at 30 L/min.  The ECG was monitored by
 6      telemetry, blood pressure by cuff measurement, and a venous catheter was inserted for
 7      measurement of routing blood chemistries and cardiac enzymes. Pulmonary artery and radial
 8      artery catheters were placed percutaneously for additional blood sampling and for measurement
 9      of hemodynamic pressures, cardiac output, and SaO2. Other hemodynamic variables were
10      calculated, including cardiac index,  stroke volume, pulmonary and systemic vascular resistance,
11      left and right ventricular stroke-work indices, and rate-pressure product. Spirometric volumes
12      (FVC, FEVj) and symptoms of breathing discomfort were measured before and after the
13      O3 exposures.  There were significant O3-induced FEVj decrements in both subject groups that
14      did not defer between groups (hypertensive, 7.6%; healthy, 6.7%). The overall results did not
15      indicate any major acute cardiovascular effects of O3 in either the hypertensive or normal
16      subjects. However,  statistically significant O3 effects for both groups combined were  increases
17      in HR, rate-pressure product, and the alveolar-to-arterial PO2 gradient,  suggesting that impaired
18      gas exchange was being compensated for by increased myocardial work.  These effects might be
19      more important in some patients with severe cardiovascular disease.  [See Section AX6.10 for
20      discussion of extrapulmonary effects ofO3 exposure.]
21
22
23      AX6.4   INTERSUBJECT VARIABILITY AND REPRODUCIBILITY
24               OF RESPONSE
25           Analysis of the factors that contribute to intersubject variability is important for  the
26      understanding of individual responses, mechanisms of response, and health risks associated with
27      acute O3 exposures.  Bates et al. (1972) noted that variation between individuals in sensitivity
28      and response was evident in respiratory symptoms and pulmonary function following  O3
29      exposure. A large degree of intersubject variability in response to O3 has been consistently
30      reported in the literature (Adams et al.,  1981; Aris et al., 1995; Folinsbee et al., 1978;  Kulle
31      et al., 1985;  McDonnell et al., 1983). Kulle et al.  (1985) noted that the magnitude of variability

        August 2005                            AX6-40      DRAFT-DO NOT QUOTE OR CITE

-------
 1     between individuals in FEVj responses increases with O3 concentration. Similarly, McDonnell
 2     et al. (1983) observed FEVj decrements ranging from 3 to 48% (mean 18%) in 29 young adult
 3     males exposed to 0.40 ppm O3 for 2 h during heavy IE. At a lower O3 concentration of
 4     0.18 ppm, 20 similarly exposed subjects had FEVj decrements ranging from 0 o 23%
 5     (mean = 6%), while those exposed to FA (n = 20) had decrements ranging from -2% to 6%
 6     (mean = 1%) (McDonnell et al., 1983). All of the subjects in these studies were young adult
 7     males. (Intersubject variability related to age and gender is discussed in Sections AX6.5.1 and
 8     AX6.5.2, respectively.)
 9          More recently, McDonnell (1996) examined the FEVj response data from three 6.6 h
10     exposure studies of young adult males conducted at the EPA Health Effects Research Laboratory
11     in Chapel Hill, NC (Folinsbee et al., 1988; Horstman et al.,  1990; McDonnell et al., 1991).
12     The response distributions for subjects at each of four O3 concentrations (0.0, 0.08, 0.10, and
13     0.12 ppm) are illustrated in Figure AX6-6. It is apparent that the FEVj responses in FA are
14     small with most tightly grouped around zero. With increasing O3 concentration, the mean
15     response increases as does the variability about the mean. At higher O3 concentrations, the
16     distribution of response becomes asymmetric with a few individuals experiencing large FEVj
17     decrements. The response distribution in Figure AX6-6 allows estimates of the number or
18     percentage of subjects responding in excess of a certain level.  With FA exposure, none of
19     87 subjects had a FEVj decrement in excess of 10%; however, 26%, 31%, and 46% exceeded a
20     10% decrement at 0.08, 0.10, and 0.12 ppm, respectively. FEVj decrements as large as 30 to
21     50% were even observed in some individuals. In 6.6-h face mask exposures of young adults
22     (half women) to 0.08 ppm O3, Adams (2002) found that 6 of 30 subjects (20%) had >10%
23     decrements in FEVj. The response distributions in Figure AX6-6 underlines the wide range of
24     response to O3 under prolonged exposure conditions and reinforces the observations by others
25     consequent to 2 h IE exposures at higher O3 concentrations (Horvath et al.,1981; McDonnell
26     etal., 1983).
27          Some of the intersubject variability in response to O3 inhalation may be due to intrasubject
28     variability, i.e., how reproducible the measured responses are in an individual between several
29     O3 exposures. The more reproducible the  subject's response, the more precisely it indicates
30     his/her intrinsic responsiveness. McDonnell et al. (1985a) examined the reproducibility of
31     individual responses to O3 in healthy human subjects (n = 32) who underwent repeated

       August 2005                            AX6-41      DRAFT-DO NOT QUOTE OR CITE

-------
                                 40-
                                 20-
                                                   0.12 ppm
                                  -15-10 -5  0  5 10 15 20 25 30 35 40 45 50
                                 50-
                                 40-
                            W
                            +J

                            o
                                                   0.10 ppm
                                  -15-10 -5  0  5 10 15 20 25 30 35 40 45 50
                                 50-
                                 20-
                                                   0.08 ppm
                                               II  I  ni—in
                                  -15-10 -5  0  5 10 15 20 25 30 35 40 45 50
                                 20-
                                                      0 ppm
                                  -15-10 -5  0  5 10 15 20 25 30 35 40 45 50



                                     FEV-, (% Decrement)
Figure AX6-6. Frequency distributions of percent decrements in FEVt for 6.6-h exposure

               to four concentrations of ozone.




Source: McDonnell (1996).
August 2005
AX6-42      DRAFT-DO NOT QUOTE OR CITE

-------
 1      exposures within a period of 21 to 385 days (mean = 88 days; no median reported) at one of
 2      five O3 concentrations ranging from 0.12 to 0.40 ppm. Reproducibility was assessed using the
 3      intraclass correlation coefficient (R).  The most reproducible responses studied were FVC
 4      (R = 0.92) and FEVj (R = 0.91).  However, at the lowest concentration, 0.12 ppm, relatively
 5      poor FEVj reproducibility was observed (R = 0.58) due, in part, to a lack of specific  O3 response
 6      or a uniformly small response in the majority of subjects.  McDonnell et al. (1985a)  concluded
 7      that for 2 h IE O3 exposures equal to or greater than 0.18 ppm, the intersubject differences in
 8      magnitude of change in FVC and FEVj are quite reproducible over time and likely due to
 9      differences in intrinsic responsiveness of individual  subjects. Hazucha et al. (2003)  exposed
10      47 subjects on three occasions for 1.5 h, with moderate intensity IE, to 0.40 to 0.42 ppm O3.
11      Reproducibility of FEVj responses was related to the length of time between re-exposures,
12      with a Spearman correlation R of 0.54 obtained between responses for exposures 1 and
13      2 (median = 105 days), and an R of 0.85 between responses for exposures 2 and 3
14      (median = 7 days).
15           Identification of mechanisms of response and health risks associated with acute O3
16      exposures are complicated by a poor association between various O3-induced responses.
17      For  example, McDonnell et al. (1983) observed a very low correlation between changes in sRaw
18      and  FVC (r = -0.16) for 135 subjects exposed to O3 concentrations ranging from 0.12 to
19      0.40 ppm for 2.5 h with IE. In a retrospective study of 485 male subjects (ages 18 to 36 yrs)
20      exposed for 2 h to one of six O3 concentrations at one of three activity levels, McDonnell et al.
21      (1999) observed significant, but low,  Spearman rank order correlations between FEVj response
22      and  symptoms of cough (R = 0.39), shortness of breath (R = 0.41), and pain on deep inspiration
23      (R = 0.30). The authors concluded from their data that the O3-induced responses are related
24      mechanistically to some degree, but that there is not a single factor which is responsible for the
25      observed individual differences in O3 responsiveness across the spectrum of symptom and lung
26      function responses.  This  conclusion is supported by differences in reproducibility observed by
27      McDonnell et al., (1985a). Compared to the intraclass correlation coefficient for FEVj
28      (R = 0.91), relatively low but statistically significant R values for symptoms ranged from 0.37 to
29      0.77, with that for sRaw being 0.54.  The reproducibility correlations for fB (R = -0.20) and VT
30      (R = -0.03) were not statistically significant.
        August 2005                             AX6-43      DRAFT-DO NOT QUOTE OR CITE

-------
 1           The effect of this large intersubject variability on the ability to predict individual
 2      responsiveness to O3 was demonstrated by McDonnell et al. (1993). These investigators
 3      analyzed the data of 290 male subjects (18 to 32 years of age) who underwent repeat 2 h IE
 4      exposures to one or more O3 concentrations ranging from 0.12 to 0.40 ppm in order to identify
 5      personal characteristics (i.e., age, height, baseline pulmonary functions, presence of allergies,
 6      and past smoking history) that might predict individual differences in FEVj response. Only age
 7      contributed significantly to intersubject responsiveness (younger subjects were more
 8      responsive), accounting for just 4% of the observed variance.  Interestingly, O3 concentration
 9      accounted for only 31% of the variance, strongly suggesting the importance of as yet undefined
10      individual characteristics that determine FEVj responsiveness to O3. A more general form of
11      this model was developed to investigate the O3 exposure FEVj response relationship (McDonnell
12      et al., 1997). These authors used data from 485 male  subjects (age = 18 to 36 years) exposed
13      once for 2 h to one of six O3 concentrations (ranging from 0.0 to 0.40 ppm) at one of 3 activity
14      levels (rest, n = 78; moderate IE, n = 92; or heavy IE,  n = 314).  In addition to investigating the
15      influence of subject's age, the model focused on determining whether FEVj response was more
16      sensitive to changes in C than to changes in VE, and whether the magnitude of responses is
17      independent of differences in lung size.  It was found  that the unweighted proportion of the
18      variability in individual responses explained by C, VE, T, and age was 41%, with no evidence
19      that the sensitivity of FEVj response to VE was different than changes in C, and no evidence that
20      magnitude of response was related to measures of body or lung size. The authors concluded that
21      much inter-individual variability in FEVj response to  O3 remains unexplained.
22
23
24      AX6.5   INFLUENCE OF AGE, GENDER, ETHNIC, ENVIRONMENTAL
25               AND OTHER FACTORS
26      AX6.5.1   Influence of Age
27           On the basis of results reported from epidemiologic studies, children and adolescents are
28      considered to be at increased risk, but not necessarily  more responsive, to ambient oxidants than
29      adults. However, findings of controlled laboratory studies that have examined the acute effects
30      of O3 on children and adolescents do not completely support this assertion (Table  AX6-5).
31      Children experience about the same decrements in spirometric endpoints as young adults

        August 2005                            AX6-44      DRAFT-DO NOT QUOTE OR CITE

-------
                             Table AX6-5. Age Differences in Pulmonary Function Responses to Ozonea
to
o
o
X
H

6
o


o
H

O

O
H
W

O


O
HH
H
W
Ozone
Concentration1"
ppm
0.40


0.42


0.0
0.40



0.0
0.24



0.0
0.12
0.18
0.24
0.30
0.40

0.0
0.12
0.18
0.24
0.30
0.40



ug/m3
784


823


0
784



0
470



0
235
353
471
589
784

0
235
353
471
589
784



Exposure Duration
and Activity
2 h IE (15' ex/1 5' rest)
VE - 33-45 L/min
(47 subjects only)
1. 5 h IE (20' ex/10' rest)
VE * 33-45 L/min
(All subjects)
2 h, IE (15' ex/1 5' rest)
VE- 18 L/min/m2 BSA
2 exposures: 25% of subj .
exposed to air-air,
75% exposed to O3-O3
4 h, IE (15' ex/1 5' rest)
VE = 20 L/min



2 h rest or IE
(4x15 min
at VE = 25 or 35
L/min/m2 BSA)



2. 33 h IE
(4x15 min
atVE = 25
L/min/m2 BSA)





Exposure
Conditions
-22 °C
40% RH
treadmill



21 °C
40% RH
treadmill


24 °C
40% RH



22 °C
40% RH





22 °C
40% RH







Number and
Gender of
Subjects
146 M
94 F




28 M
34 F



10 M

9M


485 WM (each
subject
exposed at one
activity level
to one O3
concentration)

371 (WM,
BM, WF, BF;
-25% per
group) each
subject
exposed to one
03
concentration

Subject
Characteristics
Healthy NS
18 to 60 years old




Healthy NS
18 to 57 years old
Healthy NS
18 to 59 years old

Healthy NS
60 to 69 years old
COPD
59 to 71 years old

Healthy NS
18 to 36 years old
mean age 24 years




Healthy NS
18 to 35 years old
mean age 24 years






Observed Effect(s)
Young individuals of both gender (<35 years)
significantly more responsive than older
subjects. Strong responses are less common
over the age of 35 years, especially in women.
The variability of an individual's responsiveness
to repeated exposures to O3 decreases with age.
Significant decrements in spirometric lung
function in all groups. Young males and females
(<35 years) were significantly more responsive
than older individuals (>35 years).

Healthy: small, 3.3%, decline in FEVj (p = 0.03 [not
reported in paper], paired-t on O3 versus FA pre-post
FEVj). COPD: 8% decline in FEVj (p = ns, O3 versus
FA). Adjusted for exercise, ozone effects did not differ
significantly between COPD patients and healthy
subjects.
Statistical analysis of 8 experimental chamber
studies conducted between 1980 and 1993 by the
U.S. EPA in Chapel Hill, NC. O3-induced
decrement in FEV[ predicted to decrease with
age. FEV; response of a 30 year old predicted to
be 50% the response of a 20 year old. Also see
Table 6-1
Statistical analysis of experimental data
collected between 1983 and 1990 by the U.S.
EPA in Chapel Hill, NC. Cyinduced decrement
in FEV; predicted to decrease with age. FEV;
response of a 30 year old predicted to be 65%
the response of a 20 year old. No effect of
menstrual cycle phase on FEV; response.
Inconsistent effect of social economic status on
FEV[ response.
Reference
Hazucha et al.
(2003)




Passannante
etal. (1998)



Gong et al.
(1997a)



McDonnell
etal. (1997)





Seal et al.
(1996)








-------
I
S-
to
o
o














X
_k
ON




O
>
H
6
o
2|
O
H
O
O
H
W
O
O
H
W
Table AX6-5 (cont'd). Age Differences
Ozone
Concentration1"

ppm
0.18
0.24
0.30
0.40
0.45



0.45




0.45





0.12


0.20
0.30




0.113C +
other
ambient
pollutants




ug/m3
353
470
588
784
882



882




882





235


392
588




221







Exposure Duration
and Activity
2.33h
IE
VE = 20 L/min/m2 BSA

Ih, CE
VE - 26 L/min
2h,IE
VE - 26 L/min
2 h, IE (20' ex/20' rest)
Male: VE = 28.5L/min
Female: VE = 26.1 L/min


2 h, IE (20' ex/20' rest)
VE - 26 L/min




1 h IE (mouthpiece)
VE = 4 to 5 x resting

1 h (mouthpiece)
50' rest/10' ex for first 7
males, 20' rest/10' ex for
remaining subjects
Male: VE - 29 L/min
Female: VE - 23 L/min
1 h CE (bicycle)
VE - 22 L/min





Number and
Exposure Gender of
Conditions Subjects
NA 48 WF, 55 BF



-23 °C 7M
58% RH 5 F
cycle/treadmill

23 °C 10 M,
46% RH 6 F
cycle/treadmill


-24 °C 8 M
63% RH
cycle
8F


22 °C 5 M, 7 F
75% RH
treadmill
-22 °C 9M, 10 F
>75% RH
treadmill



32.7 °C 33 M, 33 F
-43% RH
cycle




in Pulmonary Function Responses to Ozone"

Subject
Characteristics
Healthy NS,
18 to 35 years old,
black and white

Healthy NS,
60 to 79 years old
(all in 60s except
one 79 years old)
Healthy NS,
60 to 89 years old



Healthy NS,
51 to 69 years old

Healthy NS,
56 to 76 years old

Healthy NS,
12 to 17 years old

Healthy NS,
55 to 74 years old




NS for both groups,
mean age =
9.4 years old






Observed Effect(s)
Older women had smaller changes in FEV[ than
younger women. No age- related differences in
SRaw or cough score.

Comparison of 1-h CE protocol and 2-h IE
protocol indicated no difference between the
changes in pulmonary function following the
two protocols.
Mean decrement in FEV; = 5.7%; eight subjects
had a 5% or greater difference between their
response to O3 and FA, and the other eight had
less than a 5% difference between their
responses to FA and 0.45 ppm O3.
13 subjects had decrements in FEV[ on three
separate exposures to 0.45 ppm within 5% of
their mean response to the three exposures. The
other three subjects were not reproducible.
Symptom reports did not correlate well with
pulmonary function changes.
No significant changes in any pulmonary
function in healthy subjects.

No spirometic changes for either group.
Females had 13% increase in RT at 3 and
22 min after 0.30-ppm exposure.



No differences in responses of boys and girls.
Similar decrements (<5% on average)
following both purified air and ambient air (O3
at 0. 1 1 ppm) exposures.





Reference
Seal et al.
(1993)


Drechsler-
Parks et al.
(1990)

Bedi et al.
(1989)



Bedi et al.
(1988)




Koenig et al.
(1988)

Reisenauer
etal. (1988)




Avol et al.
(1987)






-------

-------
 1      exposed to comparable O3 doses (McDonnell et al., 1985b; Avol et al., 1987).  In contrast to
 2      young adults, however, they had no symptomatic response, which may put them at an increased
 3      risk for continued exposure. Similarly, young adults (Linn et al., 1986; Avol et al., 1984) have
 4      shown comparable spirometric function response when exposed to low O3 dose under similar
 5      conditions.  Among adults, however, it has been repeatedly demonstrated that older individuals
 6      respond to O3 inhalation with less intense lung function changes than younger adults. Thus,
 7      children, adolescents, and young adults appear to be about equally responsive to O3, but more
 8      responsive than middle-aged and older adults when exposed to a comparable dose of O3 (U.S.
 9      Environmental Protection Agency, 1996).
10           Gong et al. (1997a) studied ten healthy men (60 to 69 years old) and nine COPD patients
11      (59 to 71 years old) from the Los Angeles area who were exposed to 0.24 ppm O3 while
12      intermittently exercising every 15 min at a light load (-20 L/min) for 4 h.  Healthy subjects
13      showed a small but significant O3-induced FEVj decrement of 3.3% (p = 0.03 [not reported in
14      paper] paired-t on O3 versus FA pre-post FEVj)2. Small but statistically nonsignificant changes
15      were also observed for respiratory symptoms, airway resistance and arterial O2 saturation. In the
16      COPD patients, there was an 8% FEVj decrement due to O3 exposure which was not
17      significantly different from the response in the healthy subjects. The authors have concluded
18      that typical ambient concentrations of O3 are unlikely to induce "a clinically significant acute
19      lung dysfunction" in exposed older men. However, they also acknowledged that the "worst
20      case" scenario of O3 exposure used in their study causes acute spirometric responses.
21           Although Gong et al.  (1997a) and others (see Table 6-5) have examined responses to O3
22      exposure in subjects of various ages, the exposure conditions differ between most studies
23      so that age effects remain uncertain.  Three recent studies, which analyzed large data sets
24      (>240 subjects) of similarly exposed subjects, show clearly discernable changes in FEVj
25      responses to O3 as a function of age.
26           Seal et al. (1996) analyzed O3-induced spirometric responses in 371 young nonsmokers
27      (18 to 35 years of age). The subject population was approximately 25% white males, 25% white
28      females, 25% black males,  and 25% black females.  Each subject was exposed once to 0.0, 0.12,
              2Personal communication from authors, correction to Table 2 in Gong et al. (1997a), the %FEVl change at
        the end of the ozone exposure for subject ID 2195 should read 4.9 and not the published value of -4.3, the mean and
        standard deviation reported in the table are correct.

        August 2005                             AX6-48      DRAFT-DO NOT QUOTE OR CITE

-------
 1      0.18, 0.24, 0.30, or 0.40 ppm ozone for 2.3 h during IE at a VE of 25 L/min/m2 BSA. A logistic
 2      function was used to model and test the significance of age, socioeconomic status (SES), and
 3      menstrual cycle phase as predictors of FEVj response to O3 exposure.  Menstrual cycle phase
 4      was not a significant. SES was inconsistent with the greatest response observed in the medium
 5      SES and the lowest response in high SES.  FEVj responses decreased with subject age. On
 6      average, regardless of the O3 concentration, the response of 25, 30, and 35 year old individuals
 7      are predicted to be 83, 65, and 48% (respectively) of the response in 20 year olds. For example,
 8      in 20 year old exposed to 0.12 ppm ozone (2.3 h IE, VE = 25 L/min/m2 BSA) a 5.4% decrement
 9      in FEVj is predicted, whereas, a similarly exposed 35 yr old is only predicted to have a 2.6%
10      decrement. The Seal et al. (1996) model is limited to predicting FEVj responses immediately
11      postexposure in individuals exposed for 2.3 h during IE at a VE of 25 L/min/m2 BSA.
12          McDonnell et al. (1997) examined FEVj responses in 485 healthy white males (18 to
13      36 years of age) exposed once for 2 h to an O3 concentration of 0.0, 0.12, 0.18, 0.24, 0.30, or
14      0.40 ppm at rest or one of two levels of IE (VE of 25 and 35 L/min/m2 BSA). FEVj was
15      measured preexposure, after 1 h of exposure, and immediately postexposure.  Decrements
16      in FEVj were modeled by sigmoid-shaped curve as a function of subject age, O3 concentration,
17      VE, and duration of exposure. Regardless of the O3 concentration or duration of exposure, the
18      average responses of 25, 30, and 35 year old individuals are predicted to be 69, 48, and 33%
19      (respectively) of the response in 20 year olds.  The McDonnell et al. (1997) model is best suited
20      to predicting FEVj responses in while males exposed to O3 for 2 h or less under IE conditions.
21          Hazucha et al. (2003) analyzed the distribution of O3 responsiveness in  subjects (146 M,
22      94 F) between 18 and 60 years of age. Subjects were exposed to 0.42 ppm O3 for 1.5 h with IE
23      at VE = 20 L/min/m2 BSA. Figure AX6-7 illustrates FEVj responses to O3 exposure as a
24      function of subject age. Consistent with the discussion in Section 6.4, a large degree of
25      intersubject variability is evident in Figure AX6-7. Across all ages, 18% of subjects were weak
26      responders (<5% FEVj decrement), 39% were moderate responders, and 43% were strong
27      responders (> 15% FEVj decrement).  Younger subjects (<35 years of age) were predominately
28      strong responders, whereas,  older subjects (>35 years of age) were mainly weak responders.
29      In males, the FEVj responses of 25, 35, and 50 year olds are predicted to be 94, 83,  and 50%
30      (respectively) of the average response in 20 year olds.  In females, the FEVj responses of 25, 35,

        August 2005                            AX6-49     DRAFT-DO NOT QUOTE OR CITE

-------
         in
         03
         J2
         LLI
    110-
    100-
     90-
     80-
     70-
     60-
     50-
     40-
            30-I
              15   20   25   30   35  40  45  50  55  60
                           Age (years)
                                                      LLI
110-
100-


 80-
 70-
 60-
 50-
 40-
                                                  30-I
                                                                     °^ J3-6*
                                                           15   20   25   30  35  40  45  50  55   60
                                                                 Age (years)
       Figure AX6-7.  Effect of O3 exposure (0.42 ppm for 1.5 h with IE) on FEVt as a function of
                       subject age. Left panel data for males (n = 146; 19 to 60 yrs old), right
                       panel data for females (n = 94; 18 to 59 yrs old).
       Source: Adapted from Hazucha et al. (2003).
 1
 2
 3
 4
 5
 9
10
11
12
13
14
15
16
and 50 year olds are predicted to be 82, 46, and 18% (respectively) of the average response in
20 year olds.  The Hazucha et al. (1996) model is limited to predicting FEVj responses
immediately postexposure in individuals exposed to 0.42 ppm O3 for 1.5 h during IE at a VE of
20 L/min/m2 BSA.
     The pathophysiologic mechanisms behind the pronounced age-dependent, gender-
differential rate of loss of O3 responsiveness are unclear.  Passannante et al. (1998) have
previously demonstrated that O3-induced spirometric decrements (FEVj) in healthy young and
middle-aged adults are principally neural in origin, involving opioid-modulated sensory
bronchial C-fibers. (The methodological details of this study are presented in Section AX6.2.3 of
this chapter.) The peripheral afferents are most likely the primary site of action, which would be
compatible with a reflex action as well as a cortical mechanism. The pattern of progressive
decline, as well as the subsequent rate of recovery of spirometric lung function, suggest
involvement of both direct and indirect (possibly by PGE2(X) stimulation and/or sensitization of
vagal sensory fibers.  (For details, see Section AX6.2.3.1  of this chapter)
     The additional pulmonary function data published since the release of last O3 criteria
document (U.S. Environmental Protection Agency, 1996) and reviewed in this section reinforce
       August 2005
                                         AX6-50
     DRAFT-DO NOT QUOTE OR CITE

-------
 1     the conclusions reached in that document.  Children and adolescents are not more responsive
 2     to O3 than young adults when exposed under controlled laboratory conditions. However, they
 3     are more responsive than middle-aged and older individuals.  Young individuals between the age
 4     of 18 and 25 years appear to be the most sensitive to O3. With progressing age, the sensitivity
 5     to O3 declines and at an older age (>60 yrs) appears to be minimal except for some very
 6     responsive individuals. Endpoints other than FEVj may show a different age-related pattern
 7     of responsiveness.
 8
 9     AX6.5.2  Gender and Hormonal Influences
10          The few late 1970 and early 1980 studies specifically designed to determine symptomatic
11     and lung function responses of females to O3 were inconsistent. Some studies have concluded
12     that females might be more sensitive to O3 than males, while others found no gender differences
13     (U.S. Environmental Protection Agency, 1996). During the subsequent decade, seven studies
14     designed to systematically explore gender-based differences in lung function following O3
15     exposure were completed (Table AX6-6).  Protocols included mouthpiece and chamber
16     exposures, young and old individuals, normalization of ventilation to BSA or FVC, continuous
17     and intermittent exercise,  control for menstrual cycle phase, and the use of equivalent effective
18     dose of O3 during exposures. These studies have generally reported no statistically significant
19     differences in pulmonary function between males and females (Adams et al., 1987; Drechsler-
20     Parks et al.,  1987a; Messineo and Adams, 1990; Seal et al., 1993; Weinmann et al., 1995)
21     although in some  studies females  appeared to experience a slightly greater decline then males
22     (Drechsler-Parks et al., 1987a; Messineo and Adams, 1990). The comparative evaluations were
23     based on responses that included spirometry, airway resistance, nonspecific bronchial
24     responsiveness  (NSBR) determinations, and changes in frequency and severity of respiratory
25     symptoms. However, depending on how the O3 dose was calculated and normalized, the
26     findings of at least three studies may be interpreted as showing that females are more sensitive
27     to O3 than males.  The findings of the seven studies  are presented in detail in Section 7.2.1.3 of
28     the previous O3 criteria document (U.S. Environmental Protection Agency, 1996).
29          Some support for a possible increased sensitivity of females to O3 comes from a study of
30     uric acid concentration in nasal lavage fluid (NLF).  Housley et al. (1996) found that the NLF of
31     females contains smaller amounts of uric acid than the NLF of males.  The primary source of

       August 2005                            AX6-51      DRAFT-DO NOT QUOTE OR CITE

-------
to
o
o
                    Table AX6-6. Gender and Hormonal Differences in Pulmonary Function Responses to Ozone"
Ozone
Concentration11
ppm ug/m3 and Activity Conditions'
0.0 490 1 h CE NA
0.25 VE = 30 L/min Face mask
exposure







Number and Gender Subject
of Subjects Characteristics
32 M, 28 F Healthy NS
22.6 ±0.6 years old








Observed Effect(s)
Mean O3-induced FEVj decrements of 15.9%
in males and 9.4% in females (gender
differences not significant). FEV; decrements
ranged from -4 to 56%; decrements >15% in
20 subjects and >40% in 4 subjects. Uptake
of O3 greater in males than females, but
uptake not correlated with spirometric
responses.


Reference
Ultman et al.
(2004)






X
ON

t^ft
to
H

6
o


o
H

O

o
H
W

O


O
HH
H
W
0.40

0.42


0.0
0.35





0.0
0.4



0
0.12
0.24
0.30
0.40

784

823


0
686





0
784



0
235
470
588
784

2 h, IE (15' ex/1 5' rest)
VE = 33-45 L/min
1. 5 h IE (20' ex/10' rest)
VE = 33-45 L/min

1.25h, IE
(30' ex/1 5' rest/30' ex)
VE = 40 L/min




2 h, IE (15' ex/1 5' rest)
VE» 18L/min/m2BSA
2 exposures: 25% of
subj. exposed to air-air,
75% exposed to O3-O3
2.33 h IE (15' ex/15' rest)
VE = 20 L/min/m2 BSA
one exposure per subject



22 °C
40% RH
treadmill


22 °C
40% RH
treadmill




21 °C
40% RH
treadmill


22 °C
40% RH
treadmill



146 M Healthy NS, No significant gender differences in FEVj
94 F 1 8 to 60 years old among young (<3 5 years) and older
individuals. Strong responses are less
common over the age of 35 years, especially
in women.
19 F O3 responders FVC and FIVC changes about the same,
22.1 ± 2.7 years old - 13%, FEV; -20%. Increased airway
responsiveness to methacholine. Persistence
of small effects on both inspired and expired
spirometry past 18 h. Chemoreceptors not
activated but ventilatory drive was
accelerated.
28 M Healthy NS, 20-59 years old Significant decrements in spirometric lung
34 F function. No significant differences in FEV;
between young females and males and older
females and males either in responders or
nonresponders subgroups.
48WF, 55BF Healthy NS, Significant menstrual cycle phase * race
18 to 35 years old interaction for FEVj. No significant
menstrual cycle phase effect when blacks and
whites were analyzed separately.
No significant menstrual phase effects for
SRaw or cough score.
Hazucha
et al. (2003)



Folinsbee
and Hazucha
(2000)




Passannante
etal. (1998)



Seal et al.
(1996)





-------
                Table AX6-6 (cont'd). Gender and Hormonal Differences in Pulmonary Function Responses to Ozone"
«i Ozone
^j Concentration11

(^ ppm ug/m3
0.0 0
0.35 686




Exposure Duration
and Activity
2.15 h, IE
(30' ex/30' rest)




Exposure Number and Gende" tx~l ^
Conditions' of Subjects
19-24 °C 12 M
48-55% RH 12 F
treadmill


I 13UUJCCI
Characteristics
Healthy NS,
5 F follicular and 7 luteal
phase exposure,
regular menstrual cycles, 18
to 35 years old

Observed Effect(s)
Changes in FVC, FEV1; FEF25.75, Vma


X50%, and
Vrmx25% were similar during both the follicular

Reference
Weinmann
etal. (1995)
and luteal phases. No significant difference
between males and females.





X
H

6
o


o
H

O

O
H
W

O


O
HH
H
W
0.3





0
0.12
0.18
0.24
0.30
0.40
0
0.18
0.30


0.0
0.45



0
0.20
0.30


588





0
235
353
470
588
784
0
353
588


0
882



0
392
588


IhCE
VE = 50 L/min




2.33 h (15' ex/1 5' rest)
VE = 25 L/min/m2 BSA
(one exposure/subject)



1 h (mouthpiece), CE
VE ~ 47 L/min
exposures >4 days apart


2 h, IE (20' ex/20' rest)
VE = 28.5 L/min for M
VE = 26.1 L/min for F
repeated O3 exposures

1 h (mouthpiece) IE
(50' rest/10' ex first 7 M)
(20' rest/ 10' ex all others)
VE ~ 28 L/min for M
VE ~ 23 L/min for F
NA





22 °C
40% RH
treadmill



21 to 25 °C
45 to 60% RH
cycle


23.1 °C
46.1%RH
cycle/treadmill


=22 °C
>75% RH
treadmill


9F





30 to 33 F and 30 to
33 M in each
concentration group;
total of
372 individuals
participated
14 F


14 F

10M


6F

9M, 10 F




Healthy NS, regular
menstrual cycles,
20 to 34 years old



Healthy NS, 18 to 35 years
old, blacks and whites




FVC = 5. 11 ±0.53L,
NS, 20 to 24 years old

FVC = 3.74 ± 0.30 L,
NS, 19 to 23 years old
Healthy NS,
60 to 89 years old

Healthy NS,
64 to 71 years old
Healthy NS,
55 to 74 years old



FEV; decreased 13.1% during the mid-luteal
phase and 18.1% during the follicular phase.
Decrement in FEF25.75 was significantly larger
during the follicular phase than the mid-luteal
phase. Changes in FVC were similar in
both phases.
Decrements in FEV1; increases in SRaw and
cough, correlated with O3 concentration.
There were no significant differences between
the responses of males and females.


Small lung group, FVC = 3.74 ± 0.30 L.
Large lung group, FVC = 5.1 1 ± 0.53 L.
Significant concentration-response effect on
FVC and FEVjj lung size had no effect on
percentage decrements in FVC or FEVj.
Mean decrement in FEVj = 5.7%.
Decrements in FVC and FEVj were the only
pulmonary functions significantly altered by
O3 exposure. No significant differences
between responses of men and women.
No change in any spirometic measure for
either group. Females had 13% increase in RT
after 0.30-ppm exposure. Gender differences
not evaluated.

Fox etal.
(1993)




Seal et al.
(1993)




Messineo
and Adams
(1990)


Bedi et al.
(1989)



Reisenauer
etal. (1988)




-------
6-
fin
c
^
OJ
to
o















ON
Table AX6-6 (conf

Ozone
Concentration11

ppm ug/m3 and Activity
0.3 588 1 h (mouthpiece), CE
VE ~ 70 L/min for men
VE ~ 50 L/min for
women
0.0 0 2 h, IE (20' rest/ 20' ex)
0.45 882 VE» 27.9 L/min for M
VE = 25 .4 L/min for F
repeated O3 exposures

0.48 941 2h, IE
VE = 25 L/min




d). Gender



Exposure
Conditions'
21 to 25 °C
45 to 60% RH
cycle

24 °C
58% RH
cycle


21 °C
(WBGT)
cycle



and Hormonal Differences in Pulmonary



Number and Gender Subject
of Subjects Characteristics
20 M NS, 18 to 30 years old

20 F NS, 19 to 25 years old

8M Healthy NS,
5 1 to 69 years old

8F Healthy NS,
56 to 76 years old
10 F Healthy NS,
19 to 36 years old




Function Responses to Ozone"




Observed Effect(s)
Significant decrements in FVC, FEVj, and
FEF25.75 following O3 exposure.
No significant differences between men and
women for spirometry or SRaw.
Range of responses in FEVj:
0 to - 12% (mean = -5.6%).
No significant difference in responses of men
and women.
Tendency for women to have greater effects.
Mean decrement in FEV; = 22.4%.
Significant decrements in all spirometric
measurements. Results not significantly
different from a similar study on males
(Drechsler-Parks et al., 1984).






Reference
Adams et al.
(1987)


Drechsler-
Parks et al.
(1987a,b)


Horvath
etal. (1986)




          ' See Appendix A for abbreviations and acronyms.

          b Listed from lowest to highest O3 concentration.

          'WBGT = 0.7 Twetbulb + 0.3 T^^.
 H

 6
 o


 o
 H

O

 O
 H
 W

 O
O
HH
H
W

-------
 1     uric acid is plasma; therefore, lower nasal concentrations would reflect lower plasma
 2     concentrations of this antioxidant. The authors have speculated that in females, both lower
 3     plasma and NLF levels (of uric acid) can plausibly make them more susceptible to oxidant
 4     injury, since local antioxidant protection may not be as effective as with higher levels of uric
 5     acid, and consequently more free O3 can penetrate deeper into the lung.
 6           Several studies also have suggested that anatomical differences in the lung size and the
 7     airways between males and females, and subsequent differences in  O3 distribution and
 8     absorption, may influence O3 sensitivity and potentially differential O3 response.  The study of
 9     Messinio and Adams (1990) have, however, convincingly demonstrated that the  effective dose to
10     the lung, and not the lung size,  determines the magnitude of (FEVj) response.  Furthermore,
11     the O3 dosimetry experiments of Bush et al. (1996) have shown that despite gender differences in
12     longitudinal distribution of O3,  the absorption distribution in conducting airways was the same
13     for both sexes when expressed as a ratio of penetration to anatomic dead space volume. This
14     implies that gender differences, if any, are not due to differences in (normal) lung anatomy.
15     The data also have shown that routine adjustment of O3 dose for body size and gender
16     differences would be more important if normalized to anatomic dead space rather than the usual
17     FVCorBSA.
18           One of the secondary objectives of a study designed to examine the role of neural
19     mechanisms involved in limiting maximal inspiration following O3 exposure has been to
20     determine if gender differences occur. A group of healthy males (n = 28) and females (n = 34)
21     were exposed to 0.42 ppm O3 for 2 h with IE. The methodological  details of the study are
22     presented in Section AX6.2.5.1 of this document. As Figure AX6-4 shows, the differences
23     between males and females were, at any condition, measurement point, and O3 sensitivity status
24     only minimal and not significant (Passannante et al., 1998).
25           In another investigation, Folinsbee and Hazucha (2000) exposed  a group of
26     19 O3-responsive young females (average age of 22 years, prescreened for O3 responsiveness by
27     earlier exposure) to air and 0.35 ppm O3. The randomized 75-min exposures included two
28     30-min exercise periods at a VE of 40 L/min. In addition to standard pulmonary function tests,
29     they employed several techniques used for the first time in human air pollution studies
30     assessment of O3 effects. The average lung function decline from a pre-exposure value was 13%
31     for FVC, 19.9 % for FEVj, and 30% for FEF25.75.  The infrequently measured  forced inspiratory

       August 2005                            AX6-55      DRAFT-DO NOT QUOTE OR CITE

-------
 1      vital capacity (FIVC) was the same as FVC suggesting that the lung volume limiting
 2      mechanisms are the same. The reduction in peak inspiratory flow (PIF) most likely reflects an
 3      overall reduction in inspiratory effort associated with neurally mediated inhibition of inspiration.
 4      Persistence of small inspiratory and expiratory spirometric effects, airway resistance, and airway
 5      responsiveness to methacholine for up to 18 h postexposure suggests that recovery of pulmonary
 6      function after O3 exposure involves more than the simple removal of an irritant. Incomplete
 7      repair of damaged epithelium and still unresolved airway inflammation are the likely causes of
 8      the residual effects that in some individuals persisted beyond 24 h postexposure.  However, by
 9      42 hours no residual effects were detected. No significant changes were found in ventilatory
10      response to CO2 between air and O3 exposures, suggesting that chemoreceptors were not affected
11      by O3. However, O3 inhalation did result in accelerated timing of breathing and a modest
12      increase in inspiratory drive.  These observations are consistent with, and further supportive of,
13      the primary mechanisms of O3-induced reduction in inspiratory lung function, namely an
14      inhibition of inspiration elicited by stimulation of the C-fibers and other pulmonary receptors.
15      Because the measures of inspiratory and chemical drive to assess O3 effects were not reported in
16      any previous human study, no comparisons are possible. Because no male subjects were
17      recruited for the study, it is not possible to compare gender effects.  Despite being O3-responsive,
18      however, the average post-O3 decline in expiratory lung function from preexposure (13% for
19      FVC;  19.9% for FEV^ 30% for FEF25.75) was similar to that seen in female cohorts studied by
20      other investigators under similar conditions of exposure. These were the  same studies that found
21      no gender differences in O3 sensitivity (Adams et al., 1987; Messineo and Adams, 1990).
22           The study by Hazucha et al. (2003),  discussed in the previous section, has in addition to
23      aging also examined gender differences in O3 responsiveness. The male (n = 146) and female
24      (n = 94) cohorts were classified into young (19 to 35 year-old) and middle-aged (35 to 60 year-
25      old) groups. This classification was selected in order to facilitate comparison with data reported
26      previously by other laboratories.  Using a  linear regression spline model (with a break point at
27      35 years), the authors reported that the rate of loss of sensitivity is about three times as high in
28      young females  as in young males (p < 0.003). In young females, the average estimated  decline
29      in FEVj response is 0.71% per year, while in young males it is 0.19% per year. Middle-aged
30      groups of both  genders show about the same rate of decline (0.36 to 0.39%, respectively).
31      At 60 years of age, the model estimates about a 5% post-O3 exposure decline in FEVj for males,

        August 2005                            AX6-56      DRAFT-DO NOT QUOTE  OR CITE

-------
 1      but only a 1.3% decline for females. These observations suggest that young females lose O3
 2      sensitivity faster than young males, but by middle age, the rate is about the same for both
 3      genders. Descriptive statistics show that there were practically no differences in the mean value,
 4      standard error of the mean, and coefficient of variation for % FEVj decrement between the group
 5      of young males (n = 125; 83.7 ± 1.1%; CV = 13.5%) and young  females (n = 73; 83.4 ± 1.25%;
 6      CV = 12.8%). A straight linear regression model of these data was illustrated in Figure AX6-7.
 7      The slopes, significant in both males (r = 0.242; p = 0.003) and females (r = 0.488; p = 0.001),
 8      represent the decline in responsiveness of 0.29% and 0.55% per  year respectively, as assessed
 9      by FEVi.
10           Two earlier studies of the effects of the menstrual cycle phase on O3 responsiveness have
11      reported conflicting results (U.S. Environmental Protection Agency,  1996). Weinmann et al.
12      (1995) found no significant lung function effects related to menstrual cycle, although during the
13      luteal phase the effects were slightly more pronounced than during the follicular phase; while
14      Fox et al., (1993) reported that follicular phase enhanced O3 responsiveness. In a more recent
15      investigation of possible modulatory effects of hormonal changes during menstrual cycle on O3
16      response, young women (n = 150) 18 to 35 years old were exposed once to one of multiple O3
17      concentrations (0.0, 0.12, 0.18, 0.24, 0.30, 0.40 ppm) for 140 min with IE at 35 L/min/m2 BSA.
18      The women's menstrual cycle phase was determined immediately prior to O3 exposure. Post-O3,
19      no significant differences in % predicted FEVj changes that could be related to the menstrual
20      cycle phase were found. Admittedly, a less precise method of determining menstrual cycle
21      phase used in this study could have weakened the statistical power. Unfortunately, the direction
22      and magnitude of O3 response as related to the menstrual cycle phases were not reported (Seal
23      et al., 1996).  Considering the inconclusiveness of findings of this study and the inconsistency of
24      results between the two earlier studies, it is not possible to make any firm conclusions about the
25      influence of the menstrual cycle on responses to O3 exposure.
26           Additional studies presented in this section clarify an open-ended conclusion reached in the
27      previous O3 criteria document (U.S. Environmental Protection Agency, 1996) regarding the
28      influence of age on O3  responsiveness. Healthy young males and females are about equally
29      responsive to O3, although the rate of loss of sensitivity is higher in females than in males.
30      Middle-aged men and women are generally much less responsive to O3 than younger individuals.
31      Within this range, males appear to be slightly more responsive than females, but the rate of age-

        August 2005                             AX6-57     DRAFT-DO NOT QUOTE OR CITE

-------
 1      related loss in FEVj is about the same. The O3 sensitivity may vary during the menstrual cycle;
 2      however, this variability appears to be minimal.
 3
 4      AX6.5.3  Racial, Ethnic, and Socioeconomic Status Factors
 5           In the only laboratory study designed to compare spirometric responses of whites and
 6      blacks exposed to a range of O3 concentrations (0 to 0.4 ppm), Seal et al. (1993) reported
 7      inconsistent and statistically insignificant FEVj differences between white and black males and
 8      females within various exposure levels. Perhaps, with larger cohorts the tendency for greater
 9      responses of black than white males may become significant. Thus, based on this study it is still
10      unclear if race is a modifier of O3 sensitivity, although the findings of epidemiologic studies
11      reported in the previous criteria document "can be considered suggestive of an ethnic difference"
12      (U.S. Environmental Protection Agency, 1996).  However, as Gwynn and Thurston (2001)
13      pointed out, it appears that it is more the socioeconomic status (SES) and overall  quality of
14      healthcare that drives PM10- and O3-related hospital admissions than an innate or  acquired
15      sensitivity to pollutants.
16           This assertion is somewhat supported by the study of Seal et al. (1996) who employed a
17      family history questionnaire to examine the influence of SES on the O3 responsiveness of
18      352 healthy,  18- to 35-year-old black and white subjects. Each subject was exposed once under
19      controlled laboratory conditions to either air or 0.12, 0.18, 0.24, 0.30, 0.40 ppm O3 for 140 min
20      with 15 min IE at 35  L/min/m2 BSA.  An answer to the "Education of the father"  question was
21      selected  as a surrogate variable for SES status. No other qualifying indices of SES were used or
22      potential bial examined. Of the three SES categories, individuals in the middle SES category
23      showed greater concentration-dependent decline in % predicted FEVj (4-5% @ 0.4 ppm O3) than
24      low and high SES groups.  The authors did not have an "immediately clear" explanation for this
25      finding.  The SES to  %predicted FEVj relationship by gender-race group was apparently
26      examined as well; however, these results were not presented. Perhaps a more comprehensive
27      and quantitative evaluation of SES status would have identified the key factors and clarified the
28      interpretation of these findings.  With such a paucity of data it is not possible to discern the
29      influence of racial  or other related factors on O3 sensitivity.
30
        August 2005                            AX6-58      DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX6.5.4  Influence of Physical Activity
 2           Apart from the importance of increased minute ventilation on the inhaled dose of O3 during
 3      increased physical activity, including work, recreational exercise, and more structured exercise
 4      like sports, no systematic effort has been made to study other potential physical factors that may
 5      modulate O3 response. The typical physiologic response of the body to exercise is to increase
 6      both the rate and depth of breathing, as well as increase other responses such as heart rate, blood
 7      pressure, oxygen uptake, and lung diffusion capacity.
 8           Physical activity increases minute ventilation in proportion to work load.  At rest, and
 9      during light exercise, the dominant route of breathing is through the nose. The nose not only
10      humidifies air, among other physiologic functions, but also absorbs O3 thus decreasing the
11      overall  dose. As the intensity of exercise increases, the minute ventilation increases and the
12      breathing switches from nasal to oronasal mode.  There is considerable individual variation in
13      the onset of oronasal breathing, which ranges from 24 to 46 L/min (Niinimaa et al., 1980).
14      During heavy exercise, ventilation is dominated by oral breathing.  Consequently, the residence
15      time of inhaled air in the nose and the airways is shorter, reducing the uptake of O3 (Kabel et al.,
16      1994).  Moreover, increasing inspiratory flow and tidal volume shifts the longitudinal
17      distribution of O3 to the peripheral airways, which are more sensitive to injury than the larger,
18      proximal airways. Ozone uptake studies of human lung  showed that at simulated quiet
19      breathing, 50% of O3 was absorbed in the upper airways, 50% in the conducting airways, and
20      none reached the small airways (Hu et al. 1994).  With ventilation simulating heavy exercise
21      (60 L/min), the respective O3 uptakes were 10% (upper airways), 65% (conducting airways), and
22      25% (small airways). These observations imply that equal O3 dose (C x T x VE) will have a
23      greater effect on pulmonary function and inflammatory responses when inhaled during heavy
24      physical activity than when inhaled during lighter activity. Although, Ultman  et al. (2004)
25      recently reported that spirometric response are not correlated with O3 uptake. (See Chapter 4 of
26      this document for more information on the  dosimetry ofO3.)
27           Other physiologic factors activated in response to physical activity are unlikely to have as
28      much impact on O3 responsiveness as does minute ventilation; however, their potential influence
29      has not been investigated.
30
        August 2005                             AX6-59      DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX6.5.5  Environmental Factors
 2           Since the 1996 O3 criteria document, not a single human laboratory study has examined the
 3     potential influence of environmental factors such as rural versus urban environment, passive
 4     cigarette smoke exposure, and bioactive admixtures such as endotoxin on healthy individual's
 5     pulmonary function changes due to O3 (U.S. Environmental Protection Agency, 1996).
 6           Some of the unresolved issues, e.g., health effects of ETS and O3 interaction, which need to
 7     be examined in human studies were explored very recently in laboratory animal studies (see
 8     Chapter 5 for more details). In one study on mice, preexposure of animals to sidestream
 9     cigarette smoke (ETS surrogate), which elicited no immediate effects, resulted in a potentiation
10     of subsequent  O3-induced inflammatory response.  This finding suggests that typical adverse
11     effects of ETS do not necessarily have to elicit an immediate response to ETS, but may in fact
12     potentiate the effects of a subsequent exposure to another pollutant like O3 (Yu et al., 2002). The
13     key mechanism by which smoke inhalation may potentiate subsequent oxidant injury appears to
14     be damage to cell membranes and the resulting increase in epithelial permeability. Disruption of
15     this protective layer may facilitate as well as accelerate injury to subepithelial structures when
16     subsequently exposed to other pollutants (Bhalla, 2002). Although this may be a plausible
17     mechanism in  nonsmokers and acute smokers exposed to ETS and other pollutants, studies
18     involving chronic smokers who most likely already have chronic airway inflammation do not
19     seems to show exaggerated response with exposure to O3.
20           More than 25 years ago, Hazucha et al. (1973) reported that the spirometric lung function
21     of smokers declined significantly less than that of nonsmokers when exposed to 0.37 ppm O3.
22     The findings of this study have been confirmed and expanded (Table AX6-7). Frampton et al.
23     (1997a) found that exposure of current smokers (n = 34) and never smokers (n = 56) to
24     0.22 ppm O3 for 4 h with IE for 20 min of each 30 min period at 40 to 46 L/min, induced a
25     substantially smaller decline in FVC, FEVj and SGaw of smokers than never smokers. Smokers
26     also demonstrated a much narrower distribution of spirometric endpoints than never smokers.
27     Similarly, nonspecific airway responsiveness to methacholine was decreased in smokers.
28     However, both groups showed the consistency of response from exposure to exposure. It should
29     be noted that despite seemingly lesser response, the smokers were more symptomatic post air
30     exposure than  never smokers, but the opposite was true for O3 exposure. This would suggest
31     that underlying chronic airway inflammation present in smokers has blunted stimulation of

       August 2005                             AX6-60      DRAFT-DO NOT QUOTE OR CITE

-------
                               Table AX6-7. Influence of Ethnic, Environmental, and Other Factors
to
o
o
X
Oi
H

6
o


o
H

O

O
H
W

O


O
HH
H
W
Ozone
Concentration
ppm Hg/m3
0.0 0
0.4 780







0.0 0
0.12 235



0.0 0
0.22 431
0.22 431



0.0 0
0.12- 235-470a
0.24a





Exposure
Duration
and Activity
2h
IE, 15' ex/15' rest
VE = 20 L/min/m2
BSA





0.75 h
IE, 15' ex/15' rest
VE = 40-46 L/min


4h
IE, 20' ex/10' rest
VE = 40-46 L/min



2.17 h
IE, 10' ex/ 10' rest
VE = 36-39 L/min





Number
Exposure and Gender Subject
Conditions of Subjects Characteristics
20 °C 15 M, 1 F Placebo group:
40% RH healthy NS
avg age 27 yrs.

13 M, 2 F Antiox. Suppl.
Gr: Healthy NS
avg age 27 yrs.


60% RH 5 M, 12 F Asthmatics
sensitive to SO2
19 to 38 yrs old


21°C 25(M/F) Healthy NS
37% RH O3 responders and
nonresponders
18 to 40 yrs old


22 °C or 30 °C 5 M, 4 F Healthy NS
45-55% RH 24 to 32 yrs old






Observed Effect(s)
PF decrements in the supplementation
group were signif. smaller for FVC
(p < 0.046) and near significant for FEV{
(p < 0.055). The inflammatory response
(BAL) showed no significant differences
between the two groups either in the
recovery of cellular components or the
concentrations and types of inflammatory
cytokines.
No significant differences due to O3
between placebo and antioxidant
supplement cohort in either
spirometric responses or bronchial
hyperresponsiveness to 0.1 ppm SO2.
Glutathione peroxidase (GPx) activity and
eGPx protein level were significantly
(p = 0.0001) depleted in ELF for at least
18 h postexp. In BAL both endpoints
were elevated (ns). No association
between cell injury, PF, or GPx activity.
FEVj decreased (p < 0.5) by -8% at
22 °C and -6.5% at 30 °C. 19 h postexp
decline of 2.3% still signif. (p < 0.05).
SGaw signif. (p < 0.05) declined at 30 °C
but not at 22 °C. The BHR assessed 19 h
postexp. as PC50 sGaw methacholine
signif. (p < 0.05) higher at both
temperatures.
Reference
Samet et al.
(2001)







Trenga et al.
(2001)



Avissar et al.
(2000)b




Foster et al.
(2000)







-------
to
o
o
                                Table AX6-7 (cont'd). Influence of Ethnic, Environmental, and Other Factors
Ozone
Concentration
ppm Hg/ni3
0.0 0
0.40 780
Exposure Number
Duration Exposure and Gender Subject
and Activity Conditions of Subjects Characteristics
2 h 6 M, 9 F Healthy NS
IE, 20' ex/ 10' rest avg age 3 1 yrs.
Observed Effect(s)
Corticosteroid pretreatment had no effects
on post-O3 decline in PF, PMN response,
Reference
Nightingale
et al. (2000)
                           VE = mild to mod.
                                                                                         and sputum cell count under both the
                                                                                         placebo and treatment conditions.
                                                                                         Methacholine PC20 FEV{ was equally
                                                                                         decreased in both cond. 4 h after
                                                                                         exposure. No changes in exhaled NO
                                                                                         and CO.
X
Oi
ON
to
H
6
o

o
H
O
O
H
W
O
O
HH
H
W
0.0 0
0.22 431



0.0 0
0.22 431
0.22 431






4h 21 °C
IE 37% RH
20' ex/10' rest
VE = 40-46 L/min

4h 21 °C
IE, 20' ex/10' rest 37% RH
VE = 25 L/min/m2
BSA





90 M 56 never smokers
34 current
smokers
18 to 40 yrs. old

10 M, 2 F NS, O3 nonresp.,
avg age 25 yrs.;
10 M, 3 F NS, O3 resp.,
avg age 25 yrs;
1 1 M, 2 F smokers avg
age 28 yrs



Smokers are less responsive to O3 as
assessed by spirometric and
plethysmographic variables. Neither age,
gender, nor methacholine responsiveness
were predictive of O3 response.
Neither O3 responsiveness nor smoking
has altered the magnitude and the time
course of O3-induced airway
inflammation. Inflammation involved
all types of cells accessible by B AL.
The recovery profile of these cells over
time was very similar for all groups
showing highest values 18 h
postexposure.
Frampton et al.
(1997a,c)b



Torres et al.
(1997)b
Frampton et al.
(1997a,c)b





        aRamp exposure from 0.12 ppm to 0.24 ppm and back to 0.12 ppm at the end of exposure.
        bRelated studies, sharing of some subjects .

-------
 1     bronchial C-fibers and other pulmonary receptors, the receptors substantially responsible for
 2     post-O3 lung function decrements. In addition to desensitization, the other "protective"
 3     mechanisms active in smokers may be an increase in the mucus layer conferring not only a
 4     mechanical protection, but also acting as an O3 scavenger. Another plausible explanation of a
 5     diminished responsiveness of smokers may be related to elevated levels of reduced glutathione
 6     (GSH), an antioxidant, found in epithelial lining fluid of chronic but not acute smokers (MacNee
 7     etal., 1996).
 8          Despite some differences in a release of proinflammatory cytokines and subsequent
 9     recruitment of inflammatory cells, both smokers and nonsmokers developed airway
10     inflammation following O3 exposure.  This was demonstrated by the Torres et al. (1997) study
11     that involved exposures of about equal size cohorts of otherwise healthy young smokers,
12     nonsmoker O3 nonresponders (<5% FEVj post-O3 decrement) and nonsmoker O3 responders
13     (>15% FEVj post-O3 decrement) to air and two 0.22 ppm O3 atmospheres for 4 hours, alternating
14     20 min of moderate exercise (25 L/min/m2 BSA) with 10 min of rest. Both O3 exposures were
15     followed by nasal lavage (NL) and bronchoalveolar lavage (BAL) performed immediately post
16     one of exposures and 18 hr later following the other exposure.  Neither O3 responsiveness nor
17     smoking alters the magnitude or the time course of O3-induced airway inflammation.  The
18     overall cell recovery was lower immediately postexposure but higher, particularly in
19     nonsmokers, 18 h post-O3 exposure when compared to control (air) in all groups. Recovery of
20     lymphocytes, PMNs and AMs in both alveolar and bronchial lavage fluid showed the largest
21     increase in response to O3 in all groups, with nonsmokers showing greater relative increases than
22     smokers. Of the two cytokines, IL-6  and IL-8, IL-6 was substantially and significantly
23     (p < 0.0002) elevated immediately postexposure but returned back to control 18 h later in all
24     groups; but only nonsmokers' effects were significantly higher (p < 0.024). IL-8 showed a
25     similar pattern of response but the increase in all groups, though still significant (p < 0.0001),
26     was not as high as for IL-6. Between group differences were not significant. This inflammatory
27     response involved all types of cells present in BAL fluid and the recovery profile of these cells
28     over time was very similar for all  groups. In contrast to BAL, NL did not prove to be a reliable
29     marker of airway inflammation. The lack of association between lung function changes
30     (spirometry) and airway inflammation for all three groups confirms similar observations
       August 2005                            AX6-63      DRAFT-DO NOT QUOTE OR CITE

-------
 1      reported from other laboratories.  This divergence of mechanisms is further enhanced by an
 2      observation that a substantially different spirometric response between O3 responders and
 3      nonresponders, the airway inflammatory response of the two groups was very similar, both in
 4      terms of magnitude and pattern (Torres et al., 1997).
 5          The influence of ambient temperature on pulmonary effects induced by O3 exposure in
 6      humans has been studied infrequently under controlled laboratory conditions. Several
 7      experimental human studies published more than 20 years ago reported additive effects of heat
 8      and O3 exposure (see U.S. Environmental Protection Agency, 1986, 1996). In the study of
 9      Foster et al. (2000) 9 young (mean age 27 years) healthy subjects (4F/5M) were exposed for
10      130 min (IE 10 min @ 36 to 39 1/min) to filtered air and to ramp profile O3 at 22° and 30 °C,
11      45-55% RH.  The order of exposures was randomized. The O3 exposure started at 0.12 ppm,
12      reached the peak of 0.24 ppm mid-way through and subsequently declined to 0.12 ppm at the
13      end of exposure. Ozone inhalation decreased VT and increased fB as compared to baseline at
14      both temperatures. At the end of exposure FEVj decreased significantly  (p < 0.5) by -8% at
15      22 °C and -6.5% at 30 °C. One day (19 h) later, the decline of 2.3% from baseline was still
16      significant (p < 0.05). FVC decrements were smaller and  significant only at 22 °C immediately
17      postexposure. SGaw significantly (p < 0.05) declined at 30 °C but not at 22 °C. A day later,
18      sGaw was elevated above the baseline for all conditions.  The nonspecific bronchial
19      responsiveness (NSBR) to methacloline assessed as PC50 sGaw was significantly (p < 0.05)
20      higher one day  following O3 exposure at both temperatures but more so at 30 °C. Thus, these
21      findings indicate that elevated temperature has partially attenuated spirometric response but
22      enhanced airway reactivity.  Numerous studies have reported an increase in NSBR immediately
23      after exposure to O3.  Whether the late NSBR reported in this study is a persistent residual effect
24      of an earlier increase in airway responsiveness, or is a true one day  lag effect cannot be
25      determined from this study.  Whatever the origin, however, a delayed increase in airway
26      responsiveness raises a question of potentially increased susceptibility of an individual to
27      respiratory impairment, particularly if the suggested mechanism of disrupted epithelial
28      membrane holds true.
29
        August 2005                            AX6-64      DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX6.5.6  Oxidant-Antioxidant Balance
 2           Oxidant-antioxidant balance has been considered as one of the determinants of O3
 3     responsiveness. Amateur cyclists who took antioxidant supplements (vitamins C, E, and
 4     p-carotene) for three months showed no decrements in spirometric lung function when cycling
 5     on days with high O3 levels.  In contrast, matched control group of cyclists not pretreated with
 6     vitamin supplements experienced an almost 2% decline in FVC and FEVj and >5% reduction in
 7     PEF during the same activity period. Adjustment of data for confounders such as PM10 and NO2
 8     did not change the findings. Apparently, substantially elevated levels of plasma antioxidants
 9     may afford some protection against lung function impairment (Grievink et al.,1998,  1999).
10           Both laboratory animal and human studies have repeatedly demonstrated that antioxidant
11     compounds present the first line of defense against the oxidative stress.  Thus, upregulation of
12     both enzymatic and nonenzymatic antioxidant systems is critical to airway epithelial protection
13     from exposure to oxidants such as O3 and NO2 (see  Table AX6-7). As an extension of an earlier
14     study focused  on pulmonary function changes (Frampton et al., 1997a), Avissar et al. (2000)
15     hypothesized that concentration of glutathione peroxidase (GPx), one of the antioxidants in
16     epithelial lining fluid (ELF), is related to O3 and NO2 responsiveness. They exposed healthy
17     young nonsmokers (n = 25), O3-responders, and nonresponders to filtered air and twice to
18     0.22 ppm O3 for 4 h (IE, 20' ex 710' rest, @ VE 40 to 46 L/min). In the NO2 part of the study,
19     subjects were exposed to air and twice to NO2 (0.6 and 1.5 ppm) for 3 h, with IE of 10 min of
20     each 30 min @ VE of 40 L/min.  Ozone exposure elicited a typical pulmonary function response
21     with neutrophilic airway inflammation in both responders and nonresponders.  The GPx activity
22     was significantly reduced (p = 0.0001) and eGPx protein significantly depleted (p =  0.0001) in
23     epithelial lining fluid (ELF) for at least 18 h postexposure. In contrast, both GPx and eGPx were
24     slightly elevated in bronchoalveolar lavage fluid (BALF). However, neither of the two NO2
25     exposures had a significant effect on pulmonary function, airway neutrophilia, epithelial
26     permeability, GPx activity, or eGPx protein level in either ELF or BALF.  The lack of a
27     significant response to NO2 has been attributed to the weak oxidative properties of this gas.
28     No association has been observed between cell injury, assessed by ELF albumin, or pulmonary
29     function and GPx activity for O3 exposure. Thus, it is unclear what role antioxidants may have
30     in modulation  of O3-induced lung function and inflammatory responses. The  authors found a
31     negative association between lower baseline eGPx protein concentration in ELF and post-O3

       August 2005                             AX6-65      DRAFT-DO NOT QUOTE OR CITE

-------
 1      neutrophilia to be an important predictor of O3-induced inflammation; however, the causal
 2      relationship has not been established.
 3           The effects of dietary antioxidant supplementation on O3-induced pulmonary and
 4      inflammatory response of young healthy individuals has been investigated by Samet et al.
 5      (2001). Under controlled conditions, subjects received ascorbate restricted diet for three weeks.
 6      After the first week of prescribed diet, subjects were randomly assigned into two groups, and
 7      exposed to air (2 h, IE every 15  min at 20 L/min/m2 BSA). Thereafter, one group received daily
 8      placebo pills and the other a daily supplement of ascorbate, a-tocopherol and a vegetable juice
 9      for the next two weeks.  At the end of a two week period subjects were exposed to 0.4 ppm O3
10      under otherwise similar conditions as in sham exposures. Serum concentration of antioxidants
11      determined prior to O3 exposure showed that  subjects receiving supplements had substantially
12      higher concentrations of ascorbate, tocopherol and carotenoid in blood than the control group.
13      Plasma levels of glutathione and uric acid (cellular antioxidants) remained essentially the same.
14      Ozone exposure reduced spirometric lung function in both groups; however, the average
15      decrements in the supplementation group were smaller for FVC (p = 0.046) and FEVj
16      (p = 0.055) when compared to the placebo group. There was no significant correlation between
17      individual lung function changes and respective plasma levels of antioxidants. Individuals in
18      both groups experienced typical post-O3  subjective symptoms of equal severity.  Similarly, the
19      inflammatory response as assessed by BALF  showed no significant differences between the two
20      groups  either in the recovery of cellular components or the types and concentrations of
21      inflammatory cytokines.  Because of the complexity of protocol, the study was not designed as a
22      cross-over type.  However, it is unlikely that the fixed air-O3 sequence of exposures influenced
23      the findings in any substantial way.  Although the study did not elucidate the protective
24      mechanisms, it has demonstrated the value of dietary antioxidants in attenuating lung function
25      effects of O3. This observation may appear to contradict the findings of Avissar's and colleagues
26      study (2000) discussed above; however, neither study found association between lung function
27      changes and glutathione levels.  The lack of such association suggests that activation of
28      antioxidant protective mechanisms is seemingly independent of mechanisms eliciting lung-
29      function changes and that dietary antioxidants afford protection via a different pathway than
30      tissue-dependent antioxidant enzymes. Moreover, the findings of this study have provided
        August 2005                             AX6-66      DRAFT-DO NOT QUOTE OR CITE

-------
 1      additional evidence that symptomatic, functional, inflammatory, and antioxidant responses are
 2      operating through substantially independent mechanisms.
 3           Further evidence that the levels and activity of antioxidant enzymes in ELF may not be
 4      predictive or indicative of O3-induced lung function or inflammatory effects has been provided
 5      by a study of Blomberg et al. (1999). No association was found between the respiratory tract
 6      lining fluid redox potential level, an indicator of antioxidants balance, and either spirometric or
 7      inflammatory changes induced by a moderate exposure of young individuals to O3 (0.2 ppm/2 h,
 8      intermittent exercise at 20 L/min/m2 BSA). However, O3 exposure caused a partial depletion of
 9      antioxidants (uric acid, GSH, EC-SOD) in nasal ELF and a compensatory increase in plasma uric
10      acid,  affording at least some local protection (Mudway et al. 1999). More recently, Mudway
11      et al.  (2001) investigated the effect of baseline antioxidant levels on response to a 2-h exposure
12      to 0.2 ppm O3 in 15 asthmatic and 15 healthy subjects.  In the BALF of 15 healthy subjects,
13      significant O3-induced reductions in ascorbate and increases in glutathione disulphide and
14      EC-SOD were observed, whereas, levels were unaffected by O3 exposure in the asthmatics.
15      In both groups, BALF levels of uric acid and a-Tocopherol were unaffected by O3.
16           Trenga et al. (2001) studied the potential protective effects of dietary antioxidants (500 mg
17      vitamin C and 400 IU of vitamin E) on bronchial responsiveness of young to middle-aged
18      asthmatics.  Recruited subjects were prescreened by exposure to 0.5 ppm SO2 for 10 min while
19      exercising on a treadmill and selected for study participation if they experienced a >8% decease
20      in FEVj. Prior to the 1st exposure, subjects took either two supplements or two placebo pills at
21      breakfast time for 4 weeks.  They continued taking respective pills for another week when the
22      2nd exposure took place.  The 45-min exposures to air and 0.12 ppm O3 (15 min IE, VE « 3*
23      resting rate) via mouthpiece were randomized. Each exposure was followed by two 10-min
24      challenges to 0.10 and 0.25  ppm SO2 with exercise to determine bronchial hyperresponsiveness.
25      Due to potential variability  of baseline lung function between days, and the way the data have
26      been  presented, it is difficult to interpret the results. All spirometric measures (FEVl3
27      FVC, FEF25.75, and PEF) were significantly decreased from baseline at subsequent time points
28      following both the FA and O3  exposures.  Exposure to O3 caused significant decrements in FEVj
29      and PEF. The post FA and  O3 exposure decrements in lung function were not affected by the
30      treatment regimen (placebo versus vitamin). Bronchial hyperresponsiveness to 0.1 ppm SO2 was
31      also unaffected by treatment regimen.  Based on the prescreening SO2 challenge, subjects were

        August 2005                            AX6-67      DRAFT-DO NOT QUOTE OR CITE

-------
 1     ranked by their bronchial responsiveness to SO2 as "less-severe" (8 to 16% FEVj decrements)
 2     and "more-severe" (27 to 44% FEVj decrements).  The authors concluded O3 exposure increases
 3     bronchial responsiveness to SO2 in asthmatics and that antioxidant supplementation has a
 4     protective effect against this responsiveness, especially in the "more-severe" responders.
 5
 6     AX6.5.7  Genetic Factors
 7          It has been repeatedly postulated that genetic factors may play an important role in
 8     individual responsiveness to ozone. Recent studies (Bergamaschi et al., 2001; Corradi et al,
 9     2002; Romieu et al, 2004) have indeed found that genetic polymorphisms of antioxidant
10     enzymes, namely NAD(P)H:quinone oxidoreductase (NQO1) and glutathione-S-transferase Ml
11     (GSTM1), may play an important role in attenuating oxidative stress of airway epithelium.
12     Bergamaschi and colleagues (2001) studied young  nonsmokers (15 F, 9 M; mean age 28.5 years)
13     who cycled for two hours on a cycling circuit in a city park on days with the average ozone
14     concentration ranging from 32 to 103 ppb. There was no control study group nor the intensity of
15     bicycling has been reported . Since spirometry was done within 30 min post-ride, it is difficult
16     to gage how much of the statistically significant (p = 0.026) mean decrement of 160 ml in FEVj
17     of 8/24 individuals with NQO1 wild type (NQOlwt) and GSTMlnull (GSTMlnull) genotypes
18     was due to ozone.  Individuals with other genotype combinations including GSTMlnull had a
19     mean post-ride decrement of FEVj of only 40 mL.  The post-ride serum level of Clara cell
20     protein (CC16), a biomarker of airway permeability, has been elevated in both subgroups.  Only
21     a "susceptible" subgroup carrying NQOlwt in combination with GSTMlnull genotype, serum
22     concentration of CC16 showed positive correlation with ambient concentration of ozone and
23     negative correlation with FEVj changes. Despite some interesting observations, the study results
24     should be interpreted cautiously.
25          A subsequent study from the same laboratory was conducted in a more controlled
26     environment (Corradi et al., 2002).  Healthy young (mean 30.1 yrs) individuals (12 M, 10 F)
27     underwent a single exposure to 0.1 ppm O3 for 2 h  while intermittently exercising at a moderate
28     load on a bicycle ergometer. The study design did  not incorporate sham exposure, though the
29     authors have stated that in a separate experiment the effects of exercise on markers of
30     inflammation in blood and EEC were negligible. The eight subjects with NQOlwt and
31     GSTMlnull genotype, the "susceptible" group, indeed showed an increase in markers of

       August 2005                           AX6-68     DRAFT-DO NOT QUOTE OR CITE

-------
 1      inflammation (IL-6, IL-8, TEARS, LTB4) and oxidative stress (8-isoprostane, H2O2)
 2      immediately post and 18 hrs postexposure.  The fourteen subjects with other combination of
 3      genotypes showed small and inconsistent response in EEC and blood biomarkers, though PMN
 4      activity in both groups was significantly increased by exposure. The DNA adduct 8-hydroxy-2'-
 5      deoxyguanosine (8-OhdG),  a marker of oxidative DNA damage, was elevated immediately
 6      postexposure in both groups but only in the "susceptible" group the increase became significant.
 7      The spirometric endpoints (not reported) were not affected by the exposure at any time point,
 8      which contrasts the previous study.  The incomplete study design calls for a careful
 9      interpretation of the findings.
10           It is of interest to note, that human nasal mucosa biopsies of GSTM1 deficient subjects
11      showed higher antioxidant enzymes activity than biopsies of GSTM1 positive individuals when
12      incubated for 24 h in 120 ppb O3 environment (Otto-Knapp et al., 2003).
13           The influence of functional polymorphism of inflammatory and other genes on O3
14      susceptibility was studied by Yang et al. (2005). In this study 54 nonsmoking subjects
15      (11 healthy subjects, 15 mild asthmatics, 25 with rhinitis) were exposed to 250 ppb O3 for 3 h
16      (44 subjects), 200 ppb for 4 h (4 subjects), and 400 ppb for 2 h (3 subjects). During these
17      exposures subjects intermittently exercised (-14 L/min/m2 BSA). The pooled data of the tumor
18      necrosis factor a (TNF-a), lymphotoxin-a (LTA), toll-like receptor 4 (TLR4 ), superoxide
19      dismutase (SOD2) and glutathione peroxidase (GPX1) genes appear to show only TNF-a as a
20      promising genetic factors of susceptibility.  However, as the authors stated "the functional
21      significance of individual TNF-a polymorphisms remains controversial" (Yang et al., 2005).
22           More specific genotyping has shown that O3 responsiveness and asthma risk may be related
23      to the presence of variant Ser allele for NQO1. In a field study of susceptibility to ambient O3 in
24      Mexico City, 4-17 yrs old asthmatic children (n = 218) were genotyped, including variant alleles
25      (David et al., 2003).  The risk of asthma was related to the 1-h daily  max ambient O3 which
26      ranged from 12 to 309 ppb.  Relative to Pro/Pro genotype the presence of at least one NQO1
27      Ser allele variant lowered the risk of asthma in these children (RR = 0.8). In children with
28      GSTMlnull genotype combined with at least one NQO1 Ser allele variant the decreased risk of
29      asthma became statistically  significant  (RR = 0.4).  The presence of Ser allele which renders
30      NQO1 less active,  thus affecting the conjugation of quinones and formation of ROS
        August 2005                            AX6-69      DRAFT-DO NOT QUOTE OR CITE

-------
 1      subsequently reducing the oxidative stress, may plausibly explain the protective effect of this
 2      genotypic combination.
 3           Another field study of asthmatic children (n = 158) exposed to ambient O3 (12-309 ppb
 4      1-h max during the 12 week study period) has found that in children with genetic deficiency of
 5      GSTM1 the decrements in FEF25.75 were related to the previous day 1-h daily max O3.  The
 6      association was more pronounced in moderate to severe asthmatics.  Children with GSTMlpos
 7      variant showed no significant decrement in lung function.  Randomly administered antioxidant
 8      supplementation (vit. C  250 mg/day and vit. E 50 mg/day) attenuated post-ozone lung  function
 9      response in GSTMlnull children (Romieu et al., 2004).
10           These recent studies have shown that individual's innate susceptibility to ozone may be
11      linked to genetic background of an individual. Although a number of potential ozone
12      susceptibility genes have been identified, additional better designed and controlled studies are
13      needed to ascertain the link between susceptibility and polymorphism.
14           Pretreatment of healthy young subjects with inhaled corticosteroids (2 x 800 |ig/day
15      budesonide, a maximal clinical dose) for 2 weeks prior to O3 exposure  (0.4 ppm/2 h, alternating
16      20 min exercise at SOW with 10 min rest) had no apparent effect on a typical lung function
17      decline or inflammatory response to exposure. Because of the complexity of the protocol, the
18      study was not a cross-over design and no control air exposures were conducted. Both the
19      placebo and treatment conditions caused the same magnitude of changes.  Similarly, nonspecific
20      bronchial reactivity to methacholine (PC20 FEVj) was increased about the same 4 h after
21      exposure. Neither absolute nor relative sputum cell counts were affected by budesonide
22      treatment and O3 induced a typical neutrophilic response in both groups. Upregulation of
23      pro-inflammatory mediators measured in sputum was not different between the groups either.
24      The markers of inflammation and oxidative stress, exhaled NO and CO, as well as the  reactive
25      product nitrite measured in exhaled breath condensate, respectively, were not significantly
26      influenced by budesonide.  However, considering all these findings as a whole, budesonide
27      seemed to have a moderating, although not statistically significant, effect on O3-induced
28      response (Nightingale et al., 2000). Budesonide is an anti-inflammatory drug that in laboratory
29      animal studies partially suppressed neutrophilic inflammation caused by O3 (Stevens et al.,
30      1994).  Because the dose of budesonide was at therapeutic maximal levels, the pharmacologic
31      action of this drug and the site of action of O3 do not apparently coincide.

        August 2005                             AX6-70      DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX6.6   REPEATED EXPOSURES TO OZONE
 2          Repeated daily exposure to O3 in the laboratory for 4 or 5 days leads to attenuated changes
 3     in pulmonary function responses and symptoms (Hackney et al., 1977'a; U.S. Environmental
 4     Protection Agency, 1986, 1996). A summary of studies investigating FEVj responses to
 5     repeated daily exposure for up to 5 days is given in Table AX6-8. The FEVj responses to
 6     repeated O3 exposure typically have shown an increased response on the second exposure day
 7     (Day 2) compared to the initial (Day 1) exposure response.  This is readily apparent in repeated
 8     exposures to a range of concentrations from 0.4 to 0.5 ppm O3 accompanied by moderate
 9     exercise (Folinsbee et al., 1980; Horvath et al., 1981; Linn et al., 1982), and at lower
10     concentrations, 0.20 to 0.35 ppm, when accompanied by heavy exercise (Brookes et al., 1989;
11     Folinsbee and Horvath, 1986; Foxcroft and Adams,  1986; Schonfeld et al., 1989).  Mechanisms
12     for enhanced pulmonary function responses on Day 2 have not been established, although
13     persistence of acute O3-induced damage for greater than 24 h may be important (Folinsbee et al.,
14     1993). An enhanced Day 2 FEVj response was less obvious or absent in exposures at lower
15     concentrations or those that caused relatively small group mean O3-induced decrements.
16     For example, Bedi et al. (1988) found no enhancement of the relatively small pulmonary
17     function responses in older subjects (median age, 65 years) exposed repeatedly to O3.  Three
18     reports (Bedi et al., 1985; Folinsbee and Horvath, 1986; Schonfeld et al., 1989) demonstrated
19     that enhanced pulmonary function responsiveness was present within 12 h, lasted for at least
20     24 h and possibly 48 h, but was absent after 72 h.
21          After 3  to 5 days of consecutive daily exposures to O3, FEVj responses are markedly
22     diminished or absent. One study (Horvath et al., 1981) suggested that the rapidity of this decline
23     in FEVj response was related to the magnitude of the subjects' initial responses to O3  or their
24     "sensitivity."  A summary of studies examining the effects of repeated exposures to O3 on FEVj
25     and other pulmonary function, symptoms, and airway inflammation is given in Table AX6-9.
26     Studies examining persistence of the attenuation of pulmonary function responses following
27     4 days of repeated exposure (Horvath et al., 1981; Kulle et al., 1982; Linn et al., 1982) indicate
28     that attenuation is relatively short-lived, being partially reversed within 3 to 7 days and typically
29     abolished within 1 to 2 weeks. Repeated exposures separated by 1 week (for up to 6 weeks)
30     apparently do not induce attenuation of the pulmonary function response (Linn et al.,  1982).
31     Gong et al. (1997b)  studied the effects of repeated exposure to 0.4 ppm O3 in a group  of mild

       August 2005                             AX6-71      DRAFT-DO NOT QUOTE OR CITE

-------
6-
c
S-
to
o
o
^











X
ON
to



o
l>
H
6
o

H
/O
r*^S
O
H


o
i — 1
H
W
Table AX6-8. Changes in Forced Expiratory
Ozone
Concentration1"
ppm ug/m3

0.12 235
0.20 392
0.20 392
0.20 392
0.20 392
0.25 490

0.35 686
0.35 686
0.35 686

0.35 686
0.40 784
0.40 784
0.4 784
0.4 784
0.42 823
0.45 882
0.45 882
0.47 921
0.5 980
0.5 980



Volume in One Second After Repeated Daily Exposure to Ozone"

Exposure Duration Number and Gender
and Activity0

6.6 h, IE (40)
2 h, IE (30)
2 h, IE (18 and 30)
2 h, IE (18 and 30)
1 h, CE (60)
1 h, CE (63)

2 h, IE (30)
1 h, CE (60)
1 h, CE (60)

1 h, CE (60)
3 h, IE (4-5 x resting)
3 h, IE (4-5 x resting)
2 h, IE (65)
3 h, IE (32)
2 h, IE (30)
2 h, IE (27)
2 h, IE (27)
2 h, IE (3 x resting)
2 h, IE (30)
2. 5 h, IE (2 x resting)

of Subjects

17 M
10M
8M, 13 F
9
15M
4M,2F
5M,2F
10M
8M
10M
10M
15M
13M5
HFf
8M
8 M, 2 Fh
24 M
1M, 5F
10M,6F
8 M, 2 F8
8M
6


First
-12.79
+1.4
-3.0
-8.7
-5.02
-20.2
-18.8
-5.3
-31.0
-16.1
-14.4
-15.9
-9.2
-8.8
-18.0
-34.7
-21.1
-13.3
-5.8
-11.4
-8.7
-2.7



Percent Change in FEV, on

Second
-8.73
+2.7
-4.5
-10.1
-7.8
-34.8
—
-5.0
-41.0
-30.4
—
-24.6
-10.8
-12.9
-29.9
-31.1
-26.4
—
-5.6
-22.9
-16.5
-4.9

Exposure Days
Third
-2.54
-1.6
-1.1
-3.2
—
—
-22.3
-2.2
-33.0
—
-20.6
—
-5.3
-4.1
-21.1
-18.5
-18.0
-22.8
-1.9
-11.9
-3.5
-2.4


Consecutive
References'1
Fourth Fifth
-0.6 0.2 Folinsbeeetal. (1994)
— — Folinsbeeetal. (1980)
— — Glineretal. (1983)
— — Glineretal. (1983)
— — Brookes etal. (1989)
— — Folinsbee and Horvath( 1986)
— —
— — Folinsbeeetal. (1980)
-25.0 — Foxcroft and Adams (1986)
— — Schonfeld etal. (1989)
— —
— — Brookes etal. (1989)
-0.7 -1.0 Kulle etal. (1982)
-3.0 -1.6 Kulle etal. (1982)
-7.0 -4.4 Folinsbeeetal. (1998)
-12.0 -6.2 Gong etal. (1997b)
-6.3 -2.3 Horvath etal. (1981)
— — Bedi etal. (1985)
— — Bedi etal. (1989)
-4.3 — Linn etal. (1982)
— — Folinsbeeetal. (1980)
-0.7 — Hackney etal. (1977a)

"See Appendix A for abbreviations and acronyms.
""Listed from lowest to highest O3 concentration.
'Exposure duration and intensity of IE or CE were variable;
VE (number in parentheses) given
*For a more complete discussion of these studies, see Table AX6-9 and U.S.
'Subjects were especially sensitive on prior exposure to 0.42
21 individuals used in this
study.
in liters per minute or as a multiple of resting ventilation.
Environmental Protection Agency (1986).
ppm O3 as evidenced by a decrease in FEV; of more than 20%.




