vvEPA
United States
Environmental Protection
Agency
Environmental Criteria and
Assessment Office
Research Triangle Park, NC 27711
EPA-600/8-78-004
April 1978
Research and Development
Air Quality Criteria
For Ozone and
Other Photochemical
Oxidants
-------
NOTICE
This document is available through the Library Services Office, MD-35, U.S.
Environmental Protection Agency, Research Triangle Park, N.C. 27711. It isalso
available from the Superintendent of Documents, U.S. Government Printing
Office, Washington, D.C. 20402. Correspondence relating to the subject matter
of the document should be directed to:
Project Officer for Ozone - Photochemical Oxidants
Environmental Criteria and Assessment Office (MD-52)
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
-------
EPA-600/8-78-004
APRIL 1978
AIR QUALITY CRITERIA
FOR
OZONE AND OTHER
PHOTOCHEMICAL OXIDANTS
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
-------
FOREWORD
This document has been prepared pursuant to Section 108(c)of the Clean Air
Act, as amended, which requires that the Administrator from time to time review
and, as appropriate, modify and reissue criteria issued pursuant to Section
108(a). Air quality criteria are required by Section 108(a)to reflect accurately the
latest scientific information useful in indicating the kind and extent of all
identifiable effects on public health and welfare that may be expected from the
presence of a pollutant in the ambient air in varying quantities.
The original criteria document for photochemical oxidants, AP-63, was issued
in 1970. Since that time, new information has been developed, and this
document represents the modification and reissuance of the air quality criteria
for photochemical oxidants.
The regulatory purpose of these criteria is to serve as the basis for national
ambient air quality standards promulgated by the Administrator under Section
109 of the Clean Air Act, as amended. Accordingly, as provided by Section
109(d), the Administrator has reviewed the national ambient air quality
standards for photochemical oxidants based on these revised criteria and is
proposing appropriate action with respect to those standards concurrently with
the issuance of this document.
The Agency is pleased to acknowledge the efforts and contributions of all
persons and groups who have contributed to this document as participating
authors or reviewers. In the last analysis, however, the Environmental
Protection Agency is responsible for its content.
DOUGLAS M.ICOSTLE
Administrator
U.S. Environmental Protection Agency
-------
PREFACE
This document consolidates and assesses current knowledge regarding the
origin of ozone and other photochemical oxidants and discusses their effect on
health, vegetation, certain ecosystems, and materials.
Photochemical oxidants are products of atmospheric reactions involving
hydrocarbons (HC), nitrogen oxides (N0»), oxygen, and sunlight. Oxidants consist
mostly of ozone, nitrogen dioxide (NOa), and peroxyacetylnitrate (PAN), with
smaller amounts of other peroxyacylnitrates, other oxy- and peroxy-compounds,
formic and nitric acid, and formaldehyde. Usually referred to collectively as
"oxidants," they originate mainly from human activities that produce HC and
NO* emissions.
This document summarizes current data on the effects of oxidant/ozone in the
ambient air on man, vegetation, and ecosystems. The effects that have been
observed will form the scientific basis for supporting the present National
Ambient Air Quality Standard of 160 /jg/m3 (0.08 ppm) or a revised standard.
Although nitrogen dioxide is considered one of the photochemical oxidants,
oxides of nitrogen are-the subject of a separate report and are therefore
discussed in this document only as they participate in the formation and
reactions of other photochemical oxidants. Hydrocarbons and other organics are
important air pollutants because they too are precursors of other compounds
formed in the atmospheric photochemical system. In this document, toxic
organics are considered only with respect to eye irritation.
The studies and data cited constitute the best available basis for specific
standards aimed at protecting human health and the environment from
photochemical oxidants in ambient air.
-------
CONTRIBUTORS AND REVIEWERS
The following individuals participated in the preparation of this document.
Authors
Dr. J. H. B. Garner
Project Officer for Ozone and Other
Photochemical Oxidants
Environmental Criteria and Assessment Office
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
Mr. Frank Black
Environmental Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
Dr. Robert Chapman
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
Dr. Kenneth L. Demerjian
Environmental Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
Dr. Basil Dimitriades
Environmental Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
Dr. Donald E. Gardner
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
Ms. Judith A. Graham
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
Mr. Fred H. Haynie
Environmental Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C. '
Dr. Milan J. Hazucha
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
-------
Mr. John H. Margeson
Environmental Monitoring and Support Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
Dr. Frederick J. Miller
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
Dr. Richard J. Paur
Environmental Monitoring and Support Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
Mr. Larry Purdue
Environmental Monitoring and Support Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
Mr. Robert K. Stevens
Environmental Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
Environmental Protection Agency Review Group
Dr. J. H. B. Garner, Chairman
Environmental Criteria and Assessment Office
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
Dr. John Clements
Environmental Monitoring and Support Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
Mr. Robert Fankhauser
Environmental Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
Mr. Thomas B. Feagans
Strategies and Air Standards Division
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
Mr. Charles R. Hosier
Environmental Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
Mr. Michael H. Jones
Strategies and Air Standards Division
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
Dr. John H. Knelson
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
-------
Mr. Ted Ripberger
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
Dr. David Tingey
Ecological Effects Division
U.S. Environmental Protection Agency
Corvallis, Ore.
Dr. David Weber
Ecological Effects Division
U.S. Environmental Protection Agency
Corvallis, Ore.
Dr. Raymond Wilhour
Ecological Effects Division
U.S. Environmental Protection Agency
Corvallis, Ore.
Consultants
Dr. Stephen M. Ayres
Chairman, Department of Internal Medicine
St. Louis University School of Medicine
St. Louis, Mo.
Dr. David V. Bates
Dean, School of Medicine
University of British Columbia
Vancouver, B C., Canada
Dr. T. Timothy Crocker
Chairman, Department of Community and
Environmental Medicine
University of California College of Medicine
Irvine, Calif.
Dr. Richard Ehrlich
Director, Life Sciences Division
Illinois Institute of Technology Research Institute
Chicago, III.
Dr. Gustave Freeman
Director, Department of Medical Sciences
Life Sciences Division
Stanford Research Institute
Menlo Park, Calif.
Dr. Bernard D. Goldstein
Associate Professor, Department of Medicine
and Department of Environmental Medicine
New York University School of Medicine
New York, N.Y.
Dr. John R. Goldsmith
Environmental Epidemiology Unit
California State Department of Public Health
Sacramento, Calif.
-------
Dr. Jack D. Hackney
Chief, Environmental Health Laboratories
Associate Professor of Medicine
Rancho Los Amigos Hospital Campus
of the University of Southern California
Downey, Calif.
Dr. Sagar V. Krupa
Assistant Professor, Department of Plant Pathology
University of Minnesota
St. Paul, Minn.
Dr. Daniel B. Menzel
Associate Professor of Pharmacology and Medicine
and Director, Laboratory of Environmental Pharmacology
and Toxicology
Duke University Medical Center
Durham, N.C.
Dr. Carl M. Shy
Institute of Environmental Health
University of North Carolina
Chapel Hill, N.C.
Dr. Boyd R. Strain
Professor, Department of Botany
Duke University
Durham, N.C.
Technical Assistance
Environmental Protection Agency
Research Triangle Park, N.C.
Ms. F. Vandiver Duffield
Mr. Douglas B. Fennel!
Mr. R. Wayne Fulford
Mr. Allen Hoyt
Ms. Evelynne R. Rash
Ms. Beverly A. Tilton
DOT Systems, Inc.
Vienna, Va.
Mr. Frank Broadwell
Dr. John Preston
Dr. Carolyn Pyrek
Mr. Arthur Willey
-------
-------
CONTENTS
Page
FIGURES xiv
TABLES xviii
ABBREVIATIONS AND SYMBOLS xxii
ABSTRACT xxix
ASSOCIATION OF OZONE WITH OTHER CONSTITUENTS OR
MANIFESTATIONS OF SMOG
1. SUMMARY AND CONCLUSIONS 1
INTRODUCTION 1
NATURE AND ATMOSPHERIC CONCENTRATIONS OF PHOTO-
CHEMICAL OXIDANTS 1
SOURCES AND SINKS OF OXIDANTS 1
OXIDANT PRECURSORS 2
RELATIONSHIPS BETWEEN AMBIENT OXIDANT AND PRECURSOR
EMISSIONS 3
MEASUREMENT METHODS FOR OXIDANT AND OXIDANT PRE-
CURSORS 3
HEALTH EFFECTS OF OZONE AND OTHER PHOTOCHEMICAL
OXIDANTS 4
Discussion of Threshold Concentrations 4
Human Studies 4
Animal Toxicology Studies 10-
Ozone Versus Oxidant Health Effects 11
EFFECTS OF PHOTOCHEMICAL OXIDANTS ON VEGETATION AND
CERTAIN MICROORGANISMS 11
EFFECTS OF PHOTOCHEMICAL OXIDANTS ON ECOSYSTEMS .. 14
EFFECTS OF PHOTOCHEMICAL OXIDANTS ON MATERIALS .... 16
2. INTRODUCTION 17
3. NATURE AND ATMOSPHERIC CONCENTRATIONS OF PHOTO-
CHEMICAL OXIDANTS 19
NATURE OF OXIDANT 19
OXIDANT CONCENTRATIONS AND THEIR PATTERNS 20
Introduction 20
Oxidant Concentrations in Urban Atmospheres 20
Oxidant Concentrations in Rural Atmospheres 21
Patterns of Variation in Oxidant Concentrations 23
ASSOCIATION OF OZONE WITH OTHER CONSTITUENTS OR
MANIFESTATIONS OF SMOG 24
SUMMARY 27
REFERENCES FOR CHAPTER 3 27
4. SOURCES AND SINKS OF OXIDANTS 29
PHOTOCHEMICAL FORMATION OF OXIDANTS 29
-------
Page
Introduction 29
Atmospheric Reaction Mechanisms 30
Effects of MeteorologiL'al Factors 35
NATURAL SOURCES OF OXIDANTS 47
Introduction 47
Stratospheric Ozone Intrusion 47
Photochemistry of Natural Organics NO* 52
SINKS OF OXIDANTS AND OF OXIDANT PRECURSORS 55
Introduction 55
Sinks for Oxidants and Oxidant Precursors 56
SUMMARY 57
REFERENCES FOR CHAPTER 4 58
5. OXIDANT PRECURSORS 65
INTRODUCTION 65
AMBIENT LEVELS AND VARIATIONS OF OXIDANT/OZONE PRE-
CURSORS 65
Organic Compounds in Urban Atmospheres 65
Organic Compounds in Rural and Remote Areas 70
Nitrogen Oxides 71
SOURCES OF OXIDANT PRECURSORS 71
Introduction 71
Summary of Organic Emissions Data 77
Hydrocarbon Emissions From Natural Sources 78
Hydrocarbon Emissions From Anthropogenic Sources 81
Emissions of Nitrogen Oxides 82
REACTIVITY OF ORGANIC EMISSIONS 83
SUMMARY 88
REFERENCES FOR CHAPTER 5 89
6. RELATIONSHIPS BETWEEN AMBIENT OXIDANTS AND PRECURSOR
EMISSIONS 93
INTRODUCTION 93
MODELS BASED ON EMPIRICAL RELATIONSHIPS 94
Rollback Model 94
Modified Rollback: Observational Model 95
Statistical-Empirical Models 98
MECHANISTIC MODELS OF Ox/HC/NOx RELATIONSHIPS 101
AIR QUALITY SIMULATION MODELS (AQSM) 105
SUMMARY 111
REFERENCES FOR CHAPTER 6 113
7. MEASUREMENT METHODS FOR OZONE, OXIDANTS, AND THEIR
PRECURSORS 116
INTRODUCTION 116
SAMPLING FACTORS IN AMBIENT AIR MONITORING 116
MEASUREMENT OF OZONE 117
Gas-Phase Chemiluminescence (EPA Reference Methods) .. 117
Gas-Solid Chemiluminescence 118
Ultraviolet Photometry 118
MEASUREMENT OF TOTAL OXIDANTS 119
OZONE CALIBRATION PROCEDURES 120
Kl Calibration Procedures 1 20
Recently Developed Calibration Procedures 121
-------
Page
RELATIONSHIPS BETWEEN AMBIENT OXIDANT AND OZONE
DATA 122
MEASUREMENT OF PEROXYACYLNITRATES 125
MEASUREMENT OF HYDROCARBONS 1 26
Measurement of Nonmethane Hydrocarbons 126
Measurement of Individual Hydrocarbons 1 28
Measurement of Reactive Hydrocarbons 1 29
Calibration 129
MEASUREMENT OF NITROGEN OXIDES 130
Measurement of N02 1 30
Measurement of NO 131
SUMMARY 131
REFERENCES FOR CHAPTER 7 131
8. TOXICOLOGICAL APPRAISAL OF PHOTOCHEMICAL OXIDANTS .. 136
INTRODUCTION 136
EFFECTS OF OZONE ON EXPERIMENTAL ANIMALS 1 36
Respiratory Tract Transport and Absorption 136
Mortality 1 38
Pulmonary Effects 1 38
Extrapulmonary Effects 1 58
Summary and Conclusions 1 62
EFFECTS OF PHOTOCHEMICAL OXIDANTS ON EXPERIMENTAL
ANIMALS 1 65
Experimental Data 1 65
Summary 176
EFFECTS OF PEROXYACETYLNITRATE ON EXPERIMENTAL
ANIMALS 1 77
Experimental Data 1 77
Summary 177
REFERENCES FOR CHAPTER 8 1 77
9. CLINICAL APPRAISAL OF THE EFFECTS OF OXIDANTS 184
OCCUPATIONAL AND ACCIDENTAL EXPOSURES TO OZONE ... 184
CONTROLLED STUDIES OF HUMAN HEALTH EFFECTS 186
Hematology 195
Mutagenesis 196
Controlled Studies of Human Health Effects of Peroxyacetylnitrate .. 1 97
SUMMARY 1 97
REFERENCES FOR CHAPTER 9 1 97
10. EPIDEMIOLOGIC APPRAISAL OF PHOTOCHEMICAL OXIDANTS . 200
INTRODUCTION 200
EFFECTS OF SHORT-TERM PHOTOCHEMICAL OXIDANT EXPOSURES 200
Daily Mortality in Relation to Variations in Oxidant Levels .. 200
Hospital Admissions in Relation to Oxjdant Levels 206
Aggravation of Existing Respiratory Diseases by Oxidant
Pollution 208
Effects of Oxidant on the Promotion of Symptoms and Illness in
Healthy Populations 212
Impairment of Performance Associated with Oxidant Pollution ... 221
Eye Irritation in Relation to Variations in Oxidant Levels .... 230
EFFECTS OF CHRONIC PHOTOCHEMICAL OXIDANT EXPOSURES .. 236
Introduction 236
-------
Page
Mortality in Areas of High- and Low-Oxidant Pollution 237
General Morbidity in Areas of High and Low Oxidant Pollution 239
Chronic Oxidant Exposure and Pulmonary Function 244
ATTITUDES OF LAYMEN AND PHYSICIANS TOWARD OXIDANT
AIR POLLUTION , 246
State of California General Health Survey 246
Survey of Los Angeles Physicians 246
Discussion of Attitudes of Respondents and Physicians Toward
Oxidant Air Pollution 247
SUMMARY OF EPIDEMIOLOGIC APPRAISAL OF PHOTOCHEMICAL
OXIDANTS 247
Effects of Short-Term Photochemical Oxidant Exposures .... 248
Effects of Chronic Photochemical Oxidant Exposures 249
Attitudes of Laymen and Physicians Toward Oxidant Air
Pollution , 250
REFERENCES FOR CHAPTER 10 250
11. EFFECTS OF PHOTOCHEMICAL OXIDANTS ON VEGETATION
AND CERTAIN MICROORGANISMS , 253
INTRODUCTION 253
VASCULAR PLANT RESPONSE TO PHOTOCHEMICAL OXIDANTS 253
Physiological Processes ., 254
Visible Symptoms 255
Growth and Yield 257
Factors Affecting Plant Response 270
RESPONSES OF MOSSES, FERNS, AND MICROORGANISMS ... 280
SUMMARY 281
REFERENCES FOR CHAPTER 11 284
1 2. ECOSYSTEMS 294
INTRODUCTION 294
GENERAL RESPONSES OF NATURAL AND AGROECOSYSTEMS
TO STRESS BY OXIDANTS 297
Agroecosystems • 297
Natural Ecosystems 297
ORIGIN OF INJURIOUS CONCENTRATIONS OF OZONE AND
OXIDANTS 298
Advection From Urban Centers to Remote Areas in Southern
California 298
San Joaquin Valley and Adjacent Sierra Nevada Mountains . 301
Seasonal and Daily Variations of Injurious Concentrations—
Synoptic Weather Patterns Associated With Episodes
of High Pollution 301
Annual Trends of Total Oxidant Concentrations at a San Bernardino
Station and the Nearby City of San Bernardino 304
INVESTIGATING THE EFFECT OF OXIDANT STRESS ON ECOSYSTEMS 307
Ecosystem Modeling 307
Modeling the Effects of Oxidant Stress on a Western Mixed
Conifer Forest Ecosystem 308
Effects on Primary Producers (Green Plants) 308
EFFECTS ON CONSUMER POPULATIONS 319
Vertebrate Populations 319
Plant Parasites and Symbionts 321
-------
Page
EFFECTS ON DECOMPOSERS 322
SUMMARY 324
REFERENCES FOR CHAPTER 12 326
13. EFFECTS OF OZONE ON MATERIALS 329
INTRODUCTION 329
MECHANISMS OF OZONE ATTACK 329
Effects on Elastomers 329
Effects on Textiles 333
Effects on Paints 337
Effects on Other Materials 340
SUMMARY 340
REFERENCES FOR CHAPTER 13 340
-------
FIGURES
Figure Page
3-1. Trends in NOa air quality and in HC and N0» emission, Los
Angeles Basin, 1965-74 25
3-2, Pollutant trends in Los Angeles, annual and 3-year moving
averages , 25
3-3. Correlation between Bscat and maximal ozone concentration 26
4-1. Chemical changes occurring during photoirradiation of hydro-
carbon-nitrogen oxide-air systems 30
4-2. The major reaction paths for the degradation of frans-2-butene
in an irradiated NOn-polluted atmosphere 32
4-3, Isopleths (m x 102) of mean summer morning mixing heights 38
4-4, Isopleths (m x 102) of mean summer afternoon mixing heights 38
4-5. Isopleths (m x sec 1) of mean summer wind speed averaged
through the afternoon mixing layer 39
4-6. Isopleths (m x sec 1) of mean summer wind speed averaged
through the morning, mixing layer 39
4-7. Isopleths of total number of forecast days of high meteoro-
logical potential for air pollution 40
4-8 Monthly distribution of number of cases of 4 or more days of
atmospheric stagnation, 1936-75 (September) 41
4-9. Monthly distribution number of cases of 4 or more days of
atmospheric stagnation, 1936-75 (August) , 41
4-10. Diurnal variations m actinic irradiance 42
4-11. Long-term ozone variations at Quillayute • 50
4-12. Monthly ozone variations at Mauna Loa, Hawaii 51
4-13. Average monthly ozone concentrations recorded at summit of
Mount Whiteface 52
4-14. Effect of NO« concentration increase on ozone formation at
White River, Utah 55
5-1. Distribution of hydrocarbons in diesel exhaust gas 67
5-2. Distribution of hydrocarbons in jet-aircraft engine exhaust . 68
5-3 Nonmethane hydrocarbon trends in Los Angeles, 1963-72 . 70
5-4. Oxides of nitrogen trends in Los Angeles, 1963-72, 6-to9-a.m.
and maximum 1 -hr concentrations 77
5-5. Correlation of observed and calculated Oa reactivities 87
6-1. Maximum daily 1 -hr average oxidants as a function of 6- to 9-
a.m. averages of nonmethane hydrocarbon (CAMP data from
four U.S. cities 96
6-2, Upper limit oxidant values in the Los Angeles south coast air
basin as a function of average 6- to 9-a.m, concentration 97
6-3, Oxidant-hydrocarbon relationships from smog chamber and
aerometric data , 97
-------
Page
6-4, Merz, Painter, and Ryason's relation of NO* and NMHC assumed
as 50 percent of total HC and oxidant for downtown Los
Angeles , 99
6-5. California Air Resources Board aerometric results, relation
between 6- to 9-a.m. NO*, 6- to 9-a.m. HC, and maximum
hourly oxidant concentrations in downtown Los Angeles . 99
6-6, California Air Resources Board aerometric results, relation
between 6- to 9-a.m. NO,, 6- to 9-a.m. HC, and maximum
hourly oxidant concentration in Azusa 100
6-7. California Air Resources Board aerometric results, relation
between 6- to 9-a.m. NO*, 6- to 9-a.m. HC, and maximum
hourly oxidant concentrations in San Bernardino 100
6-8, California Air Resources Board aerometric results, relation
between 6- to 9-a.m. NO,, 6- to 9-a.m. HC, and maximum
hourly oxidant concentrations in Anaheim 101
6-9. Expected number of days per year exceeding 0.10 ppm versus
NO* and reactive hydrocarbon emission levels for central
Los Angeles 102
6-10. Oxidant/Os isopleths derived from combined use of smog
chamber and photochemical and modeling techniques ... 103
7-1. Diurnal ozone-oxidant averages from September 4 through
September 30, 1971 123
7-2. Ozone-oxidant ratios in Musashino, Japan 1 24
7-3. Comparison of the Kl, Mast, and chemiluminescence methods
for measuring ozone in irradiated exhaust mixtures 1 24
1-&, Relationship of total oxidant and ozone values greater than 10
ppb, August through December 1973 1 25
7-&. Comparative measurements of ambient oxidant/ozone by the
potassium iodide (Kl) chemiluminescence (Chem), and
ultraviolet (UV) methods 1 26
9-'. Changes with time in dynamic lung function tests in non-
smokers during a 2-hour exposure to 0.75 ozone and during
recovery 189
9-,'!, Changes with time of minute ventilation, respiratory rate, and
tidal volume in ozone-exposure (0.75 ppm) and subsequent
recovery period in exercising nonsmokers 1 90
9-3. Dose-response curves for Los Angeles and Montreal subjects 195
10-1. Comparison of deaths of persons aged 65 years and over, and
maximum daily temperature, Los Angeles County, July 1 to
November 30, 1955 202
10-2. Comparison of nursing home deaths, maximum daily tempera-
ture, and smog-alert days in Los Angeles County, July
through December 1955 203
10-3. Comparison of maximum concentrations of oxidant and carbon
monoxide, maximum temperature, and daily death rate for
cardiac and respiratory causes, Los Angeles County, 1956-
58 204
10-4. Monthly changes in symptom and complaint rates with daily
maximum hourly oxidant content above and below 0.10 ppm 21 8
10-5. Comparison of various symptom and complaint rates on days
when oxidant concentrations were over 0.15 pphm and
below 0.10 ppm 219
-------
Page
10-6. Relationship between oxidant concentration in the hour before
an athletic event and percentage of the team failing to im-
prove running time 222
10-7, Regression curves relating eye irritation scores and simul-
taneous oxidant concentrations from a number of stations
in the Los Angeles area, 1954 231
10-8. Variation of mean maximum eye irritation, as judged by a panel
of scientists, with maximum oxidant concentrations,
Pasadena, August-November 1955 232
10-9. Mean index of eye irritation versus oxidant concentration .. 234
10-10. Monthly regression lines between oxidant levels and eye
irritation 235
10-11. Combined weekly incidence rates of selected conditions
(colds, hay fever, asthma, and other respiratory conditions)
in persons of all ages, Los Angeles County and the remainder
of California, August 7 to November 28, 1954 240
10-12. Weekly incidence rates of illness and injury for persons aged
65 and over in Los Angeles County and the remainder of
California, August 2 to November 28, 1954 ^. 240
10-13. Percentage of respondents in Los Angeles County and San
Francisco Bay area attributing exacerbation of respiratory
conditions to air pollution and other factors, 1 956 242
11-1. Sequence of ozone-induced responses 254
11-2. Relation between ozone concentration, foliar injury, and a
reduction in plant growth or yield 264
11-3. Limiting values for foliar injury to trees and shrubs by ozone 265
11 -4. Limiting values for foliar injury to agricultural crops by ozone 266
11-5. Ozone concentration versus duration of exposure required to
produce a 5% response in three different plant susceptibility
groupings 267
11-6, Dose-response relationships and limiting values for foliar
injury to vegetation by peroxyacetylnitrate (PAN) 269
11-7. Conceptual model of factors involved in air pollution effects
on vegetation ., , 271
12-1. Law of tolerance 295
12-2. Major topographic features of the Los Angeles Basin witn
inland valleys and mountains 299
12-3. Altitudinal sequence of ecosystems in the San Bernardino
Mountains 300
12-4. Daytime changes in oxidant concentrations along a west-to-
east transect in the southern coastal air basin, including the
slopes of the San Bernardino Mountains 3Q1
12-5. The relationship of time of occurrence of the daily peak oxidant
concentration to temperature and vapor-pressure gradients
in an elevational sequence 302
12-6. Monthly summation of'total oxidant concentration data at Rim
Forest/Sky Forest, San Bernardino Mountains, California,
June through September, 1968-75 305
12-7. Number of hours of total oxidant, July through September,
greater than or equal to 392 fjg/m3 (0.20 ppm) at the down-
town San Bernardino County Air Pollution Control District
Station, 1963-74, and Rim Forest/Sky Forest, 1968-74 .. 306
-------
Page
12-8. Hypothetical relationships of oxidant dose, natural ecosystems,
agroecosystems and the relative values of their vegetation
components along a transect of southern coastal air basin of
California 307
12-9. Organism-level interactions in a mixed conifer forest 309
12-10. Tree-level interactions in a mixed-conifer forest ecosystem . 311
12-11. Community-level interactions in a mixed-conifer forest
ecosystem 313
12-12. Stand-level interactions in a mixed-conifer forest ecosystem 315
12-13. Community-succession interactions in a mixed-conifer forest
ecosystem 317
12-14. Topographic projection, San Bernardino Mountains, with
comparison of oxidant injury to black oaks at major study
sites, August 31,1974, with accumulated total oxidant dose
for June-August measured at nearby monitoring stations 318
12-15. Topographic projection, San Bernardino Mountains, showing
how ponderosa pines and Jeffrey pines in major study sites
are distributed in six injury classes according to seasonal
dose of total oxidant 318
12-16. Annual growth of the terminal shoot and first-order branches
in upper half of ponderosa pine saplings maintained in
filtered or unfiltered (ambient) air greenhouses, or in outside
ambient air (1968-73) 320
12-17. Relationship of degree of oxidant injury in ponderosa pines
with bark beetle attack and number of trees killed by western
pine beetle, mountain pine beetle or both species 323
13-1. Effect of annual average ozone concentration on added costs
resulting from damage to materials and preventive meas-
ures 340
-------
TABLES
Table Page
3-1. Summary of maximum oxidant concentrations in selected
urban areas 20
3-2. Oxidant concentrations observed in selected urban areas of,
the United States, 1974-75 21
3-3. PAN and oxidant (0») measurements (10 a.m. to 4 p.m.), Los
Angeles, Calif,, 1968 21
3-4. PAN and ozone measurements (10 a.m. to 4 p.m.), Hoboken,
N.J., 1970 22
3-5. PAN, oxidant, and ozone measurements (10 a.m. to 4 p.m.), St.
Louis, Mo., 1973 22
3-6. Concentrations of tropospheric ozone before 1962 ...., 22
3-7. Summary of ozone data from 1973-1975 oxidant studies.
Research Triangle Institute 23
4-1. Calculated rates of attack on fram-2-butene by various reactive
species in simulated smog system at several irradiation
times 31
4-2. Compounds observed in photochemical smog 35
4-3. Compounds that may be formed in photochemical smog .... 35
4-4. Secondary organic aerosols 36
4-5. Relative importance of aliphatic and aromatic precursors ... 37
4-6. Percentage difference between the Peterson and the Leighton
values at the Earth's surface over selected wave-length
intervals and solar zenith angles 42
4-7. Calculated values of NO2 photodissociation rate constant at
Earth's surface and percentage increase of the rate constant
from the surface to various heights 43
5-1. Hydrocarbons identified in ambient air 66
5-2. Frequency distributions for 6- to 9-a.m. nonmethane hydro-
carbon concentrations at CAMP sites, 1 967-72 69
5-3. Volatile plant products identified by Rasmussen 70
5-4. Frequency distribution data for 6- to 9-a,m. nitric oxide
concentrations at CAMP sites, 1962-72 72
5-5. Frequency distribution for 6- to 9-a.m. nitrogen dioxide con-
centrations at CAMP sites, 1962-72 73
5-6. Nitric oxide concentration in California by averaging time and
frequency, 1963-67 75
5-7. Nitrogen dioxide concentration in California by averaging time
and frequency, 1963-67 76
5-8. Estimates of total hydrocarbon emissions from manmade and
natural sources 77
5-9. Total yearly hydrocarbon emission rate for the continental
United States, based on leaf biomass 77
-------
Page
5-10. 1974 nationwide estimates of total hydrocarbon sources and
emissions ., 78
5-11. Nationwide total hydrocarbon emission trends, 1970-75 ... 78
5-12, Hydrocarbon emissions from mobile sources 78
5-13. Organic emission inventory for the metropolitan Los Angeles
air quality control region, 1972 79
5-14. Reactive emission inventories for the metropolitan Los Angeles
air quality control region '. 80
5-15. 1974 nationwide estimates of nitrogen oxide sources and
, emissions 83
5-16.-Comparison of reactivities of different types of organics 84
5-17. Reactivity data 84
5-18. Summary of data from studies on reactivities (toluene equivalents)
and classification of organics 85
5-19. Classification of organics with respect to their oxidant-related
reactivity in urban atmospheres 86
6-1, Forms of air quality simulation models 107
6-2. Currently available air quality simulation models for photo-
chemical oxidant 108
6-3. Treatment of meteorological variables in air quality simulation
' models 109
6-4. Uncertainties in predicted ozone concentrations from all
mechanisms and input uncertainties 111
6-5, Prior validation studies of air quality simulation models 112
7-1. Performance specifications for automated methods 118
7-2. List of designated reference methods ., 119
7-3. List of designated equivalent methods 119
7-4. Percent difference from known concentrations of nonmethane
hydrocarbons obtained by sixteen users 127
8-1. Pulmonary effects of ozone; Host defense mechanisms .,.,, 166
8-2. Pulmonary effects of ozone: Biochemistry ................. 167
8-3. Pulmonary effects of ozone: Morphology , 170
8-4. Pulmonary effects of ozone: Pulmonary function 171
8-5. Pulmonary effects of ozone: Edema and tolerance 172
8-6. Extrapulmonary effects of ozone: Hematology and serum
chemistry 173
8-7. Extrapulmonary effects of ozone: Central nervous system and
behavior 174
8-8. Extrapulmonary effects of ozone: Morphology 1 74
8-9, Extrapulmonary effects of ozone: Miscellaneous 1 74
8-10. Atmospheric mean concentrations and their standard deviations
administered from 8 a.m. to midnight each day 175
8-11. Effects of oxidants on animals 176
9-1. Summary of available data on occupational exposure of
humans to ozone 185
9-2. Summary of data on human experimental exposure to ozone
before 1970 187
1G-1. Average number of deaths per day resulting from cardiac and
respiratory causes among residents of Los Angeles County
aged 65 and over, as related to oxidant concentrations and
maximum daily temperature, by month 1954-55 201
-------
Paae
10-2. Comparison of respiratory and cardiac deaths during smog and
smog-free periods occurring in Los Angeles County, August-
November 1954 206
10-3. Correlation of morning and early afternoon oxidant levels with
oxygen consumption and airway resistance of 15 patients
with chronic respiratory disease 211
10-4. Relationship of average daily simple and adjusted percentage
of student nurses report to photochemical oxidant levels
for 868 days, November 1961 through May 1964 213
10-5. Sign-test data for testing the association of oxidant levels with
accidents in Los Angeles, August through October, 1963
and 1965 224
10-6. Number of subjects for whom correlation coefficients between
environmental and respiratory function measurements are
significant at p < 0.05, Japan, 1972 227
10-7. Correlation of eye irritation with simultaneous oxidant con-
centrations, in order of decreasing eye irritation score, for
a number of stations in the Los Angeles area, 1954 231
10-8. Correlation between eye irritation and simultaneous environ-
mental measurements, as judged by a panel of scientists, 1954 232
10-9. Effect of filter on sensory irritation and chemical measurements 233
10-10. Pearson product moment correlation of coefficients between
eye irritation and environmental factors in a nonfiltered
room 234
10-11. Total lung cancer mortality in an American Legion study popu-
lation, California, 1958-62 237
10-12. Lung cancer deaths and relative risks per 100,000 man-years
in an American Legion study population, by extent of cigarette
smoking and residence, California, 1958-62 2b8
10-13. Total chronic respiratory disease mortality in an American
Legion study population, California, 1958-62 239
10-14. Selected respiratory conditions reported by general population
sample, California, May 1956 241
10-15. Percent of survey response of general and working population
bothered by air pollution, by major geographic areas in
California, May 1956 246
10-16. Air pollution effects reported in general population survey, by
type of community and by major geographic areas in Califor-
nia, May 1956 247
11-1. Effects of oxidants (ozone) in ambient air on growth, yield, and
foliar injury in selected plants 258
11-2. Effects of acute exposure on growth and yield of selected
plants 260
11-3. Effects of ozone on selected understory species from
an aspen community 261
11-4. Effects of long-term, controlled ozone exposures on
growth, yield, and foliar injury to selected plants 262
11 -5. Concentration, time, and response equations for three suscep-
tibility groups and for selected plants or plant types with
respect to ozone 268
11-6. Ozone concentrations for short-term exposures that produce
5 or 20 percent injury to vegetation grown under sensitive
conditions 269
-------
Page
11-7. Estimates of economic losses to crops and vegetation in the
United States attributable to oxidant air pollution ., 270
11-8. Response of plants to ozone, as conditioned by humidity during
growth and exposure 274
11-9. Effects of soil moisture on response of selected plants to
oxidant stress 275
11-10. Effects of various nutrients on response of selected plants to
ozone (oxidant) stress 276
11-11. Summary effects of sulfur dioxide and ozone mixtures on foliar
injury , 277
11-12. Foliar response of selected plants to sulfur dioxide and ozone
mixtures 279
11-13. Growth response of selected plants to sulfur dioxide and ozone
mixtures 279
12-1. Composition of ecosystems 294
12-2. Descriptions of meteorologic patterns for five classes of spring
and summer days in southern California 304
12-3. Changes in oxidant injury scores and mortality rates of ponder-
osa and Jeffrey pines at 18 major study plots, 1973-74 .. 319
12-4. Changes of timber volume and percentage of total Jeffrey
pines at Barton Flats in the San Bernardino National Forest 319
12-5. Tree species and size composition in a study area affected by
oxidant air pollution 321
12-6. Injury thresholds for 2-hour exposures to ozone 322
13-1. Formulation of highly ozone-sensitive natural rubber 331
13-2. Effect of ozone on natural rubber 331
13-3. Tire sidewall formulation 332
13-4. Effects of ozone on sidewall formulations containing various
antiozonant concentrations 332
13-5. Cracking rates of white sidewall tire specimens 332
13-6. Paint erosion rate coefficients for the effects of ozone in
laboratory-controlled environments 338
13-7. Paint erosion rates at field exposure sites , 339
-------
ABBREVIATIONS AND SYMBOLS
A
AaDo,
AChE
AlChE
AM
APCD
AQSM
Ar
A203
ATPase
Avg
BAKI
bdft
7Be
Br2
Bscat
°C
14C
CAMP
GARB
Ci-C3
Ci-C4
C2+
C2H2
C4+
C4-Ce
C6H5COO2NO2
CC
v-*dyn
CH(CH3)CHO
CHsCHCHO
CH2CH2
CH2CO
CH2O
CH3
CHaCHO
CH3CO02N02
CH4
Chem
CI2
Angstrom
Alveolar-arterial gradient
Acetylcholinesterase
American Institute of Chemical Engineers
Alveolar macrophages
Air pollution control district
Air quality simulation model
Argon
Arsenous oxide
Adenosine triphosphatase
Average
Potassium iodide solution acidified with boric acid
Board foot
Radioisotope of beryllium
Bromine
Extinction coefficient due to scatter by aerosols
Degrees Celsius
Radioisotope of carbon
Community Air Monitoring Program
California Air Resources Board
Hydrocarbons containing one to three carbon atoms
Hydrocarbons containing one to four carbon atoms
Hydrocarbons containing more than two carbon
atoms
Acetylene
Hydrocarbons containing more than four carbon
atoms
Hydrocarbons containing four or five carbon atoms
Hydrocarbons containing four to six carbon atoms
Hydrocarbons containing five to ten carbon atoms
Benzyl
Peroxybenzoylnitrate (PBzN)
Closing capacity
Dynamic compliance
Propionaldehyde
Acrolein
Ethylene
Ketene
Formaldehyde; also HCHO and H2CO
Methyl group
Acetaldehyde
Peroxyacetylnitrate (PAN)
Methane
Chemiluminescence
Chlorine
-------
cm
cm3
CNS
CO
COz
Coh
COMT
cone.
COOH
COONO2
Cro3
Cs,
Cyt. P45o
2,4-D
dbh
DEN
DiFKIN
DLco
DNA
DO LA
DW
EA
EC
EH
EKMA
EPA
ESR
Et
F,
F2
op
FEF 25-50%
FEF 25-75%
FEF 50%
FEFmax
Fe2O3
FEV
FEVo.75
FEV,,0
FEV3,0
FID
FRC
FS
FT
ft
ft2
ft-c
FVC
9
Gaw/Vtg
GC
GDI
gf
GH
Centimeter
Cubic centimeter
Central nervous system
Carbon monoxide
Carbon dioxide
Coefficient of haze
Catechol-o-methyltransf erase
Concentration
Carboxyl group
Peroxynitrate group
Chromium trioxide (chromic anhydride, chromic acid)
Static lung compliance
Cytochrome Paso
2,4-DichIorophenoxyacetic acid
Diameter breast height
Diethylnitrosamine
Diffusion Kinetics Model
Ccirbon monoxide diffusing capacity of lung
Deoxyribonucleic acid
Downtown Los Angeles
Dry weight
Environment Agency (Japan)
Electron capture detection
Humidity during exposure
Empirical Kinetic Modeling Approach
U, S. Environmental Protection Agency
Electron spin resonance
Ethyl
First filial generation
Second filial generation
Degrees Fahrenheit
Mean expiratory flow during middle half of FVC
Maximal midexpiratory flow rate
Instantaneous expiratory flow at 50% of FVC
Peak expiratory flow
Ferric oxide
Forced expiratory volume
0.75-second forced expiratory flow
1 -second forced expiratory flow
3-second forced expiratory flow
Flame ionizatjon detection (or detector)
Functional residual capacity
Fourier-transform spectroscopy; also FT
Fourier-transform spectroscopy; also FS
Foot
Square foot
Foot-candle
Forced vital capacity
Gram
Specific conductance
Gas chromatography
Gas distribution index
Emission growth factor
Humidity during growth
-------
GLC
GM
G-6-P
G-6-PD
GPT
GSC
GSH
GSSG
GSSRase
H
H+
3H
ha
HC
HCHO
HCOOH
He
Hg
HI
HMP
HN02
HN03
HO
H02
HONO
HON02
H02NO
H02N02
HOON02
hr
hv
H2CO
H20
H202
H2S
H2S04
I
I,
la"
131,
in.
IR
J*
k
K
K+
KBr
kg
KH
Kl
K,
km
kPa
kV
Gas-liquid chromatography
General Motors Corporation
Glucose-6-phosphate
Glucose-6-phosphate dehydrogenase
Gas-phase titration
Gas-solid chromatography
Reduced glutathione
Oxidized glutathione
Glutathione reductase
Hydrogen
Hydrogen ion
Tritium
Hectare
Hydrocarbons
Formaldehyde; also CH20 and H2CO
Formic acid
Helium
Mercury
Hemagglutination inhibition
Hexose monophosphate
Nitrous acid; also HONO
Nitric acid; also HON02
Hydroxyl group; also OH
Hydropheroxy radical
Nitrous acid
Nitric acid
Pernitrous acid
Pernitric acid; also HOON02
Pernitric acid
Hour
Photon
Formaldehyde; also HCHO and CH20
Water
Hydrogen peroxide
Hydrogen sulfide
Sulfuric acid
Intensity of sunlight
Iodine
Triiodide ion
Radioisotope of iodine
Inch
Infrared
Actinic irradiance
Rate constant; or, dissociation constant
Potassium
Potassium ion
Potassium bromide
Kilogram
Horizontal eddy diffusivity coefficient
Potassium iodide
Photodissociation rate constant
Kilometer
Kilopascal
Kilovolt
-------
Ky
LA
LARPP
ib
LDH
LiRAQ
I/min
In
LT50
Ix
M
m
m
m2
m3
MAO
max
mb
MBC
MCP
Me
MEFR
mg
mi
mi2
ml
min
min.
mm
MMC
MMEF
mmol
mo
mph
MS
n
H2
AN2
n/a
NAD
NAAQS
NADB
NADH
NADP*
NADPH
NAS
NATO/CCMS
NBKi
NBS
ND
Vertical eddy diffusivity coefficient
Los Angetes
Los Angeles Reactive Pollutant Project
Pound
Dose lethal to 50% of recipients
Lactic acid dehydrogenase
Livermore Regional Air Quality (Model)
Liters per minute
Natural logarithm
Time in which 50% of recipients of stated dose died
Lux
Third body (in a reaction)
Meter
Meta
Square meter
Cubic meter
Monoamine oxidase
Maximum
Millibar
Maximum breathing capacity
McKee, Childers, and Parr (test)
Methyl
Maximum expiratory flow rate
Milligram
Mile
Square mile
milliliter
Minute
Minimum
Millimeter
Mean meridional circulation
Mid-maximal expiratory flow rate
Millimole
Manganese dioxide
Month
Miles per hour
Mass spectroscopy
Normal
Nitrogen
Single breath Na alveolar plateau
Not applicable
Nicotinamide adenine dinucleotide
National Ambient Air Quality Standard
National Air Data Bank
Reduced nicotinamide adenine dinucleotide
Nicotinamide adenine dinucleotide phosphate
Reduced nicotinamide adenine dinucleotide
phosphate
National Academy of Sciences
North Atlantic Treaty Organization/Committee on
Challenges of Modern Society
Neutral buffered potassium iodide
National Bureau of Standards
No data; or, not detectable
-------
Ne
NEDS
NH3
nm
NMHC
NO
N02
NO*
N20
N205
NPSH
o
0
02
03
OAQPS
O.D.
0('D)orO(1A)
OH
OH°
03.0lefin
0(3P)
0,
P
P
32P
PAN
Pa02
PAQ
PBN
PBzN
6-PG
6-P-GD
PH
P.soBN
PMN
PNH
Po,
PP
ppb
ppb C
pphm
ppm
ppm C
PPN
ppt
psi
PST
Pst TLC
r
R
RAPS
Raw
Neon
National Emissions Data System
Ammonia
Nanometer
Nonmethane hydrocarbons
Nitric oxide
vi;trogen dioxide
Nitrogen oxides
Nitrous oxide
Nitrogen pentoxide
Nonprotein sulfhydryls
Ortho
Atomic oxygen
Oxygen
Ozone
Office of Air Quality Planning and Standards, EPA
Orthostatic dysregulation
Excited atomic oxygen
Hydroxyl group; also HO
Hydroxyl radical
Ozonides
Ground-state atomic oxygen
Photochemical oxidants
Probability (of observed results occurring by chance)
Para
Radioisotope of phosphorus
Peroxyacetylnitrate
Arterial oxygen tension
Present air quality
Peroxybutyrylnitrate
Peroxybenzoylnitrate
6-phosphogluconate
6-phosphogluconate dehydrogenase
Measure of acidity (negative logarithm of hydrogen
ion concentration)
Peroxyisobutyrylnitrate
Polymorphonuclear leukocytes
Paroxysmal nocturnal hemoglobinuria
Oxygen tension
Photoperiod
Parts per billion
Parts per billion expressed as carbon
Parts per hundred million
Parts per million
Parts per million expressed as carbon
Peroxypropionylnitrate
Parts per trillion
Pounds per square inch
Pacific standard time
Maximal transpulmonary pressure
Correlation coefficient
Roentgen
(St. Louis) Regional Air Pollution Study
Airway resistance
-------
RBC
RCHO
RC(0)02
RCOO2
RCOO2H
RC(O)02NO2
ROW
RH
RHC
RL
RO2
RONO2
RO2NO2
ROOM
rpm
Rt
RTI
RTP
RV
S
-S-
SAI
SBR
S.D.
sec
SEM
S.F.
SGaw
SGPT
SH
SMSA
S02
S03
SO,
SOD
sp. or spp.