These nine subjects are a subset of the total group of

'Bronchial reactivity to a methacholine challenge also was studied.
8Seven subjects completed entire experiment.
hSubjects had mild asthma.







-------
                                    Table AX6-9.  Pulmonary Function Effects with Repeated Exposures to Ozone"
to
o
o
X
H
6
o

o
H
O
O
H
W
O
O
HH
H
W
Ozone
Concentration" Exposure
ppm Mg/ni3 Activity
0.25 490 2 h IE, (30 min
rest, 30 min
exercise),
VE = 39 L/min

0.2 392 4 h IE
(4 x 30 min
exercise), VE =
14.8 L/min/m2
BSA
0.2 392 4 h IE
(4 x 30 min
exercise), VE =
25 L/min/m2
BSA

0.4 784 3h/day for 5
days
IE (15 min rest,
15 min
exercise)
VE = 32 L/min
0.12 235 6.6 h
50 min
Number and
Exposure Gender of
Conditions Subjects
21.4 °C 5M, 3F
43.9% RH
4 days consecutive
FA exposure; 4 days
consecutive O3
exposure
1 day FA, 1 day, O3; 15 M, 8F
4 days consecutive
exposure to O3
20 °C 50% RH 9 M, 6 F
(1 day, O3; 4 days
consecutive exposure
toO3

31 °C 8M,2F
35% RH
5 consecutive days
plus follow up @ 4 or
7 days
18 °C 17 M
40% RH
Subject
Characteristics Observed Effect(s)
Healthy, NS FVC and FEVj decrements were significantly
attenuated on Day 4 of O3 exposure compared to day
1 of O3 exposure. Significant small airway function
depression accompanied by significant neutrophilia
in BALF one day following the end of O3 exposure.

Healthy, NS FEVj decrement and symptoms significantly
21 to 35 years old reduced on Day 4 of O3 exposure compared to Day
1 of O3 exposure. Airway inflammation of mucosa
persisted on Day 4 although some inflammatory
markers in BALF attenuated significantly.
Healthy, NS Significant decrease in FVC, FE Vl , SRaw, and
23 to 37 years old symptoms on Day 4 of O3 exposure compared to a
single day of O3 exposure. Number of PMNs,
fibronectin, and IL6 in BALF were significantly
decreased on Day 4 compared to a single day of O3
exposure.
Mild asthma FEV; decreased 35% on day 1 and only 6% on day
adult 5. Bronchial reactivity increased after day 1 and
remained elevated. Adaptation of asthmatics is
similar to healthy subjects but may be slower and
less complete.
Healthy NS FEV; responses were maximal on first day of
exposure (-13%), less on second day (-9%), absent
Reference
Frank et al.
(2001)

Jorres et al.
(2000)
Christian et al.
(1998)

Gong et al.
(1997b)
Folinsbee et al.
(1994)
                            exercise/10 min  five consecutive
                            rest, 30 min     daily exposures
                            lunch
                            VE=38.8
                            L/min
thereafter. Symptoms only the first 2 days.
Methacholine airway responsiveness was at least
doubled on all exposure days, but was highest on the
second day of O3. Airway responsiveness was still
higher than air control after 5 days of O3 exposure.
Trend to lessened response, but it was not achieved
after 5 days.

-------
                             Table AX6-9 (cont'd). Pulmonary Function Effects with Repeated Exposures to Ozonea
to
o
o
X
Ozone
Concentrationb
ppm ug/m3 and Activity
0.4 784 2 h IE (15 min rest,
1 5 min exercise
VE « 60 L/min

0.45 882 2 h
IE
(3 x 20 min
exercise)
VE = 27 L/min



0.20/0.20 392/392 1 h
0.35/0.20 686/392 CE at 60 L/min
0.35/0.35 686/686






0.35 686 60minCE
VE = 60 L/min
Number and
Gender of
Exposure Conditions Subjects
5 days consecutive O3 16 M
exposure


23.3 °C 10M,6F
63% RH
Exposed for
3 consecutive days,
not exposed for 2 days,
then exposed to 0.45 ppm
again for 1 day

21 to 25 °C 15 M
40 to 60% RH
(three 2-day sets
of exposures)





21 to 25 °C 40 M
40 to 60% RH (4 groups
Subject
Characteristics
Healthy NS



Healthy NS
60 to 89 years old
median 65 years
old; mean FVC =
3.99 L; mean FEV,
= 3.01L;
FEVj/FVC range =
61 to 85%
Healthy aerobically
trained NS,
FVC = 4.24 to
6.98L





NS; nonallergic,
non-Los Angeles
Observed Effect(s)
O3-exposure FEV[ decrement was greater on
day 2, 29.9%, than day 1, 18.0%, then
decreased on day 3, 21.1%, day 4, 7% and
day 5, 4.4%
Overall increase in symptoms, but no single
symptom increased significantly. FVC
decreased 111 mL and 104 mL on Days 1
and 2, respectively. FEV[ fell by 171 and
164 mL, and FEV3 fell by 185 and 172 mL.
No significant changes on Days 3 and 4 or
with FA. FEV[ changes were -5.8, -5.6,
-1.9, and - 1 .7% on the four O3 days.
Consecutive days of exposure to 0.20 ppm
produced similar FEV; responses on each day
(-5.02, -7.80); 0.35/0.20 ppm pair caused
increased response to 0.20 ppm on second
day (-8.74); 0.35/0.35 ppm caused much
increased response on Day 2 (- 15.9, -24.6).
Symptoms were worse on the second
exposure to 0.35 ppm, but not with second
exposure to 0.20 ppm.
No differences between responses to
exposures separated by 72 or 120 h.
Reference
Folinsbee et al.
(1998)
Devlin et al.
(1997)
Bedi et al.
(1989)






Brookes et al.
(1989)







Schonfeld et al.
(1989)
                                            (two exposures for each
                                            subject separated by 24,
                                            48, 72, or 120 h)
of 10)        residents for >6 mo;
             =25 years old
Enhanced FEV[ response at 24 h (-16.1% vs.
- 30.4%). Possible enhanced response at 48 h
(-14.4% vs. -20.6%). Similar trends
observed for breathing pattern and SRaw.
0.45 882 2 h IE
(3 x 20 min
exercise)
VE = 26 L/min
23.3 °C 8M, 8F
62.5% RH
(three exposures with
a minimum 1-week
interval)
Healthy NS, 61
years old for M and
65 years old for F
(FVC = 4.97 L for
MandS.llLforF)
Spirometric changes were not reproducible
from time to time after O3 exposure
R < 0.50). Repeat exposures to air yielded
consistent responses.
Bedi et al.
(1988)

-------
                        Table AX6-9 (cont'd). Pulmonary Function Effects with Repeated Exposures to Ozonea
to
o
o
X
O
HH
H
W
Ozone
Concentrationb
ppm MS/m3 and Activity
0.18 353 2h
IE (heavy)
VE *> 60 to 70 L/min
(35 L/min/m2 BSA)







0.45 882 2 h
(+0.30 IE (20 min rest,
PAN) 20 min exercise)
VE = 27 L/min





0.35 686 «lhCE
(see paper
Exposure Conditions
31 °C
35% RH
(screen exposures
in spring 1986;
second exposures in
summer/fall 1986 and
winter 1 987 and spring
1987 for responders and
nonresponders only)


22 °C
60% RH
5 days consecutive
exposure to PAN + O3





22 to 25 °C
35 to 50% RH
Number and
Gender of Subject
Subjects Characteristics
59 adult Responders:
Los Angeles 1 9 to 40 years old
residents 6 atopic,
12 responsive 2 asthmatic,
13 4 normal
nonresponsive
Nonresponders:
18 to 39 years old,
1 3 normal


3 M, 5 F Healthy NS,
Mean age =
24 years






8 M Aerobically trained
healthy NS (some
Observed Effect(s)
Responders had AFEVj = - 12.4% after initial
screening; nonresponders had no change.
Responders had nonsignificant response in
late summer or early winter, but were
responsive again in early spring (spring 1986,
-385 mL; Autumn 1986, - 17 mL; winter
1987, +16 mL; spring 1987, -347 mL).
Nonresponders did not change with season.
Suggests that responders responses may vary
with ambient exposure, but nonresponders
generally remain nonresponsive.
FEV; decreased «19% with O3 alone, «15%
on Day 1 of O3 + PAN, «5% on Day 5 of O3
+ PAN, -7% 3 days after 5 days of O3 +
PAN, * 1 5% after 5 days of O3 + PAN.
Similar to other repeated O3 exposure studies,
O3 responses peaked after 2 days, were
depressed 3 days later, and responses
returned 7 days later. PAN probably had no
effect on repeated to O3 exposure responses.
Largest FEV[ decrease on second of 4 days
O3 exposure (-40% mean decrease). Trend
Reference
Linn et al.
(1988)
(also see
Hackney et al.,
1989)






Drechsler-
Parks et al.
(1987b)
(also see
Table AX6- 14)




Foxcroft and
Adams (1986)


c
2
1
o

o
H

O
H
W
O
0.35 686 «lhCE 22 to 25 °C 8M
(see paper 35 to 50% RH
for details) (1 day FA; 1 day O3;
4 days consecutive
exposure to O3)





"See Appendix A for abbreviations and acronyms.
bListed from lowest to highest O3 concentration.

Aerobically trained
healthy NS (some
were known O3
sensitive),
22.4 ± 2.2 years old








Largest FEV[ decrease on second of 4 days
O3 exposure (-40% mean decrease). Trend
for attenuation of pulmonary function
response not complete in 4 days. VO2maj,
decreased with single acute O3 exposure
(-6%) but was not significant after 4 days of
O3 exposure (-4%). Performance time was
less after acute O3 (21 1 s) exposure than after
FA (253 s).




Foxcroft and
Adams (1986)












-------
 1      asthmatics and observed a similar pattern of responses as those seen previously in healthy
 2      subjects.  The attenuation of pulmonary responses reached after 5 days of consecutive O3
 3      exposure was partially lost at 4 and 7 days postexposure.
 4           In addition to the significant attenuation or absence of pulmonary function responses after
 5      several consecutive daily O3 exposures, symptoms of cough and chest discomfort usually
 6      associated with O3 exposure generally are substantially reduced or absent (Folinsbee et al., 1980,
 7      1994; Foxcroft and Adams, 1986; Linn et al.,  1982). Airway responsiveness to methacholine is
 8      increased with an initial O3 exposure (Holtzman et al., 1979; Folinsbee et al., 1988), may be
 9      further increased with subsequent exposures (Folinsbee et al.,1994), and shows a tendency for
10      the increased response to diminish with repeated exposure (Dimeo et al., 1981; Kulle et al.,
11      1982). The initially enhanced and then lessened response may be related to changes that occur
12      during the repair of pulmonary epithelia damaged  as a consequence of O3 exposure.
13      Inflammatory responses (Koren et al., 1989a), epithelial damage, and changes in permeability
14      (Kehrl et al.,  1987) might explain a portion of these responses. By blocking pulmonary function
15      responses and symptoms with indomethacin pretreatment, Schonfeld et al. (1989) demonstrated
16      that in the absence of an initial response, pulmonary function and symptoms effects were not
17      enhanced on Day 2 by repeated exposure to 0.35 ppm O3. These results suggest that airway
18      inflammation and the release of cyclooxygenase products of arachidonic acid play a role in the
19      enhanced pulmonary function responses and symptoms observed upon reexposure to O3 within
20      48 h.
21           Response to laboratory O3 exposure as a function of the season of the year in the South
22      Coast Air Basin of Los Angeles, CA, has been examined in several studies (Avol et al., 1988;
23      Hackney et al., 1989; Linn et al.,  1988). Their primary purpose was to determine whether O3
24      responsive subjects would remain responsive after regular ambient exposure during the "smog
25      season". The subjects were exposed to 0.18 ppm O3 for 2 h with heavy IE on four occasions,
26      spring, fall, winter, and the following spring.  The marked difference in FEVj response between
27      responsive and nonresponsive subjects seen initially (-12.4% versus +1%) no longer was present
28      after the summer smog season (fall test) or 3 to 5 months later (winter test). However, when the
29      subjects were exposed to O3 during the following spring, the responsive subjects again had
30      significantly larger changes in FEVl3 suggesting a seasonal variation in response.
        August 2005                            AX6-76      DRAFT-DO NOT QUOTE OR CITE

-------
 1           Brookes et al. (1989) and Gliner et al. (1983) tested whether initial exposure to one O3
 2      concentration could alter response to subsequent exposure to a different O3 concentration.
 3      Gliner et al. (1983) showed that FEVj response to 0.40 ppm O3 was not influenced by previously
 4      being exposed to 0.20 ppm O3 for 2 h on 3 consecutive days. Brookes et al. (1989) found
 5      enhanced FEVj and symptoms upon exposure to 0.20 ppm after previous exposure to
 6      0.35 ppm O3.  These observations suggest that, although preexposure to low concentrations of O3
 7      may not influence responses to higher concentrations, preexposure to a high concentration of O3
 8      can significantly increase responses to a lower concentration on the following day.
 9           Foxcroft and Adams (1986) demonstrated that decrements in exercise performance seen
10      after  1 h of exposure to 0.35 ppm O3 with heavy CE were significantly less after 4 consecutive
11      days  exposure than they were after a single acute exposure.  Further, exercise performance,
12      VO2max, VEmax and HR^ were not significantly different after 4 days of O3 exposure compared to
13      those observed in a FA exposure. Despite the change in exercise performance, Foxcroft and
14      Adams (1986) did not observe a significant attenuation of FEVj response, although symptoms
15      were significantly reduced. However, these investigators selected known O3-sensitive subjects
16      whose FEVj decrements exceeded 30% on the first 3 days of exposure. The large magnitude of
17      these responses, the trend for the responses to decrease on the third and fourth day, the decreased
18      symptoms, and the observations by Horvath et al. (1981) that O3-sensitive subjects adapt slowly,
19      suggest that attenuation of response would have occurred if the exposure series had been
20      continued for another  1 or 2 days. These observations support the contention advanced by
21      Horvath et al. (1981) that the progression of attenuation of response is a function of initial "O3
22      sensitivity."
23           Drechsler-Parks et al. (1987b) examined the response to repeated exposures to 0.45 ppm O3
24      plus 0.30 ppm peroxyacetyl nitrate (PAN) in 8 healthy subjects and found similar FEVj
25      responses to exposures to O3 (-19%) and to O3 plus PAN (-15%).  Thus, PAN did not increase
26      responses to O3. Further, repeated exposure to the PAN plus O3 mixture resulted in similar
27      changes to those seen with repeated O3 exposure alone.  The FEVj  responses fell to less  than
28      - 5% after the fifth day, with the attenuation of response persisting  3 days after the repeated
29      exposures,  but being absent after 7 days. These observations suggest that PAN does  not
30      influence the attenuation of response to repeated O3 exposure.  If the PAN responses  are
31      considered negligible, this study confirms the observation that the attenuation of O3 responses

        August 2005                            AX6-77      DRAFT-DO NOT QUOTE OR CITE

-------
 1      with chamber exposures lasts no longer than 1 week. [More discussion on the interaction ofO3
 2      with other pollutants can be found in Section AX6.11.]
 3           Folinsbee et al. (1993) exposed a group of 16 healthy males to 0.4 ppm O3 for 2 h/day on
 4      5 consecutive days. Subjects performed heavy IE ( VE = 60 to 70 L/min).  Decrements in FEVj
 5      averaged 18.0, 29.9, 21.1, 7.0, and 4.4% on the 5 exposure days. However, baseline preexposure
 6      FEVj decreased from the first day's preexposure measurement and was depressed by an average
 7      of about 5% by the third day. This study illustrates that, with high-concentration and heavy-
 8      exercise exposures, pulmonary function responses may not be completely recovered within 24 h.
 9      During this study, BALF also was obtained immediately after the Day 5 exposure, with results
10      reported by Devlin et al. (1997).  These authors found that some inflammation and cellular
11      responses associated with acute O3 exposure were also attenuated after 5 consecutive days of O3
12      exposure (compared to historical data for responses after a single-day exposure), although
13      indicators of epithelial cell damage—not seen immediately after acute exposure—were present
14      in BALF after the fifth day of exposure. When reexposed again 2 weeks later, changes in BALF
15      indicated that epithelial cells appeared fully repaired (Devlin et al., 1997).
16           Frank et al.  (2001) exposed 8 healthy young adults to 0.25 ppm O3 for 2 h with moderate
17      IE (exercise VE = 40 L/min) on 4 consecutive days. In addition to standard pulmonary function
18      measures, isovolumetric FEF25.75, Vmax50 and Vmax75 were grouped into a single value representing
19      small airway function (SAWgrp). Exercise ventilatory pattern was also monitored each day,
20      while peripheral airway resistance was measured by bronchoscopy followed by lavage on Day 5.
21      The authors observed two patterns of functional response in their subjects— attenuation and
22      persistent. Values of FVC and FEVj showed significant attenuation by Day 4 compared to Day
23      1 values. However, SAWgrp and rapid shallow breathing during exercise persisted on Day 4
24      compared to Day 1, and were accompanied by significant neutrophilia in BALF 1 day following
25      the end of O3 exposure. Frank et al. (2001) suggested that both types of functional response (i.e.,
26      attenuation and persistence) are linked causally to inflammation. They contend that the
27      attenuation component is attributable at least in part to a reduction in local tissue dose during
28      repetitive exposure that is likely to result from the biochemical, mechanical, and morphological
29      changes set in motion by  inflammation.  They speculated that the persistent component
30      represents the inefficiencies incurred through inflammation. Whether the persistent small airway
        August 2005                             AX6-78      DRAFT-DO NOT QUOTE OR CITE

-------
 1     dysfunction is a forerunner of more permanent change in the event that oxidant stress is extended
 2     over lengthy periods of time is unknown.
 3          Early repeated multihour (6 to 8 h) exposures focused on exposures to low concentrations
 4     of O3 between 0.08 and 0.12 ppm (Folinsbee et al., 1994; Horvath et al.,  1991; Linn et al., 1994).
 5     Horvath et al. (1991) exposed subjects for 2 consecutive days to 0.08 ppm using the 6.6-h
 6     prolonged exposure protocol  (see Table AX6-2). They observed small pre- to postexposure
 7     changes in FEVj (-2.5%) on  the first day, but no change on the second day. Linn et al. (1994)
 8     observed a 1.7% decrease in FEVj in healthy subjects after 6.6 h exposure to 0.12 ppm O3.
 9     A second consecutive day exposure to O3 yielded even smaller (<1%) responses. In a group of
10     asthmatics exposed under similar conditions (Linn et al., 1994), the FEVj response on the first
11     day was -8.6% which was reduced to -6.7% on day 2, both significantly greater than those
12     observed for the nonasthmatics group.  The observations of Horvath et al. (1991) and Linn et al.
13     (1994) elicited a somewhat different pattern of response (no enhancement of response after the
14     first exposure) than that seen  at higher concentrations in 2 h exposures with heavy exercise
15     (Tables AX6-8 and AX6-9).  However, the subjects studied by Horvath et al. (1991) were
16     exposed only to 0.08 ppm O3 and were somewhat older (30 to 43 yrs) than the subjects studied
17     by Folinsbee et al. (1994), mean  age of 25 yrs, while the nonasthmatic subjects studied by Linn
18     et al. (1994) were also older (mean = 32 yrs), had lower exercise VE (-20%) and were residents
19     of Los Angeles who often encountered ambient levels of O3 at or above 0.12 ppm.
20          Folinsbee et al. (1994) exposed 17 subjects to 0.12 ppm O3 for 6.6 h, with 50 min of
21     moderately heavy exercise (VE = 39 L/min) each hour, on 5 consecutive days.  Compared with
22     FA, the percentage changes in FEVj over the five days were -12.8%, -8.7%, -2.5%, -0.06%,
23     and +0.18%.  A parallel attenuation of symptoms was observed, but the effect of O3 in enhancing
24     airway responsiveness (measured by increase in SRaw upon methacholine challenge) over
25     5 days was not attenuated (3.67,  4.55, 3.99, 3.24, and 3.74, compared to 2.22 in FA control).
26     Nasal lavage revealed no increases in neutrophils except on the first O3 exposure day.
27          Christian et al. (1998) exposed 15 adults (6 females and 9 males; mean age = 29.1 yrs) to
28     4 consecutive days at 0.20 ppm O3 for 4 h, with 30 mm of IE (exercise VE = 25 L/min/m2)  each
29     hour. Measures of FEVj, FVC, and symptoms were all significantly reduced on Day  1, further
30     decreased on Day 2, and then attenuated to near FA control values on Day 4. The pattern of
31     SRaw response was similar, being greatest on Day 2 and no different from FA control on Day 4.

       August 2005                             AX6-79      DRAFT-DO NOT QUOTE OR CITE

-------
 1      BAL was done on Day 5 and showed that neutrophil recruitment to the respiratory tract was
 2      attenuated with repeated short-term exposures, compared to Day 1 control O3 exposure, while
 3      airway epithelial injury appeared to continue as reflected by no attenuation of IL-6, IL-8, total
 4      protein, and LDH. The authors concluded that such injury might lead to airway remodeling,
 5      which has been observed in several animal studies (Brummer et al., 1977; Schwartz et al., 1976;
 6      Tepper et al., 1989; Van Bree et al., 1989).  In a similar study to that of Christian et al. (1998),
 7      Torres et al. (2000) exposed 23 adults (8 females and 15 males; mean age = 27.9 yrs) on
 8      4 consecutive days to 0.20 ppm O3 for 4 h, with 30 min of IE (exercise VE = 26 L/min) each
 9      hour. The authors observed that FEVj was significantly reduced and symptoms were
10      significantly increased on Day 1. On Day 2, FEVj was further decreased, while symptoms
11      remained unchanged. By Day 4, both FEVj and symptoms were attenuated to near FA, control
12      values.  Twenty hours after the Day 4 exposure, BAL and bronchial mucosal biopsies were
13      performed. These authors found via bronchial mucosal biopsies that inflammation of the
14      bronchial mucosa persisted after repeated O3 exposure, despite attenuation of some inflammatory
15      markers in BALF and attenuation of lung function responses and symptoms.  Further, Torres
16      et al. (2000) observed persistent although small decrease in baseline FEVj measured before
17      exposure, thereby suggesting that there are different time scales of the functional responses
18      to O3, which may reflect different mechanisms. The levels of protein remaining elevated after
19      repeated exposures confirms the findings of others (Christian et al., 1998; Devlin et al., 1997),
20      and suggests that there is ongoing cellular damage irrespective of the attenuation of cellular
21      inflammatory responses with the airways. [Further discussion on the inflammatory responses to
22      O3 can be found in Section AX6.9. ]
23           Based on studies cited here and in the previous O3 criteria documents (U.S. Environmental
24      Protection Agency, 1986, 1996), several conclusions can be drawn about repeated  1- to 2-h O3
25      exposures. Repeated exposures to O3 can cause an enhanced (i.e., greater) lung function
26      response on the second day of exposure.  This enhancement appears to be dependent on the
27      interval between the  exposures (24 h is associated with the greatest increase) and is absent with
28      intervals >3 days. As shown in Figure AX6-8, an enhanced response also appears  to depend
29      on O3 concentration and to some extent on the magnitude  of the initial response.  Small
30      responses to the first O3 exposure are less likely to result in an enhanced response on the second
31      day of O3 exposure.  Repeated daily exposure also results  in attenuation of pulmonary function

        August 2005                            AX6-80      DRAFT-DO NOT QUOTE OR CITE

-------
           c
           o
           o
           •s
           HI
100-

 95-

 90-

 85-

 80-

 75-

 70-

 65-
                   T        I        I         I        I         I
                  0.0      0.5      1.0      1.5      2.0       2.5
                                               Time (days)
                                                       i
                                                      3.0
                    I
                   3.5
4.0
      Figure AX6-8. Regression curves were fitted to day-by-day postexposure FEVt values
                     obtained after repeated daily acute exposures to O3 for 2 to 3 h with
                     intermittent exercise at a VE of 24 to 43 L/min (adaptation studies).
                     Symbols represent the results from individual studies conducted at 0.2 ppm
                     for 2 h (+), 0.35 ppm for 2 h (•), 0.4 ppm for 2 h (+), 0.5 ppm for 2 h (#),
                     and 0.54 ppm for 3 h (A). Also shown for comparison are the FEVt values
                     obtained after exposure to 0.12 ppm O3 for 10 h (•).

      Source: Modified from Hazucha (1993).
1     responses, typically after 3 to 5 days of exposure.  This attenuated response persists for less than
2     1 week or as long as 2 weeks. In temporal conjunction with the pulmonary function changes,

3     symptoms induced by O3, such as cough and chest discomfort, also are attenuated with repeated

4     exposure. Ozone-induced changes in airway responsiveness attenuate more slowly than

5     pulmonary function responses and symptoms. Attenuation of the changes in airway

6     responsiveness appear to persist longer than changes in pulmonary function, although this has

7     been studied only on a limited basis. In longer-duration (6.6 h), lower-concentration studies that

8     do not cause an enhanced second-day response, the attenuation of response to O3 appears to
      August 2005
                               AX6-81
DRAFT-DO NOT QUOTE OR CITE

-------
 1     proceed more rapidly. Inflammatory markers from BALF on the day following both 2 h (Devlin
 2     et al., 1997) and 4 h (Christian et al., 1998; Torres et al., 2000) repeated O3 exposure for 4 days
 3     indicate that there is ongoing cellular damage irrespective of the attenuation of some cellular
 4     inflammatory responses of the airways, lung function responses and symptoms.
 5
 6
 7     AX6.7  EFFECTS ON EXERCISE PERFORMANCE
 8     AX6.7.1  Introduction
 9          In an early epidemiologic study examining race performances in Los Angeles area high
10     school cross-country runners, Wayne et al. (1967) observed that endurance exercise performance
11     was depressed by inhalation of ambient oxidant air pollutants. The authors concluded that the
12     detrimental effects of oxidant air pollutants on race performance might have been related to the
13     associated discomfort in breathing, thus limiting the runners' motivation to perform at high
14     levels, although physiologic effects limiting O2 availability could not be ruled out.
15     Subsequently, the effects of acute O3 inhalation on endurance exercise performance have been
16     examined in numerous controlled laboratory studies. These studies were discussed in the
17     previous O3 criteria document (U.S. Environmental Protection Agency, 1996) in two categories:
18     (1) those that examined the effects of acute O3 inhalation on maximal oxygen uptake (VO2max)
19     and (2) those that examined the effects of acute O3 inhalation on the ability to complete
20     strenuous continuous exercise protocols of up to 1 h in duration. In this section, major
21     observations in these studies are briefly reviewed with emphasis on reexamining the primary
22     mechanisms causing decrements in VO2max and endurance exercise performance consequent
23     to O3 inhalation. A summary of major studies of O3 inhalation effects on endurance exercise
24     performance, together with observed pulmonary function and symptoms of breathing discomfort
25     responses, is given in Table AX6-10.
26
27     AX6.7.2  Effect on Maximal Oxygen Uptake
28          Three early studies (Folinsbee et al., 1977; Horvath et al., 1979; Savin and Adams, 1979)
29     examining the effects of acute O3 exposures on VO2max were reviewed in an earlier O3 criteria
30     document (U.S. Environmental Protection Agency, 1986).  Briefly, Folinsbee et al. (1977)


       August 2005                           AX6-82      DRAFT-DO NOT QUOTE OR CITE

-------
I
S-
to
o















^>
X
o\
1

!-rj
H
6
o
2!
-^— i
o
H
O
o
NW'
H
W
O
O
HH
H
W
Table AX6-10. Ozone Effects on Exercise Performance"
Ozone
Concentration b

ppm
0.06-0.07
0.12-0.13



0.18


0.35


0.12
0.20

0.12
0.18
0.24


0.21

0.20
0.35



0.25
0.50
0.75

0.15
0.30


0.75




Hg/m3
120-140
245-260



353


686


235
392

235
353
470


412

392
686



490
980
1,470

294588



1,470



Exposure Duration Exposure
and Activity Conditions
CE 23 to 24.5 °C
(VE = 30 to 120 L/min) 50 to 53% RH
16 to 28 min progressive
maximum exercise
protocol
1 h CE or competitive NA
simulation at mean
VE = 94 L/min
50 min CE 22 to 25 °C
VE = 60 L/min 35 to 50% RH

IhCE 31 °C
VE = 89 L/min

1 h competitive 23 to 26 °C
simulation exposures at 45 to 60% RH
mean
VE = 87 L/min

1 h CE at 75% V02max 19 to 21 °C
60 to 70% RH
1 h CE or competitive 23 to 26 °C
simulation at mean 45 to 60% RH
VE = 77.5 L/min


2 h rest NA



—30 min, progressively 23 °C
incremented exercise to 50% RH
voluntary exhaustion

2 h IE NA
(4x 15 min light [50 W]
bicycle ergometry)

Number and
Gender of Subject
Subjects Characteristics
12 M, 12 F Athletic




# not given; Well-trained
all males distance runners

8 M Trained
nonathletes

15M, 2F Highly trained
competitive
cyclists
10 M Highly trained
competitive
cyclists


6 M, 1 F Well-trained
cyclists
10 M Well-trained
distance runners



8M, 5F



9 M Healthy, NS
21 to 44 years old


13 M 4 light S,
9NS



Observed Effect(s)
Reduced maximum performance time and increased symptoms
of breathing discomfort during O3 exposure.



Maximal treadmill run time reduced from 71.7 min in FA to
66.2 min during O3 exposure with no decease in arterial O2
saturation.
VT decreased, fB increased with 50-min O3 exposures; decrease
in FVC, FEVj, FEF25.75, performance time, VO2max, V^^, and
HR,,,,,, from FA to 0.35-ppm O3 exposure.
Decrease in VEmax, VO2max, VTmax, workload, ride time, FVC,
and FEVj with 0.20 ppm O3 exposure, but not significant with
0. 12-ppm O3 exposure, as compared to FA exposure.
Decrease in exercise time of 7.7 min and 10. 1 min for subjects
unable to complete the competitive simulation at 0.18 and 0.24
ppm O3, respectively; decrease in FVC and FEVj for 0.18-and
0.24-ppm O3 exposure compared with FA exposure.

Decrease in FVC, FEVj, FEF25.75, and MVV with 0.21 ppm O3
compared with FA exposure.
VT decreased and fB increased with continuous 50-min O3
exposures; decrease in FVC, FEVj, and FEF25.75 from FA to
0.20 ppm and FA to 0.35-ppm O3 exposure in all conditions;
three subjects unable to complete continuous and competitive
protocols at 0.35 ppm O3.
FVC decreased with 0.50- and 0.75-ppm O3 exposure
compared with FA; 4% nonsignificant decrease in mean
VO2max following 0.75 ppm O3 compared with FA exposure.

Exposure to 0. 15 and 0.30 ppm O3 did not decrease maximal
exercise performance or VO2max compared to FA.
No significant pulmonary function or symptom responses were
observed, although a trend (P < .10) was evident.
Decrease in FVC, FEVj, ERV, 1C, and FEF50% after 1-h
0.75-ppm O3 exposure; decrease in VO2max, VTmax, VEmax,
maximal workload, and HRmax following 0.75-ppm O3
exposure compared with FA.

Reference
Linder et al.
(1988)



Folinsbee et al.
(1986)

Foxcroft and
Adams (1986)

Gong et al.
(1986)

Schelegle and
Adams (1986)



Folinsbee et al.
(1984)
Adams and
Schelegle (1983)



Horvath et al.
(1979)


Savin and
Adams (1979)


Folinsbee et al.
(1977)


"See Appendix A for abbreviations and acronyms.
""Listed from
lowest to highest O3 concentration.

-------
 1     observed that VO2max was significantly decreased (10.5%) following a 2-h exposure to
 2     0.75 ppm O3 with light (50 Watts) IE. Reduction in VO2max was accompanied by a decrease in
 3     maximal ventilation, maximal heart rate, and a large decrease in maximal tidal volume.
 4     In addition, the 2-h IE O3 exposure resulted in a 22.3% decrease in FEVj and significant
 5     symptoms of cough and chest discomfort.  In contrast, Horvath et al. (1979) did not observe a
 6     change in VO2max or other maximal cardiopulmonary endpoints in subjects exposed for 2 h at rest
 7     to either 0.50 or 0.75 ppm, although FVC was significantly decreased 10% following the latter
 8     exposure. Without preliminary exposure to O3, Savin and Adams (1979) examined the effects of
 9     a 30-min exposure to 0.15 and 0.30 ppm O3 while performing a progressively incremented
10     exercise test to volitional fatigue (mean  = 31.5 min in FA). No significant effect on maximal
11     work time or VO2max was observed compared to that observed upon FA exposure.  Further, no
12     significant effect on pulmonary function, maximal heart rate, and maximal tidal volume was
13     observed, although maximal VE was significantly reduced 7% in the 0.30 ppm O3 exposure.
14     Results  of these early studies suggest that VO2max is reduced if the incremented maximal exercise
15     test is preceded by an O3 exposure of sufficient total inhaled dose of O3 to result in significant
16     pulmonary function decrements and symptoms of breathing discomfort.
17          Using trained nonathletes, Foxcroft and Adams (1986) observed significant (p < 0.05)
18     reductions in rapidly incremented VO2max exercise performance time (-16.7%), VO2max (-6.0%),
19     maximal VE (-15.0%), and maximal heart rate (-5.6%) immediately following an initial 50-min
20     exposure to 0.35 ppm O3 during heavy CE (VE  = 60 L/min). These decrements were
21     accompanied by a significant reduction in FEVj (-23%) and the occurrence of marked
22     symptoms of breathing discomfort. Similarly, Gong et al. (1986) found significant reductions in
23     rapidly incremented VO2max exercise performance time (-29.7%), VO2max (-16.4%), maximal VE
24     (-18.5%), and maximal workload  (-7.8%) in endurance cyclists immediately following a 1-h
25     exposure to 0.20 ppm O3 with very heavy exercise ( VE 89 L/min), but not following exposure to
26     0.12 ppm. Gong et al. (1986) observed only a 5.6% FEVj decrement and mild symptoms
27     following exposure to 0.12 ppm, but a large decrement in FEVj (-21.6%) and substantial
28     symptoms of breathing discomfort following the 0.20 ppm exposure, which the authors
29     contended probably limited maximal performance and VO2max.
30

       August 2005                            AX6-84      DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX6.7.3  Effect on Endurance Exercise Performance
 2          A number of studies of well trained endurance athletes exposed to O3 have consistently
 3     observed an impairment of 1-h continuous heavy exercise performance of some individuals
 4     (Adams and Schelegle, 1983; Avol et al., 1984; Folinsbee et al., 1984; Gong et al., 1986).  The
 5     performance impairment is indicated by an inability to complete the prescribed O3 exposures
 6     (even at concentrations as low as 0.16 ppm) that subjects were able to complete in FA (Avol
 7     et al., 1984).  Other indications of impaired endurance exercise performance upon exposure to O3
 8     include a -7.7% reduced endurance treadmill running time when exposed to 0.18 ppm O3
 9     (Folinsbee et  al., 1986), which was accompanied by significantly decreased FEVj and
10     significantly elevated symptoms of breathing discomfort.  Another study (Schelegle and Adams,
11     1986) observed the failure of some trained endurance athletes to complete a 1-h competitive
12     simulation protocol upon exposure to O3 (30 min warm-up, followed immediately by 30 min at
13     the maximal workload that each subject could just maintain in FA; mean VE = 120 L/min).
14     In this study,  all subjects (n = 10) completed the FA exposure, whereas one, five, and seven
15     subjects could not complete the 0.12, 0.18,  and 0.24 ppm O3 exposures, respectively. Following
16     the 0.18 ppm  and 0.24 ppm O3 exposures, but not the 0.12 ppm exposure, FEVj was reduced
17     significantly and symptoms were significantly increased. Linder et al. (1988) also observed
18     small decrements in performance time (1 to 2 min) during a progressive maximal exercise test
19     (mean = 21.8 min) at O3 concentrations of 0.065 and 0.125 ppm. These small effects were
20     accompanied by a significant increase in subjective perception of overall effort at 0.125 ppm, but
21     with no significant reduction in FEVj at either O3 concentration. Collectively, reduced
22     endurance exercise performance and associated pulmonary responses are clearly related to the
23     total inhaled dose of O3 (Adams and Schelegle, 1983; Avol et al.,  1984; Schelegle and Adams,
24     1986).
25          Mechanisms  limiting VO2max and maximal exercise performance upon O3 exposure have
26     not been precisely identified. Schelegle and Adams (1986) observed no significant effect of O3
27     on cardiorespiratory responses, and there was no indirect indication that arterial O2 saturation
28     was affected.  The  latter is consistent with the observation that measured arterial O2 saturation at
29     the end of a maximal endurance treadmill run was not affected by O3 (Folinsbee  et al., 1986).
30     In studies in which O3 inhalation resulted in a significant decrease in VO2max, and/or maximal
31     exercise performance,  significantly decreased FEVj and marked symptoms of breathing

       August 2005                             AX6-85     DRAFT-DO NOT QUOTE OR CITE

-------
 1      discomfort were observed (Adams and Schelegle, 1983; Avol et al., 1984; Folinsbee et al., 1977,
 2      1984, 1986; Foxcroft and Adams, 1986; Gong et al., 1986; Schelegle and Adams, 1986).
 3      However, Gong et al. (1986) observed rather weak correlations between FEVj impairment and
 4      physiological variable responses during maximal exercise (R = 0.26 to 0.44). Rather, these
 5      authors concluded that substantial symptoms of breathing discomfort consequent to 1 h of
 6      very heavy exercise while exposed to 0.20 ppm O3, probably limited maximal performance
 7      and VO2max either voluntarily or involuntarily (Gong et al., 1986). Strong support for this
 8      contention is provided by the observation of significant increases in VO2max (4.7%) and maximal
 9      performance time (8.8%) following four consecutive days of 1 h exposure to 0.35 ppm O3 with
10      heavy exercise (VE  =60 L/min) compared to initial O3 exposure (Foxcroft and Adams, 1986).
11      These improvements, which were not significantly different from those for FA, were
12      accompanied by a significant reduction in symptoms of breathing discomfort with no significant
13      attenuation of FEVj and other pulmonary function responses. In this regard, Schelegle et al.
14      (1987) observed a disparate effect of indomethacin pretreatment (an inhibitor of the cyclo-
15      oxygenation of arachidonic acid to prostaglandins associated with inflammatory responses)
16      on O3-induced pulmonary function response (significant reduction) and an overall rating of
17      perceived exertion and symptoms of pain on deep inspiration and shortness of breath (no
18      significant effect).
19
20
21      AX6.8   EFFECTS ON AIRWAY RESPONSIVENESS
22           Increased airway responsiveness, also called airway hyperresponsiveness (AHR) or
23      bronchial hyperreactivity, indicates that the airways are more reactive to bronchoconstriction
24      induced by a variety of stimuli (e.g., specific allergens, exercise, SO2, cold air) than they would
25      be when normoreactive. In order to determine the level of airway responsiveness, airway
26      function (usually assessed by spirometry or plethysmography) is measured after the inhalation of
27      small amounts of an aerosolized specific (e.g., antigen, allergen) or nonspecific (e.g.,
28      methacholine, histamine) bronchoconstrictor agent or measured stimulus (e.g., exercise, cold
29      air). The dose or  concentration of the agent or stimulus is increased from a control, baseline
30      level (placebo) until a predetermined degree of airway response, such as a 20% drop in FEVj or
31      a  100% increase in Raw, has occurred (Cropp et al., 1980; Sterk et al., 1993). The dose or

        August 2005                            AX6-86      DRAFT-DO NOT QUOTE OR CITE

-------
 1      concentration of the bronchoconstrictor agent that produced the increased responsiveness often is
 2      referred to as the "PD^FEVj" or "PC2QFEVJ" (i.e., the provocative dose or concentration that
 3      produced a 20% drop in FEVj) or the "PD100SRaw" (i.e., the provocative dose that produced a
 4      100% increase in SRaw). A high level of bronchial responsiveness is a hallmark of asthma.
 5      The range of nonspecific bronchial responsiveness, as expressed by the PD20 for example, is at
 6      least 1,000-fold from the most sensitive asthmatics to the least sensitive healthy subjects.
 7      Unfortunately, it is difficult to compare the PD^FEVj or PD100SRaw across studies because of
 8      the many different ways of presenting dose response to bronchoconstrictor drugs, for example,
 9      by mg/mL, units/mL, and molar solution; or by cumulative dose (CIU or CBU) and doubling
10      dose (DD).  Further confounding comparisons by affecting the site of drug delivery, dose, and
11      ultimately bronchial responses, the size of aerosolized agents used in challenges can vary
12      between nebulizers and as a function of supply air pressure in otherwise identical systems.
13      Other typical bronchial challenge tests with nonspecific bronchoconstrictor stimuli are based
14      on exercise intensity or temperature of inhaled cold air.
15           Increases in nonspecific airway responsiveness were previously reported as an important
16      consequence of exposure to O3 (e.g., Golden et al., 1978; Table AX6-11). Konig et al. (1980)
17      and Holtzman et al.  (1979) found the increased airway responsiveness after O3 exposure in
18      healthy subjects appeared to be resolved after 24 h. Because atopic subjects had similar
19      increases in responsiveness to histamine after O3 exposure as nonatopic subjects, Holtzman et al.
20      (1979) concluded that the increased nonspecific bronchial responsiveness after O3 exposure was
21      not related to atopy.  Folinsbee and Hazucha (1989) showed increased airway responsiveness in
22      18 female subjects 1  and 18 h after exposure to 0.35 ppm O3. Taken together, these studies
23      suggest that O3-induced increases in  airway responsiveness usually resolve  18 to 24 h after
24      exposure, but may persist in some individuals for longer periods.
25           Gong et al. (1986) found increased nonspecific airway responsiveness in elite cyclists
26      exercising at competitive levels with O3 concentrations as low as 0.12 ppm. Folinsbee et al.
27      (1988) found an approximate doubling of the mean methacholine responsiveness in a group of
28      healthy volunteers exposed for 6.6 h to 0.12 ppm O3.  Horstman et al. (1990) demonstrated
29      significant decreases in the PD100SRaw in 22 healthy subjects immediately after a 6.6-h exposure
30      to concentrations  of O3 as low as 0.08 ppm. No relationship was found between O3-induced
31      changes in airway responsiveness and changes in FVC or FEVj (Folinsbee et al., 1988; Aris

        August 2005                            AX6-87      DRAFT-DO NOT QUOTE OR CITE

-------
                                     TABLE AX6-11. Airway Responsiveness Following Ozone Exposures"
S-
to

o
        Ozone

        Concentrationb
        ppm
                  ug/m3
Exposure Duration     Exposure     Number and          Subject

   and Activity        Conditions   Gender of Subjects    Characteristics
Observed Effect(s)
                                                                                                                                   Reference
0.125
0.250
0.125
245 3hIE
490 (10 min rest, 15 min
exercise on bicycle)
VE = 30 L/min
245 3hIEx4days
27 °C 5 F, 6 M Mild bronchial
50 % RH 20-53 years old asthma
6F, 16 M
1 9-48 years old Allergic rhinitis
Mean early-phase FEV; response and number
of >20% reductions in FEV! were
significantly greater after 0.25 ppm O3 or
4 x 0.125 ppm O3. Most of the >15% late-
phase FEV; responses occurred after
exposure to 4 x 0.125 ppm O3, as well as
significant inflammatory effects, as indicated
by increased sputum eosinophils (asthma and
allergic rhinitis) and increased sputum
lymphocytes, mast cell tryptase, histamine,
and LDH (asthma only).
Holz et al.
(2002)
X
ON

oo
oo
H

6
o


o
H

/O

o
H
W

O


O
HH
H
W
0.4



0.12




0.2





0.4



0.16



784 2 h IE NA
VE = 20 L/min/m2
BSA

235 45 min IE 60% RH
exercise, rest, exercise
VE = 3 x resting


392 4 h IE 20 °C
VE = 25 L/min/m2 62% RH
BSA



784 2 h IE NA
40 min/h @ 50 W


314 7.6 h IE 22°C
50 min/h 40% RH
VE s25 1/min

6F
1M
19-26 years old

12 F
5M
19-3 8 years old


4F
8M
23-47 years old



15 healthy subjects ;
9F,6M;31.1±2.1
years old

5F
4M


Stable mild asthma;
no meds 8 h
preexposure

Physician diagnosed
asthma; SO2-induced
airway
hyperreactivity

Healthy nonsmokers





Healthy; nonatopic



Mild atopic
asthma, HMD
sensitive, 20-35 years
old
Increased bronchial responsiveness to
methacholine 16 h after exposure; inhaled
apocynin treatment significantly reduced O3-
induced airway responsiveness.
The authors concluded O3 exposure increases
bronchial responsiveness to SO2 in asthmatics
and that antioxidant supplementation has a
protective effects against this responsiveness,
especially in the "more-severe" responders.
Increased sputum total cells, % neurtophils,
IL-6, and IL-8 at 18 h after exposure;
increased airway responsiveness to
methacholine 2 h after postexposure FEV1
returned to 5% of base-line; no anti-
inflammatory effect of azithromycin.
Decreased FEV; and FVC; increased
bronchial reactivity to methacholine
4 h postexposure; no protection from
inhaled corticosteroid, budesonide.
Mean 9.1% FEV; decrease 18 h after O3
exposure; provocative dose of dust mite
allergen decreased from 10.3 to 9.7 dose
units.
Peters et al.
(2001)


Trenga et al.
(2001)



Criqui et al.
(2000)




Nightingale
et al. (2000)


Kehrl et al.
(1999)



-------
                                    TABLE AX6-11 (cont'd).  Airway Responsiveness Following Ozone Exposures3
&
S-
to
o
o
Ozone
Concentration1"
ppm Mg/m3
0.2 392

and Activity
4hIE
40 min/h @ 50 W
Exposure Number and
Conditions Gender of Subj ects
NA 1 0 asthmatic (6 F, 4
M),
26.6 ± 2.3 years old;
1 0 healthy (4 F, 6
M),
27.3 ±1.4 years old.
Subject
Characteristics
Mild atopic asthma;
nonatopic healthy
subjects; no meds
8 weeks pre-exposure
Observed Effect(s)
Decreased FEV; in asthmatic (9.3%)
and healthy (6.7%) subjects; increased
sputum neutrophils in both groups (NS);
no change in methacholine airway reactivity
24 h postexposure.
Reference
Nightingale
etal. (1999)
X
oo
VO
0.12       235
Air-antigen
        0.4
        0.2
784
392
                            1 h rest
2hIE
 VE= 20 L/min/m2
BSA
                    4hIE
                    50 min/h
                    VE = 25 L/min/m2
                    BSA
                                NA
                                            NA
                        20 °C
                        50% RH
6F
9M

5F
1M
18-27 years old

6F
12 M
18-36 years old
Mild allergic asthma;
18 to 49 years if age

Stable mild asthma;
no meds 8 h
preexposure

Phy sician-diagno sed
mild asthma; no meds
prior to exposure
No effect of O3 on airway response to grass    Hanania
or ragweed allergen.                       et al. (1998)

Increased airway responsiveness to           Hiltermann
methacholine 16 h postexposure;             et al. (1998)
no effect of proteinase inhibitor, rALP.

Decreased FEV; and FVC, increased SRaw;    Balmes et al.
lower respiratory Sx; increased              (1997);
% neutrophils, total protein, LDH,            Scannell
fibronectin, IL-8, GM-CSF, and MPO         et al. (1996)
in BAL. Correlation between pre-exposure
methacholine challenge and O3-induced
SRaw increase.
0.4


C
— ]
i
O
§ °'12
\~S
H
c °-25
o
H
W
O
o
H
W
784 3 h/d for 5 days;
alternating 15 min of rest
and exercise at
VE = 32 L/min



236 Rest

490 3 h IE
VE = 30 L/min
1 5 min ex/
10 min rest/
5 min no O3; every
30 min.


31 °C
35% RH





22 °C
40% RH
27 °C
54% RH
mouthpiece
exposure




2F
8M
19-48 years old




5F
10M
24 mild asthmatics
11F/13M
12 allergic rhinitics
6M/6F




Mild asthma
requiring only
occasional
bronchodilator
therapy


atopic
asthma
atopic mild asthmatic
NS






Significant FEV[ and Sx response on 1 st and
2nd O3 exposure days, then diminishing with
continued exposure; tolerance partially lost
4 and 7 days postexposure; bronchial
reactivity to methacholine peaked on 1 st O3
exposure day, but remained elevated with
continued exposure.
No effect of O3 on airway response to grass
allergen.
Increased allergen responsiveness afer O3
exposure.






Gong et al.
(1997b)





Ball et al.
(1996)
Jorres et al.
(1996)







-------
 X
 H

 6
 o


 o
 H

O

 O
 H
 W

 O


 O
 HH
 H
 W
Ozone
Concentrationb
ppm
0.2




0.12






0.12





0.10
0.25
0.40


Air-antigen
0.12 ppm
O3-antigen
0.08
0.10
0.12

ug/m3
392




235






235





196
490
785





157
196
235

Exposure Duration
and Activity
4hIE
50min/10min
exercise/rest each hour


IhR






6.6 h, IE x 5 days
50 min exercise/10 min
rest, 30 min lunch
VE=38.8L/min


1 h light IE
2x15 min on treadmill
VE = 27 L/min


1 h at rest


6.6 h
IE at « 39 L/min


Exposure
Conditions
22 °C
50% RH



Ambient T
&RHfor
exposure;
23°C&
50% RH for
exercise
challenge
18 °C
40% RH




21 °C
40% RH



NA


18 °C
40% RH


Number and Subject
Gender of Subj ects Characteristics
42M/24F 18-50 years
NS healthy



8 F Mild stable asthma
7M
19-45 years old




17 M Healthy nonsmokers
25 ± 4 years old




9 F Stable mild
12 M asthmatics with FEV;
1 9-40 years old >70% and
methacholine
responsiveness
4 M, 3 F Asthmatic,
21 to 64 years old

22 M Healthy NS,
18 to 32 years old


Observed Effect(s)
FEV^- 18.6%), FVC (- 14.6%), decreased
after O3. Baseline PC100 for methacholine
was not related to changes in FVC, FEVl5
a weak association was seen for PC100 and
increased SRaw.
No significant difference in % fall FEV[ or
V40p; no increase in bronchial responsiveness
to exercise challenge.




FEV[ responses were maximal on 1st day of
exposure (-13%), less on second day (-9%),
absent thereafter. Sx responses only the first
2 days. Methacholine airway responsiveness
was at least doubled on all exposure days,
but was highest on the second day of O3.
No significant differences in FEV[ or FVC
were observed for 0.10 and 0.25 ppm O3-FA
exposures or postexposure exercise challenge;
12 subjects exposed to 0.40 ppm O3 showed
significant reduction in FEV[.
Increased bronchoconstrictor response to
inhaled ragweed or grass after O3 exposure
compared to air.
33, 47, and 55% decreases in cumulative dose
of methacholine required to produce a 100%
increase in SRaw after exposure to O3 at 0.08,
0.10, and 0.12 ppm, respectively.
Reference
Aris et al.
(1995)



Fernandes
etal. (1994)





Folinsbee
etal. (1994)




Weymer
etal. (1994)



Molfino
etal. (1991)

Horstman
etal. (1990)



-------
&
^
O
O
                           TABLE AX6-11 (cont'd). Airway Responsiveness Following Ozone Exposures'1
X
H

6
O


O
H

O

O
H
W

O


O
HH
H
W
Ozone
Concentrationb
ppm Mg/m3
0.12 ppm
O3-100ppb
S02
0.12 ppm
O3-0.12ppm
03
Air-100ppbSO2
0.35 686


0.40 784





0.12 235




0.12 235
0.20 392

0.40 784


0.20 392
0.40 784
0.40 784




— Exposure Duration Exposure
and Activity Conditions
45 min in first atmosphere 22 °C
and 1 5 min in second IE 75% RH





70 min with IE at NA
40 L/min

2 h with IE at 22 °C
VE = 53 to 55 L/min 50% RH




6. 6 h with IE at NA
«25 L/min/m2 BSA



lhat VE= 89 L/min 31 °C
followed by 3 to 4 min 35% RH
at -150 L/min
3 h/day for 5 days in a
row

2 h with IE at 2 x resting 22 °C
2 h with IE at 2 x resting 55% RH
2 h/day for 3 days




Number and Subject
Gender of Subj ects Characteristics
8 M, 5 F Asthmatic,
12 to 18 years old





18 F Healthy NS,
19 to 28 years old

8M, 10 F 9 asthmatics (5 F,
4M),
9 healthy (5 F, 4 M),
18 to 34 years old


10 M Healthy NS,
18 to 33 years old



15M,2F Elite cyclists,
19 to 30 years old

13M, 11 F Healthy NS,
1 9 to 46 years old

12 M, 7 F Healthy NS,
21 to 32 years old





Observed Effect(s)
Greater declines in FEV[ and Vmaj,50o/0 and
greater increase in respiratory resistance after
O3-SO2 than after O3-O3 or air-SO2.




PD100 decreased from 59 CIU after air
exposure to 41 CIU and 45 CIU, 1 and
18 h after O3 exposure, respectively.
Decreased PC100SRsw from 33 mg/mL to
8.5 mg/mL in healthy subjects after O3.
PC100SRllw fell from 0.52 mg/mL to
0.19 mg/mL in asthmatic subjects after
exposure to O3 and from 0.48 mg/mL
to 0.27 mg/mL after exposure to air.
Approximate doubling of mean
methacholine responsiveness after exposure.
On an individual basis, no relationship
between O3-induced changes in airway
responsiveness and FEV; or FVC.
Greater than 20% increase in histamine
responsiveness in one subject at 0.12 ppm
O3 and in nine subjects at 0.20 ppm O3.
Enhanced response to methacholine after
first 3 days, but this response normalized by
Day5.
110% increase in ASRaw to a 10-breath
histamine (1.6%) aerosol challenge after
exposure to O3 at 0.40 ppm, but no change
at 0.20 ppm. Progressive adaptation of this
effect over 3-day exposure.


Reference
Koenig et al.
(1990)





Folinsbee
and Hazucha
(1989)
Kreit et al.
(1989)




Folinsbee
etal. (1988)



Gong et al.
(1986)

Kulle et al.
(1982)

Dimeo et al.
(1981)




-------
c
S-
to
O
O















^b
to


O
H
1
O
O
0
H
O
O
H
W
O
O
H
W
i/vuj^iL /VAO-II ^coiii u). /virway responsiveness roiiowmg uzoiie exposures
Ozone
Concentration1"


ppm Mg/m3 and Activity Conditions Gender of Subjects Characteristics Observed Effect(s)
0.10 196 2h NA 14 HealthNS, Increased airway responsiveness
0.32 627 24 ± 2 years old to methacholine immediately after exposure
1.00 1,960 at the two highest concentrations of O3.
0.60 1,176 2 h with IE at 2 x resting 22 °C 11M, 5F 9 atopic, Ten-breath methacholine or histamine
55% RH 7 nonatopic, challenge increased SRaw > 150% in
NS, 21 to 35 years 16 nonasthmatics after O3. On average, the
old atopic subjects had greater responses than
the nonatopic subjects. The increased
responsiveness resolved after 24 h. Atropine
premedication blocked the O3-induced
increase in airway responsiveness.
0.6 1,176 2 hat rest NA 5 M, 3 F Healthy NS, 300% increase in histamine-induced ARaw
22 to 30 years old 5 min after O3 exposure; 84 and 50%
increases 24 h and 1 week after exposure
(p > 0.05), respectively. Two subjects had
an increased response to histamine 1 week
after exposure.
aSee Appendix A for abbreviations and acronyms.
bListed from lowest to highest O3 concentration.


















Reference
Konig et al.
(1980)

Holtzman
etal. (1979)






Golden et al.
(1978)





















-------
 1      et al., 1995), suggesting that changes in airway responsiveness and spirometric volumes occur by
 2      different mechanisms.
 3           Dimeo et al. (1981) were the first to investigate attenuation of the O3-induced increases in
 4      nonspecific airway responsiveness after repeated O3 exposure. Over 3 days of a 2 h/day
 5      exposure to 0.40 ppm O3, they found progressive attenuation of the increases in airway
 6      responsiveness such that, after the third day of O3 exposure, histamine airway responsiveness
 7      was no longer different from the sham exposure levels. Kulle et al. (1982) found that there was
 8      a significantly enhanced response to methacholine after the first 3 days of exposure, but this
 9      response slowly normalized by the end of the fifth day. Folinsbee et al. (1994) found a more
10      persistent effect of O3 on airway responsiveness which was only partially attenuated after
11      5 consecutive days of O3 exposure.
12           The occurrence and duration of increased nonspecific airway responsiveness following O3
13      exposure could have important clinical implications for asthmatics. Kreit et al. (1989)
14      investigated changes in airway responsiveness to methacholine that occur after O3 exposure in
15      mild asthmatics.  They found that the baseline PC100SRaw declined from 0.52 to 0.19 mg/mL
16      after a 2-h exposure to 0.40 ppm O3  as compared to a decline from 0.48 to 0.27 mg/mL after air
17      exposure; however, because of the large variability in responses of the asthmatics, the percent
18      decrease from baseline in mean PC100SRaw was not statistically different between healthy and
19      asthmatic subjects (74.2 and 63.5%, respectively).
20           Two studies examined the effects of preexposure to O3 on exacerbation of exercise-induced
21      bronchoconstriction (Fernandes et al.,  1994; Weymer et al., 1994). Fernandes et al. (1994)
22      preexposed subjects with stable mild asthma and a history of >15% decline in FEVj after
23      exercise to 0.12 ppm O3 for  1 h at rest  followed by a 6-min exercise challenge test and found no
24      significant effect  on either the magnitude or time course of exercise-induced
25      bronchoconstriction. Similarly, Weymer et al. (1994) observed that preexposure to either 0.10 or
26      0.25 ppm O3 for 60 min while performing light IE did not enhance or produce exercise-induced
27      bronchoconstriction in otherwise healthy adult subjects with stable mild asthma. Although the
28      results suggested  that preexposure to O3 neither enhances nor produces exercise-induced asthma
29      in asthmatic subjects, the relatively low total inhaled doses of O3 used in these studies limit the
30      ability to draw any definitive conclusions.
        August 2005                             AX6-93      DRAFT-DO NOT QUOTE OR CITE

-------
 1           Gong et al. (1997b) found that subjects with asthma developed tolerance to repeated
 2      O3 exposures in a manner similar to normal subjects; however, there were more persistent effects
 3      of O3 on airway responsiveness, which only partially attenuated when compared to filtered air
 4      controls.  Volunteer subjects with mild asthma requiring no more than bronchodilator therapy
 5      were exposed to filtered air or 0.4 ppm O3, 3 h/d for 5 consecutive days, and follow-up
 6      exposures 4 and 7 days later.  Symptom and FEVj responses were large on the 1st and 2nd
 7      exposure days,  and diminished progressively toward filtered air responses by the 5th exposure
 8      day.  A methacholine challenge was performed when postexposure FEVj returned to within 10%
 9      of preexposure baseline levels.  The first O3 exposure significantly decreased PD2QFEVJ by an
10      order of magnitude and subsequent exposures resulted in smaller decreases, but they were still
11      significantly different from air control levels. Thus, the effects of consecutive O3 exposures on
12      bronchial reactivity differ somewhat from the effects on lung function. The same conclusion
13      was drawn by Folinsbee et al. (1994) after consecutive 5-day O3 exposures in healthy subjects,
14      despite a much lower bronchial reactivity both before and after O3 exposure.
15           A larger number of studies examined the effects of O3 on exacerbation of antigen-induced
16      asthma.  Molfino et al. (1991) were the first to report the effects of a 1-h resting exposure to
17      0.12 ppm O3 on the response of subjects with mild, stable atopic asthma to a ragweed or grass
18      allergen inhalation challenge. Allergen challenges were performed 24 h after air and
19      O3 exposure. Their findings suggested that allergen-specific airway responsiveness of mild
20      asthmatics is increased after O3 exposure. However, Ball et al. (1996) and Hanania et al. (1998)
21      were unable to  confirm the findings of Molfino et al. (1991) in a group of grass-sensitive mild
22      allergic asthmatics exposed to 0.12 ppm O3 for 1 h.  The differences between Hanania et al.
23      (1998) and Molfino et al. (1991), both conducted in the same laboratory, were due to better, less
24      variable control of the 1 h 0.12 ppm O3 exposure and better study design by Hanania and
25      colleagues.  In the original, Molfino et al. (1991) study, the control (air) and experimental (O3)
26      exposures were not randomized after the second  subject because of long-lasting (3 months),
27      O3-induced potentiation of airway reactivity in that subject. For safety reasons, therefore, the air
28      exposures were performed prior to the O3 exposures for the remaining 5 of 7 subjects being
29      evaluated. It is possible that the first antigen challenge caused the significant increase in the
30      second (post-O3) antigen challenge.
        August 2005                             AX6-94      DRAFT-DO NOT QUOTE OR CITE

-------
 1           Torres et al. (1996) later confirmed that higher O3 concentrations cause increased airway
 2      reactivity to specific antigens in subjects with mild allergic asthma, and to a lesser extent in
 3      subjects with allergic rhinitis, after exposure to 0.25 ppm O3 for 3 h. The same laboratory
 4      repeated this study in separate groups of subjects with asthma and rhinitis and found similar
 5      enhancement of allergen responsiveness after O3 exposure (Holz et al.,  2002); however, the
 6      effects of a 3-h exposure to 0.25 ppm O3 were more variable, most likely due to performing the
 7      allergen challenges 20 h after exposure, rather than the 3 h used in the first study.
 8           The timing of allergen challenges in O3-exposed subjects with allergic asthma is important.
 9      Bronchial provocation with allergen, and subsequent binding with IgE antibodies on mast cells
10      in the lungs, triggers the release of histamine and leukotrienes and a prompt early-phase
11      contraction of the smooth muscle cells of the bronchi, causing a narrowing of the lumen of the
12      bronchi  and a decrease in  bronchial airflow (i.e., decreased FEVj). In many asthma patients,
13      however, the release of histamine and leukotrienes from the mast cells also attracts an
14      accumulation of inflammatory cells, especially eosinophils, followed by the production of mucus
15      and a late-phase decrease  in bronchial airflow for 4 to 8 h.
16           A  significant finding from the study by Holz et al. (2002) was that clinically relevant
17      decreases in FEVj (>20%) occurred during the early-phase allergen response in subjects with
18      rhinitis after a consecutive 4-day exposure to 0.125 ppm O3.  Kehrl et al. (1999) previously
19      found an increased reactivity to house dust mite antigen in asthmatics 16 to 18 h after exposure
20      to 0.16 ppm O3 for 7.6 hours.  These important observations indicate that O3 not only causes
21      immediate increases in airway-antigen reactivity, but that this effect may persist for at least 18 to
22      20 h.  Ozone exposure, therefore, may be a clinically important co-factor in the response to
23      airborne bronchoconstrictor substances in individuals with pre-existing allergic  asthma.  It is
24      plausible that this phenomenon could contribute to increased symptom  exacerbations and, even,
25      consequent increased physician or ER visits, and possible hospital admissions (see Chapter T).
26           A number of human studies, especially more recent ones,  have been undertaken to
27      determine various aspects of O3-induced increases in nonspecific airway responsiveness, but
28      most studies have been conducted in laboratory animals (See the toxicology chapter, Section
29      5.3.4.4.). In humans, increased airway permeability (Kehrl et al., 1987; Molfmo et al., 1992)
30      could play a role in increased airway responsiveness.  Inflammatory cells and mediators also
31      could affect changes in airway responsiveness.  The results of a multiphase study (Scannell

        August 2005                             AX6-95      DRAFT-DO NOT QUOTE OR CITE

-------
 1      et al., 1996; Balmes et al., 1997) showed a correlation between preexposure methacholine
 2      responsiveness in healthy subjects and increased SRaw caused by a 4 h exposure to 0.2 ppm O3,
 3      but not with O3-induced decreases in FEVj and FVC. The O3-induced increase in SRaw, in turn,
 4      was correlated with O3-induced increases in neutrophils and total protein concentration in BAL
 5      fluid.  Subjects with asthma had a significantly greater inflammatory response to the same O3
 6      exposures, but it was not correlated with increased SRaw, and nonspecific airway provocation
 7      was not measured. Therefore, it is difficult to determine from this series of studies if underlying
 8      airway inflammation plays a role in increased airway responsiveness to nonspecific
 9      bronchoconstrictors. The study, however, confirmed an earlier observation (e.g., Balmes et al.,
10      1996) that O3-induced changes in airway inflammation  and lung volume measurements are not
11      correlated.
12           Hiltermann et al. (1998) reported that neutrophil-derived serine proteinases associated
13      with O3-induced inflammation are not important mediators for O3-induced nonspecific airway
14      hyperresponsiveness.  Subjects with mild asthma, prescreened for O3-induced airway
15      responsiveness to methacholine, were administered an  aerosol of recombinant antileukoprotease
16      (rALP) or placebo at hourly intervals  two times before and six times after exposure to filtered air
17      or 0.4 ppm O3 for 2 h. Methacholine  challenges were performed 16 h after exposure. Treatment
18      with rALP had no effect on the O3-induced decrease in FEVj or PC2QFEVJ in response to
19      methacholine challenge. The authors speculated that proteinase-mediated tissue injury caused
20      by O3 may not be important in the development of airway hyperresponsiveness of asthmatics
21      to O3. In a subsequent study using a similar protocol (Peters et al., 2001), subjects with mild
22      asthma were administered an aerosol  of apocynin, an inhibitor of NADPH oxidase present in
23      inflammatory cells such as eosinophils and neutrophils, or a placebo. In this study, methacholine
24      challenge performed 16 h after O3 exposure showed treatment-related effects on PC^EVj,
25      without an effect on FEVj. The authors concluded that apocynin could prevent O3-induced
26      bronchial hyperresponsiveness in subjects with asthma, possibly by preventing superoxide
27      formation by eosinophils and neutrophils in the larger airways.
28           Nightingale et al. (1999) reported that exposures of healthy subjects and subjects with mild
29      atopic asthma to a lower O3 concentration (0.2 ppm) for 4 h  caused a similar neutrophilic lung
30      inflammation in both groups but no changes in airway responsiveness to methacholine measured
31      24 h after O3 exposure in either group. There were, however, significant decreases in FEVj of

        August 2005                            AX6-96     DRAFT-DO NOT QUOTE OR CITE

-------
 1      6.7 and 9.3% immediately after O3 exposure in both healthy and asthmatic subjects, respectively.
 2      In a subsequent study, a significant increase in bronchoresponsiveness to methacholine was
 3      reported 4 h after healthy subjects were exposed to 0.4 ppm O3 for 2 h (Nightingale et al., 2000).
 4      In the latter study, preexposure treatment with inhaled budesonide (a corticosteroid) did not
 5      protect against O3-induced effects on spirometry, methacholine challenge, or sputum neutrophils.
 6      These studies also confirm the earlier reported findings that O3-induced increases in airway
 7      responsiveness usually resolve by 24 h after exposure.
 8           Ozone-induced airway inflammation and hyperresponsiveness were used by Criqui et al.
 9      (2000) to evaluate anti-inflammatory properties of the macrolide antibiotic, azithromycin.  In a
10      double-blind, cross-over study, healthy volunteers were exposed to 0.2 ppm O3 for 4 h after
11      pretreatment with azithromycin or a placebo.  Sputum induction 18 h postexposure resulted in
12      significantly increased total cells, percent neutrophils, IL-6, and IL-8 in both azithromycin- and
13      placebo-treated subjects. Significant pre- to postexposure decreases in FEVj and FVC also were
14      found in both subject groups. Airway responsiveness to methacholine was not significantly
15      different between azithromycin-treated and placebo-treated subjects when they were challenged
16      2 h after postexposure FEVj decrements returned to within 5 % of baseline.  Thus,  azithromycin
17      did not have anti-inflammatory effects in this study.
18           The effects of dietary antioxidants on O3-induced bronchial responsiveness to SO2
19      provocation were evaluated in adult asthmatic subjects by Trenga et al. (2001).  This study and
20     potential interpretative problems are discussed in detail in Section AX6.5.6. Briefly, 17 adult
21      asthmatic subjects sensitive to SO2 provocation took vitamin supplements (400 IU vitamin E and
22      500 mg vitamin C) or placebo once a day for 5 weeks.  After the fourth and fifth weeks of
23      vitamin or placebo, subjects were randomly exposed to FA and 0.12 ppm O3 for 45 min during
24      IE (VE « 3x resting rate) followed by two sequential 10 min exposures to 0.1 and 0.25 ppm SO2.
25      Vitamin treatment was not associated with decreased bronchial responsiveness following the
26      0.1 ppm SO2 challenge.  However, the change in spirometric responses (FEVl3 FVC, FEF25.75,
27      and PEF) between the 0.1 and 0.25 ppm  SO2 challenges were more severe for the placebo than
28      the vitamin treatment regimen (p = 0.009). The authors concluded O3 exposure increases
29      bronchial responsiveness to SO2 in asthmatics and that antioxidant supplementation has a
30      protective effect against this responsiveness.
31

        August 2005                             AX6-97      DRAFT-DO NOT QUOTE OR CITE

-------
 1     AX6.9  EFFECTS ON INFLAMMATION AND HOST DEFENSE
 2     AX6.9.1  Introduction
 3           In general, inflammation can be considered as the host response to injury, and the
 4     induction of inflammation can be accepted as evidence that injury has occurred. Several
 5     outcomes are possible:  (1) inflammation can resolve entirely; (2) continued acute inflammation
 6     can evolve into a chronic inflammatory state; (3) continued inflammation can alter the structure
 7     or function of other pulmonary tissue, leading to diseases such as fibrosis or emphysema;
 8     (4) inflammation can alter the body's host defense response to inhaled microorganisms;  and
 9     (5) inflammation can alter the lung's response to other agents such as allergens or toxins.
10     At present, it is known that short-term exposure of humans to O3 can cause acute inflammation
11     and that long-term exposure of laboratory animals results in a chronic inflammatory state (see
12     Chapter 5). However, the relationship between repetitive bouts of acute inflammation in humans
13     caused by O3 and the development of chronic respiratory disease is unknown.
14           Bronchoalveolar lavage (BAL) using fiberoptic bronchoscopy has been utilized to sample
15     cells and fluids lining the respiratory tract primarily from the alveolar region, although the use of
16     small volume lavages or balloon catheters permits sampling of the airways. Cells  and fluid can
17     be retrieved from the nasal passages using nasal lavage (NL) and brush or scrape biopsy.
18           Several studies have analyzed BAL and NL fluid and cells from O3-exposed humans for
19     markers of inflammation and lung damage (see Tables AX6-12 and AX6-13). The presence of
20     neutrophils (PMNs) in the lung has long been accepted as a hallmark of inflammation and is an
21     important indicator that O3 causes inflammation in the lungs.  It is apparent, however, that
22     inflammation within airway tissues may persist beyond the point that inflammatory cells are
23     found in BAL fluid.  Soluble mediators of inflammation such as the cytokines (IL-6, IL-8) and
24     arachidonic acid metabolites  (e.g., PGE2, PGF2(X, thromboxane, and leukotrienes [LTs] such as
25     LTB4) have been measured in the BAL fluid of humans exposed to O3. In addition to their role
26     in inflammation, many of these compounds have bronchoconstrictive properties and may be
27     involved in increased airway responsiveness following O3 exposure.
28           Some recent evidence suggests that changes in small airways function may provide a
29     sensitive indicator of O3 exposure and effect (see Section AX6.2.5), despite the fact that inherent
30     variability in their measurement by standard spirometric approaches make their assessment
31     difficult. Observations of increased functional responsiveness of these areas relative to the more

       August 2005                             AX6-98      DRAFT-DO NOT QUOTE OR CITE

-------
             Table AX6-12. Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozone"
to
o
o
X
Ozone Concentration11
ppm |ig/m3 Duration
Activity Level
(VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Upper Airway Studies
0.4 784 2 h


0.2 392 2h
0.4 980 2h
0.12 235 1.5 h
0.24 470
0.5 980 4h
0.4 784 2 h

0.5 980 4hon
2 consecutive days
At rest


IE
(15 min/30 min);
(VE) « 20
L/min/m2 BSA
At rest
IE
(20 L/min) at
15 -min intervals
Resting
IE
(70 L/min) at
15 -min intervals
Resting
12 mild,
asymptomatic dust
mite-sensitive
asthmatics;
18-35 years of age

8 M, 5 F healthy NS
20-3 1 years of age
10 mild NS
asthmatics
18-3 5 years old
5 M, 5 F, asthmatic;
4 M, 4 F,
nonasthmatic;
18 to 41 years old
6 M, 6 F,
allergic rhinitics,
31.4 ± 2.0 (SD)
years old
11 M,
18 to 35 years old
41 M
(21 O3-exposed,
20 air-exposed),
18 to 35 years old
Release of early-onset mast cell-derived mediators into NL in
response to allergen not enhanced following O3 exposure.
Neutrophil and eosinophil inflammatory mediators were not
increased after O3 exposure or enhanced after allergen
challenge. O3 increased eosinophil influx following allergen
exposure.
No neutrophilia in NL samples by 1 .5 h postexposure.
Depletion of uric acid in NL fluid by 30% during h 2 of
exposure with increase in plasma uric acid levels.
No depletion of ascorbic acid, reduced glutathione,
extracellular superoxide dismutase.
Response to allergen increased (NS). PMN and
eosinophils increased after O3 plus allergen challenge.
Ozone alone increased inflammation in the nose.
NL done immediately and 24 h after exposure.
Increased number of PMNs at both times in asthmatic subjects
exposed to 0.24 ppm O3; no change in nonasthmatic subjects.
No change in lung or nasal function.
NL done immediately after exposure. Increased upper
and lower respiratory symptoms and increased levels of PMNs,
eosinophils, and albumin in NL fluid.
NL done immediately before, immediately after, and 22 h after
exposure. Increased numbers of PMNs at both times after
exposure; increased levels of tryptase, a marker of mast cell
degranulation, immediately after exposure; increased levels
of albumin 22 h after exposure.
NL done immediately before and after each exposure
and 22 h after the second exposure. Increased levels of PMNs
at all times after the first exposure, with peak values occurring
immediately prior to the second exposure.
Michelson
etal. (1999)

Mudway et al.
(1999)
Peden et al.
(1995)
McBride et al.
(1994)
Bascom et al.
(1990)
Graham and
Koren(1990)
Koren et al.
(1990)
Graham et al.
(1988)

-------
         Table AX6-12 (cont'd).  Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozonea
to
o
o
X
O
O
Ozone Concentrationb
ppm ug/m3 Duration (VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Lower Airway Studies
0.2 392 2h IE
(15 min/30 min);
(VE) - 20
L/min/m2 BSA

0.1 196 2h mild IE





0.2 392 2h IE
(15 min/30 min);
(VE) - 20
L/min/m2 body
surface area
0.27 529 2h IE
(20 min/ 60 min);
(VE) - 25
L/min/m2 BSA

0.22 431 4 h IE
(15 min/30 min);
(VE) = 25
L/min/m2 BSA





6M, 6F healthy,
nonatopic and
9 M, 6F mild
asthmatic subjects,
19-48 years of age
12 M, 10 F
healthy subjects
mean age -30
years


6M, 9F healthy
subjects and
9 M, 6F mild
asthmatics

12 subjects with
intermittent-mild
asthma exhibiting a
dual response; 18-37
years of age
12 nonsmoker,
nonresponders;
1 3 nonsmoker,
responders;
13 smokers;
1 8-40 years of age



Significantly higher baseline expression of IL-4 and IL-5 in
bronchial mucosal biopsies from asthmatic vs. healthy subjects
6 h postexposure. Following O3 exposure, epithelial
expression of IL-5, GM-CSF, ENA-78, and IL-8 increased
significantly in asthmatics, as compared to healthy subjects.
Markers of exposure in exhaled breath condensate, including
increased 8-isoprostane, TEARS and LTB-4, and a marker of
ROS-DNA interaction in peripheral blood leukocytes
(8-OHdG), were increased in a sub-set of subjects bearing the
wild genotype for NAD(P)H:quinone oxidoreductase and the
null genotype for glutathione-S-transferase Ml .
No evidence seen for increased responsiveness to the
inflammatory effects of O3 in mild asthmatics versus healthy
subjects at 6 h following exposure. Used neutrophil
recruitment and exacerbation of pre-existing inflammation.

Exposure to O3 24 h following allergen challenge resulted in a
significant decrease in FEV1, FVC and VC and increase in
symptom scores compared to air exposure. The percentage of
eosinophils, but not neutrophils, in induced sputum was higher
6 h after O3 than after air.
Recovery of AM was approximately 3-fold higher in BAL
from smokers versus nonsmokers. Unstimulated AM from
smokers released ~2-fold greater amounts of superoxide anion
than from nonsmokers at 30 min and 18 h postexposure, but
release was not further enhanced by stimulation of the cells.
ROS generation by AM from nonsmokers decreased following
exposure at 18 h; markers of epithelial permeability increased.
No relationship was found between measures of ROS
production and lung function responsiveness to O3.
Bosson et al.
(2003)



Corradi et al.
(2002)




Stenfors et al.
(2002)



Vagaggini
et al. (2002)



Voter et al.
(2001)








-------
         Table AX6-12 (cont'd).  Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozonea
to
o
o
X
Ozone Concentrationb
ppm ug/m3 Duration
Activity Level
(VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Lower Airway Studies (cont'd)
0.2 392 2h





0.4 784 2 h



0.4 784 2 h





0.4 784 1 h




0.2 392 2h

0.4 784 2 h/day for 5 days,
2 h either 10 or
20 days later




IE
(15 min/30 min);
(VE) - 20
L/min/m2 BSA


IE
(15 min/30 min);
(VE) - 20
L/min/m2 BSA
At rest





Continuous
exercise;
(VE) - 30
L/min/m2 BSA



IE
(40 L/min) at
15 -min intervals




8M, 5F healthy
nonsmokers;
20-3 1 years of age



10M, 6F subjects
with intermittent
asthma;
19-35 years of age
12 mild,
asymptomatic dust
mite-sensitive
asthmatics;
18-35 years of age

4 healthy subjects




15 healthy
nonsmokers
16 M; 18 to 35 years
of age





Early (1 .5 h postexposure) increase in adhesion molecule
expression, submucosal mast cell numbers and alterations in
lining fluid redox status. No clear relationship between early
markers of response and lung function deficits. 2.5-fold
increase in % human leukocyte antigen (HLA)-DR+ alveolar
macrophages in BAL.
In a cross-over study, levels of eosinophil cationic protein, IL-8
and percentage eosinophils were found to be highly correlated
in induced sputum and BAL 16 h following O3 exposure.

Release of early-onset mast cell-derived mediators into NL in
response to allergen not enhanced following O3 exposure.
Neutrophil and eosinophil inflammatory mediators were not
increased after O3 exposure or enhanced after allergen
challenge. O3 increased eosinophil influx following allergen
exposure.
Apoptosis was observed in cells obtained by airway lavage
6 h following exposure. AM obtained by BAL showed the
presence of a 4-hydroxynonenal (HNE) protein adduct and the
stress proteins, 72-kD heat shock protein and ferritin. These
effects were replicated by in vitro exposure of AM to HNE.
Increased numbers of CD3+, CD4+, and CD8+ T lymphocyte
subsets, in addition to neutrophils, in BAL 6 h postexposure.
BAL done immediately after fifth day of exposure and again
after exposure 10 or 20 days later. Most markers of
inflammation (PMNs, IL-6, PGE2, fibronectin) showed partial
to complete attenuation; markers of damage (LDH, IL-8,
protein, al-antitrypsin, elastase,) did not. Reversal of
attenuation was not complete for some markers, even after
20 days.
Blomberg
etal. (1999)




Hiltermann
etal. (1999)


Michelson
etal. (1999)




Hamilton et al.
(1998)



Blomberg
etal. (1997)
Devlin et al.
(1997)






-------
         Table AX6-12 (cont'd).  Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozone"
to
o
o
X
O
to
Ozone Concentration11
ppm
Hg/m3
Exposure
Duration
Activity Level
(VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Lower Airway Studies (cont'd)
0.22







0.12



0.16


0.2


0.4



0.4





0.2


431







235



314


392


784



784





392


4h







2h



7.6 h


4h
T = 20 °C
RH = 50%
2h
T = 22 °C
RH = 50%

2h
1 5 min,
ex/ 15 min, rest



4h
T = 20 °C
RH = 50%
IE
20 min ex/1 9 min
rest
(VE) * 39-45 1/min




IE
(15 min/30 min);
(VE) - 20
L/min/m2 BSA
IE
50 min/hr
(VE) = 25L/min
IE
(50 min/60 min);
(VE)«44L/min
1 5 min rest
1 5 min exercise
cycle ergometer
(VE) « 55 1/min
(VE) = 66 1/min





IE
(50 min/60 min);
(VE)«44L/min
31M,7F
smokers and
nonsmokers





9M, 3F healthy
nonsmokers; mean
age -28 years

8 asthmatics sensitive
to dust mites

14 M, 6 F
healthy NS

1 1 healthy
nonsmokers;
18-3 5 years

8M
healthy nonsmokers




17M,6F
mild asthmatics

Post-O3 exposure FEV; in 3 groups: Smokers (-13.9%);
nonresponders (- 1 .4%) and responders (-28.5%). PMN's
increased immediately and at 18 h in all groups. Eosinophils
and lymphocytes increased after O3. IL-6 increased more in
nonsmokers. No relationship of symptoms with inflammation,
lung function changes not related to inflammation. Nasal
lavage indicators did not predict bronchial or alveolar
inflammation.
Increase in the percentage of vessels expressing P-selectin in
bronchial biopsies at 1 .5 h postexposure. No changes in FEVj,
FVC, inflammatory cells or markers in BAL, or vessels
expressing VCAM-1, E-selectin or ICAM-1 in biopsies.
Increased numbers of eosinophils in BAL after O3 exposure.