SPM
90Sr
SRI
SRM
-SS-
SSET
STA
Std
t
TDW
TF
TGV
TLC
UKi
UV
V25
V50
VC
Red blood cells; erythrocytes
Higher aldehydes
Peroxyacyl radical
Organic peroxy compounds
Organic peracids
Peroxynitrates
Root dry weight
Relative humidity
Reactive hydrocarbons
Pulmonary resistance
Alkoperoxy radical
Organic nitrates
Organic peroxynitrates
Organic hydroperoxides
Revolutions per minute
Flow resistance
Research Triangle Institute
Research Triangle Park
Residual volume
Sulfur
Sulfide linkage
Systems Applications, Inc. (Model)
Styrene-butadiene rubber
Standard deviation
Second
Scanning electron microscopy
San Francisco
Specific airway conductance
Serum glutamic-pyruvic transaminase
Sulfhydryl group
Standard metropolitan statistical area
Sulfur dioxide
Sulfur trioxide
Sulfur oxides
Superoxide dismutase
Species
Suspended particulate matter
Radioisotope of strontium
Stanford Research Institute
Standard reference material
Disulfide linkage
Small-scale eddy transport
Seasonal tropopause adjustment
Desired air quality (standard)
Student's statistic
Top dry weight
Tropopause folding
Thoracic gas volume
Total lung capacity
Unbuffered potassium iodide
Ultraviolet
Maximum expiratory flow rate at 25% of VC
Maximum expiratory flow rate at 50% of VC
Vital capacity
-------
VMT
Vo,
V02 max
v/v
W
wk
wt
XO, X02
a
or ptmole
Vehicle miles traveled
Oxygen uptake
Maximum oxygen uptake
Volume/volume
Watts
Week
Weight
Compounds in which X represents hydrogen or an
organic radical (R or RO)
Year
Level of statistical significance set by investigator
Microgram
Microgram per cubic meter
Micrometer
Micromole
Greater than
Less than
Approximately
-------
ABSTRACT
This document is an evaluation and assessment of scientific information
relative to determination of health and welfare effects associated with exposure
to various concentrations of ozone and other photochemical oxidants in ambient
air. The-document is not intended as a complete, detailed literature review. It
does not cite every published article relating to oxidants in the environment and
their effects. The literature through 1976-1977 has been reviewed thoroughly
for infprmation relative to criteria. An attempt has been made to identify the
major discrepancies in our current knowledge, again relative to criteria.
Though the emphasis is on presentation of health and welfare effects data,
other scientific data are presented and evaluated to provide a better
understanding of the pollutants in the environment. To this end, separate
chapters are included on the nature and atmospheric concentrations.of
photochemical oxidants, their sources and removal processes, oxidant
precursors, the relationships between ambient oxidants and precursor
emissions, and measurement methods for ozone, oxidants, and their precursors.
Specific areas addressed within a general area of health or welfare effects are
as follows: Toxicological, clinical, and epidemiologic appraisals, and effects on
vegetation, ecosystems, and materials.
-------
-------
1. SUMMARY AND CONCLUSIONS
INTRODUCTION
This document consolidates and assesses
current knowledge regarding the origin of ozone
and other photochemical oxidant pollutants and
their effects on health, vegetation, certain
ecosystems, and materials. This chapter
summarizes the information contained in this
document and includes conclusions that are
believed to provide a reasonable basis for
evaluating the effects on health or welfare that are
produced by various concentrations of ozone and
other photochemical oxidants. Although nitrogen
dioxide is considered one of the photochemical
oxidants, oxides of nitrogen are the subject of a
separate report and are therefore discussed in this
document only as they participate in the formation
and reactions of other photochemical oxidants.
Hydrocarbor s (HC) and other organics are
important air nollutants almost entirely because
they are preci1 -sors of other compounds formed in
the atmosph sric photochemical system, not
because they produce any direct effects
themselves, "i axic organics are considered only
with respect to eye irritation.
The studies and data cited constitute the best
available basii for specific standards aimed at
protecting hu nan health and the environment
from photochemical oxidants in ambient air.
NATURE AND ATMOSPHERIC CONCEN-
TRATIONS OF PHOTOCHEMICAL OXIDANTS
Photochemical oxidants are products of at-
mospheric reactions involving organic pollutants,
nitrogen oxides (NO*), oxygen, and sunlight. They
consist mostly of ozone, NOz, and peroxy-
acetylnitrate (PAN), with smaller amounts of other
peroxyacylnitrates and other peroxy compounds,
and they are formed along with other
photochemical products such as aldehydes,
nitrous acid, nitric acid, and formic acid.
Photochemical oxidants originate mainly from
volatile organic and NOX emissions produced by
human activities. Photochemical oxidant
formation is a complex function of emissions and
meterological patterns.
Peak concentrations of oxidant, expressed as
ozone, are generally higher in urban and suburban
areas than in rural areas, reaching levels i excess
of 590 /ug/m3 (0.3 ppm). In rural areas, peak
concentrations are lower but often exceed the 1-hr
National Ambient Air Quality Standard of 160
/ug/m3 (0.08 ppm). However, dosages or average
concentrations in rural areas are comparable to or
even higher than those in urban areas. Because of
pollutant transport, oxidant pollution is a regional
rather than a local problem.
SOURCES AND SINKS OF OXIDANTS
All of the evidence presently available shows
that in and around urban centers that have severe
oxidant/ozone pollution, photochemical oxidant
formed from anthropogenic organics and NOX is
the major contributor. Since Air Quality Criteria for
Photochemical Oxidants was issued in 1970, the
mechanisms of atmospheric oxidant/ozone for-
mation have been studied intensively and are now
understood in greater detail. Most noteworthy are
recent findings pertaining to the roles of hydroxyl
(OH-) and hydroperoxy(HO-z) radicals. The reaction
with OH- has been established to be a major-
hydrocarbon-consuming process, and OH- and
H0'2 have been identified as having major roles in
the atmospheric oxidation of NO to NOz. Aldehydes
and PAN have also been found to play important
roles in the atmospheric reaction. For olefins and
paraffins, at least, these reactions are now
understood to the extent that the kinetics of
photochemical hydrocarbon-NOx reaction
systems, as observed in the laboratory, can be
described with reasonable accuracy. Additional
research is needed to explain the atmospheric
reactions of aromatic hydrocarbons and to clarify
further the differences between laboratory and
ambient atmospheric chemical systems.
The photochemical formation of oxidant/ozone
is the result of two coupled processes: (1) a
physical process involving dispersion and
transport of precursors to oxidants (e.g., HC and
-------
NO,), and {2} the photochemical reaction process.
Both processes are strongly influenced by
meteorological factors such as dispersion, solar
radiation, temperature, and humidity. Recent data
on wind velocity and mixing height show that
episodes of limited dispersion are most common in
the Far West and in the Rocky Mountains region,
least common over the Plains States, and of
intermediate frequency east of "the Mississippi.
New measurements of solar radiation indicate that
the distribution of light intensity among respective
wavelength intervals is different from that
previously reported, which results in higher
photodissociation constants for NOz in the ambient
atmosphere. Recent field and laboratory studies
suggest that at temperatures below approximately
55° to 60°F, concentrations of photochemical
ozone are unlikely to exceed the national 1-hr
standard of 160//g/m3 (0.08 ppm).
For understanding the sources of photochemical
oxidants or ozone (oxidant/ozone), perhaps the
most important recent development is the
identification of short-and intermediate-range and
synoptic-scale transport of photochemical
oxidant/ozone. Short-range (urban-scale)
transport causes the highest ozone concentrations
some distance downwind from the core area
(region of highest emission) of an urban center.
Intermediate-range (mesoscale) transport occurs
as urban oxidant/ozone plumes extending as far
as 100 miles or more downwind and is also
involved in land-sea breeze circulation. Finally,
synoptic-scale transport over several hundred
miles, associated with high-pressure systems, has
been found to occur extensively. These findings
have significant implications with respect to the
location of oxidant/ozone monitoring stations.
Conditions during long-range transport are such
that ozone production per unit of precursor is
enhanced. Also, many organics previously thought
to be unreactive are now believed to have
significant ozone-producing potential.
In addition to photochemical reactions of
anthropogenic emissions, potential sources of
oxidant/ozone in the troposphere are the intrusion
of stratospheric ozone and the photochemical
reactions of natural organic and N0« emissions.
Estimates of ground-level concentrations of ozone
originating in the stratosphere are based on two
types of evidence: (1) global circulation patterns,
namely, patterns in air interchange between
Stratosphere and troposphere; and (2) data on
variations of ozone concentrations in remote rural
areas. Based on the evidence of stratosphere-
troposphere interchange, the annual average
stratospheric contribution to ozone con-
centrations at ground level is estimated to be 43 to
98 fjg/m3 (0.022 to 0.05 ppm). The highest
concentrations, at or above 160 /ug/m3 (0.08 ppm),
from that source are expected to occur mainly
during April and May. Analysis of ozone data for
rural areas indicated that major intrusions of
stratospheric ozone also occur during the spring
months in mid-latitudes. The highest 1-hr
concentration of stratospheric ozone reaching
ground level during the smog season (usually late
summer or early fall) has been found to range from
29 to 78 //g/m3 (0,015 to 0.040 ppm), depending
on the investigator. More recent data obtained at
Whiteface Mountain, New York, suggest a
maximum 24-hr concentration of 72^g/m3 (0.037
ppm) stratospheric ozone in July,
Certain organic emissions from vegetation
(terpenes) play the dual role of oxidant/ozone
precursor and scavenger. Despite the substantial
rates at which they are emitted in forested areas,
the ambient concentrations of Such organics,
because of their reactivity and the areal dispersion
of their sources, seldom exceed a few parts per
billion (ppb). At these concentrations, the direct
potential of terpenes for photochemical ozone
formation is estimated to be negligible. It is
conceivable, however, that the products of
atmospheric reactions involving large amounts of
terpenes do have a significant impact on
oxidant/ozone-related air quality.
OXIDANT PRECURSORS
Organic pollutants in urban atmospheres
consist mainly of hydrocarbons from automobile
exhaust and fuel evaporation, and of oxygenated
hydrocarbons and halocarbons emitted mainly
from manufacturing and from the use of organic
chemicals. In urban atmospheres, total organic
concentrations, reported usually as 6- to 9-a.m.
averages of total nonmethane hydrocarbons
(NMHC), are typically in the 1 -ppm range and can
be as high as 10 ppm or even higher, as in Los
Angeles. In rural and remote atmospheres, the
composition and concentration of organic pollut-
ants are uncertain, largely because of deficiencies
in the analytical methods available for determining
the concentrations involved. Evidence suggests
that NMHC levels are generally less than 0.1 ppm,
a fraction of which consists of vegetation-related
terpenes.
Concentrations of N0« in urban atmospheres
vary within a wide range, with highest values
-------
exceeding 1 ppm. Such concentrations appear to
decrease rapidly as the air mass moves away from
the city In rural and remote areas, ambient
concentrations do not exceed a few ppb and are
often below the 5 ppb detection limit of current
commercial N0» analyzers.
Hydrocarbons and NO* are emitted to the
atmosphere from both natural and manmade
sources, with natural sources probably con-
tributing more on a global scale. Natural and
anthropogenic sources are, however, generally
segregated geographically, so that anthropogenic
emissions are concentrated in the populated ur-
ban areas. Thus it is the latter that are most
relevant to oxidant/ozone pollution problems
downwind from populous areas. At present,
mobile sources appear to account for the major
part of the organic compounds emitted in most
urban areas. The imposition of emission standards
for mobile sources has reduced the reactive
hydrocarbon component of emissions from
gasoline-powered vehicles.
RELATIONSHIPS BETWEEN AMBIENT
OXIDANT AND PRECURSOR EMISSIONS
Quantitative relationships between ambient
oxidant/ozone and precursor emissions are
needed for predicting the impact of these
emissions on air quality. Such relationships
represent, with varying degrees of complexity, the
physical and chemical processes taking place in
the atmosphere.
Air quality simulation models (AQSM's), a
number of which are available, represent an
approach to relating precursor emissions to
oxidant air quality. However, AQSM's require a
great deal of computation and have not been
adequately evaluated.
In summary, satisfactory methods of relating the
emission of precursors to oxidant concentrations
in ambient air are notyetavailable. For this reason,
the U.S. Environmental Protection Agency has
been sustaining an intensive research effort to
develop and validate satisfactory models that
relate emissions to air quality. Based on results of
this research and on model performance standards
yet to be defined, the Agency will recommend
specific models for use in formulating optimum
ozone control strategies.
MEASUREMENT METHODS FOR OXIDANT
AND OXIDANT PRECURSORS
The chemiluminescence method, based on the
gas-phase reaction of ozone with ethylene, was
adopted by EPA as the reference method and is
now being used extensively. Several commercial
instruments based on reference methods specified
in EPA's Ambient Air Monitoring Reference and
Equivalent Methods have been found to perform
better than required by these EPA regulations.
Techniques for continuous measurement, based
on gas-solid chemiluminescence and ultraviolet
(UV) photometry, have been developed and
designated as equivalent methods.
Extensive studies in recent years have shown
that various traditional Kt calibration procedures
for oxidant or ozone analyzers are deficient, and
that these methods suffer from lack of both
precision and accuracy.
Methods for measuring total oxidants are based
on the oxidation of potassium iodide (Kl) and the
electrochemical or colorimetric detection of iodine
(I2). Nitrogen dioxide, however, causes a positive
interference, and S02 causes a negative
interference in these methods; these interferences
are very difficult to correctfor. EPA intends to study
and evaluate several new calibration procedures
and to amend Appendix D of 40 CFR Part 50 to
revise the 1 percent neutral buffered potassium
iodide (NBKI) calibration procedure or to replace it
with one or more of several alternatives under
consideration. These alternatives are: (1) gas-
phase titration (GPT) with excess NO; (2) UV
photometry; (3) GPT with excess Oa; and (4) boric
acid Kl.
Commercial instruments for monitoring total
nonmethane hydrocarbon (NMHC) in ambient air
have been tested extensively and found to be
inaccurate at NMHC levels approaching the 0.24-
ppm C air quality standard. At present,
hydrocarbons in ambient air can be accurately
measured only by sophisticated gas chroma-
tography. Routine methods are under devel-
opment.
On December 1, 1976, EPA promulgated a new
principle and calibration procedure for measuring
N02. The measurement principle is the gas-phase
chemiluminescent reaction of 03 and NO. Two
calibration procedures are prescribed. One is a
GPT procedure referenced to an NO-in-nitrogen
standard. The other calibration procedure is
referenced to an N02 permeation device. Before
the development of chemiluminescence analyses,
most N02 data were collected using methods
based on variations of the Gness-Saltzman
procedure. For all practical purposes, these
methods have been replaced by the
chemiluminescence method
-------
HEALTH EFFECTS OF OZONE AND OTHER
PHOTOCHEMICAL OXIDANTS
In this section, the influence of exposure to
ozone and other photochemical oxidants on
physiology and health is assessed, beginning with
a brief discussion of the concept of "threshold
pollutant concentrations" and itsapplicationtothe
protection of public health. Then, the strength of
association between exposure to ozone and other
oxidants and changes in several types of
biomedical indicators is evaluated.
For each type of indicator, the discussion
addressesfour main topics: (1 )thedegreetowhich
changes in each type constitute impairments in
public health; (2) the available scientific evidence
relating ozone and other oxidant exposures to
changes within each type; (3) the reliability of
existing scientific evidence; and (4) where
appropriate, the confidence with which findings
may be attributed to ozone alone, as opposed to
other substances or combinations of substances.
Whenever possible, clinical (human experi-
mental) and epidemiologic studies will be
discussed together. However, because of the great
uncertainty inherent in the quantitative
extrapolation of results of animal studies to
humans, human and animal studies will be
discussed separately.
Discussion of Threshold Concentrations
The Clean Air Act directs that National Primary
Ambient Air Quality Standards be such that their
attainment and maintenance shall, in the
judgment of the Administrator, and allowing an
adequate margin of safety, be requisite to protect
the public health. The confidence with which a
margin of safety can be defined depends on the
precision with which a threshold pollutant
concentration, (the level above which exposure to
ozone or other oxidants promotes impairment of
health and below which it does not) can be
determined.
In practice, no single overall threshold
concentration for ozone or other pollutants exists.
Thresholds have been shown to vary widely with
the population segment studied and the biologic
indicators measured. Also, since the great majority
of known dose-response relationships do not show
sharp discontinuities, it is most unlikely that a
discrete threshold pollutant concentration can be
established even for a single population segment
and a single biomedical indicator.
Despite these limitations, the environmental
decisionmaker may find it useful to incorporate the
concept of threshold concentrations into the
standard-setting process. For a given population
segment, a threshold concentration may be
operationally defined as occurring somewhere
between a concentration at which no effect on
health or 'function has been observed, and a
concentration at which such an effect has been
demonstrated. In protecting public health, the
population segments of primary concern are the
most susceptible groups, in whom exposure to
ozone or other pollutants is most likely to promote
impairment of health. Such groups may include
those with underlying illness, the very old, the very
young, and the pregnant, (It has not been
determined whether such susceptible population
segments differ from the most sensitive population
segments, which comprise those individuals most
likely to show measurable responses to very low
pollution concentrations.)
Ideally, the environmental decisionmaker would
know the pollutant concentrations with which no
adverse effects are associated in susceptible
groups, as well as the concentrations with which
such adverse effects are unambiguously
associated. Unfortunately, knowledge in both of
these areas remains sparse. Recent experimental
studies of healthy people and animals have greatly
advanced our knowledge of the health effects of
ozone. However, threshold concentrations
deducible from such studies may not apply to the
potentially susceptible groups described above.
The opportunity to study such groups experi-
mentally is severely limited by practical and ethical
constraints. The relatively few epidemiologic
studies of these groups that are extant have
yielded inconclusive results.
Human Studies
SHORT-TERM OXIDANT EXPOSURES
Mechanical Function Of The Lung
Assessment Of Health Effect. There is con-
siderable room for honest disagreement on
whether pollution-induced alterations in
mechanical lung function constitute bona fide
impairments of health in and of themselves. In the
great majority of experimental studies in which
oxidant exposures have produced changes m lung
function in healthy subjects, function has returned
to normal within a few hours. Thus there is no
reason to suspect that in healthy individuals such
changes promote any measurable increase in risk
-------
of future illness. Nor does any available evidence
suggest that in healthy individuals a small change
m ventilatory function, unaccompanied by
symptoms or impairment of oxygen uptake or work
capacity, would interfere with normal activity or
task performance.
However, three considerations suggest that
oxidant-associated changes in lung function may
signal impairment of public health. First, in per-
sons with underlying respiratory illness such as
asthma, chronic bronchitis, and emphysema, even
small decrements in lung function often interfere
with normal activity. Second, at experimental
ozone concentrations as low as 0.30 ppm,
decrements in lung function have usually been
accompanied by physical discomfort, as
manifested in symptoms such as sore throat, chest
pain, cough, and headache. At times this
discomfort has been great enough to prevent the
completion of experimental protocols, particularly
when subjects have been exercising vigorously. It
appears quite likely that the pulmonary irritant
properties of ozone (and perhaps other oxidants)
underlie both the discomfort and the decrements
in function. Thus, at least when associated with
ozone exposure, changes in lung function often
represent a level of discomfort that, even among
healthy people, may restrict normal activity or
impair the performance of tasks.
Summary Of Data. Human experimental studies
have demonstrated that the subject's level of
exercise during ozone exposure is directly related
to the magnitude of change in lung function and
the severity of symptoms at any given ozone
concentration. During exercise, subjects increase
their expiratory flow rates, andthey tendto breathe
through their mouths. These factors increase the
total dose of ozone delivered to the lung and may
increase the depth to which it is delivered in the
respiratory tree.
After 2 hr of resting exposure to 1470 fjg/m3
(0.75 ppm), healthy young adults showed small
changes in lung function in a study by Bates et al.
Folmsbee et al. observed changes in respiratory
pattern (increased respiratory frequency and
decreased tidal volume) and reductions of vital
capacity in healthy young adults exercising
submcximally after a similar resting exposure.
Immediately after 2 hr of exposure to 1470A/g/m3
(0.75 ppm) ozone, during which they performed
intermittent light exercise, subjects showed quite
pronounced changes in lung function in studies by
Folinsbee et al. and Hazucha et al. In subjects
exposed under the same conditions, Folinsbee et
al. observed changes in respiratory pattern, though
not in minute volume or oxygen uptake.
In subjects exercising maximally after 2 hr of
exposure to 1470 fjg/m3 (0.75 ppm) ozone and
intermittent light exercise, Folinsbee et al.
observed decrements in maximum work load, tidal
volume, heart rate, and oxygen uptake.
Folinsbee et al. observed decreases in tidal
volume and maximum expiratory flow rate at 50
percent of vital capacity (V5o) in subjects exercising
submaximally after 2 hr of exposure to 980 pg/m3
(0.5 ppm) ozone and intermittent light exercise. In
subjects giving no history of cough, chest
discomfort, or wheezing in response to allergy or
air pollution exposure (unreactive subjects), few
changes in lung function occurred after 4 hr of
exposure to 980 pg/m3 (0.50 ppm) and
intermittent light exercise (Hackney et al.).
However, in subjects giving such a history (reactive
subjects) and receiving an identical exposure, the
same investigators observed decrements in 8 of 14
measured parameters. Interestingly, in a second
group of unreactive subjects, the same in-
vestigators observed substantial decrements in
lung function after only 2 hr of exposure to 980
pg/m3 (0.5 ppm) ozone and intermittent light
exercise.
Statistically significant changes in forced vital
capacity, maximum mid-expiratory flow rate, and
airway resistance in 22 young males after 2 hr of
exposure to 0.4 ppm ozone and intermittent
moderate exercise were observed by Knelson etal.
After subjects had been exposed for 4 hr, changes
in these parameters had increased, and several
other flow parameters had also changed
significantly in comparison with control values.
In studies by Hazucha and Bates, Hazucha et al.,
and Folinsbee et al., subjects showed changes in
lung function after 2 hr of exposure to 730pg/m3
(0.37 ppm) ozone and intermittent light exercise.
Hackney et al. observed such changes in reactive
subjects but not in unreactive subjects. In a
separate study by Hackney and colleagues, four
Canadians and four southern Cahfornians were
exposed in the Los Angeles area to 730 fjg/m3
(0.37 ppm) ozone for 2 hr. The Californians showed
few changes in lung function, whereas the
Canadians showed decrements in most
parameters measured. (At least partly because of
the small sample sizes, no observed changes were
statistically significant.) In subjects exposed to 730
/jg/m3 (0.37ppm)ozone together with 1000A/g/m3
-------
(0.37 ppm) SOz, Hazucha and Bates observed an
effect on lung function substantially larger than
the sum of the separate effects of the individual
pollutants.
DeLucia and Adams observed changes in lung
function and respiratory pattern in healthy
subjects exercising steadily and fairly heavily over
a 1-hr exposure to 590 fjg/m3 (0.30 ppm) ozone.
Two of six subjects experienced such discomfort as
to prevent them from completing the experimental
protocol.
Hazucha observed small changes in lung
function in three nonsmokers exposed for 2 hr to
490 fjg/m3 (0.25 ppm) ozone and intermittent light
exercise. No lung function changes of note were
seen by Hackney et al. even among reactive
subjects who were similarly exposed.
After 1 hr of exposure to 290 fjg/m3 (0.1 5 ppm)
ozone and steady, fairly heavy exercise, subjects
observed by DeLucia and Adams showed changes
in respiratory pattern. In two of six subjects, the
same investigators noted inconsistent increases in
residual volume.