Ozone increased PMN, protein, IL-8, for all subjects.
No relationship of inflammation with spirometric responses.

Mean FEVl5 change = - 10%. BAL occurred at 0, 2, or 4 h
postexposure. Small n limits statistical inference. Trend for
PMN's to be highest at 4 h. LTC4 increased at all time points.
No change in PGE2 or thromboxane.
Comparison of BAL at 1 h postexposure vs. 18 h postexposure.
At 1 h, PMN's, total protein, LDH, al-antitrypsin, fibronectin,
PGE2, thromboxane B2, C3a, tissue factor, and clotting factor
VII were increased. IL-6 and PGE2 were higher after 1 h than
18 h. Fibronectin and tissue plasminogen activator higher after
18 h. No time differences for PMN and protein.
Increased PMN, protein, IL-8, LDH, in BAL. Inflammatory
responses were greater than a group of nonasthmatics
(Balmes et al., 1996)
Frampton
etal. (1997a)
Torres et al.
(1997)




Krishna et al.
(1997b)


Peden et al.
(1997)

Balmes et al.
(1996)

Coffey et al.
(1996)


Devlin et al.
(1996)
(compare with
Koren et al.
(1989a)

Scannell et al.
(1996)


-------
         Table AX6-12 (cont'd).  Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozone"
to
o
o
X
Ozone Concentration11
ppm ug/m3
Exposure
Duration
Activity Level
(VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Lower Airway Studies (cont'd)
0.4 784
0.2 392
0.08 157
0.10 196
0.4 784
0.3 588
0.40 784
0.40 784
0.4 784
2 h mouthpiece
exposure
20 °C
42% RH
4h
6.6 h
2h
Ih
(mouth-piece)
2h
2h
2h
1 5 min exercise
1 5 min rest
(VE) « 40 1/min
IE
(50 min/60 min);
(VE) = 40 L/min
IE
(50 min/60 min) +
35 min lunch;
(VE) = 40 L/min
IE
(15 min/30min);
(VE) = 70 L/min
CE (60 L/min)
IE
(15 min/30 min);
(VE) = 70 L/min
IE
(15 min/30 min);
(VE) = 70 L/min
IE
(15 min/30 min);
(VE) = 70 L/min
5M, 5F
healthy; age ~ 30
15 M, 13 F,
21 to 39 years old
18 M,
18 to 35 years of age
10 M,
18 to 35 years old
5M
11M, 18 to
35 years old
11 M,
18 to 35 years old
11 M,
18 to 35 years old
Sputum induction 4 h after O3 exposure 3-fold increase in
neutrophils and a decrease in macrophages after O3 exposure.
IL-6, IL-8, and myeleperoxidase increased after O3. Possible
relationship of IL-8 and PMN levels.
Bronchial lavage, bronchial biopsies, and BAL done 1 8 h
after exposure. BAL shows changes similar to other studies.
Airway lavage shows increased cells, LDH, IL-8. Biopsies
show increased number of PMNs.
BAL fluid 18 h after exposure to 0.1 ppm O3 had significant
increases in PMNs, protein, PGE2, fibronectin, IL-6, lactate
dehydrogenase, and a-1 antitrypsin compared with the same
subjects exposed to FA. Similar but smaller increases in all
mediators after exposure to 0.08 ppm O3 except for protein and
fibronectin. Decreased phagocytosis of yeast by alveolar
macrophages was noted at both concentrations.
BAL fluid 1 h after exposure to 0.4 ppm O3 had significant
increases in PMNs, protein, PGE2, TXB2, IL-6, LDH, a-1
antitrypsin, and tissue factor compared with the same subjects
exposed to FA. Decreased phagocytosis of yeast by alveolar
macrophages.
Significantly elevated PMNs in the BAL fluid 1, 6, and 24 h
after exposure, with peak increases at 6 h.
Macrophages removed 18 h after exposure had changes in the
rate of synthesis of 123 different proteins as assayed by
computerized densitometry of two-dimensional gel protein
profiles.
BAL fluid 18 h after exposure contained increased levels of the
coagulation factors, tissue factor, and factor VII. Macrophages
in the BAL fluid had elevated tissue factor mRNA.
BAL fluid 18 h after exposure had significant increases in
PMNs, protein, albumin, IgG, PGE2, plasminogen activator,
elastase, complement C3a, and fibronectin.
Fahy et al.
(1995)
Aris et al.
(1993)
Devlin et al.
(1990,1991)
Koren et al.
(1991)
Koren et al.
(1991)
Schelegle
etal. (1991)
Devlin and
Koren (1990)
McGee et al.
(1990)
Koren et al.
(1989a,b)

-------
         Table AX6-12 (cont'd).  Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozonea
to
o
o
X
Ozone Concentration11
ppm ug/m3 Duration
Activity Level
(VE)
Number and Gender
of Subjects
Observed Effect(s)
Reference
Repeated Exposure Studies
0.125 245 3 h exposures to
0.25 490 both O3 cones, and
to FA;
3 h on four
consecutive days to
0.125; study arms
separated by
>4 wks
0.25 490 2 h on four
consecutive days;
O3 and FA exposure
study arms
separated by
>3 wks


0.2 392 single,
4 h exposures to O3
and to FA;
4 h on four
consecutive days to
O3; study arms
separated by
>4 wks

0.2 392 single, 4 h
exposure;
4 h exposures on
four consecutive
days; study arms
separated by
>4 wks
IE
(15 min/30 min)






IE
(30 min/60 min);
(VE) ~ 8 times the
FVC/min




IE
(15 min/30 min);
(MeanVE) =
14.8 L/min/m2
BSA




IE
(30 min/60 min);
(MeanVE) = 25
L/min/m2 BSA



5M, 6F allergic
asthmatic and
16M, 6F allergic
rhinitic subjects;
19-53 years of age



5M, 3F healthy
subjects;
25-31 years of age





15M, 8F healthy
subjects;
21-35 years of age






9M, 6F healthy NS
23-37 years of age





All subjects underwent 4 exposure arms and were challenged
with allergen 20 h following the last exposure in each. Sputum
was induced 6-7 h later. In rhinitics, but not asthmatics, the
incidence and magnitude of early phase FEV; decrements to
Ag were greater after 0.25 and 4x 0. 125 ppm O3. Repeated
exposure caused increases in neutrophil and eosinophil
numbers in both subject groups, as well as increased
percentage and number of lymphocytes in the asthmatics.
Maximal mean reductions in FEV; and FVC were observed on
day 2, and became negligible by day 4. FEF25.75, VmaxSO, and
Vmax75 were combined into a single value representing small
airway function (SAWgrp). This variable was the only one to
show persistent depression of the 24 h postexposure baseline
from day 2 to day 5 measurements. Numbers of PMNs in BAL
fluid on day 5 were significantly higher in subjects following
O3, compared to air, exposures.
All subjects underwent 3 exposure arms with BAL and
bronchial mucosal biopsies performed 20 h following the last
exposure in each. After repeated exposure, functional and
BAL cellular responses were not different from those after FA,
whereas total protein, IL-6, IL-8, reduced glutathione and
ortho-tyrosine remained elevated. Also at this time,
macroscopic scores of inflammation and tissue neutrophils
were increased in mucosal biopsies. IL-10 was detected only
in BAL fluid following repeated O3 exposure.
Subjects were randomly assigned to each of the exposure
regimens in a crossover design. Compared to single exposure,
repeated exposure resulted in an initial progression followed by
an attenuation of decrements in FEVj, FVC and specific
airways resistance by day 4. Bronchial and BAL washings
showed decreases in the numbers of PMNs and fibronectin
levels and IL-6 was decreased in BAL fluid on day 4.
Holz et al.,
(2002)






Frank et al.
(2001)






Jorres et al.
(2000)







Christian et al.
(1998)






-------
to
o
o
            Table AX6-12 (cont'd). Studies of Respiratory Tract Inflammatory Effects from Controlled Human Exposure to Ozonea
Ozone Concentration11
ppm |ig/m3 Duration
Repeated Exposure Studies (cont'd)
0.4 784 2 h/day for 5 days,
2 h either 10 or
20 days later
Activity Level
(VE)
IE
(60 L/min) at
1 5 -min intervals
Number and Gender
of Subjects
16 M;
18 to 35 years of age
Observed Effect(s)
BAL done immediately after fifth day of exposure and again
after exposure 10 or 20 days later. Most markers of
inflammation (PMNs, IL-6, PGE2, fibronectin) showed
Reference
Devlin et al.
(1997)
                                                                                   complete attenuation; markers of damage (LDH, IL-8, protein,

                                                                                   al-antitrypsin, elastase) did not. Reversal of attenuation was

                                                                                   not complete for some markers, even after 20 days.
X
0.40 784 2 h
0.60


IE
(83 W for women,
1 00 W for men)
at 1 5-min intervals
7M, 3F
23 to 41 years of age


BAL fluid 3 h after exposure had significant increases
in PMNs, PGE2, TXB2, and PGF2ct at both O3 concentrations.


Seltzer et al.
(1986)


        * See Appendix A for abbreviations and acronyms.

        b Listed from lowest to highest O3 concentration.
H

b
o


o
H

O


O
H
W

O


O

-------
         Table AX6-13. Studies of Effects on Host Defense, on Drug Effects and Supportive In Vitro Studies Relating to Controlled

                                                   Human Exposure to Ozone"
to
o
o
X
Oi
O
Oi
H

b
o


o
H

O

o
H
W

O


O
HH
H
W
Ozone
Concentration13
ppm |ig/m3 Duration
Activity Level Number and
(VE) Gender of Subjects
Observed Effect(s)
Reference
Host Defense
0.2 392 2 h
0.3 588 6 h/day for
5 consecutive
days
0.2 382 2 h

IE(15min/30 4M, 5F
min); mild atopic
(VE) ~ 20 asthmatics;
L/min/m2 BSA 2 1 -42 years of age
IE (light treadmill) 24 M
(12 O3, 12 air)
IE 10M,2F
(15 min/30 min); healthy NS
( VE) -30 L/min mean -28 years
of age

A significant decline in FEV; and VC immediately following exposure.
A 2-fold increase in percent PMNs, with no changes in other
biomarkers, was observed at 6 h postexposure. By 24 h postexposure,
PMNs had decreased, but albumin, total protein, myeloperoxidase and
eosinophil cationic protein had increased.
Subjects inoculated with type 39 rhinovirus prior to exposure. NL was
performed on the morning of Days 1 to 5, 8, 15, and 30. No difference
in virus titers in NL fluid of air and O3 -exposed subjects at any time
tested. No difference in PMNs or interferon gamma in NL fluid, or in
blood lymphocyte proliferative response to viral antigen.
Subjects were exposed to O3 and FA in a cross-over design and
underwent BAL 6 h postexposure. O3 exposure induced a 3-fold
increase in % PMNs and epithelial cells, and increased IL-8, Gro-a,
and total protein in BAL fluid. % PMNs correlated positively with
chemokine levels. Exposure also resulted in a significant decrease in
the CD4+/CD8+ ratio and the % of activated CD4+ and CD8+ T cells
in BAL fluid.
Newson et al.
(2000)
Henderson
etal. (1988)
Krishna et al.
(1998)

Host Defense - Mucous Clearance
0.4 784 1 h
0.20 392 2 h
0.40 784
CE (40 L/min) 1 5 healthy NS
18 to 35 years old
IE (light treadmill) 7 M,
27.2 ± 6.0 (SD)
years old
Subjects inhaled radiolabeled iron oxide particles 2 h after exposure.
No significant O3 -induced effect on clearance of particles during the
next 3 h or the following morning.
Subjects inhaled radiolabeled iron oxide particles immediately before
exposure. Concentration-dependent increase in rate of particle
clearance 2 h after exposure, although clearance was confined
primarily to the peripheral airways at the lower O3 concentration.
Gerrity et al.
(1993)
Foster et al.
(1987)

-------
          Table AX6-13 (cont'd). Studies of Effects on Host Defense, on Drug Effects and Supportive In Vitro Studies Relating to

                                              Controlled Human Exposure to Ozone"
to
o
o
X

 I

o
H

6
o


o
H

O

o
H
W

O


O
HH
H
W
Ozone
Concentration13


ppm
77-,^*,^^-,-.-.^^

ug/m3 Duration
Activity Level
(VE)
Number and
Gender of Subjects

Observed Effects)

Reference
Host Defense - Epithelial Permeability
0.15
0.35


0.5


0.4


294 130min
686


784 2.25 h


784 2h


IE
10 exercise/
10 rest
(VE) - SxFVC
IE
(70 L/min) at
1 5-min intervals
IE
(70 L/min) at
1 5-min intervals
8M,1F
NS


16 M,
20 to 30 years old

8M,
20 to 30 years old

Subjects inhaled 99mTc-DTPA 19 h after exposure to O3. Clearance
was increased in the lung periphery. Clearance was not related to
spirometry.

Similar design and results as earlier study (Kehrl et al., 1987).
For the combined studies the average rate of clearance was 60% faster
in O3-exposed subjects.
Subjects inhaled "°Tc-DTPA 75 min after exposure. Significantly
increased clearance of 99mTc-DTPA from the lung in O3-exposed
subjects. Subjects had expected changes in FVC and SRaw.
Foster and
Stetkiewicz
(1996)

Kehrl et al.
(1989)

Kehrl et al.
(1987)

Drug Effects on Inflammation
0.25








0.4






490 3hIE
1 5-min
intervals

403
exposures:
screening,
placebo, and
two treatments
784 2h






27 °C
56 % RH
(values from Holz
etal. 1999)












14M, 4F
Healthy NS
ozone responders
3 1.4 ±8.4 years old





23 healthy adults






On average, the screening and placebo O3 exposures caused greater
than a 9-fold increase in sputum neutrophils relative to baseline levels.
Relative to placebo, the inhaled or oral corticosteroids significantly
reduced neutrophil levels by 62 and 64%, respectively. Post-O3,
spirometry not significantly different from baseline.




Subjects were exposed to O3 following random selection for a 2 wk
daily regimen of antioxidants, including vegetable juice high in the
carotenoid, lycopene, or placebo. Concentrations of lycopene in the
lungs of supplemented subjects increased by 12% following treatment.
Supplemented subjects showed a 20% decrease in epithelial cell DNA
damage as assessed by the Comet Assay. Effects attributable to
lycopene could not be separated from those of other antioxidants.
Holz et al.
(2005)







Arab et al.
(2002)






-------
          Table AX6-13 (cont'd). Studies of Effects on Host Defense, on Drug Effects and Supportive In Vitro Studies Relating to

                                              Controlled Human Exposure to Ozone"
to
o
o
X
O
OO
H

6
o


o
H

O

o
H
W

O


O
HH
H
W
Oz one
Concentration13
ppm |ig/m3 Duration
Activity Level
(VE)
Number and
Gender of Subjects
Observed Effect(s)
Reference
Drug Effects on Inflammation (cont'd)
0.0 0 2hIE
0.4 784 20 min
mild-mod.
exercise,
10 min rest


0.2 392 2 h
All exposures
separated by at
least 2 wks
(mean ~ 30d)


0.4 784 2 h





0.27 529 2 h
All exposures
separated by at
least 1 wk
(mean ~ 14 d)
0.4 784 2 h






4M, 5F






IE
(15 min/30min);
(VE) - 20
L/min/m2 BSA



IE
1 5-min intervals;
VE ~ 20 L/min/m2
BSA


Continuous
exercise;
(VE) - 25
L/min/m2 BSA

IE
1 5-min intervals
VEmin
~ 30 L/min



Healthy NS
30 ± 3 years old





Healthy (6 M, 9 F)
and mild asthmatic
(9 M, 6 F) subjects




Placebo group
15 M, IF
Antioxidant group
13M,2F
Mean age 27 years

7 M, F subjects
with mild asthma;
20-50 years of age


5 M, 4 F healthy
6 M, 7 F asthmatics





Subjects previously in Nightingale et al. (2000) study. Placebo-
control: Immediately postexposure decrements in FVC (9%) and
FEV; (14%) relative to pre-exposure values. FEV; decrement only
9% at 1 hr postexposure. By 3 h postexposure, recovery in FVC to
97% and FEV[ to 98% of preexposure values. Significant increases in
8-isoprostane at 4 h postexposure. Budesonide for 2 wk prior to
exposure did not affect responses.
Comparison was made of responses in healthy subjects, who had
higher basal ascorbate (ASC) levels and lower glutathione disulfide
(GSSG) levels than those of asthmatics. 6 h after exposure, ASC levels
were decreased and GSSG levels were increased in BAL fluid of
normals, but not asthmatics. Despite these differences in basal
antioxidant levels and response to O3, decrements in FEV; and
neutrophil influx did not differ in the two subject groups.
All subjects were exposed to FA and then entered a 2 wk regimen
of placebo or 250 mg Vit C, 50IU a-tocopherol, and 12 oz veg.
cocktail/day prior to O3 exposure. O3-induced decrements in FEV[ and
FVC were 30% and 24% less, respectively, in supplemented subjects.
Percent neutrophils and IL-6 levels in BAL fluid obtained 1 h
postexposure were not different in the two treatment groups.
Subjects were randomly exposed to FA and to O3 before and after
4 wks of treatment with 400 |ig budesonide, b.i.d. Budesonide did not
inhibit the decrement in FEV[ or increase in symptom scores, but
significantly reduced the increase in % neutrophils and IL-8 in sputum
induced 6 h postexposure.
Subjects were pretreated for 3 days prior to exposure with
indomethacin (75 mg/day) or placebo. Similar reductions in FEV; and
FVC were seen in both groups following placebo, whereas mid-flows
showed greater decline in asthmatics than normals. Indomethacin
attenuated decrements in FEV[ and FVC in normals, but not
asthmatics. Attenuation of decrements was seen for FEF60p in
asthmatics and for FEF50 in normals.
Montuschi et al.
(2002)





Mudway et al.
(2001)





Samet et al.
(2001)
Stech-Scott
et al. (2004)


Vagaggini et al.
(2001)



Alexis et al.
(2000)






-------
S-
to

o
          Table AX6-13 (cont'd). Studies of Effects on Host Defense, on Drug Effects and Supportive In Vitro Studies Relating to

                                              Controlled Human Exposure to Ozonea
X
O
VO
H

6
o


o
H

/O

o
H
W

O


O
HH
H
W
Ozone
Concentration11
ppm ug/m3 Duration
Activity Level
(VE)
Number and
Gender of Subjects
Observed Effect(s)
Reference
Drug Effects on Inflammation (cont'd)
0.4 784 2 h






0.0 784 2hIE
0.4 4xl5minat
VE = 18
L/min/m2 BSA

2 exposures:
25% subjects
exposed to air-
air, 75% to
03-03
0.4 784 2 h





0.4 784 2 h



0.35 686 1 h



IE
(20 min/30 min);
workload @
50 watts



21 °C
40% RH








IE
(60 L/min) at
1 5-min intervals



IE
(15 min/ 30 min);
(VE)=30
L/min/m2 BSA
Continuous
exercise;
(VE)35 years). Sufentanil, a narcotic analgesic, largely
abolished symptom responses and improved FEV; in strong
responders. Naloxone, an opioid antagonist, did not affect O3 effects in
weak responders. See Section AX6. 2. 5. 1




Subjects given 800 mg ibuprofen or placebo 90 min before exposure.
Subjects given ibuprofen had less of a decrease in FEV[ after O3
exposure. BAL fluid 1 h after exposure contained similar levels of
PMNs, protein, fibronectin, LDH, a-1 antitrypsin, LTB4, and C3a in
both ibuprofen and placebo groups. However, subjects given ibuprofen
had decreased levels of IL-6, TXB2, and PGE2.
Subjects received either placebo or 150 mg indomethacin/day four
days prior to O3 exposure. Indomethacin treatment attenuated the
O3-induced decrease in FEV1; but had no effect on the O3-induced
increase in Mch responsiveness.
In a placebo- and air-controlled random design, subjects were treated
with 75 mg indomethacin every 12 h for 5 days prior to exposure.
Indomethacin significantly reduced O3-induced decrements in FEV;
and FVC.
Nightingale
et al. (2000)





Passannante
etal. (1998)








Hazucha et al.
(1996)




Ying et al.
(1990)


Schelegle et al.
(1987)



-------
S-
to

o
          Table AX6-13 (cont'd). Studies of Effects on Host Defense, on Drug Effects and Supportive In Vitro Studies Relating to

                                              Controlled Human Exposure to Ozonea
X
H

6
o


o
H

/O

o
H
W

O


O
HH
H
W
Ozone
Concentrationb
ppm
ug/m3 Duration
Activity Level
(VE)
Number and
Gender of Subjects
Observed Effect(s)
Reference
Supportive In Vitro Studies
0.01 to
0.10

0.1

0.2
1
0.01 to
0.10
0.12
0.24
0.50

19.6 to 6h
196

196 24 h

392 3h
1,690 4h
19.6 to 6h
196
235 3h
470
980

bronchial
epithelial cells

Nasal mucosa

Nasal epithelial
cells and airway
epithelial cell line
Macrophage-like
THP-1 cells
bronchial
epithelial cells
Nasal epithelial
cells

Nonatopic,
nonasthmatic and
atopic, mild
asthmatic bronchial
biopsy samples
Allergic and
nonallergic patients



Nonatopic,
nonasthmatic and
atopic, mild
asthmatic bronchial
biopsy samples


Exposure to 0.01-0.10 ppm O3 significantly decreased the electrical
resistance of cells from asthmatic sources, compared to nonasthmatic
sources. This range of O3 concentrations also increased the movement
of 14C-BSA across the confluent cultures of "asthmatic" cells to an
extent that was greater than that in "nonasthmatic" cells.
Increased concentrations of neurokinin A and substance P in medium
following O3 exposure. Levels of release of both neuropeptides were
higher from tissues derived from allergic compared to nonallergic
patients.
Synergistic effect of O3 exposure on rhino virus- induced release of IL-8
at 24 h through mechanisms abrogated by antioxidant pretreatment.
Additive enhancement of ICAM-1 expression.
THP-1 cells were treated with samples of human surfactant protein A
(SP-A) genetic variants (SP-A1 and SP-A2) that had been previously
exposed to O3. O3-exposed variants differed in their ability to stimulate
the production of TNF-a and IL-8 by these cells.
No difference in constitutive release of IL-8, GM-CSF, si CAM- 1 and
RANTES from cells from nonasthmatic and asthmatic sources, except
for detection of RANTES in latter cells only. Increased release of all
mediators 24 h after 0.05 to 0.10 ppm O3 in "asthmatic" cells, but only
IL-8 and si CAM- 1 in "nonasthmatic" cells.
Small dose-response activation of NF-KB coinciding with O3-induced
production of free radicals assessed by electron spin resonance.
Increased TNF-a at two higher concentrations of O3 at 16 h
postexposure.
Bayram et al.
(2002)

Schierhorn
et al. (2002)

Spannhake
et al. (2002)
Wang et al.
(2002)
Bayram et al.
(2001)
Nichols et al.
(2001)


-------
          Table AX6-13 (cont'd). Studies of Effects on Host Defense, on Drug Effects and Supportive In Vitro Studies Relating to

                                              Controlled Human Exposure to Ozone"
to
o
o
X
H

6
o


o
H

O

o
H
W

O


O
HH
H
W
Ozone
Concentration13
ppm
ug/m3 Duration
Activity Level
(VE)
Number and
Gender of Subjects Observed Effect(s)
Reference
Supportive In Vitro Studies (cont'd)
0.06 to 118 to 24 h
0.20 392
0.5
0.5
0.4

0.25
0.50
0.25
0.50
1.00
980 Ih
980 Ih
784 Ih

490 6h
980
490 Ih
980
1,960
Nasal mucosa
Lung fibroblast
cell line with an
airway epithelial
cell line
tracheal epithelial
cells
Lung fibroblasts;
airway epithelial
cell line

Human nasal
epithelial cells
Airway epithelial
cell line and
alveolar
macrophages
105 surgical Increased histamine release correlated with mast cell degranulation.
samples from Increased release of IL-1, IL-6, IL-8 and TNF-a following O3 exposure
atopic and at 0. 10 ppm. Release of IL-4, IL-6, IL-8 and TNF-a at
nonatopic patients this concentration was significantly greater from tissues from atopic
versus nonatopic patients.
BEAS-2B cells in the presence or absence of HFL-1 cells were
exposed and incubated for 11 or 23 h. Steady-state mRNA levels
of alpha 1 procollagens type I and II, as well as TGF(31, were increased
in O3-exposed co-cultured fibroblasts compared to air controls. Data
support interactions between the cell types in the presence and the
absence of O3-exposure.
O3 exposure caused an increase in ROS formation and a decline in
PGE2 production. No differences in mRNA and protein levels of
prostaglandin endoperoxide G/H synthase 2 (PGHS-2) or the rate of its
synthesis were detected, suggesting a direct effect of Degenerated
oxidants on PGHS-2 activity.
Cells incubated with O3-exposed arachidonic acid (AA) were found to
contain DNA single strand breaks. Pretreatment of the exposed AA
solution with catalase eliminated the effect on DNA, indicating its
dependence on H2O2 production. The effect was potentiated by the
non-carbonyl component of ozonized AA.
Increased in ICAM-1, IL-6, IL-1, and TNF expression at 0.5 ppm.
No increase in IL-8 expression. No increases at 0.25 ppm.
Increased secretion of IL-6, IL-8, and fibronectin by epithelial cells,
even at lowest O3 concentration. No O3-induced secretion of these
compounds by macrophages.
Schierhorn
etal. (1999)
Lang et al.
(1998)
Alpert et al.
(1997)
Kozumbo et al.
(1996)

Beck et al.
(1994)
Devlin et al.
(1994)

-------
to
o
o
X
ON
           Table AX6-13 (cont'd). Studies of Effects on Host Defense, on Drug Effects and Supportive In Vitro Studies Relating to

                                                   Controlled Human Exposure to Ozone"
Ozone
Concentration13
ppm
|ig/m3 Duration
Activity Level
(VE)
Number and
Gender of Subjects Observed Effect(s)
Reference
Supportive In Vitro Studies (cont'd)
0.20 to
1.0



0.25
0.50
1.00
0.30
1.00

392 to 2hor4h
1960



490 Ih
980
1,960
588 Ih
1,960

Airway epithelial
cell line



Airway epithelial
cell line

Alveolar
macrophages

O3 caused a dose-related loss in cellular replicative activity at exposure
levels that caused minimal cytotoxicity. DNAsingle strand breaks
were not detected. These effects were different from those of H2O2
and, thus, not likely related to production of this oxidant within the
cells.
Concentration-dependent increased secretion of PGE2, TXB2,
PGF2ct, LTB4, and LTD4. More secretion basolaterally than apically.

Concentration-dependent increases in PGE2 production, and decreases
in phagocytosis of sheep erythrocytes. No O3-induced secretion of IL-
l,TNF,orIL-6.
Gabrielson
etal. (1994).



McKinnon
etal. (1993)

Becker et al.
(1991)

       "See Appendix A for abbreviations and acronyms.

       bListed from lowest to highest O3 concentration.
H

6
o


o
H

O

o
H
W

O


O
HH
H
W

-------
 1      central airways, and of persistent effects following repeated exposure, may indicate that further
 2      investigation of inflammatory processes in these regions is warranted.
 3           Under normal circumstances, the epithelia lining the large and small airways develop tight
 4      junctions and restrict the penetration of exogenous particles and macromolecules from the
 5      airway lumen into the interstitium and blood, as well as restrict the flow of plasma components
 6      into the airway lumen. O3 disrupts the integrity of the epithelial cell barrier in human airways, as
 7      measured by markers of plasma influx  such as albumin, immunoglobulin, and other proteins into
 8      the airways.  Markers of epithelial cell damage such as lactate dehydrogenase (LDH) also have
 9      been measured in the BAL fluid of humans exposed to O3. Other soluble factors that have been
10      studied include those involved with fibrin deposition and degradation (Tissue Factor, Factor VII,
11      and plasminogen activator), potential markers of fibrogenesis (fibronectin,  platelet derived
12      growth factor), and components of the  complement cascade (C3a).
13           Inflammatory cells of the lung such as alveolar macrophages (AMs),  monocytes, and
14      PMNs also constitute an important component of the pulmonary host defense system.  Upon
15      activation, they are capable of generating free radicals and enzymes with microbicidal
16      capabilities, but they also have the potential to damage nearby  cells. More recently published
17      studies since the last literature review (U.S. Environmental Protection Agency, 1996) observed
18      changes in T lymphocyte subsets in the airways following exposure to O3 that suggest
19      components of the immune host defense also may be affected.
20
21      AX6.9.2  Inflammatory Responses in the Upper Respiratory Tract
22           The nasal passages constitute the primary portal for inspired air at rest and, therefore, the
23      first region of the respiratory tract to come in contact with airborne pollutants. Nikasinovic et al.
24      (2003) recently reviewed the literature of laboratory-based nasal inflammatory studies published
25      since 1985.  Nasal lavage (NL) has provided a useful tool for assessing O3-induced inflammation
26      in the nasopharynx. Nasal lavage is simple and rapid to perform, is noninvasive, and allows
27      collection of multiple sequential samples.  Graham et al. (1988) reported increased levels of
28      PMNs in the NL fluid of humans exposed to 0.5 ppm O3 at rest for 4 h on 2 consecutive days,
29      with NL performed immediately before and after each exposure, as well as 22 h after the second
30      exposure. Nasal lavage fluid contained elevated numbers  of PMNs at all postexposure times
31      tested, with peak values occurring immediately prior to the second day of exposure. Bascom

        August 2005                            AX6-113     DRAFT-DO NOT QUOTE OR CITE

-------
 1      et al. (1990) exposed subjects with allergic rhinitis to 0.5 ppm O3 at rest for 4 h, and found
 2      increases in PMNs, eosinophils, and mononuclear cells following O3 exposure.  Graham and
 3      Koren (1990) compared inflammatory mediators present in both the NL and BAL fluids of
 4      humans exposed to 0.4 ppm O3 for 2 h. Increases in NL and BAL PMNs were similar (6.6- and
 5      eightfold, respectively), suggesting a qualitative correlation between inflammatory changes in
 6      the lower airways (BAL) and the upper respiratory tract (NL), although the PMN increase in NL
 7      could not quantitatively predict the PMN increase in BAL.  Torres et al. (1997) compared NL
 8      and BAL in smokers and nonsmokers exposed to 0.22 ppm O3 for 4 h.  In contrast to Graham
 9      and Koren (1990), they did not find a relationship between numbers or percentages of
10      inflammatory cells (PMNs) in the nose and the lung, perhaps in part due to the variability
11      observed in their NL recoveries.  Albumin, a marker of epithelial cell permeability, was
12      increased 18  h later, but not immediately after exposure, as seen by Bascom et al. (1990).
13      Tryptase, a constituent of mast cells, was also elevated after O3 exposure at 0.4 ppm for 2 h
14      (Koren et al., 1990). McBride et al. (1994) reported that asthmatic subjects were more sensitive
15      than nonasthmatics to upper airway inflammation at an O3 concentration (0.24 ppm for 1.5 h
16      with light IE) that did not affect lung or nasal function or biochemical mediators. A significant
17      increase in the number of PMNs in NL fluid was detected in the asthmatic subjects both
18      immediately  and 24 h after exposure. Peden et al. (1995) also found that O3 at a concentration of
19      0.4 ppm had  a direct nasal inflammatory effect, and reported a priming effect on the response to
20      nasal allergen challenge, as well.  A subsequent study in dust mite-sensitive asthmatic subjects
21      indicated that O3 at this concentration enhanced eosinophil influx in response to allergen, but did
22      not promote early mediator release or enhance the nasal response to allergen (Michelson et al.,
23      1999). Similar to observations made in the lower airways, the presence of O3 molecular
24      "targets" in nasal lining fluid is likely to provide some level of local protection against exposure.
25      In a study of healthy subjects exposed to 0.2 ppm O3 for 2 h, Mudway and colleagues (1999)
26      observed a significant depletion of uric acid in NL fluid at 1.5 h following exposure.
27           An increasing number of studies have taken advantage of advances in cell and tissue
28      culture techniques to examine the role of upper and lower airway epithelial cells and mucosa in
29      transducing the effects of O3 exposure.  Many of these studies have provided important insight
30      into the basis of observations made in vivo. One of the methods used enables the cells or tissue
31      samples  to be cultured at the air-liquid interface (ALI), allowing cells to establish apical and

        August 2005                            AX6-114      DRAFT-DO NOT QUOTE OR CITE

-------
 1      basal polarity, and both cells and tissue samples to undergo exposure to O3 at the apical surfaces
 2      as would occur in vivo. Nichols and colleagues (2001) examined human nasal epithelial cells
 3      grown at the ALI for changes in free radical production, based on electron spin resonance, and
 4      activation of the NF-KB transcription factor following exposure to O3 at 0.12 to 0.5 ppm for 3 h.
 5      They found a dose-related activation of NF-KB within the cells that coincided with O3-induced
 6      free radical production and increased release of TNF-a at levels above 0.24 ppm.  These data
 7      confirm the importance of this oxidant stress-associated pathway in transducing the O3 signal
 8      within nasal epithelial cells and suggest its role in directing the inflammatory response. In a
 9      study of nasal mucosal biopsy plugs, Schierhorn et al. (1999) found that tissues exposed to O3 at
10      a concentration of 0.1 ppm induced release of IL-4, IL-6, IL-8, and TNF-a that was significantly
11      greater from tissues from atopic patients compared to nonatopic controls. In a subsequent study,
12      this same exposure regimen caused the release of significantly greater amounts of the
13      neuropeptides, neurokinin A and substance P, from allergic patients,  compared to nonallergic
14      controls, suggesting increased activation of sensory nerves by O3 in the allergic tissues
15      (Schierhorn et al., 2002).
16
17      AX6.9.3   Inflammatory Responses in the Lower Respiratory Tract
18           Seltzer et al.  (1986) were the first to demonstrate that exposure of humans to O3 resulted in
19      inflammation in the lung.  Bronchoalveolar lavage fluid (3 h postexposure) from subjects
20      exposed to O3 contained increased PMNs as well as increased levels of PGE2,  PGF2a, and TXB2
21      compared to fluid from air-exposed subjects.  Koren et al. (1989a,b) described inflammatory
22      changes 18 h after O3 exposure. In addition to an eightfold increase in PMNs, Koren et al.
23      reported a two-fold increase in BAL fluid protein, albumin, and immunoglobulin G (IgG) levels,
24      suggestive of increased epithelial cell permeability.  There was a 12-fold increase in IL-6 levels,
25      a two-fold increase in PGE2,  and a two-fold increase in the complement component, C3a.
26      Evidence for stimulation of fibrogenic processes in the lung was shown by significant increases
27      in coagulation components, Tissue Factor and Factor VII (McGee et al.,  1990), urokinase
28      plasminogen activator and fibronectin (Koren et al.,  1989a).  Subsequent studies by Lang et al.
29      (1998), using co-cultures of cells of the BEAS-2B bronchial epithelial line and of the HFL-1
30      lung fibroblast line, provided additional information about O3-induced fibrogenic processes.
31      They demonstrated that steady-state mRNA levels of both alpha 1 and procollagens type I and

        August 2005                            AX6-115     DRAFT-DO NOT QUOTE OR CITE

-------
 1      III in the fibroblasts were increased following O3 exposure and that this effect was mediated by
 2      the O3-exposed epithelial cells.  This group of studies demonstrated that exposure to O3 results in
 3      an inflammatory reaction in the lung, as evidenced by increases in PMNs and proinflammatory
 4      compounds.  Furthermore, they demonstrated that cells and mediators capable of damaging
 5      pulmonary tissue are increased after O3 exposure and provided early suggestion of the potential
 6      importance of the epithelial cell-myofibroblast "axis" in modulating fibrotic and fibrinolytic
 7      processes in the airways.
 8           Isolated lavage of the mainstream bronchus using balloon catheters or BAL using small
 9      volumes of saline have been used to assess O3-induced changes in the large airways. Studies
10      collecting lavage fluid from isolated airway segments after O3 exposure indicate increased
11      neutrophils in the airways (Aris et al.,  1993; Balmes et al., 1996; Scannell et al., 1996). Other
12      evidence of airway neutrophil increase comes from studies in which the initial lavage fraction
13      ("bronchial fraction") showed increased levels of neutrophils (Schelegle et al.,  1991; Peden
14      et al., 1997; Balmes et al., 1996; Torres et al., 1997). Bronchial biopsies show increased PMNs
15      in airway tissue (Aris et al.,1993) and, in sputum collected after O3 exposure, neutrophil numbers
16      are elevated (Fahy et al., 1995).
17           Increased BAL protein, suggesting O3-induced changes in epithelial permeability (Koren
18      et al., 1989a, 1991 and Devlin et al., 1991) supports earlier work in which increased epithelial
19      permeability, as measured by increased clearance of radiolabled diethylene triamine pentaacetic
20      acid (99mTc-DTPA) from the lungs of humans exposed to O3, was demonstrated (Kehrl et al.,
21      1987).  In addition, Foster and Stetkiewicz (1996) have shown that increased permeability
22      persists for at least 18-20 h and the effect is greater at the lung apices than at the base. In a study
23      of mild atopic asthmatics exposed to 0.2 ppm O3 for 2 h,  Newson et al. (2000) observed a 2-fold
24      increase in the percentage of PMNs present at 6 hours postexposure, with no change in markers
25      of increased permeability as assessed by sputum induction. By 24 h, the neutrophilia was seen
26      to subside while levels of albumin, total protein, myeloperoxidase, and eosinophil cationic
27      protein increased significantly.  It was concluded that the transient PMN influx induced by acute
28      exposure of these asthmatic subjects was followed by plasma extravasation and the activation of
29      both PMNs and eosinophils within the airway tissues. Such changes in permeability associated
30      with acute inflammation may provide better access  of inhaled antigens, particulates, and other
31      substances to the submucosal region.

        August 2005                            AX6-116     DRAFT-DO NOT QUOTE OR CITE

-------
 1           Devlin et al. (1991) reported an inflammatory response in subjects exposed to 0.08
 2      and 0.10 ppm O3 for 6.6 h.  Increased numbers of PMNs and levels of IL-6 were found at
 3      both O3 concentrations, suggesting that lung inflammation from O3 can occur as a consequence
 4      of prolonged exposure to ambient levels while exercising. Interestingly, those individuals who
 5      had the largest increases in inflammatory mediators in this study did not necessarily have the
 6      largest decrements in pulmonary function, suggesting that separate mechanisms underlie these
 7      two responses.  The absence of a relationship between spirometric responses and inflammatory
 8      cells and markers has been reported in  several studies, including Balmes et al.,  1996; Schelegle
 9      et al., 1991; Torres et al., 1997; Hazucha et al., 1996; Blomberg et al., 1999.  These observations
10      relate largely to disparities in the times of onset and duration following single exposures. The
11      relationship between inflammatory and residual functional responses following repeated or
12      chronic exposures may represent a somewhat different case  (see Section AX6.9.4).
13           As indicated above, a variety of potent proinflammatory mediators have been reported to
14      be released into the airway lumen following O3 exposure. Studies of human  alveolar
15      macrophages (AM) and airway epithelial cells exposed to O3 in vitro suggest that most mediators
16      found in the BAL fluid of O3-exposed humans are produced by epithelial cells. Macrophages
17      exposed to O3 in vitro showed only small increases in PGE2 (Becker et al., 1991). In contrast,
18      airway epithelial cells exposed in vitro to O3 showed large concentration-dependent increases
19      in PGE2, TXB2, LTB4, LTC4, and LTD4 (McKinnon et al., 1993) and increases in IL-6, IL-8, and
20      fibronectin at O3 concentrations as low as 0.1 ppm (Devlin et al.,  1994). Macrophages lavaged
21      from subjects exposed to 0.4 ppm (Koren et al., 1989a) showed changes in the rate of synthesis
22      of 123 different proteins, whereas AMs exposed to O3 in vitro showed changes in only six
23      proteins, suggesting that macrophage function was altered by mediators released  from other
24      cells. Furthermore, recent evidence suggests that the release of mediators from AMs may be
25      modulated by the products of O3-induced oxidation of airway lining fluid components, such as
26      human surfactant protein A (Wang et al., 2002).
27           Although the release of mediators has been demonstrated to occur at exposure
28      concentrations and times that are minimally cytotoxic to airway cells, potentially detrimental
29      latent effects have been demonstrated in the absence of cytotoxicity. These include the
30      generation of DNA single strand breaks (Kozumbo et al., 1996) and the loss  of cellular
31      replicative activity (Gabrielson et al., 1994) in bronchial epithelial cells exposed in vitro, and the

        August 2005                            AX6-117     DRAFT-DO NOT QUOTE OR CITE

-------
 1      formation of protein and DNA adducts. A highly toxic aldehyde formed during O3-induced lipid
 2      peroxidation is 4-hydroxynonenal (HNE). Healthy human subjects exposed to 0.4 ppm O3 for
 3      1 h underwent BAL 6 h later. Analysis of lavaged alveolar macrophages by Western blot
 4      indicated increased levels of a 32-kDa HNE-protein adduct, as well as 72-kDa heat shock protein
 5      and ferritin, in O3- versus air-exposed subjects (Hamilton et al., 1998).  In a recent study of
 6      healthy subjects exposed to 0.1 ppm O3 for 2 h (Corradi et al., 2002), formation of 8-hydroxy-2'-
 7      deoxyguanosine (8-OHdG), a biomarker of reactive oxidant species (ROS)-DNA interaction,
 8      was measured in peripheral blood lymphocytes.  At 18 h postexposure, 8-OHdG was
 9      significantly increased in cells compared to pre-exposure levels, presumably linked to concurrent
10      increases in chemical markers of ROS. Of interest, the increase in 8-OHdG was only significant
11      in a subgroup of subjects with the wild genotype for NAD(P)H:quinone oxidoreductase and the
12      null genotype for glutathione-S-transferase Ml, suggesting that polymorphisms in redox
13      enzymes may confer "susceptibility' to O3 in some individuals. The generation of ROS
14      following exposure to O3 has been shown to be associated with a wide range of responses.  In a
15      recent study, ROS production by alveolar macrophages lavaged from subjects exposed to
16      0.22 ppm for 4 h was assessed by flow cytometry (Voter et al., 2001). Levels were found to be
17      significantly elevated  18 h postexposure and associated with several markers of increased
18      permeability. An in vitro study of human tracheal epithelial cells exposed to O3 indicated that
19      generation of ROS resulted in decrease in synthesis of the bronchodilatory prostaglandin, PGE2,
20      as a result of inactivation of prostaglandin endoperoxide G/H synthase 2 (Alpert et al.,  1997).
21      These and similar studies indicate that the responses to products of O3 exposure in the airways
22      encompass a broad range of both stimulatory and inhibitory activities, many of which may be
23      modulated by susceptibility factors upstream in the exposure process, at the level of
24      compensating for the imposed oxidant stress.
25           The inflammatory responses to O3 exposure also have been studied in asthmatic subjects
26      (Basha et al., 1994; Scannell et al., 1996; Peden et al., 1997).  In these studies, asthmatics
27      showed significantly more neutrophils in the BAL (18 h postexposure) than similarly exposed
28      healthy individuals. In one of these studies (Peden et al.,  1997), which included only allergic
29      asthmatics who tested positive for Dematophagoides farinae antigen, there was an eosinophilic
30      inflammation (2-fold increase), as well as neutrophilic inflammation (3-fold increase).  In a
31      study of subjects with intermittent asthma that utilized a 2-fold higher concentration of O3

        August 2005                            AX6-118      DRAFT-DO NOT QUOTE OR CITE

-------
 1      (0.4 ppm) for 2 h, increases in eosinophil cationic protein, neutrophil elastase and IL-8 were
 2      found to be significantly increased 16 h postexposure and comparable in induced sputum and
 3      BAL fluid (Hiltermann et al, 1999). Scannell et al. (1996) also reported that IL-8 tended to be
 4      higher in post-O3 exposure BAL in asthmatics compared to nonasthmatics (36 vs. 12 pg/mL,
 5      respectively) suggesting a possible mediator for the significantly increased neutrophilic
 6      inflammation in asthmatics relative to healthy subjects (12 vs. 4.5%, respectively).  In a recent
 7      study comparing the neutrophil response to O3 at a concentration and exposure time similar to
 8      those of the latter three studies, Stenfors and colleagues (2002) were unable to detect a
 9      difference in the increased neutrophil numbers between 15 mild asthmatic and  15 healthy
10      subjects by bronchial wash at the 6 h postexposure time point. These results suggest that, at least
11      with regard to neutrophil influx, differences between healthy and asthmatic individuals develop
12      gradually following exposure and may not become evident until later in the process.
13           In another study, mild asthmatics who exhibited a late phase underwent allergen challenge
14      24 hrs before a 2 h exposure to 0.27 ppm O3 or filtered air in a cross-over design  (Vagaggini
15      et al., 2002). At 6 h postexposure, eosinophil numbers in induced sputum were found to be
16      significantly greater after O3 than after air.  Studies such as these suggest that the time course of
17      eosinophil and neutrophil influx following O3 exposure can occur to levels detectable within the
18      airway lumen by as early as 6 h. They also suggest that the previous or concurrent activation of
19      proinflammatory pathways within the airway epithelium may enhance the inflammatory effects
20      of O3. For example, in an in vitro study  of epithelial cells from the upper and lower respiratory
21      tract, cytokine production induced by rhinovirus infection was enhanced synergistically by
22      concurrent exposure to O3 at 0.2 ppm for 3 h (Spannhake et al, 2002).
23           The use of bronchial mucosal biopsies has also provided important insight into the
24      modulation by O3 of existing inflammatory processes within asthmatics. In a study of healthy
25      and allergic asthmatic  subjects exposed to 0.2 ppm O3 or filtered air for 2 h, biopsies were
26      performed 6 hr following exposure (Bosson et al., 2003).  Monoclonal antibodies were used to
27      assess epithelial expression of a variety of cytokines and chemokines.  At baseline (air
28      exposure), asthmatic subjects showed significantly higher expression of interleukins (IL)-4
29      and -5. Following O3 exposure, the epithelial expression of IL-5, IL-8, granulocyte-macrophage
30      colony-stimulating factor (GM-CSF) and epithelial cell-derived neutrophil-activating peptide
        August 2005                             AX6-119     DRAFT-DO NOT QUOTE OR CITE

-------
 1      78 (ENA-78) was significantly greater in asthmatic subjects, as compared to healthy subjects.
 2      In vitro studies of bronchial epithelial cells derived by biopsy from nonatopic, nonasthmatic
 3      subjects and asthmatic subjects also demonstrated the preferential release of GM-CSF and also
 4      of regulated on activation, normal T cell-expressed and -secreted (RANTES) from asthmatic
 5      cells following O3 exposure.
 6           The time course of the inflammatory response to O3 in humans has not been explored fully.
 7      Nevertheless,  studies in which BAL was performed 1-3 h (Devlin et al., 1990; Koren et al.,
 8      1991; Seltzer et al.,  1986) after exposure to 0.4 ppm O3 demonstrated that the inflammatory
 9      response is quickly initiated, and other studies (Koren et al., 1989a,b; Torres et al., 1997;
10      Scannell et al., 1996; Balmes et al., 1996) indicated that, even 18 h after exposure, inflammatory
11      mediators such as IL-6 and PMNs were still elevated.  However, different markers show peak
12      responses at different times.  Ozone-induced increases in IL-8, IL-6, and PGE2 are greater
13      immediately after O3 exposure, whereas BAL levels of fibronectin and plasminogen activator are
14      greater after 18 h. PMNs and some products (protein, Tissue Factor) are similarly elevated both
15      1  and 18 h after O3 exposure (Devlin et al., 1996; Torres et al., 1997).  Schelegle et al. (1991)
16      found increased PMNs in the "proximal airway" lavage at 1, 6, and  24 h after O3 exposure, with
17      a peak response at 6 h. In a typical BAL sample, PMNs were elevated only at the later time
18      points.  This is consistent with the greater increase 18 h after exposure seen by Torres et al.
19      (1997). In addition to the influx of PMNs and (in allergic asthmatics) eosinophils, lymphocyte
20      numbers in BAL were also seen to be elevated significantly at 6 h following exposure of healthy
21      subjects to 0.2 ppm O3 for 2 h (Blomberg et al., 1997).  Analysis of these cells by flow cytometry
22      indicated the increased presence of CD3+, CD4+ and CD8+ T cell subsets.  This same laboratory
23      later demonstrated that within 1.5 h following exposure of healthy subjects to the same O3
24      regimen, expression of human leukocyte antigen (HLA)-DR on lavaged macrophages underwent
25      a significant, 2.5-fold increase (Blomberg et al., 1999). The significance of these alterations in
26      immune system components and those in IL-4 and IL-5 expression described above in the
27      studies of Bosson et al. (2003) has not been fully explored and may  suggest a role for O3 in the
28      modulation of immune inflammatory processes.
29
        August 2005                            AX6-120      DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX6.9.4  Adaptation of Inflammatory Responses
 2           Residents of areas with high oxidant concentrations tend to have somewhat blunted
 3      pulmonary function responses and symptoms to O3 exposure (Hackney et al., 1976, 1977b, 1989;
 4      Avol et al., 1988; Linn et al., 1988).  Animal studies suggest that while inflammation may be
 5      diminished with repeated exposure, underlying damage to lung epithelial cells continues (Tepper
 6      et al., 1989).  Devlin et al. (1997) examined the inflammatory responces of humans repeatedly
 7      exposed to 0.4 ppm O3 for 5 consecutive days. Several indicators of inflammation (e.g., PMN
 8      influx, IL-6, PGE2, fibronectin, macrophage phagocytosis) were attenuated after 5 days of
 9      exposure (i.e., values were not different from FA).  Several markers (LDH, IL-8, total protein,
10      epithelial cells) did not show attenuation, indicating that tissue damage probably continues to
11      occur during repeated exposure. The recovery of the inflammatory response occurred for some
12      markers after 10 days, but some responses were not normalized even after 20 days.  The
13      continued presence of markers of cellular injury indicates a persistent but not necessarily
14      perceived response to O3.
15           Christian et al.  (1998) randomly subjected heathy subjects to a single exposure and to
16      4 consecutive days of exposure to 0.2 ppm O3 for 4 h.  As reported by  others, they found an
17      attenuation of FEVl3 FVC and specific airway resistance when comparing the single exposure
18      with day 4 of the multiday exposure regimen. Similarly, both "bronchial" and "alveolar"
19      fractions of the BAL showed decreased numbers of PMNs and fibronectin concentration at day
20      4 versus the single exposure, and a decrease in IL-6 levels in the alveolar fraction.  Following a
21      similar study design and exposure parameters, but with single day filtered air controls, Torres
22      et al. (2000) found a decrease in FEVj and increases in the percentages of neutrophils  and
23      lymphocytes, in concentrations of total protein, IL-6, IL-8, reduced glutathione,  ortho-tyrosine
24      and urate in BAL fluid, but no changes in bronchial biopsy histology following the single
25      exposure. Twenty hours after the day 4 exposure, both functional and BAL cellular responses
26      to O3 were abolished. However, levels of total protein, IL-6, IL-8, reduced glutathione and
27      ortho-tyrosine were still increased significantly.  In addition, following the day 4 exposure,
28      visual scores for bronchitis, erythema and the numbers of neutrophils in the mucosal biopsies
29      were increased.  Their results indicate that, despite reduction of some markers of inflammation
30      in BAL and measures of large airway function, inflammation within the airways persists
31      following repeated exposure to O3.

        August 2005                            AX6-121     DRAFT-DO NOT QUOTE OR CITE

-------
 1          In another study, Frank and colleagues (2001) exposed healthy subjects to filtered air and
 2     to O3 (0.25 ppm, 2 h) on 4 consecutive days each, with pulmonary function measurements being
 3     made prior to and following each exposure. BAL was performed on day 5, 24 h following the
 4     last exposure.  On day 5, PMN numbers remained significantly higher in the O3 arm compared to
 5     air control. Of particular note in this study was the observation that small airway function,
 6     assessed by grouping values for isovolumetric FEF25.75, VmaxSO and Vmax75 into a single
 7     value, showed persistent reduction from day 2 through day 5.  These data suggest that methods
 8     to more effectively monitor function in the most peripheral airway regions, which are known to
 9     be the primary sites of O3 deposition in the lung, may provide important information regarding
10     the cumulative effects of O3 exposure. Holz et al. (2002) made a comparison of early and late
11     responses to allergen challenge following O3 in subjects with allergic rhinitis or allergic asthma.
12     With some variation, both early and late FEVj and cellular responses in the two subject groups
13     were significantly enhanced by 4 consecutive days of exposure to 0.125 ppm O3 for 3 h.
14
15     AX6.9.5  Effect of Anti-Inflammatory and Other Mitigating Agents
16          Studies have shown that indomethacin, a non-steroidal anti-inflammatory agent (NSAID)
17     that inhibits the production of cyclooxygenase products of arachidonic acid metabolism, is
18     capable of blunting the well-documented decrements in pulmonary function observed in
19     humans exposed to O3 (Schelegle et al., 1987; Ying et al., 1990). Indomethacin did not alter
20     the O3-induced increase in bronchial responsiveness to methacholine (Ying et al.,  1990).
21     Pretreatment of healthy subjects  and asthmatics with indomethacin prior to exposure to 0.4 ppm
22     for 2 h significantly attenuated decreases in FVC and FEVj in  normals, but not asthmatics
23     (Alexis et al., 2000).  Subjects have  also been given ibuprofen, another NSAID agent that blocks
24     cyclooxygenase metabolism, prior to O3 exposure. Ibuprofen blunted decrements in lung
25     function following O3 exposure (Hazucha et al., 1996).  Subjects given ibuprofen also had
26     reduced BAL levels of the cyclooxygenase product PGE2 and thromboxane B2, as well as IL-6,
27     but no decreases were observed in PMNs, fibronectin, permeability, LDH activity, or
28     macrophage phagocytic function. These studies suggest that NSAIDs can blunt O3-induced
29     decrements in FEVj with selective (perhaps drug-specific) affects on mediator release and other
30     markers of inflammation.
       August 2005                           AX6-122     DRAFT-DO NOT QUOTE OR CITE

-------
 1           At least two studies have looked at the effects of the inhaled corticosteroid, budesonide, on
 2      the effects of O3, with differing outcome perhaps associated with the presence of preexistent
 3      disease.  Nightingale and colleagues (2000) exposed healthy nonsmokers to 0.4 ppm O3 for 2 h
 4      following 2 wk of treatment with budesonide (800 micrograms, twice daily) or placebo in a
 5      blinded, randomized cross-over study. This relatively high O3 exposure resulted in significant
 6      decreases in spirometric measures and increases in methacholine reactivity and neutrophils and
 7      myeloperoxidase in induced sputum.  No significant differences were observed in any of these
 8      endpoints following budesonide treatment versus placebo. In contrast, Vagaggini et al. (2001)
 9      compared the effects of treatment with budesonide (400 micrograms, twice daily) for 4 wk on
10      the responses of mild asthmatic subjects to exposure to 0.27 ppm O3for 2 h.  Prior to exposure,
11      at the midpoint and end of exposure, and at 6 h postexposure, FEVj was measured and a
12      symptom questionnaire was administered; at 6 h postexposure, sputum was induced.
13      Budesonide treatment did not inhibit the decrement in FEVj or alter symptom score, but
14      significantly blunted the increase in percent PMNs and concentration of IL-8  in the sputum. The
15      difference in subject health status between  the two studies (healthy versus mild asthmatic) may
16      suggest a basis for the differing outcomes;  however, because of differences in the corticosteroid
17      dosage and O3 exposure levels, that basis remains unclear.
18           Holz et al. (2005) investigated the mitigation of O3-induced inflammatory responses in
19      subjects pretreated with single doses of inhaled fluticasone and oral prednisolone. Eighteen
20      healthy ozone-responders (>10% increase in sputum neutrophils from O3 exposure) received
21      corticosteroid treatment or placebo 1-h before being exposed for 3-h with IE (15 min periods
22      rest/exercise) to 0.25 ppm O3.  Sputum was collected 3-h  post-O3 exposure. The 18 ozone-
23      responders were selected from 35 screened subjects.  Twelve subjects were disqualified from the
24      study (6 produced insufficient sputum and  6 had inadequate neutrophil responses to O3), the
25      remaining 5 subjects were [presumably] qualified but did not participate. The O3 exposure
26      caused small changes in FEVj (-3.6% ±6.8%) that were not significantly different from baseline
27      or between treatment groups (i.e., prescreening, placebo,  fluticasone, and prednisolone).
28      On average, the prescreening and placebo O3 exposures caused greater than a 9-fold increase in
29      sputum neutrophils relative to baseline levels. Relative to placebo, the inhaled or oral
30      corticosteroids significantly reduced neutrophil levels by  62 and 64%, respectively.  Total
31      protein levels were not altered by O3 or corticosteroid treatment.  Authors concluded that the

        August 2005                            AX6-123      DRAFT-DO NOT QUOTE OR CITE

-------
 1      pronounced anti-inflammatory effect of steroids in their study was due to administering the
 2      highest single doses shown to be safe and well tolerated. Furthermore, steroids were
 3      administered so that maximal plasma level would be reached at approximately the beginning of
 4      the O3 exposure.
 5           Because the O3 exerts its actions in the respiratory tract by virtue of its strong oxidant
 6      activity, it is reasonable to assume that molecules that can act as surrogate targets in the airways,
 7      as constituents of either extracellular fluids or the intracellular milieus, could abrogate the effects
 8      of O3. Some studies have examined the ability of dietary "antioxidant" supplements to reduce
 9      the risk of exposure of the lung to oxidant exposure. In a study of healthy, nonsmoking adults,
10      Samet and colleagues (2001) restricted dietary ascorbate and randomly treated subjects for
11      2 weeks with a mixture of vitamin C, a-tocopherol and vegetable  cocktail high in carrot and
12      tomato juices or placebo. Responses to 0.4 ppm O3 for 2 h were assessed in both groups at the
13      end of treatment.  O3 -induced decrements in FEVj and FVC were significantly reduced in the
14      supplemented group, whereas the inflammatory response, as assessed by percentage neutrophils
15      and levels of IL-6 in BAL fluid, were unaffected by antioxidant supplementation. In a  study that
16      focused on supplementation with a commercial vegetable cocktail high in the carotenoid,
17      lycopene, healthy subjects were exposed for 2 h to 0.4 ppm O3 after 2 wk of antioxidant
18      supplementation or placebo (Arab  et al., 2002). These investigators observed that lung epithelial
19      cell DNA damage, as measured by the Comet Assay,  decreased by 20% in  supplemented
20      subjects.  However, the relationships between the types  and levels of antioxidants in airway
21      lining fluid and responsiveness to O3 exposure is likely to be complex. In another study where
22      differences in ascorbate and glutathione concentrations between healthy  and mild asthmatic
23      subjects were exploited, no relationship between antioxidant levels and spirometric or cellular
24      responses could be detected (Mudway et al., 2001).
25
26      AX6.9.6  Changes in Host Defense Capability Following Ozone Exposure
27           Concern about the effect of O3 on human host defense  capability derives from numerous
28      animal studies demonstrating that acute exposure to as little  as 0.08 ppm O3 causes decrements
29      in antibacterial host defenses (see Chapter 5). A study of experimental rhinovirus infection in
30      susceptible human volunteers failed to  show any effect of 5 consecutive days of O3 exposure on
31      the clinical evolution of, or host response to, a viral challenge (Henderson et al., 1988). Healthy

        August 2005                            AX6-124     DRAFT-DO NOT QUOTE OR CITE

-------
 1      men were nasally inoculated with type 39 rhinovirus (103 TCID50). There was no difference
 2      between the O3-exposed and control groups in rhinovirus liters in nasal secretions, in levels of
 3      interferon gamma or PMNs in NL fluid, or in blood lymphocyte proliferative response to
 4      rhinovirus antigen.  However, subsequent findings that rhinovirus can attach to the intracellular
 5      adhesion molecule (ICAM)-l receptor on respiratory tract epithelial cells (Greve et al., 1989)
 6      and that O3 can up-regulate the ICAM-1 receptor on nasal epithelial cells (Beck et al., 1994)
 7      suggest that more studies are  needed to explore the possibility that prior O3 exposure can
 8      enhance rhinovirus binding to, and infection of, the nasal epithelium.
 9           In a single study, human AM host defense capacity was measured in vitro in AMs removed
10      from subjects exposed to 0.08 and 0.10 ppm O3 for 6.6 h while undergoing moderate exercise.
11      Alveolar macrophages from O3-exposed subjects had significant decrements in complement-
12      receptor-mediated phagocytosis of Candida albicans (Devlin et al., 1991).  The impairment of
13      AM host defense capability could potentially result in decreased ability to phagocytose and kill
14      inhaled microorganisms in vivo.  A concentration-dependent decrease in phagocytosis of AMs
15      exposed to 0.1 to 1.0 ppm O3  in vitro has also been shown Becker et al. (1991). Although the
16      evidence is inconclusive at present, there is a concern that O3 may render humans and animals
17      more susceptible to a subsequent bacterial challenge.
18           Only two studies (Foster et al., 1987; Gerrity et al., 1993) have investigated the effect
19      of O3 exposure on mucociliary particle clearance  in humans. Foster et al. (1987) had seven
20      healthy subjects inhale radiolabeled particles (5 jim MMAD) and then exposed these subjects to
21      FA or O3 (0.2 and 0.4 ppm) during light IE for 2 h.  Gerrity et al. (1993) exposed 15 healthy
22      subjects to FA or 0.4 ppm O3  during CE (40 L/min) for 1 h; at 2 h post-O3 exposure, subjects
23      then inhaled radiolabeled particles  (5 jim MMAD). Subjects in both studies had similar
24      pulmonary function responses (average FVC decrease of 11 to 12%) immediately postexposure
25      to 0.4 ppm O3.  The Foster et  al. (1987) study suggested there is a  stimulatory affect of O3
26      on mucociliary clearance; whereas, Gerrity et al.  (1993) found that in the recovery period
27      following O3 exposure, mucus clearance is similar to control, i.e., following a FA exposure.
28      The clearance findings in these studies are complementary not conflicting.  Investigators in both
29      studies suggested that O3-induced increases in mucociliary clearance could be mediated by
30      cholinergic receptors. Gerrity et al. (1993) further suggested that transient clearance increases
        August 2005                            AX6-125     DRAFT-DO NOT QUOTE OR CITE

-------
 1      might be coincident to pulmonary function responses; this supposition based on the return of
 2      sRaw to baseline and the recovery of FVC to within 5% of baseline (versus an 11% decrement
 3      immediately postexposure) prior to clearance measurements.
 4           Insofar as the airway epithelial surface provides a barrier to entry of biological, chemical
 5      and particulate contaminants into the submucosal region, the maintenance of barrier integrity
 6      represents a component of host defense.  Many of the studies of upper and lower respiratory
 7      responses to O3 exposure previously cited above have reported increases in markers of airway
 8      permeability after both acute exposures and repeated exposures. These findings suggest that O3
 9      may increase access of airborne agents. In a study of bronchial epithelial cells obtained from
10      nonatopic and mild atopic  asthmatic subjects (Bayram et al., 2002), cells were grown to
11      confluence and transferred to porous membranes. When the cultures again reached confluence,
12      they were exposed to 0.01-0.1 ppm O3 or air and their permeability was assessed by measuring
13      the paracellular flux of 14C-BSA. The increase in permeability 24 h following O3 exposure was
14      observed to be significantly greater in  cultures of cells derived from asthmatics, compared to
15      healthy subjects. Thus, the late increase  in airway  permeability following exposure of asthmatic
16      subjects to O3, of the sort described by Newson et al. (2000), may be related to an inherent
17      susceptibility of 'asthmatic' cells to the barrier-reducing effects of O3.
18           As referenced in Section 6.9.3, the  O3-induced increase in the numbers of CD8+ T
19      lymphocytes in the airways of healthy subjects reported by Blomberg et al.  (1997) poses several
20      interesting questions regarding possible alterations in immune surveillance processes following
21      exposure. In a subsequent study from  the same group, Krishna et al. (1998) exposed healthy
22      subjects to 0.2 ppm O3 or filtered air for 2 h followed by BAL at 6 h.  In addition to increased
23      PMNs and other typical markers of inflammation, they found a significant decrease in the
24      CD4+/CD8+ T lymphocyte ratio and in the proportion of activated CD4+ and CD8+ cells.
25      Studies relating to the effects of low-level O3 exposure on the influx and activity of
26      immunocompetent cells in the upper and lower respiratory tracts may shed additional light on
27      modulation of this important area of host defense.
28
29
        August 2005                            AX6-126      DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX6.10   EXTRAPULMONARY EFFECTS OF OZONE
 2           Ozone reacts rapidly on contact with respiratory system tissue and is not absorbed or
 3      transported to extrapulmonary sites to any significant degree as such.  Laboratory animal studies
 4      suggest that reaction products formed by the interaction of O3 with respiratory system fluids or
 5      tissues may produce effects measured outside the respiratory tract—either in the blood, as
 6      changes in circulating blood lymphocytes, erythrocytes, and serum, or as changes in the structure
 7      or function of other organs, such as the parathyroid gland, the heart, the liver, and the central
 8      nervous system. Very little is known, however, about the mechanisms by which O3 could cause
 9      these extrapulmonary effects. (See Section 5.4 for a discussion of the systemic effects
10      ofO3 observed in laboratory animals.)
11           The results from human exposure studies discussed in the previous criteria documents
12      (U.S. Environmental Protection Agency, 1986, 1996) failed to demonstrate any consistent
13      extrapulmonary effects. Early studies on peripheral blood lymphocytes collected from human
14      volunteers did not find any significant genotoxic or functional changes at O3 exposures of 0.4 to
15      0.6 ppm for up to 4 h/day. Limited data on human subjects indicated that 0.5 ppm O3 exposure
16      for over 2 h caused transient changes in blood erythrocytes and sera (e.g., erythrocyte fragility
17      and enzyme activities), but the physiological significance of these studies remains questionable.
18      The conclusions drawn from these early studies raise doubt that cellular damage or altered
19      function is occurring to circulating cells at O3 exposures under 0.5 ppm.
20           Other human exposure studies have attempted to identify specific markers of exposure
21      to O3 in blood. For example, Schelegle et al. (1989) showed that PGF2a was elevated after O3
22      exposure (0.35 ppm); however, no increase in a-1 protease inhibitor was observed by Johnson
23      et al.  (1986). Foster et al. (1996) found  a reduction in the serum levels of the free radical
24      scavenger a-tocopherol after O3 exposure.  Vender et al. (1994) failed to find any changes in
25      indices of red blood cell antioxidant capacity (GSH, CAT) in healthy male subjects exposed to
26      0.16 ppm O3 for 7.5 h while intermittently exercising.  Liu et al. (1997, 1999) used a salicylate
27      metabolite, 2,3, dehydroxybenzoic acid (DFffiA), to indicate increased levels of hydroxyl radical
28      which hydroxylates salicylate to DHBA. Increased DHBA levels after exposure to 0.12 and
29      0.40 ppm suggest that O3 increases production of hydroxyl radical.  The levels of DHBA were
30      correlated with changes in spirometry.
        August 2005                            AX6-127      DRAFT-DO NOT QUOTE OR CITE

-------
 1           Only a few experimental human studies have examined O3 effects in other nonpulmonary
 2      organ systems besides blood. Early studies on the central nervous system (Gliner et al., 1979,
 3      1980) were not able to find significant effects on motor activity or behavior (vigilance and
 4      psychomotor performance) from O3 exposures at rest up to 0.75 ppm (U.S. Environmental
 5      Protection Agency, 1986).  Drechsler-Parks et al.  (1995) monitored ECG, HR, cardiac output,
 6      stroke volume, and systolic time intervals in healthy, older subjects (56 to 85 years of age)
 7      exposed to 0.45 ppm O3 using a noninvasive impedance cardiographic method. No changes
 8      were found at this high O3 concentration after 2 h of exposure while the subjects exercised
 9      intermittently at 25 L/min.
10           Gong et al. (1998) monitored ECG, HR, cardiac output, blood pressure, oxygen saturation,
11      and chemistries, as well as calculating other hemodynamic variables (e.g., stroke volume,
12      vascular resistance, rate-pressure products) in both healthy (n = 6) and hypertensive (n = 10)
13      adult males (41-78 years old). Subjects were exposed for 3 h with IE (VE « 30 L/min) to FA
14      and on the subsequent day to 0.3 ppm O3.  See Section AX6.3 for more details about this study.
15      Statistically significant O3 effects for both groups combined were increases in HR, rate-pressure
16      product, and the alveolar-to-arterial PO2 gradient. Gong et al. (1998) suggested that by
17      impairing alveolar-arterial oxygen transfer, the O3 exposure could potentially lead to adverse
18      cardiac events by decreasing oxygen supply to the myocardium. The subjects in the Gong et al.
19      (1998) study had sufficient functional reserve so as to not experience significant ECG changes or
20      myocardial ischemia and/or injury. However, Gong et al. (1998) concluded that O3 exposure
21      could pose a cardiopulmonary risk to persons with preexisting cardiovascular disease, with or
22      without concomitant respiratory  disease.
23           The mechanism for the decrease in arterial oxygen tension in the Gong et al. (1998) study
24      could be due to an O3 induced ventilation-perfusion mismatch. It is well recognized and
25      accepted that ventilation and perfusion per unit lung volume increase with progression from the
26      apex to the base of the lung in normal upright healthy humans (Inkley and Maclntyre, 1973;
27      Kaneko et al., 1966).  But, Foster et al. (1993) demonstrated that even in relatively young
28      healthy adults (26.7 ± 7 years old), O3 exposure can cause ventilation to shift away from the well
29      perfused basal lung (see Section AX6.2.3 for more details).  This effect of O3 on ventilation
30      distribution [and by association the small airways] may persist beyond 24-h postexposure (Foster
31      et al., 1997). Hypoxic pulmonary artery vasoconstriction acts to shift perfusion away from areas

        August 2005                            AX6-128     DRAFT-DO NOT QUOTE OR CITE

-------
 1      of low ventilation and moderate ventilation-perfusion mismatches (Santak et al., 1998).  This
 2      arterial vasoconstriction is thought to be mediated by protein kinase C, (Barman, 2001; Tsai
 3      et al., 2004). A more generalized (i.e., not localized to poor ventilated areas) increase in
 4      pulmonary vascular resistance in response to O3 exposure would presumably act against the
 5      ability of the hypoxic vasoconstriction in mediating ventilation-perfusion mismatches.  Acute
 6      arterial vasoconstriction has been observed clinically in humans (15 M, 10 F; 34.9 ± 10 years
 7      old) exposed for 2-h to O3 (0.12 ppm) in tandem with fine particulate («150 |ig/m3) (Brook et al.,
 8      2002).  Delaunois et al. (1998) also found that O3 exposure increases total (arterial, capillary, and
 9      venous segments) pulmonary vascular resistance in rabbits.  Hence, vasoconstriction could
10      potentially be induced by mechanisms other than regional hypoxia during O3 exposure.  This
11      notion is consistent with the O3-induced reduction in alveolar-arterial oxygen transfer observed
12      by Gong et al. (1998).
13           Effects of O3 exposure on alveolar-arterial oxygen gradients may be more pronounced in
14      patients with preexisting obstructive lung diseases.  Relative to healthy elderly subjects, COPD
15      patients have increased heterogeneity in both regional ventilation and perfusion (Kronenberg
16      et al., 1973). King and Briscoe (1968) examined the distribution of ventilation and perfusion in
17      a group of eight patients with severe COPD (mean FEVj/FVC = 36%).  In these patients, 68% of
18      the lung by volume received 45% of the cardiac output, but only 10% of the total alveolar
19      ventilation. This distribution of ventilation and perfusion in the patients contributed to their low
20      mean SaO2 of only 82% (inspired oxygen, 20.93%).  Thus,  even prior to O3 exposure, COPD
21      patients may have reduced gas exchange and low SaO2. Based on model predictions, increasing
22      tidal volume increases the O3 dose to the proximal alveolar region (Overton et al.,  1996).
23      Similarly, with 90%  of the alveolar ventilation supplied to only 32% of lung's volume, the well
24      ventilated regions of the COPD lung will be subjected to increased peripheral O3 doses.  Any
25      inflammatory or edematous responses due to O3 delivered to the well ventilated regions of the
26      COPD  lung will likely further inhibit gas exchange and reduce oxygen saturation.
27           In addition to reducing alveolar-arterial oxygen transfer, O3 induced vasoconstriction could
28      also acutely induce pulmonary hypertension. Individuals with COPD and coexisting pulmonary
29      hypertension might be subpopulations sensitive to cardiac effects as a consequence of O3
30      exposure.  Acute pulmonary hypertension could potentially affect cardiac function by increasing
31      right ventricular workloads.  Oral or inhaled vasodilators are used in patients to reduce

        August 2005                            AX6-129     DRAFT-DO NOT QUOTE OR CITE

-------
 1      pulmonary artery pressure to improve right ventricular function (Santak et al., 1998).
 2      Consequently, inducing pulmonary vasoconstriction in these patients would perhaps worsen their
 3      condition, especially if their right ventricular function was already compromised.  There are
 4      reduced spirometric and symptom responses to O3 exposure with age (see Section AX6.5.7). It is
 5      conceivable, therefore, that COPD patients and elderly individuals (due to their decreased
 6      symptomatic responses to ambient O3) might further increase their risk of adverse
 7      cardiopulmonary responses by continuing their exposures beyond the point where young healthy
 8      adults might experience discomfort and cease exposure.
 9
10
11      AX6.11    OZONE MIXED WITH OTHER POLLUTANTS
12           Controlled laboratory studies simulating conditions of ambient exposures have failed for
13      the most part to demonstrate  significant adverse effects either in healthy subjects, atopic
14      individuals, or in young and middle-aged asthmatics.
15
16      AX6.11.1  Ozone and Sulfur Oxides
17           The difference in solubilities and other chemical properties of O3 and SOX seems to limit
18      chemical interaction and formation of related species in the mixture of these pollutants either in
19      liquid or gaseous phase. Laboratory studies reviewed in the previous O3 criteria document
20      (Table AX6-14) reported, except for one study (Linn et al., 1994), no significant effects on
21      healthy individuals exposed to mixtures of O3 and SO2 or H2SO4 aerosol. In the study of Linn
22      et al. (1994), which was a repeated 6.5 h exposure protocol, O3 alone and O3 + H2SO4 induced
23      significant spirometric decrements in healthy adults and asthmatics, but the magnitude of effects
24      between exposure atmospheres was not significant. Asthmatic and atopic subjects showed
25      somewhat enhanced or potentiated response to mixtures or sequential exposure, respectively;
26      however, the observed effects were almost entirely attributable to O3 (U.S. Environmental
27      Protection Agency,  1996). Thus, in both healthy and asthmatic subjects, the interactive effects
28      of O3 and other pollutants were marginal and the response was dominated by O3.
29           Since 1994, the only laboratory study that examined the health effects of a mixture of O3
30      and sulfur oxides (SO2 and H2SO4) has been that of Linn et al. (1997).  In this study, the
31      investigators closely simulated ambient summer haze air pollution conditions in Uniontown, PA

        August 2005                           AX6-130      DRAFT-DO NOT QUOTE OR CITE

-------
                                                       Table AX6-14.  Ozone Mixed with Other Pollutants3
to
O
O
X
 H
 b
 o
 O
 H
O
 o
 H
 W
 O
 O
 HH
 H
 W
Concentration11 Number and
Exposure Duration Exposure Gender of Subject
ppm Hg/in3 Pollutant and Activity Conditions' Subjects Characteristics
Observed Effect(s)
Reference
Sulfur-Containing Pollutants
0.0
0.1 +
0.1 +





0.2
0.3

0.12
0.30

0.05


0.12









0 Air 4h 25 °C
196"+ 03+ IE 15' ex/ 15' rest 50% RH
262" + S02 + VE = 22 L/min
101" H2S04




392 03 90 min. 21 °C
564 NO2 VE ~ 32 L/min 50% RH
H2SO4 IE 3x15 min
235 03 1. 5 h with IE for 22 °C
564 NO2 2 consecutive days; 65% RH
70 H2SO4 VE = 23.2L/min
HNO3


235 O3 6.5 h 21 °C
100 H2SO4 2 consecutive days 50% RH
50 min exercise/h
VE = 29 L/min






8 M, 7 F Healthy
1 M, 4 F Asthmatic
10M, 11 F Allergic
All NS,
9 to 12 yrs. old



24 Asthmatic NS,
(17 M, 7F) 11 to 18 years old

22 completed Asthmatic NS,
study; adolescents; NS, 12 to
15M, 7F 19 years old



8 M, 7 F Nonasthmatic NS,
22 to 41 years old

13M, 17 F Asthmatic NS,
18 to 50 years old





Spirometry, PEER and subjective
symptoms score showed no meaningful
changes between any condition for a
total study population. The symptoms
score reported by a subset of
asthmatics/allergies were positively
associated with inhaled concentration
ofH2SO4(p = 0.01).
H2SO4/O3/NO2, O3/NO2 and clean air
produced similar responses

No significant pulmonary function
changes following any exposure
compared to response to clean air.
Six additional subjects started the
study, but dropped out due to
uncomfortable symptoms.
Exposure to O3 or O3 + H2SO4 induced
significant decrements in
forced expiratory function. Differences
between O3 and O3 + H2SO4 were, at
best, marginally significant. O3 is the
more important pollutant for inducing
respiratory effects. A few asthmatic
and nonasthmatic subjects were more
responsive to O3 + H2SO4 than to
O, alone.
Linn et al.
(1997)






Linnetal.
(1995)

Koenig et al.
(1994)




Linn et al.
(1994)








         0.08         157     O3         3-h exposure to aerosol,
         0.12         235     O3         followed 24 h later by a
         0.18         353     O3         3-h exposure to O3. IE
                      100     NaCl       (10 min per half hour)
                      100     H2SO4      VE = 4 times resting
                                        (30 to 364 min)
21 °C           16 M, 14 F       Nonasthmatic NS,       No significant changes in symptoms or
-40% RH                        18 to 45 years old       lung function with any aerosol/O3
                10M, 20 F       Asthmatic NS,          combination in the healthy group.
                                21 to 42 years old       In asthmatics, H2SO4 preexposure
                                                      enhanced the small decrements in FVC
                                                      that occurred following exposure to
                                                      0.18ppmO3. Asthmatics had no
                                                      significant changes on FEV; with any
                                                      O3 exposures, but symptoms were
                                                      greater.
Utell et al.
(1994)
Frampton
etal. (1995)

-------
                                     Table AX6-14 (cont'd).  Ozone Mixed with Other Pollutants3
to
o
o
X
to
H

6
o


o
H

O

o
H
W

O


O
HH
H
W
Concentration11

ppm

Hg/m3

Pollutant
Number and
Exposure Duration Exposure Gender of
and Activity Conditions' Subjects

Subject
Characteristics


Observed Effect(s)


Reference
Sulfur-Containing Pollutants (cont'd)
0.12
0.10



0.25



235
262



490
1,200
to
1,600
03
SO2



03
H2SO4


1 h (mouthpiece) 22 °C 8 M, 5 F
IE VE ~ 30 L/min 75% RH
45-min exposure to air or
O3, followed by 1 5-min
exposure to O3 or SO2
2h 35 °C 9M
IE 83% RH
VE = 30 to 32 L/min

Allergic asthmatics,
12 to 18 years old,
medications withheld
for at least 4 h before
exposures
Healthy NS,
19 to 29 years old


Prior exposure to O3 potentiated
pulmonary function responses to SO2;
decrements in FEVj were -3, -2, and
-8% for the air/O3, O3/O3, and O3/SO2
exposures, respectively.
No significant effects of exposure to O3
alone or combined with H2SO4 aerosol.


Koenig et al.
(1990)



Horvath
etal. (1987)


Nitrogen-Containing Pollutants
0.0
0.2
0.4
0.2+0.4

0.1
0.2
0.1+0.2




0.0
0.36
0.36
0.75
0.36

0.45
0.60


0
392
752


196
376





0
706
677
1,411
943

883
1,129


Air
03
NO2
O3+NO2

03
N02
03+N02




Air
03
NO2
NO2
SO2

03
NO2


3 h 9 M, 2 F
IE 10' ex/ 20' rest
VE =32 L/min


6h
IE 10' ex/ 20' rest
VE =32 L/min
T = 25°C
RH = 50%


2 h Head only 6 M, 6 F
rest exposure




2-h random exposures to 23.6 °C 6 M
FA, O3, NO2, and O3 + 62% RH 2 F
NO2; IE; VE = 26-29
L/min
Atopic asthmatics
22 to 41 yrs. old










Healthy NS
19 to 33 yrs. old




Healthy, NS, 56 to
85 years old


Exposure to NO2 alone had minimal
effects on FEVj. However, O3 alone or
in combination elicited significantly
greater decline in FEVj in a short (3 h)
exposure (higher concentrations) than a
long (6 h) exposure where the effects
were nonsignificant. Allergen
challenge inhalation significantly
reduced PD20 FEV; in all short but not
the long exposures. No additive or
potentiating effects have been
observed.
For NO2 and SO2 the absorbed fraction
of O3 increased relative (to baseline)
whereas after O3 exposure it decreased.
The differences explained by an
increased production of O3-reactive
substrate in ELF due to inflammation.
Exercise-induced cardiac output was
smaller with O3 + NO2 exposure
compared to FA or O3 alone.

Jenkins et al.
(1999)










Rigas et al.
(1997)




Drechsler-
Parks et al.
(1995)


-------
                                                 Table AX6-14 (cont'd). Ozone Mixed with Other Pollutants3
to
o
o
X
H
6
o
o
H
O
o
H
W
O
O
HH
H
W
Concentration11
ppm
(ig/m3 Pollutant
Exposure Duration
and Activity
Exposure
Conditions'
Number and
Gender of
Subjects
Subject
Characteristics
Observed Effect(s)
Reference
         Nitrogen-Containing Pollutants (cont'd)
0.30
0.60









0.2





0.0
0.12
0.30
0.12+0.30

589 O3
1,129 N02









392 O3
500 HNO3
H20



0 Air
235 03
564 N02
03 + N02

2-h exposure to NO2 or 21 °C
FA, followed 3 h later by 40% RH
2-h exposure to O3, IE
VE = 20 L/min/m2 BSA







5 h 20 °C
IE (50 min/h exercise) 5% RH
VE ~ 40 L/min
2 h HNO3 or H2O fog or
air, followed by 1-h break,
followed by 3 h O3
1 h (mouthpiece) 22 °C
IE 75% RH
VE = 33 L/min
VE = 35 L/min

21 F Healthy NS,
18 to 34 years old









6 M, 4 F Healthy NS,
minimum of 1 0%
decrement in FEV;
after 3 h exposure to
0.20 ppm O3 with
50 min exercise/h
5 M, 7 F Healthy NS,
12 to 17 years old

9 M, 3 F Asthmatic
13 to 18 years old
No significant effect of NO2 exposures
on any measured parameter. Sequential
exposure of NO2 followed by O3
induced small but significantly larger
decrements in FEVj and FEF25.75 than
FA/O3 sequence. Subjects had
increased airway responsiveness to
methacholine after both exposures, with
significantly greater responsiveness
after the NO2/O3 sequences than after
the FA/O3 sequence.
Exposure to HNO3 or H2O fog followed
by O3 induced smaller pulmonary
function decrements than air followed
by03.


Findings inconsistent across cohorts
and atmospheres. No significant
differences in FEV; and RT between
asthmatics and healthy, or between
atmospheres and cohorts.
Hazucha
etal. (1994)









Aris etal.
(1991)




Koenig et al.
(1988)



          0.30        589    O3          1 h (mouthpiece)
          0.60        1,129   N02        CE
                                        VE ~ 70 L/min for men
                                        VB ~ 50 L/min for women
20 M, 20 F
Healthy NS,
21.4±1.5(SD)
years old for F,
22.7 ± 3.3 (SD) years
old for M
No differences between responses to
O3 and NO2 + O3 for spirometric
parameters. Increase in SRaw with
NO2 + O3 was significantly less than for
O, alone.
Adams etal.
(1987)
0.30
0.30

0.15
0.15

0.15
0.15
0.15

589
564
200
294
284
200
294
282
393
200
03
N02
H2S04
03
NO2
H2SO4
03
N02
S02
H2S04
2 hCE for 20 min 28 to 29 °C 6M
V = 25 L/min 50 to 60% RH

2 h, 60 min 6 M
total exercise
V ~ 25 L/min
2 h, 60 min 3 M
total exercise
V = 25 L/min

Healthy subjects, Possible small decrease in SGaw
some smokers

Possible small decrease in SGaw


Possible small decrease in FEV;



Kagawa
(1986)









-------
6-
Table AX6-14 (cont'd). Ozone Mixed With Other Pollutants3
M Concentration11
to
O ppm H2/m
,_, FF TO
Pollutant
Number and
Exposure Duration Exposure Gender of Subject
and Activity Conditions' Subjects Characteristics
Observed Effect(s)
Reference
Peroxyacetyl Nitrate
0.45
0.60
0.13
0.45
0.30
0.485
0.27
883
1,129
644
883
1,485
952
1,337
03
N02
PAN
03
PAN
03
PAN
2h
IE
VE = 25 L/min
2h
IE
VE = 27 L/min
2h
IE
VE = 25 L/min
24 °C 8M, 8F Healthy NS;
55 to 58% RH 8 M, 8 F 19 to 26 years old;
51 to 76 years old
22 °C 3 M, 5 F Healthy NS,
60% RH mean age = 24 years
21 °C 10 F Healthy NS,
WBGT 19 to 36 years old
No differences between
responses to O3 alone, O3 + NO2, O3 +
PAN, or 03 + N02 + PAN.
No differences between
responses to exposure to
O3 alone and O3 + PAN.
Exposure to the mixture of PAN + O3
induced decrements in FVC and FEVj
averaging 10% greater than observed
following exposure to O3 alone.
Drechsler-
Parks et al.
(1989)
Drechsler-
Parks et al.
(1987b)
Horvath
etal. (1986)
Particle-Containing Pollutants
> 0.0
X 0.12
0
235b +
153"
Air
03 +
PM25
2-2.5 h
rest
22 °C 15M, 10 F Healthy NS
30% RH 18to50yrs. old
Neither systolic nor diastolic pressure
has been affected by pollutants
exposure despite a significant brachial
Brook et al.
(2002)
 H
 6
 o
 o
 H
O
 o
 H
 W
 O
                                                                                                                                     artery constriction and a reduction in
                                                                                                                                     arterial diameter when compared to
                                                                                                                                     filtered air (p = 0.03). Absence of
                                                                                                                                     flow- and nitroglycerin-mediated
                                                                                                                                     brachial artery dilatation.
           "See Appendix A for abbreviations and acronyms.
           'Grouped by pollutant mixture.
O
HH
H
W

-------
 1      as well as controlled the selection of study subjects with the objective to corroborate earlier
 2      reported findings of an epidemiologic study of Neas et al. (1995).  The subjects were 41 children
 3      (22F/19M) 9 to 12 yrs old. Of these, 26 children had history of asthma or allergy. During a
 4      14-day study period, children were exposed on the 4th and  11th day for 4 hrs (IE, 15 min @ avg.
 5      VE 22 L/min) in random order to air and a mixture of 0.10 ppm O3, 0.10 ppm SO2 and 42 to
 6      198 mg/m3 H2SO4 (mean cone. 101 mg/m3, 0.6 mm MMAD). The effects of controlled
 7      exposures were assessed by spirometry. Except for exposure days, children used diaries to
 8      record activity, respiratory symptoms, location, and PEFR.  Thus,  every exposure day was
 9      bracketed by 3 days of monitoring.  Spirometry, PEFR, and respiratory symptoms score showed
10      no meaningful changes between any condition for a total study population. The symptoms score
11      reported by a subset of asthmatic/allergic subjects was positively associated with the inhaled
12      concentration of H2SO4 (p = 0.01). However, the reported symptoms were different from the
13      ones reported in the Uniontown study (Neas et al., 1995). Although retrospective statistical
14      power calculations using these study observations for the symptoms score, PEFR, and
15      spirometric endpoints were sufficient to detect with >80% probability the  same magnitude of
16      changes as observed in Uniontown, the effects were minimal and not significant.  The divergent
17      observations of the two studies have been explained by the  presence of an unidentified
18      environmental factor in Uniontown, differences in physico-chemical properties of acid,
19      differences in time course of exposure and history of previous exposure of children to pollutants,
20      psychological and physiological factors related to chamber exposures,  and by other conjectures.
21
22      AX6.11.2  Ozone and Nitrogen-Containing Pollutants
23           Nitrogen dioxide is a key component of the photooxidation cycle and formation of O3.
24      Both  gases are almost invariably present in ambient atmosphere. Compared to O3, NOX species
25      have limited solubility and moderate oxidizing capability.  Both O3 and NO2 are irritants and
26      tissue oxidants and exert their toxic actions through many common mechanisms.  The regional
27      dosimetry and the primary sites of action of O3 and NO2 overlap but are not the same.  Since
28      these gases are relatively insoluble in water, they will likely penetrate into the peripheral airways
29      that are more sensitive to damage than better protected conducting airways. The controlled
30      studies reviewed in the previous O3 criteria document (Table AX6-14) generally reported only
31      small pulmonary function changes after combined exposures of NO2 or nitric acid (FDSTO3) with

        August 2005                           AX6-13 5      DRAFT-DO NOT QUOTE OR CITE

-------
 1      O3, regardless if the interactive effects were potentiating or additive. In two of these studies, the
 2      effects reached statistical significance, but they were not coherent. Preexposure with NO2
 3      potentiated both spirometric and nonspecific airway reactivity response following subsequent O3
 4      exposure (Hazucha et al., 1994); however, exposure to NO2 + O3 mixture blunted SRaw increase
 5      as compared to O3 alone (Adams et al.,1987). As with O3 and SOX mixtures, the effects have
 6      been dominated by O3 (U.S. Environmental Protection Agency, 1996).
 7           Combined exposure to O3 and NO2 also blunted the exercise-induced increase in cardiac
 8      output found with FA and O3 exposures alone (Drechsler-Parks, 1995). Eight healthy older
 9      subjects (56 to 85 years of age) were exposed for 2 h to FA, 0.60 ppm NO2, 0.45 ppm O3, and to
10      0.60 ppm NO2 + 0.45 ppm O3 while alternating 20-min periods of rest and exercise.  Cardiac
11      output, FIR, stroke volume, and systolic time intervals were measured by noninvasive impedance
12      cardiography at the beginning of each exposure, while the subjects were at rest, and again during
13      the last 5 min  of exercise. Metabolic  exercise data (VE, VO2, fB) also were  measured.  There
14      were no statistically significant differences between exposures for HR, VE,  VO2, fB,  stroke
15      volume, or systolic time intervals. Exercise increased cardiac output after all exposures;
16      however, the incremental increase over rest was significantly smaller for the combined O3
17      and NO2 exposures.  The authors speculated that nitrate and nitrite reaction  products from the
18      interaction of  O3 and NO2 cross the air/blood interface in the lungs, causing peripheral
19      vasodilation and a subsequent drop in cardiac output. No major cardiovascular effects of O3
20      only exposures have been reported in human subjects (see Section AX6.10).
21           Despite  suggested potentiation of O3 response by NO2 in healthy subjects, it is unclear
22      what response, and at what dose, either sequential or combined gas exposures will induce
23      in asthmatics.  Jenkins et al. (1999) exposed 11 atopic asthmatics  in random order to air,
24      0.1 ppm O3, 0.2 ppm NO2, and 0.1 ppm O3 + 0.2  ppm NO2 for 6 h (IE for 10 min @ 32 L/min
25      every 40 min). Two weeks later, 10 of these subjects were exposed for 3 h to doubled
26      concentrations of these gases (i.e., 0.2 ppm O3, 0.4 ppm NO2, and 0.2 ppm O3 + 0.4 ppm NO2)
27      employing the same exercise regimen. Immediately following each exposure, subjects were
28      challenged with allergen (D. pteronyssinus) and PD20 FEVj was determined. Exposure to NO2
29      alone had minimal effects on FEVj or airway responsiveness.  However, O3 alone or in
30      combination with NO2 elicited a significantly (p < 0.05) greater decline in FEVj in a short (3 h)
        August 2005                            AX6-136      DRAFT-DO NOT QUOTE OR CITE

-------
 1      exposure (higher concentrations) than the long (6 h) exposure, where the effects were not
 2      significant.  Allergen challenge inhalation significantly (p = 0.018 to 0.002) reduced PD20 FEVj
 3      in all short, but not the long, exposures. No associations were observed between pollutant
 4      concentrations and physiologic endpoints.  The statistical analyses of these data suggest that the
 5      combined effect (O3 + NO2) on lung function (FVC, FEVj) was not significantly greater than the
 6      effect of individual gases for 6-h exposures, thus no additive or potentiating effects have been
 7      observed. Shorter 3-h exposures using twice as high NO2 concentrations, however, showed
 8      significant FEVj decrements following exposures to atmospheres containing O3.  The analysis
 9      also suggests that it is the inhaled concentration, rather than total dose, that determines lung
10      airway responsiveness to allergen.
11           The potential for interaction between O3 and other gas mixtures was studied by Rigas et al.
12      (1997). They used an O3 bolus absorption technique to determine how exposures to O3, NO2,
13      and SO2 will affect distribution of O3 adsorption by airway mucosa.  The selected O3 bolus
14      volume was set to reach lower conducting airways. Healthy young nonsmokers (6F/6M) were
15      exposed on separate days at rest in a head dome to 0.36 ppm O3, 0.36 ppm NO2, 0.75 ppm NO2
16      and 0.75 ppm SO2 for 2 h. The rationale for the selection of these gases was their differential
17      absorption.  Because O3 and NO2 are much less soluble in liquid (i.e., ELF) than  SO2, they are
18      expected to penetrate deeper into the lung than SO2 which is absorbed more quickly in the
19      epithelial lining fluid of the upper airways. The actual experimental measurements have shown
20      that during continuous NO2 and SO2 exposure the absorbed fraction of an O3 bolus in lower
21      conducting airways increased relative to baseline, whereas during continuous O3 exposure the O3
22      bolus fraction in lower conducting airways decreased. The authors attempted to explain the
23      differences by suggesting that there may be increased  production of an O3-reactive substrate
24      in epithelial lining fluid due to airway inflammation.  As interpreted by the investigators,
25      during NO2 and SO2 exposures the substrate was not depleted by these gases and so could react
26      with the O3 bolus, whereas during O3 exposure the substrate was depleted, causing the fractional
27      absorption of the O3 bolus to decrease. Greater absorption in males than females for all gases
28      was attributed to anatomical differences in  the bronchial tree.
29
        August 2005                            AX6-137     DRAFT-DO NOT QUOTE OR CITE

-------
 1      AX6.11.3  Ozone and Other Pollutant Mixtures Including Particulate Matter
 2           Almost all of the studies published over the last twenty years investigating the health
 3      effects of mixtures of O3 with other air pollutants involved peroxyacetyl nitrate (PAN). These
 4      studies on healthy individuals exposed under laboratory conditions came from the Horvath
 5      laboratory at UC Santa Barbara (Table AX6-13). In the last of this series of studies, Drechsler-
 6      Parks and colleagues (1989) found the same equivocal interaction of O3 and PAN as in previous
 7      studies, which is attributable to O3 exposure alone (U.S. Environmental Protection Agency,
 8      1996). Subsequently, only a couple of studies have investigated the effects of more complex air
 9      pollutant mixtures on human pathophysiology under controlled conditions.
10           It is not only the interaction between air pollutants in ambient air; but, as Rigas et al.
11      (1997) has found, an uneven distribution of O3,  SO2, and NO2 absorption in the lower conducting
12      airways of young healthy subjects may modulate pathophysiologic response as well.  Exposure
13      to SO2 and NO2 increased, while exposure to O3 decreased, the absorbing capacity of the airways
14      for O3. The authors have suggested that SO2 or NO2 -inflamed airways release additional
15      substrates into the epithelial lining fluid that react with O3, thus progressively removing O3 from
16      the airway lumen.  This mechanism may explain findings  of antagonistic response (e.g., Adams
17      et al., 1987; Dreschler-Parks, 1995) when the two gases are combined in an exposure
18      atmosphere.
19           The mechanisms by which inhalation exposure to other complex ambient atmospheres
20      containing particulate matter (PM) and O3 induce cardiac  events frequently reported in
21      epidemiologic studies are rarely studied in human subjects under laboratory conditions.
22      Recently, Brook et al. (2002) have reported changes in brachial artery tone and reactivity in
23      healthy nonsmokers following 2-h exposures to a mixture of 0.12 ppm O3 and 153 |ig/m3 of
24      concentrated ambient PM25, and a control  atmosphere of filtered air with a trace of O3
25      administered in random order.  Neither systolic nor diastolic pressure was affected by pollutant
26      exposure despite a significant brachial artery constriction  and a reduction in arterial diameter
27      when compared to filtered air (p = 0.03). The authors postulate that changes in arterial tone may
28      be a plausible mechanism of air pollution-induced cardiac events. However, the observations of
29      no changes in blood pressure, and an absence of flow- and nitroglycerin- mediated brachial
30      artery dilatation, cast some doubt on the plausibility of this mechanism.  A number of other
        August 2005                            AX6-13 8     DRAFT-DO NOT QUOTE OR CITE

-------
 1      proposed mechanisms advanced to establish a link between cardiac events due to pollution and
 2      changes in vasomotor tone based on the findings of this study are purely speculative.
 3
 4
 5      AX6.12  CONTROLLED STUDIES OF AMBIENT AIR EXPOSURES
 6           A large amount of informative O3 exposure-effects data has been obtained in controlled
 7      laboratory exposure studies under a variety of different experimental conditions. However,
 8      laboratory simulation of the variable pollutant mixtures present in ambient air is not practical.
 9      Thus, the exposure effects of one or several artificially generated pollutants (i.e., a simple
10      mixture) on pulmonary function and symptoms may not explain responses to ambient air where
11      complex pollutant mixtures exist. Epidemiologic studies, which do investigate ambient air
12      exposures, do not typically provide the level of control and monitoring necessary to adequately
13      characterize short term responses.  Thus, controlled exposures to ambient air using limited
14      numbers of volunteers have been used to try and bridge the gap between laboratory and
15      community exposures.
16
17      AX6.12.1   Mobile Laboratory Studies
18           As presented in previous criteria documents (U.S. Environmental Protection Agency, 1986;
19      1996), quantitatively useful information on the effects of acute exposure to photochemical
20      oxidants on pulmonary function and symptoms responses originated from field studies using a
21      mobile laboratory. These field studies involved subjects exposed to ambient air, FA without
22      pollutants, or FA containing artificially generated concentrations of O3 that are comparable to
23      those measured in the ambient environment. As a result, measured pulmonary responses in
24      ambient air can be directly compared to those found in more artificial or controlled conditions.
25      However, the mobile laboratory shares some of the same limitations of stationary exposure
26      laboratories (e.g., limited number of both subjects and artificially generated pollutants for
27      testing). Further, mobile laboratory ambient air studies are dependent on ambient outdoor
28      conditions which can be unpredictable, uncontrollable, and not completely characterizable.
29           As summarized in Table AX6-15, investigators in California used a mobile laboratory and
30      demonstrated that pulmonary effects of ambient air in Los Angeles residents are related to O3
31      concentration and level of exercise (Avol et al.,  1983, 1984, 1985a,b,c, 1987; Linn et al., 1980,

        August 2005                             AX6-139     DRAFT-DO NOT  QUOTE OR CITE

-------
I
S-
to
o
^













X
o^

o




O
3>
'•Tj
H
I
0
o
2|
0
h-j


H
W
O
O
H
W
Table AX6-15. Acute Effects of Ozone in Ambient Air in Field
Mean Ozone
Concentration" Ambient

ppm MS/m3 (°C) Duration (VE)
0.113 ±.033 221 ±65 33 ± 1 Ih CE (22 L/min)



0.144 ±.043 282 ±84 32 ± 1 Ih CE (32 L/min)




0.153 ±.025 300 ±49 32 ±2 Ih CE (53 L/min)





0.156 ±.055 306 ±107 33 ±4 Ih CE (38 L/min)




0.165 ±.059 323 ±115 33 ± 3 Ih CE (42 L/min)





0.174 ±.068 341 ±133 33 ±2 2h IE (2 times resting)
at 15-min intervals





aSee Appendix A for abbreviations and acronyms.
bRanked by lowest level of O3 in ambient air, presented as the mean ± SD.
cMean±SD.





Number of Subjects
66 healthy children,
8 to 1 1 years old


59 healthy
adolescents,
12 to 15 years old


50 healthy adults
(competitive
bicyclists)



48 healthy adults,
50 asthmatic adults



60 "healthy" adults
(7 were asthmatic)




34 "healthy" adults,
30 asthmatic adults










Studies with a Mobile Laboratory"



Observed Effect(s)
No significant changes in forced expiratory
function and symptoms of breathing
discomfort after exposure to 0. 1 1 3 ppm O3 in
ambient air.
Small significant decreases in FVC (-2.1%),
FEV075 (-4.0%), FEV; (-4.2%), andPEFR
(-4.4%) relative to control with no recovery
during a 1-h postexposure rest; no significant
increases in symptoms.
Mild increases in symptoms scores and
significant decreases in FEV; (-5.3%) and
FVC; mean changes in ambient air were not
statistically different from those in purified air
containing 0.16 ppm O3.

No significant changes for total symptom
score or forced expiratory performance in
normals or asthmatics; however, FEV;
remained low or decreased further (-3%) 3 h
after ambient air exposure in asthmatics.
Small significant decreases in FEV[ (-3.3%)
and FVC with no recovery during a 1-h
postexposure rest; TLC decreased and AN2
increased slightly.


Increased symptom scores and small
significant decreases in FEV[ (-2.4%), FVC,
PEFR, and TLC in both asthmatic and healthy
subjects; however, 25/34 healthy subjects were
allergic and "atypically" reactive to polluted
ambient air.










Reference
Avol et al.
(1987)


Avol et al.
(1985a,b)



Avol et al.
(1984, 1985c)




Linn et al.
(1983)
Avol et al.
(1983)

Linn et al.
(1983)
Avol et al.
(1983)


Linn et al.
(1980, 1983)











-------
 1      1983). Avol et al. (1987) observed no significant pulmonary function or symptoms responses in
 2      children (8 to 11 years) engaged in moderate continuous exercise for 1 h while breathing
 3      ambient air with an O3 concentration of 0.113 ppm. However, significant pulmonary function
 4      decrements and increased symptoms of breathing discomfort were observed in healthy
 5      exercising (1 h continuous) adolescents (Avol et al., 1985a,b), athletes, (Avol et al., 1984, 1985c)
 6      and lightly exercising asthmatic subjects (Linn et al., 1980, 1983) at O3 concentrations averaging
 7      from 0.144 to 0.174 ppm. Many of the healthy subjects with a history of allergy appeared to be
 8      more responsive to O3 than "nonallergic" subjects (Linn et al., 1980, 1983), although a
 9      standardized evaluation of atopic status was not performed.  Comparative studies of exercising
10      athletes (Avol et al., 1984, 1985c) with chamber exposures to oxidant-polluted ambient air
11      (mean O3 concentration of 0.153 ppm) and purified air containing a controlled concentration of
12      generated O3 at 0.16 ppm showed similar pulmonary function responses and symptoms,
13      suggesting that acute exposures to coexisting ambient pollutants had minimal contribution to
14      these responses under the typical summer ambient conditions in Southern California. This
15      contention is similar to, but extends, the laboratory finding of no significant difference in
16      pulmonary function effects between O3 and O3 plus PAN exposures (Drechsler-Parks, 1987b).
17      Additional supporting evidence is provided in Section AX6.11.
18
19      AX6.12.2  Aircraft Cabin Studies
20          Respiratory symptoms and pulmonary function effects resulting from exposure to O3 in
21      commercial aircraft flying at high altitudes, and in altitude-simulation studies, have been
22      reviewed elsewhere (U.S. Environmental Protection Agency, 1986, 1996). Flight attendants,
23      because of their physical activities at altitude, tend to receive higher exposures. In a series of
24      hypobaric chamber studies of nonsmoking subjects exposed to 1,829 m (6,000 ft) and O3 at
25      concentrations of 0.2 and 0.3 ppm for 3 or 4 h (Lategola et al., 1980a,b),  increased symptoms
26      and pulmonary function decrements occurred at 0.3 ppm but not at 0.2 ppm.
27          Commercial aircraft cabin O3 levels were reported to be very low (average concentration
28      0.01 to 0.02 ppm) during 92 randomly selected smoking and nonsmoking flights in 1989 (Nagda
29      et al., 1989). None of these flights recorded O3 concentrations exceeding the 3-h time-weighted
30      average (TWA) standard of 0.10 ppm promulgated by the Federal Aviation Administration
31      (FAA, 1980), probably due to the use of O3-scrubbing catalytic filters (Melton,  1990). However,

        August 2005                           AX6-141     DRAFT-DO  NOT QUOTE OR CITE

-------
 1     in-flight O3 exposure can still occur because catalytic filters are not necessarily in continuous use
 2     during flight. Other factors to consider in aircraft cabins, however, are erratic temperature
 3     changes, lower barometric pressure and oxygen pressure, and lower humidity, often reaching
 4     levels between 4 and 17% (Rayman, 2002).
 5          Ozone contamination aboard high-altitude aircraft also has been an interest to the U.S. Air
 6     Force because of complaints by crew members of frequent symptoms of dryness and irritation of
 7     the eyes, nose, and throat and an occasional cough (Hetrick et al., 2000). Despite the lack of
 8     ventilation system modifications as used in commercial aircraft, the O3 concentrations never
 9     exceeded the FAA ceiling limit of 0.25 ppm and exceeded the 3-h TWA of 0.10 ppm only 7% of
10     the total monitored flight time (43 h).  The authors concluded that extremely low average
11     relative humidity (12%) during flight operations was most likely responsible for the reported
12     symptoms.
13
       August 2005                            AX6-142     DRAFT-DO NOT QUOTE OR CITE

-------
  1       REFERENCES

  2       Adams, W. C. (2000a) Feasibility study of prolonged ozone inhalation exposure via face mask. Inhalation Toxicol.
  3             12:299-313.
  4       Adams, W. C. (2000b) Ozone dose-response effects of varied equivalent minute ventilation rates. J. Exposure Anal.
  5             Environ. Epidemiol. 10: 217-226.
  6       Adams, W. C. (2002) Comparison of chamber and face-mask 6.6-hour exposures to ozone on pulmonary function
  7             and symptoms responses. Inhalation Toxicol. 14: 745-764.
  8       Adams, W. C. (2003a) Comparison of chamber and face mask 6.6-hour exposure to 0.08 ppm ozone via
  9             square-wave and triangular profiles on pulmonary responses. Inhalation Toxicol. 15: 265-281.
10       Adams, W. C. (2003b) Relation of pulmonary responses induced by 6.6-h exposures to 0.08 ppm ozone and 2-h
11             exposures to 0.30 ppm ozone via chamber and face-mask inhalation. Inhalation Toxicol. 15: 745-759.
12       Adams, W. C.; Ollison, W. M. (1997) Effects of prolonged simulated ambient ozone dosing patterns on human
13             pulmonary function and symptomatology. Presented at: 90th annual meeting of the Air & Waste Management
14             Association; June; Toronto, Ontario,  Canada. Pittsburgh, PA: Air & Waste Management Association; paper
15             no. 97-MP9.02.
16       Adams, W. C.; Schelegle, E. S. (1983) Ozone and high ventilation effects on pulmonary function and endurance
17             performance. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 55: 805-812.
18       Adams, W. C.; Savin, W. M.; Christo, A. E. (1981) Detection of ozone toxicity during continuous exercise via the
19             effective dose concept. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 51: 415-422.
20       Adams, W. C.; Brookes, K.  A.; Schelegle, E. S. (1987) Effects of NO2 alone and in combination with O3 on young
21             men and women. J. Appl. Physiol. 62: 1698-1704.
22       Alexis, N.; Urch, B.; Tarlo,  S.; Corey, P.; Pengelly, D.; O'Byrne, P.; Silverman, F. (2000) Cyclooxygenase
23             metabolites play a different role in ozone-induced pulmonary function decline in asthmatics compared to
24             normals. Inhalation Toxicol. 12: 1205-1224.
25       Alexis, N. E.; Becker, S.; Bromberg, P. A.; Devlin, R.; Peden, D. B. (2004) Circulating GDI Ib expression correlates
26             with the neutrophil response and airway mCD14 expression is enhanced following ozone exposure in
27             humans. Clin. Immunol. Ill: 126-131.
28       Alpert, S. E.; Walenga, R. W.; Jaspers, I.; Qu, Q.; Chen, L. C. (1997) Ozone inactivates Cyclooxygenase in human
29             tracheal epithelial cells without altering PGHS-2 mRNA or protein. Am. J. Physiol. 272: L879-L887.
30       Arab, L.; Steck-Scott, S.; Fleishauer, A. T. (2002) Lycopene and the lung. Exp. Biol. Med. 227: 894-899.
31       Aris, R.; Christian, D.; Sheppard, D.; Balmes, J. R. (1991) The effects of sequential exposure to acidic fog and
32             ozone on pulmonary function in exercising subjects. Am. Rev. Respir. Dis. 143: 85-91.
33       Aris, R. M.; Christian, D.; Hearne, P. Q.; Kerr, K.; Finkbeiner, W. E.; Balmes, J. R. (1993) Ozone-induced airway
34             inflammation in human subjects as determined by airway lavage and biopsy. Am. Rev. Respir. Dis.
35             148: 1363-1372.
36       Aris, R. M.; Tager, L; Christian, D.; Kelly, T.; Balmes, J. R.  (1995) Methacholine responsiveness is not associated
37             with O3-induced decreases inFEVj. Chest 107: 621-628.
38       Avissar, N. E.; Reed, C. K.; Cox, C.; Frampton, M. W.; Finkelstein, J. N. (2000) Ozone, but not nitrogen dioxide,
39             exposure  decreases glutathione peroxidases in epithelial lining fluid of human lung. Am. J. Respir. Crit. Care
40             Med. 162: 1342-1347.
41       Avol, E. L.; Linn, W. S.;  Shamoo, D. A.; Venet, T. G.; Hackney, J. D. (1983) Acute respiratory effects of Los
42             Angeles smog in continuously exercising adults. J. Air Pollut. Control Assoc. 33: 1055-1060.
43       Avol, E. L.; Linn, W. S.;  Venet, T. G.; Shamoo, D. A.; Hackney, J. D. (1984) Comparative respiratory effects of
44             ozone and ambient oxidant pollution  exposure during heavy exercise. J. Air Pollut. Control Assoc.
45             34:804-809.
46       Avol, E. L.; Linn, W. S.;  Shamoo, D. A.; Valencia, L. M.; Anzar, U. T.; Venet, T. G.;  Hackney, J. D. (1985a)
47             Respiratory effects of photochemical oxidant air pollution in exercising adolescents. Am. Rev. Respir. Dis.
48             132:619-622.
49       Avol, E. L.; Linn, W. S.;  Shamoo, D. A.; Valencia, L. M.; Anzar, U. T.; Hackney, J. D. (1985b) Short-term health
50             effects of ambient  air pollution in adolescents. In: Lee, S. D., ed. Evaluation of the scientific basis for
51             ozone/oxidants standards: transactions of an APCA international specialty conference; November 1984;
52             Houston,  TX.  Pittsburgh, PA: Air Pollution Control Association; pp. 329-336. (APCA international specialty
53             conference transactions: TR-4).
54
         August 2005                                 AX6-143      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Avol, E. L.; Linn, W. S.; Venet, T. G.; Shamoo, D. A.; Spier, C. E.; Hackney, J. D. (1985c) Comparative effects of
  2             laboratory generated ozone and ambient oxidant exposure in continuously exercising subjects. In: Lee, S. D.,
  3             ed. Evaluation of the scientific basis for ozone/oxidants standards: proceedings of an APCA international
  4             specialty conference; November 1984; Houston, TX. Pittsburgh, PA: Air Pollution Control Association; pp.
  5             216-225. (APCA international specialty conference transactions: TR-4).
  6       Avol, E. L.; Linn, W. S.; Shamoo, D. A.; Spier, C. E.; Valencia, L. M; Venet, T. G.; Trim, S. C.; Hackney, J. D.
  7             (1987) Short-term respiratory effects of photochemical oxidant exposure in exercising children. JAPCA
  8             37: 158-162.
  9       Avol, E. L.; Linn, W. S.; Shamoo, D. A.; Hackney, J. D. (1988) Seasonal ozone reactivity  in Los Angeles residents.
10             Presented at: 81st annual meeting of the Air Pollution Control Association; June; Dallas, TX. Pittsburgh, PA:
11             Air Pollution Control Association; paper no. 88-122.6.
12       Ball, B. A.; Folinsbee, L. J.; Peden, D. B.; Kehrl, H. R. (1996) Allergen bronchoprovocation of patients with mild
13             allergic asthma after ozone exposure. J. Allergy Clin. Immunol. 98:  563-572.
14       Balmes, J. R.; Chen, L. L.; Scannell, C.; Tager, I.; Christian, D.; Hearne, P. Q.; Kelly, T.; Aris, R. M. (1996)
15             Ozone-induced decrements in FEVj and FVC do not correlate with measures of inflammation. Am. J. Respir.
16             Crit. Care Med. 153: 904-909.
17       Balmes, J. R.; Aris, R. M.; Chen, L. L.; Scannell, C.; Tager, I. B.; Finkbeiner, W.;  Christian, D.; Kelly, T.; Hearne,
18             P. Q.; Ferrando, R.; Welch, B. (1997) Effects of ozone on normal and potentially sensitive human subjects.
19             part I: airway inflammation and responsiveness to ozone in normal and asthmatic subjects. Cambridge, MA:
20             Health Effects Institute. Research report no. 78; pp 1-37, 81-99.
21       Barman,  S. A. (2001) Effect of protein kinase C inhibition on hypoxic pulmonary vasoconstriction. Am. J. Physiol.
22             Lung Cell Mol. Physiol. 280: L888-L895.
23       Bascom,  R.; Naclerio, R. M.; Fitzgerald, T. K.; Kagey-Sobotka, A.; Proud, D. (1990) Effect of ozone inhalation on
24             the response to nasal challenge with antigen of allergic subjects.  Am. Rev. Respir. Dis. 142: 594-601.
25       Basha, M. A.; Gross, K. B.; Gwizdala, C. J.; Haidar, A. H.; Popovich, J., Jr. (1994) Bronchoalveolar lavage
26             neutrophilia in asthmatic and healthy volunteers after controlled  exposure to ozone and filtered purified air.
27             Chest 106: 1757-1765.
28       Bates, D. V; Bell, G. M.; Burnham, C. D.; Hazucha, M.; Mantha, J.; Pengelly, L. D.; Silverman, F. (1972)
29             Short-term effects of ozone on the lung. J. Appl. Physiol. 32: 176-181.
30       Bayram, H.; Sapsford, R. J.; Abdelaziz, M. M.; Khair, O. A. (2001) Effect of ozone and nitrogen dioxide on the
31             release of proinflammatory mediators from bronchial epithelial cells of nonatopic nonasthmatic subjects and
32             atopic asthmatic patients in vitro. J. Allergy Clin. Immunol. 107: 287-294.
33       Bayram, H.; Rusznak, C.; Khair, O. A.; Sapsford, R. J.; Abdelaziz, M. M. (2002) Effect of ozone and nitrogen
34             dioxide on the permeability of bronchial epithelial cell cultures of non-asthmatic and asthmatic subjects.
35             Clin. Exp. Allergy 32: 1285-1292.
36       Beck, N.; Koenig, J. Q.; Luchtel, D. L.; Altaian, L. C.; Orsborn, M. T.; Kenney, J.  S. (1994) Ozone can increase the
3 7             expression of intercellular adhesion molecule-1 and the synthesis of cytokines by human nasal epithelial cells.
38             Inhalation Toxicol. 6: 345-357.
39       Becker, S.; Madden, M. C.; Newman, S. L.; Devlin, R. B.; Koren, H. S. (1991) Modulation of human alveolar
40             macrophage properties by ozone exposure in vitro. Toxicol. Appl. Pharmacol. 110:  403-415.
41       Bedi, J. F.; Drechsler-Parks, D. M.; Horvath, S. M. (1985) Duration of increased pulmonary function sensitivity to
42             an initial ozone exposure. Am. Ind. Hyg. Assoc. J. 46: 731-734.
43       Bedi, J. F.; Horvath, S. M.; Drechsler-Parks, D. M. (1988) Reproducibility of the pulmonary function response of
44             older men and women to a 2-hour ozone exposure. JAPCA 38: 1016-1019.
45       Bedi, J. F.; Horvath, S. M.; Drechsler-Parks, D. M. (1989) Adaptation by older individuals repeatedly exposed to
46             0.45 parts per million ozone for two hours. JAPCA 39: 194-199.
47       Bergamaschi, E.; De Palma, G.; Mozzoni, P.; Vanni, S.; Vettori, M. V.; Broeckaert, F.; Bernard, A.; Mutti, A.
48             (2001) Polymorphism of quinone-metabolizing enzymes and susceptibility to ozone-induced acute effects.
49             Am. J. Respir. Crit. Care Med. 163: 1426-1431.
50       Bhalla, D. K. (2002) Interactive effects of cigarette smoke and ozone in the induction of lung injury. Toxicol. Sci.
51             65:1-3.
52       Blomberg, A.; Helleday, R.; Pourazar, J.; Stenfors, N.; Kelly, F. J.; Frew, A. J.; Holgate, S. T.; Sandstrom, T. (1997)
53             Early airway and peripheral blood cell responses to 0.20 ppm ozone in healthy human subjects. Eur. Respir. J.
54             10(suppl 25): 274S.
         August 2005                                  AX6-144      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Blomberg, A.; Mudway, I. S.; Nordenhall, C.; Hedenstrom, H.; Kelly, F. I; Frew, A. I; Holgate, S. T.;
  2             Sandstrom, T. (1999) Ozone-induced lung function decrements do not correlate with early airway
  3             inflammatory or antioxidant responses. Eur. Respir. J.  13: 1418-1428.
  4       Blomberg, A.; Mudway, I.; Svensson, M; Hagenbjork-Gustafsson, A.; Thomasson, L.; Helleday, R.; Dumont, X.;
  5             Forsberg, B.; Nordberg, G.; Bernard, A. (2003) Clara cell protein as a biomarker for ozone-induced lung
  6             injury in humans. Eur. Respir. J. 22: 883-888.
  7       Bosson, J.; Stenfors, N.; Bucht, A.; Helleday, R.; Pourazar, J.; Holgate, S. T.; Kelly, f. J.; Sandstrom, T.; Wilson, S.;
  8             Frew, A. J.; Blomberg, A. (2003) Ozone-induced bronchial epithelial cytokine expression differs between
  9             healthy and asthmatic subjects. Clin. Exp. Allergy 33:  777-782.
10       Brook, R. D.; Brook, J. R.; Urch, B.; Vincent, R.; Rajagopalan, S.; Silverman, F. (2002) Inhalation of fine paniculate
11             air pollution and ozone causes acute arterial vasoconstriction in healthy adults. Circulation 105: 1534-1536.
12       Brookes, K. A.; Adams, W. C.; Schelegle, E. S. (1989) 0.35 ppm O3 exposure induces hyperresponsiveness on 24-h
13             reexposure to 0.20 ppm O3. J. Appl. Physiol. 66: 2756-2762.
14       Brummer, M. E. G.; Schwartz, L. W.; McQuillen, N. K. (1977) A quantitative study of lung damage by scanning
15             electron microscopy. Inflammatory cell response to high-ambient levels of ozone. In: Johari,  O.; Becker,
16             R. P., eds. Scanning electron microscopy/1977/11 - biological applications of the SEM: proceedings of the
17             workshops on advances inbiomedical applications of the SEM & STEM; March-April; Chicago, IL.
18             Chicago, IL: IIT Research Institute; pp. 513-518.
19       Bush, M. L.; Asplund, P. T.; Miles, K. A.; Ben-Jebria, A.; Ultman, J. S. (1996) Longitudinal distribution of O3
20             absorption in the lung: gender differences and intersubject variability. J. Appl. Physiol. 81: 1651-1657.
21       Centers for Disease Control and Prevention. (2003) Key clinical activities for quality asthma care: recommendations
22             of the National Asthma Education and Prevention Program. Morb. Mortal. Wkly. Rep. 52(RR-6): 1-8.
23             Available: http://www.cdc.gov/mmwr/PDF/rr/rr5206.pdf (11 April 2003).
24       Chen, L. L.; Tager, I. B.; Peden, D. B.; Christian, D. L.; Ferrando, R. E.; Welch, B. S.; Balmes, J. R. (2004) Effect
25             of ozone exposure on airway responses to inhaled allergen in asthmatic subjects. Chest 125: 2328-2335.
26       Christian, D. L.; Chen, L. L.; Scannell, C. H.; Ferrando, R. E.; Welch, B. S.; Balmes, J. R. (1998) Ozone-induced
27             inflammation is attenuated with multiday exposure. Am. J. Respir. Crit. Care Med. 158: 532-537.
28       Coffey, M. J.; Wheeler, C. S.; Gross, K. B.; Eschenbacher, W. L.; Sporn, P. H. S.; Peters-Golden, M. (1996)
29             Increased 5-lipoxygenase metabolism in the lungs of human subjects exposed to ozone. Toxicology
30             114: 187-197.
31       Coleridge, J. C. G.; Coleridge,  H. M.; Schelegle, E. S.; Green, J.  F. (1993) Acute inhalation of ozone stimulates
32             bronchial C-fibers and rapidly adapting receptors in dogs. J. Appl. Physiol. 74: 2345-2352.
33       Collins, D. V.; Cutillo, A. G.; Armstrong, J. D.; Crapo, R. O.; Kanner, R. E.; Tocino, I.; Renzetti, A. D., Jr. (1986)
34             Large airway size, lung  size, and maximal expiratory flow in healthy nonsmokers. Am. Res. Respir. Dis.
35             134:951-955.
36       Corradi, M.; Alinovi, R.; Goldoni, M.; Vettori, M.; Folesani,  G.; Mozzoni, P.; Cavazzini, S.; Bergamaschi, E.; Rossi,
37             L.; Mutti, A. (2002) Biomarkers of oxidative stress after controlled human exposure to ozone. Toxicol. Lett.
38             134:219-225.
39       Criqui, G. I.; Solomon, C.; Welch, B. S.; Ferrando, R.  E.; Boushey, H. A.; Balmes, J. R. (2000) Effects of
40             azithromycin on ozone-induced airway neutrophilia and cytokine release. Eur. Respir. J. 15: 856-862.
41       Cropp, G. J. A.; Bernstein, I. L.; Boushey, H. A., Jr.; Hyde, R. W.; Rosenthal, R. R.; Specter, S. L.; Townley, R. G.
42             (1980) Guidelines for bronchial inhalation challenges with pharmacologic and antigenic agents. ATS News
43             (Spring): 11-19.
44       David, G. L.; Romieu, I.; Sienra-Monge, J. J.; Collins, W. J.;  Ramirez-Aguilar, M.; Del Rio-Navarro, B. E.;
45             Reyes-Ruiz, N. I.; Morris, R.  W.; Marzec, J. M.; London, S. J. (2003) Nicotinamide adenine  dinucleotide
46             (Phosphate) reduced:quinone  oxidoreductase and glutathione s-transferase ml polymorphism and childhood
47             asthma. Am. J. Respir. Crit. Care Med. 168: 1199-1204.
48       Delaunois, A.; Segura, P.; Montano, L. M.; Vargas, M. H.; Ansay, M.; Gustin, P. (1998) Comparison of
49             ozone-induced effects on lung mechanics and hemodynamics in the rabbit. Toxicol. Appl. Pharmacol.
50             150:58-67.
51       Devlin, R. B.; Koren, H. S.  (1990) The use of quantitative two-dimensional gel electrophoresis to analyze changes in
52             alveolar macrophage proteins in humans exposed to ozone. Am. J. Respir. Cell Mol. Biol. 2:  281-288.
53       Devlin, R. B.; McDonnell, W. F.; Koren, H. S.; Becker, S. (1990) Prolonged exposure of humans to 0.10 and
54             0.08 ppm ozone results in inflammation in the lung. Presented at:  83rd annual meeting of the Air & Waste
55             Management Association; June; Pittsburgh, PA. Pittsburgh, PA: Air & Waste Management Association;
56             paper no. 90-150.2.
         August 2005                                  AX6-145      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Devlin, R. B.; McDonnell, W. F.; Mann, R.; Becker, S.; House, D. E.; Schreinemachers, D.; Koren, H. S. (1991)
  2             Exposure of humans to ambient levels of ozone for 6.6 hours causes cellular and biochemical changes in the
  3             lung. Am. J. Respir. Cell Mol. Biol. 4: 72-81.
  4       Devlin, R. B.; McKinnon, K. P.; Noah, T.; Becker, S.; Koren, H.  S. (1994) Ozone-induced release of cytokines and
  5             fibronectinby alveolar macrophages and airway epithelial cells. Am. J. Physiol. 266: L612-L619.
  6       Devlin, R. B.; McDonnell, W. F.; Becker, S.; Madden, M. C.; McGee, M. P.; Perez, R.; Hatch, G.; House, D. E.;
  7             Koren, H. S. (1996) Time-dependent changes of inflammatory mediators in the lungs of humans exposed to
  8             0.4 ppm ozone for 2 hr: a comparison of mediators found in bronchoalveolar lavage fluid 1 and 18 hr after
  9             exposure. Toxicol. Appl. Pharmacol. 138:  176-185.
10       Devlin, R. B.; Folinsbee, L. J.; Biscardi, F.; Hatch, G.; Becker, S.; Madden, M. C.; Robbins, M.; Koren, H. S. (1997)
11             Inflammation and cell damage induced by repeated exposure of humans to ozone. Inhalation Toxicol.
12             9:211-235.
13       Dimeo, M. J.; Glenn, M. G.; Holtzman, M. J.; Sheller, J. R.; Nadel, J. A.; Boushey, H.  A. (1981) Threshold
14             concentration of ozone causing an increase in bronchial reactivity in humans and adaptation with repeated
15             exposures. Am. Rev. Respir. Dis. 124: 245-248.
16       Drechsler-Parks, D. M. (1995) Cardiac output effects of O3  and NO2 exposure in healthy older adults. Toxicol. Ind.
17             Health 11: 99-109.
18       Drechsler-Parks, D. M.; Bedi, J. F.; Horvath, S. M. (1984) Interaction of peroxyacetyl nitrate and ozone on
19             pulmonary functions. Am. Rev. Respir. Dis. 130: 1033-1037.
20       Drechsler-Parks, D. M.; Bedi, J. F.; Horvath, S. M. (1987a) Pulmonary function responses of older men and women
21             to ozone exposure. Exp. Gerontol. 22: 91-101.
22       Drechsler-Parks, D. M.; Bedi, J. F.; Horvath, S. M. (1987b) Pulmonary function desensitization on repeated
23             exposures to the combination of peroxyacetyl nitrate and ozone. JAPCA 37: 1199-1201.
24       Drechsler-Parks, D. M.; Bedi, J. F.; Horvath, S. M. (1989) Pulmonary function responses of young and older adults
25             to mixtures of O3, NO2 and PAN. Toxicol. Ind. Health 5: 505-517.
26       Drechsler-Parks, D. M.; Horvath, S. M.; Bedi, J. F. (1990) The "effective dose" concept in older adults exposed to
27             ozone. Exp. Gerontol. 25: 107-115.
28       Eschenbacher, W.  L.; Ying, R. L.; Kreit, J. W.; Gross, K. B. (1989) Ozone-induced lung function changes in normal
29             and asthmatic subjects and the effect of indomethacin. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.;
30             Grant, L. D., eds. Atmospheric ozone research and its policy implications: proceedings of the 3rd US-Dutch
31             international symposium; May 1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier
32             Science Publishers; pp. 493-499. (Studies in environmental science 35).
33       Fahy, J.  V.; Wong, H. H.; Liu, J. T.; Boushey, H. A. (1995) Analysis of induced sputum after air and ozone
34             exposures in healthy subjects. Environ. Res. 70: 77-83.
35       Federal Aviation Administration. (1980) Airplane cabin ozone contamination. F. R. (January 21) 45: 3880-3883.
36       Fernandes, A. L. G.; Molfino, N. A.; McClean, P. A.; Silverman, F.; Tarlo, S.; Raizenne, M.; Slutsky, A. S.; Zamel,
37             N. (1994) The effect of pre-exposure to 0.12 ppm of ozone on exercise-induced  asthma. Chest
38             106: 1077-1082.
39       Folinsbee, L. J.; Hazucha, M. J. (1989) Persistence of ozone-induced changes in lung function and airway
40             responsiveness. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone research
41             and its policy implications: proceedings of the 3rd US-Dutch international symposium; May 1988; Nijmegen,
42             The Netherlands. Amsterdam, The Netherlands: Elsevier Science Publishers; pp. 483-492. (Studies in
43             environmental science 35).
44       Folinsbee, L. J.; Hazucha, M. J. (2000) Time course of response to ozone exposure in healthy adult females.
45             Inhalation Toxicol. 12: 151-167.
46       Folinsbee, L. J.; Horvath, S. M. (1986) Persistence of the acute effects of ozone exposure. Aviat. Space Environ.
47             Med.  57: 1136-1143.
48       Folinsbee, L. J.; Silverman, F.; Shephard, R. J.  (1975) Exercise responses following ozone exposure. J. Appl.
49             Physiol. 38: 996-1001.
50       Folinsbee, L. J.; Silverman, F.; Shephard, R. J.  (1977) Decrease of maximum work performance following ozone
51             exposure. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 42: 531-536.
52       Folinsbee, L. J.; Drinkwater, B. L.;  Bedi, J. F.;  Horvath, S. M. (1978) The influence of exercise on the pulmonary
53             function changes due to exposure to low concentrations of ozone. In: Folinsbee, L. J.; Wagner, J. A.; Borgia,
54             J. F.; Drinkwater, B. L.; Gliner, J. A.; Bedi, J. F., eds. Environmental  stress: individual human adaptations.
55             New York, NY: Academic Press; pp. 125-145.
         August 2005                                 AX6-146       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Folinsbee, L. I; Bedi, J. F.; Horvath, S. M. (1980) Respiratory responses in humans repeatedly exposed to low
  2             concentrations of ozone. Am. Rev. Respir. Dis. 121: 431-439.
  3       Folinsbee, L. J.; Bedi, J. F.; Horvath, S. M. (1984) Pulmonary function changes after 1 h continuous heavy exercise
  4             in0.21 ppmozone. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 57: 984-988.
  5       Folinsbee, L. J.; Horstman, D. H.; Vorona, R. D.; Prince, J. M.; Berry, M. (1986) Determinants of endurance
  6             performance during ozone inhalation. Presented at: 30th International Union of Physiological Sciences
  7             congress; July; Vancouver, BC, Canada. Proc. Int. Union Physiol. Sci.  16: 176.
  8       Folinsbee, L. J.; McDonnell, W. F.; Horstman, D. H. (1988) Pulmonary function and symptom responses after
  9             6.6-hour exposure to 0.12 ppm ozone with moderate exercise. JAPCA 38: 28-35.
10       Folinsbee, L. J.; Devlin, R. B.; Abdul-Salaam, S.; Koren, H. S. (1993) Repeated severe ozone exposure causes
11             depressed baseline spirometry. Am. Rev. Respir. Dis. 147: A638.
12       Folinsbee, L. J.; Horstman, D. H.; Kehrl, H. R.; Harder, S.; Abdul-Salaam, S.; Ives, P. J. (1994) Respiratory
13             responses to repeated prolonged exposure to 0.12 ppm ozone. Am. J. Respir. Crit. Care Med. 149: 98-105.
14       Folinsbee, L. J.; Devlin, R. B.; Robbins, M. K.; Biscardi, F. H.; Abdul-Salaam, S.; Koren, H. S. (1998) Repeated
15             exposure of humans to ozone: pulmonary function and symptom responses. Research Triangle Park, NC: U.S.
16             Environmental Protection Agency; National Center for Environmental Assessment; unpublished data.
17       Foster, W. M.; Stetkiewicz, P. T.  (1996) Regional clearance of solute from the respiratory epithelia: 18-20 h
18             postexposure to ozone. J. Appl. Physiol. 81: 1143-1149.
19       Foster, W. M.; Costa, D. L.; Langenback, E. G. (1987) Ozone exposure alters tracheobronchial mucociliary function
20             in humans. J. Appl. Physiol. 63: 996-1002.
21       Foster, W. M.; Silver, J. A.; Groth, M. L. (1993) Exposure to ozone alters regional function and particle dosimetry in
22             the human lung. J. Appl. Physiol. 75: 1938-1945.
23       Foster, W. M.; Wills-Karp, M.; Tankersley, C. G.; Chen, X.; Paquette, N. C. (1996) Bloodborne markers in humans
24             during multiday exposure to ozone. J. Appl. Physiol. 81: 794-800.
25       Foster, W. M.; Weinmann, G. G.; Menkes, E.; Macri, K. (1997) Acute exposure of humans to ozone impairs small
26             airway function. Ann. Occup. Hyg. 41(suppl. 1): 659-666.
27       Foster, W. M.; Brown, R. H.; Macri, K.; Mitchell, C. S. (2000) Bronchial reactivity of healthy subjects: 18-20 h
28             postexposure to ozone. J. Appl. Physiol. 89: 1804-1810.
29       Fox, S. D.; Adams, W. C.; Brookes, K. A.; Lasley, B. L. (1993) Enhanced response to ozone exposure during the
30             follicular phase of the menstrual cycle. Environ. Health Perspect. 101: 242-244.
31       Foxcroft, W. J.; Adams, W. C. (1986) Effects of ozone exposure on four consecutive days on work performance and
32             VO2max. J. Appl. Physiol. 61: 960-966.
33       Frampton, M. W.;  Morrow, P. E.; Cox, C.; Levy, P. C.; Condemi, J. J.; Speers, D.; Gibb, F. R.; Utell, M. J. (1995)
34             Sulfuric acid aerosol followed by ozone exposure in healthy and asthmatic subjects. Environ. Res. 69: 1-14.
35       Frampton, M. W.;  Morrow, P. E.; Torres, A.; Cox, C.; Voter, K. Z.; Utell, M.  J.; Gibb, F. R.; Speers, D. M. (1997a)
36             Ozone responsiveness in smokers and nonsmokers. Am. J. Respir. Crit. Care Med. 155: 116-121.
37       Frampton, M. W.;  Balmes, J. R.;  Cox, C.; Krein, P. M.; Speers, D. M.; Tsai, Y.; Utell, M. J. (1997b) Effects of
3 8             ozone on normal and potentially sensitive human subjects. Part III: mediators of inflammation in
39             bronchoalveolar lavage fluid from nonsmokers, smokers, and asthmatic subjects exposed to ozone:
40             a collaborative study. Cambridge, MA: Health Effects Institute. Research report 78; pp. 73-99, 81-99.
41       Frank, R.; Liu, M.  C.; Spannhake, E. W.; Mlynarek, S.; Macri, K.; Weinmann, G. G. (2001) Repetitive ozone
42             exposure of young adults:  evidence of persistent small airway dysfunction. Am. J. Respir. Crit. Care Med.
43             164: 1253-1260.
44       Freed, A. N.; Chou, C. L.; Fuller, S. D.; Croxton, T. L. (1996) Ozone-induced vagal reflex modulates airways
45             reactivity in rabbits. Respir. Physiol. 105: 95-102.
46       Gabrielson, E. W.; Yu, X.-Y.; Spannhake, E. W. (1994) Comparison of the toxic effects of hydrogen peroxide and
47             ozone on cultured human bronchial epithelial cells. Environ. Health Perspect. 102: 972-974.
48       Gerrity, T. R.; Bennett, W. D.; Kehrl, H.; DeWitt, P. J. (1993) Mucociliary clearance of inhaled particles measured
49             at 2 h after ozone exposure in humans. J. Appl. Physiol.  74: 2984-2989.
50       Gliner, J. A.; Matsen-Twisdale, J. A.; Horvath, S. M. (1979) Auditory and visual sustained attention during ozone
51             exposure. Aviat. Space Environ. Med. 50: 906-910.
52       Gliner, J. A.; Horvath, S. M.; Sorich, R. A.; Hanley, J. (1980) Psychomotor performance during ozone exposure:
53             spectral and discriminant function analysis of EEG. Aviat. Space Environ. Med. 51: 344-351.
54       Gliner, J. A.; Horvath, S. M.; Folinsbee, L. J. (1983) Preexposure to low ozone concentrations does not diminish the
55             pulmonary function response on exposure to higher ozone concentrations. Am. Rev. Respir. Dis. 127: 51-55.
         August 2005                                 AX6-147       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Golden, J. A.; Nadel, J. A.; Boushey, H. A. (1978) Bronchial hyperirritability in healthy subjects after exposure to
  2             ozone. Am. Rev. Respir. Dis. 118: 287-294.
  3       Gomez, F. P.; Rodriguez-Roisin, R. (2002) Global initiative for chronic obstructive lung disease (GOLD) guidelines
  4             for chronic obstructive pulmonary disease. Curr. Opin. Pulm. Med. 8: 81-86.
  5       Gong, H., Jr.; Tierney, D. F.  (1995) Evaluation of chronic obstructive pulmonary disease patients for ozone
  6             sensitivity: validation of health advisories. Sacramento, CA.: California State Air Resources Board, Research
  7             Division; CARB-A133-123. Available from: NTIS, Springfield, VA; PB98-153620.
  8       Gong, H., Jr.; Bradley, P. W.; Simmons, M. S.; Tashkin, D. P. (1986) Impaired exercise performance and pulmonary
  9             function in elite cyclists during low-level ozone exposure in a hot environment. Am. Rev. Respir. Dis.
10             134:726-733.
11       Gong, H., Jr.; Shamoo, D. A.; Anderson, K. R.; Linn, W. S. (1997a) Responses of older men with and without
12             chronic obstructive pulmonary disease to prolonged ozone exposure. Arch. Environ. Health 52: 18-25.
13       Gong, H., Jr.; McManus, M.  S.;  Linn, W. S. (1997b) Attenuated response to repeated daily ozone exposures in
14             asthmatic subjects. Arch.  Environ. Health 52: 34-41.
15       Gong, H., Jr.; Wong, R.; Sarma, R. J.; Linn, W.  S.;  Sullivan, E. D.; Shamoo, D. A.; Anderson, K. R.; Prasad, S. B.
16             (1998) Cardiovascular effects of ozone exposure in human volunteers. Am. J. Respir. Crit. Care Med.
17             158: 538-546.
18       Graham, D. E.; Koren, H.  S.  (1990) Biomarkers of inflammation in ozone-exposed humans: comparison of the nasal
19             and bronchoalveolar lavage. Am. Rev. Respir. Dis. 142: 152-156.
20       Graham, D.; Henderson, F.; House, D.  (1988) Neutrophil influx measured in nasal lavages of humans exposed to
21             ozone. Arch. Environ. Health 43: 228-233.
22       Greve, J. M.; Davis, G.; Meyer,  A. M.; Forte, C. P.; Yost, S. C; Marlor, C. W.; Kamarck, M. E.; McClelland, A.
23             (1989) The major human rhinovirus receptor is ICAM-1. Cell 56: 839-847.
24       Grievink, L.; Jansen, S. M. A.; Van't Veer, P.; Brunekreef, B. (1998) Acute effects of ozone on pulmonary function
25             of cyclists receiving antioxidant supplements. Occup. Environ. Med. 55: 13-17.
26       Grievink, L.; Zijlstra, A. G.; Ke, X.; Brunekreef, B. (1999) Double-blind intervention trial on modulation of ozone
27             effects on pulmonary function by antioxidant supplements. Am. J. Epidemiol. 149:  306-314.
28       Gwynn, R. C.; Thurston, G. D. (2001) The burden of air pollution: impacts among racial minorities. Environ. Health
29             Perspect. 109(suppl. 4): 501-506.
30       Hackney, J. D.; Linn, W. S.;  Buckley, R. D.; Hislop, H. J. (1976) Studies in adaptation to ambient oxidant air
31             pollution: effects of ozone exposure in Los Angeles residents vs. new arrivals. Environ. Health Perspect.
32             18: 141-146.
33       Hackney, J. D.; Linn, W. S.;  Mohler, J. G.; Collier, C. R. (1977a) Adaptation to short-term respiratory effects of
34             ozone in men exposed repeatedly. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 43: 82-85.
35       Hackney, J. D.; Linn, W. S.;  Karuza, S. K.; Buckley, R. D.; Law, D. C.; Bates, D. V.; Hazucha, M.; Pengelly, L. D.;
36             Silverman, F. (1977b) Effects of ozone exposure in Canadians and southern Californians: evidence for
37             adaptation? Arch. Environ. Health 32: 110-116.
38       Hackney, J. D.; Linn, W. S.;  Shamoo, D. A.; Avol,  E. L. (1989) Responses of selected reactive and nonreactive
39             volunteers to ozone exposure in high and low pollution seasons. In: Schneider, T.; Lee, S. D.; Welters, G. J.
40             R.; Grant, L. D., eds. Atmospheric ozone research and its policy implications: proceedings of the 3rd
41             US-Dutch international symposium; May 1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands:
42             Elsevier Science Publishers; pp. 311-318. (Studies in environmental science 35).
43       Hamilton, R. F.; Li, L.; Eschenbacher,  W. L.; Szweda, L.; Holian, A. (1998) Potential involvement of
44             4-hydroxynonenal in the response of human  lung cells to ozone. Am. J. Physiol. 274: L8-L16.
45       Hanania, N.  A.; Tarlo, S. M.; Silverman, F.; Urch, B.; Senathirajah, N.; Zamel, N.; Corey, P. (1998) Effect of
46             exposure to low levels of ozone on the response to inhaled allergen in allergic asthmatic patients. Chest
47             114:752-756.
48       Hazbun, M. E.; Hamilton, R.; Holian, A.; Eschenbacher, W. L. (1993) Ozone-induced increases in substance P and
49             8-epi-prostaglandin F2a in the airways of human subjects. Am. J. Respir. Cell Mol. Biol. 9:  568-572.
50       Hazucha, M. J. (1987) Relationship between ozone exposure and pulmonary function changes. J. Appl. Physiol.
51             62:1671-1680.
52       Hazucha, M. J. (1993) Meta-analysis and "effective dose" revisited. In: Mohr, U., ed. Proceedings of the  3rd
53             International Inhalation Symposium on Advances in Controlled Clinical Inhalation  Studies; October,  1991.
54             Berlin, Germany: Springer-Verlag; pp. 247-256.
55       Hazucha, M. J.; Sant'Ambrogio, G. (1993) Effects of ozone on the activity of slowly (SAR) and rapidly adapting
56             (RAR) receptors in cats. FASEB J. 7: 407A.
         August 2005                                  AX6-148       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Hazucha, M; Silverman, F.; Parent, C.; Field, S.; Bates, D. V. (1973) Pulmonary function in man after short-term
  2             exposure to ozone. Arch. Environ. Health 27: 183-188.
  3       Hazucha, M. J.; Bates, D. V.; Bromberg, P. A. (1989) Mechanism of action of ozone on the human lung. J. Appl.
  4             Physiol. 67: 1535-1541.
  5       Hazucha, M. J.; Folinsbee, L. J.; Seal, E., Jr. (1992) Effects of steady-state and variable ozone concentration profiles
  6             on pulmonary function. Am. Rev. Respir. Dis. 146: 1487-1493.
  7       Hazucha, M. J.; Folinsbee, L. J.; Seal, E.; Bromberg, P. A. (1994) Lung function response of healthy women after
  8             sequential exposures to NO2 and O3. Am. J. Respir. Crit. Care Med. 150: 642-647.
  9       Hazucha, M. J.; Madden, M.; Pape, G.; Becker, S.; Devlin, R.; Koren, H. S.; Kehrl, H.; Bromberg, P. A. (1996)
10             Effects of cyclo-oxygenase inhibition on ozone-induced respiratory inflammation and lung function changes.
11             Eur. J. Appl. Physiol. Occup. Med. 73: 17-27.
12       Hazucha, M. J.; Folinsbee, L. J.; Bromberg, P. A. (2003) Distribution and reproducibility of spirometric response to
13             ozone by gender and age. J. Appl. Physiol. 95: 1917-1925.
14       Henderson, F. W.; Dubovi, E. J.; Harder, S.; Seal, E., Jr.; Graham, D. (1988) Experimental rhinovirus infection in
15             human volunteers exposed to ozone. Am. Rev. Respir. Dis. 137: 1124-1128.
16       Hetrick, S. M.; Gould, W. D.; Christensen, D. E. (2000) Inflight cabin ozone aboard long duration C-5  airlift
17             missions: a historical issue revisited. Aviat. Space Environ. Med. 71: 408-414.
18       Hiltermann, T. J. N.; Stolk, J.; Hiemstra, P. S.; Fokkens, P. H. B.; Rombout, P. J. A.; Sont, J. K.; Sterk, P. J.;
19             Dijkman, J. H. (1995) Effect of ozone exposure on maximal airway narrowing in non-asthmatic  and asthmatic
20             subjects. Clin. Sci. 89: 619-624.
21       Hiltermann, T. J. N.; de Bruijne, C. R.;  Stolk, J.; Zwinderman, A. H.; Spieksma, F. Th. M.; Roemer,  W.;
22             Steerenberg, P. A.; Fischer, P. H.; van Bree, L.; Hiemstra, P. S. (1997) Effects of photochemical air pollution
23             and allergen exposure on upper respiratory tract inflammation in asthmatics. Am. J. Respir.  Crit. Care Med.
24             156: 1765-1772.
25       Hiltermann, T. J. N.; Peters, E. A.; Alberts, B.; Kwikkers, K.; Borggreven, P. A.; Hiemstra, P. S.; Dijkman, J. H.;
26             van Bree, L. A.; Stolk, J. (1998)  Ozone-induced airway hyperresponsiveness in patients with asthma: role of
27             neutrophil-derived serine proteinases. Free Radical Biol. Med. 24: 952-958.
28       Hiltermann, J. T. N.; Lapperre, T.  S.; Van Bree, L.; Steerenberg, P. A.; Brahim, J. J.; Sont, J. K.; Sterk, P. J.;
29             Hiemstra, P. S.; Stolk, J.  (1999) Ozone-induced inflammation assessed in sputum and bronchial  lavage fluid
30             from asthmatics: a new noninvasive tool in epidemiologic studies on air pollution and asthma. Free Radical
31             Biol. Med. 27: 1448-1454.
32       Holtzman, M. J.; Cunningham, J. H.; Sheller, J. R.; Irsigler, G. B.; Nadel, J. A.; Boushey, H. A. (1979) Effect of
33             ozone on bronchial reactivity in atopic and nonatopic subjects. Am. Rev. Respir. Dis. 120: 1059-1067.
34       Holz, O.; Jorres, R. A.; Timm, P.;  Mticke, M.; Richter, K.; Koschyk, S.; Magnussen, H. (1999) Ozone-induced
35             airway inflammatory changes differ between individuals and are reproducible. Am. J. Respir.  Crit. Care Med.
36             159:776-784.
37       Holz, O.; Mticke, M.; Paasch, K.; Bohme, S.; Timm, P.; Richter, K.; Magnussen, H.; Jorres, R. A.  (2002) Repeated
3 8             ozone exposures enhance bronchial allergen responses in subjects with rhinitis or asthma. Clin. Exp. Allergy.
39             32:681-689.
40       Holz, O.; Tal-Singer, R.; Kanniess, F.; Simpson, K. J.; Gibson, A.; Vessey, R. S. J.; Janicki, S.; Magnussen, H.;
41             Jorres, R. A.; Richter, K. (2005)  Validation of the human ozone challenge model as a tool for assessing
42             anti-inflammatory drugs in early development. J. Clin. Pharmacol. 45: 498-503.
43       Horstman, D. H.; Folinsbee, L. J.; Ives, P. J.; Abdul-Salaam, S.; McDonnell, W. F. (1990) Ozone concentration and
44             pulmonary response relationships for 6.6-hour exposures with five hours of moderate exercise to 0.08, 0.10,
45             and 0.12 ppm. Am. Rev.  Respir.  Dis. 142: 1158-1163.
46       Horstman, D. H.; Ball, B. A.; Brown, J.; Gerrity, T.; Folinsbee, L. J. (1995) Comparison of pulmonary  responses of
47             asthmatic and nonasthmatic subjects performing light exercise while exposed to a low level of ozone.
48             Toxicol. Ind. Health 11: 369-385.
49       Horvath, S. M.; Gliner, J. A.; Matsen-Twisdale, J. A. (1979) Pulmonary function and maximum exercise responses
50             following acute ozone exposure. Aviat. Space Environ. Med. 50: 901-905.
51       Horvath, S. M.; Gliner, J. A.; Folinsbee, L. J. (1981) Adaptation to ozone: duration of effect. Am. Rev. Respir. Dis.
52             123:496-499.
53       Horvath, S. M.; Bedi, J. F.; Drechsler-Parks, D. M. (1986) Effects of peroxyacetyl nitrate alone and in combination
54             with ozone in healthy young women. J. Air Pollut. Control Assoc.  36: 265-270.
55       Horvath, S. M.; Folinsbee, L. J.; Bedi, J. F. (1987) Combined effect of ozone and sulfuric acid on pulmonary
56             function in man. Am. Ind. Hyg. Assoc. J.  48: 94-98.
         August 2005                                  AX6-149      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Horvath, S. M; Bedi, J. F.; Drechsler-Parks, D. M.; Williams, R. E. (1991) Alterations in pulmonary function
  2             parameters during exposure to 80 ppb ozone for 6.6 hours in healthy middle aged individuals. In: Berglund,
  3             R. L.; Lawson, D. R.; McKee, D. J., eds. Tropospheric ozone and the environment: papers from an
  4             international conference; March 1990; Los Angeles, CA. Pittsburgh, PA: Air & Waste Management
  5             Association; pp. 59-70. (A&WMA transactions series: no. TR-19).
  6       Housley, D. G.; Eccles, R.; Richards, R. J. (1996) Gender difference in the concentration of the antioxidant uric acid
  7             in human nasal lavage. Acta Oto-Laryngol. 116: 751-754.
  8       Hu, S.-C.; Ben-Jebria, A.; Ultman, J. S. (1994) Longitudinal distribution of ozone absorption in the lung: effects of
  9             respiratory flow. J. Appl. Physiol. 77: 574-583.
10       Inkley, S. R.; Maclntyre, W. J. (1973) Dynamic measurement of ventilation-perfusion with xenon-133 at resting
11             lung volumes. Am. Rev. Respir. Dis. 107: 429-441.
12       Jenkins, H. S.; Devalia, J. L.; Mister, R. L.; Bevan, A. M.; Rusznak, C.; Davies, R. J. (1999) The effect of exposure
13             to ozone and nitrogen dioxide on the airway response of atopic asthmatics to inhaled allergen: dose- and
14             time-dependent effects. Am. J. Respir. Crit. Care Med. 160: 33-39.
15       Joad, J. P.; Kott, K. S.; Brie, J. M. (1996) The local C-fiber contribution to ozone-induced effects on the isolated
16             guinea pig lung. Toxicol. Appl. Pharmacol. 141: 561-567.
17       Johnson, D. A.; Winters, R. S.; Woolley, T.; Graham, D.; Henderson, F. W. (1986) Ozone effects on
18             alpha-1-proteinase inhibitor in vivo: blood plasma inhibitory activity is unchanged. Exp. Lung Res.
19             11:95-103.
20       Jorres, R.; Nowak, D.; Magnussen, H.; Speckin, P.; Koschyk, S. (1996) The effect of ozone exposure  on allergen
21             responsiveness in subjects with asthma or rhinitis. Am. J. Respir. Crit. Care Med. 153: 56-64.
22       Jorres, R. A.; Holz, O.; Zachgo, W.; Timm, P.; Koschyk, S.; Mtiller, B.; Grimminger, F.; Seeger, W.; Kelly, F. J.;
23             Dunster, C.; Frischer, T.; Lubec, G.; Waschewski, M.; Niendorf, A.; Magnussen, H. (2000) The effect of
24             repeated ozone exposures on inflammatory markers in bronchoalveolar lavage fluid and mucosal biopsies.
25             Am. J. Respir. Crit. Care Med. 161: 1855-1861.
26       Kabel, J. R.; Ben-Jebria, A.; Ultman, J. S. (1994) Longitudinal distribution of ozone absorption in the lung:
27             comparison of nasal and oral quiet breathing. J. Appl. Physiol. 77: 2584-2592.
28       Kagawa, J. (1986) Experimental studies on human health effects of aerosol and gaseous pollutants.  In: Lee, S. D.;
29             Schneider, T.; Grant, L. D.; Verkerk, P. J., eds. Aerosols:  research, risk assessment and control strategies -
30             proceedings of the second U.S.-Dutch international symposium;  May 1985; Williamsburg, VA. Chelsea, MI:
31             Lewis Publishers, Inc.; pp.  683-697.
32       Kaneko, K.; Milic-Emili, J.; Dolovich, M. B.; Dawson, A.; Bates, D. V. (1966) Regional distribution of ventilation
33             and perfusion as a function of body position. J. Appl. Physiol. 21: 767-777'.
34       Kehrl, H. R.; Vincent, L. M.; Kowalsky, R. J.; Horstman, D. H.; O'Neil, J. J.; McCartney, W. H.; Bromberg, P. A.
35             (1987) Ozone exposure increases respiratory epithelial permeability in humans. Am. Rev. Respir. Dis.
36             135: 1124-1128.
37       Kehrl, H. R.; Vincent, L. M.; Kowalsky, R. J.; Horstman, D. H.; O'Neil, J. J.; McCartney, W. H.; Bromberg, P. A.
38             (1989) Ozone-induced changes in the pulmonary clearance of 99mTc-DTPA in man. In: Schneider, T.; Lee, S.
39             D.; Wolters,  G. J. R.; Grant, L. D., eds. Atmospheric ozone research and its policy implications: proceedings
40             of the 3rd US-Dutch international symposium; May 1988; Nijmegen, The Netherlands. Amsterdam,
41             The Netherlands: Elsevier Science Publishers; pp. 343-351.  (Studies in environmental science 35).
42       Kehrl, H. R.; Peden, D. B.; Ball, B. A.; Folinsbee, L. J.; Horstman, D. H. (1999) Increased specific  airway reactivity
43             of persons with mild allergic asthma after 7.6 hours of exposure  to 0.16 ppm ozone. J. Allergy. Clin.
44             Immunol.  104: 1198-1204.
45       King, T. K. C.; Briscoe, W. A. (1968) The distribution of ventilation, perfusion, lung volume and transfer factor
46             (diffusing  capacity) in patients with obstructive lung disease. Clin. Sci. 35: 153-170.
47       Kleinman, M. T.; Mautz, W. J.; Bjarnason, S. (1999) Adaptive and non-adaptive responses in rats exposed to ozone,
48             alone and in mixtures, with acidic aerosols. Inhalation Toxicol. 11: 249-264.
49       Koenig, J. Q.; Covert, D. S.; Marshall, S. G.; Van Belle,  G.; Pierson, W. E. (1987) The effects of ozone and nitrogen
50             dioxide on pulmonary function in healthy and in asthmatic adolescents. Am. Rev. Respir. Dis.
51             136:1152-1157.
52       Koenig, J. Q.; Covert, D. S.; Smith, M. S.; Van Belle, G.; Pierson, W. E. (1988) The pulmonary effects of ozone and
53             nitrogen dioxide alone and combined in healthy and asthmatic adolescent subjects. Toxicol. Ind. Health
54             4:521-532.
55       Koenig, J. Q.; Covert, D. S.; Hanley, Q. S.; Van Belle, G.; Pierson, W.  E. (1990) Prior exposure to ozone potentiates
56             subsequent response to sulfur dioxide in adolescent asthmatic subjects. Am. Rev. Respir. Dis. 141: 377-380.
         August 2005                                  AX6-150      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Koenig, J. Q.; Covert, D. S.; Pierson, W. E.; Hanky, Q. S.; Rebolledo, V.; Dumler, K.; McKinney, S. E. (1994)
  2             Oxidant and acid aerosol exposure in healthy subjects and subjects with asthma. Part I: effects of oxidants,
  3             combined with sulfuric or nitric acid, on the pulmonary function of adolescents with asthma. Cambridge, MA:
  4             Health Effects Institute; pp. 1-36; research report no. 70.
  5       Konig, G.; Rommelt, H.; Kienle, H.; Dirnagl, K.; Polke, H.; Fruhmann, G. (1980) Anderung der
  6             bronchomotorischen Reagibilitat des Menschen durch Einwirkung von Ozon [Changes in the bronchial
  7             reactivity of humans caused by the influence of ozone]. Arbeitsmed. Sozialmed. Praeventivmed.
  8             151:261-263.
  9       Koren, H. S.; Devlin, R. B.; Graham, D. E.; Mann, R.; McGee, M. P.; Horstman, D. H.; Kozumbo, W. J.; Becker, S.;
10             House, D. E.; McDonnell, W. F.; Bromberg, P. A. (1989a) Ozone-induced inflammation in the lower airways
11             of human subjects. Am. Rev. Respir. Dis. 139: 407-415.
12       Koren, H. S.; Devlin, R. B.; Graham, D. E.; Mann, R.; McDonnell, W. F. (1989b) The inflammatory response in
13             human lung exposed to ambient levels of ozone. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.; Grant, L. D.,
14             eds. Atmospheric ozone research and its policy implications: proceedings of the 3rd US-Dutch international
15             symposium; May 1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier Science
16             Publishers; pp. 745-753. (Studies in environmental science 35).
17       Koren, H. S.; Hatch, G. E.; Graham, D. E. (1990) Nasal lavage as a tool in assessing acute inflammation in response
18             to inhaled pollutants. Toxicology 60: 15-25.
19       Koren, H. S.; Devlin, R. B.; Becker, S.; Perez, R.; McDonnell, W. F. (1991) Time-dependent changes of markers
20             associated with inflammation in the lungs of humans exposed to ambient levels of ozone. Toxicol. Pathol.
21             19:406-411.
22       Kozumbo, W. J.; Hanley, N. M.; Agarwas, S.; Thomas, M. J.; Madden, M. C. (1996) Products of ozonized
23             arachidonic acid potentiate the formation of DNA single strand breaks in cultured human lung cells. Environ.
24             Mol. Mutagen. 27: 185-195.
25       Kreit, J. W.; Gross, K. B.; Moore, T. B.; Lorenzen, T. J.; D'Arcy, J.; Eschenbacher, W. L. (1989) Ozone-induced
26             changes in pulmonary function and bronchial responsiveness in asthmatics. J. Appl. Physiol. 66: 217-222.
27       Krishna, M. T.; Springall, D.; Meng, Q.-H.; Withers, N.; Macleod, D.; Biscione,  G.; Frew, A.; Polak, J.; Holgate, S.
28             (1997a) Effects of ozone on epithelium and sensory nerves in the bronchial mucosa of healthy humans. Am.  J.
29             Respir. Crit. Care Med.  156: 943-950.
30       Krishna, M. T.; Blomberg, A.; Biscione, G. L.; Kelly, F.; Sandstrom, T.; Frew, A.; Holgate, S. (1997b) Short-term
31             ozone exposure upregulates P-selectin in normal human airways. Am. J. Respir. Crit. Care Med. 155:
32             1798-1803.
33       Krishna, M. T.; Madden, J.; Teran, L.  M.; Biscione, G. L.; Lau, L. C. K.; Withers, N. J.; Sandstrom,  T.; Mudway, L;
34             Kelly, F. J.; Walls, A.; Frew, A. J.; Holgate, S. T. (1998) Effects of 0.2 ppm  ozone onbiomarkers of
3 5             inflammation in bronchoalveolar lavage fluid and bronchial mucosa of healthy subjects. Eur. Respir. J.
36             11: 1294-1300.
37       Kronenberg, R. S.; Drage, C. W.; Ponto, R. A.; Williams, L. E. (1973) The effect of age on the distribution of
3 8             ventilation and perfusion in the lung. Am. Rev. Respir. Dis. 108: 576-586.
39       Kulle, T. J.; Sauder, L. R.; Kerr, H. D.; Farrell, B. P.; Bermel, M. S.; Smith, D. M. (1982) Duration of pulmonary
40             function adaptation to ozone in humans. Am. Ind. Hyg. Assoc. J. 43: 832-837.
41       Kulle, T. J.; Sauder, L. R.; Hebel, J. R.; Chatham, M. D. (1985) Ozone response relationships in healthy
42             nonsmokers. Am. Rev. Respir. Dis. 132: 36-41.
43       Lang, D. S.; Torres, R. A.; Mticke, M.; Siegfried, W.; Magnussen, H. (1998) Interactions between human
44             bronchoepithelial cells and lung fibroblasts after ozone exposure in vitro. Toxicol. Lett. 96-97: 13-24.
45       Larsen, R. I.; McDonnell, W. F.; Horstman, D. H.; Folinsbee, L. J. (1991) An air quality data analysis system for
46             interrelating effects, standards, and needed source reductions: part 11. a lognormal model relating human lung
47             function decrease to O3 exposure. J. Air Waste Manage. Assoc. 41: 455-459.
48       Lategola, M. T.; Melton, C. E.; Higgins, E.  A. (1980) Effects of ozone on symptoms and cardiopulmonary function
49             in a flight attendant surrogate population. Aviat. Space Environ. Med. 51:  237-246.
50       Lategola, M. T.; Melton, C. E.; Higgins, E.  A. (1980) Pulmonary and symptom threshold effects of ozone in airline
51             passenger and cockpit crew surrogates. Aviat. Space Environ. Med. 51: 878-884.
52
53       Lefohn, A. S.; Foley, J. K. (1993) Establishing relevant ozone standards to protect vegetation and human health:
54             exposure/dose-response considerations. Air Waste 43: 106-112.
55       Linder, J.; Herren, D.; Monn, C.; Wanner, H.-U. (1988) Die Wirkung von Ozon auf die korperliche
56             Leistungsfa'higkeit [The effect of ozone on physical activity]. Schweiz Z. Sportmed. 36: 5-10.
         August 2005                                 AX6-151       DRAFT-DO NOT QUOTE OR CITE