Small but statistically significant increases in
airway resistance, as measured by plethysmo-
graphy, were observed in two of four healthy
subjects immediately after a 1 -hr exposure to 200
/jg/m3 (0.1 ppm) ozone. The investigators
(Goldsmith and Nadel) did not state whether the
subjects exercised during exposure.
Von Nieding and Wagner reported that subjects
showed decrements in arterial oxygen pressure
(Pa02> and airway resistance after 2 hr of exposure
to 200 A
-------
adaptation in lung function is of any long-term
consequence to health.
Attributability Of Effects To Ozone. Experimental
studies have shown that ozone at concentrations
observed in the ambient air can produce changes
in mechanical lung function. The epidemiologic
studies of Kagawa and Toyama, in which lung
function was correlated more strongly with ozone
exposure than with total oxidant exposure, are
consistent with this finding. However, the degree
to which epidemiologic observations may be
attributed specifically to ozone remains in doubt,
since Kagawa and Toyama were not able fully to
separate the effects of ozone from the effects of
other environmental factors.
Finally, the work of Hazucha and Bates suggests
that the effect on lung function of ozone at 730
)Ug/m3(0.37 ppm) may be enhanced by an identical
concentration of sulfur dioxide. Observation of this
enhancement argues in favor of providing a margin
of safety in a primary National Ambient Air Quality
Standard for ozone. However, this observation
does not support a quantitative recommendation
for a safety margin, since it has not yet been
determined whether such enhancement occurs at
lower ozone concentrations or with substances
other than sulfur dioxide.
Impairment Of Physical Performance —
Decrements in physical performance are
deterrents to personal satisfaction. In this respect,
such decrements constitute impairment of public
health. The degree to which oxidant exposures
may promote decrements in physical performance
has not been determined. However, Wayne et al.
assessed the association between hourly oxidant
concentrations and the proportion of a high school
cross-country team failing to improve running
times between successive track meets in southern
California. Over six cross-country seasons, the
correlation of average oxidant concentration in the
hour before the race with the proportion of runners
failing to improve times was 0.88. The
corresponding correlation for both the first and
second three-season periods was 0.945. A
correlation this high denotes a very close
numerical relationship between two variables.
During the period studied, hourly oxidant
concentrations ranged from approximately 60 to
590 Aig/m3 (0.03 to 0.30 ppm). Inspection of the
data of Wayne et al. reveals no obvious
relationship between unimproved running time
and oxidant concentrations below 200 to 290
3 (0.10 to 0.15 ppm), in spite of the high
overall correlations mentioned above. Also, since
the authors did not consider ozone separately from
other oxidants, the specific contribution of ozone to
the observed results cannot be determined from
this study.
As far as can be ascertained, no replication of the
study of Wayne eta I. has appeared in the published
literature. For three reasons, however, the data
from this study are more trustworthy than most
results of a single epidemiologic study. First, the
correlations between hourly oxidant concentration
and unimproved running time were unusually
high. Second, as mentioned in Chapter 10 of this
document, Herman, at the University of North
Carolina, has analyzed the data of Wayne et al. as
well as data from two additional seasons, and he
has observed results similar to those of Wayne et
al. Third, the results of Wayne et al. are
qualitatively consistent with the results of the
experimental lung function studies mentioned
above, especially those of Folinsbeeetal. In view of
these experimental studies, it would also appear
plausible that ozone contributed significantly to
the results of Wayne et al.
Oxidant Effects In Asthmatics - In the United
States, there are an estimated 6 to 8 million
asthmatics, about 70 percent of whom are
estimated to live in urban areas. Thus, the number
of asthmatics who may be exposed to elevated
oxidant concentrations is substantial.
Available epidemiologic evidence on the
relationship between oxidant exposure and
exacerbation of asthma is very limited. In 1961,
Schoettlin and Landau reportedthatthe proportion
of selected asthmatics in the Pasadena area
having attacks was significantly greater (p < 0.05)
on days when the maximum hourly oxidant
concentration exceeded 490 /ug/m3 (0.25 ppm)
than on days when the corresponding
concentration was below this level. However, the
proportion of asthmatics having attacks on days of
maximum hourly oxidant concentration above 250
/jg/m3 (0.13 ppm) was not significantly different
from the corresponding proportion when the
maximum hourly concentration was below this
level. The authors did not state the actual
percentage of asthmatics having attacks on days in
any exposure category, though they did state that
asthma attacks tended to coincide with elevated
oxidant levels in 8 (6 percent) of 137 patients
studied.
Because it does not present asthma attack rates,
the Schoettlin and Landau report gives no
-------
indication whether increases in attack rates might
have been expected at maximum hourly oxidant
concentrations below 490 /*g/m3 (0.25 ppm). Nor
does it allow any judgments as to the extent to
which increased attack rates might be attributable
specifically to ozone. Despite the considerable
attention it has received, this report should- be
considered the preliminary investigation its
authors intended it to be.
Respiratory Symptoms and Headache
Assessment Of Health Effect. As mentioned
above, increased rates of respiratory symptoms
and headache constitute impairment of public
health. Even when mild, such symptoms are
annoying. Even when reversible, they may restrict
normal activity or limit the performance of tasks.
Summary Of Data. In nearly all experimental
studies in which ozone exposures have been
sufficient to produce changes in lung function,
most subjects have reported respiratory
symptoms. The most common symptoms have
been throat tickle, substernal tightness, pain on
deep inspiration, and cough. Wheezing, dyspnea,
and headache have occurred less commonly.
Symptom severity has increased with ozone
concentration and exercise. In studies of heavy
exercise, symptoms have occasionally been severe
enough to prevent subjects from completing
experimental protocols.
In an epidemiologic study in southern California,
Hammer et al. assessed the association between
daily maximum hourly oxidant concentration and
rates of chest discomfort, cough, and headache
among student nurses. Rates of each of these
symptoms, whether unadjusted or adjusted for
fever, began to increase in the following oxidant
concentration ranges: chest discomfort, 490 to
570 pg/m3 (0.25 to 0.29 ppm); cough, 590 to 760
/jg/m3 (0.30 to 0.39 ppm); and headache, 290 to
370 fjg/m3 (0.15 to 0.19 ppm). Adjusted and
unadjusted rates of headache, however, were not
unequivocally elevated below maximum hourly
oxidant concentrations of 590 to 760 /jg/m3 (0.30
to 0.39 ppm).
Several Japanese investigators have assessed
the association between daily pollutant
concentrations and symptom rates among
students. In one Japanese study, rates of sore
throat, dyspnea, and headache were somewhat
higher during the summer months on days when
the oxidant concentration exceeded 200 /jg/m3
(0.10 ppm) than on days when it did not. Over a 1 -
year period, rates of respiratory symptoms and
headache were higher on days when the oxidant
concentration exceeded 290 /jg/m3 (0.15 ppm)
than on days when it was lower than 200/jg/m3
(0.10 ppm). Though there is reason to believe that
the stated oxidant concentrations were daily
maximum hourly averages, averaging times were
not clearly presented in the report of this study.
Reliability Of Evidence. Because of their close
correlation with the results of experimental
studies, the results of Hammer et al. appear to be
quite reliable, even though unconfirmed in this
country. Their reliability is enhanced by the large
number of person-days of observation (about
53,000) L , .compassed by the study. Though ozone
levels were not considered m the Hammer et al.
study, it is reasonable to hypothesize, in view of
experimental studies, that ozone contributed
substantially to observed increases in rates of
cough, chest discomfort, and headache.
As nearly as can be determined from
translations of original articles, the Japanese
epidemiologic studies cited in this document were
appropriately designed. However, it is very difficult
to interpret their results. Also, at least at present,
the applicability of these results to any oxidant
pollution problem in the United States must be
considered very limited.
In data analyses, Japanese investigators (like
many U.S. investigators) have not been able fully to
separate the effects of individual pollutants, It is
conceivable that combinations of pollutants
unique to Japan were necessary to promote the
increased symptom rates observed there.
Averaging times for pollutant measurements were
not clearly stated in the Japanese studies.
Therefore, it is often impossible to draw specific
inferences as to dose and response. In addition, the
degree to which differences in Japanese and U.S.
cultural responses to air pollution may
differentially affect symptom perception m the two
countries has not been determined
A problem more fundamental than any of these
specific reservations, however, is that Japanese
and U.S. investigators as yet have great difficulty in
exchanging specific scientific concepts. Until the
quality of scientific communication between these
groups of investigators increases, the ability to
interpret the Japanese studies and to apply them to
situations in the United States will remain severely
limited.
Oxidants And Eye Irritation - In that it is annoying
and uncomfortable, the reversible eye irritation
-------
produced by exposure to ambient photochemical
oxidants may legitimately be considered a
marginal impairment of public health. However,
whether such eye irritation is sufficient to impair
performance or restrict normal activity has not
been determined. Nor has any association
between oxidant-mediated eye irritation and
chronic eye disease been observed.
In epidemiologic studies, no symptom has been
more consistently linked to oxidant exposure than
eye irritation. In most studies reported before the
publication in 1970 of Air Quality Criteria for
Photochemical Oxidants, rates of eye irritation
were observed to increase fairly steadily when
oxidant concentrations ranged from 200 to 880
/ug/m3 (0,10 to 0.45 ppm). Hammer et al. found that
rates of eye discomfort began to increase at
oxidant concentrations of 290 to 370 /ug/m3 (0.15
to 0.19 ppm).
Evidence linking ambient photochemial oxidant
exposures to eye irritation is convincing. However,
the specific etiologic agent or agents remain
unknown. Experimental studies have shown quite
conclusively that ozone at ambient concentrations
is not an eye irritant.
Oxidants and Mortality - Review of existing studies
shows no consistent association between daily
oxidant concentrations and daily mortality rates.
As far as can be ascertained, no studies of oxidant
exposures and mortality have been performed
since the publication in 1970 of Air Quality Criteria
for Photochemical Oxidants.
Other Effects Of Short-Term Ozone And Oxidant
Exposure
Changes In Erythrocytes. In an experimental
study, Buckley et al. observed increased rates of
erythrocyte lysis in HZ02 after a 2%-hr exposure of
healthy subjects to 980 jug/m3 (0.5 ppm) ozone and
intermittent light exercise. These investigators
also noted changes in the" activity of several
erythrocytic enzymes. Hackney et al. observed an
increased in vitro lysis rate in erythrocytes of Cana -
dians, but not in that of Californians following a 2-
hr exposure of subjects to 730 (ig/m3 (0.37 ppm)
ozone and intermittent light exercise.
Chromosomal Aberrations. Merz et a I. repdrted
chromosomal abnormalities in the lymphocytes of
six subjects after they were exposed to 980/ug/m3
(0.5 ppm) ozone for 6 or 10hrs. However, a study by
McKenzie et al. showed no increased rate of
leukocyte chromosomal aberration in the
lymphocytes of 30 subjects exposed for 4 hr to 780
£fg/m3 (0.4 ppm) ozone. In the few epidemiologic
studies of oxidant exposure and chromosome
morphology performed to date, factors other than
differences in oxidant exposure have confounded
observed results to the extent that no inferences
can be drawn.
Biochemical Parameters. Buckley et al. observed
reduced glutathione reductase activity and
increased vitamin E and lipid peroxidation in
human serum following exposure of subjects to
980 fjg/m3 (0.5 ppm) ozone and intermittent light
exercise for 2% hr. Following a 2-hr exposure to
730 ^g/m3 (0.37 ppm) ozone and intermittent light
exercise, the Canadians and Californians studied
by Hackney et al. showed increases in serum
vitamin E levels; but only in the Canadians were
these increases statistically significant.
After subjects exercised steadily and vigorously
throughout a 1-hr exposure to 590 /jg/m3 (0.30
ppm) ozone, DeLucia and Adams observed no
changes in the following biochemical blood
parameters! hemoglobin level, nonprotein
sulfhydryl level, erythrocyte glucose-6-phosphate
dehydrogenase activity, and glutathione reductase
activity.
Clinical Significance. The clinical significance of
ozone-mediated changes observed in studies of
blood is not yet known. As yet, an epidemiologic
study of oxidant effects on anemic individuals has
not been done. Whether or not oxidant exposures
promote changes in leukocyte chromosome
morphology, the significance of the changes
themselves is unknown. Finally, changes in serum
parameters of the magnitude observed in
experimental ozone studies have not yet been
linked to any clinical diseases.
LONG-TERM OXIDANT EXPOSURES
With the exception of the Hackney et al study of
Californians and Canadians, no experimental
studies of humans have as yet assessed the effects
of long-term oxidant exposures. The few available
epidemiologic studies of such exposures have
yielded inconclusive results, fvlahoney has ob-
served an association between broad patterns of
oxidant distribution and annual respiratory disease
mortality in the Los Angeles area. However, no
convincing association between lung cancer
mortality and oxidant exposure has been shown. In
some studies, a limited association between the
frequency of chronic obstructive lung disease and
oxidant exposure has been observed. In other
studies, however, no such association has been
-------
apparent. No relationship between long-term
oxidant exposure and acute respiratory disease
incidence or change in lung function has been
observed, though neither of these areas has been
extensively investigated.
Most available epidemiologic studies of long-
term oxidant exposure are difficult to interpret. As
usually acknowledged by the authors themselves,
factors other than pollution exposure may often
have influenced the results observed. Some
studies have proved difficult to evaluate because
areas in which health variables were compared did
not show clear-cut differences in pollution
exposure. Thus human studies cannot yet provide
the environmental decisionmaker with concrete
information as to the effects of long-term oxidant
exposures on public health,
Animal Toxicology Studies
ANIMAL INFECTIVITY EXPERIMENTS
Increased susceptibility of animals to bacterial
infection following ozone exposure at 196 fjg/m3
(0.1 ppm) is described by several investigators
(Coffin et a!., 1968; Ehrlich et al., 1976; Gardner et
al , 1974; Coffin and Gardner, 1 972; Miller et al.,
1978). Others (Goldstein et al., 1971a, 1971b;
Goldstein and Hoeprich, 1972; Goldstein et al.,
1974), using bacterial infection, have developed
indices of infection for measuring the effects of
ozone on the lungs of rodents. The effective
concentration of ozone at which susceptibility to
infection is increased is lower when either other
pollutants or stress is combined with ozone
exposure. These findings have definite human
health implications, although different exposure
levels may be associated with such effects in
humans. These reactions in mice represent effects
on basic biological responses to infectious agents,
and there is no reason to believe that the pollutant-
induced alterations of basic defense mechanisms
that occur in mice could not occur in humans who
possess equivalent defense systems. The
extrapolation of these data to man is not supported
by direct epidemiologic evidence that susceptibility
to infection increases in persons exposed to ozone
and other photochemical materials. However, the
biochemical and cellular alterations described
below for animals suggest that multiple epithelial
and biochemical targets are perturbed by ozone
exposure. Ozone-induced irritation of the major
bronchi in man does occur at ozone concentrations
approximating 490/ug/m3 (0.25 ppm). Fortracheal
ozone concentrations greater than 100 /L/g/m3
(0.05 ppm), it is predicted that a smaller inspired
concentration would be required in man than in
rabbits and guinea pigs to result in a given
respiratory bronchiolar dose (Miller et al., 1 977). A
3-hr exposure to 490 ^g/m3 ozone (0.25 ppm) has
been shown to be sufficiently high to injure rabbit
alveolar macrophages. Thus, depending on the
extent of nasopharyngeal removal of ozone in man,
a given level may be expected to produce similar
effects in man and certain animals.
MORPHOLOGICAL ABNORMALITIES OF ANIMAL
RESPIRATORY SYSTEMS
A range of morphological effects is noted in
association with experimental ozone exposures of
392 to 1960 fig/m3 (0.2 to 1.0 ppm). These effects
vary from the reversible replacement of Type 1
with Type 2 alveolar cells (e.g., 392 ^g/m3 or 0,2
ppm ozone for 3 hr/day over 7 days) to emphyse-
matous changes and terminal bronchiolar and
alveolar damage (e.g., 784 fjg/m3, or 0.4 ppm,
ozone for 6 hr/day, 5 day/wk, for 10 months; and
1058 yug/m3, or 0.54 ppm ozone continuous for 3
months). Occurrence of these effects after long-
term exposure to low concentrations increases the
expectation that repeated or chronic exposures
may have the potential for inducing similar effects
in humans.
BIOCHEMICAL EFFECTS
There is an impressive variety of biochemical
alterations associated with ozone exposures over
the range of 1 96 to 1960 ^g/m3 (0.1 to 1.0 ppm).
Though effects caused by levels of 980/Kj/m3 (0.5
ppm) and greater have definite toxic potential, the
biological significance of those changes detected
in exposures to 196 to 392/ug/m3 (0.1 to 0.2 ppm)
are more difficult to assess. These biochemical
changes are significant, however, in
demonstrating that there are ozone-induced
effects at cellular sites and in organ systems
distant from the lung. Biological reactions to ozone
include an increase in the activity of enzymes that
detoxify oxidizing chemical species. Increased
enzymatic activity, as well as increased oxygen
consumption, has been shown in some instances
to be prevented, reversed, or diminished by
increased vitamin E levels (vitamin E is an
antioxidant). Thus increased enzymatic activity
may represent a reaction to potential toxicity rather
than being a toxic response itself. Nevertheless,
such reactions represent the organism's response
to ozone. Such perturbations in biological systems
may pose a health risk to the population,
particularly for susceptible individuals.
10
-------
GENETIC AND TERATOGENIC POTENTIAL
A report by Zelac (1971) on chromosomal
abnormalities in peripheral leukocytes of intact
hamsters which were exposed to 392 pg/m3 (0.2
ppm) ozone for 5 hr was not confirmed by Gooch et
al. (1976). Although confirmation of the results
may suggest another category of ozone-induced
effects, the preliminary evidence does not suggest
that these effects occur at lower levels of exposure.
IMPLICATIONS OF ANIMAL TOXICOLOGY DATA
Note that the difference between concentrations
of ozone that produce toxicological effects in
animals (as well as symptomatic and lung function
changes in humans) and ambient air levels of
ozone is much smaller than for nearly any other
atmospheric pollutant. There is an unusual
clustering and convergence of various
toxicological, experimental human, and
epidemiologically observed human effects in
association with ozone concentrations ranging
from 392 to 11 76 fjg/m3 (0.2 to 0.6 ppm).
Chronic effects of long-term ozone exposure in
man cannot be quantitatively related to specific
ozone concentrations of short (hourly or daily)
averaging times because of the long period of
disease induction and the varied exposures of
individuals during the period of induction.
Epidemiologic studies may establish relationships
between long-term ozone exposure and the risk of
human chronic disease, but toxicological studies
must be relied on in quantifying the ozone levels
that may induce chronic effects.
Ozone Versus Oxidant Health Effects
Two characteristics of ozone and oxidant
exposures should be cited with reference to public
health: (1) ozone itself is a primary cause of most of
the health effects reported in toxicological and
experimental human studies, and the evidence for
attributing many health effects to this substance
alone is very compelling; and (2) the complex of
atmospheric photochemical substances is known
to produce health effects, some of which (eye
irritation, for example) are not attributable to pure
ozone but may be caused by other photochemical
substances in combination with ozone.
EFFECTS OF PHOTOCHEMICAL OXIDANTS
ON VEGETATION AND CERTAIN MICRO-
ORGANISMS
Since injury to vegetation by oxidants was first
identified in 1944 in the Los Angeles Basin, our
understanding of oxidant effects and of the
widespread nature of their occurrence has
increased substantially. The major phytotoxic
components of the photochemical oxidant complex
are ozone and peroxyacetylnitrate (PAN), although
some data suggest that other phytotoxicants are
also present. The peroxyacylnitrates are the most
phytotoxic of the known photochemical oxidants;
but because ozone is ubiquitous and associated
with widespread injury to vegetation, it is the most
important phytotoxic component of the
photochemical oxidant complex.
The effects of photochemical oxidants on
vascular plants occur at several levels, ranging
from the subcellular to the organismic, depending
on the concentration and duration of exposure to
the pollutant and the interval between cessation of
exposure and examination of the plant.
The earliest effect is an increase in cell
membrane permeability. Following that, cellular
and biochemical changes take place that are
ultimately expressed on the organismic level as
visible foliar injury, increased leaf drop, reduced
plant vigor, reduced plant growth, and death. Such
biochemical modifications in an individual plant
are manifested by changes in plant communities
and, finally, in whole ecosystems.
Leaf stomata are the principal sites through
which ozone and PAN enter plants. Oxidants affect
photosynthesis, respiration, transpiration,
stomatal opening, and metabolic pool
development, as well as biochemical pathways
and enzyme systems.
Visible injury is identifiable as pigmented,
chlorotic, or necrotic foliar patterns. Metabolic
cellular disturbances can occur without visible
injury and may be reversible. However, most of the
growth effects reported until recently were
accompanied by visible injury.
Classic ozone injury is demonstrated by the
upper-surface leaf fleck of tobacco and the leaf
stipple of grape. Many plants show an upper-
surface response with no associated injury to the
lower surface of leaves. However, in mono-
cotyledonous plants such as grasses or cereals,
and in some non-monocotyledonous plants, there
is no division of mesophyll tissue, and bifacial
necrotic spotting (flecking) is a common symptom
of ozone injury.
Coniferous trees exhibit different symptoms.
Ozone is probably the cause of emergence tipburn
in eastern white pine (white pine needle dieback)
and of chlorotic decline, a needle injury of
ponderosa pine
11
-------
Classic PAN injury appears as a glaze followed
by bronzing of the lower leaf surface of many
plants. Complete collapse of leaf tissue can occur if
concentrations are sufficiently high. Early leaf
senescence and abscission usually follow the
chronic symptoms. Patterns of chronic injury are
generally not characteristic and may be confused
with symptoms caused by biotic diseases, insect
infestation, nutritional disorders, or other
environmental stresses,
A great deal of research has been done to define
more accurately the effects of oxidants on plant
growth and yield. Studies comparing the growth of
plants in field chambers provided with carbon-
filtered or nonfiltered ambient air containing
oxidants have reported up to 50-percent decreases
m yield of citrus (orange and lemon) exposed to
oxidants; a 10- to 1 5-percent suppression in grape
yield in the first year and a 50- to 60-percent
reduction over the following 2 years; and a 5- to
29-percent decrease in yield of cotton lint and seed
in California. Losses of 50 percent in some
sensitive potato, tobacco, and soybean cultivars
have been reported in the eastern United States, It
is apparent that oxtdants in the ambient air reduce
the yields of many sensitive plant cultivars.
Experimental chambers with controlled
environments have been used to study both short-
term and long-term effects of exposure; to ozone
(Tables 11-2, 11-4, Chapter 11). Radishes given
one, two, or three acute exposures (785 /jg/m3, or
0.40ppm)of 1.5hr each at 7,14, and/or 21 days of
age exhibited reductions in root growth. The
reductions in root growth from the multiple ozone
exposures were equal to the additive effects of the
single exposures, In other words, the temporal
distribution was not a significant factor. When soy-
bean plants were exposed to 1468 /jg/m3 (0 75
ppm) ozone for 1 hr, root growth was consistently
reduced more than top growth. There were also
reductions in nodule weight and number. The
greater reduction of root growth compared to top
growth is related to the transport of photosynthate.