-------
  1       Linn, W. S.; Buckley, R. D.; Spier, C. E.; Blessey, R. L.; Jones, M. P.; Fischer, D. A.; Hackney, J. D. (1978) Health
  2             effects of ozone exposure in asthmatics. Am. Rev. Respir. Dis. 117: 835-843.
  3       Linn, W. S.; Jones, M. P.; Bachmayer, E. A.; Spier, C. E.; Mazur, S. F.; Avol, E. L.; Hackney, J. D. (1980)
  4             Short-term respiratory effects of polluted ambient air: a laboratory study of volunteers in a high-oxidant
  5             community. Am. Rev. Respir. Dis. 121: 243-252.
  6       Linn, W. S.; Fischer, D. A.; Medway, D. A.; Anzar, U. T.; Spier, C. E.; Valencia, L. M.; Venet, T. G.; Hackney,
  7             J. D. (1982) Short-term respiratory effects of 0.12 ppm ozone exposure in volunteers with chronic obstructive
  8             pulmonary disease. Am. Rev. Respir. Dis. 125: 658-663.
  9       Linn, W. S.; Avol, E. L.; Hackney, J. D.  (1983) Effects of ambient oxidant pollutants on humans: a movable
10             environmental chamber study. In:  Lee, S. D.; Mustafa, M. G.; Mehlman, M. A., eds. International symposium
11             on the biomedical effects of ozone and related photochemical oxidants; March 1982; Pinehurst, NC.
12             Princeton, NJ: Princeton Scientific Publishers, Inc.; pp. 125-137. (Advances in modern environmental
13             toxicology: v. 5).
14       Linn, W. S.; Avol, E. L.; Shamoo, D. A.; Spier, C. E.; Valencia, L. M.; Venet, T. G.; Fischer, D. A.; Hackney, J. D.
15             (1986) A dose-response study of healthy, heavily exercising men exposed to ozone at concentrations near the
16             ambient air quality standard. Toxicol. Ind. Health 2: 99-112.
17       Linn, W. S.; Avol, E. L.; Shamoo, D. A.; Peng, R.-C.; Valencia, L. M.; Little, D. E.; Hackney, J. D. (1988) Repeated
18             laboratory ozone exposures of volunteer Los Angeles residents: an apparent seasonal variation in response.
19             Toxicol. Ind. Health 4: 505-520.
20       Linn, W. S.; Shamoo, D. A.; Anderson, K. R.; Peng, R.-C.; Avol, E. L.; Hackney, J. D. (1994) Effects of prolonged,
21             repeated exposure to ozone, sulfuric acid, and their combination in healthy and asthmatic volunteers. Am. J.
22             Respir. Crit. Care Med. 150: 431-440.
23       Linn, W. S.; Anderson, K. R.; Shamoo, D. A.; Edwards, S. A.; Webb, T. L.; Hackney, J. D.; Gong, H., Jr. (1995)
24             Controlled exposures of young asthmatics to mixed oxidant gases and acid aerosol. Am. J. Respir. Crit. Care
25             Med. 152: 885-891.
26       Linn, W. S.; Gong, H., Jr.; Shamoo, D. A.;  Anderson, K. R.; Avol, E. L. (1997) Chamber exposures of children to
27             mixed ozone, sulfur dioxide, and sulfuric acid. Arch. Environ. Health 52: 179-187.
28       Liu, L.; Leech, J. A.; Urch, R. B.;  Silverman, F. S. (1997) In vivo salicylate hyroxylation: a potential biomarker for
29             assessing acute ozone exposure and effects in humans. Am. J. Respir. Crit. Care Med.  156: 1405-1412.
30       Liu, L.; Leech, J. A.; Urch, R. B.; Poon, R.; Zimmerman, B.; Kubay, J. M.; Silverman, F. S. (1999) A comparison of
31             biomarkers of ozone exposure in human plasma, nasal lavage, and sputum. Inhalation Toxicol. 11: 657-674.
32       MacNee, W.;  Morrison, D.; Rahman, I.; Li, X. Y.; Donaldson, K. (1996) Cigarette smoke and ozone-induced
33             epithelial perturbation in vivo and in vitro: the  role  of glutathione. Chest 109(suppl.): 39S.
34       Magnussen, H.; Nowak, D.; Jorres, R. (1994) Experimentelle Schadstoffeffekte bei Gesundenund Asthmatikern
3 5             [Experimental effects of air pollutants in healthy and asthmatic subjects]. Pneumologie 48: 85-88.
36       Marthan, R.; Roux, E.; Savineau, J.-P. (1996) Human bronchial smooth muscle responsiveness after in vitro
37             exposure to oxidizing pollutants. Cell Biol. Toxicol. 12: 245-249.
38       Martin, T. R.; Castile, R. G.; Fredberg, J. J.; Wohl, M. E. B.; Mead, J.  (1987) Airway size is related to sex but not
39             lung size in normal adults. J. Appl. Physiol. 63: 2042-2047.
40       McBride, D. E.; Koenig, J. Q.; Luchtel, D. L.; Williams, P. V; Henderson, W. R., Jr. (1994) Inflammatory effects of
41             ozone in the upper airways  of subjects with asthma. Am. J. Respir.  Crit. Care Med. 149:  1192-1197.
42       McDonnell, W. F. (1996) Individual variability in human lung function responses to ozone exposure. Environ.
43             Toxicol. Pharmacol. 2: 171-175.
44       McDonnell, W. F.; Smith, M. V. (1994) Description of acute ozone response as a function of exposure rate and total
45             inhaled dose. J. Appl. Physiol. 76: 2776-2784.
46       McDonnell, W. F.; Horstman, D. H.; Hazucha, M. J.; Seal, E., Jr.; Haak, E. D.; Salaam, S. A.; House, D. E. (1983)
47             Pulmonary effects of ozone exposure during exercise: dose-response characteristics. J. Appl. Physiol.:
48             Respir. Environ. Exercise Physiol. 54: 1345-1352.
49       McDonnell, W. F., Ill; Horstman,  D. H.;  Abdul-Salaam, S.; House, D. E. (1985a) Reproducibility of individual
50             responses to ozone exposure. Am. Rev. Respir. Dis. 131: 36-40.
51       McDonnell, W. F., Ill; Chapman, R. S.; Leigh, M. W.; Strope, G. L.; Collier, A. M. (1985b) Respiratory responses
52             of vigorously exercising children to 0.12 ppm ozone exposure. Am. Rev. Respir. Dis. 132: 875-879.
53       McDonnell, W. F.; Horstman, D. H.; Abdul-Salaam, S.; Raggio, L. J.;  Green, J. A. (1987) The respiratory  responses
54             of subjects with allergic rhinitis to ozone exposure and their relationship to nonspecific airway reactivity.
55             Toxicol. Ind. Health 3: 507-517.
         August 2005                                  AX6-152      DRAFT-DO NOT QUOTE OR CITE

-------
  1       McDonnell, W. F.; Kehrl, H. R.; Abdul-Salaam, S.; Ives, P. I; Folinsbee, L. I; Devlin, R. B.; O'Neil, J. I;
  2             Horstman, D. H. (1991) Respiratory response of humans exposed to low levels of ozone for6.6 hours.
  3             Arch. Environ. Health 46: 145-150.
  4       McDonnell, W. F.; Muller, K. E.; Bromberg, P. A.; Shy, C. M. (1993) Predictors of individual differences in acute
  5             response to ozone exposure. Am. Rev. Respir. Dis. 147: 818-825.
  6       McDonnell, W. F.; Stewart, P. W.; Andreoni, S.; Seal, E., Jr.; Kehrl, H. R.; Horstman, D. H.; Folinsbee, L. J.; Smith,
  7             M. V. (1997) Prediction of ozone-induced FEVj changes: effects of concentration, duration, and ventilation.
  8             Am. J. Respir. Crit. Care Med. 156: 715-722.
  9       McDonnell, W. F.; Stewart, P. W.; Smith, M. V.; Pan, W. K.; Pan, J. (1999) Ozone-induced respiratory symptoms:
10             exposure-response models and association with lung function. Eur. Respir. J. 14: 845-853.
11       McGee, M. P.; Devlin, R.; Saluta, G.; Koren, H. (1990) Tissue factor and factor VII messenger RNAs in human
12             alveolar macrophages: effects of breathing ozone. Blood 75: 122-127.
13       McKinnon, K. P.; Madden, M. C.; Noah, T. L.; Devlin, R. B. (1993) In vitro ozone exposure increases release of
14             arachidonic acid products from a human bronchial epithelial cell line. Toxicol. Appl. Pharmacol.
15             118:215-223.
16       Melton, C. E. (1990) Airliner cabin ozone: an updated review. Washington, DC: Federal Aviation Administration,
17             Office of Aviation Medicine; report no. DOT/FAA/AM-89/13. Available from: NTIS, Springfield, VA;
18             AD-A219264.
19       Messineo, T. D.; Adams, W. C. (1990) Ozone  inhalation effects in females varying widely in lung size: comparison
20             with males. J. Appl. Physiol. 69: 96-103.
21       Michelson, P. H.; Dailey, L.; Devlin, R. B.; Peden, D.  B. (1999) Ozone effects on the immediate-phase response to
22             allergen in the nasal airways of allergic asthmatic subjects. Otolaryngol. Head Neck Surg.  120: 225-232.
23       Mohammed, S. P.; Higenbottam, T. W.; Adcock, J. J. (1993) Effects of aerosol-applied capsaicin, histamine and
24             prostaglandin-E2 on airway  sensory  receptors of anaesthetized cats. J. Physiol. Lond. 46: 951-966.
25       Molfino, N. A.; Wright, S. C.;  Katz, L; Tarlo, S.; Silverman, F.; McClean, P. A.; Szalai, J. P.; Raizenne, M.; Slutsky,
26             A. S.; Zamel, N. (1991) Effect of low concentrations of ozone on inhaled allergen responses in asthmatic
27             subjects. Lancet 338(8761):  199-203.
28       Molfino, N. A.; Slutsky, A. S.; Zamel, N. (1992) The effects of air pollution on allergic bronchial responsiveness.
29             Clin. Exp.  Allergy 22: 667-672.
30       Montuschi, P.;  Nightingale, J. A.; Kharitonov, S. A.; Barnes, P. J. (2002) Ozone-induced increase in exhaled
31             8-isoprostane in healthy subjects is resistant to inhaled budesonide. Free Radical Biol. Med. 33: 1403-1408.
32       Mudway, I. S.; Kelly, F. J. (2000) Ozone and the lung: a sensitive issue. Mol. Aspects. Med. 21: 1-48.
33       Mudway, I. S.; Blomberg, A.; Frew, A. J.; Holgate, S.  T.;  Sandstrom, T.; Kelly, F. J. (1999) Antioxidant
34             consumption and repletion kinetics in nasal lavage fluid following exposure of healthy human volunteers to
35             ozone. Eur. Respir. J. 13: 1429-1438.
36       Mudway, I. S.; Stenfors, N.; Blomberg, A.; Helleday, R.; Dunster, C.; Marklund, S. L.; Frew, A. J.; Sandstrom, T.;
3 7             Kelly, F. J. (2001) Differences in basal airway antioxidant concentrations are not predictive of individual
3 8             responsiveness to ozone: a comparison of healthy and mild asthmatic subjects. Free Radical Biol. Med.
39             31:962-974.
40       Nagda, N. L.; Fortmann, R. C.; Koontz, M. D.; Baker,  S. R.; Ginevan, M. E. (1989) Airliner cabin environment:
41             contaminant measurements,  health risks, and mitigation options. Washington, DC: U.S. Department of
42             Transportation, Office of the Secretary.  Available from: NTIS, Springfield, VA; PB91 -1593 84.
43       National Institutes of Health. (1997) Guidelines for the diagnosis and management of asthma: expert panel report 2.
44             Bethesda, MD: U.S. Department of Health and Human Services, National Heart, Lung, and Blood Institute;
45             publication no. 97-4051. Available: http://www.nhlbi.nih.gov/guidelines/asthma/asthgdln.pdf (11 April 2003).
46       National Institutes of Health. (2003) Expert panel report: guidelines for the diagnosis and management of asthma.
47             Update on selected topics 2002. Bethesda, MD: U.S. Department of Health and Human Services, National
48             Heart, Lung, and Blood Institute; NIH publication no. 02-5074. Available:
49             http://www.nhlbi.nih.gov/guidelines/asthma/asthmafullrpt.pdf [26 August, 2004].
50       Neas, L. M.; Dockery, D. W.; Koutrakis, P.; Tollerud,  D. J.; Speizer, F. E. (1995) The association of ambient air
51             pollution with twice daily peak expiratory flow  rate measurements  in children. Am. J. Epidemiol.
52             141: 111-122.
53       Newson, E. J.; Krishna, M. T.; Lau, L. C. K.; Howarth, P. H.; Holgate, S.  T.; Frew, A. J. (2000) Effects of
54             short-term exposure to 0.2 ppm ozone on biomarkers of inflammation in sputum, exhaled nitric oxide, and
55             lung function in subjects with mild atopic asthma. J. Occup. Environ. Med. 42: 270-277.
         August 2005                                  AX6-153      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Nichols, B. G.; Woods, J. S.; Luchtel, D. L.; Corral, I; Koenig, J. Q. (2001) Effects of ozone exposure on nuclear
  2             factor-KB activation and tumor necrosis factor-a expression in human nasal epithelial cells. Toxicol. Sci.
  3             60:356-362.
  4       Nightingale, J. A.; Rogers, D. F.; Barnes, P. J. (1999) Effect of inhaled ozone on exhaled nitric oxide, pulmonary
  5             function, and induced sputum in normal and asthmatic subjects. Thorax 54: 1061-1069.
  6       Nightingale, J. A.; Rogers, D. F.; Chung, K. F.; Barnes, P. J. (2000) No effect of inhaled budesonide on the response
  7             to inhaled ozone in normal subjects. Am. J. Respir. Crit. Care Med. 161: 479-486.
  8       Niinimaa, V.; Cole, P.; Mintz, S.; Shephard, R. J. (1980) The switching point from nasal to oronasal breathing.
  9             Respir. Physiol. 42:61-71.
10       Nikasinovic, L.; Momas, L; Seta, N. (2003) Nasal epithelial and inflammatory response to ozone exposure: a review
11             of laboratory-based studies published since 1985. J. Toxicol. Environ. Health B 6: 521-568.
12       Otto-Knapp, R.; Jurgovsky, K.; Schierhorn, K.; Kunkel, G. (2003) Antioxidative enzymes in human nasal mucosa
13             after exposure to ozone. Possible role of GSTM1 deficiency. Inflamm. Res. 52: 51-55.
14       Passannante, A. N.; Hazucha, M. J.; Bromberg, P. A.; Seal, E.; Folinsbee, L.; Koch, G. (1998) Nociceptive
15             mechanisms modulate ozone-induced human lung function decrements. J. Appl. Physiol. 85:  1863-1870.
16       Peden, D. B. (2001a) Air pollution in asthma: effect of pollutants on airway  inflammation. Ann. Allergy Asthma
17             Immunol. 87(suppl. 3): 12-17.
18       Peden, D. B. (200 Ib) Effect of pollutants in rhinitis. Curr. Allergy Asthma Rep. 1: 242-246.
19       Peden, D. B.; Setzer, R. W., Jr.; Devlin, R. B. (1995) Ozone exposure has both a priming effect on allergen-induced
20             responses and an intrinsic  inflammatory action in the nasal airways of perennially allergic asthmatics. Am. J.
21             Respir. Crit. Care Med. 151: 1336-1345.
22       Peden, D. B.; Boehlecke, B.; Horstman, D.; Devlin, R. (1997) Prolonged acute exposure to 0.16 ppm ozone induces
23             eosinophilic airway inflammation in asthmatic subjects with allergies. J. Allergy Clin. Immunol.
24             100:802-808.
25       Peters, E. A.; Hiltermann, J. T.; Stolk, J. (2001) Effect of apocynin on ozone-induced airway hyperresponsiveness to
26             methacholine in asthmatics. Free Radical Biol. Med. 31: 1442-1447.
27       Rayman, R. B. (2002) Cabin air quality: an overview. Aviat. Space Environ. Med. 73: 211-215.
28       Reisenauer, C. S.;  Koenig, J. Q.;  McManus, M. S.; Smith, M. S.; Kusic, G.; Pierson, W. E. (1988) Pulmonary
29             response to ozone exposures in healthy individuals aged 55 years or greater. JAPCA 38: 51-55.
30       Riediker, M.; Monn, C.; Koller, T.; Stahel, W. A.; Wuthrich, B. (2001) Air pollutants enhance rhinoconjunctivitis
31             symptoms in pollen-allergic individuals. Ann. Allergy Asthma Immunol. 87: 311-318.
32       Rigas, M. L.; Ben-Jebria, A.; Ultman, J. S. (1997) Longitudinal distribution of ozone absorption in the lung: effects
33             of nitrogen  dioxide, sulfur dioxide, and ozone exposures. Arch. Environ. Health 52:  173-178.
34       Rombout, P. J. A.; Dormans, J. A. M. A.; Marra, M.; VanEsch, G. J. (1986) Influence of exposure regimen on
3 5             nitrogen dioxide-induced morphological changes in the rat lung. Environ. Res. 41: 466-480.
36       Romieu, L; Sienra-Monge, J. J.; Ramirez-Aguilar, M.; Moreno-Macias, H.; Reyes-Ruiz, N. L; Estela del
37             Rio-Navarro, B.; Hernandez-Avila, M.; London, S. J. (2004) Genetic polymorphism of GSTM1 and
3 8             antioxidant supplementation influence  lung function in relation to ozone exposure in asthmatic children in
39             Mexico City. Thorax 59: 8-10.
40       Samet, J. M.; Hatch, G. E.; Horstman, D.; Steck-Scott, S.; Arab, L.; Bromberg, P. A.; Levine, M.; McDonnell,
41             W. F.; Devlin, R. B.  (2001) Effect of antioxidant supplementation on ozone-induced lung injury in human
42             subjects. Am. J. Respir. Crit. Care Med. 164: 819-825.
43       Santak, B.; Calzia, E.; Radermacher, P. (1998) Ventilation-perfusion relationships: clinical implications.
44             Acta Anaesthesiol. Scand. 42[suppl. 112]: 18-20.
45       Savin, W. M.; Adams, W. C. (1979) Effects of ozone inhalation on work performance and VO2max. J. Appl. Physiol.:
46             Respir. Environ. Exercise Physiol. 46:  309-314.
47       Scannell, C.; Chen, L.; Aris, R. M.; Tager, I.; Christian, D.; Ferrando, R.; Welch, B.; Kelly, T.; Balmes, J. R. (1996)
48             Greater ozone-induced inflammatory responses in subjects with asthma. Am.  J. Respir. Crit. Care Med.
49             154:24-29.
50       Schelegle, E. S.; Adams, W. C. (1986) Reduced exercise time in competitive simulations consequent to low level
51             ozone exposure. Med. Sci. Sports Exercise  18: 408-414.
52       Schelegle, E. S.; Adams, W. C.; Siefkin, A. D. (1987) Indomethacin pretreatment reduces ozone-induced pulmonary
53             function decrements in human subjects. Am. Rev.  Respir. Dis. 136: 1350-1354.
54       Schelegle, E. S.; Adams, W. C.; Giri, S. N.; Siefkin, A. D. (1989) Acute ozone exposure increases plasma
55             prostaglandin F2a  in ozone-sensitive human subjects. Am. Rev. Respir. Dis.  140: 211-216.
         August 2005                                  AX6-154      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Schelegle, E. S.; Siefkin, A. D.; McDonald, R. J. (1991) Time course of ozone-induced neutrophilia in normal
  2             humans. Am. Rev. Respir. Dis. 143: 1353-1358.
  3       Schelegle, E. S.; Carl, M. L.; Coleridge, H. M; Coleridge, J. C. G.; Green, J. F. (1993) Contribution of vagal
  4             afferents to respiratory reflexes evoked by acute inhalation of ozone in dogs. J. Appl. Physiol. 74: 2338-2344.
  5       Schelegle, E. S.; Eldridge, M. W.; Cross, C. E.; Walby, W. F.; Adams, W. C. (2001)  Differential effects of airway
  6             anesthesia on ozone-induced pulmonary responses in human subjects. Am. J. Respir. Crit. Care Med.
  7             163: 1121-1127.
  8       Schierhorn, K.; Zhang, M.; Matthias, C.; Kunkel, G. (1999) Influence of ozone and nitrogen dioxide on histamine
  9             and interleukin formation in a human nasal mucosa culture system. Am. J. Respir. Cell Mol. Biol.
10             20: 1013-1019.
11       Schierhorn, K.; Hanf, G.; Fischer, A.; Umland, B.; Olze, H.; Kunkel, G. (2002) Ozone-induced release of
12             neuropeptides from human nasal mucosa cells. Int. Arch. Allergy Immunol. 129: 145-151.
13       Schonfeld, B. R.; Adams, W. C.; Schelegle, E. S. (1989) Duration of enhanced responsiveness upon re-exposure to
14             ozone. Arch. Environ. Health 44: 229-236.
15       Schwartz, L. W.; Dungworth, D. L.; Mustafa, M. G.; Tarkington, B. K.; Tyler, W. S. (1976) Pulmonary responses of
16             rats to ambient levels of ozone: effects of 7-day intermittent or continuous exposure. Lab. Invest. 34: 565-578.
17       Seal, E., Jr.; McDonnell, W. F.; House, D. E.; Salaam, S. A.; Dewitt, P. J.; Butler, S.  O.; Green, J.; Raggio, L.
18             (1993) The pulmonary response of white and black adults to six concentrations of ozone. Am. Rev.  Respir.
19             Dis.  147: 804-810.
20       Seal, E., Jr.; McDonnell, W. F.; House, D. E. (1996) Effects of age, socioeconomic status,  and menstrual cycle on
21             pulmonary response to ozone. Arch. Environ. Health 51: 132-137.
22       Sekizawa, S. L; Tsubone, H. (1994) Nasal receptors responding to noxious chemical  irritants. Respir. Physiol.
23             96:37-48.
24       Seltzer, J.; Bigby, B. G.; Stulbarg, M.; Holtzman, M. J.; Nadel, J. A.; Ueki, I. F.; Leikauf, G. D.; Goetzl, E. J.;
25             Boushey, H. A. (1986) O3-induced change in bronchial reactivity to methacholine and airway inflammation in
26             humans. J. Appl. Physiol. 60: 1321-1326.
27       Solway, J.; Leff, A. R. (1991) Sensory neuropeptides and airway function. J. Appl. Physiol. 71: 2077-2087.
28       Spannhake, E. W.; Reddy, S. P. M.;  Jacoby, D. B.; Yu, X.-Y.; Saatian, B.; Tian, J. (2002) Synergismbetween
29             rhinovirus infection and oxidant pollutant exposure enhances airway epithelial cell cytokine production.
30             Environ. HealthPerspect. 110: 665-670.
31       Spit, B. J.; Bretschneider, F.; Hendriksen, E. G. J.; Kuper, C. F. (1993) Ultrastructure of free nerve endings in
32             respiratory and squamous epithelium on the rat nasal septum. Cell Tissue Res. 274:  329-335.
33       Steck-Scott, S.; Arab, L.; Craft, N. E.; Samet, J. M. (2004) Plasma and lung macrophage responsiveness to
34             carotenoid supplementation and ozone exposure in humans. Eur. J. Clin. Nutr. 1-9.
35       Stenfors, N.; Pourazar, J.; Blomberg, A.; Krishna, M. T.; Mudway, L; Helleday, R.; Kelly, F. J.; Frew, A. J.;
3 6             Sandstrom, T. (2002) Effect of ozone on bronchial mucosal inflammation in asthmatic and healthy subjects.
37             Respir. Med. 96: 352-358.
38       Sterk, P. J.; Fabbri, L. M.; Quanjer, P. H.; Cockcroft, D. W.; O'Byrne, P. M.; Anderson, S. D.; Juniper, E. F.;
39             Malo, J.-L. (1993) Airway responsiveness.  Standardized challenge testing with pharmacological, physical
40             and sensitizing stimuli in adults. Report Working Party Standardization of Lung Function Tests, European
41             Community for Steel and Coal. Eur. Respir. J. 6(suppl. 16): 53-83.
42       Stevens, W. H. M.; Adelroth, E.; Wattie, J.; Woolley, M. J.; Ellis, R.; Dahlback, M.;  O'Byrne, P. M. (1994) Effect of
43             inhaled budesonide on ozone-induced airway hyperresponsiveness and bronchoalveolar lavage cells in dogs.
44             J. Appl. Physiol.  77: 2578-2583.
45       Superko, H. R.; Adams, W. C.; Daly, P. W. (1984) Effects of ozone inhalation during exercise in selected patients
46             with heart disease. Am. J. Med. 77: 463-470.
47       Tepper, J. S.; Costa, D. L.; Lehmann, J. R.; Weber, M. F.; Hatch, G. E.  (1989) Unattenuated structural and
48             biochemical alterations in the rat lung during functional adaptation to ozone. Am. Rev. Respir. Dis.
49             140:493-501.
50       Tepper, J. S.; Wiester, M. J.; Weber, M. F.; Menache, M. G. (1990) Measurements of cardiopulmonary response in
51             awake rats during acute exposure to near-ambient concentrations of ozone. Fundam. Appl. Toxicol. 10: 7-15.
52       Torres, A.; Utell, M. J.;  Morow, P. E.; Voter, K. Z.; Whitin, J. C.; Cox, C.; Looney, R. J.; Speers, D. M.; Tsai, Y.;
53             Frampton, M. W. (1997) Airway inflammation in smokers and nonsmokers with varying responsiveness to
54             ozone. Am. J. Respir. Crit.  Care Med. 156: 728-736.
55       Trenga, C. A.; Koenig, J. Q.; Williams, P. V.  (2001) Dietary antioxidants and ozone-induced bronchial
56             hyperresponsiveness in adults with asthma. Arch. Environ. Health 56: 242-249.
         August 2005                                  AX6-155      DRAFT-DO NOT QUOTE OR CITE

-------
  1       Tsai, B. M; Wang, M; Pitcher, J. M.; Meldrum, K. K.; Meldram, D. R. (2004) Hypoxic pulmonary vasoconstriction
  2             and pulmonary artery tissue cytokine expression are mediated by protein kinase C. Am. J. Physiol. Lung Cell
  3             Mol. Physiol. 287: L1215-L1219.
  4       U.S. Environmental Protection Agency. (1986) Air quality criteria for ozone and other photochemical oxidants.
  5             Research Triangle Park, NC: Office of Health and Environmental Assessment, Environmental Criteria and
  6             Assessment Office; report nos. EPA-600/8-84-020aF-eF. 5v. Available from: NTIS, Springfield, VA;
  7             PB87-142949.
  8       U.S. Environmental Protection Agency. (1996) Air quality criteria for ozone and related photochemical oxidants.
  9             Research Triangle Park, NC: Office of Research and Development; report nos. EPA/600/AP-93/004aF-cF. 3v.
10             Available from: NTIS, Springfield, VA; PB96-185582, PB96-185590, and PB96-185608. Available:
11             www.epa.gov/ncea/ozone.htm.
12       Ultman, J. S.; Ben-Jebria, A.; Arnold, S. F. (2004) Uptake distribution of ozone in human lungs: intersubject
13             variability in physiologic response. Boston, MA: Health Effects Institute.
14       Utell, M.  J.; Frampton, M. W.; Morrow, P. E.; Cox, C.; Levy, P. C.; Speers, D. M.; Gibb, F. R. (1994) Oxidant and
15             acid aerosol exposure in healthy subjects and subjects with asthma. Part II: effects of sequential sulfuric acid
16             and ozone exposures on the pulmonary function of healthy subjects and subjects with asthma. Cambridge,
17             MA: Health Effects Institute; pp. 37-93; research report no. 70.
18       Vagaggini, B.; Carnevali, S.; Macchioni, P.; Taccola, M.; Fornai, E.; Bacci, E.; Bartoli, M. L.; Cianchetti, S.; Dente,
19             F.  L.; Di Franco, A.; Giannini, D.; Paggiaro. P. L. (1999) Airway inflammatory response to ozone in subjects
20             with different asthma severity. Eur. Respir. J. 13: 274-280.
21       Vagaggini, B.; Taccola, M.;  Conti, I.; Carnevali, S.; Cianchetti,  S.; Bartoli, M. L.; Bacci, E.; Dente, F. L.;
22             Di Franco, A.; Giannini, D.; Paggiaro, P. L. (2001) Budesonide reduces neutrophilic but not functional
23             airway response to ozone in mild asthmatics. Am. J. Respir. Crit. Care Med. 164: 2172-2176.
24       Vagaggini, B.; Taccola, M.;  Cianchetti, S.; Carnevali, S.; Bartoli, M. L.; Bacci, E.; Dente, F. L.; Di Franco, A.;
25             Giannini, D.; Paggiaro, P. L. (2002) Ozone exposure increases eosinophilic airway response induced by
26             previous allergen challenge. Am. J. Respir. Crit. Care Med. 166: 1073-1077.
27       Van Bree, L.; Rombout, P. J. A.; Rietjens, I. M. C. M.; Dormans, J. A. M. A.; Marra, M. (1989) Pathobiochemical
28             effects in rat lung related to episodic ozone exposure. In: Schneider, T.; Lee, S. D.; Wolters, G. J. R.;
29             Grant, L. D., eds. Atmospheric ozone research and its policy implications: proceedings of the 3rd US-Dutch
30             international symposium; May 1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier
31             Science Publishers; pp. 723-732. (Studies in environmental science 35).
32       Vender, R. L.; Horstman, D. H.; Mangione, S. (1994) Red blood cell antioxidants in human volunteers exposed to
33             ozone.  Toxicol. Ind. Health 10: 53-58.
34       Voter, K. Z.; Whitin, J. C.; Torres, A.; Morrow, P. E.; Cox, C.; Tsai, Y.; Utell, M. J.; Frampton, M.  W. (2001)
3 5             Ozone  exposure and the production of reactive oxygen species by bronchoalveolar cells in humans. Inhalation
36             Toxicol. 13:465-483.
37       Wang, G.; Umstead, T. M.; Phelps, D. S.; Al-Mondhiry, H.; Floras, J. (2002) The effect of ozone exposure on the
3 8             ability of human surfactant protein A variants to stimulate cytokine production. Environ. Health Perspect.
39             110:79-84.
40       Wayne, W. S.; Wehrle, P. F.; Carroll, R. E. (1967) Oxidant air pollution and athletic performance. JAMA J. Am.
41             Med. Assoc. 199:  151-154.
42       Weinmann, G. G.; Weidenbach-Gerbase, M.;  Foster, W. M.; Zacur, H.; Frank, R. (1995) Evidence for
43             ozone-induced small-airway dysfunction: lack of menstrual-cycle and gender effects. Am. J. Respir. Crit.
44             Care Med. 152: 988-996.
45       Weymer, A. R.; Gong, H., Jr.; Lyness, A.; Linn, W. S. (1994) Pre-exposure to ozone does not enhance or produce
46             exercise-induced asthma. Am. J. Respir. Crit. Care Med.  149: 1413-1419.
47       Yang, I. A.; Holz, O.; Jorres, R. A.; Magnussen, H.; Barton, S. J.; Rodriguez, S.; Cakebread, J.  A.; Holloway, J. W.;
48             Holgate, S. T. (2005) Association of tumor necrosis factor-a polymorphisms and ozone-induced change in
49             lung function. Am. J.  Respir. Crit. Care Med. 171: 171-176.
50       Yeadon, M.; Wilkinson, D.;  Darley-Usmar, V.; O'Leary, V. J.; Payne, A. N. (1992) Mechanisms contributing to
51             ozone-induced bronchial hyperreactivity in guinea-pigs. Pulm. Pharmacol. 5: 39-50.
52       Ying, R. L.; Gross,  K. B.; Terzo, T. S.; Eschenbacher, W. L. (1990) Indomethacin does not inhibit the
53             ozone-induced increase in bronchial responsiveness in human subjects. Am. Rev. Respir. Dis. 142: 817-821.
54       Yu, M.; Pinkerton, K. E.; Witschi, H. (2002) Short-term exposure to aged and diluted sidestream cigarette smoke
55             enhances ozone-induced lung injury inB6C3Fl mice. Toxicol. Sci. 65: 99-106.
56
         August 2005                                  AX6-156      DRAFT-DO NOT QUOTE OR CITE

-------
               CHAPTER 7 ANNEX
      EPIDEMIOLOGIC STUDIES OF HUMAN
  HEALTH EFFECTS ASSOCIATED WITH AMBIENT
               OZONE EXPOSURE
August 2005                AX7-1    DRAFT-DO NOT QUOTE OR CITE

-------
 AX7-1. Tables of Epidemiologic Studies of Human Health
     Effects Associated with Ambient Ozone Exposure
August 2005                   AX7-2    DRAFT-DO NOT QUOTE OR CITE

-------
OQ
 to
 o
 o
                    Table AX7-1.  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
           Reference, Study
                                                                   Copollutants
Location and Period     Outcomes and Methods     Mean O3 Levels    Considered
                                        Findings, Interpretation
                                                                                                                                             Effects
 X
        United States

        Mortimer et al. (2002)
        Eight urban areas in
        the U.S.:
        St. Louis, MO;
        Chicago, IL;
        Detroit, MI;
        Cleveland, OH;
        Washington, DC;
        Baltimore, MD;
        East Harlem, NY;
        Bronx, NY
        Jun-Aug 1993
                      National Cooperative Inner
                      City Asthma Study
                      (NCICAS) cohort.  This
                      panel study examined 846
                      asthmatic children aged
                      4-9 years for O3 exposure
                      effects on PEF and
                      morning symptoms using
                      linear mixed effects models
                      and GEE.
8-h avg O3
(10 a.m.-6p.m.):
48ppb
SD not provided.

Range of medians
across cities
shown in a figure:
Approximately
34 to 58 ppb.

<5%ofdays
exceeded 80 ppb.
PM10, NO2,      No associations were seen between single
SO2            or multiday O3 measures and any evening
               outcome measure.  The effects of O3 on
               morning outcomes increased over several
               days with the strongest associations seen
               for multiday lags.  Joint modeling of O3
               with NO2 or SO2 resulted in slightly
               reduced estimates for each pollutant.
               Closed cohort. Approximately 60% of
               NCICAS cohort returned diary,
               characterizations similar to entire cohort.
8-h avg O3 (per 15 ppb):

Percent change in morning PEF:
Lag 1-5:
All areas: -0.59% (-1.05,-0.13)
St. Louis: -0.86% (-2.10, 0.38)
Chicago:  -0.62% (-2.41,1.16)
Detroit: -0.75% (-2.36, 0.86)
Cleveland: -0.62% (-2.23, 0.99)
Washington, DC: -0.54% (-2.02,
0.93)
Baltimore: 0.24% (-0.95, 1.43)
EastHarlem: -0.73% (-1.63, 0.17)
Bronx: -0.69% (-1.54, 0.15)

Odds ratios:
Morning symptoms:
Lag 1-4:
All areas: 1.16(1.02,1.30)
St. Louis: 0.82(0.59,1.14)
Chicago:  1.09(0.69,1.72)
Detroit:  1.72(1.12,2.64)
Cleveland: 1.20(0.81,1.79)
Washington, DC: 1.11 (0.72, 1.72)
Baltimore: 1.19(0.89,1.60)
EastHarlem: 1.22(0.97,1.53)
Bronx: 1.23(0.98,1.54)

-------
OQ
 to
 o
 o
               Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
           Reference, Study                                                   Copollutants
          Location and Period     Outcomes and Methods     Mean O, Levels     Considered
                                                                   Findings, Interpretation
                                                                                     Effects
 X
United States (cont'd)

Mortimer et al. (2000)
Eight urban areas in the
U.S.:
St. Louis, MO;
Chicago, IL;
Detroit, MI;
Cleveland, OH;
Washington, DC;
Baltimore, MD;
East Harlem, NY;
Bronx, NY
Jun-Aug 1993
        Avoletal. (1998)
        Southern California
        communities
        Spring-summer 1994
                                A cohort of 846 asthmatic
                                children aged 4-9 years
                                examined for effects of
                                summer O3 exposure on
                                PEF and morning
                                symptoms. Two subgroups
                                were compared:  (l)low
                                birth weight or premature
                                and (2) normal birth weight
                                or full-term. Analysis using
                                GEE and linear mixed
                                effects models. Panel
                                study.
Three panels of children
(age 10-12 years):
(1) asthmatic (n= 53);
(2) wheezy (n = 54); and
(3) healthy (n = 103).
Examined for symptoms,
medication use, outdoor
time, physical activity, and
pulmonary function
measures in relation to O3
exposure, via logistic
regression and GLM.
                           8-h avg O3
                           (10 a.m.-6 p.m.):
                           48ppb
                           SD not provided.

                           See Mortimer
                           et al. (2002).
                  None
Stratified analysis
of low and high
24-h avg O3:

Fixed site O3:
Low:  <100 ppb
High: > 100 ppb

Personal O3:
Low:  < 15.6 ppb
High: > 32.4 ppb
None
               Low birth weight and premature
               asthmatic children had greater declines in
               PEF and higher incidence of morning
               symptoms than normal birth weight and
               full-term asthmatic children.
                                                                                   The three groups responded similarly.
                                                                                   Few pulmonary function or symptom
                                                                                   associations. Asthmatic children had
                                                                                   the most trouble breathing, the most
                                                                                   wheezing, and the most inhaler use on
                                                                                   high O3 day sin the spring. Ozone levels
                                                                                   were considered too low during the
                                                                                   period of the study.  Noncompliance
                                                                                   by subjects may have been a problem.
                                                                                   Other analysis methods may have been
                                                                                   more appropriate.
8-h avg O3 (per 15 ppb):

Percent change in morning PEF:
Low birth weight:
Lag 1-5:  -1.83%(-2.65,-1.01)
Normal birth weight:
Lag 1-5:  -0.30% (-0.79, 0.19)
Interaction term for birth weight,
p<0.05

Odds ratios:
Morning symptoms:
Low birth weight:
Lag 1-4:  1.42(1.10,1.82)
Normal birth weight:
Lag 1-4:  1.09(0.95,1.24)
Interaction term for birth weight,
p<0.05

Multiple endpoints analyzed. Few
consistent or statistically significant
responses to O3 exposure reported.

-------
OQ
 to
 o
 o
                Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
           Reference, Study                                                   Copollutants
          Location and Period      Outcomes and Methods     Mean O, Levels    Considered
                                                                    Findings, Interpretation
                                                                                      Effects
 X
United States (cont'd)

Gillilandetal. (2001)
12 Southern California
communities
Jan-Jun 1996
         Linn etal. (1996)
         Three towns in
         California: Rubidoux,
         Upland, Torrance
         Fall-spring 1992-1993
         and 1993-1994
Time-series study of 1,933
4th grade children (age
9-10 years) followed for
school absences. Each
absence classified as
illness-related or not.
Among former, classified
into respiratory
or gastrointestinal.
Respiratory absences
further classified into upper
or lower. Pollution
measured in central site in
each town. Analysis of
distributed lag effects
controlling for time, day of
week, and temperature in a
Poisson model.

Panel study of 269 school
children (age unspecified),
each followed for
morning/afternoon lung
function and symptoms for
one week in fall, winter,
and spring over 2 school
years. Personal exposure
monitoring in a subset.
Analyzed afternoon
symptoms versus same day
pollution and morning
symptoms versus 1-day
lag pollution.
                                                            8-h avg O3
                                                            (10 a.m.-6 p.m.):
                                                            Levels not
                                                            reported.

                                                            Figure depicts
                                                            mean range of
                                                            approximately
                                                            35 to 55 ppb
                                                            across the
                                                            12 communities.
                  PM10,N02
24-h avg O3:

Personal:
5 ppb
SD3

Central site:
23 ppb
SD12
PM2 5, N02
                Ozone strongly associated with illness-
                related and respiratory absences. PM10
                only associated with upper respiratory
                absences. Long distributed lag effects
                for O3 raise questions about adequacy
                of control for seasonal changes.
Central site O3 correlated with personal
exposures, r= 0.61. Ozone effects
observed on lung function but only
significant for FEV; in one analysis.
No effects on symptoms. Ozone effects
were not robust to NO2 or PM2 5. Power
may have been limited by short followup
within seasons (limiting  both person-days
and variability in exposures).
                                       8-h avg O3 (per 20 ppb):

                                       Percent change in school absences:

                                       All illness:
                                       62.9% (18.4, 124.1)
                                       Nonrespiratory illnesses:
                                       37.3% (5.7, 78.3)
                                       Respiratory illnesses:
                                       82.9% (3.9, 222.0)
                                       Upper respiratory:
                                       45.1% (21.3, 73.7)
                                       Lower respiratory with wet cough:
                                       173.9% (91.3, 292.3)
                                                                                                                            Change in lung function (per ppb):

                                                                                                                            FEV[ next morning:
                                                                                                                            -0.26 mL (SE 0.25), p = 0.30
                                                                                                                            FEV; afternoon:
                                                                                                                            -0.18 mL(SE 0.26), p = 0.49
                                                                                                                            FEV[ crossday difference:
                                                                                                                            -0.58 mL(SE 0.23), p = 0.01

                                                                                                                            FVC next morning:
                                                                                                                            -0.21 mL (SE 0.22),p = 0.34
                                                                                                                            FVC afternoon:
                                                                                                                            -0.20 mL(SE 0.29), p = 0.48
                                                                                                                            FVC crossday difference:
                                                                                                                            -0.25 mL (SE 0.25), p = 0.32

-------
               Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
uw
to
O
O
Reference, Study
Location and Period

Copollutants
Outcomes and Methods Mean O3 Levels Considered

Findings, Interpretation

Effects

        United States (cont'd)
X
        Ostroetal. (2001)
        Central Los Angeles
        and Pasadena, CA
        Aug-Octl993
        Delfino et al. (2003)
        Los Angeles, CA
        Nov 1999-Jan 2000
Panel study of 138 African-
American children aged 8-
13 years with doctor
diagnosed asthma requiring
medication in past year
followed for daily
respiratory symptoms and
medication use. Lags of
0 to 3 days examined.
A panel study of
22 Hispanic children with
asthma aged 10-16 years.
Filled out symptom diaries
in relation to pollutant
levels. Analysis using
l-hmaxO3:

Los Angeles:
59.5 ppb
SD31.4

Pasadena:
95.8 ppb
SD49.0
PM10,N02,
pollen, mold
l-hmaxO3:
25.4 ppb
SD9.6
N02, S02,
CO, volatile
organic
compounds,
PM,n
Correlation between PM10 and O3 was
r = 0.35. Significant O3 effect seen for
extra medication use (above normal use).
No O3 effect on symptoms in expected
direction observed. Inverse association
seen for cough. PM10 effects  seen at a lag
of 3 days.  Time factors not explicitly
controlled in analysis; may have led to
confounding of O3 effects.
Support the view that air toxics in
the pollutant mix from traffic may
have adverse effects on asthma in
children.
1-h max O3 (per 40 ppb):

Odds ratios:

Extra medication use:
Lagl:  1.15(1.12,1.19)

Respiratory symptoms:
Shortness of breath:
Lag 3:  1.01 (0.92,1.10)
Wheeze:
Lag 3:  0.94(0.88,1.00)
Cough:
Lag 3:  0.93(0.87,0.99)

1-h max O3 (per 14.0 ppb):

Odds ratio:
Symptoms interfering with
daily activities:
LagO:  1.99(1.06,3.72)
H
6
0
2;
o
H
0
0
H
W
0
O
HH
H
W
u\e,ii moaei.
Delfino et al. ( 1 997a) Panel study of 22
Alpine, CA asthmatics aged 9-46 years
May-Augl 994 followed for respiratory
symptoms, moming-
aftemoon PEF, and (32
agonist inhaler use.
Personal O3 measured for
12 hours/day using passive
monitors. GLM mixed
model.


Ambient: PM10, pollen, No O3 effects observed. No quantitative results for O3.
12-h avg O3 fungi
(8 a.m.-8 p.m.):
64 ppb
SD17
Personal:
12-h avg O3:
(8 a.m.-8 p.m.)
18 ppb
SD14


-------
OQ
 to
 o
 o
               Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
           Reference, Study                                                   Copollutants
          Location and Period     Outcomes and Methods     Mean O, Levels     Considered
                                                                                           Findings, Interpretation
                                                                                     Effects
 X
United States (cont'd)

Delfinoetal. (1998a)
Alpine, CA
Aug-Octl995
Delfino et al. (2004)
Alpine, CA
Aug-Oct 1999, Apr-Jun
2000
                        A panel of 24 asthmatics
                        aged 9-17 years followed
                        for daily symptoms.
                        Analysis using GEE model.
                        Panel study of 19 asthmatic
                        children (age 9- 1 7 years)
                        followed daily for 2 weeks
                        to determine relationship
                        between air pollutants,
                        namely PM, and FEVj.
                        Linear mixed model used
                        for analysis.
                                                           l-hmaxO3:
                                                           90 ppb
                                                           SD 18
8-h max O3:
62 . 8 ppb
SD 15.1
IQR 22.0
                  PM,f
PM2 5, PM10,
NO2
Asthma symptoms were significantly
associated with both ambient O3 and
PM10 in single-pollutant models. Ozone
effects generally robust to PM10. Current
day O3 effects strongest in asthmatics not
on anti-inflammatory medication. Effects
of O3 and PM10 were largely independent.
The largest effects for PM10 were seen for
a 5-day distributed lag. For O3 effects,
there were no lag day effects; current day
results showed the greatest effect.

Significant declines in FEV; associated
with various PM indices (personal, indoor
home, etc.), but not ambient O3 levels.
                                                       1-h max O3 (per 58 ppb):

                                                       Odds ratios:
                                                       O3 only model:
                                                       LagO:  1.54(1.02,2.33)
                                                       O3 with PM10 model:
                                                       LagO:  1.46(0.93,2.29)
                                                                                                                                    No quantitative results for O3.
         Delfinoetal. (1996)
         San Diego, CA
         Sep-0ctl993
                        Panel study of 12 well-      Ambient:
                        characterized moderate      1-h max O3:
                        asthmatics aged 9-16 years   68 ppb
                        (7 males, 5 females)         SD 30
                        followed over 6 weeks
                        for medication use and       Ambient:
                        respiratory symptoms.       12-h avg O3:
                        Allergy measured at         43 ppb
                        baseline with skin prick      SD 17
                        tests. Personal O3
                        measured with passive       Personal:
                        badge. Analysis with        12-h avg O3:
                        GLM mixed model.          11.6 ppb
                                                  SD11.2
                  PM2 5, SO42 ,     No effect of ambient O3 on symptom
                  H+, HNO3,       score. Personal O3 significant for
                  pollen, fungal    symptoms, but effect disappeared when
                  spores           confounding day of week effect was
                                  controlled with weekend dummy
                                  variable.  (32 inhaler used among
                                  7 subjects was significantly related to
                                  personal O3. Results of this small study
                                  suggest the value of personal exposure
                                  data in providing more accurate estimates
                                  of exposures. However, nearly  50% of
                                  personal O3 measurements were below
                                  limits of detection, diminishing  value of
                                  these data. Pollen and fine particulate
                                  (low levels) were not associated with
                                  any of the outcomes.
                                                                                                                           Change in (32-agonist inhaler use
                                                                                                                           (per ppb personal O3):
                                                                                                                           0.0152 puffs/day (SE 0.0075),
                                                                                                                           p = 0.04

-------
OQ
 to
 o
 o
               Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
           Reference, Study                                                  Copollutants
          Location and Period     Outcomes and Methods    Mean O, Levels    Considered
                                                                    Findings, Interpretation
                                                                                     Effects
         United States (cont'd)
 X
 oo
         Chen et al. (2000)
         Washoe County, NV
         1996-1998
Time-series study of school
absenteeism examined
among 27,793 students
(kindergarten to 6th grade)
from 57 elementary
schools.
First-order autoregression
models used to assess
relationship between O3
and school absenteeism
after adjusting for weather,
day of week, month,
holidays, and time trends.
Ozone levels from the
current day, and
cumulative lags of
1-14 days, 1-21 days, and
1-28 days examined.
l-hmaxO3:        PM10, CO        Multipollutant models were examined.
37.45 ppb                         Ozone concentrations in the preceding
SD 13.37                          14 days were significantly associated
                                  with school absenteeism for students
                                  in grades 1 through 6, but not those in
                                  kindergarten. Both PM10 and CO
                                  concentrations on the concurrent day
                                  were associated with school absenteeism,
                                  but the estimate for PM10 was a negative
                                  value.
1-h max O3 (per 50 ppb):

Total absence rate:
O3 with PM10 and CO model:
Lag 1-14:  3.79% (1.04, 6.55)
        Newhouse et al. (2004)
        Tulsa, OK
        Sep-Oct2000
Panel study of 24 subjects
aged 9-64 years with
physician diagnosis of
asthma.  Performed PEF
twice daily (morning and
afternoon), and reported
daily respiratory symptoms
and medication use.
Forward stepwise multiple
regression models and
Pearson correlation
analyses.
24-havgO3:       PM25, CO,       Among ambient air pollutants, O3 seemed
30 ppb            SO2, pollen,      to be most significant factor. Morning
Range 10-70       fungal spores    PEF values significantly associated with
                                  average and maximum O3 levels on the
                                  previous day. Individual symptoms,
                                  including wheezing, headache, and
                                  fatigue, also significantly related to
                                  average and maximum daily O3.  Multiple
                                  regression analyses produced complex
                                  models with different predictor variables
                                  for each symptom.
Pearson correlation coefficient:

Morning PEF:
Mean O3 levels:
Lagl:  -0.274, p< 0.05
Maximum O3 levels:
Lagl:  -0.289, p< 0.05

-------
               Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
OQ
 to
 o
 o
           Reference, Study
                                                                     Copollutants
                                                                                                    Findings, Interpretation
                                                                                                                                        Effects
 X
United States (cont'd)

Ross et al. (2002)
East Moline, IL and
nearby communities
May-Octl994
Neasetal. (1995)
Uniontown, PA
Summer 1990
                                Panel study of 59
                                asthmatics aged 5-49 years
                                recruited.  19 lost to
                                follow-up, yielding
                                study population of 40.
                                Assessment of PEF
                                and respiratory symptoms.
                                Analytical methods unclear
                                in terms of control for
                                time factors.
Panel study of 83 4th and
5th grade children reported
twice daily PEF and the
presence of cold, cough, or
wheeze. Relationship to
pollutants was analyzed by
an autoregressive linear
regression model/GEE.
The number of hours each
child spent outdoors during
the preceding 12-h period
was evaluated.
                           41.5 ppb
                           SD 14.2
                           IQR20
PM10, SO2,
NO2, pollen,
fungi
                                                           12-h avg O3:
                                                           Daytime
                                                           (8 a.m.-8 p.m.):
                                                           50.0 ppb

                                                           Overnight
                                                           (8 p.m.-8 a.m.):
                                                           24.5 ppb
SO2, PM10, H+
                Saw significant associations between
                O3 and both PEF declines and
                symptom increases.  Most but not all
                effects remained after controlling for
                temperature, pollen and fungi.  Ozone
                effect on morning PEF disappeared after
                adjusting for temperature.  No PM10
                effects observed.
                Evening cough was associated with O3
                levels weighted by hours spent outdoors
                during the prior 12 hours. A decrease in
                PEF was associated with O3 levels
                weighted by hours spent outdoors. When
                particle-strong acidity was added to the
                model, the decrement was decreased and
                no longer significant.
8-h max O3 (per 20 ppb):

Change in PEF (L/min):
Morning:
Lag 0-1:  -2.29 (-4.26,-0.33)
Afternoon:
LagO:  -2.58 (-4.26,-0.89)

Symptom score (on scale of 0-3):
Morning:
Lag 1-3:  0.08(0.03,0.13)
Afternoon:
Lag 1-3:  0.08(0.04,0.12)

12-h avg O3 (per 30 ppb increment
weighted by proportion of time
spent outdoors during prior 12
hours):

Evening PEF:
-2.79 L/min (-6.7,-1.1)

Odds ratio:
Evening cough:
2.20(1.02,4.75)
        Neasetal. (1999)
        Philadelphia, PA
        Jul-Sep 1993
                        Panel study of 156 children
                        aged 6-1 1 years at two
                        summer camps followed
                        for twice-daily PEF.
                        Analysis using mixed
                        effects models adjusting for
                        autocorrelated errors.
                           Daytime
                           12-h avg O3
                           (9 a.m-9 p.m.):

                           SW camp:
                           57.5 ppb
                           IQR19.8

                           NE  camp:
                           55.9 ppb
                           IQR21.9
H+, SO/-,
PM2.5, PM10,
PMln.,<
                Some O3 effects detected as well as PM
                effects. Similar O3-related decrements
                observed in both morning and afternoon
                PEF. Ozone effects not robust to SO42~ in
                two-pollutant models, whereas SO42~
                effects relatively robust to O3.
12-h avg O3 (per 20 ppb):

Morning and evening PEF:

O3 only models:
LagO:
-1.38 L/min (-2.81, 0.04)
Lag 1-5:
-2.58 L/min (-4.81,-0.35)

O3 with SO42~ model:
Lag not specified:
-0.04 L/min

-------
OQ
 to
 o
 o
               Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
           Reference, Study
          Location and Period
Outcomes and Methods
Mean O, Levels
Copollutants
 Considered
                                                                    Findings, Interpretation
             Effects
 X
         United States (cont'd)

         Gent et al. (2003)
         Southern New England
         Apr-Sep2001
         Korricketal. (1998)
         Mount Washington,
         NH
         Summers 1991, 1992
Panel study of 271 children
(age <12 years) with
active, doctor-diagnosed
asthma followed over
183 days for respiratory
symptoms.  For analysis,
cohort split into two
groups:  130 who used
maintenance medication
during follow-up and
141 who did not, on
assumption that medication
users had more severe
asthma.  Logistic
regression analyses
performed.

Cross-sectional study
evaluating the acute effects
of ambient O3 on
pulmonary function of
exercising adults.
530 hikers (age 15-64
years) monitored before
and after their hike.
Analysis using a general
linear regression model.
                          58.6 ppb
                          SD 19.0
                                                           51.3 ppb
                                                           SD 15.5
                                            PM2.5
                          O3 level per hour
                          of hiking:
                          40 ppb
                          SD12
                          Range 21-74
                 PM25, smoke,
                 acidity
                                 Correlation between 1-h max O3 and
                                 daily PM25 was 0.77 during this warm-
                                 season study.  Large numbers of
                                 statistical tests performed.  Significant
                                 associations between symptoms and O3
                                 seen only in medication users, a subgroup
                                 considered to be more sensitive. PM2 5
                                 significant for some symptoms, but not
                                 in two-pollutant models. Ozone effects
                                 generally robust to PM2 5.  Study
                                 limitations include limited control for
                                 meteorological factors and the post-hoc
                                 nature of the population stratification by
                                 medication use.
               With prolonged outdoor exercise low-
               level exposures to O3 were associated
               with significant effects on pulmonary
               function.  Hikers with asthma had a
               four-fold greater responsiveness to
               exposure to O3.
                                                      1-h max O3 (per 50 ppb):

                                                      Odds ratios:
                                                      Regular medication users (n = 130):

                                                      Chest tightness:
                                                      O3 only model:
                                                      Lagl:  1.26(1.00,1.48)
                                                      O3 with PM2 5 model:
                                                      Lagl:  1.42(1.14,1.78)

                                                      Shortness of breath:
                                                      O3 only model:
                                                      Lagl:  1.22(1.02,1.45)
Percent change in lung function
(per 50 ppb O3):

FEVI: -2.6% (-4.7,-0.4)
FVC:  -2.2% (-3.5,-0.8)

-------
OQ
 to
 o
 o
               Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
           Reference, Study                                                  Copollutants
         Location and Period     Outcomes and Methods     Mean O, Levels    Considered
                                                                  Findings, Interpretation
                                                                                   Effects
        United States (cont'd)
 X
        Thurstonetal. (1997)
        Connecticut River
        Valley, CT
        June 1991, 1992,  1993
        Naeheretal. (1999)
        Vinton, VA
        Summers 1995, 1996
Panel study of children
(age 7-13 years) with
moderate-to-severe asthma
followed for medication
use, lung function, and
medical symptoms at a
summer asthma camp for
one week in 1991 (n = 52),
1992 (n= 58), and 1993
(n = 56). Analysis was
conducted using both
Poisson modeling and
GLM.

Panel study evaluated the
relationship between O3
and daily change in PEF
studied in a sample of
473 nonsmoking women
aged 19-43 years who
recently delivered babies.
PEF performed twice daily
for a two-week period.
Mixed linear random
coefficient model.
l-hmaxO3:

1991:  114.0ppb
1992:  52.2 ppb
1993:  84.6 ppb

1991-1993:  83.6
ppb
 +, S042-
Ozone was most consistently associated
with acute asthma exacerbation, chest
symptoms, and lung function decrements.
Pollen was poorly associated with any
adverse effect. Consistent results were
obtained between the aggregate and
individual analyses.
53.69 ppb
Range
17.00-87.63

24-h avg O3:
34.87 ppb
Range 8.74-56.63
PM25,PM10,
SO/-, H+
Ozone was the only exposure related to
evening PEF with 5-day cumulative lag
exposure showing the greatest effect.
1-h max O3 (per 83.6 ppb):

Relative risks:
P2-agonist use: 1.46, p< 0.05
Chest symptoms:  1.50, p< 0.05

Change in PEF (per ppb):
-0.096 L/min,p< 0.05
24-h avg O3 (per 30 ppb):

Evening PEF:
Lag 1-5:
-7.65 L/min(-13.0,-2.25)

-------
OQ
 to
 o
 o
               Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
           Reference, Study                                                  Copollutants
          Location and Period     Outcomes and Methods    Mean O, Levels    Considered
                                                                                           Findings, Interpretation
                                                                                                                Effects
 X
 to
         Canada

         Braueretal. (1996)
         Fraser Valley, British
         Columbia, Canada
         Jun-Aug 1993
Brauer and Brook
(1997)
Fraser Valley, British
Columbia, Canada
Jun-Aug 1993
Panel study of 58 berry
pickers aged 10-69 years
had lung function measured
before and after a series
of outdoor work shifts
(average duration
= 11 hours) over 59 days.
Analysis using pooled
regression with subject-
specific intercepts, with
and without temperature
control.

Additional analysis of
Braueretal., 1996 with
personal exposure
presented for three groups,
stratified by time spent
outdoors.

Group 1:  25 individuals
who spent most of the day
indoors.

Group 2:  25 individuals
who spent much of the day
indoors, but still spent
several daylight hours
outdoors.

Group 3:  15 individuals
who spent the entire work
day outdoors.
                  PM2 5, SO42~,     End shift FEV; and FVC significantly
40.3 ppb          NO3~, NH4+,     diminished in relation to O3 levels.
SD 15.2           H+              PM25 also related to lung function
                                  declines, but O3 remained significant in
Work shift O3:                      two-pollutant models. Next morning
26.0 ppb                          lung function remained diminished
SD 11.8                           following high O3 days. Ozone effects
                                  still evident at or below 40 ppb. There
                                  was an overall decline of lung function of
                                  roughly 10% over course of study,
                                  suggesting subchronic effect.  Levels of
                                  other pollutants low during study.
l-hmaxO3:

Ambient:
40 ppb
SD15
Range 13-84
PM25, SO42~,     Group 1: 9.0% sampling time (24-h)
NO3~, NH4+,     outdoors. Personal to ambient O3 ratio
H+              was 0.28.

                Group 2: 25.8% sampling time (24-h)
                outdoors. Personal to ambient O3 ratio
                was 0.48.

                Group 3: 100% sampling time
                (11-h work shift) outdoors. Personal
                to ambient O3 ratio was 0.96.

                One of the first direct demonstrations that
                magnitude of personal exposure to O3 is
                related to amount of time spent outdoors.
                Further showed that, on average, outdoor
                fixed O3 monitors were representative of
                day-to-day changes in O3 exposure
                experienced by the study population.
                                                                                                                           Change in lung function (per ppb
                                                                                                                           1-h max O3):

                                                                                                                           Endshift lung function:
                                                                                                                           FEVp -3.8mL(SE0.4)
                                                                                                                           FVC:  -5.4mL(SE0.6)

                                                                                                                           Next morning function:
                                                                                                                           FEVp -4.5 mL (SE 0.6)
                                                                                                                           FVC:  -5.2mL(SE0.7)
Same outcomes as reported in
Braueretal., 1996.

-------
OQ
 to
 o
 o
               Table AX7-1  (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
           Reference, Study
          Location and Period
                         Outcomes and Methods
                           Mean O, Levels
                   Copollutants
                   Considered
                      Findings, Interpretation
                                                   Effects
 X
         Europe

         Scarlett et al. (1996)
         Surrey, England
         Jun-Jul 1994
Ward et al. (2002)
Birmingham and
Sandwell, England
Jan-Mar 1997
May-Jul 1997
                       Panel study examined
                       154 children aged 7 years
                       in a primary school next to
                       a major motorway for O3
                       exposure effects on PEF0 75,
                       FVC, and FEV; using
                       autoregression for percent
                       change in function.
A panel study of
162 children (age 9 years at
time of enrollment in
Sept 1996). 39 of
162 children (24%)
reported wheezing in the
past 12 months. Examined
association of ambient acid
species with PEF and
symptoms.  Single-day lags
of 0 to 3 days and a 7-day
cumulative lag were
investigated. Linear
regression used for PEF
and logistic regression used
for symptoms.
                           8-h max O3:
                           50.7 ppb
                           SD 24.48
                  PM10,N02,
                  pollen
24 h-avg O3:

Winter:
Median 13.0 ppb
Range 2-33

Summer:
Median 22.0 ppb
Range 10-41
PM10,PM25,
SO2, H+, CL,
HC1, HN03,
NH3, NH4+,
NO3- SO42~
              No significant association was seen
              between pulmonary function measures
              and O3 levels. No pollen effects.
Pollutants levels were generally low,
even in the summer.  Significant
associations were noted between
respiratory health outcomes and air
pollutants, but no consistent patterns were
identified.  The association between O3
and PEF was generally negative in the
summer and positive in the winter. More
associations between O3 and symptoms
were observed in the winter. Ozone was
associated with a significant increase in
cough, shortness of breath, and wheeze
during the winter. Results did not
indicate that children with atopy or a
history of recent wheezing were more
susceptible to short-term effects of air
pollutants.
                                       Change in lung function (per ppb O3
                                       weighted by inverse of variance):
Lagl:  0.01 mL (-0.12, 0.13)
FVC:
Lagl:  0.07 mL (-0.09, 0.23)
FEV075/FVC:
Lagl:  -0.1%(-5.1,4.8)

24-h avg O3 (per 21.5 ppb for
winter; per 10.2 ppb for summer):

Change in PEF (L/min):
Morning (lag 0-6):
Winter: 17.53(6.56,28.52)
Summer:  -5.66 (-11.21,-0.09)
Afternoon (lag  0-6):
Winter: 0.28 (-9.03, 9.79)
Summer:  -0.14 (-5.34, 5.04)

Odds ratios:
Symptoms:
Cough (lag 0-6):
Winter: 0.88(0.42,1.81)
Summer:  0.95(0.76,1.19)
Shortness of breath (lag 0-6):
Winter: 2.79(1.56,4.95)
Summer:  1.35(0.95,1.94)
Wheeze (lag 0-6):
Winter: 1.59(0.77,3.31)
Summer:  0.88(1.38,0.57)

-------
OQ
to
o
o
           Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
X


£
Reference, Study
Location and Period
Europe (cont'd)
Taggartetal. (1996)
Runcom and Widnes
in NW England
M-Sep 1993










Desqueyroux et al.
(2002a)
Paris, France
Nov 1995-Nov 1996






Desqueyroux et al.
(2002b)
Paris, France
Oct 1995-Nov 1996






Outcomes and Methods

Panel study investigated
the relationship of
asthmatic bronchial
hyperresponsiveness and
pulmonary function with
summertime ambient air
pollution among 38 adult
nonsmoking asthmatics
(age 18-70 years) using
log-linear models.
Analysis limited
to investigation of within
subject variance of the
dependent variables.
Panel study of 60 severe
asthmatics (mean age 55
years) were monitored by
their physicians for asthma
attacks. Asthma attacks
were based on medical data
collected by a pulmonary
physician at time of clinical
examination. Analysis
using GEE.
Panel study of 39 adult
patients with severe COPD
(mean age 67 years)
followed over 14 months
by physicians
for exacerbations.
Logistic regression with
GEE, examining exposure
lags of 0 to 5 days.

Copollutants
Mean O3 Levels Considered

l-havgO3: SO2,NO2,
Maximum smoke
61 ng/m3

24-h avg O3:
Maximum
24.5 ng/m3







8-havg03 PM10
(10 a.m.-6 p.m.):

Summer:
41 ng/m3
SD18

Winter:
1 1 ng/m3
SD10
8-h avg O3 PM10, SO2,
(10a.m.-6p.m.): NO2

Summer:
41 |ig/m3
SD18

Winter:
11 |ig/m3
SD10
Findings, Interpretation

No association found for O3. Changes
in bronchial hyperresponsiveness were
found to correlate significantly with
change in the levels of 24-h mean SO2,
NO2, and smoke.









Significant associations between PM10,
O3, and incident asthma attacks were
found. Low O3 levels raise plausibility
concerns.






50 COPD exacerbations observed over
follow-up period. 1-, 2-, and 3-day lag
O3 significantly related to exacerbations.
No other pollutants significant. Low O3
levels raise plausibility and confounding
concerns.




Effects

24-h avg O3 (per 10 |ig/m3):

Percent change in bronchial
hyperresponsiveness:
Lagl: 0.3% (-16.6, 20.6)
Lag 2: 2.6% (-22.1, 34.9)








8-havgO3(perlO|ig/m3):

Odds ratio:
Lag 2: 1.20(1.03,1.41)






8-h avg O3 (per 10 ng/m3):

Odds ratio:
Lagl: 1.56(1.05,2.32)

Effects appeared larger among
smokers and those with worse gas
exchange lung function.



-------
OQ
 to
 o
 o
                Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
           Reference, Study
          Location and Period
                         Outcomes and Methods
                            Mean O3 Levels
                   Copollutants
                    Considered
                                         Findings, Interpretation
             Effects
 X
         Europe (cont'd)

         Just et al. (2002)
         Paris, France
         Apr- Jim  1996
Lagerkvist et al. (2004)
Brussels, Belgium
May 2002
                        Panel study of 82
                        medically diagnosed
                        asthmatic children (mean
                        age 10.9 years) followed
                        for O3 exposure and PEF,
                        asthmatic attacks, cough,
                        supplementary use of
                        P2-agonists, and symptoms
                        of airway irritation.
                        Analysis by GEE.
Panel study of 57 children
(mean age 10.8 years)
stratified by swimming
pool attendance.
Pulmonary function test
performed and Clara cell
protein levels measured in
blood before and after light
exercise outdoors for two
hours. Analysis using
student's t-test and Pearson
correlation test.  For dose
calculations, O3 levels
indoors assumed to be 50%
of the mean outdoor O3
concentration.
24-havgO3:
58.9 |ig/m3
SD 24.5
Range 10.0-121.0
                  PM10,NO2
Daytime outdoor
03:
Range 77- 116
                                                           Exposure dose:
                                                           Range 352-914
                                                           Hg/m3-hour
                                                                              None
                                                             In asthmatic children, O3 exposure was
                                                             related to the occurrence of asthma
                                                             attacks and additional bronchodilator use.
                                                             O3 was the only pollutant associated with
                                                             changes in lung function, as shown by an
                                                             increase in PEF variability and decrease
                                                             in PEF.
                                  Ozone levels did not have any adverse
                                  effect on FEV[ after 2 hours of outdoor
                                  exercise. In addition,  no significant
                                  differences were observed between Clara
                                  cell protein levels before and after
                                  exercise. A marginally significant
                                  positive correlation between ambient O3
                                  dose and Clara cell protein levels
                                  observed among the nonswimmers,
                                  indicating increased antioxidant activity
                                  following O3 exposure in this group.  The   r = -0.08, p
                                  lack of a clear relationship between Clara
                                  cell protein levels and O3 dose may be
                                  attributable to the short period of time
                                  between measurements and diurnal
                                  variability of the protein levels.
24-havgO3(perlO|ig/m3):

Percent change in daily PEF
variability:
Lag 0-2:  2.6%, p = 0.05

Odds ratio:
Supplementary use of p2-agonist
on days on which no steroids
were used:
LagO:  1.41 (1.05,1.89)

Pearson correlation:
O3 exposure dose and Clara cell
protein levels in serum:

All subjects (n = 54):
r = 0.17, p = 0.21
Nonswimmers (n = 33):
r=0.34,p = 0.06
Swimmers (n = 21):
r= -008 n = 0.74

-------
OQ
 to
 o
 o
               Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
           Reference, Study                                                   Copollutants
          Location and Period     Outcomes and Methods     Mean O, Levels     Considered
                                                                    Findings, Interpretation
                                                                                      Effects
 X
Europe (cont'd)

Schindler et al. (2001)
Eight communities
of Switzerland
May-Sep 1991
         Frischer et al. (1993)
         Umkirch, Germany
         May-Octl991
A random sample of 3,912
adult never-smokers, aged
18 to 60 years, examined
for short-term O3-related
changes in lung function.
Natural logarithms of FVC,
FEY^ and FEF25.75 were
regressed against the
individual predictor
variables and O3. Spline
functions were used to
control potential trends.
Sensitivity analyses for
grass and pollens, and NO2
and TSP involved linear
time-trend variables.

Panel study of nasal lavage
repeatedly performed on
44 school children (age
9-11 years) according to
protocol published by
Korenetal. (1990).
Samples collected morning
after "low"  and "high"
O3 days. Nasal lavage
samples analyzed for
polymorphonuclear
leukocyte counts, albumin,
tryptase, eosinophil
cationic protein, and
myeloperoxidase. Analysis
using individual regression
methods.
                                                           8 h-avg O3
                                                           (10 a.m.-6 p.m.):
                                                           90.3 ug/m3
                                                           Range 2.9-247.1
                    NO,, TSP
Stratified analysis
of half hour avg
O3 at 3 p.m.:

Low:
<140 ug/m3
High:
>180 ug/m3
None
                Daily average concentrations of O3 were
                associated with daily sample means of
                FEV[ and FEF25.75 in this random adult
                cross-sectional sample.  The associations
                between daily O3 levels and daily means
                of lung function were smaller in
                magnitude than the association between
                annual O3 levels in the previous analyses
                (Ackermann and Liebrich et al., 1997).
                This analytic approach was designed to
                filter out long-term components.
                Sensitive analyses indicated that major
                confounding by uncontrolled effects of
                pollen, NO2, and TSP was unlikely.
Significant higher polymorphonuclear
leukocyte counts after high O3 days.
In children without symptoms of rhinitis,
significantly elevated myeloperoxydase
and eosinophil cationic protein
concentrations detected. Results suggest
that ambient O3 produces an
inflammatory  response in the upper
airways of healthy children.
                                       8-havgO3(perlOug/m3):

                                       % change in lung function:

                                       FEVI:  -0.51% (-0.88,-0.13)
                                       FVC: -0.24% (-0.59, 0.11)
                                       FEF25.75:  -1.04% (-1.85,-0.22)
                                                                                                                            Children without symptoms of
                                                                                                                            rhinitis (n = 30):

                                                                                                                            Myeloperoxydase:
                                                                                                                            LowO3:  median 77.39 ug/L
                                                                                                                            HighO3: median 138.60 ug/L
                                                                                                                            p < 0.05; Wilcoxon sign rank test

                                                                                                                            Eosinophilic cationic protein:
                                                                                                                            LowO3:  median 3.49  ug/L
                                                                                                                            High O3: median 5.39 ug/L
                                                                                                                            p < 0.05; Wilcoxon sign rank test

-------
OQ
 to
 o
 o
               Table AX7-1  (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
           Reference, Study
          Location and Period
 Outcomes and Methods
                                             Copollutants
 Mean O3 Levels    Considered
                                         Findings, Interpretation
                                                                   Effects
 X
         Europe (cont'd)

         Frischeretal. (1997)
         Umkirch, Germany
         May-Octl991
        Hoppeetal. (1995a,b)
        Munich, Germany
        Apr-Sep 1992-1994
Panel study examined 44
school children aged 9-11
years for ratio ofortho-
tyrosine to/>ara-tyrosine in
nasal lavage as a marker of
hydroxyl radical attack.
Nasal lavage performed
according to protocol
published by Koren et al.
(1990).  Concomitant lung
function tests performed.
Analysis using individual
regression methods.

Panel study of five study
groups (age 12-95 years):
(1) senior citizens (n = 41);
(2) juvenile asthmatics
(n = 43); (3) forestry
workers (n = 41);
(4) athletes (n = 43); and
(5) clerks (n = 40) as a
control group. Examined
for lung function (FVC,
FEVj, PEF) and questions
on irritated airways. Each
subject tested 8 days, 4
days with elevated or high
O3 and 4 days with low O3.
Analysis using Wilcoxon
matched pairs signed rank
test and linear regression.
Stratified analysis   None
of !/2-h avg O3 at
3 p.m.:

Low:
<140 |ig/m3
High:
>180 ue/m3
!/2-h max O3
(1 p.m.-4 p.m.):

Seniors:
High: 70ppb
Low:  31 ppb
                                                          Asthmatics:
                                                          High:  74 ppb
                                                          Low: 34 ppb

                                                          Forestry workers:
                                                          High:  64 ppb
                                                          Low: 32 ppb

                                                          Athletes:
                                                          High:  71 ppb
                                                          Low: 28 ppb

                                                          Clerks:
                                                          High:  68 ppb
                                                          Low:  15 ppb
None
                                  Ambient O3 was associated with the
                                  generation of hydroxyl radicals in the
                                  upper airways of healthy children and
                                  significant lung function decrements.
                                  However, the ortho/para ratio was not
                                  related to polymorphonuclear leukocyte
                                  counts.  Passive smoking was not related
                                  to outcomes.
                                  No indication that senior citizens
                                  represent a risk group in this study.
                                  Senior citizens had the lowest ventilation
                                  rate (mean 10 L/min).  Athletes and
                                  clerks experienced significant decrements
                                  in lung function parameters. Well-
                                  medicated juvenile asthmatics have a
                                  trend towards large pulmonary
                                  decrements.  Forestry workers were
                                  exposed to motor tool exhaust, which
                                  might be a potential promoting factor.
                                                      FEV; (% predicted):
                                                      Low: 105.4 (SD 15.6)
                                                      High: 103.9 (SD 15.0)
                                                      A:  1.5,p = 0.031

                                                      Ortho/para ratio:
                                                      Low: 0.02(SD0.07)
                                                      High: 0.18 (SD 0.16)
                                                      A:  0.17, p = 0.0001
!/2-h max O3 (per 100 ppb):

Change in lung function:

Seniors:
FEVj:  0.034 L(SD 0.101)
PEF: 0.006 L/s (SD 0.578)
Asthmatics:
FEVj:  -0.210 L(SD 0.281)
PEF: -0.712 L/s (SD 0.134)*
Forestry workers:
FEVI:  -0.140 L(SD 0.156)
PEF: -1.154 L/s (SD 0.885)*
Athletes:
FEVI:  -0.152 L(SD 0.136)*
PEF: -0.622 L/s (SD 0.589)*
Clerks:
FEVj:  -0.158 L(SD 0.114)*
PEF: -0.520 L/s (SD 0.486)*

*p<0.05

-------
OQ
 to
 o
 o
               Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
           Reference, Study
          Location and Period
 Outcomes and Methods
 Mean O3 Levels
 Copollutants
 Considered
       Findings, Interpretation
             Effects
 X
 oo
         Europe (cont'd)

         Hoppe et al. (2003)
         Munich, Germany
         Apr-Sep 1992-1995
        Koppetal. (1999)
        Two towns in Black
        Forest, Germany
        Mar-Oct 1994
Three of the same study
groups examined in Hoppe
etal. (1995a,b) —
asthmatics (n = 43),
athletes (n = 43), and
elderly = 41).  One
additional risk group,
children (n = 44), was
examined.

Over 80%  of the elderly
and asthmatic groups took
medications on a daily
basis. Eye and airway
symptoms  were assessed,
as were pulmonary
function test. GLM
analyses was conducted.
Panel study of 170 school
children (median age
9.1 years) followed over
11 time points with nasal
lavage sampling.  Subjects
were not sensitive to
inhaled allergens. Nasal
lavage samples analyzed
for eosinophil cationic
protein, albumen, and
leukocytes. Analysis
using GEE.
!/2-h max O3
(1 p.m.-4 p.m.):

Asthmatics:
High: 66.9ppb
Low:  32.5 ppb

Athletes:
High: 65.9 ppb
Low:  27.2 ppb

Children:
High: 70.4 ppb
Low:  29.8 ppb

Elderly:
High: 66.1 ppb
Low:  30.6 ppb
1/2-hmaxO3:

Villingen:
64 ug/m3
5th %-95th %
1-140

Freudenstadt:
105 ug/m3
5th %-95th %
45-179
    NO,
PM10,N02.
S02, TSP
For the group mean values there are
hardly any O3 effects detectable at the
concentration level of this study; lack of
power may have made it difficult to
detect small O3 effects.  Analysis on an
individual basis shows clearly different
patterns of O3 sensitivity. Ozone
responders are defined as individuals
with relevant lung function changes of at
least 10% for FEVl5 FVC, and PEF, and
20%forsRaw.  Most of the responders
were found in the asthmatic and children
groups. The sample size may limit
quantitative extrapolation to larger
populations, but may allow cautious
first estimates.
Eosinophil cationic protein and leukocyte
levels peaked soon after first major
O3 episode of summer, but did not
show response to later, even higher,
O3 episodes.  These observations are
consistent with an adaptive response
in terms of nasal inflammation.
!/2-h max O3 (per 50 ppb):

% change in lung function, lag 0:

Asthmatics:
FEVI: 4.26%(-3.13,11.66)
PEF: 6.67% (-1.55,  14.89)
Athletes:
FEVj: 0.01% (-0.13, 0.11)
PEF: -0.13% (-0.29, 0.03)
Children:
FEVj: -1.81%(-5.34,1.73)
PEF: -11.88% (-18.98,-4.78)
Elderly:
FEVI: 2.10% (-4.65, 8.84)
PEF: 7.29% (-2.84,  17.43)

Ozone responders:

Asthmatics: 21%
Athletes: 5%
Children: 18%
Elderly:  5%

Change in log eosinophil cationic
protein concentration (per ug/m3
03):

Early summer:
0.97(0.03,1.92)
Late summer:
-0.43 (-1.34, 0.47)

-------
               Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
OQ
 to
 o
 o
           Reference, Study
          Location and Period
                                                                     Copollutants
                                                                                            Findings, Interpretation
                                                                                                                Effects
 X
 VO
         Europe (cont'd)

         Ulmeretal. (1997)
         Freudenstadt and
         Villingen, Gennany
         Mar-Oct 1994
Cuijpers et al. (1994)
Maastricht, the
Netherlands
Nov-Dec 1990
(baseline),
Aug 8-16 1991 (smog
episode)
                        Panel study of 135 children
                        aged 8-11 years in two
                        towns were evaluated.
                        Pulmonary function was
                        associated with the highest
                        O3 concentration in the
                        previous 24 hours.
                        An initial cross-sectional
                        analysis was followed by
                        a longitudinal analysis
                        using GEE with the data
                        at four time periods
                        (Apr, Jun, Aug, Sep).
During episode,
212 children (age
unspecified) randomly
chosen from
535 reexamined for lung
function and symptoms.
Corrected baseline lung
function compared by
paired t-test. Difference in
prevalence of respiratory
symptoms examined.
                                             None
                           Freudenstadt:
                           Median 50.6 ppb
                           10th%-90th%
                           22.5-89.7

                           Villingen:
                           Median 32.1 ppb
                           10th%-90th%
                           0.5-70.1
Baseline:          PM10, SO2f
8-h avg O3:        NO2
Range 2-56 ug/m3

Smog episode:
l-hmaxO3:
Exceeded
160 ug/m3 on
11 days
                                  In the cross-sectional analysis,
                                  a significant negative association between
                                  O3 exposure and FVC was only shown
                                  at the June testing. ForFEVl5
                                  no significant associations were detected.
                                  In contrast, the longitudinal analysis
                                  obtained a statistically significant
                                  negative correlation between O3
                                  exposure, and FVC and FEV[ for the
                                  subpopulation living in the town with
                                  higher O3 levels, Freudenstadt.  The
                                  associations were more pronounced
                                  in males than females.
                Small decrements in FEV[ and FEF25.75
                were found in the 212 children.
                However, significant decreases in
                resistance parameters also were noted.
                Each day a different group of 30 children
                were measured. The results of the lung
                function are contradictory in that
                spirometry suggest airflow obstruction
                while impedance measurement suggest
                otherwise.  Respiratory symptoms
                impacted by low response rate of 122 of
                212 children due to summer holidays.
                No increase was observed.
                                                       Change in lung function (per ug/m3
                                                       '/2-h max O3):
Freudenstadt:
-1. 13 mL,p = 0.002
Villingen:
-0.19mL,p = 0.62

FVC:
Freudenstadt:
- 1. 23 mL,p = 0.002
Villingen:
0.02 mL,p = 0.96

Change in lung function and
impedance between baseline and
smog episode:
                                                                                                                                    -0.032 L(SD 0.226), p< 0.05
                                                                                                                                    FEF25.75:
                                                                                                                                    -0.086 L/s (SD 0.415), p < 0.01

                                                                                                                                    Resistence at 8 Hz:
                                                                                                                                    -0.47 cmH20/(L/s) (SD 1.17),
                                                                                                                                    p < 0.05
         Gielenetal. (1997)
         Amsterdam, the
         Netherlands
         Apr-Jul 1995
                        Panel study of 61 children
                        aged 7-13 years from two
                        special schools for
                        chronically ill children,
                        followed for twice-daily
                        PEF, symptoms, and
                        medication usage. 77% of
                        cohort had doctor-
                        diagnosed asthma.
                           l-hmaxO3:
                           77.3 ug/m3
                           SD 15.7

                           8-h max O3:
                           67.0 ug/m3
                           SD 14.9
PM10, BS,       Morning PEF significantly associated
pollen           with 8-h max O3 at a lag of 2 days.
                BS also associated with PEF. Among
                14 symptom models tested, only one
                yielded a significant O3 finding (for
                upper respiratory symptoms). PM10
                and BS, but not O3, were related to
                P2-agonist inhaler use.
                                                                         8-h max O3 (per 83.2 ug/m3):

                                                                         Percent change in PEF:
                                                                         Morning:
                                                                         Lag 2:  -1.86% (-3.58,-0.14)
                                                                         Afternoon:
                                                                         Lag 2:  -1.88% (-3.94, 0.18)

-------
               Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
OQ
 to
 o
 o
           Reference, Study
         Location and Period
                        Outcomes and Methods
                           Mean O3 Levels
 Copollutants
 Considered
       Findings, Interpretation
            Effects
 X
 to
 o
Europe (cont'd)

Hilterman et al. (1998)
Bilthoven, the
Netherlands
M-Octl995
        Hoek and Brunekreef
        (1995)
        Deume and Enkhuizen,
        the Netherlands
        Mar-M 1989
Panel study of 60 adult
nonsmoking intermittent to
severe asthmatics (age
18-55 years) followed over
96 days. Measured
morning and afternoon
PEF, respiratory
symptoms, and medication
use. Analysis controlled
for time trends,
aeroallergens,
environmental tobacco
smoke exposures, day of
week, temperature.
Lags of 0 to 2 days
examined.