Ozone also affects nitrogen fixation in clover,
soybean, and pinto bean through reduction in
nodule number, even though nodule size and
efficiency of nitrogen fixation are not influenced.
The effect of ozone on the number of nodules
formed by legumes, if widespread, could have a
major impact on plant communities and could
affect fertilizer requirements. There are
indications that the effect of ozone on nodulation
may be related to the carbohydrate supply in the
host plant.
Experimental long-term exposures to ozone of a
variety of crops, as well as of ornamental and
native plants, have resulted in a reduction in
growth and/or yield. Exposure of 14 species
representative of the aspen plant community to
ambient air containing 98 to 1 37 /jg/m3 (0.05 to
0.07 ppm) ozone and to ozone concentrations of
290 and 588 /jg/m3 (0.15 and 0.30 ppm) for 3
hr/day, 5 days/week, and to charcoal-filtered air
throughout the growing season, resulted m foliar
injury to all species at the highest pollutant
concentration. The growth of two soybean
cultivars (Hood and Dare) was inhibited by
intermittent exposure to ozone at 196 vg/m3 (0.10
ppm) for 3 weeks. Both root and top growth were
decreased. Similar results were noted with radish,
except that a lower concentration of ozone (98
/jg/m3, or 0,05 ppm) inhibited growth. In these
studies, the reduced growth occurred even though
there were few visible symptoms of plant injury.
A 30-percent reduction in the yield of wheat
occurred when wheat at anthesis was exposed to
ozone at 392 /vg/m3 (0.2 ppm) for 4 hr/day for 7
days. A significant reduction in the yield of tomato
was noted when plants were experimentally
treated with ozone at 686/jg/m3(0.35 ppm);fewer
fruit set, and thus fewer fruit were harvested.
Chronic exposures to ozone at 98 to 290 /jg/m3
(0.05 to 0.15 ppm)for 4 to 6 hr/day reduced yields
in soybean and corn grown under field conditions.
The threshold for measurable effects for ozone
appears to be between 98 to 196 /jg/m3 (0.05 to
0.1 0 ppm) for sensitive plant cultivars. This is well
within the range of ozone levels monitored in the
eastern United States. Growth orflowering effects
were reported for carnation, geranium, radish, and
pinto bean grown in greenhouse chambers and
exposed to ozone at 98 to 294 fjg/m3 (0,05 to 0.15
ppm) for 2 to 24 hr/day.
The two most critical factors in determining
plant response to air pollution are duration of
exposure and concentration of pollutants. These
two factors describe exposuredose. Indetermining
the response of vegetation to oxidants,
concentration is more important than time.
The concept of limiting values was used by
Jacobson to define a boundary between doses of a
pollutant that are likely to injure vegetation
measurably and those that are not. Foliar injury
was used as the index of plant response. The
ranges for limiting values for effects of ozone are:
1. Trees and shrubs -
400 to 1,000 fjg/m3 (0.2 to 0.51 ppm)
for 1 hr
12
-------
200 to 500 ^g/m3 (0,1 to 0.25 ppm) for
2 hr
120 to 340 pg/m3 (0.06 to 0.17 ppm)
for 4 hr
2. Agricultural crops -
400 to 800 /jg/m3 (0.2 to 0.41 ppm) for
0.5 hr
196 to 500 ^g/m3 (0.1 to 0.25 ppm) for
1 hr
75 to 180 A/g/m3 (0.04 to 0.09 ppm) for
4 hr
Limiting values for PAN are:
1000 fjg/m3 (0.2 ppm) for 0.5 hr
500 fig/m3 (0.1 ppm) for 1 hr
175 fjg/m3 (0.035 ppm) for 4 hr
Doses of ozone or PAN greater than the upper
limiting values are likely to cause foliar injury.
The data points used to determine the limiting
values listed above are not necessarily threshold
values, but are based on available published
research data. More than 200 studies were
surveyed. Any limitations that were present in the
experimental techniques used in the studies are
therefore expressed in the data points. The number
of studies used to derive the data points for ozone
exposure of trees and shrubs (19) and for PAN is
another limitation on the values given above.
For agricultural crops, the inaccuracies m
measurements make the interpretation of results
of repeated long-duration exposuresdifficult.Thus
limiting values for ozone concentrations below
100 /jg/m3 (0.05 ppm) are not useful.
An ozone concentration of 98 to 137 ^g/m3
(0.05 to 0.07 ppm) for 4 to 6 hr/day for 1 5 to 133
days can significantly inhibit plant growth and
yield of certain species (Table 11 -4, Chapter 11).
Plant sensitivity to ozone and PAN is conditioned
by many factors, Genetic diversity in sensitivity to
ozone between species and between cultivars
within a species is well documented. Variations in
sensitivity to ozone within a natural species are
well known for several pine species, including
white, loblolly, and ponderosa. Plant sensitivity to
oxidants can be changed by both climatic and
edaphic factors. A change in environmental
conditions can initiate a change in sensitivity at
once, but it will be 3 to 5 days before the response
of the plant is totally modified. Plants generally are
more sensitive to ozone when grown under short
photoperiods, medium light, medium temperature,
high humidity, and high soil moisture. Injury from
PAN may increase with an increase in light
intensity. Conditions during exposure and growth
affect the response of plants to oxidants in similar
ways. In general, environmental conditions
optimum for plant growth tend to increase the
sensitivity to ozone. Factors that increase water
stress at the time of exposure tend to make plants
more tolerant to ozone. Soil moisture is probably
the most important environmental factor that
affects plant response to oxidants during the
normal growing season. Physiologic age affects
the response of the leaf to oxidants. Young leaf
tissue is most sensitive to PAN, whereas newly
expanding and maturing tissue is most sensitive to
ozone. Light is required for plant tissue to respond
to PAN; a similar light requirement is not needed
for plants to respond to ozone.
The majority of effects observed, such as
suppression of root growth, mineral uptake, and
nitrogen fixation, apparently result from a
suppression of photosynthesis and modifications
in photosynthate distribution. This suppression of
metabolic reserves ultimately slows plant growth
and renders the plant more sensitive to other
stresses. Physiological changes can provide a
sensitive means of.monitoring the health and vigor
of the plant with or without visible injury. Ozone
affects pollen germination in some species and
may affect yield through incomplete pollination of
flowers. Investigations with Arabidopsis that/ana
showed no mutagenic effects from ozone over
seven generations.
Mixtures of ozone and SOa can cause effects
below the levels caused by either gas alone;
however, there is some disagreement concerning
the interactions of ozone with other gases. Ratios
of gas ' mixtures, intermittent exposures,
sequential exposures to pollutants, and pre-
disposition by one pollutant to the effects of a
second pollutant may be important factors in
nature, but insufficient knowledge is available for
elucidation of the effects.
The response of plants to oxidants may be
conditioned by the presence or absence of biotic
pathogens. Depending on the plant and the
pathogen, oxidants may cause more or less injury
to a given species. Oxidant injury to ponderosa
pine predisposes the trees to later invasion by bark
beetles. Ozone and ozone/sulfur dioxide mixtures
can decrease the population of some plant-
parasitic nematodes. Variable plant responses
were noted when herbicides were used in the
presence of high oxidant concentrations.
Little research on the effects of oxidants on
ferns, nonvascular green plants, and micro-
organisms has been reported. Lichens and mosses
13
-------
are responsive to acid gases, but there is no
definite evidence that they respond to oxidants
Ferns may be especially sensitive, but their injury
response is different from that of higher vascular
plants. Growth and sporulation of fungi on
surfaces are usually, but not always, affected.
Ozone from 0.1 to several milligrams per liter of
solution is required to kill many microorganisms in
liquid media. Most work with microorganisms has
been done to study the effectiveness of ozone as a
biocide in the storage of vegetation or in the
treatment of water or sewage.
EFFECTS OF PHOTOCHEMICAL OXIDANTS
ON ECOSYSTEMS
Plants, animals, and microorganisms usually do
not live alone but exist as populations. Populations
live together and interact as communities.
Communities, because of the interactions of their
populations and of the individuals that make them
up, respond to pollutant stress differently from
individuals. Man is an integral part of these
communities, and as such, he is directly involved in
the complex ecological interactions that occur
within the communities and within the ecosystem
of which the communities are a part.
The stresses placed on the ecosystems and their
communities can be far-reaching, inasmuch asthe
changes that occur may be irreversible. For
example, it has been suggested that the arid lands
of India are the result of defoliation and elimination
of vegetation, which in turn induced local climatic
changes that were not conducive to the
reestablishment of the original vegetation.
An ecosystem (e.g., the planet earth, a forest, a
pond, or a fallen log) is a major ecological unit
made up of living (biotic) and physical (abiotic)
components through which the cycling of energy
and nutrients occurs. A structured relationship
exists among the various components. The biotic
units are linked together by functional
interdependence, and the abiotic units constitute
all of the physical factors and chemical substances
that interact with the biotic units. The processes
occurring within the biotic and abiotic units and
the interactions among them can be influenced by
the environment.
Ecosystems tend to change with time.
Adaptation, adjustment, and evolution are
constantly taking place as the biotic component,
the populations, and the communities of living
organisms interact with the abiotic component in
the environment. Recognizable sequential
changes occur. With time, populations and
communities may replace one another This
sequential change, termed succession, may
culminate in climax communities. Climax
communities are structurally complex, are more or
less stable, and are held m a steady state through
the operation of a particular combination of biotic
and abiotic factors The disturbance or destruction
of a climax community or ecosystem results in its
being returned to a simpler stage. Existing studies
indicate that changes occurring within
ecosystems, in response to pollution or other
disturbances, follow definite patterns that are
similar even in different ecosystems. It is therefore
possible to predict broadly the basic biotic
responses to the disturbance of an ecosystem.
Diversity and structure are most changed by
pollution as a result of the elimination of sensitive
species of flora and fauna and of the selective
removal of the larger overstory plants in favor of
plants of small stature. The result is a shift from the
complex forest community toward the less
complex hardy shrub and herb communities. The
opening of the forest canopy changes the
environmental stresses on the forest floor, causing
differential survival and, consequently, changed
gene frequencies in subcanopy species.
Associated with the reduction in diversity and
structure is a shortening of food chains, a
reduction m the total nutrient inventory, and a
return to a simpler successional stage.
It should be emphasized that ecosystems are
usually being subjected to a number of stresses at
the same time, not just a single perturbation such
as oxidant pollution.
The effects of oxidants on the mixed-conifer
forest of the San Bernardino Mountains
graphically demonstrate the changes that occur in
natural ecosystems as discussed above. Since the
early 1940's, the San Bernardino Forest has been
undergoing stress from oxidants transported long-
range from Los Angeles, 144 km (90 miles) away.
Losses of ponderosa and Jeffrey pines, the
overstory vegetation, have increased dramatically
as pollutant levels have risen. Black oak has also
suffered oxidant injury. The composition of both
plant and animal populations has been altered by
the death of the ponderosa and Jeffrey pines.
The interaction of pollutant and inversion layers
at the heated mountain slope results in the vertical
venting of oxidants over the mountain crest by up-
slope flow, thus establishing an elevational
gradient of oxidant concentrations. Oxidant
concentrations ranging from 100 to 200 //g/m3
(0.05 to 0.10 ppm) at altitudes as high as 2432 m.
14
-------
approximately 1033 m above the mountain crest,
have been measured by aircraft.
Total oxidant concentrations in the San
Bernardino Mountains have been measured
continuously from May through September since
1968 at the Rim Forest-Sky Forest monitoring
station. During each of the first 7 years of
monitoring, between June and September, the
total number of hours in which concentrations of
ozone were 160 jug/m3 (0.08 ppm) or more was
never less than 1300, The number of hours in
which the total oxidant concentration was 390
/jg/m3 (0.20 ppm) or higher increased from fewer
than 100 in 1969 to nearly 400 in 1 974. It was not
uncommon to observe momentary oxidant peaks
as high as 1180/jg/m3 (0.60 ppm). Theduration of
oxidant concentrations exceeding 200 /jg/m3
(0.10 ppm) was 9, 1 3, 9, and 8 hr/day going from
the lower- to the higher-altitude stations.
The most recent data firmly indicate that oxidant
concentrations in the San Bernardino Forest will
either increase annually or oscillate around the
mean of present high concentrations in the
foreseeable future.
The transport of the urban plume from the coast
northeastward to the mountains can be readily
demonstrated. Because of this transport, the
permanent vegetation constituting natural
ecosystems receives much greater chronic
exposure, while the short-lived vegetation
constituting the economically more valuable
agroecosystem of the Los Angeles coastal plain
can be subject to injurious doses, but in
intermittent, short-term fumigations. Each
situation has measurable economic and aesthetic
consequences, but on different time scales. The
single-species agricultural ecologic system (the
agroecosystem) has little resilience to pollutant
stress. Losses are sometimes immediate and
occasionally catastrophic. The complex natural
ecosystem is initially more resistant to pollutant
stress, but the longer chronic exposures cause
disruption of both structure and function in the
system that may be irreversible.
The oxidant injury to the mixed conifer stands of
the San Bernardino Mountains that began in the
early 1940's, as indicated above, is well advanced.
A similar problem is developing in the forests of the
southern Sierra Nevada Mountains. Both areas
show direct as well as indirect effects on all
subsystems of the forest ecosystem producers,
consumers, and decomposers
In summary:
1. Ozone injury has limited the growth and
caused the death of ponderosa and Jeffrey
pines. An estimated 1.3 million trees have
been affected. Decrease in cone production
has resulted in a decrease in reproduction.
Black oak has also suffered injury from
ozone.
2. Reduction in fruits and seeds that make up
the diet of most of the common small
mammals has influenced the populations
of these organisms.
3. Essential processes, such as recycling of
nutrients, may have been disrupted,
causing a limitation in the growth of
vegetation.
4. Death of the predominant vegetation has
caused an alteration in the species
composition and a change m the wildlife
habitat.
The San Bernardino Mountain study illustrates
the complexity of the problems caused by
environmental pollution. The changes that have
occurred in this mountain ecosystem as the result
of oxidant transport have already influenced the
importance and value of this natural resource to
the residents of Southern California.
The injury to the eastern white pine in the
Appalachian Mountains caused by oxidant
transport from the urban northeast has begun a
similar sequential change thai could degrade this
important recreational area. Total oxidant peaks as
high as 220 /jg/m3 (0.11 ppm) were recorded for
July 1975. Concentrations exceeding 160 /jg/m3
(0.08 ppm) were measured in June 1976. These
episodes resulted m significant increases in
oxidant injury to three categories of eastern white
pine in the Blue Ridge Mountains (the eastern
range of the Appalachian Mountains).
Ecosystems are usually evaluated by modern
man solely on the basis of their economic value.
This economic value in turn depends on the extent
to which man can manipulate the ecosystem for
his own purposes. This single-purpose point of
view makes it difficult to explain the many benefits
of a natural ecosystem to ma n's welfare m terms of
the conventional cost-benefit analysis. Gosselink,
Odum, and Pope have, however, placed a value on
a tidal marsh by assigning monetary values to its
multiple contributions to man's welfare such as
fish nurseries, food suppliers, and waste-
treatment functions of the marsh. They estimate
15
-------
the total social value to range from $50,000 to
$80,000 per acre.
Westman also evaluated the benefits of natural
ecosystems by estimating the monetary costs
associated with the loss of the free services
(absorption of air pollution, regulation of global
climate and radiation balance, and soil bioding)
provided by the ecosystems. Westman estimated
that the oxidant damage to the San Bernardino
National Forest could result in a cost of $27 million
per year (1973 dollars) just for removal of
sediments resulting from erosion, as long as the
forest remained in the early stages of succession.
Estimates of the cost in currency of the values of
items and qualities such as clean air and water,
untamed wildlife, and wilderness, once regarded
as priceless, are an attempt to rationalize the
activities of civilization. When estimating the
monetary cost in currency of the values lost
through damage to ecosystems, the assumption is
usually made that the decisionmakers will choose
the alternative that is most socially beneficial as
indicated by costs compared to benefits. As
Westman points out, the assumption that
decisions maximizing benefit/cost ratios
simultaneously optimize social equity and utility is
based on certain inherent corollaries:
(1) The human species has the exclusive right to use and
manipulate nature for its own purposes (2) Monetary units
are socially acceptable as means to equate the value of
natural resources destroyed and those developed (3) The
value of services lost during the interval before the
replacement or substitution of the usurped resource has
occurred is included in the cost of the damaged resource f4)
The amount of compensation in monetary units accurately
reflects the full value of the loss to each loser m the
transaction. (5) The value of the item to future generations
has been judged and included man accurate way m the total
value (6) The benefits of development accrue to the same
sectors of society, and in the same proportions, as the
sectors on whom the costs are levied, or acceptable
compensation has been transferred Each of these
assumptions, and others not listed, can and have been
challenged
In the case of (4) above, for example, the losses
incurred when developing natural ecosystems are
involved affect species other than man. These
losses are seldom, if ever, compensated The
public at large also is usually not consulted to
determine whether the dollar compensation is
adequate and acceptable. Frequently, there is no
direct compensation. Corollary (5) can never be
fulfilled because it is impossible to determine
accurately the value to future generations.
Evaluating the contribution of functioning
natural ecosystems to human welfare is a very
complex task and involves weighing both
economic and human social values. As life support
systems, they should not be evaluated in economic
terms.
With the passage of time, man has destroyed
many of the naturally occurring ecosystems of
which he was a part and has replaced them with
simplified ecosystems wholly dependent on his
care and protection and requiring a large input of
energy.
Man favors the simple unstable and synthetic
ecosystems because when extensively managed,
and subsidized by the use of fossil fuels, they are
highly productive. An agricultural ecosystem
(agroecosystem) is an example of such a simplified
ecosystem. The effects of oxidants on
agroecosystems have been under study for more
than 20 years. The study of effects on natural
ecosystems is much more recent.
Plants grown in agroecosystems are largely
annuals and can be replaced when they are
susceptible to pollutant stress. Natural ecosystems
remain in place year after year. Manmade
pollutants are undoing relationships developed
within these ecosystems over millions of years.
EFFECTS OF PHOTOCHEMICAL OXIDANTS
ON MATERIALS
Ozone is a major factor in the overall
deterioration of several different types of organic
materials. In fact, certain specific organic
compounds are more sensitive to ozone attack
than are humans or animals. The magnitude of
damage is difficult to assess because ozone is one
of many oxidizing chemicals m the atmosphere
that contributes to the weathering of materials.
Nevertheless, researchers have shown that ozone
accelerates the deterioration of several classes of
materials, including elastomers (rubber), textile
dyes and fibers, and certain types of paints and
coatings.
Although many organic materials have been
shown to be susceptible to ozone attack, only
certain paints, elastomers, and dyes sustain
damage representing significant economic loss.
Even the measures to prevent ozone damage to
elastomers and dyes constitute a major cost.
16
-------
2. INTRODUCTION
Air quality criteria are expressions of the
scientific knowledge of the relationships among
various concentrations, averaged over an
appropriate time period, of air pollutants in the
atmosphere and their adverse effects on human
health and the environment. Criteria are issued to
assist in the formulation of decisions regardingthe
need for control of a pollutant and the development
of air quality standards governing the pollutant. Air,
quality criteria are descriptive; that is, they
describe the effects that have been observed when
the ambient air concentration of a pollutant has
reached or exceeded a specific figure for a
specified period of time.
Many factors must be considered when
developing criteria. Consideration must be given to
the chemical and physical characteristics of the
pollute its, the techniques available for measuring
these characteristics, and the exposure time,
relative humidJty, and other environmental
conditiDns, Natural and anthropogenic emissions
must 13 considered and assessed. The number and
distrib ition of sources and the extent to which the
specif'.; pollutant is emitted into the environment
direct!" affect the levels to which the receptor is
expose d. In addition, the criteria must also include
consid sration of the influence of all such variables
on the effects of air pollution on human health,
agriculture, vegetation, wildlife, visibility, and
climate. Furthermore, the individual charac-
teristics of the receptor must be taken into account.
Air quality standards are prescriptive. They
prescribe pollutant exposures or levels of effect
that scientific judgment determines should not be
exceeded in a specified geographic area, and they
are used as one of several factors in designing
legally enforceable pollutant emission standards.
This document focuses on photochemical
oxidants and their precursors, HC and NO,, asthey
are found in the ambient air. The nitrogen oxides
are examined, but only for their oxidant-precursor
role; a detailed examination of NO, as pollutants in
their own right ca n be found in Air Quality Criteria
for Nitrogen Oxides (U.S. Environmental
Protection Agency, Publication No. AP-84,
January 1971) and Air Quality Criteria for Nitrogen
Oxides (Luxembourg: NATO/CCMS, Document N.
15, June 1973). The discussion of HC in this
document rests almost entirely on their role as
precursors of other compounds formed in the
atmospheric reaction system and not on the direct
effects of HC themselves. Gas-phase HC and
certain of their oxidation products that are
associated with thi manifestations of photo-
chemical air pollution are discussed. Toxic
organics will not be discussed here. Establishing
health data for toxic organics isextremelycomplex
and will be handled at a later date as more studies
are completed.
This publication reviews the sources of oxidants
and of oxidant precursors and discusses the
chemistry of the atmospheric oxidant formation
process and the relationships between oxidants
and their precursors. Air quality data, control
techniques, and source and emission inventories
are discussed briefly for orientation purposes only,
inasmuch as these topics are discussed in detail in
other documents. Measurement techniques and
fate and transport of pollutants are discussed to
the extent necessary to give a clear understanding
of the health and welfare effects and of the bases
for sound oxidant control strategies.
The status of control technology for oxidants and
their precursors has not been treated. For
information on this subject, the reader is referred
to Control Techniques for /VO* and Hydrocarbons
from Mobile Sources (U.S. Department of Health,
Education, and Welfare, Publication No. AP-66,
March 1977), and Control Techniques for
Hydrocarbons and Organic Solvents (U.S.
Department of Health, Education, and Welfare,
Publication No. AP-68, March 1977). The subject
of adequate margin of safety stipulated in Section
108 of the Clean Air Act also has not been treated.
Again, the reader is referred to documentation
prepared by OAQPS.
Methods and techniques for controlling the
sources of photochemical oxidants as well as the
17
-------
costs of applying these techniques are discussed in
other documents.
The scientific literature has been reviewed
through 1976-77. This document is not intended
as a complete, detailed literature review, and it
does not cite every published article relating to
oxidants in the environment and their effects. The
literature has been reviewed thoroughly for
information relative to criteria. Chapters 1 1
(Effects of Photochemical Oxidants on Vegetation
and Certain Microorganisms) and 1 2 (Ecosystems)
in this document are based on the chapters by the
same numbers and titles in the National Academy
of Sciences publication, Ozone and Other
Photochemical Oxidants (Washington, D.C.,
1977). An attempt has been made to identify the
major deficiencies in our current knowledge
relative to criteria. Between the issuance of the
preprint volumes and the final publication of this
document, a few minor changes were made to
correct inadvertent omissions.
18
-------
3. NATURE AND ATMOSPHERIC CONCENTRATIONS
OF PHOTOCHEMICAL OXIDANTS
NATURE OF OXIDANT
Photochemical oxidants are chemical entities of
concern because of their detrimental effects on
biological systems and on certain materials. They
are products formed in the atmosphere by
sunlight-driven, chemical reactions that involve
HC* and NO*.
The term "photochemical oxidants" is used here
to define those atmospheric pollutants that are
photochemical reaction products and are capable
of oxidizing neutral iodide ions. Extensive research
has unequivocally identified several components
of the photochemical oxidants mixture. Thus
oxidants in ambient air are known to consist
mainly of ozone, PAN, and nitrogen dioxide (N02)
and are suspected to include also (but in smaller
amounts) other peroxyacylnitrates, hydrogen
peroxide, alkyl hydroperoxides, nitric and nitrous
acids, peracids, and ozonides. Collectively, they are
measured by potassium iodide procedures and are
referred to in this report as "oxidant." An
important distinction to be made here is that the
formation of N02 in ambient air clearly precedes
the formation of the other oxidants. For this
reason, the relative levels of nitrogen dioxide and
of the aggregate of the other oxidants vary
considerably during the day, with N02 being
invariably the dominant oxidant earlier in the day.
Because of this difference in the variation pattern
and the differences in effects between NO2 and the
aggregate of the other oxidants, the NO2 pollution
problem has been treated independently of the
other photochemical oxidant problem. Further-
more, because of its predominance among
oxidants other than NO2, ozone has been singled
out and treated as the sole representative of such
oxidants and has been given most of the research
attention.
"The term ' hydrocarbon' is meant to include also all nonhydrocarbon organic
compounds capable of participating in the atmospheric oxidant formation process
Accordingly, the terms hydrocarbons" and ' organics' are often used
interchangeably
Photochemical air pollution is customarily
defined in terms of the concentrations of ozone and
N02 only. This should not be interpreted to suggest
that other photochemical pollutants are thought to
be of less concern. Rather, it reflects (1) the fact
that ozone and N02 pollution problems are more
easily quantified and hence are more amenable to
research than problems associated with the other
pollutants (e.g., eye irritation and visibility
reduction), and (2) the assumption that abatement
of the ozone and NOs pollution problems will, in all
probability, alleviate the other photochemical
pollution problems.
The ozone found in the lower levels of the earth's
atmosphere has been traced to both natural and
anthropogenic sources. One natural source of
tropospheric ozone is the ozone abundantly
present in the stratosphere, which can be
transported into the biosphere. Ozone can also
form naturally from electrical discharges in the
atmosphere and from atmospheric photochemical
reactions involving naturally emitted organic
vapors and NOX. Obviously, the levels of ozone
resulting from all such uncontrollable sources
must be known if the anthropogenic sources and
the effects of their control are to be assessed
reliably.
Ozone and other oxidants from anthropogenic
sources are products of atmospheric photo-
chemical reactions involving primary organic and
inorganic pollutants and atmospheric oxygen.
More specifically, the oxidant formation process is
initiated by the photolysis of light-absorbing air
contaminants such as NO2, aldehydes, and ozone,
resulting in the formation of highly reactive
radicals that react subsequently with organic
pollutants. The net result of this photochemical
activity is the accumulation of ozone, and other
oxidants, to concentrations that depend on many
factors, including local meteorological conditions
(sunlight intensity, air stagnation, temperature,
etc.) as well as the concentrations, makeup, and
variation patterns of the primary pollutants
19
-------
present. Such dependence of oxidant formation on
multiple factors makes this pollution problem
immensely complex.
OXIDANT CONCENTRATIONS AND THEIR
PATTERNS
Introduction
Although buildup of photochemical oxidants
occurs in nearly every urban center in the United
States, no pollution episodes involving a sudden
and massive assault on human health have been
attributed solely to photochemical oxidants.
Accordingly, no case studies have been reported in
which pollutants and their effects were com-
prehensively and systematically examined during
an oxidant pollution episode. Photochemical
oxidants are viewed as an air pollution problem
mainly on the bases of (1) observations of humans,
animals, plants, and materials in areas known to
have high oxidant levels, and (2) results from
studies in which human subjects, animals, plants,
and materials were exposed to smoggy
atmospheres or to synthetic mixtures containing
oxidant or ozone under controlled conditions. To
present convincingly the case that photochemical
oxidant air pollution is a problem, it would be
necessary and sufficient to present (1) data on the
concentration and frequency of occurrence of
photochemical oxidants, and (2) evidence
regarding the adverse effects of oxidants. Data on
the occurrence of oxidants are given in this
chapter, but only in a brief summary form, since
the occurrence of oxidant at problem levels is well
established. Evidence on the effects of
oxidant/ozone is presented and discussed in detail
in subsequent chapters. For a more detailed
discussion of ambient levels (and variations in
those levels) of ozone and other oxidants and of
some nonoxidant photochemical pollutants, the
reader is referred to the literature.26"27
Oxidant Concentrations in Urban Atmospheres
Concentrations of oxidant in ambient air were
measured in 1974 at some 340 monitoring
stations operated by state and local control
agencies. Such data are gathered by EPA, stored in
EPA's National Air Data Bank, analyzed for trends,
and reported in summary form as EPA
reports.26'29'30'31'32
Table 3-1 shows the maximum hourly average
as well as the number of days when the maximum
hourly average oxidant concentration was equal to
or exceeded 294, 196, and 98/yg/m3 (0,1 5, 0.10,
and 0.05 ppm) for 1 2 monitoring sites in urban
areas. Higher readings of 0.52 ppm in Philadelphia
and 0.35 ppm in St. Louis were judged tc be
spurious and were discarded. More recent data for
1 6 urban areas are given in Table 3-2.
In general, locations within the Los Angeles
Basin urban area have the highest peak oxidant
concentrations (represented by the 99th percentile
values) as well as the most frequent violations of
the oxidant standard. High levels are also observed
in the northeast corridor between Washington,
D.C., and Boston, Mass. In general, most cities for
which extensive ozone data are available have
been shown to exceed the national air quality
standard for ozone of 160 /jg/m3 (0.08 ppm),
A number of these urban areas have
experienced very high levels of oxidant, exceeding
600 /jg/m3 (0.30 ppm). Los Angeles, for example,
recorded maximum 1 -hr values in excess of 1 200
/jg/m3 (0.60 ppm). Denver, Philadelphia, Houston,
and the area just east of New York - northeastern
TABLE 3-1. SUMMARY OF MAXIMUM OXIDANT CONCENTRATIONS IN SELECTED URBAN AREAS
Urban area
Los Angeles, Calif
Pasadena, Calif
San Diego, Calif
Sacramento, Calif
Santa Barbara, Calif
San Francisco, Calif
Philadelphia, Pa
St Louis, Mo
Cincinnati, Ohio
Denver, Colo
Washington, D C
Chicago, III
Year
1964-1967
1964-1967
1964-1967
1964-1967
1964-1967
1964-1967
1964-1972
1964-1972
1964-1972
1964-1972
1964-1972
1964-1972
Total
valid
data
730
728
623
7T1
723
647
1783
2014
1668
1542
2157
2042
Number of days with at leasl 1 hourly
average equal lo or exceeding
98 j/g/m3
(005 ppm)
540
546
440
443
510
185
723
1042
749
934
1032
828
196 W/mJ
(0 10 ppm)
354
401
130
104
76
29
186
152
105
179
193
85
294 f/g/mj
(01 5 ppm)
220
299
35
16
11
6
39
27
14
35
25
14
Maximum
hourly
average.
ppm
058
046
038
0 26
025
022
033
0 22
026
036
0 25
020
fjg/mj
1 137
902
745
5TO
490
431
647
431
510
510
490
392
20
-------
TABLE 3-2. OXIDANT CONCENTRATIONS OBSERVED IN SELECTED URBAN AREAS
OF THE UNITED STATES, 1974-75
Urban area
New York, N Y - Northeastern N. J
Los Angeles - Long Beach, Calif
Chicago, II! - Northwestern Ind
Philadelphia, Pa.
Detroit, Mich.
Boston, Mass
Washington, D C
Cleveland, Ohio
Minneapolis - St Paul, Minn
Houston - Galveston, Tex
Baltimore, Md
Dallas - Fort Worth, Tex
Milwaukee - Racine, Wis
Seattle - Tacoma, Wash
Cincinnati, Ohio - Northern Ky.
Denver, Colo
Total
no of
valid
sites8
8
3
6
10
2
7
8
5
2
4
2
2
7
4
6
6
No of Sites
exceeding
oxidant
standard0
8
33
6
10
2
7
8
5
1
4
2
2
7
3
6
6
Range of
2nd max
1-hr
values,
^g/m3 (ppm|
259-510(013-026)
255-784(013-040)
163-427(008-0.22)
216-625(0.1 1-032)
455-514(023-026)
186-376(0.09-019)
363-451 (0 18-023)
245-41 , (0 12-021)
141-206(007-0.10)
304-588(0 16-030)
314-372(0 16-0 19)
274-323 (0 14-0 16)
332-425(017-022)
1 18-235(006-0 12)
284-412 (0 14-021)
212-349(0.11-018)
% of days
OKtdant
standard
violated
at the
worst site
197
38.6
80
382
3 1
8 2
274
36
42
235
154
148
145
78
23 1
275
"Only sites having a minimum of 4OOO observations were included »n this summary
''The ojcidant standard is a 1 -hr average of 160 y/g/m3, not to be exceeded more than once per year
New Jersey have experienced levels in excess of
600 Aig/m3 (0.30 ppm). Hourly values exceeding
400/ug/m3 (0,20 ppm) haveoccurred in most of the
major urban areas.
The data in Table 3-1 and some of those in Table
3-2 were obtained by the potassium iodide
method,26 and therefore they include both ozone
and nonozone oxidant species. Data on ozone
alone have been obtained only for the more recent
years, after the chemiluminescence method for
measurement of ozone was introduced. National
summaries of such data are not available in a
reported form, but detailed tabulations are
available.28'29'30'31'32
Nonozone oxidants are certain to exist in urban
atmospheres, but in concentrations considerably
lower than those of ozone.7'26'27 Measurements of
such oxidants in ambient air are scant and consist
mostly of PAN data.21'22 Some recent data on
absolute concentrations of PAN and on ozone-to-
PAN or oxidant-to-PAN ratios both in urban and
rural atmospheres are shown in Tables 3-3
through 3-5.21 The data currently available
indicate, in general, that in urban atmospheres,
PAN concentrations are considerably lower than
those of oxidant or ozone, but nevertheless, they
are not negligible. Hydrogen peroxide (H2O2) also
has been reported to occur in smog chamber and
ambient atmospheres at concentrations as high as
0.18 ppm.2 Peroxybenzoyinitrate (CeHsCOOzNOz)
has been reported to occur in urban ambient air in
the Netherlands,23 but no evidence exists for its
occurrence in U.S. urban ambient air. The
significance of such PAN and other oxidant
concentrations can only be assessed on the basis
of their adverse effects on humans, vegetation,
and materials.
Oxidant Concentrations in Rural Atmospheres
Ozone concentrations in rural atmospheres
have been of interest because they were thought to
be a measure of the composite contributions of
TABLE 3-3. PAN AND OXIDANT (O») MEASUREMENTS (10 a.m. to 4 p.m.), LOS ANGELES, CALIF., 1968"
No of
samples
19
59
27
6
5
2
PAH
cone ,
ppb
0-10
10-20
20-30
30-40
40-50
> 50
Avg
PAN cone ,
ppb
8.65
130
24.0
32.6
47 1
65.5
Range at
oxidant cone .
ppb
30-190
44-220
70-290
1 00-400
1 60-285
243-410
Avg
oxidant cone ,
ppb
79.0
97.0
1440
1680
209.0
3270
Ratio
avg O,/avg PAN
9.13
758
604
523
4.49
4,94
Range of
observed
O,/PAN
5 1-244
3 3-18 3
28-130
30-97
3.2-6.1
3 9-6.0
a 1 96 j/g/m3 - 1 ppb as ozone, 5 pg/m3 ~ 1 ppb as PAN
21
-------
TABLE 3-4. PAN AND OZONE MEASUREMENTS (10 a.m. to 4 p.m.), HOBOKEN, N.J., 1970'
No of
samples
14
15
4
8
2
PAN
cone ,
ppb
0-2
2-4
4-6
6-8
8-10
Avg
PAN cone ,
ppt>
1 5
2 8
47
7 1
99
Range of
ozone cone .
ppb
15,0-1120
22.0-1500
100.0-135.0
97 0-2190
2140-2780
Avg
ozone cone .
ppb
420
71 0
1130
1620
245 0
Raiio
avg Oa/avg PAN
280
253
240
228
25 1
Range of
observed
Oy PAN
56-775
10 1-62 5
22 9-25 0
12 6-34 1
20 7-29 6
s 1 96 t/g/mj = 1 ppb as ozone, 5 pg/m3 ~ 1 ppb as PAN
TABLE 3-5. PAN, OXJDANT. AND OZONE MEASUREMENTS (10 a.m. to 4 p.m.), ST. LOUIS, MO., 19732
No of
samples
3
31
60
31
14
5
10
PAN
cone .
ppb
0-2
2-4
4-6
6-8
8-10
10-12
> 12
Avg PAN
cane ,
ppb
1.7
3.0
5 0
68
93
108
186
Range of
oxidant
cone .
ppb
46-100
46-216
30-250
39-224
40-180
30-90
52-100
Avg
oxidant
cone ,
ppb
80.6
84 6
673
68 7
846
734
856
Ratio
avg
O,/avg
PAN
475
283
134
10 1
81
68
46
Range of
ozone
cone ,
ppb
10-40
24-116
10-140
20-120
10-96
30-85
32-80
Avg
ozone
cane ,
ppb
286
48 1
502
50.3
502
540
59.5
Ratio
avg
Qa/avg
PAN
163
16.2
100
74
54
50
32
a 1 96 ^g/m3 - 1 ppb as ozone, 5 /jg/m3 - 1 ppb as PAN
natural sources to the ambient ozone problem.
Recent studies, however, have established the
occurrence of pollutant transport, thus refuting
this early notion; and it is now widely accepted that
rural areas are not necessarily unaffected by
anthropogenic pollution. This conclusion has led
air pollution specialists to use the term "rural" to
signify areas or atmospheres that are nonurban
but are occasionally susceptible to anthropogenic
pollutants, and the term "remote" to signify areas
or atmospheres so far removed from anthro-
pogenic pollutant sources that their contam-
ination by such pollutants is highly unlikely.
Before 1962, ozone concentrations both in rural
and in remote areas appeared to be considerably
below the NAAQS level of 160/jg/m3 or 0.08 ppm
(Table 3-6).17 Comparable levels were observed
also in rural North Carolina during 1 964-67.34
Measurements in the last 5 to 6 years, however,
showed that the ozone concentrations m several
rural areas exceeded 160 /ug/m3 (0.08 ppm) with a
frequency comparable to or even surpassing that
observed in many urban areas.
This fact is illustrated by the data in Table 3-7,
which were obtained by recent multiyear field
studies sponsored by EPA.35 Similar data were also
reported by New York State investigators.6 Such
increased ozone levels (i.e.,>160/L/g/m3, or>0.08
ppm) in rural areas are almost certain to be caused
largely by anthropogenic ozone and/or precursors
generated locally or transported into such areas
from urban centers.20'33'35'36'40'43 There is evidence
TABLE 3-6. CONCENTRATIONS OF TROPOSPHERIC OZONE BEFORE 19621
Observer
Cotz and Volz (1951)
Regener (1957)
Regener (1957)
Ehmerl(1952)
Teichert (1955)
Kay (1953)
Brewer (1955)
Rice and Pales (1959)
Wexler el al. (1960)
Location, time, and remarks
Arosa, Switzerland, 1950-51, high
valley, daiiy maximum values
Mt Capillo and Albuquerque,
New Mexico, 1951-52
O'IMeil, Nebraska, 1953
Weissenau, Bodensee, Germany, 1952
Lindenberg Obs., Germany, 1953-54
Farnborough, England, 1952-53
Tromsfl, Norway, 1954
Mauna Loa Observatory, Hawaii
Little America Station, Antarctica
Altitude
1860 m
3100 m
1600 m
125m above ground
20m
above ground
80 m
above ground
0-12,000 m
0-10,000 m
3000 m
100 m
Ql.lU
Range"
19-90
18-85
3-120
30-100
0-90
0-70
0-50
0-50
26-50
60-70
30-62
20-60
g/m3
Average3
50
45
36
60
35
30
30
27
38
65
45
45
a As interpreted from the published data The values sometimes represent absolute maxima, sometimes mean maxima
-------
TABLE 3-7. SUMMARY OF OZONE DATA FROM 1973-75 OXIDANT
STUDIES, RESEARCH TRIANGLE INSTITUTE36
Station
McHenry, Md.
Kane, Pa.
Coshocton, Ohio
Lewisburg, W Va
Wilmington, Ohio
McConnelsville, Ohio
Wooster, Ohio
McHenry, Md.
DuBois, Pa.
Canton, Ohio
Cincinnati, Ohio
Cleveland, Ohio
Columbus, Ohio
Dayton, Ohio
Pittsburgh, Pa
Bradford, Pa
Lewisburg, W Va.
Creston, Iowa
Wolf Point, Mont
DeRidder, La
Pittsburgh, Pa
Columbus, Ohio
Poynette, Wis
Cedar Rapids, Iowa
Des Moines, Iowa
Omaha, Neb.
Nederland, Tex.
Port O'Conner, Tex.
Austin, Tex.
Houston, Tex
Year
1973
1973
1973
1973
1974
1974
1974
1974
1974
1974
1974
1974
1974
1974
1974
1975
1975
1975
1975
1975
1975
1975
1975
1975
1975
1975
1975
1975
1975
1975
Type of
station
Rural
Rural
Rural
Rural
Rural
Rural
Rural
Rural
Rural
Urban
Urban
Urban
Urban
Urban
Urban
Rural
Rural
Rural
Rural
Rural
Urban
Urban
Rural
Urban
Urban
Urban
Urban
Rural
Urban
Urban
Average O3
ppm
0.074
0.065
0056
0.0952
0.052
0.057
0.047
0.057
0056
0035
0.025
0.031
0033
0035
0.028
0040
0.038
0035
0.028
0.030
0030
0.022
0.038
0025
0036
0035
0027
0027
0025
0.026
l/g/m3
145
127
110
187
102
112
92
112
110
69
49
61
65
69
55
78
74
69
55
59
59
43
74
49
70
69
53
53
49
51
No hours
20 08 ppm
600
639
357
249
259
262
262
262
341
148
54
51
113
114
106
100
59
17
0
38
227
43
126
6
124
64
138
99
19
141
No
hours
1662
2131
1785
1663
1751
2011
1878
2011
1667
1829
1548
1652
1935
1576
1622
2332
2386
2117
2160
2994
2841
2885
2663
2781
2528
1787
2714
2912
2504
2104
% Hours
2008
ppm
37.0
300
200
15.0
149
130
140
130
205
8.0
3.5
30
58
7.2
65
43
25
08
0.0
1 3
80
1.5
4.7
0.2
49
36
5.1
3.4
0.8
6.7
that some rural oxidant is anthropogenic in nature.
This evidence is based on measurement of Freons
or of acetylene, or on results of wind back-
trajectory analysis, both of which are indications
of the passage of the rural air over urban sources.
20,33,36,40,43 Additional data on ozone con-
centrations in remote areas are presented in
Chapter 4.
Data on nonozone oxidants in rural or remote
atmospheres are scarce. Recent data on PAN in the
rural atmosphere at Wilmington, Ohio, showed
levels considerably lower than those found in
urban atmospheres, both absolutely and relative to
ozone concentrations. Earlier as well as recent2
measurements of PAN and oxidant/ozone showed
that (1) the PAN concentrations are much lower
than those of oxidant/ozone and (2) the PAN-to-
oxidant/ozone ratio varies with location, the rural
areas showing lower values than the urban
centers.21 The maximum PAN concentration from
1 500 samples taken in August 1974 was 4.1 ppb,
with the daily maximum rarely exceeding 3.0
ppb,21 even though the ozone concentration
frequently exceeded 1 60 /vg/m3 (0.08 ppm). Such
low rural PAN concentrations, expressed in
absolute terms as well as in terms relative to
ozone, can be explained in different ways.
Accordingly, one explanation is that the chemical
mechanism of oxidant formation is such that the
lower reactant concentrations (of HC and NO*) and
the prolonged irradiation conditions that
characterize rural pollutant mixtures result in
lower PAN yields and lower PAN-to-ozone product
ratios.21 Another explanation lies in the reversible
decomposition of PAN into NC>2 and peroxyacyl
radical; in rural areas where NO? is nearly absent,
PAN decomposition is enhanced.13 A third
conceivable explanation is that the lower PAN-to-
ozone ratios in rural atmospheres may reflect a
greater proportion of stratospheric ozone relative
to the photochemically produced ozone. The
explanations based on the chemical mechanism
are believed to be the more plausible ones, at least
for the Wilmington case.21 The stratospheric ozone
explanation may be a valid one in the cases in
which PAN is undetectable.
Patterns of Variation in Oxidant Concentrations
Patterns of variation in oxidant or ozone
concentrations are of interest for at least two
23
-------
reasons. They provide a more detailed description
of the oxidant problem, which in turn serves to
guide the control effort better. Such patterns also
help to explain the chemical and physical
mechanisms by which emissions of precursors (or
of ozone) disperse, react, and ultimately cause the
observed oxidant problems.
The most conspicuous variation patterns are the
seasonal and diurnal ones. Such patterns result
largely from (1) variations in emissions of oxidant-
forming pollutants, (2) variations in atmospheric
transport and the dilution processes, and (3)
variations in other atmospheric variables involved
in the photochemical formation of oxidants.