The occurrence of acute
respiratory symptoms
investigated in children
aged 7-11 years (Deume
n = 241; Enkhuizen
n = 59). Symptoms
included cough, shortness
of breath, upper and lower
respiratory symptoms,
throat and eye irritation,
headache and nausea.
Ozone-related symptom
prevalence and incidence
were examined.  Lags of
0 and 1 day, and mean O3
concentration from
previous week were
investigated.  Analyses
using Ist-order
autoregressive models and
logistic regression models.
                                                          80.1 ug/m3
                                                          Range 6-94
PM10,N02,
S02, BS
                                                  l-hmaxO3:

                                                  Deume:
                                                  57ppb
                                                  SD20
                                                  Range 22-107

                                                  Enkhuizen:
                                                  59ppb
                                                  SD14
                                                  Range 14-114
PM10,N02,
SO,
Ozone had strongest association with
symptoms of any pollutant analyzed.
PEF lower with O3 but not statistically
significant.  No effect on medication use.
No effect modification by steroid use or
hyperresponsiveness.
No consistent association between
ambient O3 concentrations and the
prevalence or incidence of symptoms in
either city. The one significant positive
coefficient in Enkhuizen for prevalence
of upper respiratory symptoms was not
confirmed by the Deurne results.
No associations of daily symptom
prevalence or incidence found with any
of the other copollutants examined.
8-h max O3 (per 100 ug/m3):

Odds ratios:
Respiratory symptoms:

Shortness of breath:
LagO:  1.18(1.02,1.36)
Sleep disturbed by breathing:
LagO:  1.14(0.90,1.45)
Pain on deep inspiration:
LagO:  1.44(1.10,1.88)
Cough of phlegm:
LagO:  0.94(0.83,1.07)
Bronchodilator use:
LagO:  1.05(0.94,1.19)
1-h max O3 (per 50 ppb):

Prevalence of symptoms:

Deume:
Any respiratory symptom:
LagO:  -0.06 (SE 0.04)
Cough:
LagO:  -0.07 (SE 0.07)

Upper respiratory symptoms:
LagO:  -0.06 (SE 0.05)

Enkhuizen:
Any respiratory symptom:
LagO:  0.12(SE0.07)
Cough:
LagO:  -0.07 (SE 0.18)
Upper respiratory symptoms:
LagO:  0.18(SE0.09)*

*p<0.05

-------
OQ
 to
 o
 o
               Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
           Reference, Study
                                                                    Copollutants
         Location and Period     Outcomes and Methods    Mean O3 Levels    Considered
                                                                                         Findings, Interpretation
                                                                                    Effects
Latin America
        Castillejos et al. (1995)
        SW Mexico City
        Augl990-Octl991
                       Panel study of children
                       aged 7/2-11 years
                       (22 males, 18 females)
                       tested up to 8 times for
                       FEV; and FVC, before and
                       after exercise. Target
                       minute ventilation was
                       35 L/min/m2. Analysis
                       using GEE models.
                  PM,f
112.3ppb
Range 0-365

5th quintile mean
229.1 ppb
The mean % decrements in lung function   Percent change with exercise in
were significantly greater than zero only
in the fifth quintile of O3 exposure
(183-365 ppb).
5th quintile of O3 exposure
(183-365 ppb):
                                                                                                                                        -2.85%(-4.40, -1.31)
                                                                                                                                 FVC: -1.43% (-2.81, -0.06)
 X
 to
        Gold etal. (1999)
        SW Mexico City
        1991
                       Panel study of 40 school
                       children aged 8-11 years in
                       polluted community
                       followed for twice-daily
                       PEF and respiratory
                       symptoms. PEF deviations
                       in morning/afternoon from
                       child-specific means
                       analyzed in relation to
                       pollution, temperature,
                       season, and time trend.
                       Morning symptoms
                       analyzed by Poission
                       regression.
24-h avg O3:        PM2 5, PM10     Reported significant declines in PEF
52.0 ppb                          in relation to 24-h avg O3 levels.
IQR 25                           Effects did not vary by baseline symptom
                                 history.  Lags chosen to maximize effects
                                 and varied by outcome. Ozone generally
                                 robust to PM25. Morning phlegm
                                 significantly related to 24-h avg O3 at
                                 a 1-day lag.
                                      24-h avg O3 (per 25 ppb):

                                      Percent change in PEF:
                                      Morning:
                                      Lag 1-10:  -3.8%(-5.8,-1.8)
                                      Afternoon:
                                      Lag 0-9:  -4.6% (-7.0,-2.1)

                                      Percent change in phlegm:
                                      Morning:
                                      Lagl:  1.1% (1.0, 1.3)

-------
OQ
 to
 o
 o
               Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
           Reference, Study
                                                                   Copollutants
         Location and Period     Outcomes and Methods     Mean O3 Levels    Considered
                                                                                         Findings, Interpretation
                                                                                   Effects
 X
 to
 to
Latin America (cont'd)

Romieuetal. (1996)
N Mexico City
Apr-Jul 1991,
Nov 1991-Feb 1992
                               Panel study of 71 mildly
                               asthmatic children aged
                               5-13 years followed for
                               PEF and respiratory
                               symptoms.  Lower
                               respiratory symptoms
                               included cough, phlegm,
                               wheeze and/or difficulty
                               breathing.
l-hmaxO3:
190ppb
SD80
PM25, PM10,     Ozone effects observed on both PEF and    l-hmaxO3 (per 50 ppb):
NO2, SO2       symptoms.  Symptom, but not PEF,
               effects robust to PM10 in two-pollutant
               models.  Symptoms related to O3
               included cough and difficulty breathing.
                                      Change in PEF (L/min):
                                      Morning:
                                      Lag 0:  -2.44 (-4.40, -0.49)
                                      Lagl:  -0.23 (-0.41, 1.62)
                                      Lag 2:  -1.49 (-3.80, 0.80)
                                      Afternoon:
                                      LagO:  -0.56 (-2.70, 1.60)
                                      Lagl:  -1.27 (-3.20, 0.62)
                                      Lag 2:  -1.92 (-4.50, 0.66)

                                      Odds ratios:
                                      Lower respiratory symptoms:
                                      LagO:  1.09(1.03,1.15)
                                      Lagl:  1.10(1.04,1.17)
                                      Lag 2:  1.04(0.97,1.12)
        Romieuetal. (1997)
        SW Mexico City
        Apr-Jul 1991,
        Nov 1991-Feb 1992
                       Same period as Romieu
                       et al., 1996, but in different
                       section of city. 65 mildly
                       asthmatic children aged
                       5-13 years followed for
                       twice-daily PEF, and
                       respiratory symptoms.
                       Up to 2 months follow-up
                       per child. Analysis
                       included temperature and
                       looked at 0- to 2-day lags.
                       No time controls. Lower
                       respiratory symptoms
                       included cough, phlegm,
                       wheeze and/or difficulty
                       breathing. Panel study.
l-hmaxO3:
196 ppb
SD78
PM,f
Ozone had significant effects on PEF and   1-h max O3 (per 50 ppb):
symptoms, with largest effects at lags 0
and 1 day.  Symptoms related to O3        Change in PEF (L/min):
included cough and phlegm.  Ozone        Morning:
effects stronger than those for PM10.        Lag 0:  -1.32 (-3.21, 0.57)
                                      Lagl:  -0.39 (-2.24, 1.47)
                                      Lag 2:  -0.97 (-2.94, 0.99)
                                      Afternoon:
                                      LagO:  -1.81 (-3.60,-0.01)
                                      Lagl:  -2.32 (-4.17,-0.47)
                                      Lag 2:  -0.21 (-2.44,2.02)

                                      Odds ratios:
                                      Lower respiratory symptoms:
                                      LagO:  1.11 (1.05,1.19)
                                      Lagl:  1.08(1.01,1.15)
                                      Lag 2:  1.07(1.02,1.13)

-------
               Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
OQ
 to
 o
 o
           Reference, Study
                                                                    Copollutants
          Location and Period     Outcomes and Methods     Mean O3 Levels     Considered
                                                                                          Findings, Interpretation
                                                                                                               Effects
 X
 to
Latin America (cont'd)

Romieuetal. (1998)
Mexico City
Mar-May 1996
(1st phase)
Jun-Aug 1996
(2nd phase)
        Romieu et al. (2002)
        Mexico City
        Octl998-Apr2000
Panel study of 47 street
workers aged 18-58 years
randomly selected to take a
daily supplement (vitamin
C, vitamin E, and beta
carotene) or placebo during
1 st phase of study.
Following washout period,
the use of supplements and
placebos was reversed
during 2nd phase.
Pulmonary function test
performed twice a week at
end of workday. Plasma
concentrations of beta
carotene and a-tocopherol
measured. Analysis using
GEE models.

Panel study of 158
asthmatic children aged
6-16 years randomly given
a vitamin (C and E)
supplement or placebo
followed for 12 weeks.
Peak flow was measured
twice a day and spirometry
was performed twice per
week in the morning.
Double blind study.
Plasma concentration of
vitamin E levels measured.
Analysis using GEE
models.
                                                           l-hmaxO3:
                                                           123 ppb
                                                           SD40

                                                           55% of days >110
                                                           ppb.

                                                           Workday hourly
                                                           average during
                                                           workday prior to
                                                           pulmonary
                                                           function test:
                                                           67.3 ppb
                                                           SD24
PM10,N02
                                                  l-hmaxO3:
                                                  102 ppb
                                                  SD47
PM10,N02
During the 1 st phase, O3 levels were
significantly associated with declines in
lung function parameters.  No
associations were observed in the daily
supplement group. A significant
supplement effect was observed. Ozone-
related decrements also were observed
during the 2nd phase, however the
associations were not significant.
Supplementation with antioxidants during
the 1 st phase may have had a residual
protective effect on the lung.
Ozone levels were significantly
correlated with decrements in FEF25.75
in the placebo group, but not in the
supplement group. When analysis was
restricted to children with moderate-to-
severe asthma, amplitudes of decrements
were larger and significant for FEV1;
FEF25.75, and PEF in the placebo group.
Supplementation with antioxidants may
modulate the impact of O3 exposure on
the small airways of children with
moderate to severe asthma.
1-h max O3 (per 10 ppb):

Placebo group:

1st phase:
LagO:  - 17.9 mL (SE 5.4)*
FVC:
LagO:  - 14.8 mL(SE 7.1)*

2nd phase:
FEVI:
LagO:  -3.3mL(SE6.5)
FVC:
LagO:  -0.27 mL (SE 7.8)

No significant associations with O3
observed when taking supplements.

1-h max O3 (per 10 ppb):

Children with moderate to severe
asthma:

Placebo group:
O3 with PM10 and NO2 models:
FEVI:
Lagl:  -4.59 mL, p = 0.04
FEF2.5.75:
Lagl:  -13.32mL/s,p<0.01
PEF:
Lagl:  - 15.01 mL/s, p = 0.04

No association observed in the
vitamin supplement group.

-------
OQ
 to
 o
 o
               Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
           Reference, Study
                                                                     Copollutants
          Location and Period     Outcomes and Methods    Mean O3 Levels    Considered
                                                                                           Findings, Interpretation
                                                                                     Effects
Latin America (cont'd)
 X
 to
         Romieu et al. (2004)
         Mexico City
         Octl998-Apr2000
                       Additional analysis of
                       Romieu et al., 2002 with
                       data on glutathion
                       S-transferase Ml
                       polymorphism (GSTM1
                       null genotype) in
                       158 asthmatic children.
                       Analysis performed using
                       GEE models, stratified by
                       GSTM1 genotype (null
                       versus positive) within the
                       two treatment groups
                       (placebo and antioxidant
                       supplemented). Panel
                       study.
l-hmaxO3:
102ppb
SD47
None
In the placebo group, O3 exposure was
significantly and inversely associated
with FEF2 5.75 in children who had the
GSTM1 null genotype, with larger effects
seen in children with moderate-to-severe
asthma. No significant decrements were
seen in the GSTM1 positive children.
These results provide preliminary
evidence that asthmatic children who
may be genetically impaired to handle
oxidative stress are most susceptible to
the effect of O3 on small airways
function.
1-h max O3 (per 50 ppb):
                                                      FEF25.75 in children with moderate
                                                      to severe asthma:
                                                                         Placebo group:
                                                                         GSTM1 null:
                                                                         Lagl: -80.8mL/s,p = 0.002
                                                                         GSTM1 positive:
                                                                         Lag 1: -34.4 mL/s, p> 0.10

                                                                         Supplement group:
                                                                         GSTM1 null:
                                                                         Lag 1: -7.3 mL/s, p> 0.10
                                                                         GSTM1 positive:
                                                                         Lagl: 2.0 mL/s, p> 0.10
         Australia
         Jalaludin et al. (2000)
         Sydney, Australia
         Feb-Dec 1994
                       Panel study of three groups
                       of children (mean age 9.6
                       years): (1) n = 45 with
                       history of wheeze
                       12 months, positive
                       histamine challenge, and
                       doctor-diagnosed asthma;
                       (2) n = 60 with history of
                       wheeze and doctor-
                       diagnosed asthma;
                       (3) n = 20 with only history
                       of wheeze.  Examined for
                       evening PEF and daily O3
                       using GEE model and
                       population regression
                       models.
Mean daytime O3   PM10, NO2       A significant negative association was
(6 a.m.-9 p.m.):                     found between daily mean deviation in
12 ppb                            PEF and same-day mean daytime O3
SD 6.8                            concentration after adjusting for
                                  copollutants, time trend, meteorological
Maximum                         variables, pollen count, w&Alternaria
daytime O3                        count. The association was stronger in
(6 a.m.-9 p.m.):                     a subgroup of children with bronchial
26 ppb                            hyper-reactivity and doctor-diagnosed
SD 14.4                           asthma. In contrast, the same-day
                                  maximum O3 concentration was not
                                  statistically associated.
                                                       Change in PEF (per 10 ppb mean
                                                       daytime O3):

                                                       All children (n = 125):

                                                       O3 only model:
                                                       -0.9178 (SE 0.4192), p = 0.03
                                                       O3 with PM10 model:
                                                       -0.9195 (SE 0.4199), p = 0.03
                                                       O3 with PM10 and NO2 model:
                                                       -0.8823 (SE 0.4225), p = 0.04

-------
             Table AX7-1 (cont'd).  Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
X
to
Reference, Study
Location and Period
Australia (cont'd)
Jalaludin et al. (2004)
Sydney, Australia
Feb-Dec 1994














Asia
Park et al. (2002)
Seoul, Korea
Mar 1996-Dec 1999





Outcomes and Methods

Same three groups of
children as studied in
Jalaludin et al., 2000.
Examined relationship
between O3 and evening
respiratory symptoms
(wheeze, dry cough, and
wet cough), evening
asthma medication use
(inhaled p2-agonist and
inhaled corticosteroids),
and doctor visits for
asthma. Analysis using
GEE logistic regression
models. Panel study.



Time-series study. Poisson
GAM with default
convergence criteria used
in analysis. Children from
1st to 6th grade at one
elementary school located
in high traffic area
Copollutants
Mean O3 Levels Considered

Mean daytime O3 PM10, NO2
(6 a.m.-9 p.m.):
12 ppb
SD6.8

Maximum
daytime O3
(6 a.m.-9 p.m.):
26 ppb
SD 14.4








8-h avg O3 PM10, NO2,
(10a.m.-6p.m.): SO2, CO
22.86 ppb
Range 3. 13-69. 15




Findings, Interpretation

No significant O3 effects observed on
evening symptoms, evening asthma
medication use, and doctors visits.
Also, no differences in the response of
children in the three groups. A potential
limitation is that the use of evening
outcome measures rather than morning
values may have obscured the effect of
ambient air pollutants. Only consistent
relationship was found between mean
daytime PM10 concentrations and doctor
visits for asthma.






Ozone positively associated with illness-
related absences at a lag of 0-day. For
non-illness-related absences, inverse
relationship with O3 observed. PM10 and
SO2 also associated with illness-related
absences. Ozone effects were robust in
two-pollutant models.

Effects

Mean daytime O3 (per 8.3 ppb):

Odds ratios:
All children (n = 125):

Wheeze:
Lagl: 1.00(0.93,1.08)
Dry cough:
Lagl: 1.03(0.96,1.11)
Wet cough:
Lagl: 0.97(0.92,1.03)
Inhaled p2-agonist use:
Lagl: 1.02(0.97,1.07)
Inhaled corticosteroid use:
Lagl: 1.02(0.99,1.04)
Doctor visit for asthma:
Lagl: 1.05(0.77,1.43)

8-h avg O3 (per 15.94 ppb):

Relative risks:

All absences:
1.01(0.99,1.03)
Illness-related absences:
                             followed for school
                             absences. Average
                             enrollment count was
                             1,264. Each absence
                             classified as illness-related
                             or not. Single-day lags of
                             0 and 1 day, and a
                             cumulative 7-day lag
                             considered.
1.08(1.06,1.11)
Non-illness-related absences:
0.84 (0.80, 0.87)

-------
OQ
 to
 o
 o
               Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
           Reference, Study                                                   Copollutants
          Location and Period     Outcomes and Methods     Mean O, Levels     Considered
                                                                   Findings, Interpretation
                                                   Effects
        Asia (cont'd)
 X
 to
         Chen etal. (1998)
         Six communities in
         Taiwan
         1994-1995
4,697 school children (age
unspecified) from a rural
area (Taihsi), urban areas
(Keelung and Sanchung),
and petrochemical
industrial areas (Jenwu,
Linyuan, and Toufen)
cross-sectionally examined
for respiratory symptoms
and diseases using parent-
completed questionnaires.
Multiple logistic regression
models used to compare
odds of symptoms and
diseases in urban or
petrochemical areas to the
rural area after controlling
for potential confounding
factors.  Cross-sectional
panel study.
24-h avg O3:       SO2, CO,
                  PM10,N02
Rural area:
52.56 ppb

Urban area:
Mean range
38.34-41.90 ppb

Petrochemical
industrial area:
Mean range
52.12-60.64 ppb
School children in urban communities,
but not in petrochemical industrial areas,
had significantly more respiratory
symptoms and diseases compared to
those living in the rural community.
However, mean O3 levels in the urban
communities were lower than that of the
rural community. No causal relationship
could be derived between O3 and
respiratory symptoms and diseases in this
cross-sectional study.
                                                                        Urban areas compared to rural area:

                                                                        Odds ratios:
                                                                        Respiratory symptoms:

                                                                        Morning cough:
                                                                        1.33(0.98,1.80)
                                                                        Day or night cough:
                                                                        1.67(1.21,2.29)
                                                                        Shortness of breath:
                                                                        1.40(1.04,1.91)
                                                                        Wheezing or asthma:
                                                                        1.68(1.11,2.54)
         Chen etal. (1999)
         Three towns in Taiwan:
         Sanchun, Taihsi,
         Linyuan
         May 1995-Jan 1996
Valid lung function data
obtained once on each of
895 children (age 8-13
years) in three towns.
Examined relation between
lung function and pollution
concentrations on same day
and over previous 1, 2, and
7 days. Multipollutant
models examined.
Cross-sectional panel
study.
1 -h max O3:        SO2, CO,        FEV1 and FVC significantly associated
Range 19.7-110.3   PM10, NO2       with 1-day lag O3. FVC also related to
ppb                               NO2, SO2, and CO. No PM10 effects
SD not provided.                    observed. Only O3 remained significant
                                  in multipollutant models. No PM10
                                  effects. A significant O3 effect was not
                                  evident at O3 levels below 60 ppb.
                                       Change in lung function:

                                       O3 only models:
                                       Lagl:
                                       FEVI: -0.64 mL/ppb(SE 0.34)*
                                       FVC:  -0.79 mL/ppb (SE 0.32)*

                                       O3 with NO2 models:
                                       Lagl:
                                       FEVp -0.85 mL/ppb (SE 0.34)*
                                       FVC:  -0.91 mL/ppb (SE 0.37)*

                                       *p<0.05

-------
OQ
 to
 o
 o
               Table AX7-1 (cont'd). Effects of Acute O3 Exposure on Lung Function and Respiratory Symptoms in Field Studies
           Reference, Study                                                  Copollutants
         Location and Period     Outcomes and Methods     Mean O, Levels    Considered
                                                                  Findings, Interpretation
                                                                                   Effects
        Asia (cont'd)
 X
 to
        Chan and Wu (2000)
        Taichung City, Taiwan
        Sep2001
        (Questionnaire survey)
        Nov-Dec 2001
        (Field study)
A cohort of mail carriers
(mean age 39 years)
examined for effects of O3
exposure on PEF. Their
exposure periods were
between 9 a.m. and 5 p.m.
every working day. PEF
measurements taken twice
daily. Single-day lags
from 0 to  3 days examined.
A two-step statistical
model was used, a multiple
linear regression without
air pollutants followed by a
linear mixed effects model
to estimate pollution
effects.
8-h avg O3         PM10, NO2      Significant associations observed
(9 a.m.-5 p.m.):                    between evening PEF and O3
35.6 ppb                          concentrations at lags of 0, 1 and 2 days.
SD 12.1                          Largest effect observed at a 1-day lag.
Range 7.6-65.1                    Similar O3 effects on morning PEF also
                                 noted (data not presented). Neither PM10
                                 nor NO2 showed consistent associations
                                 with PEF. Ozone results were robust to
                                 adjustment for PM10 and NO2.
8-h avg O3 (per 10 ppb):

% change in evening PEF:
LagO: 0.54%,p<0.05
Lagl: 0.69%,p<0.05
Lag 2: 0.52%,p<0.05

-------
                               Table AX7-2.  Effects of Acute O3 Exposure on Cardiovascular Outcomes in Field Studies
OQ
 to
 o
 o
           Reference, Study
          Location and Period
  Outcomes and Methods     Mean O3 Levels
                   Copollutants
                   Considered
                       Findings, Interpretation
                                                    Effects
 X
 to
 oo
         United States

         Liao et al. (2004)
         Three locations in U.S.:
         Minneapolis, MN;
         Jackson, MS; Forsyth
         County, NC
         1996-1998
Cross-sectional study of
5,431 cohort members of
the Atherosclerosis Risk in
Communities study, men
and women aged 45-64
years at entry in 1987.
Association between O3 and
cardiac autonomic control
assessed using 5-minute
heart rate variability indices
collected over a 4-hour
period. Analysis using
multivariable linear
regression models, adjusting
for individual
cardiovascular disease risk
factors and meteorological
factors.
8-h avg O3         PM10, CO,       Significant interaction between O3 and
(10 a.m.-6 p.m.):   SO2, NO2        ethnicity in relation to high-frequency
41 ppb                            power (p < 0.05).  Ambient O3
SD 16                             significantly associated with high-
                                  frequency  power among whites, but not
                                  blacks. No significant O3 effect on other
                                  heart rate variability indices, including
                                  low-frequency power and SD of normal
                                  R-R intervals.  More consistent
                                  relationships observed between PM10
                                  and heart rate variability indices.
                                                        8-h avg O3 (per 16 ppb):

                                                        Log-transformed high-frequency
                                                        power:
                                                        White race:
                                                        Lagl:  -0.069 (SE 0.019)*
                                                        Black race:
                                                        Lagl:  0.047 (SE 0.034)

                                                        Log-transformed high-frequency
                                                        power:
                                                        Lagl:  -0.010 (SE 0.016)

                                                        SD of normal R-R intervals:
                                                        Lagl:  -0.336 (SE 0.290)

                                                        *p<0.05
         Peters et al. (2000a)
         Eastern Massachusetts
         1995-1997
Records of detected
arrhythmias and therapeutic
interventions were
downloaded from
defibrillators implanted in
cardiac clinic patients aged
22-85 years (n =  100).
Analysis was restricted to
defibrillator discharges
precipitated by ventricular
tachycardias or fibrillation.
Data were analyzed by
logistic regression models
using fixed effect models
with individual intercepts.
24-h avg O3:
18.6 ppb
IQR 14.0
PM25,PM10,
BC, CO, N02,
SO,
No significant O3 effects observed for
defibrillator discharge interventions.
For patients with ten or more
interventions, increased arrhythmias
were associated significantly with
PM2 5, CO, and NO2 at various lag
periods, but not O3.
24-h avg O3 (per 32 ppb):

Odds ratios:
Defibrillator discharges:

Patients with at least one event:
LagO:  0.96(0.47,1.98)
Patients with at least ten events:
LagO:  1.23(0.53,2.87)

-------
OQ
 to
 o
 o
                         Table AX7-2 (cont'd).  Effects of Acute O3 Exposure on Cardiovascular Outcomes in Field Studies
           Reference, Study
          Location and Period
                          Outcomes and Methods     Mean O3 Levels
                                               Copollutants
                                               Considered
       Findings, Interpretation
                                                                                      Effects
United States (cont'd)

Peters etal. (2001)
Greater Boston area,
MA
Janl995-May 1996
                                 Case-crossover study
                                 design used to investigate
                                 association between air
                                 pollution and triggering of
                                 myocardial infarction in
                                 772 patients (mean age
                                 6 1.6 years). For each
                                 subject, one case period
                                 was matched to three
                                 control periods 24 hours
                                 apart.  Conditional logistic
                                 regression used for analysis.
                                                    19.8ppb

                                                    24-h avg O3:
                                                    19.9ppb
                                              PM25,PM10,
                                              PM10.25,BC,
                                              CO, NO,, SO,
                                                                                      None of the gaseous pollutants, including   Odds ratios:
O3, were significantly associated with
the triggering of myocardial infarctions.
Significant associations reported for
PM2 5 and PM10.
                                                                          Myocardial infarctions:

                                                                          2-h avg O3 (per 45 ppb):
                                                                          Lag 1 hour:
                                                                          1.31 (0.85,2.03)

                                                                          24-h avg O3 (per 30 ppb):
                                                                          Lag 24 hours:
                                                                          0.94(0.60,1.49)
 X
 to
 VO
Park et al. (2005)
Greater Boston area,
MA
Nov 2000-Oct 2003
Cross-sectional study
examining the effect of O3
on heart rate variability in
497 adult males (mean age
72.7 years).  Subjects were
monitored during a 4-
minute rest period between
8 a.m. and 1  p.m.  Ozone
levels measured at central
site 1 km from study site.
Exposure averaging times
of 4-hours, 24-hours, and
48-hours investigated.
Modifying effects of
hypertension, ischemic
heart disease, diabetes,
and use of cardiac/
antihypertensive
medications  also examined.
Linear regression analyses.
24-h avg O3:       PM2 5, particle    Of the pollutants examined, only PM25
23.0 ppb          number         and O3 showed significant associations
SD 13.0           concentration,    with heart rate variability outcomes.
                  BC, NO2, SO2,   The 4-hour averaging period was most
                  CO             strongly associated with heart rate
                                  variability indices. The O3 effect was
                                  robust in models including PM2 5. The
                                  associations between O3 and heart rate
                                  variability indices were stronger in
                                  subjects with hypertension (n = 335) and
                                  ischemic heart disease (n = 142).
                                  In addition, calcium-channel blockers
                                  significantly influenced the effect of O3
                                  on low frequency power. Major
                                  limitations of this study are the use  of a
                                  short 4-minute period to monitor heart
                                  rate variability and the lack of repeated
                                  measurements for each subject.
                                        4-h avg O3 (per 13 ppb):

                                        Change in low frequency power:

                                        All subjects:
                                        -11.5% (-21.3, -0.4)
                                        Subjects with hypertension:
                                        -12.6% (-25.0, 1.9)
                                        Subjects without hypertension:
                                        -5.4% (-21.6, 14.1)
                                        Subjects with ischemic heart
                                        disease:
                                        -25.8% (-41.9,-5.3)
                                        Subjects without ischemic heart
                                        disease:
                                        -4.8% (-16.7, 8.8)

-------
                         Table AX7-2 (cont'd).  Effects of Acute O3 Exposure on Cardiovascular Outcomes in Field Studies
OQ
 to
 o
 o
           Reference, Study
          Location and Period
                         Outcomes and Methods    Mean O3 Levels
                                              Copollutants
                                              Considered
                                        Findings, Interpretation
Effects
 X
United States (cont'd)

Gold et al. (2000;
reanalysis Gold et al.,
2003)
Boston, MA
Jun-Sep 1997
Panel study of repeated
measurements of heart rate
variability in subjects aged
53-87 years (n = 2 1,1 63
observations). Twenty-five
minute protocol included
5 minutes each of rest,
standing, exercise outdoors,
recovery, and 20 cycles of
slow breathing. Ozone
levels measured at central
site 4.8 miles from study
site.  Analyses using
random effects models and
GAM with stringent
convergence criteria.
l-hmaxO3:        PM25, NO2,     Increased levels of O3 were associated
25.7 ppb           SO2            with reduced r-MSSD (square root of the
IQR 23.0                         mean of the squared differences between
                                 adjacent normal RR intervals) during the
                                 slow breathing period after exercise
                                 outdoors. The estimated O3 effects were
                                 similar to those of PM25.  Results suggest
                                 that O3 exposure may decrease vagal
                                 tone, leading to reduced heart rate
                                 variability.
                                                                                                                                    1-h max O3 (per 23.0 ppb):

                                                                                                                                    Change in r-MSSD:

                                                                                                                                    During first rest period:
                                                                                                                                    O3 only model:
                                                                                                                                    -3.0 ms(SE 1.9), p = 0.12

                                                                                                                                    During slow breathing period:
                                                                                                                                    O3 only model:
                                                                                                                                    -5.8ms(SE2.4),p = 0.02
                                                                                                                                    O3 with PM2 5 model:
                                                                                                                                    -5.4 ms(SE 2.5), p = 0.03
         Schwartz et al. (2005)
         Boston, MA
         12 weeks during the
         summer of 1999
                        A panel study of 28 elderly
                        subjects (age 61-89 years).
                        Various HRV parameters
                        were measured for 30
                        minutes once a week.
                        Analysis using linear mixed
                        models with log-
                        transformed HRV
                        measurements. To examine
                        heterogeneity of effects,
                        hierarchical model was
                        used.
l-havgO3:         BC, PM25,     HRV parameters examined included:
Median 34 ppb     CO, SO2, NO2   standard deviation of normal RR intervals
IQR 26                           (SDNN), root mean squared differences
                                 between adjacent R-R intervals
                                 (r-MSSD), proportion of adjacent NN
                                 intervals differing by more than 50 ms
                                 (PNN50), and low frequency/high
                                 frequency ratio (LFHFR).  Ozone was
                                 weakly associated with HRV parameters.
                                 Strongest association seen for BC, an
                                 indicator of traffic  particles. The random
                                 effects model indicated that the negative
                                 effect of BC on HRV was not restricted
                                 to a few subjects. Subjects with MI
                                 experienced greater BC-related
                                 decrements in HRV parameters.  Authors
                                 noted that in this study ambient O3 might
                                 represent a secondary particle effect and
                                 not a true O3 effect, and suggested that
                                 personal exposure measurements might
                                 be necessary to assess the effect of O3 on
                                 cardiovascular outcomes.
                                                                                                    1-h avg O3 (per 26 ppb):

                                                                                                    Change in HRV parameters:

                                                                                                    SDNN:
                                                                                                    -3.1ms (-7.0, 0.9)
                                                                                                    r-MSSD:
                                                                                                    -8.5ms (-16.6, 0.3)
                                                                                                    PNN50:
                                                                                                    -6.5% (-18.9, 7.8)

-------
                         Table AX7-2 (cont'd).  Effects of Acute O3 Exposure on Cardiovascular Outcomes in Field Studies
OQ
 to
 o
 o
           Reference, Study
          Location and Period
  Outcomes and Methods     Mean O3 Levels
                  Copollutants
                   Considered
                       Findings, Interpretation
                                                  Effects
 X
         United States (cont'd)

         Dockery et al. (2005)
         Boston, MA
         Ml 995-Jul 2002
Effect of air pollution on
incidence of ventricular
arrhythmias was examined
in 203 patients with
implantable cardioverter
defibrillators using time-
series methods. Mean
follow-up period was
3.1 years/subject. All
subjects located <40 km of
air pollution monitoring
site.  Two-day mean air
pollution level used in
analysis. Results analyzed
by logistic regression using
GEE with random effects.
Modifying effects of
previous arrhythmia within
3 days also examined.
48-h avg O3:
Median 22.9 ppb
IQR 15.4
PM25, BC,
SO42~, particle
number, CO,
NO2, SO2
No associations were observed between
air pollutants and ventricular arrhythmias
when all events were considered.
When only examining ventricular
arrhythmias within 3 days of a prior
event, positive associations were found
for most pollutants except for O3.
Suggestive evidence of a concentration-
response relationship between ventricular
arrhythmias and increasing quintiles
of03.
48-h avg O3 (per 15.4 ppb):

Odds ratios:

All events:
1.09(0.93,1.29)
Prior arrhythmia event < 3 days:
1.01  (0.76,1.35)
Prior arrhythmia event > 3 days:
1.14(0.92,1.40)

-------
                         Table AX7-2 (cont'd).  Effects of Acute O3 Exposure on Cardiovascular Outcomes in Field Studies
OQ
 to
 o
 o
           Reference, Study
          Location and Period
  Outcomes and Methods    Mean O3 Levels
                   Copollutants
                   Considered
                       Findings, Interpretation
                                                    Effects
 X
 to
         United States (cont'd)

         Rich et al. (2005)
         Boston, MA
         Ml 995- Jul 2002
Same study population as
Dockery et al. (2005).
Case-crossover study design
used to examine association
between air pollution and
ventricular arrhythmias.
For each case period,
3-4 control periods were
selected.  Various moving
average concentrations of
exposure considered - lags
0-2, 0-6, 0-23, and
0-47 hours. Analysis using
conditional logistic
regression models.
1 -h avg O3:        PM2 5, BC,       Associations observed for PM2 5 and O3
Median 22.2 ppb   CO, NO2, SO2    with a 24-h moving average, and for NO2
IQR21.7                          and SO2 with a 48-h moving average. In
                                  two-pollutant analyses, only PM25 and O3
24-h avg O3:                       appeared to act independently. In
Median 22.6 ppb                   contrast to results from other pollutants,
IQR 15.7                          stratified analyses indicated that O3 was
                                  associated with increased risk among
                                  subjects without a recent event, but not
                                  those with recent events. The odds ratio
                                  for the 24-h moving average
                                  concentration was larger than that for the
                                  same-calendar day concentration. This
                                  suggested that using calendar day
                                  concentrations might result in greater
                                  exposure misclassification which could
                                  lead to underestimation of risk. Findings
                                  of an association with 24-h moving
                                  average concentrations but not with
                                  shorter time periods could imply  a
                                  cumulative effect across the previous 24
                                  hours.
                                                        Odds ratios:

                                                        24-h moving average O3 (per
                                                        15.9 ppb):

                                                        O3 only model:
                                                        All events:
                                                        1.21 (1.00,1.45)
                                                        Prior arrhythmia event < 3 days:
                                                        1.04(0.78,1.37)
                                                        Prior arrhythmia event > 3 days:
                                                        1.28(1.05,1.58)

                                                        O3 with PM2 5 model:
                                                        All events:
                                                        1.18(0.94,1.49)

                                                        24-h calendar day O3 (per 15.7
                                                        ppb):

                                                        O3 only model:
                                                        All events:
                                                        0.96(0.80,1.15)
         Canada
         Rich et al. (2004)
         Vancouver, British
         Columbia, Canada
         Feb-Dec 2000
Case-crossover study
design used to investigate
association between air
pollution and cardiac
arrhythmia in patients aged
15-85 years (n = 34) with
implantable cardioverter
defribillators. Controls
periods were selected 7 days
before and after each case
day. Analysis using
conditional logistic
regression.
l-hmaxO3
27.5 ppb
SD9.7
IQR 13.4
PM25,PM10,
EC, OC,
so/-, co,
NO2, SO2
No consistent association between any
of the air pollutants, including O3, and
implantable cardioverter defribillators
discharges. No significant association
observed in all year data, however,
significant relationship found in winter
months at a 3-day lag.  Overall, little
evidence that air pollutants affect risk
of cardiac arrhythmias, however, power
was limited to study subtle effects.
No quantitative results for O3.

-------
                         Table AX7-2 (cont'd).  Effects of Acute O3 Exposure on Cardiovascular Outcomes in Field Studies
OQ
 to
 o
 o
           Reference, Study
          Location and Period
                          Outcomes and Methods     Mean O3 Levels
                  Copollutants
                   Considered
                       Findings, Interpretation
                                                   Effects
         Canada (cont'd)

         Vedal et al. (2004)
         Vancouver, British
         Columbia, Canada
         1997-2000
                        Retrospective, longitudinal
                        panel study of 50 patients
                        (age 12-77 years) with
                        implantable cardioverter
                        defribillators. Occurrence
                        of cardiac arrhythmia was
                        associated with air
                        pollutants over four-year
                        period. GEE used for
                        analysis.
28.2 ppb
SD 10.2
IQR 13.8
PM10, CO,
N02, S02
No consistent association between any
of the air pollutants and percent change in
arrhythmia. Among patients with at least
2 arrhythmia event-days per year, a
significant negative relationship between
O3 and arrhythmias was observed at lag
3-day during the summer, but no
associations were found during the
winter. These results do not provide
evidence for an O3 effect on cardiac
arrhythmias in susceptible patients.
                                                                         No quantitative results for O3.
 X
Europe

Ruidavets et al. (2005)
Toulouse, France
Jan 1997 - Jun 1999
                                 MONICA Project.
                                 Examined short-term effects
                                 of pollution on acute MI
                                 using case-crossover study
                                 design. The study
                                 population included
                                 395,744 inhabitants aged
                                 35 to 64 years. Acute MI
                                 was examined using
                                 clinical, electrocardiograms,
                                 and enzymatic data. Four
                                 case definitions were used;
                                 there were a total of 635
                                 cases for the most inclusive
                                 definition.  Deaths were
                                 validated. Age, gender,
                                 history of ischemic heart
                                 disease, and survival status
                                 evaluated.  Analyses using
                                 fixed-effects method with
                                 conditional logistic
                                 regression.
8-h max O3:          NO2, SO2      After adjustment for temperature, relative
74.8  ug/m3                        humidity, and influenza epidemics, an
SD 28.1                           association between O3 and acute MI was
Range 3.8-160.2                    found for 0- and 1-day lags, but not for
                                  longer lags. Older age was an important
                                  risk factor. Subjects with no personal
                                  history of ischemic heart disease yielded
                                  a stronger association. Moderate levels
                                  of NO2 and SO2 were observed. NO2 and
                                  SO2 were not associated with acute MI.
                                  No PM data was reported.
                                                        8-h max O3 (per 5 ug/m3):

                                                        Relative risk:

                                                        All cases (n = 635):
                                                        LagO:  1.05(1.01,1.08)
                                                        Lagl:  1.05(1.01,1.09)

                                                        Age group:
                                                        Age 35-54 years (n = 281):
                                                        LagO:  1.04(0.99,1.09)
                                                        Age 55-64 years (n = 283):
                                                        LagO:  1.06(1.01,1.12)

                                                        History of ischemic heart disease:
                                                        Yes (n = 127):
                                                        LagO:  1.03(0.96,1.12)
                                                        No (n = 437):
                                                        LagO:  1.05(1.01,1.09)

                                                        Age 55-64 years with no history
                                                        of ischemic heart disease (n =
                                                        225):
                                                        LagO:  1.07(1.01,1.13)
                                                        Lagl:  1.11 (1.04,1.19)

-------
OQ
 to
 o
 o
                         Table AX7-2 (cont'd).  Effects of Acute O3 Exposure on Cardiovascular Outcomes in Field Studies
 Reference, Study                                                   Copollutants
Location and Period     Outcomes and Methods    Mean O3 Levels    Considered
                                        Findings, Interpretation
                                                                                                                                                Effects
         Latin America
 X
         Holguin et al. (2003)
         Mexico City
         Feb-Apr 2000
                       Panel study of the
                       association between O3
                       and heart rate variability
                       examined in 34 elderly
                       subjects (mean age
                       79 years) in a nursing home.
                       Subjects were monitored
                       during a 5-minute rest
                       period between 8 a.m. and
                       1 p.m. every other day for a
                       3-month period. A total of
                       595 observations were
                       collected. Ambient O3
                       levels obtained from central
                       site 3 km upwind from
                       study site. Analysis
                       performed using GEE
                       models adjusting for
                       potential confounding
                       factors including age and
                       average heart rate.
l-hmaxO3
149 ppb
SD40
PM2 5 (indoor,
outdoor,
total), NO2,
SO2, CO
Only PM2 5 and O3 were significantly
associated with heart rate variability
outcomes. A significant effect of O3 on
heart rate variability was limited to
subjects with hypertension (n = 21).
In two-pollutant models, the magnitude
of the PM25 effect decreased slightly but
remained significant, whereas O3 was
no longer associated with heart rate
variability indices.
1-h max O3 (per 10 ppb):

Log10 high frequency
power/100,000 ms2:
All subjects:
-0.010 (-0.022, 0.001)
Subjects with hypertension:
-0.031 (-0.051,-0.012)
Subjects without hypertension:
0.002 (-0.012, 0.016)

Log10 low frequency
power/100,000 ms2:
All subjects:
-0.010 (-0.021, 0.001)
Subjects with hypertension:
-0.029 (-0.046,-0.011)
Subjects without hypertension:
0.001 (-0.012, 0.015)

-------
                                            Table AX7-3. Effects of O3 on Daily Emergency Department Visits
OQ
 to
 o
 o
           Reference, Study
          Location and Period
                           Outcomes and
                               Design
Mean O3 Levels
Copollutants
 Considered
Lag Structure
  Examined
     Method, Findings,
       Interpretation
          Effects
(Relative Risk and 95% CI)
 X
United States
Jaffe et al. (2003)
Cincinnati, Cleveland,
and Columbus, OH
Jun-Aug 1991-1996
         Jones etal. (1995)
         Baton Rouge, LA
         Jun-Aug 1990
                                 Daily time-series
                                 study of emergency
                                 department visits for
                                 asthma among
                                 Medicaid recipients
                                 aged 5-34 years.
                        Daily emergency
                        department visits for
                        respiratory complaints
                        over a 3-month period
                        in pediatric
                        (age 0-17 years), adult
                        (age 18-60 years),
                        and geriatric
                        (age >60 years)
                        subgroups.
                        Time-series study.
8-h max O3:

Cincinnati:
60ppb
SD20

Cleveland:
SOppb
SD17

Columbus:
57ppb
SD16
PM10,N02,
SO,
    1,2,3
l-hmaxO3:
69.1 ppb
SD28.7
                                                        24-h avg O3:
                                                        28.2 ppb
                                                        SD11.7
Mold, pollen     Not specified.
Poisson regression with control
for city, day of week, week,
year, minimum temperature,
overall trend, and a dispersion
parameter.  No specific effort
to control cycles, but
regression residuals were
uncorrelated, presumably due
to seasonal restriction.  Results
shown for individual cities and
overall. PM10 available only
every 6th day. Positive
relationships between
emergency department visits
for asthma and 8-h max O3
levels lagged 2 to 3 days.
Results of borderline statistical
significance. Other pollutants
also related to asthma
emergency department visits
in single-pollutant models.

Relatively simple statistical
approach using ordinary least
squares regression to model
effects of O3 by itself and of O3
along with pollen counts, mold
counts, temperature, and
relative humidity. No direct
control of cycles but authors
reported nonsignificant
autocorrelations among model
residuals. Data restriction to
3-month period may have
reduced any cyclic behavior.
Significant associations
between respiratory emergency
department visits and O3
observed for adult age group
only in multiple regression
models.
8-h max O3 (per 30 ppb):

Cincinnati:
Lag 2:  1.16(1.00,1.37)
Cleveland:
Lag 2:  1.03(0.92,1.16)
Columbus:
Lag 3:  1.16(0.98,1.37)

Three cities: 1.09(1.00,
1.19)
                                               24-h avg O3 (per 20 ppb):

                                               Pediatric: 0.87(0.65,1.09)
                                               Adult:  1.20(1.01,1.39)
                                               Geriatric: 1.27(0.93,1.61)

-------
                       Table AX7-3 (cont'd).  Effects of O3 on Daily Emergency Department Visits
JW
to
o
o
Reference, Study
Location and Period

Outcomes and
Design

Mean O3 Levels

Copollutants
Considered

Lag Structure
Examined

Method, Findings,
Interpretation

Effects
(Relative Risk and 95% CI)

United States (cont'd)












X

OJ
ON



O
^
'•Tj
H
6
o
1 — ^
2,
o
NW'
H
0

0
H
W
O

O
1 — I
H
W
Wilson et al. (2005) Daily emergency
Portland, ME room visits for total
1998-2000 respiratory and asthma
Manchester, NY examined. Time-
1996-2000 series study.











Cassino et al. (1999) Daily time-series
New York City study of emergency
Aug 1992-Dec 1995 department visits in a
cohort of 1 ,1 1 5 cohort
of 1,115 adult
asthmatics aged 1 8-
84. Stratified into
552 never-smokers,
278 light smokers, and
285 heavy smokers.









8-h max O3: SO2 0

Portland:
All available
data:
43.1 ppb
SD 13.5
Manchester:
Summer:
42.8 ppb
SD 14.6
Fall:
30.6 ppb
SDH. 5


l-hmax03: CO,NO2, 0,1,2,3
37.2 ppb S02
IQR28

24-h avg O3:
17.5 ppb
IQR14












Poisson GAM with stringent
convergence criteria. Positive
associations for asthma in the
larger city, Portland. Authors
expressed the view that larger
cities might be needed to
conduct such studies.









Used Poisson GAM with
default convergence criteria.
No warm season results
presented. Significant O3
effects seen only at lag 2
among heavy-smokers.
Copollutants did not have
effects. Short-term cycles and
episodic variations in asthma
may not have been controlled
adequately with 3 -month
period LOESS. Multiple tests
performed, and inconsistent
results across smoking strata
and lags raise possibility of
chance findings. No PM
results included.


8-h max O3 (per 30 ppb3):

Portland:
Total respiratory:
0.970(0.915,1.029)
Asthma:
1.094(1.032,1.160)
Manchester:
Total respiratory:
0.970(0.915,1.029)
Asthma:
0.970(0.863,1.092)




O3 24-h avg (per 14 ppb):

Heavy smoker subgroup:
LagO: 0.87(0.75-1.02)
Lagl: 1.07(0.93-1.24)
Lag 2: 1.26(1.10-1.44)
Lag 3: 0.96(0.83-1.10)













-------
X
Reference, Study
Location and Period
United States (cont'd)
Weisel et al. (2002)
New Jersey
May-Aug 1995







Friedman etal. (2001)
Atlanta, GA
M-Aug 1996











Outcomes and
Design

Daily asthma
emergency department
visits for all ages.
Time-series study.






Emergency
department visits and
hospital admissions
for asthma in children
aged 1-16 years.
Outcomes measures
during 1996 Summer
Olympics were
compared to a
baseline period of
4 weeks before and
after the Olympic
Games. Time-series
study.
Mean O3 Levels

l-hmaxO3;
5-h avg O3
(10 a.m.-3 p.m.);
and 8-h avg O3
(2 p.m.-lO p.m.)
analyzed.

Levels not
reported.

l-hmaxO3:

Baseline:
81.3ppb
SD not given.

Intervention
period:
58.6 ppb
SD not given.




Copollutants Lag Structure Method, Findings,
Considered Examined Interpretation

Pollen, spores 0, 1, 2, 3 No control for time, but
authors report no
autocorrelation, which
alleviates concerns about lack
of control. Significant O3
effects reported, even after
adjusting for potential
confounding by pollen. All
three O3 indices gave
essentially same results.
NO2, SO2, 0,0-1,0-2 Analysis using Poisson GEE
CO, PM10, models addressing serial
mold autocorrelation. An overall
decrease was observed when
comparing the number of visits
and hospitalizations during the
Olympic Games to the baseline
period. However, significant
associations between O3 and
asthma events were found
during the Olympic Games.



Effects
(Relative Risk and 95% CI)

Slope estimate
(visits/day/ppb):

Excluding data from May
when pollen levels are high:

O3 only model:
LagO: 0.039, p = 0.049
O3 with pollen model:
LagO: 0.040, p = 0.014
1-h max O3 (per 50 ppb):

Pediatric emergency
departments:
LagO: 1.2(0.99,1.56)
Lag 0-1: 1.4(1.04,1.79)
Lag 0-2: 1.4(1.03,1.86)








-------
                                       Table AX7-3 (cont'd). Effects of O3 on Daily Emergency Department Visits
uw
to
O
O
Reference, Study
Location and Period

Outcomes and
Design

Mean O3 Levels

Copollutants
Considered

Lag Structure
Examined

Method, Findings,
Interpretation

Effects
(Relative Risk and 95% CI)

X
United States (cont'd)

Metzger et al. (2004)
Atlanta, GA
Jan 1993-Aug 2000
        Peel et al. (2005)
        Atlanta, GA
        Jan 1993-Aug 2000
                                Emergency
                                department visits for
                                total and cause-
                                specific
                                cardiovascular
                                diseases by age groups
                                19+years and 65+
                                years.  Time-series
                                study.
                       8-h max O3:
                       Median 53.9 ppb
                       10th%-90th%
                       13.2-44.7
Emergency
department visits for
total and cause-
specific respiratory
diseases by age groups
0-1,2-18,19+, and
65+ years. Time-
series study.
8-h max O3:
55.6 ppb
SD23.8
                  NO2, SO2,            0-2        Poisson GLM regression used
                  CO, PM25,                      for analysis. A priori models
                  PM10, PM10.                     specified a lag of 0 to 2 days.
                  25, ultrafine                     Secondary analyses performed
                  PM count,                       to assess alternative pollutant
                  SO42~, H+,                       lag structures, seasonal
                  EC, OC,                        influences, and age effects.
                  metals,                          Cardiovascular visits were
                  oxygenated                     significantly associated with
                  hydrocarbons                    several pollutants, including
                                                  N02, CO, and PM2 5, but
                                                                 NO2, SO2,            0-2        Poisson GEE and GLM
                                                                 CO, PM2 5,                      regression used for analysis.
                                                                 PM10, PM10.                     A priori models specified a lag
                                                                 25, ultrafine                     of 0 to 2 days.  Also performed
                                                                 PM count,                       secondary analyses estimating
                                                                 SO42~, H+,                       the  overall effect of pollution
                                                                 EC, OC,                        over the previous two weeks.
                                                                 metals,                          Seasonal analyses indicated
                                                                 oxygenated                     stronger associations with
                                                                 hydrocarbons                    asthma in the warm months.
                                                                                                 Quantitative results not
                                                                                                 presented for multipollutant,
                                                                                                 age-specific, and seasonal
                                                                                                 analyses.
8-h max O3 (per 25 ppb):

All ages:
Total cardiovascular:
1.008(0.987,1.030)
Dysrhythmia:
1.008(0.967,1.051)
Congestive heart failure:
0.965(0.918,1.014)
Ischemic heart disease:
1.019(0.981,1.059)
Peripheral vascular and
cerebrovascular disease:
1.028(0.985,1.073)

8-h max O3 (per 25 ppb):

All ages:
Total respiratory:
1.024(1.008,1.039)
Upper respiratory infections:
1.027(1.009,1.045)
Asthma:
1.022(0.996,1.049)
Pneumonia:
1.015(0.981,1.050)
COPD:
1.029(0.977,1.084)

-------
                                       Table AX7-3 (cont'd).  Effects of O3 on Daily Emergency Department Visits
JW
to
o
o
Reference, Study
Location and Period

Outcomes and
Design

Mean O3 Levels

Copollutants
Considered

Lag Structure
Examined

Method, Findings,
Interpretation

Effects
(Relative Risk and 95% CI)

X
VO
United States (cont'd)

Tolbert et al. (2000)
Atlanta, GA
Jun-Aug 1993-1995
                                Pediatric (aged
                                0-16 years) asthma
                                emergency department
                                visits over three
                                summers in Atlanta.
                                A unique feature of
                                the study was
                                assignment of O3
                                exposures to zip code
                                centroids based on
                                spatial interpolation
                                from nine O3
                                monitoring stations.
                                Time-series study.
                       l-hmaxO3:
                       68.8 ppb
                       SD21.1
                       8-h max O3:
                       59.3 ppb
                       SD19.1
                   PM10, NO2,
                   mold, pollen
                                A priori specification of
                                model, including a lag of 1 day
                                for all pollutants and
                                meteorological variables.
                                Secondary analysis using
                                logistic regression of the
                                probability of daily asthma
                                visits, referenced to total visits
                                (asthma and nonasthma).
                                Significant association with
                                O3 and PM10 in 1-, but not in
                                2-pollutant models (correlation
                                between O3 and PM10:
                                r = 0.75). Secondary analysis
                                indicated nonlinearity, with O3
                                effects clearly significant only
                                for days > 100 ppb versus days
                                <50 ppb.
                               8-h max O3 (per 20 ppb):

                               Poisson regression:
                               O3 only model:
                               1.040(1.008,1.074)
                                                                                                        Logistic regression:
                                                                                                        O3 only model:
                                                                                                        1.042(1.017,1.068)
                                                                                                        O3 with PM10 model:
                                                                                                        1.024(0.982,1.069)
        Zhu et al. (2003)
        Atlanta, GA
        Jun-Aug 1993-1995
Asthma emergency
department visits in
children (age 0-16
years) over three
summers in Atlanta.
Time-series study.
8-h max O3:
Levels not
reported.
None
Used Bayseian hierarchical
modeling to address model
variability and spatial
associations. Data were
analyzed at the zip code level
to account for spatially
misaligned longitudinal data.
A positive, but nonsignificant
relationship between O3 and
emergency room visits for
asthma.
                                                                                                                                8-h max O3 (per 20 ppb):

                                                                                                                                Posterior median:
                                                                                                                                1.016(0.984,1.049)

-------
X
Reference, Study
Location and Period
Canada
Delfinoetal. (1997b)
Montreal, Quebec,
Canada
Jun-Sep 1992-1993














Outcomes and
Design

Daily time-series
ecologic study of
emergency department
visits for respiratory
complaints within five
age strata (<2, 2-18,
19-34, 35-64,
>64 years).










Mean O3 Levels

l-hmaxO3:

1992:
33.2ppb
SD 12.6

1993:
36.2 ppb
SD 13.8

8-h max O3:

1992:
28.8 ppb
SDH. 3
1993:
30.7 ppb
SD11.5
Copollutants Lag Structure Method, Findings,
Considered Examined Interpretation

PM10, PM25, 0,1,2 Used ordinary least squares,
SO42~, H+ with 1 9-day weighted moving
average pre-filter to control
cycles; weather also controlled.
Significant effects reported for
1-day lag O3 in 1993 only for
age >64 years. This O3 effect
reported to be robust in two-
pollutant models. LowO3
levels raise plausibility
concerns. Short data series,
multiple tests performed, and
inconsistent results across
years and age groups raise
possibility of chance findings.



Effects
(Relative Risk and 95% CI)

1993 (age >64 years):

1-h max O3 (per 36.2
Lag 1: 1.214(1.084,

8-h max O3 (per 30.7
Lagl: 1.222(1.091,














ppb):
1.343)

ppb):
1.354)












-------
g>
OQ
v>
r+
to
o
o
64 years).
l-hmaxO3:

1989:
44.1 ppb
SD 18.3

1990:
35.4 ppb
SD 12.9

8-h max O3:

1989:
37.5 ppb
SD 15.5

1990:
29.9 ppb
SD11.2
Estimated
PM,,
0, 1, 2        Same analytical approach used
             in Delfino et al, 1997.  Results
             presented only for 1989.
             Significant effects reported for
             1-day lag O3 in 1989 only for
             age >64 years. This O3 effect
             reported to be robust in
             2-pollutant models.
1989 (age >64 years):

1-h max O3 (per 44.1 ppb):
Lag 1:  1.187(0.969,1.281)

8-h max O3 (per 37.5 ppb):
Lagl:  1.218(1.097,1.338)

No significant O3 effects in
other age groups or for 1990.
        Stiebetal. (1996)
        Saint John, New
        Brunswick, Canada
        May-Sep 1984-1992
Daily emergency
department visits for
asthma in all ages,
age<15 years and
15+years.  Time-
series study.
1-h max O3:         SO2, NO2,         0,1,2,3      Poisson analysis with control
41.6 ppb           SO42~, TSP                      of time based on 19-day
Range 0-160                                        moving average filter. Also
95th % 75                                          controlled day of week and
                                                   weather variables.  Ozone was
                                                   only pollutant consistently
                                                   associated with emergency
                                                   department visits for asthma,
                                                   but effect appeared nonlinear,
                                                   with health impacts evident
                                                   only above 75 ppb O3.
                                                               1-h max O3 >75 ppb:

                                                               Lag 2:  1.33(1.10,1.56)

-------
X
to
Reference, Study
Location and Period
Europe
Sunyeretal. (1997)
Four European cities:
Barcelona, Helsinki,
London, and Paris
1986-1992












Atkinson etal. (1999a)
London, England
1992-1994




Hajatetal. (1999;
2002)
London, England
1992-1994










Outcomes and
Design

Emergency
admissions for asthma
in children (<15 year)
and adults (15-64
years). Time-series
study.











Emergency
department visits
for respiratory
complaints, asthma for
all ages and age 0-14,
15-64, and 65+ years.
Time-series study.
Daily doctor consults
for asthma, lower
respiratory diseases,
and upper respiratory
diseases for age 0-14,
15-64, and >65 years.
Time-series study.







Mean O3 Levels

l-hmaxO3:

Barcelona:
Median 72 ug/m3
Range 7-283

Helsinki:
Median 27 ug/m3
Range 1-78

London:
Median 40 ug/m3
Range 1-188

Paris:
Median 36 ug/m3
Range 1-230
8-h max O3:
17.5 ppb
SD11.5




8-h max O3:

All year:
17.5 ppb
SD11.5

Warm season:
22.7 ppb
SD 12.2

Cold season:
12.1 ppb
SD7.6

Copollutants Lag Structure Method, Findings,
Considered Examined Interpretation

BS, SO2, NO2 0, 1, 2, 3 Poisson analysis using APHEA
methodology. Significant O3
effects on emergency
admissions for asthma
observed among 15-64 year
olds in Barcelona and London.
Across all cities, there was no
strong evidence for
associations involving O3.








NO2, SO2, 0, 1 , 2, 3 Poisson GLM regression used
CO, PM10 0-1,0-2,0-3 for analysis. No warm season
analysis attempted. PM10
positively associated.



BS, SO2, 0-3 Used Poisson GAM with
NO2, CO, default convergence criteria.
PM10, pollen Conducted all year and
seasonal analyses. Single- and
two-pollutant models analyzed.
Significant negative effects for
O3. This may reflect residual
confounding by seasonal
factors or highly negative
correlation with other
pollutants..



Effects
(Relative Risk and 95% CI)

1-h max O3 (per 50 ug/m3):

Weighted mean effect (best
lag selected for each city):

Age<15 years:
3-city pooled estimate
(Barcelona data not
available):
1.006(0.976,1.037)

Age 15-64 years:
4-city pooled estimate:
1.015(0.955,1.078)



8-h max O3 (per 25.7 ppb):

All ages:
Total respiratory:
Lagl: 1.017(0.991,1.043)
Asthma:
Lag 1: 1.027(0.983,1.072)
Upper respiratory diseases,
age >65 years:

All year:
8-h max O3 (per 25.7 ppb)
Lag 2: -8.3% (-13.3, -3.0)

Warm season:
8-h max O3 (per 28.5 ppb)
Lag 2: -0.6% (-6.1, 5.1)

Cool season:
8-h max O3 (per 19.8 ppb)
Lag 2: -7.9% (-12.9, 2.7)

-------
g>
OQ
v>
r+
to
o
o
 14 years. 43 ppb
Time-series study. IQR22

Winter:
29 ppb
IQR16



and month dummy variables
and extensive control for
weather factors (minimum,
maximum, mean temperature,
relative humidity, dewpoint
temperature; continuous and
categorical parameterizations)




1-h max O3 (per 12.7 ppb):

Summer:
0.991 (0.939,

Winter:
1.055(0.998,







1.045)


1.116)






-------
OQ
 to
 o
 o
                                      Table AX7-3 (cont'd).  Effects of O3 on Daily Emergency Department Visits
           Reference, Study
          Location and Period
   Outcomes and
       Design
Mean O3 Levels
Copollutants
 Considered
Lag Structure
  Examined
Method, Findings,
  Interpretation
         Effects
(Relative Risk and 95% CI)
        Europe (cont'd)

        Tobias etal. (1999)
        Barcelona, Spain
        1986-1989
Daily asthma
emergency department
visits. Investigated
sensitivity of results
to four alternative
methods for
controlling asthma
epidemics. Time-
series study.
Levels not
reported.
BS, NO2, SO2   Not specified.
                Poisson analysis using
                APHEA methodology.
                Asthma epidemics either
                not controlled, or controlled
                with one, six, or individual
                epidemic dummy variables.
                        O3 results were sensitive to
                        method used to control
                        asthma epidemics, with
                        regression coefficients
                        ranging over 5-fold
                        depending on the model.
                        Only 1 of 8 models reported
                        had a significant O3 effect.
 X
Tenias et al.
(1998; 2002)
Valencia, Spain
1994-1995









Latin America
Hemandez-Garduno
etal. (1997)
Mexico City
May 1992-Jan 1993




Daily emergency
department visits for
asthma and COPD
in persons aged
>14 years. Time-
series study.








Visits to clinics for
respiratory diseases in
persons aged 1 month
to 92 years. Time-
series study.



l-hmaxO3: BS,NO2,
S02, CO
All year:
62.8 ug/m3
Range 13.3-157.3

Warm season:
74.0 ug/m3

Cool season:
5 1.4 ug/m3



Percent time SO2, NO2,
exceeding 1-h CO
max O3 of 120
ppb:
6.1-13.2% by
location


0, 1,2, 3, 4, 5 Poisson analysis using APHEA
methodology. Compared
warm and cold season effects.
GAM explored in sensitivity
analysis. For asthma, both O3
and NO2 significant in single-
and two-pollutant models, and
O3 effect larger in warm
season. For COPD, both O3
and CO significant in both
single- and two-pollutant
models and no difference in O3
effects by season.

0,1,2,3,4,5 GLM with pre-adjustment.
Ozone at lags 0 and 5 days
significantly associated with
daily visits for all ages,
age <14 years, and 15+ years.
Neither O3 nor NO2 significant
in two-pollutant model.

1-h max O3 (per 10 ug/m3):

Asthma:
All year:
Lag 1: 1.06(1.01,1.11)
Warm season:
Lag 1: 1.08(1.02,1.05)
Cold season:
Lagl: 1.04(0.97,1.11)

COPD:
All year:
Lag 5: 1.06(1.02,1.10)

1-h max O3 (per maximum
less average, value not
given):

LagO: 1.19(SE0.08),
p<0.05
Lag 5: 1.19 (SE 0.08),
p<0.05

-------
X
Reference, Study
Location and Period
Latin America (cont'd)
Lin etal. (1999)
Sao Paulo, Brazil
May 1991 -Apr 1993



Martins et al. (2002)
Sao Paulo, Brazil
May 1996-Sep 1998





Ilabaca etal. (1999)
Santiago, Chile
Feb 1995-Aug 1996






Asia
Hwang and Chan
(2002)
50 cities in Taiwan
1998




Outcomes and
Design

Daily pediatric (age
unspecified)
respiratory emergency
department visits.
Time-series study.

Daily emergency
department visits
for chronic lower
respiratory diseases
in persons
aged >64 years.
Time-series study.

The association
between pollutant
levels and emergency
visits for pneumonia
and other respiratory
illnesses among
children. Time-series
study.


Daily clinic visits for
lower respiratory
illnesses for all ages.
Time-series study.





Mean O3 Levels

l-hmaxO3:
34ppb
SD22



l-hmaxO3:
34ppb
SD21
IQR21




O3 1-hmax:

Warm season:
66.6 ug/m3
SD25.2

Cold season:
27.6 ug/m3
SD 20.2

l-hmaxO3:
54.2 ppb
SD 10.2
Range 38.9-78.3




Copollutants Lag Structure Method, Findings,
Considered Examined Interpretation

SO2, CO, 0,1,0-1,0-2, Seasonal control using month
PM10,NO2 0-3,0-4,0-5 dummy variables. Also
controlled day of week,
temperature. Both O3 and
PM10 associated with outcome
alone and together.
CO,NO2, 0-1,0-2,0-3, Analyzed using Poisson GAM
SO2, PM10 0-4, 0-5, 0-6. with default convergence
criteria. Only O3 and SO2
significant in single-pollutant
models. Ozone effect
remained significant when SO2
included in two-pollutant
model.
PM10, PM25, 1, 2, 3, 1-7 Poisson regression analysis.
SO2,NO2








NO2, SO2, 0, 1 Analysis using general linear
PM10, CO regressions with
moving-average residual
processes and Bayesian
hierarchical modeling. All
pollutants except O3 were
associated with daily clinic
visits.
Effects
(Relative Risk and 95% CI)

1-h max O3 (per 5 ppb):

O3 only model:
Lag 0-4: 1.022(1.016,1.028)
O3 with PM10 model:
Lag 0-4: 1.015(1.009,1.021)
1-h max O3 (per 18.26 ppb):

Lag 0-3: 1.14(1.04,1.23)





Warm season:
1-h max O3 (per 30 ug/m3):
Lag 2: 1.019(1.003,1.035)

Cold season:
1-h max O3 (per 24 ug/m3):
Lag 2: 0.995(0.978,1.011)



1-h max O3 (per 40 ppb):

Bayesian hierarchical model:
1.003(0.983,1.023)





-------
 £                                   Table AX7-3 (cont'd). Effects of O3 on Daily Emergency Department Visits
OQ                                                                                                                                                   	
 03.         Reference, Study        Outcomes and                       Copollutants   Lag Structure        Method, Findings,                 Effects
 to       Location and Period          Design         Mean O3 Levels    Considered     Examined            Interpretation          (Relative Risk and 95% CI)
 o
 o
        Asia (cont'd)

        Chewet al. (1999)       Emergency             l-hmaxO3:        TSP, PM10,         0,1,2       Simplistic but probably          No quantitative results
        Singapore              department visits for     23 ppb            SO2, NO2                      adequate control for time by      presented for O3.
        Jan 1990-Dec 1994       asthma in persons       SD 15                                          including 1-day lagged
                               aged 3-21 years.                                                        outcome as covariate.
                               Time-series study.                                                      In adjusted models that
                                                                                                    included covariates, O3
                                                                                                    had no significant effect.
 X

-------
                                                Table AX7-4.  Effects of O3 on Daily Hospital Admissions
OQ
r+
O
O











Reference, Study
Location and Period
United States
Neidell (2004)
California
1992-1998








Outcomes and Design

Asthma hospital
admissions within five
age strata (0-1, 1-3,
3-6, 6-12, and
12- 18 years).
Time-series study.





Mean O3 Levels

O3 (index not
specified):
38.9 ppb
SD 17.8

Low SES:
40.1 ppb

High SES:
38.3 ppb

Copollutants
Considered

CO, NO2,
PM10;
multipollutant
models







Lag Structure Method, Findings,
Examined Interpretation

Not specified. Statistical analysis using
naturally occurring seasonal
variations in pollutant
concentrations within zip
codes. Linear regression
analyses. Consistent
significant positive effects
only observed for CO.
Negative O3 effect observed
in all age groups. Number of
smog alerts was negatively
Effects
(Relative Risk and 95% CI)

Slope estimate (adjusting for
number of smog alerts):

O3 with CO, NO2, and PM10
models:
Age 3-6 years:
-0.038 (SE 0.014)
Age 6-12 years:
-0.044 (SE 0.013)
Age 12- 18 years:
-0.022 (SE 0.011)
X
O
HH
H
W
associated with asthma
hospitalizations, indicating
avoidance behavior on high
O3 days. Interaction term
with indicator variable for
low SES was positive in all
age groups and statistically
significant in age 3-6 years
and 12-18 years, after
adjusting for number of smog
alerts.
O3 x low SES interaction
term:
Age 3-6 years:
0.092 (SE 0.026)
Age 6-12 years:
0.024 (SE 0.024)
Age 12-18 years:
0.042 (SE 0.019)
-LJ
6
0
2;
o
H
O
0
H
W
0
Mann et al. (2002)
South Coast air basin,
CA
1988-1995







Ischemic heart disease
admissions for age
40+ years. Time-series
study.







8-h max O3:
50.3 ppb
SD30.1
IQR 39.6







PM10, CO, 0, 1 , 2, 3, 4, 5, Poisson GAM with cubic
NO2 0-1,0-2,0-3, B-splines; co-adjustment.
0-4 No significant O3 effects
observed overall or in warm
season. CO and NO2
significant in full-year
analyses.




O3 coefficients all negative,
but no consistent, significant
effect.









-------
g>
CfQ
wa
r-K
K>
o
(^













X
_|^
oo

o
?d
i-rj
H
o
0

o
H
0

H

0
O
H
W

Reference, Study
Location and Period

United States (cont'd)
Linn et al. (2000)
Los Angeles, CA
1992-1995













Nauenberg and Basu
(1999)
Los Angeles, CA
1991-1994

Sheppard et al.
(1999; reanalysis
Sheppard, 2003)
Seattle, WA
1987-1994








Table

Outcomes and Design


Total respiratory and
total cardiovascular
admissions for age
30+ years. Time-series
study.











Unscheduled asthma
admissions for all ages.
Time-series study.


Asthma admissions for
age <65 years. Time-
series study.










AX7-4 (cont'd).

Mean O3 Levels


24-h avg O3:

Winter:
14 ppb
SD7

Spring:
32 ppb
SD10

Summer:
36 ppb
SD8
Fall:
15 ppb
SD9
24-h avg O3:
19. 88 ppb
SD11.13


8-h max O3:
30.4
IQR20










Effects of O3 on Daily Hospital Admissions
Copollutants Lag Structure Method, Findings,
Considered Examined Interpretation


PM10, CO, 0 Poisson GLM;
NO2 co-adjustment. Only
significant O3 effects
observed were inverse
associations with total
cardiac admission in full-year
and winter season, suggesting
residual confounding.
No significant effects of O3
on respiratory admissions.






PM10 0, 0-7 Poisson GLM with pre-
adjustment. No significant
effects of O3. No warm
season results presented.

PM2 5, PM10, 1,2,3 Poisson GAM, reanalyzed
PM10_2 5, SO2, with stringent convergence
CO criteria; Poisson GLM.
Results stratified by season.
Ozone significant predictor
of outcome. No two-
pollutant model results
reported for O3.






Effects
(Relative Risk and 95% CI)


24-h avg O3 (per 10 ppb):

All year:
Respiratory:
1.008(1.000,1.016)
Cardiovascular:
0.993(0.987,0.999)









24-h avg O3 (per 20 ppb):

All insurance categories:
LagO: 1.01(0.93,1.08)

8-h max O3 (per 20 ppb):

GLM with natural splines:
Lag 2: 1.07(1.01,1.13)










-------
                                            Table AX7-4 (cont'd).  Effects of O3 on Daily Hospital Admissions
uw
to
O
O
Reference, Study
Location and Period

Outcomes and Design

Mean O3 Levels

Copollutants
Considered

Lag Structure
Examined

Method, Findings,
Interpretation

Effects
(Relative Risk and 95% CI)

VO
United States (cont'd)

Schwartz (1996)
Spokane, WA
Apr-Oct, 1988-1990
        Koken et al. (2003)
        Denver, CO
        Jul-Aug
        1993-1997
Total respiratory
admissions in persons
aged 65+ years.
                       Cause-specific
                       cardiovascular
                       admissions for age
                       >65 years.  Cause
                       categories include
                       acute MI, coronary
                       atherosclerosis,
                       pulmonary heart
                       disease, cardiac
                       dysrhythmia, and
                       congestive heart
                       failure.  Time-series
                       study.
l-hmaxO3:
79 ug/m3
IQR23

24-h avg O3
56 ug/m3
IQR17
                       24-h avg O3:
                       25.0ppb
                       SD6.61
                       Range 5.4-40.2
                                                                         PM1
                   PM10, NO2
                   SO,, CO
    2          Poisson GAM with default
               convergence criteria. Results
               available only for warm
               season. Ozone and PM10
               both significant predictors of
               outcome.  No two-pollutant
               models reported.  Ozone
               effects robust to more
               extensive temperature
               specification.

0,1, 2, 3,4     Analysis using Poisson
               GLM. Results suggest that
               O3 increases the risk of
               hospitalization for coronary
               atherosclerosis and
               pulmonary heart disease.
               No association was found for
               PM10.  Strong O3 effects
               observed in this seasonal
               study compared to other
               studies examining year-round
               data. Male gender and
               higher temperatures were
               found to be important risk
               factors for cardiovascular
               disease. No multipollutant
               analyses were reported.
1-h max O3 (per 50ug/m3):
Lag 2:  1.244(1.002-1.544)

24-h avg O3 (per 50ug/m3):
Lag 2:  1.284(0.926-1.778)
24-h avg O3 (per 9.7 ppb):

Acute MI:
0.824 (0.733, 0.925)
Coronary atherosclerosis:
1.123(1.040,1.214)
Pulmonary heart disease:
1.214(1.040,1.418)

-------
Table AX7-4 (cont'd). Effects of O3 on Daily Hospital Admissions
OQ
to
O
o















i-rj
H
6
0
o
H
0
o
H
W
O

i — i
H
W

Reference, Study
Location and Period

United States (cont'd)
Moolgavkar et al.
(1997)
Minneapolis/St. Paul,
MN and Birmingham,
AL
1986-1991




Schwartz etal. (1996)
Cleveland, OH
Apr-Oct 1988-1990




Gwynn and Thurston
(2001)
New York City
1988-1990




Gwynn et al. (2000)
Buffalo, NY
May 1988-Octl990





Outcomes and Design


Total respiratory,
pneumonia, and COPD
admissions for age
>64 years. Time-series
study.





Total respiratory
admissions for age
65+ years. Time-series
study.



Respiratory admissions
for all ages, stratified
by race and insurance
status. Time-series
study.



Total respiratory
admissions for all ages.
Time-series study.





Mean O3 Levels


24-h avg O3:

Minnesota:
26.2 ppb
IQR 15. 3

Alabama:
25.1 ppb
IQR 12.7

l-hmaxO3:
56 ppb
IQR 28




24-h avg O3:
22.1 ppb
IQR 14.1
Maximum 80.7




24-h avg O3:
26.2 ppb
IQR 14.8





Copollutants Lag Structure Method, Findings,
Considered Examined Interpretation


PM10, SO2, 0, 1 , 2, 3 Poisson GLM with co-
NO2 adjustment. Both O3 and
PM10 significant in MN; not
in AL. Ozone, but not PM10,
effects were robust to NO2
and SO2.




PM10, SO2 1-2 Poisson GLM with sinusoids;
co-adjustment. Results
available only for warm
season. Ozone and PM10
both significant predictors of
outcome. No two-pollutant
models reported.
H+, SO/", 1 GLM with high-pass filter.
PM10 Ozone associated with
respiratory admissions;
effects larger for nonwhites
and for those uninsured or on
medicaid.


PM10, SO/- , 0, 1 , 2, 3 Used Poisson with GAM
H+, COH, CO, default convergence criteria
NO2, SO2 for control of temperature;
moving average control for
time. Ozone significant
predictor of outcome. No
two-pollutant models
reported.
Effects
(Relative Risk and 95% CI)


24-h avg O3 (per 1 5 ppb):

Minnesota:

Total respiratory:
Lagl: 1.060(1.033,1.087)
Pneumonia:
Lag 1: 1.066(1.034,1.098)
COPD:
LagO: 1.045(0.995,1.067)
1-h max O3 (per 100 ug/m3):

1.09(1.02,1.16)




24-h avg O3 (per 58.6 ppb):

White:
1.032(0.987,1.079)
Non- white:
1.122(1.074,1.172)
Uninsured:
1.138(1.084,1.194)
24-h avg O3 (per 14.8 ppb):

Lag 1: 1.029(1.013-1.045)






-------
                                 Table AX7-4 (cont'd). Effects of O3 on Daily Hospital Admissions
OQ
to
o
o
X
Reference, Study
Location and Period
United States (cont'd)
Weisel et al. (2002)
New Jersey
May-Aug 1995







Canada
Burnett etal. (1997a)
16 Canadian cities
1981-1991


















Copollutants
Outcomes and Design Mean O3 Levels Considered

Asthma admissions for l-hmaxO3; Pollen, spores
all ages. Time-series 5-h avg O3
study. (10 a.m.-3 p.m.);
and 8-h avg O3
(2 p.m.-lO p.m.)
analyzed.

Levels not
reported.


Total respiratory l-hmaxO3: SO2, NO2,
admissions for all ages, CO, coefficien
age <65 years and All year: t of haze
65+ years. Time-series 31 ppb
study. 95th % 60

Mean range
across cities:
26-38 ppb
95th % 45-84

Jan-Mar:
26 ppb
Apr-Jun:
40 ppb
Jul-Sep:
26 ppb
Oct-Dec:
21 ppb
9.6% of O3 data
missing.
Lag Structure Method, Findings,
Examined Interpretation

0,1,2,3 No control for time, but
authors report no
autocorrelation, which
alleviates concerns about
lack of control. Significant
O3 effects reported after
adjusting for potential
confounding by pollen.



0, 1, 2, 0-1, Poisson GLM with co-
0-2, 1-2 adjustment. Results stratified
by season. Significant O3
effect observed in warm
season only. No O3 effects
on control outcomes. Results
consistent across cities.














Effects
(Relative Risk and 95% CI)

Slope estimate
(admissions/day /ppb):

5-h avg O3 and 8-h avg O3:
O3 only model:
Lag 2: 0.099, p = 0.057

All three O3 indices:
O3 with pollen model:
Lag 2: 0.1 l,p = 0.033

1-h max O3 (per 30 ppb):

All ages:
Jan-Mar:
Lag 1: 0.994(0.964,1.025)
Apr-Jun:
Lagl: 1.042(1.012,1.073)
Jul-Sep:
Lag 1: 1.050(1.026,1.074)
Oct-Dec:
Lagl: 1.028(0.998,1.059)











-------
X
to
Reference, Study
Location and Period
Canada (cont'd)
Burnett etal. (1995)
168 Hospitals in
Ontario, Canada
1983-1988




Burnett etal. (1997b)
Toronto, Ontario,
Canada
Summers 1992-1994





Burnett etal. (1999)
Toronto, Ontario,
Canada
1980-1994








Outcomes and Design Mean O3 Levels

Respiratory and 1-h max O3:
cardiovascular 36.3 ppb
admissions for all
ages and within age
strata. Study focused
mainly on testing for
sulfate effects. Time-
series study.
Unscheduled 1 -h max O3 :
respiratory and 41.2 ppb
cardiovascular IQR 22
admissions for all ages.
Time-series study.