Detailed descriptions of the seasonal and diurnal
patterns for oxidant/ozone concentrations and
some explanatory discussions can be found in the
literature,7'9'26
Briefly, the striking characteristics of the sea-
sonal anddiurnalpatternsare(1 )the occurrence of
higher oxidant concentrations in the summer
months, and (2) the occurrence of a daily oxidant
peak in the early afternoon hours. Of these, the
seasonal pattern is consistent with and, in fact,
supports the theory of photochemical formation of
atmospheric oxidant in that the higher tem-
peratures and sunlight intensities in the summer
enhance the photochemical process.19
Unlike the seasonal patterns, the diurnal pattern
can be explained in more than one way. Until
recently, the only explanation accepted was based
on the assumption that the daily oxidant peak at a
particular location resulted from local emission
sources or from those that had been transported to
the site. The reactions normally occurred several
hours before the oxidant peak.19
A more recent theory explains the oxidant peak
as the net result of three simultaneously occurring
processes:
.6,25
(1) downward transport of
oxidant/ozone from layers aloft; (2) destruction of
oxtdant/ozone on surfaces and in reaction with NO
at ground level; and (3) photochemical in situ
production of oxidant/ozone. Also, interpretation
of reported evidence suggests that downward
transport and destruction by reaction with NO,
rather than photochemical formation,6'25
determine the concentration of oxidant/ozone
observed at ground level for the cores of some
cities as well as remote areas. Whether the
oxidant/ozone in the layers aloft originates from
natural or anthropogenic sources is a question that
some investigators consider unresolved;5'6 the
bulk of the recent evidence on long-range pollutant
transport, however, points to the anthropogenic
origin.25'31
ASSOCIATION OF OZONE WITH OTHER
CONSTITUENTS OR MANIFESTATIONS OF
SMOG
The association of ambient levels of ozone and
N02 does not show well-defined patterns that
would suggest a certain impact of ozone-related
control on N02. Trend analysis of Los Angelesdata
shows the ambient N02 concentration to increase
and ozone to decrease during 1965-74 (see
Figures 3-1 and 3-2).42 These trends have been
explained as reflecting the effects of the HC and
NO, emission rate changes during that period. The
effect of the ozone-related HC control alone on N02
cannot be delineated from the apparently
overwhelming effect of the NO, emission change.
A recent study by Trijonis of the ambient N02-
precursor relationships provided evidence that the
hydrocarbon emissions factor had only a small
codirectional effect on ambient N02, whereas the
NOX factor had a much stronger effect, equivalent
to a 1:1 correspondence between NOX emission
change and ambient NOa concentration change.41
Laboratory (smog chamber) data also show (1) a
nearly 1:1 correspondence between fractional
changes in the N0« precursor factor and fractional
changes in "daily" average or maximum 1 -hr N02,
and (2) a small effect upon N02 from change in
HC.8'16
The association of ozone and PAN has been
explored using the limited amount of ambient data
available, and some explanations and
interpretations have been given of the results.
Earlier and recent measurements of PAN and
oxidant/ozone showed (1) the PAN concentrations
to be much smaller than those of oxidant/ozone,
and (2) the PAN-to-oxidant/ozone ratio to vary
with location, the rural areas showing lower
valiTes than the urban centers.21 The lower PAN-
to-oxidant/ozone ratios m rural areas can be
explained mechanistically either as a result of
enhanced PAN decomposition in NOg-free
atmospheres or as a result of the relatively lower
concentrations and prolonged irradiation
conditions, as a result of pollutant transport, that
characterize the reaction systems in rural
atmospheres. Overall, the ambient data
associations do not provide conclusive evidence on
the impact of ozone-related control on PAN. The
evidence, however, from laboratory and
theoretical studies is definitive and shows that for
urban atmospheres, changes in HC and NO,
24
-------
+36%*-;
BASIIMWIDE HC EMISSION CHANGE:
BASiNWIDE NOX EMISSION CHANGE: +36%
AVERAGE N02 CONCENTRATION CHANGE (11 STATIONS): (+20%
Figure 3-1. Trends in NOz air quality and in HC and NO, emission, Los Angeles Basin, 1965-74,
18
»=,_• IB
?1| "
jjo 12
10
<» » U E
< ui0l
0> Q Q? D.
' Xt °"
O Z
•is
! I CCl
S o a
Oo
z >
i
30
25 -
2"-
20 |-
1 5
1963 84 6S 66 67 68 6» 70 71
YEAR
ANNUAL AVERAGE
O THREE-YEAR MOVING AVERAGE
Figure 3-2. Pollutant trends in Los Angeles, annual and 3-
year moving average***2
emissions should have greater impacts on ambient
PAN than on ambient ozone,8
Ambient aldehydes, although nonoxidants, are
also of interest here because they are constituents
of the photochemical pollution complex and are
known to have some physiological effects
(Chapters 8 and 9), However, aldehydes are also
directly discharged into the atmosphere as
automotive emissions,10'39 Concentrations in
urban atmospheres seem to rise quickly in the
early morning hours, reach and maintain a broad
plateau through most of the day, and then
decrease in the afternoon.7 This pattern probably
reflects the dual origin (primary and secondary) of
the ambient aldehydes as well as the fact that rapid
photochemical formation of aldehydes, unlike
oxidant/ozone, begins with the onset of
irradiation.19 Ambient data associations do not
provide evidence on the impact of ozone-related
control on aldehydes; laboratory data, however, do
suggest a nearly 1:1 correspondence between
fractional changes in HC and fractional changes in
ambient (photochemical) aldehydes.8
The association of ambient oxidant/ozone with
eye irritation has been explored extensively, but
only in one location, the Los Angeles Basin, This
association has been described in detail in the
predecessor of this criteria document and need not
25
-------
be elaborated here Briefly, ambient
oxidant/ozone appears to correlate well with eye
irritation, a relationship which, however, is clearly
not of cause-effect nature. Lack of causative
relationship means that control of ozone is not
expected to have a direct effect on eye irritation.
However, the high degree of correlation between
the ambient data, the laboratory evidence of good
correlation between the photochemical process
and eye irritation,45 and, finally, the fact that some
eye irritants, (namely formaldehyde, acrolein, and
PAN) are known to be products of atmospheric
photo-oxidation of HC, when considered together
all suggest that oxidant-related emission control
will probably have a reducing effect on eye
irritation for Los Angeles. Eye irritation has been
observed also in other areas. A pollution episode,
for example, in New York City during November 23-
25, 1966, caused numerous eye irritation
complaints from residents,1 Oxidant/ozone
measurements were not made during the episode,
and therefore an association could not be
established. The problem could have been caused
either by primary pollutants (aldehydes) mainly, or
by both primary and secondary pollutants. In
conclusion, based on present-day knowledge and
understanding, the evidence and conclusions
obtained for the Los Angeles area cannot be
assumed to have universal validity. For such
assumption to be supported (or refuted), additional
evidence must be obtained.
The association of ambient ozone with visibility
or visibility-reducing aerosol is perhaps the most
interesting of the associations discussed here.
Public reaction to the visibility problem has been
stronger than to the ozone problem, and there is a
feeling among local air pollution control officials
that the often upsetting ozone control programs
will be more acceptable if they are shown to have a
beneficial impact on the haze problem also. In spite
of this strong interest in the ozone-haze
association, relatively little evidence is currently
available on this subject. The evidence available is
mostly on the Los Angeles atmosphere. A
considerable part was obtained in the course of
conducting the California Air Characterization
Study (ACHEX)14 and has been summarized
informatively in a recent National Academy of
Sciences report.42 The part of the evidence of most
interest here is shown in Figure 3-3 in terms of
diagrams indicating a good correlation for the Los
Angeles Basin between visibility reduction and
ambient ozone concentration. Somewhat similar,
but weaker, evidence was reported by Husar et
14
12
10
O
cj
cj
Z
Bsca, AT PEAK OZONE (BASED ON TWO HOUR
AVERAGED DATA FROM LOS ANGELES AREA)
(SUMMER 1973)
O ROUBIDOUX (RIVERSIDE)
a WESTCOVINA
D POMONA
• DOMINGUEZ HILLS
(TORRANCEi
(SUMMER 1972)
* PASADENA
• RIVERSIDE
• POMONA
• HARBOR FWY
(DOWNTOWN LA)
O BLIMP FLIGHT -
(9/6/73)
1315-1500 PST
BLIMP
NO AEROSOL IN THE ATMOSPHERE
1 I I I
01 0.2 0.3 0.4
03, ppm (MAXIMUM)
0.5
06
Figure 3-3. Correlation between BMa,
and maximal ozone concentration.15
al.15 and by Wolff et al.46 who showed that
synoptic-scale areas with elevated ozone
concentrations roughly corresponded in location
and magnitude with areas of reduced visibility
("haze blobs"). Such good correlations between
ozone and haze are self-explanatory in the cases in
which the haze-causing aerosols consist largely of
photochemical reaction products. This clearly is
the case with Los Angeles, as deduced from the
findings that the optically active aerosol in the Los
Angeles air consists primarily of photochemically
generated sulfate with lesser amounts of nitrates
and organics.3'11'12'44
The good ozone-haze correlation observed in Los
Angeles, the aerosol composition data, the good
correlation between haze and ambient sulfates,24
and, finally, the experimental and theoretical
evidence on the atmospheric photo-oxidation
process18'3 all point to the conclusion that oxidant-
related control should also cause some reduction
in haze in Los Angeles. This conclusion, with two
26
-------
important qualifications, is probably correct. The
qualifications are that (1) ozone-related control,
although beneficial, is not necessarily the most
effective means for controlling haze, and (2) the
Los Angeles evidence on the impact of ozone-
related control on haze does not have universal
validity. The second qualification is supported by
the case of Denver, for example, where the haze
problem appears to be caused by primary
aerosols.38'4 Such aerosols cannot possibly be
impacted by the ozone-related control except, of
course, when HC emission control measures
happen to reduce vehicular aerosol emissions
(e.g., measures reducing automobile traffic).
Houston also has been reported to show
interpollutant relationships different from those
occurring in Los Angeles.4
SUMMARY
Photochemical oxidants are products of
atmospheric reactions involving organic
pollutants, NOx, oxygen, and sunlight. They consist
mostly of ozone, NC>2, and PAN, with smaller
amounts of other peroxyacylnitrates and other
peroxy-compounds, and are formed along with
other photochemical products such as aldehydes,
gaseous and paniculate nitrates, and sulfates.
They originate mainly from volatile organic and
NOx emissions associated with human activities.
Photochemical oxidant formation is a complex
function of emissions and meteorological patterns.
Peak concentrations of oxidant expressed as
ozone are generally higher within urban and
suburban areas, reaching levels in excess of 590
/Kj/m3 (0.3 ppm). In rural areas, peak con-
centrations are lower but often exceed the 1-hr
National Ambient Air Quality Standard of 160
/yg/m3 (0.08 ppm). Dosages or average con-
centrations, however, are comparable to or even
higher than those in urban areas. Because of
pollutant transport, the oxidant problem is viewed
as having regional dimensions in the hundreds-of-
miles range.
REFERENCES FOR CHAPTER 3
1 Becker, W. H., F J. Schilling, and M. P Verna. The effect
on health of the 1966 eastern seaboard air pollution
episode Arch Environ Health 76.414-419, 1968
2 Bufalmi, J J , B W Gay, Jr., and K L Brubaker.
Hydrogen peroxide formation from formaldehyde
photooxidation and its presence in urban atmosphere
Environ. Sci Technol 6816-821,1972
3 Cass, G R The Relationship Between Sulfate Air Quality
and Visibility at Los Angeles EQL Memorandum No 18,
California Institute of Technology, Pasadena, CA , August
1976
4. City of Houston. Petition for Review and Revision of
Ambient Air Quality Standard for Photochemical
Oxidants and Requirements for Control City of Houston
Health Department, Houston, Tex., July 11, 1977
5 Coffey, P E.andW N Stasiuk Evidence of atmospheric
transport of ozone into urban areas. Environ. Sci
Technol 359-62, 1975
6. Coffey, P., W Stasiuk, and V Mohnen. Ozone in urban
and rural areas of New York State. In International
Conference on Photochemical Oxidant Pollution and Its
Control Proceedings Volume I. B Dimitnades, ed EPA-
600/3-77-001a, U.S. Environmental Protection Agency,
Research Triangle Park, N.C., January 1977 pp 89-96
7 Committee on Challenges of Modern Society Air Quality
Criteria for Photochemical Oxidants and Related
Hydrocarbons NATO/CCMS Publication No 29, North
Atlantic Treaty Organization, Brussels, Belgium,
February 1974
8. Dimitnades, B On the function of hydrocarbon and
nitrogen oxides in photochemical smog formation. Report
of Investigation Rl 7433, U S Department of the Interior.
Bureau of Mines, Washington, D.C , September 1970
9 Dimitnades, B Photochemical Oxidants in the Ambient
Air of the United States EPA-600/3-76-017, U.S
Environmental Protection Agency, Research Triangle
Park, N C , February 1976
10. Dimitnades, B , andT C Wesson Reactivities of exhaust
aldehydes J. Air Pollut. Control Assoc. 22 33-88, 1972.
11 Friedlander, S K Chemical element balances and
identification of air pollution sources Environ Sci.
Technol 7235-240, 1973.
12. Grosjean, D, and S. K Friedlander Gas-particle
distribution factors for organic and other pollutants in the
Los Angeles atmosphere J Air Pollut Control Assoc
251038-1044, 1975
13 Hendry, D G , and R A. Kenley Generation of peroxy
radicals from peroxynitrates(RO2NC>2) Decomposition of
peroxyacyl nitrates J Am Chem Soc 33.3198-3199,
1977
14 Hidy, G M , B Appel, R J Charlson, W E Clark, S K
Friedlander, R Giauque, S. Heisler, P K Mueller, R
Ragami, L W Richards, T B Smith, A. Waggoner, J J
Wesolowski, K T Whitby, and W White
Characterization of Aerosols in California (ACHEX) Final
Report Volume I Summary California Air Resources
Board Contract No. 358, Rockwell International Science
Center, Thousand Oaks, Calif , April 1975
15 Husar, R B,D E Patterson, C C Paley.andN V Gillani.
Ozone in hazy air masses In International Conference on
Photochemical Oxidant Pollution and Its Control
Proceedings Volume I B Dimitnades, ed EPA-600/3-
77-001 a, US Environmental Protection Agency,
Research Triangle Park, N C , January 1977. pp 275-
282
16. Jeffries, H, D Fox, and R Kamens Outdoor Smog
Chamber Studies Effect of Hydrocarbon Reduction in
Nitrogen Dioxide EPA-650/3-75-01 1 , US
Environmental Protection Agency, Research Triangle
Park, N C , June 1975
17. Junge, C E Atmospheric Chemistry and Radioactivity
Academic Press, Inc , New York, 1963
18 Kocmond, W C, and J Y Yang Sulfur dioxide
photooxidation rates and aerosol formation mechanisms
27
-------
A smog chamber study EPA/600-3-76-090, U S.
Environmental Protection Agency, Research Triangle
Park, N C , August 1976
19 Leighton, P A Photochemistry of Air Pollution, Academic
Press, Inc , New York, 1961
20 Lonneman.W A Ozone and hydrocarbon measurements
m recent oxidant transport studies In International
Conference on Photochemical Oxidant Pollution and Its
Control Proceedings Volume I B Dimnnades, ed EPA-
600/3-77-001a, U S Environmental Protection Agency,
Research Triangle Park, IMC, January 1977, pp 211-
223
21 Lonneman, W A , J J Bufaiim, and R L Seila PAN and
oxidant measurement in ambient atmosphere Environ
Set Techno! /0347-380, 1976
22 Mayrsohn, H , and C Brooks The analysis of PAN by
electron capture chromatography Presented at the
Western Regional Meeting, American Chemical Society,
November 18, 1965
23 Mei]er, G M , and H Nieboer Determination of
peroxybenzoyl nitrate (PBzN) in ambient air In Ozon und
Begleitsubstanzen im photo-chemischen Smog VDI Ber
(270)55-56, 1977
24 Miller, D F , and D W Joseph Smog-chamberstudieson
photochemical aerosol precursor relationships EPA-
600/ 3-76-080, U S Environmental Protection Agency,
Research Triangle Park, N C , July 1976
25 Mohnen, V , and E Reiter Internationa! Conference on
Oxidants, 1976—Analysis of Evidence and Viewpoints
Part III The Issue of Stratospheric Ozone Intrusion EPA-
600/3-77 115, US Environmental Protection Agency,
Research Triangle Park, IM C , December 1977
26 National Air Pollution Control Administration. Air Quality
Criteria for Photochemical Oxidants. NAPCA Publication
No AP-63, U S Department of Health, Education, and
Welfare, Public Health Service, Washington, D, C , March
1970
27 National Research Council. Ozone and Other
Photochemical Oxidants National Academy of Sciences,
Washington, D C , 1977
28 Office of Air Quality and Planning Standards Monitoring
and Air Quality Trends Report, 1972 EPA-450/1-73-
004, U S Environmental Protection Agency, Research
Triangle Park, N C , December 1973
29 Office of Air Quality and Planning Standards Monitoring
and Air Quality Trends Report, 1973 EPA-450/1-74-
007, U S Environmental Protection Agency, Research
Triangle Park, N C , October 1 974
30 Office of Air Quality and Planning Standards Monitoring
and Air Quality Trends Report, 1974 EPA-450/1-76-
001, US Environmental Protection Agency, Research
Triangle Park, N C , February 1976
31 Office of AirQualityand Planning Standards National Air
Quality and Emission Trends Report, 1 975 EPA-450/1 -
76-002, U S Environmental Protection Agency,
Research Triangle Park, N C , November 1976
32 Office of Air Quality Planning and Standards The
National Air Monitoring Program Air Quality and
Emissions Trends Annual Report Volumes I and II EPA-
450/1 -73-O01a and b, U S Environmental Protection
Agency, Research Triangle Park, N C , August 1973
33 Rasmussen, R A Surface ozone observations in rural
and remote areas J Occup Med 75346-350, 1976
34 Ripperton, L A, and J J. B Worth Chemical and
Environmental Factors Affecting Ozone Concentration in
the Lower Atmosphere Environmental Science and
Engineering Publication No 234, University of North
Carolina, Chapel Hill, N C,, 1969.
35 Ripperton, L A.J J B. Worth, F M. Vukovich, and C. E
Decker Research Triangle Institute studies of high ozone
concentrations in nonurban areas. In International
Conference on Photochemical Oxidant Pollution and Its
Control Proceedings Volume I B Dimitnades, ed EPA-
600/3-77-001 a, U S. Environmental Protection Agency,
Research Triangle Park,,NC, January 1977. pp 413-
424
36 Robinson, E , and R A Rasmussen Identification of
natural and anthropogenic rural ozone for control
purposes In Speciality Conference on Ozone/
Oxidants—Interactions with the Total Environment Air
Pollution Control Association, Pittsburgh, Pa., March
1976 pp 3-13
37 Russell, P A (ed ) Denver Air Pollution Study—1973
Proceedings of a Symposium Volume I. EPA-600/9-76-
007a, U S. Environmental Protection Agency, Research
Triangle Park, N C., June 1976
38 Russell, P A (ed ) Denver Air Pollution Study - 1973
Proceedings of a Symposium Volume II EPA-600/9-77-
001, US Environmental Protection Agency, Research
Triangle Park, N C , February 1977
39 Seizinger, D E, and B Dimitnades Oxygenates in
exhaust from simple hydrocarbon fuels J Air Pollut
Control Assoc 2247-51, 1972
40 Spicer, C W,J L Gemma, D. W Joseph, P S Sticksel
and G F Ward The Transport of Oxidant Beyond Urban
Areas EPA-600/3-76-018a, US Environmental
Protection Agency, Research Triangle Park, N C ,
February 1976
41 Tnjonis, J Empirical Relationships Between
Atmospheric Nitrogen Dioxide and Its Precursors EPA-
600/3-78-018, US Environmental Protection Agency,
Research Triangle Park, N C , February 1978
42 Tnjonis, J C, T K Peng, G J McRae, and L Lees
Emissions and Air Quality Trends in the South Coast Air
Basin EQL Memorandum No. 16, California Institute of
Technology, Pasadena, Calif , January 1976
43 Westberg, H , K Allwine, E Robinson, and P
Zimmerman Light Hydrocarbon and Oxidant Transport
Studies m Ohio—1974. EPA-600/3-78-007, U.S
Environmental Protection Agency, Research Triangle
Park, N C , January 1978
44 White, W H., and P T Roberts The nature and origins of
visibility-reducing aerosols in Los Angeles Presented at
the 68th Annual Meeting, Air Pollution Control
Association, Boston, Mass , June 15-20, 1975
45 Wilson, K W Survey of Eye-Irritation and Lachrymation
in Relation to Air Pollution, Coordinating Research
Council, Inc , New York, April 15,1974
46 Wolff, G T,P J Lioy, G D Wight, R F Meyers, and R.T
Cederwall An investigation of long-range transport of
ozone across the midwestern and eastern United States
In International Conference on Photochemical Oxidant
Pollution and Its Control. Proceedings Volume I. B
Dimitnades, ed EPA-600/3-77-001 a, US En-
vironmental Protection Agency, Research Triangle Park,
N C , January 1977 pp 307-317
28
-------
4. SOURCES AND SINKS OF OXIDANTS
PHOTOCHEMICAL FORMATION OF OXIDANTS
Introduction
All the evidence presently available indicates
that in the urban centers and adjacent downwind
areas where severe oxidant problems occur, the
major cause by far is photochemical oxidant
formation. In areas with less severe problems, and
especially in remote areas, nonphotochemical
sources (e.g., stratospheric ozone intrusion) may
be significant and on occasions may be the
dominant sources.
When photochemical oxidants in air were
identified as products of a photochemical process
involving primary air pollutants, it became'
immediately apparent that the problem could not
be abated by traditional methods of direct control.
The chemistry of this process had to be clearly
understood before rational abatement measures
could be devised. Accordingly, in the ensuing
years, numerous studies of the oxidant chemistry
were conducted with the following specific
objectives:
1. To identify the precursorsof photochemical
oxidants (that is, those primary pollutants
that participate as reactants in the oxidant-
forming process) for the purpose of
directing the control effort to the right
targets.
2. To determine the kinetics of the precursor
reactions mainly for the purpose of
deducing the impact of precursor control on
oxidant formation.
3. To determine the stoichiometry of the
precursor reactions for the purpose of more
fully assessing the pollution problems
caused by such precursors.
4. To elucidate in detail the oxidant-forming
mechanism for the purpose of providing
deterministic, as distinct from empirical,
input to the oxidant abatement effort, and
also for the purpose of relating oxidant
formation to other manifestations of
photochemical pollution.
Several of these objectives have been achieved,
at least to the degree of completion such that the
information generated was sufficiently compre-
hensive and reliable to permit development of a
crude but promising oxidant abatement strategy.
These chemical studies clearly are relevant to the
purposes of this document; however, they need not
all be included in this review or expanded on in
detail except where they pertain to applications
directly related to the questions of oxidant control.