Cause-specific 24-h avg O3:
respiratory and 19.5 ppb
cardiovascular IQR 19
admissions for all ages.
Cause categories
included asthma,
COPD, respiratory
infections, heart failure,
ischemic heart disease,
and cerebrovascular
disease. Time-series
study.
Copollutants Lag Structure Method, Findings,
Considered Examined Interpretation

SO42~ 1 GLM with pre-adjustment of
outcome variables. Results
stratified by season. Authors
report that O3 associated with
respiratory admission in
warm season only.


PM2 5, PM10, 0, 1 , 2, 3, 4, Poisson GLM with co-
H+, SO42~, SO2, 2 to 5 multiday adjustment. Results stratified
NO2, CO, periods lagged by season. Ozone and
coefficient 1 to 4 days coefficient of haze strongest
of haze predictors of outcomes.
Ozone effects on both
outcomes were robust to PM.
PM effects were not robust to
03.
Estimated 0,1,2,0-1, Poisson GAM with LOESS
PM25, PM10, 0-2, 1-2, 1-3, pre-filter applied to pollution
PM10_25, CO, 2-3,2-4 and hospitalization data.
NO2, SO2 Ozone effects seen for
respiratory outcomes only.
Ozone effect robust to PM;
not vice versa. No seasonal
stratification.




Effects
(Relative Risk and 95% CI)

No quantitative results
presented for O3.






12-h avg O3 (8 a.m.-8 p.m.)
(per 11.5 ppb):

Models adjusted for
temperature and dewpoint:
Respiratory :
Lag 1-3: 1.064(1.039,1.090)
Cardiovascular:
Lag 2-4: 1.074(1.035,1.115)
24-h avg O3 (per 19.5 ppb):

Asthma:
Lag 1-3: 1.063(1.036,1.091)
COPD:
Lag 2-4: 1.073(1.038,1.107)
Respiratory infection:
Lag 1-2: 1.044(1.024,1.065)





-------
                                           Table AX7-4 (cont'd). Effects of O3 on Daily Hospital Admissions
OQ
to
O
o
Reference, Study
Location and Period
Panarla (VonfrH
Outcomes and Design

Mean O3 Levels

Copollutants
Considered

Lag Structure
Examined

Method, Findings,
Interpretation

Effects
(Relative Risk and 95% CI)

        Burnett etal. (2001)
        Toronto, Ontario,
        Canada
        1980-1994
Acute respiratory
disease admissions for
age <2 years.  Time-
series study.
l-hmaxO3:
45.2ppb
IQR25
Estimated
PM25,PM10,
PM10.25,CO,
NO2, SO2
                    1,2,3,4,     Poisson GAM with LOESS
                    5, 0-4        pre-filter applied to pollution
                                 and hospitalization data.
                                 Sensitivity analyses using co-
                                 adjustment.  Results stratified
                                 by season. Ozone effects
                                 significant only in summer.
                                 Ozone effect robust to PM;
                                 not vice versa.
                             1-h max O3 (per 45.2 ppb):

                             Summer:
                             O3 only model:
                             Lag 0-4: 1.348(1.193,1.524)
                             O3 with PM2 5 model:
                             Lag 0-4: 1.330(1.131,1.565)
X
        Lin et al. (2003)
        Toronto, Ontario,
        Canada
        1981-1993
        Fung et al. (2003)
        Windsor, Ontario,
        Canada
        Apr 1995-Dec 2000
Asthma admission for
age 6-12 years.
Case-crossover design.
l-hmaxO3:
30 ppb
IQR20
CO, SO2, NO2
Cardiovascular hospital   1 -h max O3:
admissions for age       39.3 ppb
<65 and >65 years.       SD 21.4
Time-series study.       Range 1-129
                0,0-1,0-2,0-3,
                  0-4, 0-5, 0-6
NO2, SO2, CO,
PM10,
coefficient
of haze, total
reduced sulfur
compounds
                  0,0-1, 0-2
Conditional logistic
regression model analysis.
No O3 effects observed.
Positive relations to CO, SO2
and NO, observed.
                                                   Conducted both time-series
                                                   analysis using Poisson GLM
                                                   with natural splines.
                                                   Strongest effect observed for
                                                   SO2 in individuals aged
                                                   >65 years. No associations
                                                   were found any other
                                                   pollutant, including O3.
1-h max O3 (per 20 ppb):

Odds ratios:
Males:
LagO:  0.96(0.88,1.04)
Females:
LagO:  0.86(0.78,1.04)

1-h max O3 (per 29 ppb):

Age <65 years:
LagO:  0.999(0.913,1.093)
Lag 0-2:  1.042(0.923,1.177)

Age >65 years:
LagO:  0.974(0.924,1.027)
Lag 0-2:  1.014(0.941,1.092)

-------
X
Reference, Study
Location and Period
Canada (cont'd)
Luginaah et al. (2005)
Windsor, Ontario,
Canada
Apr 1995-Dec 2000












Outcomes and Design Mean O3 Levels

Respiratory hospital 1-h max O3:
admissions by gender 39.3 ppb
for all ages and age SD 21 .4
0-14, 15-64, and Range 1-129
65+ years. Time-series
study.










Copollutants
Considered

NO2, SO2, CO,
PM10,
coefficient
of haze, total
reduced sulfur
compounds










Lag Structure Method, Findings,
Examined Interpretation

0,0-1, 0-2 Conducted both time-series
analysis using Poisson GLM
with natural splines and
bidirectional case-crossover
analysis using conditional
logistic regression models.
For case-crossover analysis,
control periods selected two
weeks before and after the
case period. Results were
consistent for the time-series
and case-crossover analyses.
Significant associations were
found for all pollutants
except O3 and total reduced
sulfur compounds.
Effects
(Relative Risk and 95% CI)

1-h max O3 (per 29 ppb):

All ages:

Time-series analysis:
Males:
LagO: 1.04(0.92,1.17)
Females:
LagO: 0.95(0.82,1.10)

Case-crossover analysis:
Males:
LagO: 1.06(0.93,1.22)
Females:
LagO: 1.01(0.77,1.34)

Lin et al. (2004)
Vancouver, British
Columbia, Canada
1987-1998
                               Asthma admissions for    1-h max O3:
                   CO, SO2, NO2
                               age 6-12 years. Time-
                               series study.
28.02 ppb
SD11.54
IQR 14.81
0,0-1,0-2,0-3,
 0-4, 0-5, 0-6
Poisson GAM with LOESS
(using default convergence
criteria).  Repeated analysis
with natural cubic splines
using 1,000 iterations with
convergence criteria 10"15.
Results were similar for both
analyses. NO2 exposure
associated for males in low
SES but not high.  No
association for CO and O3 in
either SES group.
1-h max O3 (per 14.8 ppb):

Males:
Low SES:
Lag 1:  0.85(0.76,0.94)
High SES:
Lagl:  0.93(0.83,1.04)

Females:
Low SES:
Lagl:  1.11(0.97,1.28)
High SES:
Lag 1:  0.91 (0.78, 1.05)

-------
                                           Table AX7-4 (cont'd).  Effects of O3 on Daily Hospital Admissions
OQ
 to
 o
 o
          Reference, Study
         Location and Period
Outcomes and Design   Mean O3 Levels
                  Copollutants
                  Considered
Lag Structure
  Examined
Method, Findings,
  Interpretation
         Effects
(Relative Risk and 95% CI)
 X
        Canada (cont'd)

        Yang et al. (2003)
        Vancouver, British
        Columbia, Canada
        Jan 1986-Dec 1998
Daily respiratory
admissions in children
aged <3 years and
adults aged 65+ years.
Bidirectional
case-crossover.
Conditional logistic
regression.
24-h avg O3:        CO, NO2, SO2,     1,2,3,4,5     Used bidirectional case-
13.41 ppb          coefficient                       crossover analysis,
SD 66.61          of haze                          comparing air pollution on
IQR 9.74                                          day of admission to levels
                                                  one week prior and after.
                                                  SES evaluated. O3 was
                                                  positively associated with
                                                  respiratory hospital
                                                  admissions among young
                                                  children and the elderly.
                                            24-h avg O3 (per 9.74 ppb):

                                            Odds ratios:
                                            Age <3 years:
                                            Lag 4:  1.22(1.15,1.30)
                                            Age 65+ years:
                                            Lag 4:  1.13(1.09,1.18)
Europe
Anderson etal. (1997)
Five European cities:
London, Paris,
Amsterdam,
Rotterdam, Barcelona
Study periods vary by
city, ranging from
1977-1992




Atkinson et al. (2001)
Eight European cities:
Barcelona,
Birmingham, London
Milan, Netherlands,
Paris, Rome, and
Stockholm
Study periods vary by
city, ranging from
early to middle 1990s

Emergency COPD
admissions for all ages.
Each city analyzed
previously by
individual teams.
Results combined here
via meta-analysis.
Time-series study.




Total respiratory,
asthma, and COPD
admissions for all ages
and age 0-14,
15-64 and 65+ years.
Time-series study.





l-hmaxO3: TSP, SO2,
N02, BS
Median range
across five cities:

All year:
36-77 ug/m3
Warm season:
48-91 ug/m3
Cool season:
20-64 ug/m3

8-h max O3: PM10, NO2,
SO2, CO
Mean range:
26.0 ug/m3
(Rome) to 66.6
ug/m3
(Stockholm)




0, 1, 2, 3, 4, 5 Poisson GLM using APHEA
methodology. Results
stratified by season.
Ozone most consistent and
significant predictor of
admissions. Warm season
effect larger.





N/A Study focused on PM10
effects. Copollutants
included only as effect
modifiers. No direct O3
results shown. Ozone
appeared to modify the PM10
effect on respiratory
admissions for persons over
age 64 years of age.


1-h max O3 (per 50 ug/m3):

Weighted mean effect across
five cities (best lag selected
for each city):

All year:
1.03(1.01,1.05)
Warm season:
1.03(1.01,1.05)
Cool season:
1.01(0.98,1.05)
No results presented for O3.










-------
Table AX7-4 (cont'd). Effects of O3 on Daily Hospital Admissions
OQ
c
VI
to
o
o














(!/i
ON


O
^
Prj
H
6
0
2|
O

o
0
H
W
0
O
H
W

Reference, Study
Location and Period

Europe (cont'd)
Le Tertre et al.
(2002b)
Eight European cities:
Barcelona,
Birmingham, London
Milan, Netherlands,
Paris, Rome, and
Stockholm
Study periods vary by
city, ranging from
early to middle 1990s
Wong et al. (2002)
London, England
1992-1994
Hong Kong
1995-1997




















Outcomes and Design


Total cardiovascular,
ischemic heart disease,
and stroke admissions
for all ages and age
0-64 and 65+ years.
Time-series study.





Total respiratory
(>64 years), asthma
(15-64 years), total
cardiovascular (all
ages), and ischemic
heart disease (all ages)
admissions. Time-
series study.
















Copollutants
Mean O3 Levels Considered


8-h max O3: PM10, BS,
N02, S02, CO
Mean range:
26.0 ug/m3
(Rome) to 66.6
ug/m3
(Stockholm)




8-h max O3: PM10, NO2,
SO2
London:
All year:
34.9 ug/m3
SD23.1
Warm season:
45.3 ug/m3
Cool season:
24.0 ug/m3

Hong Kong:
33.5 ug/m3
SD23.0
Warm season:
32.0 ug/m3
Cool season:
35.1 ug/m3






Lag Structure Method, Findings,
Examined Interpretation


N/A Main focus on PM10 and BS.
Gaseous copollutants
evaluated as effect modifiers.
Greater PM10 effects seen in
cities with lower annual O3
levels. No risk estimates
presented for O3.




0, 1 , 2, 3 Poisson GAM with default
0-1 convergence criteria. Most
consistent associations found
with total respiratory
admissions. Ozone
associated with total
respiratory admissions in all
year and warm season
analyses in both cities.
Associations in cool season
analyses observed only in
Hong Kong. All year O3
effects robust to copollutants











Effects


(Relative Risk and 95% CI)




No results presented for O3.




















8-h max O3 (per 10 ug/m3):

Total respiratory:

London:
All year:
Lag 0-1: 1.008(1.
Warm season:
Lag 0-1: 1.010(1.
Cool season:
Lag 0-1: 1.002(0.

Hong Kong:
All year:
Lag 0-1: 1.008(1.
Warm season:
Lag 0-1: 1.008(1.
Cool season:
Lag 0-1: 1.010(1.









002,1.014)

003,1.017)

993,1.012)



003,1.013)

002,1.014)

002,1.017)





-------
                                            Table AX7-4 (cont'd).  Effects of O3 on Daily Hospital Admissions
uw
to
O
O
Reference, Study
Location and Period

Outcomes and Design

Mean O3 Levels

Copollutants
Considered

Lag Structure
Examined

Method, Findings,
Interpretation

Effects
(Relative Risk and 95% CI)

X
Europe (cont'd)

Anderson et al. (1998)
London, England
1987-1992
Atkinson etal. (1999b)
London, England
1992-1994
                              Admissions for asthma    8-h max O,:
                               in all ages and
                               age 0-14, 15-64,
                               and 65+ years.
                               Time-series study.
Total and cause-
specific respiratory and
cardiovascular
admissions in all ages
and in all ages and
age 0-14, 15-64, and
65+ years. Time-series
study.
                                              15.5 ppb
                                              IQR13

                                              l-hmaxO3:
                                              20.6 ppb
                                              IQR16
                                              8-h max O3:
                                              17.5 ppb
                                              SD11.5
SO2, NO2,         0,1,2,0-1,     Poisson GLM using APHEA
BS, pollens            0-2         method; co-adjustment.
                                 Ozone significantly
                                 associated with asthma
                                 admissions in the warm
                                 season for all ages and for
                                 age 15-64 years.  Warm
                                 season O3 effect robust in
                                 2-pollutant models. Inverse
                                 associations observed in the
                                 cool season for some age
                                 groups.

NO2, SO2, CO,    0, 1, 2, 3, 0-1,    Poisson GLM using APHEA
PM10, BS           0-2,0-3       methodology. No significant
                                 associations seen between O3
                                 and respiratory admissions.
                                 Ozone was positively
                                 associated with total
                                 cardiovascular admissions in
                                 age 65+ years. Seasonal
                                 analyses were not conducted.
                                                                                                       8-h max O3 (per 10 ppb):

                                                                                                       All ages:
                                                                                                       Warm season:
                                                                                                       Lag 1:  1.022(1.006,1.038)
                                                                                                       Cool season:
                                                                                                       Lagl:  0.968(0.946,0.992)
8-h max O3 (per 25.7 ppb):

All ages:
Total respiratory:
Lagl:  1.012(0.990,1.035)
Total cardiovascular:
Lag 2:  1.023(1.002,1.046)
        Ponce de Leon et al.
        (1996)
        London, England
        Apr 1987-Feb 1992
                       Total respiratory
                       admissions in several
                       age strata: all ages,
                       0-14, 15-64, 65+years.
                       Time-series study.
                       8-h avg O3          BS, SO2,NO2      0,1,2,0-1,    Poisson GLM using APHEA
                       (9 a.m.-5 p.m.):                         0-2,0-3,      co-adjustment methodology.
                       15.6 ppb                                           Ozone significant predictor
                       SD 12                                              overall. Effect larger and
                       IQR 14                                            more significant in warm
                                                                          season. Effect robust to
                                                                          copollutants. Effects varied
                                                                          by age.
                                                                                                                              All ages:

                                                                                                                              All year:
                                                                                                                              8-h avg O3 (per 26 ppb):
                                                                                                                              Lag 1:  1.029(1.011,1.048)

                                                                                                                              Warm season:
                                                                                                                              8-h avg O3 (per 29 ppb):
                                                                                                                              Lag 1:  1.048(1.025,1.073)

                                                                                                                              Cool season:
                                                                                                                              8-h avg O3 (per 20 ppb):
                                                                                                                              Lag 1:  0.996(0.972,1.021)

-------
                                           Table AX7-4 (cont'd). Effects of O3 on Daily Hospital Admissions
OQ
to
O
o
Reference, Study
Location and Period
F.nrnnp Tr.nnt'rn
Outcomes and Design

Mean O3 Levels

Copollutants
Considered

Lag Structure
Examined

Method, Findings,
Interpretation

Effects
(Relative Risk and 95% CI)

X
oo
        Poloniecki et al.
        (1997)
        London, England
        Apr 1987-Mar 1994
Prescott et al. (1998)
Edinburgh, Scotland
1992-1995
                       Cause-specific and total
                       circulatory admissions
                       for all ages. Time-
                       series study.
                       8-h avg O3
                       (9 a.m.-5 p.m.):
                       Median 13 ppb
                       Range 0-94
                   BS, NO2,
                   SO,, CO
Total respiratory and
cardiovascular
admissions for age
<65 years and
65+ years. Time-series
study.
24-h avg O3:
14.5 ppb
Range 1-37
                                                                        BS, PM10,
                                                                        NO,, SO,, CO
             Poisson regression using
             APHEA methodology. No
             association was found
             between O3 and circulatory
             diseases in all year analyses.
             Results from acute MI
             suggest potential seasonal
             effect.
0,1, 1-3      Poisson GLM, month dummy
             variables; co-adjustment. No
             O3 or other pollution effects
             on respiratory admissions.
             Significant inverse
             association of O3 with
             cardiac admissions in older
             age group. Very low O3
             concentrations.
8-h avg O3 (per 25 ppb):

Total circulatory:
All year:
0.9726(0.9436,1.0046)

Acute MI:
All year:
0.9825(0.9534,1.0142)
Warm season:
1.0126(0.9560,1.0228)
Cool season:
0.9680(0.9208,1.0202)

24-h avg O3 (per 10 ppb):

Respiratory:
Age <65 years:
Lag 1-3:  0.971(0.885,1.068)
Age 65+ years:
Lag 1-3:  1.009(0.916,1.111)

Cardiovascular:
Age <65 years:
Lag 1-3:  1.041(0.946,1.144)
Age 65+ years:
Lag 1-3:  0.941(0.886,0.999)

-------
Table AX7-4 (cont'd). Effects of O3 on Daily Hospital Admissions
OQ
c
VI
to
o
o












j>
X
^
VO


O
E>
H
6
0
2"
O
H
0
0
H
W
0
^
O
H
W

Reference, Study
Location and Period

Europe (cont'd)
Schouten et al. (1996)
Amsterdam and
Rotterdam,
the Netherlands
1977-1989















Hagen et al. (2000)
Drammen, Norway
Nov 1994-Dec 1997




Oftedal et al. (2003)
Drammen, Norway
1995-2000






Outcomes and Design


Unscheduled total
respiratory, asthma, and
COPD admissions in
all ages. Time-series
study.















Total respiratory
admissions for all ages.
Time-series study.




Admissions for
respiratory disease.
Time-series study.





Copollutants
Mean O3 Levels Considered


8-h max O3: SO2, NO2, BS

Amsterdam:
Summer:
86 ug/m3
5th % to 95th %
28-152
Winter:
53 ug/m3
5th % to 95th %
3-104
Rotterdam:
Summer:
81 ug/m3
5th % to 95th %
25-199
Winter:
45 ug/m3
5th % to 95th %
3-96
24-h avg O3: PM10,NO2,
44.48 ug/m3 SO2, benzene,
SD 18.40 toluene,
IQR 26.29 formaldehyde



24-h avg O3: Benzene,
44.6 ug/m3 formaldehyde,
SD19.2 toluene, PM10,
IQR 26. 9 NO2, SO2




Lag Structure Method, Findings,
Examined Interpretation


0,1,2,0-1, Poisson GLM using APHEA
0-2, 0-3, 0-4, methodology; co-adjustment.
0-5 No consistent O3 effects.
Concern regarding multiple
comparisons.















0 Poisson GAM with partial
splines; co-adjustment.
Single and multipollutant
models evaluated. No O3
effects. Ozone levels low
and cycles may not have
been adequately controlled.
0 Benzene had the strongest
association.






Effects
(Relative Risk and 95% CI)


1-h max O3 (per 100 ug/m3):

Amsterdam and Rotterdam:

Total respiratory, all ages:
Summer:
Lag 2: 1.051(1.029,1.073)
Winter:
Lag 2: 0.976(0.951,1.002)











24-h avg O3
(per 26.29 ug/m3):

LagO: 0.964(0.899-1.033)



24-h avg O3 (per 26.9 ug/m3):

0.996(0.942,1.053)





-------
                                 Table AX7-4 (cont'd). Effects of O3 on Daily Hospital Admissions
OQ
to
O
O
X
Reference, Study
Location and Period
Europe (cont'd)
Ponka and Virtanen
(1996)
Helsinki, Finland
1987-1989



Ballesteretal. (2001)
Valencia, Spain
1994-1996

Latin America
Gouveia and Fletcher
(2000a)
Sao Paulo, Brazil
Nov 1992-Sep 1994





Australia
Morgan etal. (1998a)
Sydney, Australia
1990-1994








Outcomes and Design

Asthma admissions for
age 0-14 years and
15-64 years. Time-
series study.



Emergency total
cardiovascular
admissions for all ages.
Time-series study.

Total respiratory,
pneumonia, and asthma
admissions for age
<5 years. Time-series
study.





Admissions for asthma
(age 1-14 years,
1 5-64 years), COPD
(age 65+ years), and
heart disease (all ages,
0-64 years, 65+ years).
Time-series study.




Mean O3 Levels

O3 (index not
specified):
22 ug/m3




8-h max O3:
23ppb
Range 5-64


l-hmaxO3:
63.4 ug/m3
SD38.1
IQR 50.3






l-hmaxO3:
25ppb
SD13
IQR 11






Copollutants Lag Structure Method, Findings,
Considered Examined Interpretation

TSP, SO2, NO2 0, 1, 2, 3, 4, 5 Poisson GLM using APHEA
methodology. Reported
significant O3 effect for age
0-14 years, but also for
control (digestive disease)
conditions. Ozone levels
very low.
SO2, NO2, 0, 1, 2, 3, 4, 5 Poisson GLM using APHEA
CO, BS methodology. Results
stratified by season. No O3
effects.

PM10, NO2, 0,1,2 Poisson GLM with co-
SO2, CO adjustment using sine/cosine
waves. Significant O3 effects
on total respiratory and
pneumonia admissions.
Ozone effects fairly robust
to NO2 and PM10.



Bscatter, NO2 0,1,2,0-1, Poisson with GEE.
0-2 No significant effects of O3
in single or multipollutant
models.






Effects
(Relative Risk and 95% CI)

Not quantitatively useful.






8-h max O3 (per 5 ppb):

Lag 2: 0.99(0.97-1.01)


l-hmaxO3 (per
119.6 ug/m3):

Total respiratory:
LagO: 1.054(1.003,1.107)
Pneumonia:
LagO: 1.076(1.014,1.142)
Asthma:
Lag 2: 1.011(0.899,1.136)

l-hmaxO3(per28ppb):

Asthma, age 1-14 years:
Lagl: 0.975(0.932,1.019)
Asthma, age 15-64 years:
LagO: 1.025(0.975,1.078)
COPD, age 65+ years:
LagO: 1.010(0.960,1.062)
Heart disease, all ages:
LagO: 1.012(0.990,1.035)

-------
OQ
 to
 o
 o
                                           Table AX7-4 (cont'd). Effects of O3 on Daily Hospital Admissions
 X
Reference, Study
Location and Period
Australia (cont'd)
Petroeschevsky et al.
(2001)
Brisbane, Australia
1987-1994





Asia
Lee et al. (2002)
Seoul, Korea
Dec 1997-Dec 1999




Chang et al. (2005)
Taipei, Taiwan
1997-2001


Outcomes and Design

Unscheduled asthma,
total respiratory and
total cardiovascular
admissions in several
age strata: all ages,
0-4,5-14,15-64,
65+ years.
Time-series study.


Asthma admissions
for age <1 5 years.
Time-series study.




Total cardiovascular
hospital admissions for
all ages. Cool days
(<20 °C) and warm
Copollutants
Mean O3 Levels Considered

l-hmaxO3: Bscatter, SO2,
25.3 ppb NO2
Range 2. 5-107.3

8-h avg O3
(10 a.m.-6 p.m.):
19.0 ppb
Range 1.7-64.7


l-hmaxO3: SO2,NO2,
36.0 ppb CO, PM10
SD 18.6
IQR21.7



24-h avg O3: PM10, SO2,
19.74 ppb N02,CO
IQR 10.87
Range 2. 30-53. 93
Lag Structure Method, Findings,
Examined Interpretation

0, 1 , 2, 3, 0-2, Poisson GLM using APHEA
0-4 co-adjustment methodology.
Results stratified by season.
Ozone significantly related to
asthma and total respiratory
admissions, not for cardiac
admissions. Effects varied by
age group. Ozone effects
robust to copollutants.

0, 1,2, 3, 4, Poisson GAM using default
0-1,1-2,2-3, convergence criteria. Ozone
3-4 associated with asthma
admissions in single- and
two-pollutant models.


0-2 Conditional logistic regression.
All cardiovascular admissions
chosen because similar risks
have been observed for major
Effects
(Relative Risk and 95% CI)

8-h avg O3 (per 10 ppb):

All ages:
Total respiratory:
Lag 2: 1.023(1.003,1.043)
Asthma:
Lag 0-4: 1.090(1.042,1.141)
Total cardiovascular:
Lag 3: 0.987(0.971,1.002)

1-h max O3 (per 21 .7 ppb):

O3 only model:
Lag 1: 1.12(1.07-1.16)

O3 with PM10 model:
Lag 1: 1.10(1.05,1.15)
24-h avg O3 (per 10.87 ppb):

Odds ratios:

                              days (>20 °C) were
                              evaluated.  Case-
                              crossover approach.
subcategories and combining
counts provide greater power.
Subtropical climate in Taipei.
In the analysis of warm days
only, all pollutants except SO2
were associated with
cardiovascular admissions.
Ozone effect slightly
diminished in two-pollutant
model adjusting for PM10.
O3 only models:
Warm: 1.189(1.154,1.225)
Cool:  1.073(1.022,1.127)

O3 with PM10 models:
Warm: 1.066(1.038,1.094)
Cool:  0.980(0.924,1.039)

-------
X
to
Reference, Study
Location and Period
Asia (cont'd)
Yang et al. (2004a)
Kaohsiung, Taiwan
1997-2000








Tsai et al. (2003a)
Kaohsiung, Taiwan
1997-2000















Outcomes and Design

Total cardiovascular
hospital admissions for
all ages. Cool days
(<25 °C) and warm
days (>25 °C) were
evaluated. Case-
crossover approach.




Stroke admissions
(subarachnoid
hemorrhagic stroke,
primary intracerebral
hemorrhage, ischemic
stroke, and others) for
all ages. Cool days
(<20 °C) and warm
days (>20 °C) were
evaluated. Case-
crossover approach.







Copollutants Lag Structure Method, Findings,
Mean O3 Levels Considered Examined Interpretation

24-h avg O3: PM10, SO2, 0-2 Conditional logistic regression.
25.02ppb NO2, CO All pollutants except SO2
IQR 21.20 associated with cardiovascular
Range 1 .25-83.00 admissions on warm days.
On cool days, O3 effect was
diminished. Ozone effect was
robust to adjustment to
copollutants. Results from
tropical city may restrict
generalization to other
locations.
24-h avg O3: PM10, SO2, 0-2 Conditional logistic regression.
25.02ppb NO2, CO Warm day associations were
IQR 21 .20 positive while cool days were
Range 1 .25-83.00 generally negative with some
positive associations. Ozone
effect was robust to adjustment
for SO2 and CO, but not PM10.











Effects
(Relative Risk and 95% CI)

24-h avg O3 (per 21 .20 ppb):

Odds ratios:

O3 only models:
Warm: 1.351(1.279,1.427)
Cool: 1.057(0.962,1.162)

O3 with PM10 models:
Warm: 1.308(1.219,1.404)
Cool: 0.820 (0.732, 0.912)
24-h avg O3 (per 21 .20 ppb):

Odds ratios:

O3 only models:
Primary intracerebral
hemorrhage:
Warm: 1.20(1.06,1.35)
Cool: 0.57(0.24,1.34)
Ischemic stroke:
Warm: 1.15(1.07,1.23)
Cool: 0.88(0.50,1.53)
O3 with PM10 models:
Primary intracerebral
hemorrhage:
Warm: 0.98(0.85,1.14)
Ischemic stroke:
Warm: 0.96(0.88,1.05)

-------
                                            Table AX7-4 (cont'd).  Effects of O3 on Daily Hospital Admissions
uw
to
O
O
Reference, Study
Location and Period

Outcomes and Design

Mean O3 Levels

Copollutants
Considered

Lag Structure
Examined

Method, Findings,
Interpretation

Effects
(Relative Risk and 95% CI)

Asia (cont'd)

Wongetal. (1999a)
Hong Kong
1994-1995
                               Total and cause-
                               specific respiratory and
                               cardiovascular
                               admissions in several
                               age strata: all ages,
                               0-4, 5-64, 65+ years.
                               Time-series study.
8-h max O3:
20.2 ug/m3
Median 24. 15
IQR31.63
N02, S02,
PM10

0,1,2,3,4,5,
0-1,0-2,0-3,
0-4, 0-5

                                                                                                Poisson GLM using APHEA
                                                                                                methodology. Ozone
                                                                                                significantly associated with
                                                                                                total and cause specific
                                                                                                respiratory and cardiac
                                                                                                outcomes. Ozone results
                                                                                                robust to adjustment for high
                                                                                                PM10, but not high NO2.
                                                                                                Effects of O3 persisted in
                                                                                                cold season.
                                                                                                        8-h max O3 (per 10 ug/m3):

                                                                                                        All ages:
                                                                                                        Total respiratory:
                                                                                                        Lag 0-3:  1.022(1.015,1.029)
                                                                                                        Total cardiovascular:
                                                                                                        Lag 0-5:  1.013(1.005,1.021)
X
Wongetal. (1999b)
Hong Kong
Jan 1995-Jun 1997
Cause-specific
cardiovascular
admissions for
>65 years age.
Time-series study.
                                               8-h avg O3:

                                               Warm season:
                                               31.2 ug/m3
                                               Cool season:
                                               34.8 ug/m3
                                                                         NO2, SO2,
                                                                         respirable PM
0,1,2,3,4,5,
 0-1,0-2,0-3,
   0-4, 0-5
GLM with sinusoids;
co-adjustment. Ozone
significantly associated with
total and cause-specific
cardiovascular admissions in
cool season only, when O3
levels are higher in Hong
Kong. Details missing in
brief report.
O3 (per 50 ug/m3):

O3 with NO2 models:
Total cardiovascular:
All year:
Lag 0-1:  1.03(1.00,1.07)
Warm season:
Lag 0-1:  1.01 (0.95,1.07)
Cool season:
Lag 0-1:  1.08(1.02,1.14)

-------
                                                  Table AX7-5. Effects of Acute O3 Exposure on Mortality
uw
to
O
O
Reference, Study
Location and Period

Outcome Measure Mean O3 Levels

Copollutants
Considered

Lag Structure
Reported

Method/Design

Effect Estimates

X
Meta-analysis

Bell et al. (2005)
Various U.S. and
non-U.S. cities
Varying study periods
                                   All cause;
                                   cardiovascular;
                                   respiratory; all
                                   ages; age 64+ or
                                   65+ years.
Not relevant.
Various PM
indices
0,1,2, or 0-1
Meta-analysis.
Bayesian hierarchical
model; included up to
144 estimates from
39 studies.

Risk estimates obtained
for yearly data versus
warm season; cause-
specific; PM
adjustment; U.S. versus
non-U.S.;various lags;
and GAM versus non-
GAM.

Comparisons with
NMMAPS estimates
(Bell etal., 2004).
24-h avg O3 (per 10 ppb):

Posterior means:

All cause:
All year: 0.87% (0.55,1.18)
Warm:  1.50% (0.72, 2.29)
Cardiovascular:
All year: 1.11% (0.68,1.53)
Warm: 2.45% (0.88, 4.10)
Respiratory:
All year: 0.47% (-0.51,
1.47)

O3 with PM model:
All cause:
All year: 0.97% (-0.03,
1.98)

Meta-analysis results were
consistently larger than those
from NMMAPS. In addition,
heterogeneity of city-specific
estimates
in the meta-analysis were
larger compared to
NMMAPS. These findings
indicate possible publication
bias.

-------
                                             Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
uw
to
O
O
Reference, Study
Location and Period

Outcome Measure Mean O3 Levels

Copollutants
Considered

Lag Structure
Reported

Method/Design

Effect Estimates

X
Meta-analysis (cont'd)

Ito et al. (2005)
Various U.S. and
non-U.S. cities
Varying study periods
                                   All cause
Not relevant.
Various PM
indices
 Reported lags
up to 3 day lags
Meta-analysis.
DerSimonian-Laird
approach; included up
to 43 estimates from
38 studies.

Risk estimates obtained
for yearly data versus
warm season; PM
adjustment; correction
for asymmetry in
funnel plot; and GAM
versus non-GAM.

Seven U.S. cities time-
series study with
various sensitivity
analyses.
24-h avg O3 (per 20 ppb):

O3 only model:
All year: 1.6% (1.1, 2.0)
Warm: 3.5% (2.1, 4.9)

O3 with PM model:
All year: 1.5% (0.8, 2.2)

Non-GAM-affected:
All year: 1.4% (0.8, 2.0)
GAM-affected:
All year: 1.9% (1.0, 2.8)

Correction for funnel plot
asymmetry:
All year: 1.4% (0.9, 1.9)

Seven U.S. cities analysis
found that including PM
in the model did not
substantially reduce the O3
risk estimates. However,
differences in the weather
adjustment model could
result in a two-fold difference
in risk estimates.

-------
                                             Table AX7-5 (cont'd).  Effects of Acute O3 Exposure on Mortality
uw
to
O
O
Reference, Study
Location and Period

Outcome Measure Mean O3 Levels

Copollutants
Considered

Lag Structure
Reported

Method/Design

Effect Estimates

X
Oi
Meta-analysis (cont'd)

Levy et al. (2005)
Various U.S., Canadian
and European cities
Varying study periods
                                   All cause
Not relevant.
PM10,PM25, SO2,
N02,CO
0, 1-2        Meta-analysis.
             Empirical Bayes
             metaregression;
             included up to
             48 estimates from
             28 studies.

             Risk estimates obtained
             by season, for
             copollutant adjustment,
             North America versus
             Europe; various lags;
             temperature
             adjustment; GAM
             versus non-GAM;
             annual mean O3; and
             total deaths.

             Examined relationship
             between O3
             personal exposure and
             ambient concentrations
             using cooling
             degree days and
             residential central air-
             conditioning
             prevalence.
1-h max O3 (per 10 ug/m3):

O3 only model:
All year:  0.21% (0.16, 0.26)
Warm: 0.43% (0.29, 0.56)
Cool: -0.02 (-0.17, 0.14)

Non-GAM-affected:
All year:  0.23% (0.15, 0.31)
GAM-affected:
All year:  0.20% (0.13, 0.27)

Nonlinear specification of
temperature:
All year:  0.20% (0.13, 0.27)
Linear specification of
temperature:
All year:  0.06 (-0.03, 0.16)

In the metaregression,
air-conditioning prevalence
and lag time were the
strongest predictors
of between-study variability.

-------
OQ
 to
 o
 o
                                            Table AX7-5 (cont'd).  Effects of Acute O3 Exposure on Mortality
        Reference, Study
        Location and Period
Outcome Measure     Mean O3 Levels
  Copollutants
   Considered
Lag Structure
  Reported
   Method/Design
     Effect Estimates
 X
 H
 6
 o

 o
 H
O
 O
 H
 W
 O
 O
 HH
 H
 W
United States

Bell et al. (2004)
95 U.S. communities
1987-2000
                                   All cause;
                                   cardiopulmonary;
                                   all ages; age
                                   <65 years;
                                   age 65-74 years;
                                   age  >75 years
                   24-h avg O3:
                   26 ppb across the 95
                   cities.

                   No individual city
                   data provided.
PM10, PM25; two-
pollutant models
 0,1,2,3,0-6
Poisson GLM;
Bayesian hierarchical
model. Time-series
study.
24-h avg O3 (per 10 ppb):

Posterior means:

All cause, all ages:
All available data:
LagO:  0.25% (0.12, 0.39)
Lag 0-6: 0.52% (0.27, 0.77)
Warm season:
LagO:  0.22% (0.08, 0.38)
Lag 0-6: 0.39% (0.13, 0.65)

All cause, all available data:
Age <65 years:
Lag 0-6: 0.50% (0.10, 0.92)
Age 65-74 years:
Lag 0-6: 0.70% (0.28, 1.12)
Age  >75 year:
Lag 0-6: 0.52% (0.18, 0.87)

Cardiopulmonary, all ages:
All available data:
Lag 0-6: 0.64% (0.31, 0.98)

-------
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
uw
to
o
O













X
oo


M
^^
>
H
M
^^
O

O
H
0
o
H
W
O
O
H
W
Reference, Study
Location and Period

United States (cont'd)
Samet et al. (2000;
reanalysis Dominici et al.,
2003)
90 U.S. cities (80 U.S.
cities with O3 data)
1987-1994







Huang et al. (2005)
19 large U.S. cities
Jun-Sept 1987-1 994


















Outcome Measure Mean O3 Levels


All cause; 24-h avg O3:
cardiopulmonary
Mean range:
Approximately
12 ppb
(Des Moines, IA)
to 36 ppb
(San Bernardino,
CA)




Cardiopulmonary 24-h avg O3:

Mean range:
Approximately
18 ppb (Oakland,
CA) to 56 ppb
(San Bernadino,
CA)

Daily range across
19 U.S. cities:
0-1 00 ppb








Copollutants Lag Structure
Considered Reported Method/Design Effect Estimates


PM10, NO2, SO2, 0,1,2 Poisson GAM, 24-h avg O3 (per 1 0 ppb):
CO; two-pollutant reanalyzed with
models stringent convergence Posterior means:
criteria; Poisson GLM.
Time-series study. All cause:

All available data:
Lagl: 0.19% (0.03, 0.35)

Summer:
Lagl: 0.51% (0.23, 0.78)
Winter:
Lagl: -0.53% (-1.10, 0.05)
PM10, PM2 5; two- 0, 1 , 2, 0-6 Poisson GLM; 24-h avg O3 (per 1 0 ppb):
pollutant models Bayesian hierarchical
model. Time-series Posterior means:
study.
Single-day lag models:
O3 only model:
LagO: 0.73% (0.27, 1.19)
O3 with PM10 model:
LagO: 0.74% (-0.33, 1.72)

Cumulative lag models:
O3 only model:
Lag 0-6: 1.25% (0.47, 2.03)
Model adjusted for heat
waves:
Lag 0-6: 1.11%(0.38, 1.86)





-------
                                            Table AX7-5 (cont'd).  Effects of Acute O3 Exposure on Mortality
X
vo
Reference, Study
Location and Period
United States (cont'd)
Schwartz (2005)
14 U.S. cities
1986-1993







Copollutants
Outcome Measure Mean O3 Levels Considered

All cause 1-h max O3: PM10; two-
pollutant models
Median range:
35.1 ppb (Chicago,
IL) to 60.0 ppb
(Provo, UT)




Lag Structure
Reported Method/Design

0 Case-crossover
analysis; conditional
logical regression
controlled for
temperature using
nonlinear regression
splines and matching



Effect Estimates

1-h max O3 (10 ppb):

Analysis with temperature
regression splines:
All year:
0.19% (0.03, 0.35)
Warm season:
0.26% (0.07, 0.44)
Cold season:
0% (-0.27, 0.27)
                                                                                                    Analysis with temperature
                                                                                                    matched controls:
                                                                                                    All year:
                                                                                                    0.23% (0.01, 0.44)
                                                                                                    Warm season:
                                                                                                    0.37% (0.11,0.62)
                                                                                                    Cold season:
                                                                                                    -0.13% (-0.53, 0.28)
        Kinney and Ozkaynak
        (1991)
        Los Angeles County, CA
        1970-1979
All cause;
respiratory;
circulatory
        Kinney etal. (1995)
        Los Angeles County, CA
        1985-1990
All cause
1-h max total
oxidants (Ox):
75 ppb
SD45
1-h max O3:
70 ppb
SD41
KM (particle
optical
reflectance), NO2,
SO2, CO;
multipollutant
models
PM10, NO2, SO2,
CO; two-pollutant
models
OLS (ordinary least
squares) on high-pass
filtered variables.
Time-series study.
Linear, log-linear, and
Poisson. Time-series
study.
All cause:
Multipollutant model:
Slope estimate:
0.030 deaths/ppb (SE 0.009),
p = 0.0005

Cardiovascular:
O3 only model:
Slope estimate:
0.023 deaths/ppb (SE 0.006),
p< 0.0001

1-h max O3 (per 143 ppb):

O3 only model:
2% (0, 5)
O3 with PM10 model:

-------
g>
CtQ
«J
r+
to
o
o
(^











.
X
o



O
>
hH

o

O
H
0
c|
o
H
W
O
H
W

Reference, Study
Location and Period

United States (cont'd)
Ostro(1995)
San Bernardino County and
Riverside County, CA
1980-1986
Ostro (2000)
Coachella Valley, CA
1989-1998






Fairley (1999; reanalysis
Fairley, 2003)
Santa Clara County, CA
1989-1996















Table AX7-5 (cont'd).

Outcome Measure Mean O3 Levels


All cause 1 -h max O3:
140 ppb
Range 20-370

All cause; 1 -h max O3:
respiratory;
cardiovascular Palm Springs:
67 ppb
Range 0-1 90

Indio:
62 ppb
Range 0-1 80
All cause; 8-h max O3:
respiratory; 29 ppb
cardiovascular SD 15

24-h avg O3:
16 ppb
SD9

O3 ppb-hours
>60 ppb:
Levels not reported.








Effects of Acute O3 Exposure on Mortality
Copollutants Lag Structure
Considered Reported Method/Design


PM2 5 0 Autoregressive linear;
Poisson. Time-series
study.

PM10, PM2 5, 0 Poisson GAM with
PM10_2 5, NO2, CO default convergence
criteria. Time-series
study.





PM10, PM25, 0 Poisson GAM,
PM10_25, SO42~, reanalyzed with
coefficient of stringent convergence
haze, NO3", NO2, criteria; Poisson GLM.
SO2; two-pollutant Time-series study.
models















Effect Estimates


1 -h max O3 (per 100 ppb):

Warm season:
2.0% (0.0, 5.0)
1-h max O3 (per 40 ppb):

All cause:
-l%(-4, 3)
Respiratory:
3% (-9, 16)
Cardiovascular:
-4% (-9,1)

GAM with stringent
convergence criteria:

All cause:
8-h max O3 (per 33 ppb):
3.1% (-0.3, 6. 6)
O3 ppb-hours >60 ppb
(increment not reported):
3.8% (1.4, 6.3)

Cardiovascular:
8-h max O3 (per 33 ppb):
2.6% (-2.3, 7.8)
O3 ppb-hours >60 ppb
(increment not reported):
4. 3% (0.4, 8.3)




-------
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
uw
to
o
O













^
X

 0.05)

Winter:
24-h avg O3 (per 7.7
All cause:
2.4% (p > 0.05)
Cardiovascular:
1.5%(p>0.05)
24-h avg O3 (per 20

St. Louis, MO:
0.6% (t = 0.38)
Eastern Tennessee:
-1.3%(t=-0.37)











7 ppb):






1 ppb):






ppb):




ug/m3):














-------
OQ
 to
 o
 o
                                         Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
 X
 to
H
6
o

o
H
O
O
H
W
O
O
HH
H
W
Reference, Study
Location and Period
United States (cont'd)
Ito and Thurston (1996)
Cook County, IL
1985-1990








Moolgavkar (2003)
Cook County, IL and Los
Angeles County, CA
1987-1995






Outcome Measure

All cause;
respiratory;
circulatory; cancer;
race/gender
subcategories






All cause;
cardiovascular








Copollutants Lag Structure
Mean O3 Levels Considered Reported Method/Design

l-hmaxO3: PM10, NO2, SO2, 0-1 Poisson GLM. Time-
38.1 ppb CO; two-pollutant series study.
SD 19.9 models








24-h avg O3: PM2 5, PM10, NO2, 0,1,2,3,4,5 Poisson GAM with
SO2, CO; two- default convergence
Cook County: pollutant models criteria. Time-series
Median 18 ppb study.
Range 0.2-67

Los Angeles
County:
Median 24 ppb
Range 0.6-77
Effect Estimates

1-h max O3 (per 100 ppb):

All cause:
O3 only model:
10% (6, 15)
O3 with PM10 model:
7% (1,12)
Circulatory (results given in
graphic format):
O3 only model:
12% (6, 20)
24-h avg O3 (per 100 ppb):

All cause, all year:
Cook County:
LagO: 1.4% (t= 6.3)
Los Angeles County
LagO: 0.4% (t = 2.3)

All cause, summer:
Cook County:
                                                                                                                           LagO: 2.9%(t= 7.2)
                                                                                                                           Los Angeles County
                                                                                                                           LagO: 1.0% (t = 2.8)

                                                                                                                           Cardiovascular, all year:
                                                                                                                           Cook County:
                                                                                                                           LagO: 1.8%(t=5.5)
                                                                                                                           Los Angeles County
                                                                                                                           LagO: 0.2%(t= 0.9)

                                                                                                                           Cardiovascular, summer:
                                                                                                                           Cook County:
                                                                                                                           LagO: 3.3%(t= 5.6)
                                                                                                                           Los Angeles County
                                                                                                                           LagO: 0.8%(t= 1.7)

-------
X
Reference, Study
Location and Period
United States (cont'd)
Moolgavkar (2000)
Cook County, IL; Los
Angeles County, CA; and
Maricopa County, AZ
1987-1995









Lippmann et al. (2000;
reanalysis Ito, 2003)
Detroit, MI
1985-1990
1992-1994
Outcome Measure

Cardiovascular;
cerebro vascular;
COPD











All cause;
respiratory;
circulatory;
cause-specific

Mean O3 Levels

24-h avg O3:

Cook County:
Median 1 8 ppb
Range 0.2-67

Los Angeles
County:
Median 24 ppb
Range 0.6-77

Maricopa County:
Median 25 ppb
Range 1-50
24-h avg O3:
25 ppb
Maximum 55


Copollutants
Considered

PM25, PM10, NO2,
SO2, CO; two- and
three-pollutant
models










PM10,PM25,
PM10.25,S042-,H+,
N02, S02, CO;
two-pollutant
models
Lag Structure
Reported Method/Design

0, 1, 2, 3, 4, 5 Poisson GAM with
default convergence
criteria. Time-series
study.










0, 1, 2, 3, 0-1, Poisson GAM,
0-2, 0-3 reanalyzed with
stringent convergence
criteria; Poisson GLM.
Time-series study.
Effect Estimates

24-h avg O3 (per 100 ppb):

Cook County:
Cardiovascular, all year:
LagO: 1.51% (0.78, 2.24)
COPD, all year:
LagO: 1.53% (-0.49, 3.55)

Los Angeles County and
Maricopa County:
O3 results not presented.
Noted as negative or small
and insignificant in all year
and warm season analyses.
GAM with stringent
convergence criteria:

All lags and outcomes during
both study periods (n = 140):
                                                                                                                                          24-h avg O3 (per 5th to
                                                                                                                                          95th % increment):
                                                                                                                                          Median 1.6%
                                                                                                                                          Range-1.8-2.6

                                                                                                                                          1985-1990:
                                                                                                                                          24-h avg O3 (per 36 ppb):
                                                                                                                                          All cause:
                                                                                                                                          LagO: 1.08% (-1.08, 3.30)
                                                                                                                                          Circulatory:
                                                                                                                                          LagO: 1.84% (-1.26, 5.04)

                                                                                                                                          1992-1994
                                                                                                                                          24-h avg O3 (per 28 ppb):
                                                                                                                                          All cause:
                                                                                                                                          LagO: 2.58% (-2.41, 7.82)
                                                                                                                                          Circulatory:
                                                                                                                                          LagO: 2.13% (-5.04, 9.85)

-------
OQ
 to
 o
 o
                                         Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
 X
Reference, Study
Location and Period
United States (cont'd)
Lipfert et al. (2000a)
Seven counties in
Philadelphia, PA area
1991-1995




Outcome Measure

All cause;
respiratory;
cardiovascular;
all ages; age
65+ years; age
<65 years; various
subregional
boundaries
Mean O3 Levels

1-hmax O3:
44.76 ppb
SD 25.68

24-havg O3:
23. 44 ppb
SD 13.86

Copollutants
Considered

PM10,PM25,
PM10.25,S042-,
NO3-, other PM
indices, NO2, SO2,
CO; two-pollutant
models


Lag Structure
Reported Method/Design

0-1 Linear with 1 9-day
weighted average
Shumway filters.
Time-series study.




Effect Estimates

1-h max O3 (per 45 ppb less
background, not reported):

All cause, all ages:
O3 only model:
3.19%, p< 0.055
O3 with PM2 5 model:
3. 34%, p< 0.055
        Moolgavkar et al. (1995)
        Philadelphia, PA
All cause
24-h avg O3:
TSP, SO2;
multipollutant
Poisson; GEE and
nonparametric
Cardiovascular, all ages:
O3 only model:
3.98%, p< 0.055
O3 with PM2 5 model:
5.35%, p< 0.055

24-h avg O3 (per 100 ppb):
1973-1988

O
^
^
H
6
n
s.**
o
H
0
0
H
W
0


O
H
W
Spring: models
25. 9 ppb
Range 2.9-74.0

Summer:
35.5 ppb
Range 1.3-159.0

Fall:
16. 2 ppb
Range 0.2-63.8
Winter:
11. 9 ppb
Range 0.0-32.9





bootstrap methods. O3 with TSP and SO2 models:
Time-series study. Spring:
2.0% (-6.7, 11.5)
Summer:
14.9% (6.8, 23.6)
Fall:
-4. 5% (-13. 9, 5.9)
Winter:
0.4% (-15. 6, 19.4)











-------
                                            Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
OQ
to
O
o
Reference, Study
Location and Period
TTnitprl Statp* front' rh
Copollutants
Outcome Measure Mean O3 Levels Considered

Lag Structure
Reported

Method/Design

Effect Estimates

X
        Chock et al. (2000)
        Pittsburgh, PA
        1989-1991
All cause; age
<74 years;
age 75+ years
        De Leon et al. (2003)
        New York City
        1985-1994
Circulatory and
cancer with and
without
contributing
respiratory causes
l-hmaxO3:
Levels not reported.
PM10, NO2, SO2,
CO; two-, five-,
and six-pollutant
models
            PoissonGLM. Time-
            series study.
24-h avg O3:
21.59 ppb
5th %-95th %
7.00-44.97
PM10, N02, S02,
CO; two-pollutant
models
0 or 1       Poisson GAM with
            stringent convergence
            criteria; Poisson GLM.
            Time-series study.
1-h max O3 (per 40 ppb):

Age <74 years:
O3 only model:
-1.5% (t=-0.68)
O3 with PM10 model:
-2.0% (t=-0.93)

Age 75+ years:
O3 only model:
-1.8% (t=-0.82)
O3 with PM10 model:
-2.2% (t=-0.98)
Quantitative results not given.

Circulatory deaths:
Larger O3 effect estimates
with contributing respiratory
causes than without
(RR nonsignificant).

Cancer deaths:
Smaller O3 effect estimates
with contributing respiratory
causes than without
(RR nonsignificant).

-------
Table AX7-5 (cont'd).  Effects of Acute O3 Exposure on Mortality
OQ

to
O
O











X
^j
ON


O
^
^TJ
H
6
0
2|
O

O
0
H

O
o
HH
H
W

Reference, Study
Location and Period

United States (cont'd)
Klemm and Mason (2000);
Klemm et al. (2004)
Atlanta, GA
Aug 1998- July 2000






Canada
Vedal et al. (2003)
Vancouver, British
Columbia, Canada
1994-1996






Villeneuve et al. (2003)
Vancouver, British
Columbia, Canada
1986-1999









Copollutants
Outcome Measure Mean O3 Levels Considered


All cause; 8-h max O3: PM2 5, PM10_2 5,
respiratory; 47.03 ppb EC, OC, NO2,
cardiovascular; SD 24.71 SO42~,
cancer; other; age NO3', SO2, CO
<65 years; age
65+ years





All cause; 1 -h max O3: PM10, NO2, SO2,
respiratory; 27.3 ppb CO
cardiovascular SD 10.2







All cause; 24-h avg O3: PM25, PM10,
respiratory; 13.4 ug/m3 PM25.10, TSP,
cardiovascular; Range 0.6-38.6 coefficient of
cancer; haze, SO42~, SO2,
socioeconomic NO2, CO
status







Lag Structure
Reported Method/Design


0- 1 Poisson GLM using
quarterly, monthly, or
biweekly knots for
temporal smoothing.
Time-series study.






0,1,2 Poisson GAM with
stringent convergence
criteria. Time-series
study.






0,1,0-2 Poisson GLM with
natural splines. Time-
series study.











Effect Estimates


All cause, age 65+ years:

Quarterly knots:
Slope estimate:
0.00079 (SE 0.00099),
t=0.80
Monthly knots:
Slope estimate:
0.00136 (SE 0.001 11),
t=1.22

1-h max O3 (per 10.2 ppb):

Summer:
All cause:
LagO: 4.0% (1.4, 6.7)
Respiratory:
LagO: 1.5% (-6.6, 9.6)
Cardiovascular:
LagO: 3.9% (-0.3, 8.0)

24-h avg O3 (per 21 .3 ug/m3):

All year:
All cause:
LagO: 1.4% (-0.9, 3.6)
Respiratory:
LagO: 1.6% (-4. 5, 8.1)
Cardiovascular:
LagO: 0.7% (-2.7, 4.3)
Cancer:
LagO: 2.6% (-1.2, 6.5)

                                                                                No effect modification of O3-
                                                                                mortality effects by
                                                                                socioeconomic status.

-------
OQ
to
o
o
                                  Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
X
H

6
o


o
H

O

O
H
W

O


O
HH
H
W
Reference, Study
Location and Period
Canada (cont'd)
Goldberg et al. (2003)
Montreal, Quebec, Canada
1984-1993








Goldberg et al. (2001)
Montreal, Quebec, Canada
1984-1993













Outcome Measure Mean O3 Levels

Congestive heart 24-h avg O3:
failure as 29.0 |ig/m3
underlying cause of SD 17.1
death versus those
classified as having
congestive heart
failure one year
prior to death



All cause; cause- 24-h avg O3:
specific; all ages; 29.0 |ig/m3
age <65 years; age SD 17.1
>65 years












Copollutants Lag Structure
Considered Reported Method/Design Effect Estimates

PM2 5, coefficient 0,1,0-2 Poisson GLM with 24-h avg O3 (per 21.3 |ig/m3):
of haze, SO42~, natural splines. Time-
SO2, NO2, CO series study. Congestive heart failure as
underlying cause of death:
Lag 0-2: 4. 54% (-5.64,
15.81)

Having congestive heart
failure one year prior to
death:
Lag 0-2: 2.34% (-1.78, 6.63)
PM25, coefficient 0, 1, 0-2 Poisson GAM with 24-h avg O3 (per 21 .3 |ig/m3):
of haze, SO2, NO2, default convergence
NO, CO criteria. Time-series All cause, all year:
study. All ages:
Lag 0-2: 2.26% (1.23, 3.29)
Age <65 years:
Lag 0-2: 0.1 8% (-1.79, 2.20)
Age >65 years:
Lag 0-2: 2. 84% (1.66, 4.04)
Cardiovascular, all year:
All ages:
Lag 0-2: 3.00% (1.44, 4.59)
Age <65 years:
Lag 0-2: 1.33% (-2.30, 5.09)
Age >65 years:
Lag 0-2: 3. 33% (1.62, 5.08)

-------
X
oo
Reference, Study
Location and Period
Europe
Gryparis et al. (2004)
23 European cities
Study periods vary by city,
ranging from 1990-1997















Copollutants Lag Structure
Outcome Measure Mean O3 Levels Considered Reported Method/Design

All cause; l-hmaxO3: PM10, NO2, SO2, 0-1 Poisson GAM with
respiratory; Median range: CO; two-pollutant stringent convergence
cardiovascular models criteria; Bayesian
Summer: hierarchical model.
44 ppb (Tel Aviv, Time-series study.
Israel) to 117 ppb
(Torino, Italy)
Winter:
1 1 ppb (Milan,
Italy) to 57 ppb
(Athens, Greece)

8-h max O3:
Median range:

Summer:
30 ppb (Rome,
Italy) to 99 ppb
(Torino, Italy)
Effect Estimates

8-h max O3 (per 10 ug/m3):

Weighted mean effect across
21 cities with 8-h max O3
concentrations:

Random effects model:
All cause:
All year:
0.03% (-0.18, 0.21)
Summer:
O3 only model:
0.31% (0.17, 0.52)
O3 with PM10 model:
0.27% (0.08, 0.49)
Winter:
O3 only model:
0.12% (-0.12, 0.37)
O3 with PM10 model:
                                                        Winter:
                                                        8 ppb (Milan, Italy)
                                                        to 49 ppb (Budapest,
                                                        Hungary)
0.22% (-0.08, 0.51)

Respiratory:
Summer:
1.13% (0.74, 0.51)
Winter:
0.26% (-0.50, 0.84)

Cardiovascular:
Summer:
0.46% (0.22, 0.73)
Winter:
0.07% (-0.28, 0.41)

-------
g>
OQ
v>
r+
to
o
O

-------
                                            Table AX7-5 (cont'd).  Effects of Acute O3 Exposure on Mortality
uw
to
O
O
Reference, Study
Location and Period

Outcome Measure Mean O3 Levels

Copollutants
Considered

Lag Structure
Reported

Method/Design

Effect Estimates

X
oo
o
Europe (cont'd)

Anderson et al. (1996)
London, England
1987-1992
                                   All cause;
                                   respiratory;
                                   cardiovascular
                    1-hmax O3:
                    20.6 ppb
                    SD 13.2
                                                       8-h avg O3 (9 a.m.-
                                                       5p.m.):  15.5 ppb
                                                       SD 10.9
                     BS, NO2, SO2;
                     two-pollutant
                     models
                                     Poisson GLM. Time-
                                     series study.
                                        All year:
                                        8-h avg O3 (per 26 ppb):
                                        All cause:
                                        2.43% (1.11, 3.76)
                                        Respiratory:
                                        6.03% (2.22, 9.99)
                                        Cardiovascular:
                                        1.44o/0 (-0.45, 3.36)

                                        Warm season:
                                        8-h avg O3 (per 29 ppb):
                                        All cause:
                                        3.48% (1.73, 5.26)
                                        Respiratory:
                                        5.41% (0.35, 10.73)
                                        Cardiovascular:
                                        3. 55% (1.04, 6.13)

                                        Cool season:
                                        8-h avg O3 (per 20 ppb):
                                        All cause:
                                        0.77% (-0.88, 2.44)
                                        Respiratory:
                                        6.20(1.67,10.94)
                                        Cardiovascular:
                                        -1.69% (-3.99, 0.68)
        Bremneretal. (1999)
        London, England
        1992-1994
All cause;
respiratory;
cardiovascular; all
cancer; all others;
all ages; age
specific (0-64, 65+,
65-74, 75+ years)
22.6 ppb
SD 13.4
                                                       17.5 ppb
                                                       SD11.5
BS, PM10, NO2,
SO2, CO; two-
pollutant models
 Selected best
fromO, 1,2, 3,
  (all cause);
 0,1,2,3,0-1,
   0-2, 0-3
 (respiratory,
cardiovascular)
                                                                                                        Poisson GLM. Time-
                                                                                                        series study.
8-h max O3 (per 26 ppb):

All ages:
All cause:
Lag 2:  -0.7% (-2. 3, 0.9)
Respiratory:
Lag 2:  -3.6% (-7.7, 0.8)
Cardiovascular:
Lag 2:  3. 5% (0.5, 6.7)

-------
                                 Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
X
oo
Reference, Study
Location and Period
Europe (cont'd)
Anderson et al. (2001)
West Midlands region,
England
1994-1996





Prescott et al. (1998)
Edinburgh, Scotland
1992-1995





Le Tertre et al. (2002a)
Le Havre, Lyon, Paris,
Rouen, Strasbourg, and
Toulouse, France
Study periods vary by city,
ranging from 1990-1995








Outcome Measure

All cause;
respiratory;
cardiovascular; all
ages; age 0-14
years; age 15-64
years; age 65+
years


All cause;
respiratory;
cardiovascular; all
ages; age <65
years; age >65
years


All cause;
respiratory;
cardiovascular











Copollutants Lag Structure
Mean O3 Levels Considered Reported Method/Design

8-h max O,,: PM10, PM25, 0-1 Poisson GAM with
24.0 ppb PM2 5.10, BS, SO/', default convergence
SD13.8 NO2, SO2,CO criteria. Time-series
study.





24-havgO3: BS,PM10,NO2, 0 Poisson GLM. Time-
14.5 ppb SO2, CO; two- series study.
SD 2 . 3 pollutant models





8-hmaxO3: BS,NO2, SO2 0-1 Poisson GAM with
default convergence
Le Havre: criteria. Time-series
43.4 ug/m3 study.
Lyon:
52.0 ug/m3
Paris:
26.0 ug/m3
Rouen:
57.9 ug/m3
Strasbourg:
37.0 ug/m3
Toulouse:
68.0 ug/m3
Effect Estimates

8-h max O3 (per 30.8 ppb):

All ages:
All cause:
2. 9% (-0.1, 6.0)
Respiratory:
2.2% (-5.4, 10.4)
Cardiovascular:
0.9% (-3.4, 5.4)
24-havgO3(perlOppb):

All cause, all ages:
-4.2% (-8.1, -0.1)

Cardiovascular, age >65
years:
2.2% (-5. 1,10. 3)
8-h max O3 (per 50 ug/m3):

Six-city pooled estimates:

All cause:
2.7% (1.3, 4.1)
Respiratory:
0.8% (-4.8, 6.2)
Cardiovascular:
2.4% (-0.3, 5.1)





-------
X
oo
to
Reference, Study
Location and Period
Europe (cont'd)
Dab etal. (1996)
Paris, France
1987-1992






Zmirou etal. (1996)
Lyon, France
1985-1990





Sartor etal. (1995)
Belgium
Summer 1994














Copollutants
Outcome Measure Mean O3 Levels Considered

Respiratory 1 -h max O3: BS, PM13, NO2,
43.9 ug/m3 S02, CO
5th %-99th %
6.0-147.0

8-h max O3:
27.7 ug/m3
5th %-99th %
3-110
All cause; 1 -h max O3: PM13, SO2, NO2
respiratory; 15.23 ug/m3
cardiovascular; Range 0-1 52
digestive
8-h avg O3
(9 a.m.-5 p.m.):
9.94 ug/m3
Range 0-78. 92
All cause; age 24-h avg O3: TSP, NO, NO2,
<65 years; Geometric mean: SO2
age 65+ years
During heat wave
(42 day period):
72.4 ug/m3
Range 34. 5-1 11. 5

Before heat wave
(43 day period):
52.4 ug/m3
Range 30.7-92.0

After heat wave
(39 day period):
38.6 ug/m3
Range 18.8-64.9
Lag Structure
Reported Method/Design Effect Estimates

0 Poisson autoregressive. 1-h max O3 (per 100 ug/m3):
Time-series study. 1 .074 (0.934, 1 .235)

8-h max O3 (per 100 ug/m3):
1.040(0.934,1.157)




Selected best Poisson GLM. Time- 8-h avg O3 (per 50 ug/m3):
from 0, 1, 2, 3 series study.
All cause:
LagO: 3% (-5, 12)
Respiratory:
Lagl: l%(-8, 10)
Cardiovascular:
Lagl: 0%(-11,12)
0,1,2 Log-linear regression. No individual regression
Time-series study. coefficient for O3 alone;
interaction with temperature
suggested.

24-h avg O3 (from 18.8 to
111.5 ug/m3) and temperature
(from 10.0 to 27.5°C):

Age <65 years:
Lag 1 : 16% increase in
mortality (5.3% expected)
Age 65+ years:
Lagl: 36.5% increase in
mortality (4% expected)



-------
                                            Table AX7-5 (cont'd).  Effects of Acute O3 Exposure on Mortality
X
oo
Reference, Study
Location and Period
Europe (cont'd)
Hoek et al, (2000;
reanalysis Hoek, 2003)
The Netherlands: entire
country, four urban areas
1986-1994

Hoek etal. (2001;
reanalysis Hoek, 2003)
Outcome Measure

All cause; COPD;
pneumonia;
cardiovascular

Total
cardiovascular;
Mean O3 Levels

8-h avg O3( 12p.m.-
8p.m.):
Median: 47 ug/m3
Range 1-226

8-h avg O3( 12p.m.-
8p.m.):
Copollutants
Considered

PM10, BS, SO42~,
N03-, N02, S02,
CO; two-pollutant
models

PM10, NO2, SO2,
CO
Lag Structure
Reported Method/Design

1,0-6 Poisson GAM,
reanalyzed with
stringent convergence
criteria; Poisson GLM.
Time-series study.

1 Poisson GAM,
reanalyzed with
Effect Estimates

GLM:
All cause:
8-h avg O3 (per 150 ug/m3):
Lagl: 4.3% (2.4, 6.2)
8-h avg O3 (per 120 ug/m3):
Lag 0-6: 5.9% (3.1, 8.7)
8-h avg O3(per 150 ug/m3):
The Netherlands
1986-1994
myocardial
infarction;
arrhythmia; heart
failure;
cerebro vascular;
thrombosis-related
Median: 47 ug/m3
Range 1-226
stringent convergence
criteria; Poisson GLM.
Time-series study.
        Roemer and van Wijinen
        (2001)
        Amsterdam, the
        Netherlands
        1987-1998
                          All cause
                    8-h max O3:

                    Background sites:
                    43 ug/m3
                    Maximum 221

                    Traffic sites:
                    36 ug/m3
                    Maximum 213
                    BS, PM10, NO2,
                    SO2, CO
1 , 2, 0-6
Poisson GAM with
default convergence
criteria (only one
smoother). Time-series
study.
GLM:

Total cardiovascular:
6.2% (3.3, 9.2)
Myocardial infarction:
4.3% (0.1, 8.6)
Arrhythmia:
11.4% (-1.2, 25.5)
Heart failure:
10.2% (1.2, 19.9)
Cerebrovascular:
9.1% (2.9, 15.7)
Thrombo sis-related:
16.6% (2.8, 32.2)

8-h max O3 (per 100 ug/m3):

Total population using
background sites:
Lagl:  -0.3%(-4.1,3.7)
Total population using traffic
sites:
Lagl:  0.2% (-3.6, 4.2)

-------
                                  Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
OQ
to
o
o
X
oo
H

6
o


o
H

O

O
H
W

O


O
HH
H
W
Reference, Study
Location and Period
Europe (cont'd)
Verhoeffetal. (1996)
Amsterdam, the
Netherlands
1986-1992
Peters et al. (2000b)
NE Bavaria, Germany
1982-1994
Coal basin in Czech
Republic
1993-1994


Ponkaetal. (1998)
Helsinki, Finland
1987-1993


Outcome Measure

All cause; all ages;
age 65+ years

All cause;
respiratory;
cardiovascular;
cancer


All cause;
cardiovascular;
age <65 years,
age 65+ years


Copollutants Lag Structure
Mean O3 Levels Considered Reported Method/Design

l-hmaxO3: PM10, NO2, SO2, 0,1,2 Poisson. Time-series
43 ug/m3 CO; multipollutant study.
Maximum 301 models

24-havgO3: TSP, PM10, NO2, 0,1,2,3 Poisson GLM.
SO2, CO Time-series study.
Czech Republic:
40.3 ug/m3
SD25.0
Bavaria, Germany:
38.2 ug/m3
SD21.9
24-havgO3: TSP, PM10, NO2, 0,1,2,3,4,5, Poisson GLM.
Median 18 ug/m3 SO2 6,7 Time-series study.
5th %-95th% 3-51


Effect Estimates

1-h max O3 (per 100 ug/m3)
All ages:
LagO: 1.8% (-3. 8, 7.8)
Lagl: 0.1% (-4.7, 5.1)
Lag 2: 4.9% (0.1, 10.0)
24-h avg O3 (per 100 ug/m3):
Czech Republic:
All cause:
Lag 2: 7.8% (-1.8, 18.4)
Bavaria, Germany:
All cause:
LagO: 8.2% (0.4, 16.7)
Cardiovascular:
LagO: 6.1% (-3.7, 17.0)
24-h avg O3 (per 20 ug/m3):
All cause, age <65 years:
Not significant, values not
reported.
Cardiovascular, age <65
years:
LagO: -2.0% (-9. 5, 6.1)
Lagl: 6.2% (-2.2, 15.5)

-------
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
OQ
O
O















oo
u\


O
i
H
6
0
2|
O
H
O
0
H
W
O

1 — I
H
W
Reference, Study
Location and Period
Europe (cont'd)
Saez et al. (2002)
Barcelona, Spain
1991-1995
Madrid, Spain
1992-1995
Valenica, Spain
1994-1996






Garcia-Aymerich et al.
(2000)
Barcelona, Spain
1985-1989















Saezetal. (1999)
Barcelona, Spain
1986-1989
Copollutants Lag Structure
Outcome Measure Mean O3 Levels Considered Reported Method/Design

All cause; 8-h max O3: NO2, PM, SO2, 0-5 Poisson GAM with
respiratory; CO; multipollutant default convergence
cardiovascular Barcelona: models criteria. Time-series
67.5 ug/m3 study.
SD 32.2

Madrid:
42.1 ug/m3
SD27.8

Valencia:
45.5 ug/m3
SD 19.7
All cause; l-hmaxO3: BS,NO2, SO2, Selected best Poisson GLM.
respiratory; Levels not reported. single-day lag Time-series study.
cardiovascular;
general population;
patients with COPD














Asthma mortality; l-hmaxO3: BS,NO2, SO2, 0-2 Poisson with GEE.
age 2-45 years Levels not reported. Time-series study.

Effect Estimates

8-h max O3 (per 10 ug/m3):

Three-city pooled estimates:

All cause:
0.23% (-0.15, 0.61)
Respiratory:
0.29% (-0.05, 0.63)
Cardiovascular:
0.60% (0.08, 1.13)



1-h max O3 (per 50 ug/m3):

All cause:
General population:
Lag 5: 2.4% (0.6, 4.2)
COPD patients:
Lag 3: 4.0% (-4.7, 13.4)
Respiratory:
General population:
Lag 5: 3. 5% (-1.9, 9.2)
COPD patients:
Lag 3: 5. 7% (-7.9, 21.4)

Cardiovascular:
General population:
Lagl: 2.9% (0.4, 5.4)
COPD patients:
Lag 3: !.!%(- 14.2, 19.2)

Slope estimate:
0.021 (SE 0.01 l),p = 0.054


-------
                                            Table AX7-5 (cont'd).  Effects of Acute O3 Exposure on Mortality
OQ
 to
 O
 O
Reference, Study
Location and Period
                                   Outcome Measure     Mean O3 Levels
Copollutants
 Considered
Lag Structure
  Reported
                                                                                                        Method/Design
                                                                                                       Effect Estimates
 X
Europe (cont'd)

Sunyeretal. (1996)
Barcelona, Spain
1985-1991
All cause;
respiratory;
cardiovascular;
3.11 3.gcS,
age 70+ years
                                                                          BS, SO2, NO2
                                                      Summer:
                                                      86.5 ng/m3
                                                      Range 9.5-283.5

                                                      Winter:
                                                      55.2 ng/m3
                                                      Range 7-189.2
                                                                                      Selected best
                                                                                     single-day lag
                                 Autoregressive
                                 Poisson. Time-series
                                 study.
                                       1-h max O3 (per 100 |ig/m3):

                                       All cause, all ages:
                                       All year:
                                       LagO: 4.8% (1.2, 8.6)
                                       Summer:
                                       LagO: 5.8%(1.7, 10.1)
                                       Winter:
                                       LagO: 2.6%(-3.5, 9.1)

                                       Respiratory, all ages:
                                       All year:
                                       Lag 5: 7.1% (-3.8, 19.2)
                                       Summer:
                                       Lag 5: 5.0% (-7.3, 18.8)
                                       Winter:
oo
Oi


O
^
Prj
H
1
O
0
2" Sunyer and Basagana Mortality in a 1-h max O3: PM10, NO2, CO
O (2001) cohort of patients Mean not reported
H Barcelona, Spain withCOPD IQR21 |ig/m3
O 1990-1995
i~*
0
H
W
0
O
H
W
Lag 5: 14.0% (-7.6,40.6)

Cardiovascular, all ages:
All year:
Lagl: 5.8%(0.9, 11.1)
Summer:
Lagl: 8. 8% (2. 8, 15.2)
Winter:

Lagl: -0.8% (-8. 9, 7.9)

0-2 Conditional logistic 1-h max O3 (per 21 |ig/m3):
(case-crossover)
Odds ratio:
0.979(0.919,1.065)









-------
                                            Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
OQ
 to
 o
 o
        Reference, Study
        Location and Period
                          Outcome Measure    Mean O3 Levels
                      Copollutants
                       Considered
                    Lag Structure
                      Reported
              Method/Design
     Effect Estimates
Europe (cont'd)

Sunyer et al. (2002)
Barcelona, Spain
1986-1995
                                  All cause,
                                  respiratory, and
                                  cardiovascular
                                  mortality in a
                                  cohort of patients
                                  with severe asthma
Median 69.3 ug/m3
Range 6.6-283.0
                                              Median 54.4 ug/m3
                                              Range 3.9-244.5
PM10, BS, SO2,
N02, CO, pollen
0-2        Conditional logistic
           (case-crossover)
1-hmax O3 (per 48 ug/m3):

Odds ratios:

Patients with only one
admission:
All cause:
1.096(0.820,1.466)
Cardiovascular:
1.397(0.854,2.285)

Patients with more than
one admission:
All cause:
X

oo


o
s>
ITJ
H
6
° Diaz etal. (1999)
^ Madrid, Spain
^ 1990-1992
0
0
H
W
o

O
HH
H
hi
1.688(0.9/8,2.643)
Cardiovascular:
1.331(0.529,3.344)
Patients admitted for both
asthma and COPD:
All cause:
0.946(0.674,1.326)
Cardiovascular:
0.985(0.521,1.861)

All cause; 24-h avg O3: TSP, NO2, SO2, 1,4,10 Autoregressive linear. 24-h avg O3 (per 25 ug/m3):
respiratory; Levels not reported. CO Time-series study.
cardiovascular For O3 levels higher than 35
ug/m3:
All cause:
Lag 4: 12% (p < 0.01)

U-shaped (quadratic) O3-
mortality relationship with a
minimum of 35 ug/m3.


-------
                                       Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
£ Reference, Study
j^ Location and Period
o
O Latin America
Borja-Aburtoetal. (1997)
Mexico City
1990-1992








Outcome Measure


All cause;
respiratory;
cardiovascular; all
ages; age <5 years;
age >65 years






Copollutants
Mean O3 Levels Considered


l-hmaxO3: TSP, SO2, CO;
Median 155 ppb two-pollutant
models
8-h max O3:
Median 94 ppb

1 0-h avg O3 (8a.m.-
6 D m Y
±^ V
Median 87 ppb

24-h avg O3:
Median 54 ppb
Lag Structure
Reported Method/Design


0, 1, 2 Poisson iteratively
weighted and filtered
least-squares method.
Time-series study.







Effect Estimates


l-hmaxO3 (per

All ages:

O3 only model:
All
/\11 CollSc.
LagO: 2.4% (1.
Respiratory:
LagO: 2.3% (-
Cardiovascular:
LagO: 3.6%(0.


100 ppb):





1,3.9)

1.9,6.7)

6,6.6)
X
oo
oo
O3 with TSP model:
All cause:
LagO: -1.8% (-10.0, 6.4)
Respiratory:
LagO: -1.9%(-11.0,8.2)
Cardiovascular:
LagO: 2.4 (-4.4, 9.6)
^n
H
O
o
^^
Hy
-^H
o
H
0
c
o
H
W
0
&
O
H
W
Borja-Aburto et al. (1998) All cause;
SW Mexico City respiratory;
1993-1995 cardiovascular;
other; all ages;
age >65 years










1-hmax O3:
163 ppb
SD57

24-h avg O3:
44.0 ppb
SD15.7








PM2 5, NO2, SO2; 0,1,2,3,4,5, Poisson GAM with
two-pollutant 1-2 default convergence
models criteria (only one
smoother). Time series
study.










24-h avg O3 (per 10 ppb):

All cause, all ages:
Lag 1-2: 0.6% (-0.3, 1.5)
All cause, age > 65 years:
Lag 1-2: 0.8% (-0.4, 2.0)
Respiratory, all ages:
Lag 1-2: -0.7% (-3.6, 2.1)
Cardiovascular, all ages:
Lag 1-2: 1.8% (0.1, 3.5)
Other noninjury, all ages:
Lag 1-2: 0.3% (-0.9, 1.4)




-------
                                  Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
OQ
to
o
o
X
oo
VO
H

6
o


o
H

O

O
H
W

O


O
HH
H
W
Reference, Study
Location and Period
Latin America (cont'd)
O'Neill et al. (2004)
Mexico City
1996-1998





Tellez-Rojo et al. (2000)
Mexico City
1994




Loomisetal. (1999)
Mexico City
1993-1995



Gouveia and Fletcher
(2000b)
Sao Paulo, Brazil
1991-1993









Copollutants
Outcome Measure Mean O3 Levels Considered

All cause; all ages; 24-h avg O3: PM10
age 65+ years; SES 35.3 ppb
gradient SD11.0





Respiratory; l-hmaxO3: PM10,NO2, SO2
COPD mortality; 1 34. 5 ppb
age 65+ years; SD33.4
within medical
unit; outside of
medical unit

Infant mortality 24-h avg O3: PM2 5, NO2
44.1 ppb
SD15.7



All ages (all cause); 1 -h max O3: PM10, NO2, SO2,
age <5 years 67.9 ug/m3 CO
(all cause, SD42.1
respiratory,
pneumonia);
age 65+ years
(all cause,
respiratory,
cardiovascular)




Lag Structure
Reported Method/Design

0- 1 Poisson GAM with
stringent convergence
criteria. Time-series
study.




1, 2, 3, 4, 5, 1-3, Poisson, iteratively
1-5,1-7 weighted and filtered
least-squares method.
Time-series study.



0,1,2,3,4,5, Poisson GAM with
2-3 default convergence
criteria. Time-series
study.


0,1,2 Poisson GLM. Time-
series study.











Effect Estimates

24-h avg O3 (per 10 ppb):

All ages:
0.65% (0.02, 1.28)
Age 65+ years:
1.39% (0.51, 2.28)
SES gradient did not show
any consistent pattern.
1-h max O3 (per 40 ppb):

Outside medical unit:
Respiratory:
Lag 1-5: 14.0% (4.1, 24.9)
COPD mortality:
Lag 1-5: 15. 6% (4.0, 28.4)
24-h avg O3 (per 10 ppb):

O3 only model:
2.45% (-0.54, 5.43)
O3 with PM2 5 model:
1.40% (-1.92, 4.72)
1-h max O3 (per 106 ug/m3):

All ages:
All cause:
LagO: 0.8%(-1. 1,2.7)

Age 65+ years:
All cause:
Lag 2: 2.3% (0,4.6)
Respiratory:
Lag 2: 5.1% (-0.6, 11.1)
Cardiovascular:
LagO: 3.1% (-0.4, 6.7)

-------
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
OQ
to
O
o











j>
X
Reference, Study
Location and Period

Latin America (cont'd)
Pereiraetal. (1998)
Sao Paulo, Brazil
1991-1992
Saldivaetal. (1994)
Sao Paulo, Brazil
1990-1991
Saldivaetal. (1995)
Sao Paulo, Brazil
1990-1991




Outcome Measure


Intrauterine
mortality

Respiratory; age <5
years

All cause;
age 65+ years





Mean O3 Levels


l-hmaxO3:
67.5 ug/m3
SD45.0
24-h avg O3:
12. 14 ppb
SD 9.94
1-h max O3:
38.3 ppb
SD29.7

24-h avg O3:
12.5 ppb
SD11.5
Copollutants
Considered


PM10, N02, S02,
CO

PM10, N02, S02,
CO; multipollutant
models
PM10, NO2, SO2,
CO; two-pollutant
models




Lag Structure
Reported Method/Design


0 Poisson, linear with
M-estimation. Time-
series study.
0-2 OLS of transformed
data. Time-series
study.
0-1 OLS; Poisson with
GEE. Time-series
study.




Effect Estimates


Slope estimate:
0.0000 (SE 0.0004)

Slope estimate:
0.01048 deaths/day/ppb
(SE 0.02481 ),p = 0.673
Slope estimate:

l-hmaxO3:
0.0280 deaths/day/ppb
(SE 0.0213), p> 0.05

24-h avg O3:
                                                                           0.0093 deaths/day/ppb
                                                                           (SE 0.0813), p> 0.05

o

>
H
o
0

o
H
0
c
H
W
0
T^
O

w
Cifuentes et al. (2000) All cause
Santiago, Chile
1988-1966







Ostro et al. (1 996) All cause
Santiago, Chile
1989-1991






1-h max O3:

Summer:
108.2 ppb
IQR48.0





1-h max O3:
52.8 ppb
Range 11 -264






PM2 5, PM10 25, 0,1,2,3,4,5, Poisson GAM with
CO, SO2,NO2 1-2,1-3,1-4, default convergence
1-5 criteria; Poisson GLM.
Time-series study.






PM10, NO2, SO2; 1 OLS, several other
two-pollutant methods. Time-series
models study.






1-h max O3 per (108.2 ppb):

GLM:

Summer:
O3 only model:
Lag 1-2: 0.3%(t=0.3)
Multipollutant model:
Lag 1-2: -0.1% (t= -0.1)

All year:
1-h max O3 (per 52.8 ppb):
-3% (-4, -2)

Summer:
1 -h max O3 (per 100 ppb):
4% (0, 9)



-------
                                  Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
OQ
to
O
O
X
Reference, Study
Location and Period
Australia
Morgan etal. (1998b)
Sydney, Australia
1989-1993





Simpson etal. (1997)
Brisbane, Australia
1987-1993










Asia
Kim et al. (2004)
Seoul, Korea
1995-1999











Outcome Measure Mean O3 Levels

All cause; 1 -h max O3:
respiratory; 24 ppb
cardiovascular SD 13





All cause; 8-h avg O3
respiratory; (10 a.m.-6 p.m.):
cardiovascular;
all ages; age All year:
<65 years; 18.1 ppb
age 65+ years Range 1.7-63.4

Summer:
20.2 ppb
Range 2.7-63.4
Winter:
16.1 ppb
Range 1.7-56.9

All cause 1 -h max O3:

All year:
35. 16 ppb
SD 18.31

Summer:
46.87 ppb
SD 22.46

Winter:
2 1.26 ppb
SD6.92
Copollutants
Considered

PMby
nephelometer,
NO2;
multipollutant
models



PM10,PMby
nephelometer,
NO2, SO2, CO











PM10, N02, S02,
CO; two-pollutant
models










Lag Structure
Reported Method/Design

0 Poisson with GEE.
Time-series study.






0 Autoregressive Poisson
with GEE. Time-series
study.











1 Poisson GAM with
stringent convergence
criteria (linear model);
GLM with cubic
natural spline; GLM
with B-mode spline
(threshold model).
Time-series study.






Effect Estimates

l-hmaxO3(per28ppb):

All cause:
2.04% (0.37, 3.73)
Respiratory:
-0.84% (-7.16, 5.91)
Cardiovascular:
2.52% (-0.25, 5.38)
8-h avg O3 (per 10 ppb):

All cause, all ages:
All year:
2.4% (0.8, 4.0)
Summer:
3.0% (1.0, 5.0)
Winter:
1.3% (-1.4, 4.1)





1-h max O3 (per 21 .5 ppb):

All year:
Linear model:
2.6% (1.7, 3.5)
Threshold model:
3.4% (2.3, 4.4)

Summer:
Linear model:
1.9% (0.5, 3.3)
Threshold model:
3.8% (2.0, 5.7)

-------
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
OQ
c
VI
to
o
o











^
X
^b
to


O
t?
H
6
o
!2
O
— 1
O
r^t
0
H
W
o
O
H
W

Reference, Study Copollutants
Location and Period Outcome Measure Mean O3 Levels Considered

Asia (cont'd)
Lee etal. (1999) All cause 1-h max O3: TSP, SO2
Seoul and Ulsan, Korea
1991-1995 Seoul:
32.4 ppb
10th%-90th%
14-55
Ulsan:
26.0 ppb
10th%-90th%
16-39
Lee and Schwartz (1999) All cause 1-h max O3: TSP, SO2
Seoul, Korea
1991-1995 Seoul:
32.4 ppb
10th%-90th%
14-55



Kwon et al. (200 1 ) Mortality in a 1-h max O3: PM10, NO2, SO2,
Seoul, Korea cohort of patients 3 1.8 ppb CO
1 994-1 998 with congestive IQR 20.5
heart failure Range 4.3-102.8










Lag Structure
Reported Method/Design


0 Poisson with GEE.
Time-series study.








0 Conditional logistic
regression. Case-
crossover with
bidirectional control
sampling.




0 Time-series analysis
using Poisson GAM
with default
convergence criteria;
case-crossover analysis
using conditional
logistic regression.








Effect Estimates


1-h max O3 (per 50 ppb):

Seoul:
1.5% (0.5, 2.5)
Ulsan:
2.0% (-11. 1,17.0)




1-h max O3 (per 50 ppb):

Two controls, plus and minus
one week:
1.5% (-1.2, 4.2)

Four controls, plus and minus
two weeks:
2. 3% (-0.1, 4. 8)
1-h max O3 (per 20.5 ppb):

Odds ratios from case-
crossover study design:

General population:
1.9% (1.0, 2.9)
Congestive heart failure
cohort:
5.1% (-3. 6, 14.7)




-------
Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality

£ Reference, Study
j^ Location and Period

J5 Asia (cont'd)
Hong et al. (2002)
Seoul, Korea
1995-1998



Tsai et al. (2003b)
Kaohsiung, Taiwan
1994-2000


!>
X
^o
OJ
Yang et al. (2004b)
O Taipei, Taiwan
F 1994-1998
-n
H
6
o
2|
O
H
O
0
H
W
0
O
H
W

Copollutants Lag Structure
Outcome Measure Mean O3 Levels Considered Reported Method/Design Effect Estimates


Acute stroke 8-h avg O3: PM10, NO2, SO2, 0 Poisson GAM with 8-h avg O3 (per 9.3 ppb):
mortality 22.6 ppb CO default convergence
SD 12.4 criteria. Time-series O3 only model:
IQR 9. 3 study. 2.9% (0.3, 5.5)
PM10< median: 5.5%
PM10> median: -2.5%
All cause; 24-h avg O3: PM10, SO2, NO2, 0-2 Conditional logistic 24-h avg O3 (per 19.2 ppb):
respiratory; 23. 6 ppb CO regression. Case-
cardiovascular; Range 1.2-83.0 crossover analysis. Odds ratios:
tropical area All cause:
0.994(0.995,1.035)
Respiratory:
0.996(0.848,1.169)
Cardiovascular:
1.005(0.919,1.098)
All cause; 24-h avg O3: PM10, SO2, NO2, 0-2 Conditional logistic 24-h avg O3 (per 9.34 ppb):
respiratory; 17. 18 ppb CO regression. Case-
cardiovascular; Range 2. 3-43.47 crossover analysis. Odds ratios:
subtropical area All cause:
0.999(0.972-1.026)
Respiratory:
0.991(0.897-1.094)
Cardiovascular:
1.004(0.952-1.058)









-------
                                 Table AX7-5 (cont'd). Effects of Acute O3 Exposure on Mortality
X
j^.
Reference, Study
Location and Period
Asia (cont'd)
Hedley et al. (2002)
Hong Kong
1985-1995
Intervention Jul 1990
(switch to low sulfur-
content fuel)










Outcome Measure

All cause;
cardiovascular;
respiratory;
neoplasms and
other causes;
all ages;
age 15-64 years;
age 65+ years








Mean O3 Levels

Average monthly
03:

Baseline:
18.5 lig/m3
SD7.5

1 year after
intervention:
21.3|ig/m3
SD9.1

2-5 years after
intervention:
22.1 ng/m3
SD 10.2
Copollutants
Considered

SO2 (main
pollutant of
interest, 45%
reduction observed
5 years after
intervention),
PM10, SO/', N02









Lag Structure
Reported Method/Design

Monthly Poisson regression of
averages monthly averages to
considered estimate changes in the
without lags increase in deaths from
warm to cool season.
Annual proportional
change in death rate
before and after the
intervention was also
examined.






Effect Estimates

Declines observed in all
cause (2. l%,p = 0.001),
respiratory (3.9%, p = 0.001),
and cardiovascular (2.0%, p =
0.020) mortality after the
intervention.

Analysis not specific for O3
effects. As O3 levels did not
change before and after the
intervention, O3 likely did not
play a role in the decline in
observed mortality.




-------
OQ
 to
 o
 o
                                           Table AX7-6. Effects of Chronic O3 Exposure on Respiratory Health
           Reference, Study
         Location, and Period
   Mean O3 Levels
                                                              Study Description
                                                                                            Results and Comments
 X
United States

Galizia and Kinney
(1999; expo sure data
Kinney etal, 1998)
U.S. nationwide
1995
l-hmaxO3:
10-year mean Jun-Aug:
61.2 ppb
SD15.5
Range 13-185
         Goss et al. (2004)
         U.S. nationwide
         1999-2000
l-hmaxO3
Sl.Oppb
SD7.3
                        Cross-sectional study of a nationwide sample of
                        520 young adults.  Subjects were nonsmokers, aged
                        17-21 years, 50% males.  Each subject provided one
                        spirometric lung function measurement in the spring
                        of their 1st year at Yale College in New Haven, CT,
                        and completed a questionnaire addressing
                        residential history, respiratory diseases, and activity
                        patterns. Long-term O3 exposure was treated as a
                        high/low dichotomous variable, with subjects
                        assigned to the high O3 category if they lived for
                        4+ years in counties with 10-year summer mean O3
                        levels greater than 80 ppb. Four lung function
                        variables (FVC, FEVl5 FEF25.75, FEF75) were
                        regressed on O3 exposure, controlling for
                        age, height, height squared, sex, race, parental
                        education, and maternal smoking history.
                        Respiratory symptom histories (cough, phlegm,
                        wheeze apart from colds, and composite index for
                        any of the three symptoms) were logistically
                        regressed on O3 exposure, controlling for sex, race,
                        parental education, and maternal smoking.
                                               Cohort study of 11,484 cystic fibrosis patients over
                                               the age of 6 years. Exposure to O3, PM25, PM10,
                                               NO2, SO2, and CO assessed by linking Aerometric
                                               Information Retrieval System with patients' home
                                               zip code. Studied exacerbation and lung function.
                                               Mortality was also of interest, but study was
                                               underpowered to examine this outcome. Logistic
                                               regression models were used to analyze the
                                               exacerbations and multiple linear regression was
                                               used to study lung function.  O3 monitoring season
                                               and regional effects also were examined.
                                                                                                         Significant decrements in FEV[ and FEF25.75 in relation to O3
                                                                                                         exposure were observed for all subjects and for males alone, but
                                                                                                         not for females alone. Similar patterns observed for FVC and
                                                                                                         FEF75, but not with statistical significance.

                                                                                                         Percent difference in lung function for high versus low O3
                                                                                                         exposure groups:
                                                                         All subjects: -3.07% (-0.22, -5.92)
                                                                         Females: -0.26% (3.79, -4.31)
                                                                         Males: -4.71% (-0.66, -8.76)

                                                                         FEF25.75:
                                                                         All subjects: -8.11% (-2.32, -13.90)
                                                                         Females: -1.96% (6.39, -10.30)
                                                                         Males: -13.02% (-4.87, -21.17)

                                                                         Wheeze and respiratory symptom index were significantly elevated
                                                                         for high O3 exposure group.

                                                                         Odds ratios for symptoms:

                                                                         Wheeze:  1.97(1.06,3.66)
                                                                         Respiratory symptom index: 2.00 (1.15, 3.46)

                                                                         Ozone may increase the risk for pulmonary exacerbations in cystic
                                                                         fibrosis patients.

                                                                         Odds ratios for two or more exacerbations (per 10 ppb increase in
                                                                         1-hmax O3):

                                                                         O3 only model:  1.10(1.03,1.17)
                                                                         O3 with PM2 5 model:  1.08(1.01,1.15)

                                                                         PM25, but not O3, was significantly associated with declines in
                                                                         lung function in these patients.

-------
                                     Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
OQ
 to
 o
 o
           Reference, Study
         Location, and Period
Mean O3 Levels
Study Description
                                                                                           Results and Comments
 X
         United States (cont'd)

         Kinney and Lippmann
         (2000)
         Fort Sill, OK;
         Fort Leonard Wood,
         MO; Fort Dix, NJ;
         Fort Benning, GA;
         West Point, NY
         Apr-Sep 1990
         Greeretal. (1993)
         California
         1973-1987
l-hmaxO3:
Mean during 5-week
summer training
period:

Fort Benning, GA:
55.6 ppb
(0 hours O3> 100 ppb)

Fort Dix, NJ:
71.3 ppb
(23 hours O3>100 ppb)

Fort Leonard Wood,
MO:
55.4 ppb
(1 hours O3> 100 ppb)

Fort Sill, OK:
61.7 ppb
(1 hours O3> 100 ppb)

Annual mean O3:
Levels not reported.
                     Prospective cohort study of 72 nonsmoking students
                     (mean age 20.25 years) at the U.S. Military
                     Academy at West Point, NY were measured for
                     lung function and respiratory symptoms before
                     (Apr) and after (Aug-Sep) taking part in an
                     intensive, largely outdoor, summer training over
                     five weeks (Jul 11-Aug 15) at four U.S. military
                     bases.  Ozone levels in the Fort Dix, NJ area were
                     consistently higher than at the three remaining three
                     locations. Analysis assessed change in lung
                     function and respiratory symptoms measured before
                     and soon after the summer training, and examined
                     whether adverse trends would be more pronounced
                     in students exposed to higher O3 levels during
                     summer training.
                     Prospective cohort study of 3,914 nonsmoking
                     adults aged 25+ years at enrollment in 1977
                     completed questionnaires at two time points, 1977
                     and 1987. To be eligible, subjects had to have lived
                     10 or more years within 5 miles of current
                     residence.  Residential histories used to interpolate
                     air pollution levels to zip centroids over a 20-year
                     period (1966-1987). New asthma cases defined as
                     answering yes to doctor diagnosed asthma at 1987
                     followup among those answering no at enrollment
                     in 1977.  Multiple logistic regression used to test for
                     associations between new-onset asthma and long-
                     term exposures to air pollution, controlling for age,
                     education, pneumonia or bronchitis before age 16
                     years, and years worked with a smoker through
                     1987. All models stratified by gender.
                                  Mean FEV! declined significantly over the two measurement
                                  points for all subjects combined, which may reflect combined
                                  effects of O3 with exposures to dust, vehicle exhaust, and
                                  environmental tobacco smoke as reported by subjects from all four
                                  locations in the post-summer questionnaire. However, a larger
                                  mean decline was seen at the higher O3 site, Fort Dix, than at the
                                  remaining three sites, suggesting an influence of O3 exposures.

                                  A larger decline was observed in subjects with post-summer
                                  measurements in the 1 st two weeks after returning from training
                                  compared to those measured in the 3rd and 4th weeks, which is
                                  consistent with the lung function effects being somewhat transient.

                                  Change in lung function over the summer:
                                                                                                        All locations:  -44 mL (SE21), p = 0.035
                                                                                                        Fort Dix: -78 mL (SE41), p = 0.07
                                                                                                        Forts Sill, Leonard Wood, and Benning combined:
                                                                                                        -31mL(SE24),p = 0.21
                                  There were 27 incident cases of asthma among 1,305 males and
                                  51 incident cases among 2,272 females.  In logistic regression
                                  analyses, long-term O3 exposures were associated with increased
                                  risk of incident asthma among males but not females.

                                  Relative risks for incident cases of asthma (per 10 ppb increase
                                  in annual mean O3):

                                  Males: 3.12(1.61,5.85)
                                  Females: 0.94(0.65,1.34)

-------
OQ
 to
 o
 o
                                      Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
           Reference, Study
         Location, and Period
   Mean O3 Levels
               Study Description
                   Results and Comments
 X
         United States (cont'd)

         McDonnell et al.
         (1999)
         California
         1973-1992
         Peters etal. (1999a,b)
         12 Southern
         California
         communities
         1993-1994
8-h avg O3
(9 a.m.-5 p.m):
20-year mean:
46.5 ppb
SD15.3
l-hmaxO3:
Mean range:

1986-1990:
30.2 ppb (Santa Maria)
to 109.2 ppb
(San Dimas)

1994:
35.5 ppb (Santa Maria)
to 97.5 ppb
(Lake Gregory)
This prospective cohort study continued the work of
Greer et al. (1993). 3,091 nonsmoking adults
completed questionnaires at one additional time
point, 1992. Residential histories used to interpolate
air pollution levels to zip centroids over the period
1973-1992, yielding annual mean exposure
estimates for O3, PM10, SO2, and NO2.  New asthma
cases defined as answering yes to doctor diagnosed
asthma at either 1987 or 1992. Multiple logistic
regression used to test for associations between
new-onset asthma and long-term  exposures to air
pollution, controlling for age, education, pneumonia
or bronchitis before age 16, and ever smoking.
All models run separately for males and females.

3,676 children aged 9-16 years enrolled into
the 1st cohort of the Children's Health Study in
1993. Subjects provided questionnaire data on
respiratory disease histories and symptoms.
3,293 subjects also underwent pulmonary function
testing, of which 2,781 were used in air pollution
regressions. Air pollution data for O3, PM10, PM2 5,
NO2, and inorganic acid vapors analyzed from
1986-1990 and 1994. For cross-sectional analysis
of respiratory diseases, individual pollutants were
tested for associations with ever asthma, current
asthma, bronchitis, cough, and wheeze after
controlling for covariates. For analysis of lung
function, individual pollutants and pairs of
pollutants were regressed with FVC, FEV1; FEF25.75,
and PEF, controlling for usual demographic and
anthropometric covariates.
There were 32 incident cases of asthma among 972 males and
79 incident cases among 1,786 females. In logistic regression
analyses, long-term O3 exposures were associated with increased
risk of incident asthma among males but not females. Other
pollutants were neither associated with asthma incidence nor
were confounders of the O3 association in males.

Relative risks for incident cases of asthma (per 27 ppb increase
in annual mean 8-h avg O3):

Males: 2.09(1.03,4.16)
Females: 0.86(0.58,1.26)
Acids and NO2, but not O3, were associated significantly with
prevalence of wheeze.  No associations of O3 with any of the
respiratory diseases or symptoms.

Decreased lung function was associated with multiple pollutants
among females but not males. For O3 exposure in females, all four
lung function variables declined with increasing exposure.
Associations were stronger for current (1994) exposure compared
to previous (1986-1990) exposure. In males who spent more time
outdoors, FVC and FEV[ declined significantly with higher current
exposure to O3.

Change in lung function (per 40 ppb 1-h max O3 from 1986-1990):

Females:
PEF:  -187.2mL/s(SE50.1),p<0.005
FEF25.75:  -102.2 mL/s (SE 28.8), p < 0.01

Males:
PEF:  31.1 mL/s (SE 48.8), p> 0.05
FEF25.75:  11.7 mL/s (SE 26.7), p > 0.05

-------
OQ
 to
 o
 o
                                      Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
           Reference, Study
         Location, and Period
                          Mean O3 Levels
                                       Study Description
                                                                    Results and Comments
 X
 oo
         United States (cont'd)

         Gauderman et al.
         (2000; 2004a,b)
         12 Southern
         California
         communities
         1993-2001
Gauderman et al.
(2002)
12 Southern
California
communities
1996-1999
         McConnell et al.
         (1999)
         12 Southern
         California
         communities
         1993
                       8-h avg O3
                       (10 a.m.-6 p.m.):
                       Mean range:
                       Approximately 28 ppb
                       (Long Beach) to 65 ppb
                       (Lake Arrowhead)
8-h avg O3
(10 a.m.-6 p.m.):
Mean range:
Approximately 27 ppb
(Long Beach) to 67 ppb
(Lake Gregory)
                       l-hmaxO3:
                       Estimated annual daily
                       mean:
                       65.5 ppb
                       Range 35.5-97.5
Analysis of longitudinal lung function change
in relation to long-term air pollution levels in
the Children's Health Study cohort. Children from
4th (n = 1,498), 7th (n = 802), and 10th (n = 735)
grade enrolled in 1993. Children enrolled in 7th and
10th grade were followed until 1997; 4th graders
were followed until 2001.  Baseline questionnaires
completed and annual pulmonary function tests
(FVC, FEVj, FEF25.75, FEF75) performed. Air
pollution monitoring stations established in the
12 study communities beginning in 1994 to measure
O3, NO2, PM10, PM25, and inorganic acid. Analysis
using adjusted linear regression models.

Second cohort of the longitudinal cohort Children's
Health Study. 2,081 4th graders (mean age 9.9
years) enrolled in 1996.  Baseline questionnaires
were completed and annual pulmonary function
tests (FVC, FEVj, FEF25.75, FEF25.75/FVC, PEF)
were performed.  1,672 children had at least two
pulmonary function test data. Air pollutants
examined  include O3, NO2, PM10, PM2 5, inorganic
acid, elemental carbon, and organic carbon.
Adjusted linear regression model was used.
                        First cohort of the Children's Health Study.
                        Association of O3 with prevalence of chronic lower
                        respiratory tract symptoms among children with a
                        history of asthma was examined in a cross-sectional
                        study in 12 communities. Questionnaires were
                        completed by parents of 3,676 4th, 7th, and 10th
                        graders, of which 493 had asthma. Exposure data
                        (O3 NO2, PM10, PM25, and inorganic acid vapors)
                        collected in 1994 used to estimate exposure.
                        Analysis using logistic regression method.
                                                                         In the 7th and 10th grade cohorts, difference in lung function
                                                                         growth from the least to the most polluted community was not
                                                                         associated with any of the air pollutants, including O3. Among the
                                                                         4th graders, decreased lung growth was associated with exposures
                                                                         to PM and NO2, but not with O3.
In this cohort, a significant association between O3 and PEF and
FVC was noted in children spending more time outdoors.

Percent difference in annual increases in lung function from least
to most polluted community (per 36.6 ppb increase in annual mean
8-h avg 03):

PEF:
All children: -1.21% (-2.06, -0.36)
Children more  outdoors:  -1.62% (-2.93, -0.29)
Children less outdoors: -0.87% (-2.09, 0.37)

Children with asthma were much more likely to have bronchitis or
related symptoms than children without such history. Among the
asthmatic children, significant relationship were observed between
phlegm and all pollutants studied, with the exception of O3.

-------
OQ
 to
 o
 o
                                      Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
           Reference, Study
         Location, and Period
   Mean O3 Levels
               Study Description
                   Results and Comments
 X
         United States (cont'd)

         McConnell et al.
         (2003)
         12 Southern
         California
         communities
         1996-1999
8-h avg O3
(10 a.m. -6 p.m.):

4-year average across
12 communities:
47.2 ppb
SD11.3
Range 28. 3-65. 8

Range of yearly
variability within the
12 communities:
5.3 ppb
SD3.2
Range 1.7-13.2
A total of 475 children with asthma from the 1st and
2nd cohorts of the Children's Health Study were
recruited to examine the relationship between
bronchitic symptoms and air pollutants. Analysis
involved three stages using logistic mixed effects
models. Within-community variability in air
pollution was assessed in the first stage; individual
level time-independent confounders were assessed
in the second stage; and the effects of 4-year
average air pollutants were examined in the third
stage. Other copollutants examined include NO2,
PM10, PM25, PM10_25, inorganic acid, organic acid,
EC, and OC.
Symptoms were generally associated with the various air
pollutants. Within-community effects were greater than the
between-community effects. Authors note that if the larger within-
community effect estimates are correct, then other cross-sectional
(between-community) studies may have underestimated the true
effect of air pollution on respiratory  symptoms in children. These
differences may be attributable to confounding by poorly measured
or unmeasured risk factors that vary  between communities. Ozone
effect estimates were markedly reduced in two-pollutant models
(odds ratios not provided).

Odds ratios for bronchitic symptoms (per ppb O3):

Within-community effects: 1 .06 (1 .00 -1 . 12)
Between-community effects:  0.99 (0.98, 1.01)
         McConnell et al.
         (2002)
         12 Southern
         California
         communities
         1993-1998
l-hmaxO3:
Four-year mean
(1994-1997):

Low pollution
communities (n = 6):
50.1 ppb
Range 37.7-67.9

High pollution
communities (n = 6):
75.4 ppb
Range 69.3-87.2
Prospective cohort study of 3,535 children (age
9-16 years) without a history of asthma recruited in
1993 and 1996, and followed with annual surveys
through 1998 to determine incidence of new onset
asthma. Participation in sports assessed at baseline.
Copollutants included PM10, PM25, NO2, and
inorganic acid vapors.  Asthma incidence was
examined as a function of number of sports played
in high and low pollution communities, controlling
for age, sex,  and ethnic origin.
Asthma incidence was not higher in the high pollution
communities as compared with the low pollution communities,
regardless of the pollutant used to define high/low. In fact,
the high O3 communities had generally lower asthma incidence.
However, in high O3 communities, there was an increased risk of
asthma in children playing three or more sports compared to those
playing no sports; no such increase was observed in the low O3
communities.  No other pollutant showed this association. These
results suggest that high levels of physical activity is associated
with risk of new asthma development for children living
in communities with high O3 levels.

Relative risk of developing asthma in children playing three or
more sports compared to those playing no sports:

Low pollution communities: 0.8(0.4, 1.6)
High pollution communities: 3.3 (1 .9, 5.9)

-------
OQ
 to
 o
 o
                                      Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
           Reference, Study
         Location, and Period
                          Mean O3 Levels
                                                              Study Description
                                                                    Results and Comments
 X
 o
 o
United States (cont'd)

Avoletal. (2001)
12 Southern
California
communities and six
western states
Baseline 1994
Follow-up 1998
Ritz and Yu( 1999)
Southern California
1989-1993
                                8-h avg O3
                                (10 a.m.-6 p.m.):

                                Mean range of baseline
                                levels:
                                30.4 ppb for Santa
                                Maria to 70.8 ppb for
                                Lake Gregory

                                Mean range of changes:
                                11.7 ppb increase to
                                27.0 ppb decrease in O3
                                levels
                                Data not given.
                                               110 children enrolled in Children's Health Study in
                                               1993 and 1994 followed after moving to different
                                               western states. Age 10-11 years at time of
                                               enrollment). Follow-up pulmonary function testing
                                               carried out in 1998. Change in lung function over
                                               time tested in relation to change in exposures to
                                               PM10, N02,  and O3.
                                               125,573 births within 2 miles of an air monitoring
                                               station were examined to determine associations
                                               between CO and low birth weight. Copollutants
                                               included only as potential confounders.
                                                 Negative but nonsignificant associations were found between lung
                                                 function parameters and changes in O3.  The relationship was
                                                 strongest with PM10. Subjects who moved to areas of lower PM10
                                                 showed greater increases in FEV1 compared subjects who moved
                                                 to areas with higher PM10.

                                                 Change in lung function (per 10 ppb increase in changes in annual
                                                 mean 8-h avg O3):

                                                 FEVI:  0.1 mL (-8.7, 8.9)
                                                 FVC: -1.4 mL (-10.8, 8.0)
                                                 MMEF (maximal midexpiratory flow):  -3.4 mL/s (-23.6, 16.8)
                                                 PEF:  -8.9 mL/s (-41.6, 23.8)

                                                 Exposure to higher levels of ambient CO during the last trimester
                                                 was associated with a significantly increased risk for low birth
                                                 weight. Effects of CO appeared more pronounced after adjustment
                                                 for concurrent exposures to NO2, PM10, and O3.  Ozone effect
                                                 estimates were not reported.
         Ritz et al. (2000)
         Southern California
         1989-1993
                       8-h avg O3
                       (9 a.m.-5 p.m.):
                       Six weeks before birth:
                       36.9 ppb
                       SD 19.4
                       Range 3.3-117 ppb
Data on 97,158 singleton births over period
1989-1993 linked to air monitoring data during
different periods of pregnancy to determine
associations between pollution exposures and
preterm birth. Besides O3, pollutants of interest
included PM10, NO2, and CO. Multiple regression
analysis used, controlling for maternal age, race,
education, parity, and other factors.
                                                                                                Both PM10 and CO during early or late pregnancy were associated
                                                                                                with increased risk for preterm birth. No associations observed
                                                                                                withO3.

-------
OQ
 to
 o
 o
                                      Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
           Reference, Study
         Location, and Period
   Mean O3 Levels
               Study Description
                   Results and Comments
 X
         United States (cont'd)

         Ritz et al. (2002)
         Southern California
         1987-1993
Data not given.
         Kilnzlietal. (1997);
         Tageretal. (1998)
         Los Angeles and San
         Francisco, CA;
         Berkeley, CA
         1995
8-h avg O3
(10 a.m.-6 p.m.):

Range of lifetime
mean:

Los Angeles:
25-74 ppb

San Francisco:
16-33 ppb
The effect of air pollution on the occurrence of birth
defects was examined using a case-control study
design.  Analyses focused on CO and O3. Six types
of cardiac birth defects were investigated - aortic
defects, defects of the atrium and atrium septum,
endocardial and mitral valve defects, pulmonary
artery and valve defects, conotruncal defects, and
ventricular septal defects.  A 1:10 case-control ratio
was achieved for defect-specific analyses.  Analyses
were conducted using polytomous logistic
regression.  A two-stage hierarchical regression
model was used to adjust for multiple comparisons.
A pilot cohort study of 130 freshman students
(age 17-21 years) at the University of California at
Berkeley measured for lung function and histories
of residential locations and indoor/outdoor activity
patterns and levels.  By design, students had
previously resided in one of two metropolitan areas
that differed greatly in O3 concentrations,
San Francisco or Los Angeles.  A key goal was to
test whether measures  of small airways function
(e.g., nitrogen washout, FEF25.75, FEF75) were
sensitive measures of long-term O3 impacts.
Lifetime exposures to O3, PM10 and NO2 assigned
by interpolation to sequence of residence locations
from available monitoring stations. Multiple
exposure measures were derived with varying
degrees of incorporation of time-activity
information, going from ecological concentration
to individual time-activity weighted exposure.
Performed linear regression analysis of lung
function on O3 exposures, controlling for height,
ethnicity, gender, and region.
Concentration-response patterns were observed for O3
concentrations during the 2nd month of gestation on aortic artery
and valve defects, pulmonary artery and valve anomalies, and
conotruncal defects. CO during the 2nd gestational month was
found to be associated with ventricular septal defects.  The results
were inconclusive for NO2 and PM10. Findings from this study
suggest that there may a vulnerable window of development to
human malformations.

Odds ratios for birth defects (per 10 ppb O3 during 2nd month of
gestation):

Aortic defects:  1.56(1.16,2.09)
Pulmonary valve defects:  1.34  (0.96, 1.87)
Conotruncal defects:  1.36 (0.91, 2.03)

Decreased FEF25.75 and FEF75 were associated with long-term O3
exposures.  Results were similar whether O3 exposure was purely
ecologic or incorporated time-activity information.  FVC, FEV^
and nitrogen washout were generally not associated with O3 levels.
No evidence for PM10 or NO2 main effects or confounding of O3.
Similar patterns results using O3 hours >60 ppb as exposure metric
instead of daily 8-h avg O3 (10 a.m.-6 p.m.).  Surprisingly, region
of residence was a major negative confounder as lung function was
lower on average among students from the low O3 city, San
Francisco, than among those who had lived in Los Angeles.
Ozone exposures were significant predictors only after controlling
the regional effect.

Change in lung function (per 20 ppb increase in lifetime mean
8-h avg O3):

FEF25.75:  -420 mL/s (-886, 46); 7.2% of population mean
FEF75: - 334 mL/s (-657, -11); 14% decline of population mean

-------
                                      Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
OQ
 to
 o
 o
           Reference, Study
         Location, and Period
   Mean O3 Levels
               Study Description
                   Results and Comments
         United States (cont'd)

         Sherwin et al. (2000)
         Los Angeles, CA and
         Miami, FL
         1995-1997
Levels not reported.
Lungs obtained from autopsies of young residents
(age 11-30 years) of Miami (n = 20) and Los
Angeles (n = 18) who died suddenly from homicide,
vehicular accident, or other violence.
Semiquantitative measurements of centriacinar
region alterations were compared between the
two cities.
A greater extent (p < 0.02) and severity (p < 0.02) of centriacinar
region alterations were observed in lungs of the Los Angeles
residents than the Miami residents. These differences could not be
attributed to smoking history. The higher O3 levels in Los Angeles
might be responsible for the greater centriacinar region alterations,
however correlations could not be performed due to the relatively
small number of cases available.
 X
 o
 to
         Gongetal. (1998b)
         Glendora, CA
         1977-1987
l-hmaxO3:
Annual means range
(1983-1987):
109ppbtol34ppb
Longitudinal cohort study of 164 adults (mean age
45 years; 34% males) from a high O3 community
underwent lung function testing in 1986-1987 (T3).
Subjects were recruited from a cohort of 208
nonsmoking adults who had been tested on two
previous occasions:  1977-1978 (Tl) and 1982-1983
(T2). Analyzed changes in lung function at three
time points. Subjects were also asked to undergo
controlled exposures to 0.40 ppm O3 over 2 hours
with intermittent exercise. 45 subjects agreed to
participate.  Investigators hypothesized that acutely
responsive subjects would show more rapid declines
in function over the study period.
Mean FVC and FEV! showed nonsignificant increase from T2 to
T3, whereas an earlier analysis of the Tl to T2 change had found a
significant decline in function (Betels et al., 1987).  There was
evidence for 'regression to the mean,' in that subject with larger
declines in function from Tl to T2 tended to have larger increases
in function from T2 to T3. A consistent decline in FEV/FVC ratio
was observed at all three time points (p < 0.0001 by ANOVA).

Acute changes in lung function, determined using controlled O3
exposures, were not associated with chronic lung function changes.
Chen et al. (2002)
Washoe County, NV
1991-1999

8-hmaxO3:
27.23 ppb
SD 10.62
Range 2.76-62.44
Birth weight for 36,305 single births analyzed
in relation to mean PM10, O3, and CO levels in
trimesters 1, 2, and 3.

PM10 was the only air pollutant associated with decreased birth
weights. Ozone levels quite low throughout study.


-------
                                     Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
OQ
 to
 o
 o
           Reference, Study
         Location, and Period
   Mean O3 Levels
Study Description
Results and Comments
 X
         United States (cont'd)

         Kinney etal. (1996b)
         New York City
         1992-1993
1-h max O3:              Study of 19 healthy adult joggers (age 23-38 years;
                        18 males) from the Governors Island U.S. Coast
Summer                 Guard facility in New York harbor underwent a
(Jul-Sep 1992):           series of two bronchoalveolar lavages, first in the
58 ppb                  summer of 1992 and then again in the winter of
Maximum 100           1992. Because the summer of 1992 had lower than
                        average O3 levels, six subjects underwent a third
Winter                  bronchoalveolar lavage in the summer of 1993.
(Jan-Mar 1992):          Study tested whether inflammatory markers in
32 ppb                  bronchoalveolar lavage fluid were elevated during
Maximum 64             the summer O3 season among adults who regularly
                        exercised outdoors. Outcomes included cell
Summer                 differentials, release of interleukin-8 (IL-8)
(Jul-Sep 1993):           and tumor necrosis factor-alpha (TNF-a) in
69 ppb                  bronchoalveolar lavage cells supernatants, release
Maximum 142           of reactive oxygen species by macrophages, and
                        concentrations of protein, lactate dehydrogenase,
                        IL-8, fibronectin, al-antitrypsin (al-AT),
                        complement fragments (C3a), and prostaglandin E2
                        (PGE2) in bronchoalveolar lavage fluids.
                                  There was no evidence of acute inflammation in the summer of
                                  1992 compared to the winter; i.e., neutrophil differentials, IL-8 and
                                  TNF-a showed no significant differences. However, a measure of
                                  cell damage, lactate dehydrogenase, was elevated in the summer,
                                  suggesting possible O3-mediated damage to the lung epithelium
                                  with repeated exposures to O3 while exercising.  O3 levels during
                                  the summer of 1992 were atypically low for New York City.
                                  Among six subjects who agreed to undergo a third bronchoalveolar
                                  lavage test in the summer of 1993, lactate dehydrogenase was
                                  again elevated compared to winter. In addition, IL-8 was elevated
                                  in the summer of 1993, suggesting acute inflammation.

-------
                                      Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
OQ
 to
 o
 o
           Reference, Study
         Location, and Period
   Mean O3 Levels
Study Description
Results and Comments
 X
         Europe

         Charpinetal. (1999)
         Seven towns in
         SE France
         Jan-Feb 1993
         Ramadour et al.
         (2000)
         Seven towns in
         SE France
         Jan-Feb 1993
8-h max O3:              Cross-sectional cohort study of 2,073 children (age
Range of means:         10-11 years) from 7 towns tested for atopy based on
30.2-52.1 ug/m3          skin prick testing (house dust mite, cat dander, grass
                        pollen, cypress pollen, sadAlternaria). Towns
24-h avg O3:             represented a range of ambient O3 and other
Range of means:         pollutant (NO2 and SO2) levels.  Tested hypothesis
20.1-42.1 ug/m3          that atopy is greater in towns with higher
                        photochemical pollution levels.  To be eligible,
                        subjects must have resided in current town for at
                        least 3 years. Authors stated that Jan to Feb
                        pollution levels correlated with levels observed
                        throughout the year, though no data was given to
                        support this.

8-h max O3:              Cross-sectional cohort study of 2,445 children
Range of means:         (age 13-14 years) who had lived at their current
30.2-52.1 ug/m3          residence for at least three years were recruited from
                        schools in seven towns in SE France. This region
                        has highest O3  levels in France.  Subjects completed
                        ISAAC survey of asthma and respiratory symptoms.
                        In addition to O3 also collected data on SO2 and
                        NO2.  Analyzed relationships between asthma and
                        other respiratory conditions with mean air pollution
                        levels across the seven towns using logistic
                        regression, controlling for family history of asthma,
                        personal history of early-life respiratory diseases,
                        and SES. Also did simple univariate linear
                        regressions.
                                   In this cross-sectional analysis, no differences in atopy levels were
                                   seen across the seven towns. Authors concluded that long-term
                                   exposures to oxidant pollution do not favor increased allergy to
                                   common allergens.  The very low winter O3 levels monitored and
                                   lack of long-term exposure data make it impossible to reach this
                                   conclusion in a definitive manner.
                                   In logistic regressions, no significant associations seen between O3
                                   and 12-month history of wheezing, history of asthma attack,
                                   exercise induced asthma and/or dry cough in last 12 months.

                                   In simple bivariate scatterplots of respiratory outcomes versus
                                   mean O3 levels in the seven towns, there appeared to be strong
                                   positive relationships (r = 0.71 for wheezing in last 12 months
                                   and r = 0.96 for asthma attacks). No data on slope estimates given.
                                   Concerns about potential confounding across towns
                                   limits the interpretation of this study.

-------
OQ
 to
 o
 o
                                     Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
           Reference, Study
         Location, and Period
                                                                    Study Description
                                                                                                                Results and Comments
X

o
Europe (cont'd)

Ihorst et al. (2004)
Nine communities in
Lower Austria
Apr 1994-Oct 1997
Six communities in
Germany
Febl996-0ctl999
         Kopp et al. (2000)
         Ten communities in
         Austria and SW
         Germany
         Mar 1994-Nov 1995
                               !/2-h avg O3:
                               Quartile ranges:

                               Summer:
                               1st quartile: 22-30 ppb
                               2ndquartile:  30-38
                               ppb
                               3rd quartile:  38-46 ppb
                               4th quartile:  46-54 ppb

                               Winter:
                               1st quartile: 4-12 ppb
                               2nd quartile:  12-20
                               ppb
                               3rd quartile:  20-28 ppb
                               4th quartile:  28-36 ppb
                              !/2-h avg O3:
                              Stratified by low,
                              medium, high
                              exposure:

                              Low:  24-33 ppb
                              Medium: 35-38 ppb
                              High: 44-52 ppb
                                                      Longitudinal cohort study of 2,153 children (median
                                                      age 7.6 years) were studied for the effects of semi-
                                                      annual and 3'/2-year mean O3 concentrations on
                                                      FVC and FEV^ As a measure of lung growth,
                                                      the difference between two consecutive values for
                                                      each child was divided by the number of days
                                                      between tests.  The effect of O3 exposure on lung
                                                      growth was analyzed by linear regression models,
                                                      after adjusting for sex, age, height at start of the
                                                      time period, and passive smoking exposure.
                                              Longitudinal cohort study of 797 children with a
                                              mean age of 8.2 years. Four pulmonary function
                                              tests (FVC, FEVj) performed on each child over two
                                              summers. Examined association between average
                                              daily lung function growth and exposure to O3,
                                              PM10, NO2, and SO2. Analysis using linear
                                              regression models.
Higher semi-annual mean O3 levels were associated with
diminished lung function growth during the summer, but
increased lung function growth in the winter.

Change in lung function (4th quartile compared to 1 st quartile
semi-annual O3 mean):

Summer:
FEV; (mL/100 days):  -18.5 (-27.1,-9.8)
FVC (mL/100 days):  -19.2 (-27.8, -10.6)

Winter:
FEV; (mL/100 days):  10.9(2.1,19.7)
FVC (mL/100 days):  16.4(8.3,24.6)

No associations between longer term O3 exposure (mean summer
O3 over a 3!/2-year period) and lung function growth was found.

Lower FVC and FEV[ increases were observed in children exposed
to high ambient O3 levels compared to those exposed to lower O3
levels during the summer. During the winter, children in higher O3
areas showed a slightly greater increase in FVC and FEV; than
those in the low O3 areas, which might reflect that children catch
up in lung function deficits during the winter season.

Change in lung function for high versus low O3 exposure groups
(per ppb O3):

FEVI:
Summer of 1994:  -0.303 mL/day, p = 0.007
Winter of 1994/1995:  0.158 mL/day, p = 0.006
Summer of 1995:  -0.322 mL/day, p = 0.001

-------
OQ
 to
 o
 o
                                      Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
           Reference, Study
         Location, and Period
                                       Study Description
                                                                    Results and Comments
 X
         Europe (cont'd)

         Frischeretal. (1999)
         Nine communities in
         Austria
         1994-1996
24-h avg O3:

Summer:
34.8ppb
SD8.7

Winter:
23.1ppb
SD7.7
         Frischeretal. (2001)
         Nine communities in
         Austria
         Sep-Octl997
!/2-h avg O3:
30-day mean:
31.57ppb
IQR 20.61
Longitudinal cohort study of communities from two
counties chosen to represent a broad range of O3
concentrations; a two-fold range in mean levels was
observed.  1,150 children (mean age 7.8 years;
52% males) from grades 1 and 2 performed
spirometry in spring and fall over three years (total
of six measurements per child) to determine if
seasonal exposure to O3 would be associated with
diminished lung function growth, especially over
the summer seasons. Ozone levels were low during
lung function testing periods.  Participation rates
were high. At baseline, respiratory histories were
collected and subjects were tested for allergy by
skin prick. Examined association between O3 levels
and change in lung function (FVC, FEVj, and
MEF50 [maximal expiratory flow at 50% of vital
capacity]) over each season, controlling for baseline
function, atopy, gender, site, environmental tobacco
smoke exposure, season, and change in height.
Other pollutants studied included PM10, SO2, and
NO,.
A cross-sectional cohort study of 877 school
children (mean age 11.2 years).  Analyzed
for urinary eosinophil protein as a marker
of eosinophil activation determined from a single
spot urine sample using linear regression models.
Seasonal mean O3 exposures were associated with reductions in
growth in all three lung function measures.  Inconsistent results
seen for other pollutants.  Summer season lung function growth
decrements per unit O3 were larger when data restricted to children
who spent whole summer in their community. No evidence for
nonlinear O3 effect. No confounding of O3 effect by temperature,
ETS, or acute respiratory illnesses.

Change in lung function (per ppb O3):

FEV; (mL/day):
All subjects:
Summer: -0.029 (SE 0.005), p< 0.001
Winter:  -0.024 (SE 0.006), p< 0.001
Restricted to subjects who stayed in community:
Summer: -0.034 (SE 0.009), p< 0.001

FVC (mL/day):
All subjects:
Summer: -0.018 (SE 0.005), p< 0.001
Winter:  -0.010 (SE 0.006), p = 0.08
Restricted to subjects who stayed in community:
Summer: -0.033 (SE 0.007), p< 0.001

Log-transformed urinary eosinophil protein-X concentrations were
found to be significantly associated with O3 levels, after adjusting
for gender, site, and atopy.

Change in log urinary eosinophil protein-X (per ppb O3):
0.007 ug/mmol creatinine (SE 0.02), p < 0.001

-------
OQ
 to
 o
 o
                                     Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
           Reference, Study
         Location, and Period
                                      Study Description
                                                                   Results and Comments
 X
         Europe (cont'd)

         Horak et al. (2002a,b)
         Eight communities in
         Austria
         1994-1997
         Palli et al. (2004)
         Florence, Italy
         1993-1998
Summer:
31.8ppb
Range 18.7-49.3

Winter:
19.8ppb
Range 12.7-35.9
24-h avg O3:
Range of monthly
means from 1993-1998:
Approximately
25-125 ppb
This longitudinal cohort study continued the work
of Frischer et al., 1999 by including one additional
year of data, 1997. The major hypothesis
considered PM10. For this study, 80.6% of
the 975 children (mean age 8.11 years) performed
all six lung function tests. A total of 860 children
were included in the GEE analysis.  Multipollutant
analysis for PM10, SO2, and NO2.
Cohort study of 320 residents (age 35-64 years) in
the metropolitan area of Florence enrolled in a study
investigating the correlation between levels of DNA
bulky adducts and cumulative O3 exposure.
One blood sample was collected for each subject.
Various time windows of exposure were examined,
ranging from 0-15 days to 0-90 days prior to the
blood draw.  Simple Spearman correlations between
DNA adduct levels and different O3 exposure time
windows were calculated after stratifying  by
smoking history, area of residence, and population
type (random sample or exposed workers).
Seasonal mean O3 showed a negative effect on lung function
growth, confirming the previous shorter study.  Ozone effects were
robust to inclusion of PM10 into the model. However, for FEV; in
winter, the O3 effect slightly diminished after including PM10.
Taking into account only children who stayed at home the whole
summer period did not affect the results.

Change in lung function (per ppb O3):

FEV; (mL/day):
O3 only models:
Summer: -0.021, p< 0.001
Winter: -0.020, p < 0.001
O3 with PM10 models:
Summer: -0.020, p< 0.001
Winter: -0.012, p = 0.04

Consistent relationships between O3 exposure and DNA adduct
levels were  observed only among never smokers.  Correlations
were highest among never smokers who resided in the urban area
and were not occupationally exposed to vehicle traffic pollution.
Associations were significant up to a time window of 0-60 days
prior to the blood draw in the subgroup of never smokers, with
strongest relationships observed between 30-45 days prior.

-------
OQ
 to
 o
 o
                                     Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
           Reference, Study
         Location, and Period
                                                             Study Description
                                                                                                                  Results and Comments
 X
 o
 oo
Latin America

Calderon-Garciduenas
etal. (1995)
SW Mexico City
Nov 1993
Manzanillo, Mexico
Jan 1994
Calderon-Garciduenas
etal. (1997)
SW Mexico City
Sep-Nov 1995
Manzanillo, Mexico
Jan 1995
                               SW Mexico City
                               (urban):
                               l-havgO3>120ppb:
                               4.4 hours/day
                               Maximum 307 ppb

                               Manzanillo, Pacific
                               port (control):
                               No detectable air
                               pollutants.
                       SW Mexico City
                       (urban):
                       l-havgO3>120ppb:
                       82 hours/month
                       Maximum 286 ppb

                       Manzanillo, Pacific
                       port (control):
                       No detectable air
                       pollutants.
                                               Cross-sectional cohort study in which nasal lavage
                                               samples collected from 38 urban (mean age
                                               12.2 years) and 28 control (mean age 11.7 years)
                                               children. Samples analyzed for polymorphonuclear
                                               leukocyte counts, expression of human complement
                                               receptor type 3 (GDI Ib) on nasal
                                               polymorphonuclear leukocytes, and nasal
                                               cytologies.
Longitudinal cohort study of 129 urban and 19
control children aged 6-12 years old with no history
of smoking or environmental tobacco smoke
exposure and no current medication use for atopy or
asthma. Three nasal biopsies obtained at 4-week
intervals and analyzed for DNA damage based on
the presence of DNA fragments.
Nasal cytologies revealed that children from Mexico City had
abnormal nasal mucosae, including mucosal atrophy, marked
decreases in the numbers of ciliated-type cells and goblet cells,
and squamous metaplasia.

Exposed children had significantly higher nasal
polymorphonuclear leukocyte counts (p < 0.001) and nasal GDI Ib
expression (p < 0.001) compared to controls. However, the
inflammatory response did not seem to correlate with the previous
day's O3 exposure in a dose-dependent manner, suggesting that
there might be a competing inflammatory mechanism at the
bronchoalveolar level. Overall, these results suggest that ambient
O3 produces an inflammatory response in chronically exposed
children.

Urban children had significantly more DNA fragments than did
control children (p < 0.0001).  Percentage of damaged cells was
82.2% (SE 6.4) in urban children and 17.0% (SE 6.1) in control
children. Among urban children, more upper respiratory
symptoms and DNA damage was seen with increasing age.
Older children spent more time outdoors and engaged in physical
activities (p< 0.001).

Urban children were exposed to a complex pollution mix, making
it difficult to attribute effects to O3 specifically. However, authors
noted that O3 was the pollutant with most exceedences of air
quality standard.

-------
OQ
 to
 o
 o
                                      Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
           Reference, Study
         Location, and Period
                                       Study Description
                                                                     Results and Comments
 X
         Latin America (cont'd)

         Calderon-Garciduenas   SW Mexico City
         etal. (1999)
         SW Mexico City
         May-June 1996
         Manzanillo, Mexico
         May 1996
         Calderon-Garciduenas
         etal. (2001)
         SW Mexico City and
         Veracruz, Mexico
         1984-1999
         Calderon-Garciduenas
         et al. (2003)
         SW Mexico City and
         two control cities,
         Tuxpam and Tlaxcala,
         Mexico
         Ml 999-Jul 2000
(urban):
l-havgO3>80ppb:
May:
161 hours/month
Maximum 232 ppb
June:
98 hours/month
Maximum 261 ppb

Manzanillo, Pacific
port (control):
Mean<10 ppb
SW Mexico City
(urban):
l-havgO3>80ppb:
4 hours/day
Maximum 250 ppb

Veracruz (control):
In compliance for all
air criteria pollutants.

12-h avg O3
(8 a.m.-8 p.m.):

SW Mexico City:
Jan-Jun 1999:
84.3 ppb
Jul-Dec 1999:
60.9 ppb
Jan-Jun 2000:
76.8 ppb
Cross-sectional cohort study of 86 urban and
12 control children aged 6-13 years old with no
history of smoking or environmental tobacco smoke
exposure and no use of medication for atopy or
asthma. Urban children stratified into five groups
by school grade level (1st through 5th). Nasal
epithelial biopsies obtained from inferior nasal
turbinates, and analyzed for single strand DNA
breaks and for 8-OHdG (8-hydroxy-2'-
deoxyguanosine), a mutagenic lesion produced by
G^T mutations. These outcomes relate to possible
carcinogenic effects of air pollution exposures.
Multiple air pollutants monitored in SW Mexico
City within 3 miles of urban subject residences.

Ultra structural nasal pathology in Mexico City
children (n = 15) chronically exposed to O3, PM,
and other pollutants was compared to nasal
pathology in children from a city with low pollutant
levels (n = 11). All children were clinically healthy,
aged 4-15 years.  Statistical analyses performed
using student's t-test and Fisher's exact test.
174 urban and 27 control children aged 5-17years
examined for respiratory damage from chronic
exposure to air pollutants. Outcomes included nasal
abnormalities, interstitial lung markings assessed by
chest X-ray, lung function, and serum cytokines.
This cohort study combined cross-sectional
(radiology and hematological findings) and
longitudinal (spirometry) designs. Also examined
PM10 effects on respiratory damage.
No respiratory symptoms reported by control children; urban
children reported multiple nasal and lung symptoms, including
cough and chest discomfort among 46% of urban children, with
higher rates for 5th versus 1 st graders. 8-OHdG was
approximately 3-fold higher in biopsies from urban children
(p < 0.05), however, no differences by school grade.  Single strand
DNA breaks were more common in urban versus control children,
with an age-dependent increase in the urban children (p < 0.05).
These results suggest that DNA damage is present in the nasal
epithelial cells of children living in highly polluted SW Mexico
City  and may reflect enhanced risk of cancer later in life.

Though O3 represents an important component of the pollution
mix, it is not possible to attribute effects solely to O3.

Unremarkable mucociliary epithelium in nasal biopsies of control
children. The nasal mucosa in Mexico City children were
fundamentally disordered.  The mucociliary defense mechanisms
no longer functioned optimally. Major findings included lack of
cohesion between cells, epithelial shedding, necrotic cells, PMN
epithelial infiltration, and short or absent cilia.
Mexico City children exhibited nasal abnormalities (22%),
hyperinflation (67%), interstitial markings (49%), and a mild
restrictive pattern by spirometry (10%). In children with increased
interstitial markings, FEF75 values were significantly declined
(r = 0.42, p < 0.003).  Mexico City children also had more serum
IL-10 and IL-6, and less serum IL-8 than controls.  No significant
abnormalities were observed in the control children.
These results suggest chronic lung effects of O3 and related
copollutants at the high levels experienced in Mexico City.

-------
OQ
 to
 o
 o
                                      Table AX7-6 (cont'd).  Effects of Chronic O3 Exposure on Respiratory Health
  Reference, Study
Location, and Period      Mean O3 Levels
               Study Description
                                                                                                                             Results and Comments
         Latin America (cont'd)

         Fortoul et al. (2003)
         Mexico City
         May 1997
                       9-h avg O3
                       (9 a.m.-6 p.m.):

                       South:
                       121 ppb
                       North:
                       89 ppb
Cross-sectional cohort study estimated DNA strand
breaks on nasal epithelial cells and leucocytes
sampled from asthmatic (n = 15) and nonasthmatic
(n = 224) medical students aged 18-28 years using a
single-cell gel electrophoresis assay.
Greater genotoxic damage in asthmatics' nasal epithelial cells
(p < 0.05) may reflect their higher vulnerability for DNA damage,
or a decreased ability to repair it, compared with nonasthmatic
subjects.
         Gouveia et al. (2004)
         Sao Paulo, Brazil
         1997
                       l-hmaxO3:
                       63.0 ppb
                       SD33.5
Birth weight for 179,460 single births analyzed
in relation to PM10, SO2, CO, NO2, and O3 levels in
trimester 1, 2, and 3. GAM and logistic regression
models used for analysis.
Exposures to PM10 and CO during 1 st trimester were found to
have significant negative associations with birth weight.
No associations observed for the other air pollutants, including O3.
 X
Asia

Haetal. (2001)
Seoul, Korea
1996-1997
                                8-h avg O3:

                                1 st trimester:
                                Median 22.4 ppb
                                IQR 13.6

                                3rd trimester:
                                Median 23.3 ppb
                                IQR 16.1
Examined association between air pollution
exposure during pregnancy and low birth weight
among all full-term births for a two-year period.
These associations were evaluated after adjusting
for gestational age, maternal age, parental
educational level, parity, and infant sex. Analysis
using GAM with default convergence criteria.
Exposures during the 1st and 3rd trimesters were initially
examined separately. For the 1st trimester exposure estimates,
positive associations with risk of low birth weight were observed
for CO, NO2, SO2, and TSP, but not O3. For the 3rd trimester
exposure estimates, a positive association was observed for O3, but
not the other pollutants. When exposures from both trimesters
were examined simultaneously, the risk of low birth weight
remained positive for CO, NO2, SO2, and TSP during the 1st
trimester. However for O3, the positive association with 3rd
trimester exposure was diminished. These results suggest that
exposures to CO, NO2, SO2, and TSP in the 1st trimester may be
risk factors for low birth weight.

Relative risk of low birth weight (per 13.6 ppb 8-h avg O3 for 1 st
trimester; per 16.1 ppb 8-h avg O3 for 3rd trimester):

Stratified analyses by trimester:
1st trimester: 0.92(0.88,0.96)
3rd trimester:  1.09(1.04,1.14)

Combined analyses of both trimesters:
1st trimester: 0.96 (0.87, 1.07)
3rd trimester:  1.06(0.94,1.18)

-------
OQ
 to
 o
 o
                                      Table AX7-6 (cont'd). Effects of Chronic O3 Exposure on Respiratory Health
  Reference, Study
Location, and Period      Mean O3 Levels
               Study Description
                                                                                                                           Results and Comments
 X
         Asia (cont'd)

         Kuo et al. (2002)
         Central Taiwan
         1996
                       l-hmaxO3:
                       Annual mean range
                       across 7 of 8 schools:
                       18.6-27.3 ppb
Cross-sectional study. Respiratory questionnaire
administered to 12,926 children aged 13-16 years at
eight junior high schools in central Taiwan, to
determine asthma prevalence. The association
between asthma prevalence and air pollution
exposure analyzed by simple Pearson correlations of
prevalence with annual mean air pollution levels
(O3, SO2, PM10, and NO2), and by multiple logistic
regression.  The 775 asthmatics who were identified
then provided follow-up data on symptoms and
exacerbations over a one-year period. Simple
Pearson correlations were computed between
monthly hospital admissions and air pollution
levels, not controlling for covariates such as season
or weather.
Asthma prevalence ranged from 5.5% to 14.5% across the 8
schools. Based on simple Pearson's correlations, mean O3
(r = 0.51) and NO2 (r = 0.63) levels were correlated with variations
in asthma prevalence. However, only NO2 remained significant
in multiple logistic regression analyses after adjusting for various
potential confounding factors.

Longitudinal hospital admissions results are inconclusive due
to analytical limitations.  Monthly  correlations of hospital
admissions for asthmatics yielded variable results, all of which
would be confounded by temporal factors.

-------
OQ
 to
 o
 o
                                  Table AX7-7.  Effects of Chronic O3 Exposure on Mortality and Incidence of Cancer
  Reference, Location,
      Study Period
    Mean O3 Levels
                 Study Description
                                                                                                                                Results and Comments
         United States

         Pope et al. (2002)
         U.S. nationwide
         1982-1998
                          l-hmaxO3:
                          59.7 ppb
                          SD 12.8

                          24-h avg O3:
                          45.5 ppb
                          SD7.3
                          Prospective cohort study of approximately 500,000
                          members of American Cancer Society cohort enrolled
                          in 1982 and followed through 1998 for all cause,
                          cardiopulmonary, lung cancer, and all other cause
                          mortality. Age at enrollment was 30+ years. Air
                          pollution concentrations in urban area of residence at
                          time of enrollment assessed from 1982 through 1998.
                          Other pollutants considered include TSP, PM15, PM10,
                          PM25, PM15.25, SO/', S02, N02, and CO.
                                                    No significant effect of O3 on mortality risk, though the
                                                    association of Jul-Sep O3 concentrations with all cause
                                                    and cardiopulmonary mortality were positive and nearly
                                                    significant.

                                                    Residential location was known only at enrollment to
                                                    study in 1982. Thus, exposure misclassification is
                                                    likely to be high.
 X
Lipfert et al. (2000b;
2003)
32 Veterans
Administration hospitals
nationwide in the U.S.
1976-1996
95th % O3:

1960-1974: 132 ppb
1975-1981: 140 ppb
1982-1988: 94 ppb
1989-1996: 85 ppb
Cohort study of approximately 50,000 U.S. veterans
(all males) diagnosed with hypertension.  Mean age at
recruitment was 51 years.  Exposure to O3 during four
periods (1960-1974,  1975-1981, 1982-1988, 1989-
1996) associated with mortality over three periods
(1976-1981, 1982-1988, 1989-1996).  Long-term
exposures to TSP, PM15, PM10, PM25, PM15.25, SO42~,
NO2, and CO also analyzed.  Used Cox proportional
hazards regression, adjusting for race, smoking, age,
systolic and diastolic blood pressure, body mass index,
and socioeconomic factors.
Positive average concurrent responses for TSP, SO42~,
NO2, O3 in individual period analyses, but only O3 was
significant for overall.  Two-pollutants analyses indicate
that responses to peak O3 are robust.

Relative risks (per mean 95th % O3 less estimated
background level, value not reported):

Averaged over all four periods:

Exposure concurrent with mortality:
O3 only model:
9.4% (SE 4.6), p < 0.05
O3 with NO2 model:
12.2%, p < 0.05

Exposure before mortality:
O3 only model:
-0.2% (SE 6.3) p> 0.05

Analyses were robust to the deletion of diastolic blood
pressure in the models, indicating that the association
between mortality and O3 was not mediated through
blood pressure.

-------
OQ
 to
 o
 o
                 Table AX7-7 (cont'd).  Effects of Chronic O3 Exposure on Mortality and Incidence of Cancer
Reference, Location,
    Study Period
Mean O3 Levels
Study Description
Results and Comments
 X
         United States (cont'd)

         Abbey etal. (1999)
         Three California air
         basins:  San Francisco,
         South Coast (Los Angeles
         and eastward), San Diego
         1977-1992
                        24-h avg O3:
                        26.11ppb
                        SD 7.65
                        IQR 12.0

                        O3h/year>100ppb:
                        330 h/year
                        SD295
                        IQR 551
                      Prospective cohort study of 6,338 nonsmoking non-
                      Hispanic white adult members of the Adventist Health
                      Study followed for all cause, cardiopulmonary,
                      nonmalignant respiratory, and lung cancer mortality.
                      Participants were aged 27-95 years at enrollment in
                      1977. 1,628 (989 females, 639 males) mortality events
                      followed through 1992. All results were stratified by
                      gender. Used Cox proportional hazards analysis,
                      adjusting for age at enrollment, past smoking,
                      environmental tobacco smoke exposure, alcohol use,
                      education, occupation, and body mass index. Analyzed
                      mortality from all natural causes, cardiopulmonary,
                      nonmalignant respiratory, and lung cancer. Ozone
                      results were presented for both metrics.
                                    Of 16 regressions involving O3 exposures (two genders;
                                    four mortality causes; two O3 metrics), 11 were positive
                                    and one was statistically significant, for lung cancer in
                                    males for O3 h/year >100 ppb.

                                    Relative risks for lung cancer mortality in males:

                                    24-h avg O3 (per 12.0 ppb):
                                    2.10(0.99,4.44)

                                    O3 h/year >100 ppb (per 551 hours/year):
                                    4.19(1.81,9.69)

                                    Inconsistency across  outcomes and genders raises
                                    possibility of spurious finding. The lack of
                                    cardiopulmonary  effects raises plausibility concerns.
         Beesonetal. (1998)
         Three California air
         basins:  San Francisco,
         South Coast (Los Angeles
         and eastward), San Diego
         1977-1992
                        Annual mean O3:
                        26.2 ppb
                        SD7.7
                        O3h/year>100ppb:
                        333 h/year
                        SD297
                      Prospective cohort study of 6,338 nonsmoking non-
                      Hispanic white adult members of the Adventist Health
                      Study aged 27-95 years at time of enrollment.
                      36 (20 females, 16 males) histologically confirmed lung
                      cancers were diagnosed through 1992. Extensive
                      exposure assessment, with assignment of individual
                      long-term exposures to O3, PM10, SO42", and SO2, was a
                      unique strength of this study. All results were stratified
                      by gender. Used Cox proportional hazards analysis,
                      adjusting for age at enrollment, past smoking,
                      education, and alcohol use.
                                    Males, but not females, showed moderate association
                                    for O3 and incident lung cancer risk.

                                    Relative risks for lung cancer incident in males:

                                    O3 h/year >100 ppb (per 556 hours/year):
                                    All males: 3.56(1.35,9.42)
                                    Never smokers:  4.48(1.25,16.04)
                                    Past smokers: 2.15 (0.42, 10.89)

-------
era                          Table AX7-7 (cont'd).  Effects of Chronic O3 Exposure on Mortality and Incidence of Cancer

 ^         Reference, Location,
 O            Study Period            Mean O3 Levels                       Study Description                                 Results and Comments

         Latin America

         Pereira et al. (2005)        Mean avg of days/year      Annual records on larynx and lung cancer diseases        Results from this ecologic study provide limited
         Sao Paulo, Brazil           when O3 levels exceeded    obtained from the Sao Paulo Cancer Registry. The        evaluation of the relationship between air pollution and
         Exposure period           air quality standards (units   correlation between average air pollution data from       cancer. There was a significant difference in the
         1981-1990                 not provided):              1981 to 1990 and cases of larynx and lung cancer from    incidence of larynx and lung cancer among the Sao
         Case period                                         1997 were assessed using Pearson correlation            Paulo city districts.
         1997                      Lapa: 40.2                 coefficients.  Other pollutants examined included PM10,
                                  Moema:  19.6              NO2, NOX, SO2, and CO.                              Of all the pollutants examined O3 was best correlated
                                  Mooca:  67.1                                                                  with cases of larynx and lung cancer.
                                  Se:  28.2
                                                                                                               Pearson correlation coefficient:

                                                                                                               Larynx cancer: 0.9929 (p = 0.007)
 G                                                                                                             Lung cancer: 0.7234 (p = 0.277)
 K*N     	

-------
      AX7-2.  Description of Summary Density Curves
August 2005                   AX7-115   DRAFT-DO NOT QUOTE OR CITE

-------
 1      Introduction
 2           The summary density was used in various figures in Chapter 7. It is important that the
 3      reader not confuse a summary density with an error density which was  also used in Chapter 7
 4      figures.  This section will explain the relationship between these two densities, past use of
 5      similar densities, the theory behind this density, and explain how to interpret the graph of the
 6      summary density.
 7           In explaining the formula for the summary density and the interpretation of it, the
 8      discussion will explain the need for it and its construction. First, the preliminaries will be
 9      discussed. The statistically experienced reader can skip that portion except to read the 'error
10      density'  definition.
11           The summary density has been used before (Flachaire and Nunez, 2002). There, summary
12      density was weighted by the population size, and was used in an economic context to estimate
13      income distributions. A meta-analytic method for use in the presence of non-normal
14      distributions of effects with varying precision has been developed (Burr and Doss, 2005).  This
15      was used in an analysis of effects from multiple studies concerning the  association of the Platelet
16      PI A polymorphism of Glycoprotein Ilia and risk of coronary heart disease. A related density
17      estimate, the kernel density estimate, has also been used in a publication (Kochi et al., 2003)
18      referenced in a White Paper (Dockins et al., 2004) presentation to the U.S. EPA Science
19      Advisory Board - Environmental Economics Advisory Board.
20
21      Preliminaries
22           The error density is the curve describing the  distribution of uncertainty about the mean (or
23      posterior mean) or slope estimate.  For a normal density, this is sometimes referred to as the
24      "bell-shaped curve." With a two-sided test, the area under the curve and to the left of the no-
25      effect value for a positive (or right of the value for negative) estimate is less than 2.5% when the
26      effect estimate is statistically significant.
27           The log odds and the log relative risk estimates are usually considered to have a normal
28      distribution when the estimate is based on a large number of observations. Many times when the
29      health effect is a continuous variable, the estimate (or a transformation  of it) is assumed to have
30      a normal distribution.
        August 2005                             AX7-116      DRAFT-DO NOT QUOTE OR CITE

-------
 1           The two-sided confidence limits for a health effect estimates are based on the error density
 2      of the estimate.  The confidence limits include 95% of the area under the error density with equal
 3      portions of area outside the limits.
 4           In displaying confidence limits for the same estimate obtained under different conditions
 5      (relative risk for different cities or studies), the stick diagram is often used (for an example, see
 6      Figure 7-1).  The length of the line (stick) represents the confidence limit with the mean at the
 7      midpoint of this line.  When the confidence limits are for relative risks or odds ratios, the mean
 8      estimates may not be at the midpoint because their distributions are skewed. The stick diagram
 9      displays the confidence limits side-by-side.
10
11      What is Summary Density?
12           The summary density is the average of the error densities at each possible value of the
13      estimate. Since each estimate is different, the summary density usually will not appear as a
14      normal density.  The summary density may have many modes (bumps) or appear skewed.
15
16      Need for Summary Density
17           The summary density is used to portray the distribution of the heterogeneous effects, while
18      accounting for the differing error densities.  There are other graphics used for the same purpose.
19      The stick diagram (also called Forest Plots) is a portrayal of the heterogeneity, but is not easy to
20      interpret. The stick diagram gives a distorted view because the effects with the poorest estimates
21      and consequently the least informative have the longest confidence limits; thus, they catch the
22      eye rather than the most informative effects with shorter confidence intervals. A histogram
23      could be constructed from these estimates, but it weighs each estimate equally when some are
24      more precise than others.  Summary density curves can  be viewed as smoothed histograms.
25      However, unlike a histogram, summary density curves account for varying standard errors of the
26      individual mean effect estimates.
27           Many readers interpret stick diagrams by noting the fraction of significant effects. This
28      method has limitations, since there may be an overall significant effect detectable  by meta-
29      analysis and yet there are many insignificant effects due to low power in the individual studies.
30      The summary density of these effects may show a mode at a value different from zero. This
31      suggests that the insignificant effects tend to cluster around a value different from zero. The

        August 2005                             AX7-117      DRAFT-DO NOT QUOTE OR CITE

-------
 1      summary density can also show two or more modes indicating disagreement between estimates,
 2      or the presence of a factor (or multiple factors) that varies between estimates.
 3           Any summary density that does not appear to look similar to a normal density may be
 4      reflecting a distribution that is not normal.  This is important because most meta-analyses
 5      assume that the heterogeneity distribution is normal.
 6           Another method for portraying the distribution of effects having different precision is the
 7      radial plot developed by Galbraith (1994).  This is also a good method to summarize various
 8      effect estimates, but may require more statistical experience to understand.
 9
10      Theory of Summary Density
11           Statistical science has studied Kernel density estimates (Silverman, 1986) and the summary
12      density belongs to the class called the Gaussian Variable Kernel Density.  Gaussian is another
13      name for the normal distribution.  The summary density has a variable kernel because variances
14      of each kernel differ since the lengths of the confidence intervals of the estimates are not the
15      same. The kernel is the shape of the distribution used for each estimate. For example, other than
16      the normal density a triangular density could be used.  The histogram is a kernel density
17      estimate. The kernel in this case is rectangular (a uniform density). The rectangular shape is not
18      considered a good kernel because  as the number of observations increase the histogram
19      converges more slowly to the true density than do other kernels.
20           The summary density is presented as a graphical description of the heterogeneity. It does
21      not converge to the true distribution of the heterogeneity. To do so in an optimal way, the
22      variance of the error density would have to decrease depending on the number of effects being
23      studied and the standard deviation among them.
24           Rather than the summary density only being a descriptive tool, it can be used for inference.
25      Research has yielded a formula for fixing the variance of the Gaussian kernel so that there is less
26      than a 5% chance of erroneously concluding that a sample from a normal density is multimodal
27      (has more than one bump) (Jones,  1983).  Since the kernel of the summary density does not use
28      this formula, the significance of a  multimodal summary density is unknown. Also, due to the
29      varying kernel of the summary density,  the formula does not apply and likely calculates too
30      small a value. However,  simulation could be carried out to compute the p-value for each
31      application of the summary density to data.

        August 2005                             AX7-118      DRAFT-DO NOT QUOTE OR CITE

-------
 1     Extending the Summary Density to an Optimal Kernel Density Estimate
 2          The Guassian kernel density estimate is an average of normal density estimates, and is
 3     given by the equation
                                                                                      (AX7-1)
                                                  h*/2n
 4
 5     where x is the value (in this case a possible value of the effect) at which the density is to be
 6     evaluated, ut is the effect estimate, n is the number of samples (number of effects) and h is the
 7     standard deviation of each Gaussian density. As the number of samples increases, smaller
 8     values of h are used so that the kernel density estimate converges to the true density.
 9           The definition of the summary density is the same as the Guassian kernel density except

                                           hi = CT,-                                    (AX7-2)
10
              A.                                        ,
11     where Gi is the estimate of the standard error of the i  sample (effect estimate).  In this case, the
12     summary density is called a variable kernel density estimate since the ht vary.
13           To extent the summary density to an optimal density consider ht of the form:

                                              kA   .
                                       ^^%~CTz'                                  (AX7'3)
14
15     where & is a constant to be determined, og is the geometric mean of the estimated standard
16     errors and

                                    A = min(6, IQR 11.34)                             (AX7-4)
17
              s\.
18     where (J is the standard deviation of the samples (effect estimates) and IQR is their interquartile
19     range. Both these statistics are estimates of the true sigma when the distribution of the effects is
       August 2005                            AX7-119     DRAFT-DO NOT QUOTE OR CITE

-------
 1      normal.  The extension of the summary density to an optimal kernel density will be referred to as
 2      the extension.
 3           Next, the choice of the ht for the extension will be explained. Jones (1986) has shown for
 4      the (constant) kernel that when

                                             1.25CT
                                        h =	                                     (AX?_5)
                                               nk
 5
 6      a normally distributed sample will have a multimodal (more than one mode) kernel estimate
 7      5% of the time. The choice of ht for the extension will coincide with Jones' choice when all
 8      the standard errors are equal, and the IQR/l .34 is smaller than the estimate of a, and k = 2.5.
 9      However, for the extension a larger k is required to achieve the same critical value (5%) as Jones
10      has achieved. This is due to the kernel being variable and prone to have more than one mode.
11      Also, A is used rather than the more common estimate of <7, since A is used with the optimal
12      kernel.
13           Another simpler but somewhat unsatisfactory approach is to use Jones' result directly on
14      the effect size (the ratio of the effect estimate to its standard error). The variability of this ratio
15      is approximately constant for effects estimated from long sampling periods.
16           Probably the best and yet a somewhat subjective method is to adopt aspects of Silverman's
17      method (Silverman, 1986).  A multiple of the standard deviation used with the summary density
18      is considered. This multiple is either increased or decreased until  the density becomes visually
19      multiple modal. The effects are simulated using A as defined by equation AX7-3 for the
20      standard error and the mean of the effects for the error densities.
21           Whether or not multimodal is present at, say the 5% level can be approximated by fixed or
22      sequential sampling methods.
23
24      An Example of Calculating the Summary Density
25           Table AX7-8 shows the O3-associated excess risk estimates for cardiovascular mortality in
26      the warm season from select studies (see Figure 7-22 for the stick diagram of these estimates).
        August 2005                            AX7-120      DRAFT-DO NOT QUOTE OR CITE

-------
                Table AX7-8. Ozone-Associated Cardiovascular Mortality Risk Estimates
            	(95% CI) per Standardized Increment	
                                                                          %Excess Risk in
            Reference               Study Location                           Mortality
            Moolgavkar (2003)a      Los Angeles, CA                        1.61 (-0.24,3.50)
            Moolgavkar (2003)a      Cook County, IL                        6.82 (4.38,9.32)
            Lippmann et al. (2000)    Detroit Area, MI                        2.84 (-2.39, 8.35)
            Vedal et al. (2003)        Vancouver, British Columbia, Canada     16.37 (-1.14, 36.98)
            Anderson et al. (1996)     London, England                        4.46 (1.30, 7.72)
            Sunyer et al. (1996)       Barcelona, Spain                        6.70 (2.15, 11.50)
            Simpson  et al. (2001)      Brisbane, Australia                       7.37 (-3.41, 19.36)
            a Indicates use of Poisson GAM with default convergence criteria.
 1           The estimated log relative risk is considered normally distributed, so the % change is
 2      converted to log relative risk per standard unit (RR) by Equation AX7-6:

                              log(RR) = log[(% change/ 100) +1]                       (AX7-6)
 3
 4           The standard error of the log(RR) is obtained  by first applying the above equation to the
 5      confidence limits to get confidence limits for log(RR). Then the difference of these limits
 6      divided by 3.92 is used as the standard error of the log(RR).
 7           The equation of the normal density is
 8
 9     In the above equation, for one of the effects, the log(RR) is substituted for jut and the its standard
10     error is substituted for at.  This is done for each of the seven effect estimates.  These densities are
1 1     calculated over a range of x. For the figure, x has been calculated every 0.0004 units of log(RR)
12     from -0.2 to 0.2. Then for each value of x, each of the densities is averaged, and the result is the

       August 2005                             AX7-121      DRAFT-DO NOT QUOTE OR CITE

-------
 1
 2
 3
 4
 5
 6
summary density. In cases where a log transformation was used, the normal density was divided
by ex to convert to the lognormal density.
     Figure AX7-1 shows each of the seven error densities and the summary density multiplied
by 7.  In the figure, the summary density appears bimodal, but there are too few effects to
confirm this statistically. The smaller mode reflects a clustering of three error densities.
           V)
           c
           0)
           Q
                                          0                      10
                                   % Change in Cardiovascular Mortality
                                                                                20
       Figure AX7-1. Density curves of the O3-associated excess risk of cardiovascular mortality
                      in the warm season per standardized increment (see Section 7.1.3.2). The
                      thicker curve is the summary density curve of the seven effect estimates.
 1          In Chapter 7, the summary density curves in Figures 7-26 and 7-27 were calculated using
 2     equation AX7-3 with A equal to the estimated standard deviation of the effects. Simulation
 3     indicated the all-year curve in Figure 7-26 was not significantly multimodal.
 4
 5     Significance of a Summary Density
 6          Consider the instance when there is no significant difference between effect estimates, and
 7     the pooled analysis finds the overall effect to be significantly positive. In this case, the error
 8     density of the overall effect would have less than 2.5% of its area under the curve beyond the
 9     value of no effect.  However, the  summary density of the individual error densities will generally
10     show more than 2.5% beyond the no-effect value. Results from Mortimer et al (2002) reported
11     in Chapter 7 found no significant effects of O3 on PEF and incidence of symptoms in the
       August 2005
                                       AX7-122
DRAFT-DO NOT QUOTE OR CITE

-------
 1      individual cities, but observed a significant overall effect when the data from all cities were
 2      combined (see Figures 7-4 and 7-7). The summary density of the error densities from the eight
 3      cities had about 20% to 25% of the area beyond zero.  Simulation under these conditions verified
 4      that these larger percentages of the area beyond the value of no effect are common. One of the
 5      reasons for this result is that the summary density does not treat these estimates to be derived
 6      from random estimates of the same single true estimate.
 7          Rather than treating the area beyond the value of no effect in a summary density as an
 8      indication of significance, it should be treated as an indication of disagreement. For example,
 9      there might be a mode at zero and a mode at a positive value. This could be due to some cities
10      being always below the threshold and other cities being always above the threshold. In this case,
11      there could be appreciable area below zero  and yet there is a positive effect for some of the
12      cities.  Even if this simple case were the true state of nature, the summary density might appear
13      normal due to wide error densities.
14
15      "Apples and Oranges" Issue
16          The argument that a summary analysis is comparing apples and oranges or mixing apples
17      and oranges, is based on a variety of reasons. One of these reasons is based on the uncertainty
18      arising when the analysis includes effects that vary on factors other than the factor of interest.
19      The summary density may include  such factors, but when the summary density is bimodal there
20      is not an automatic conclusion this occurrence is due to only one factor. It may be spurious, due
21      to multiple factors or an unknown factor.
22          Another basis for the "apples and oranges" criticism is the lack of commonality among the
23      effects. That is, the effects may vary on a number of factors, but through some averaging
24      process the effect of interest can be estimated,  and these other factors will average out. The
25      summary density  does not average  effects, but considers common location through a clustering
26      of effects. This is a weaker assumption concerning the commonality of the effects.
27          Consider a summary density based on estimates  of high precision that cluster together, and
28      one estimate of low precision that is significantly different from the rest (the confidence intervals
29      do not overlap) and is based on an older measurement method.  The summary density could
30      average out the outlying value while forming a high mode based on the other effects.  Such a
31      graph would lead one to ignore the results of the old measurement method.  Ignoring previous

        August 2005                            AX7-123      DRAFT-DO NOT QUOTE OR CITE

-------
 1      results when more precise measurements become available is often practical under such
 2      circumstances.
 3
 4      Masking of Heterogeneity
 5           When the variances are very large, the kernel density appears to be very similar to a normal
 6      density.  Thus, large error densities can mask the true pattern of heterogeneity. There may be
 7      good reasons to suppose that the heterogeneity is other than normal, and the failure of the
 8      summary density to show this pattern is due to wide error densities.  When such masking occurs,
 9      the summary density cannot reject the assumption of normality of heterogeneous effects in a
10      meta-analysis.
11           Another reason that masking may occur with the summary density is the use of the
12      standard error of the effect as the standard deviation of the effects error density. Kernel density
13      theory permits decreasing the standard deviation when more effects are available. Narrower
14      error densities should clarify the heterogeneity distribution.  However, the ideal reduction in
15      variance due to increasing observation size is not large for the numbers of effects usually
16      considered. For example, ideally h decreases  as n115 increases, which is rather slowly.
17
18      Conclusions
19           The summary density is not new. As it stands, it is a kernel density estimate without a
20      fixed value of h.  Others have fixed h either using graphics (Kochi et al., 2003) or ad hoc
21      (Burr and Doss, 2003).  Flachaire and Nunez (2004) used a weighted average  of error densities
22      with the weights based on the population size.  This and other types of weighting should be
23      considered in the future. Also, improvements to unmask the heterogeneity distribution in a
24      statistically justified manner should be studied.
25           The summary density is a simple graphical method for portraying the  distribution of
26      heterogeneous effects in the presence of effects estimated with different precision. It has
27      graphical advantages over both the stick diagram and  the histogram.  The summary density can
28      be put on a more firm statistical footing. Inference concerning the presence of modes could be
29      made reliably if p-values were generated from simulation.  The summary  density is a graphical -
30      diagnostic tool for the normality assumption in meta-analyses. A meta-analysis method has
31      been developed for use when the distribution of effects is not normal and  the precision varies

        August 2005                            AX7-124      DRAFT-DO NOT QUOTE OR CITE

-------
1      (Burr, 2005). There is a need to develop other improvements to unmask the heterogeneity
2      distribution in a statistically justified manner.
3           The summary density overcomes some issues with reliance on statistical tests. If effects
4      were insignificant, one would expect them to cluster on either side of the no-effect value.  If the
5      summary density indicates a mode at a positive effect value, a tentative conclusion is that there is
6      positive nonrandom effect. How one would confirm the mode is statistically significant may
7      need to be studied, and it needs to be kept in mind that effects can include spurious components.
      August 2005                            AX7-125      DRAFT-DO NOT QUOTE OR CITE

-------