To serve the purposes of this document better,
the discussion of the chemistry pertaining to
photochemical oxidants will be divided and
presented in several sections. The following
section on atmospheric reaction mechanisms
includes a general discussion of the chemical
processes and a review of recent atmospheric
chemistry studies, focusing only on those findings
that are judged to have had the strongest impact on
the understanding of the oxidant formation
mechanism. It is intended and hoped that this
introductory discussion and review will prepare
the reader to understand better the subsequent
sections dealing with the specific applications of
the information used in formulating oxidant
control strategies. In those sections, the reaction
mechanisms and other pertinent aspects of the
oxidant chemistry will be discussed critically and
more comprehensively. For additional detailed
discussions of the earlier studies of atmospheric
chemistry, the reader is referred elsewhere.54'78'92'
93
Finally, there is one important difference in
emphasis between the following reaction
mechanism section and the preceding chapters. In
the preceding chapters, ozone and other oxidants
were dealt with nearly always collectively, using
the term "oxidant." In the discussions of reaction
mechanisms, however, all mechanistic inter-
pretations regarding oxidant, in actuality, pertain
to ozone alone. Analogous mechanistic
discussions addressed to other oxidants will not be
included in this document mainly because the
29
-------
emphasis is on ozone and because the mechanistic
information available for other oxidants is
considerably deficient.
Atmospheric Reaction Mechanisms
The pioneering work of Haagen-Smit in 195254
first demonstrated through laboratory ex-
perimentation that the photochemical oxidants
present in an urban atmosphere may indeed be
products of atmospheric photochemical reactions
involving organic (e.g., hydrocarbons) and
inorganic (e.g., nitrogen oxides) pollutants. Since
then, numerous studies of oxidant formation have
been conducted,4'5'78'93 and a wealth of information
is now available regarding the stoichiometry,
kinetics, and mechanisms of these HC-NOx-air-
sunlight reactions. The following discussion
summarizes the results of these studies and
presents highlights of the chemical mechanism
that is presently thought to explain best the
phenomenon of photochemical oxidant formation.
All experimental reserach concerned with the
mechanism of atmospheric oxidant formation was
conducted in the laboratory using experimental
conditions tKat were.similar to but not nearly as
complex as those prevailing in the ambient
atmosphere. Such simplifications of the natural
system were necessary to facilitate research, but
they also limited the validity of extrapolating
research findings to ambient conditions. Thus
most mechanistic evidence obtained to date
pertains solely to the oxidant formation process
occurring in laboratory systems. The most that can
be said at this time is that all of the reaction steps
known to occur in the experimental reaction
systems are almost certain to occur in the ambient
atmosphere also, although perhaps with varying
significance. The converse, however, may not
necessarily be true; that is, unidentified reaction
steps may occur in the ambient atmosphere but not
in laboratory simulations.
The overall chemical changes that are observed
to occur when a mixture of HC and NO, pollutants
is exposed to sunlig1-.,. under conditions similar to
those in the atmosphere are illustrated in Figure 4-
1. The curves of Figure 4-1 depict the overall
photochemical process as consisting of two
distinct reaction stages occurring consecutively.
During the first stage, nitric oxide (NO) is converted
into nitrogen dioxide (NOz) without anyBppreciable
buildup of ozone (O3) or other non-NC>2 oxidants.
The second stage starts when almost all of the HO
has been converted into NOZ and is characterized
IRRADIATION TIME
IMW2 —
01 n
On 4 NO
yn \ NO
9Mn 4- n..
light
M
^~INVJ T vj
-K)3+ M
-^•NO 4- O
-^.MPi 4- vn
^HNL<2 ^ AU
-k-iNn..
Figure 4-1. Chemical changes occurring during photoirrad-
iationof hydro-carbon-nitrogen oxide-air systems
by rapid accumulation of ozone and other oxidant
and nonoxidant products.
Several mechanisms explaining the obser-
vations depicted in Figure 4-1 have been
postulated. Although these mechanisms differ
substantially in chemical detail, they all explain
ozone formation in the atmosphere to be the net
result of the following main reactions:
(4-1)
(4-2)
(4-3)
(4-4)
(4-5)
where M is the third body in a reaction, and X is
hydrogen (H) or organic radical (R or RCO). Through
this mechanistic procedure, the NO product from
the photolysis of NO2 reacts rapidly with and
consumes ozone to regenerate the photolyzed
NO2. Therefore, unless other processes convert
the NO into NO2, ozone is not allowed to
accumulate to significant levels. However, other
NO conversion processes do exist—mainly the
reactions of NO with XO2 and, to a much lesser
degree, with molecular oxygen (O2>. The reaction
with XO2 occurs only when photochemically
reactive organic compounds are present, in which
case the reaction can be sufficiently rapid to cause
atmospheric accumulation of ozone to significant
levels. In the absence of reactive organics, the only
NO conversion process parallel to the O3-NO
reaction is the reaction of NO with 02; this
reaction, however, is relatively slow and does not
cause significant ozone accumulation.
In mathematical terms, the ozone buildup in the
atmosphere obeys the equation28
[O3] = [NO2] (4-6)
[NO]
30
-------
where I is light (sunlight) intensity, and k is a
constant. This equation is derived by applying the
steady state hypothesis to chemical reaction
steps 4-1, 4-2, and 4-3. This equation illustrates
the effect on ozone buildup of any process that
converts NO into NO2 (e.g., the reaction of NO with
X02>; such effect is toward high [N02]: [NO] ratios,
and hence high levels of ozone buildup. Ambient
measurements in downtown Detroit132 in the
summer of 1973 and from the Los Angeles
Reactive Pollutant Project (LARPP)18 have
essentially verified the photostationary state
equation in tie atmosphere within the stochastic
limits of the turbulent atmosphere.124
Though the overall mechanistic scheme
depicted by processes 4-1 through 4-5 was
established long ago, the detailed reactions
explaining the consumption of the organic and
inorganic reactants and the formation of the X02
radicals (e.g., reaction 4-4) have been relatively
obscure. It is in this latter mechanistic area that
some important developments have taken place in
the past 5 to 6 years. These new findings and their
impact on current understanding are summarized
here. The significance of these findings in the
specific applications of reaction mechanisms in
the development of mathematical models will be
discussed in a subsequent section of this
document.
One important accomplishment in recent years
in the area of atmospheric reaction mechanisms is
the development of computer techniques for
simulating the atmospheric smog-forming
process.19'40 Such techniques have provided a
useful tool for selecting those reactions that play a
key mechanistic role and for predicting the
existence of potentially important but as yet
unidentified reaction products in the ambient
atmosphere.
The most important mechanistic finding,
however, in recent years pertains to the identities,
sources, and roles of the radicals responsible for
the oxidation of the organic and inorganic
reactants. Unlike the early emphasis on atomic
oxygen, 0(3P), and ozone roles, the current
thinking attributes to the hydroxyl radical (OH)
most of the hydrocarbon- and aldehyde-
consuming chemical activity.58'133
Other radicals also, most notablyatoms and RO2,
can attack and consume the organic reactant
significantly, as illustrated by the data inTable4-1.
These data, derived from a computer simulation of
an irradiated frans-2-butene/NOx reaction sys-
tem, represent calculated rates of attack of several
TABLE 4-1. CALCULATED RATES OF ATTACK ON
7"fl/l/VS-2-BUTENE BY VARIOUS REACTIVE SPECIES IN
SIMULATED SMOG SYSTEM AT SEVERAL
IRRADIATION TIMES"
Attack rate, ppb/mm
Species
0
03
OH
H02
N03
02 (cr'A)
2 mm
0013
0026
1 72
0.16
0.05 x 10^"
2.9 x 10"6
30 mm
0018
0 16
055
0.15
22x10""
2 1 x 1Q"6
60 m
0.011
016
0.27
009
2.6x
1.3 x
m
10'4
10-"
aData from Demerjian et al 40 Initial conditions [NO]. 0 075 ppm, [N02], 0 025 ppm,
[frans-2-butene], 0 1 0 ppm, [CO], 10ppm, [CH2O], 0 10 ppm, [CH3CHO], 0 06 ppm,
[CH<], 1 5 ppm. relative humiuity, 50%
reactive species on butene, at several irradiation
times.
The radicals OH, H02, and X02 have also been
identified as having major roles in the oxidation of
NO into N02 (reaction 4-4).93 Sources of these
important radicals are now believed to be the
reactions:
(Aldehydes, MONO, 03, others)
hv/ »OH, H02, R02, others
(4-7)
Oxidation of NO into N02 occurs via a chain
mechanism illustrated by the following sequence:
OH + (hydrocarbons, aldehydes)
H02, R02, others (4-8)
H02+NO - N02 +OH (4-9)
R02+N0-R0 (4-10)
RO + 02 - H02 + aldehyde (4-11)
Such chain processes are terminated when
radicals react to form more stable products, for
example.
OH + N02-HON02 (4-12)
H02 + H02 (or R02) - H202 (or ROOH) (4-13)
HO2 + NO2 - MONO (or HOON02?) (4-14)
RC(O)O2 - RC(O)02N02 (i.e., PAN) (4-15)
HO2, RO2 + surface - (H02, R02) surface (4-16)
A more complete picture of the reactions
accounting for the organic oxidant degradation,
oxidation of NO into NO2, and organic product
formation in the ambient atmosphere, is illustrated
in Figure 4-2 for the case in which the organic
reactant is an olefin.40 Comparably complete
mechanistic information is now available for
paraffins also, but not for aromatic hydrocarbons.
In the inorganic part of the atmospheric reaction
mechanism, some steps recently have becomethe
subjects of controversy because of their different
roles in the ambient atmosphere and in the
laboratory systems. Thus, the reactions
31
-------
il14.
O
CHBCH-CHCHS
CHSCH=CHCH3
-»-CH,=CHCH(O, H)CH,
0('P)
CH3CH(6)CH(6I)CH|
CH3CH(O)CHCHJ
*>CH,CO, + NO,
AS BEFORE
* These products undergo significant photodecooiposition in sunlight
L.
Figure 4-2. The major reaction paths for the degradation of ffans-2-butene in an irradiated IMO,-polluted atmosphere.
32
-------
CHjCH^HCH,
HO/""
CH,CH=CHCHa + H,O
CH,CH=CHCH,6,
O^o
I I
CH,CHCHCH,
<°"
O
CHjCHCHiO^CHj
HO
CH,CHOHCHCH,
O,
1
"
NO
CH,CHOHCH(O)CH9 ^ NO,
CHjCHOH
I'
CHaCH(0,)OH
| NO
CHjCH(6)OH -i- NOa
Figure 4-2. (Continued). The major reaction paths for the degradation of frans-2-butene in an irradiated NO.-poiluted
atmosphere.
33
-------
NO + N02 +H20 - MONO (4-17)
N205 + H20 - 2HN03 (4-18)
are known to occur both homogeneously (in the
gas phase) and heterogeneously. Both reactions
are thought to be of relatively little importance in
the ambient atmosphere but are suspected to
occur significantly on the reactor walls in
laboratory systems.1771 Questions are thus raised
about the applicability of the laboratory findings to
the ambient atmosphere, which is an issue of
crucial importance.
Another noteworthy finding is that some
nitrogenated products of the photochemical smog
system, such as PAN and possibly peroxynitric acid
(HOON02) have a greater mechanistic role than
thought earlier. Recent investigations of PAN
chemistry 34'60'98 have revealed that PAN can
thermally decompose to an acylperoxy radical and
N02. The rate of decomposition is extremely
temperature dependent. Because of this tem-
perature dependency, significant levels of PAN can
build up early in the day when temperatures are
relatively low. In the late afternoon when ambient
temperatures are higher, the decomposition of
PAN can proceed at a rapid rate, liberating N02
molecules that can lead to enhanced ozone
production.
The discovery that PAN can thermally
decompose and enhance ozone formation
suggests that there may be other peroxynitrates
that can also affect the rate of smog formation.
Recently, peroxynitric acid (HOON02) was
identified by Fourier-transform infrared spec-
troscopy (FS)
79,95
as an intermediate of
photochemical smog systems. This species and the
related peroxynitrates (ROON02) could act as
radical sinks and affect the rate of smog formation.
The important H02 and R02 free radicals, as well as
NC>2, could be temporarily stored as peroxynitrate
and then released (through thermal decomposition
of the peroxynitrate) at later stages in the reaction
to enhance ozone production. However, recent
experimental evidence33'50'125 indicates that the
decomposition of HOONOz proceeds so rapidly at
room temperature that this species will not be a
significant sink for HC>2 or N02. Consequently,
HOON02 is not likely to be of significance in
photochemical smog formation.
This role of the peroxynitrate products raises the
more general question regarding the role of
atmospheric reaction products in the chemistry of
polluted air masses irradiated for several days.
Such multiday reaction of pollutant mixtures is
now known to occur in the ambient atmosphere
and specifically, within urban air masses subjected
to long-range transport.136 While the chemical
mechanism of such aged systems cannot be
different from that of fresh systems in terms of
fundamental reaction steps involved, the relative
rates of such steps may be sufficiently different
that the relative roles of the organic and NO*
reactants, as well as the relative reactivities of the
various organics in producing oxidant/ozone, may
be substantially different. Evidence attesting to
such differences has recently been reported by
researchers who did laboratory and computer
simulation studies of fresh and aged HC-NO*
reaction systems. Thus smog chamber studies
have suggested that the oxidant-to-precursor
dependencies in aged systems may be different in
magnitude, and possibly in direction also, from the
dependencies observed in fresh systems.41'71
Current knowledge and understanding of the
atmospheric chemistry of aged polluted air are
considerably inferior to those pertaining to fresh
atmospheres.
The chemical transformation of S02 to par-
ticulate sulfates is of considerable interest, both in
terms of understanding S02 removal mechanisms
in the atmosphere as well as the potential health
and welfare effects96 associated with particulate
sulfates. Though it is beyond the scope of this
document to provide a detailed review of the state-
of-the-art of SOx chemistry, a short discussion
regarding its interaction in the photochemical
process seems warranted.
In recent years, significant progress toward
understanding the homogeneous SO2 oxidation
pathways in the atmosphere has been made.16'
119'122 One of the more important contributions has
been the experimental gas kinetic rate constant
determinations for the H02 and HO reactions with
S02: 20'99
H02 + S02- H0 + S03
HO + S02 + M - HOSO2+ M
The quantification of these reaction rates in
conjunction with theoretical estimates of the
reaction rates for the organic radical analogies
R02 and RO have explained major portions of the
S02 transformation observed in both experimental
smog chamber studies and field investigations.
16,119,142
In the presence of water vapor, S02 reacts
rapidly to form H2SO4. The details of the reaction
pathway for the HOS02 species is currently
unknown, but evidence exists indicating that
34
-------
H2S04 is the resulting end product. In the
atmosphere, H2S04 may undergo complex inter-
actions with aerosol particulate or react with
gaseous ammonia to form a host of sulfate salts.
Characterization of these condensable species via
experimental and ambient observations will be a
major area of continued research and will play a
significant role in elucidating the mechanistic
details of atmospheric SO*.
Outside the reaction mechanism area, some
significant developments have taken place in
recent years in the area of reaction product
identification. Table 4-2 lists gaseous products a nd
typical concentrations at which such products
were found to occur in the ambient atmosphere.
Table 4-3 lists gaseous products either observed in
laboratory (smog chamber) systems or unreported
yet but theoretically expected to exist in the
ambient atmosphere. Particulate products also
exist but are not easy to identify. Such products are
certain to include nitrates and a variety of organic
compounds. Tables 4-4 and 4-5 list some reaction
products identified in ambient atmosphere
aerosols.93 The significance of these findings is
that they provide a more complete description of
the photochemical pollution problem and that they
verify orfurtherelucidatethe atmospheric reaction
mechanism.
TABLE 4-2. COMPOUNDS OBSERVED
IN PHOTOCHEMICAL SMOG
TABLE 4-3. COMPOUNDS THAT MAY BE FORMED IN
PHOTOCHEMICAL SMOG
Typical (or maximal)
concentration reported,
Compound
Ozone, O3
PAN, CH3COO2NO2
Hydrogen peroxide, H2O2
Formaldehyde, CH2O
Higher aldehydes, RCHO
Acrolem, CH2CHCHO
Formic acid, HCOOH
01 (06)
0005 (0.2)
(0.18)
(016)
(0.36)
(0011)
(005)
004
In conclusion, the information generated in the
years since the issuance of the preceding criteria
document provides a much more complete picture
of the atmospheric oxidant formation process.
Nevertheless, it should be stressed that the
present situation is somewhat uncertain because
most of the supporting evidence was obtained from
laboratory atmospheres and is therefore mainly
applicable to such an environment. The
mechanism of the oxidant-forming process in the
ambient atmosphere, containing a multitude of
reacting pollutants, may include reaction steps in
addition to those presently recognized. Such
Compound
Peroxybenzoyl nitrate,
C6H5C002N02
Nitric acid, HONO2
Organic hydroperoxides,
ROOM
Organic peracids, RCOO2H
Organic peroxynitrates,
RO2NO2
Ozonides, O3-olefin
Ketene, CH2CO
Nitrous acid, MONO
Pernitnc acid, HO2NO2
Pernitrous acid, HO2NO
Organic nitrates, RONO2
Possible synthesis
0COO2 + NO2
N02 + OH
N2O5 + H2O
RO2 + HO2
RCOO2+ HO2
RO2 + NO2+ M*
O3 + olefm + M"
O3+ olefm
OH + NO
N02 + HO2+ M*
NO + HO2 + Ma
RO + NO2
RO2 + NO
Reference
62
55
91
40
40
40
7
88
8,32
50,79,95
31
37
37
aM represents any molecule that takes part in the three-body process
additional reactions, for example, known or
suspected to occur in the ambient atmosphere, are
those causing degradation of aromatic hydro-
carbons and of numerous organics into gaseous
and particulate products. To date, these
degradations are largely unexplored Important
reactions may also be occurring through energy
transfer processes promoted by pollutants or other
molecules capable of absorbing solar energy and of
transferring such energy to nonabsorbing
pollutants. Also, the nature and importance of
heterogeneous reactions occurring on the surface
of ambient aerosol particles49 and on the surface of
laboratory chambers17 have been explored but are
not well understood. Further studies are needed to
ascertain the applicability of laboratory data to
ambient atmospheres.
Effects of Meteorological Factors
INTRODUCTION
The photochemical oxidant/ozone con-
centrations observed m the ambient air above
urban and nonurban areas arethe net result of two
broadly defined processes: First, a physical process
involving dispersion and transport of the oxidant-
precursor emissions; and second, a chemical
process involving reaction of the dispersed
pollutants under the stimulus of sunlight. The
potential effects of meteorological factors on both
these processes are obvious. Thus the factors
related to atmospheric dilution and transport affect
ambient levels of pollutant accumulation and the
geographical relationship between source areas
and corresponding oxidant problem areas. Solar
radiation and ambient temperature are also
important by virtue of their effects on the chemical
35
-------
TABLE 4-4. SECONDARY ORGANIC AEROSOLS2
Compounds identified"
Possible gas-phase hydrocarbon
precursors
Aliphatic multiiunctional compounds
1. X-(CH2)n-Y (n=3,4,5)-
X Y
COOH CH2OH
COOH COM
COOH COOH
COOHb CH2ONO
or COH CH2ONO2
COH CH2OH
COH COH
COOHb COONO
or COH COONO2
COH COONO
COOH COONO2
COOH CH2ONO2
2 Others
CH2OH-CH=C(COOH)-CHO
CH2OH-CH2-CH=C(COOH)-CHO
CHO-CH=CH-CH(CH3)CHO
CH2OH-CH=CH=CHb-C(CH3)CHO
CsHeOs isomers
Nitrocresols
CeH6O2 isomers"
Aromatic monofunctional compounds
3 C6H5-(CH2)n-COOH (n = 0,1,2,3)
4 C6H5-CH2OH
C6H5CHO
Hydroxynitrobenzyl alcohol
Terpene-denved oxygenates.
5 Pinonic acid
Pinic acid
Norpmonic acid
6 Isomers of pmonic acid "
CgHitOz isomers
CioHi4O3 isomers
CioHi6O2 isomers
1 Cyclic olefms
(CH2)n
CH
CH
and/or diolefms
>C=CH-(CH2)n-CH=C<
2 Not known; possibly from aromatic ring cleavage
3 Alkenylbenzenes
C6H5-(CH2)n-CH=CHR; also toluene for C6H6COOH
4 Toluene, styrene, other monoalkylbenzenes7
5 a-Pmene
6 Other terpenes?
"Compounds identified at West Covma, Calif, July 24, 1 974
blsomers not resolved by mass spectrometry
processes. Such meteorological influences
obscure both the absolute and the relative effects
of the emission-related factors to degrees thatvary
with geographical location and season. Therefore,
for proper assessment of emission-related factors,
it is essential that meteorological factors and their
effects on the oxidant problem be well understood.
Such understanding often depends on mete-
orological details that are far beyond the scope of
this discussion, but some general climatic factors
may be discussed to illustrate their impact on air
quality.
ATMOSPHERIC MIXING
The diurnal urban emission pattern for oxidant-
forming pollutants is fairly uniform from weekday
to weekday. It is apparent, therefore, that
variations in daily oxidant/ozone accumulation
must be attributable largely to meteorological
factors. Of these factors, atmospheric mixing has
an extremely significant effect on oxidant
formation.
The rate and extent of atmospheric mixing and
diffusion depends on stability, wind speed, and
topography. Temperature inversions are relatively
stable layers that inhibit atmospheric diffusion.
They are common at night in layers near the
ground, but they may also occur at higheraltitudes
and at times other than at night. Also, several
successive atmospheric layers with different
degrees of stability are not unusual. The morning
peak concentrations of oxidant/ozone precursors
36
-------
TABLE 4-5. RELATIVE IMPORTANCE OF ALIPHATIC AND AROMATIC PRECURSORS"
Secondary organic
aerosols
X-(CH2)n-Y
Gas-phase hydrocarbon
precursors X
COOH
COOH
CH COOH
(CH2)n COOH
CH or COH
COH
COH
. . COOH
and/or or CQH
COH
COOH
COOH
Y
CH2OH
COH
COOH
CH2ONO
CH2ONO2
CH2OH
COH
COONO
COONO2
COONO
COONO2
CH2ONO2
Concentration," /yg/m3
n=3
2 18
1 39
1 35
1 01
—
0 31
0.30
0 14
1 01
—
0 12
n=4
3.40
259
0.78
0.40
—
0.40
0 24
024
0.14
—
0 15
n=5
065
082
0 15
0 27
—
013
—
—
—
—
—
>C=CH-(CH2)n-CH=C<
C6H5-CH=CHR
Total
COOH-CHz-COOH
COOH-(CH2)2-COOH
Total difunctional compounds 18.89
C6H5-(CH2)n-COOH
Total from aromatics.
781
0.15
057
834
n=0.0.38
n-1 041
n-2 0 52
n-3 0 03
1.34
2.02
'Of aerosols in Pasadena, Calif. Sept 22, 1972, sampling period, 7 30 a mjo 1 2 35 p m
The same response factor (that of adipic acid) was used for all difunctional compounds
often result from emissions from ground sources
discharged into a stable atmosphere below an
inversion layer. Typically, these low-level
inversions begin to lift and provide a greater mixing
height as the sun rises and heats the earth's
surface. Thus mixing height (that is, the height
above surface through which relatively vigorous
vertical mixing occurs) varies during the day and
with the season. Figure 4-365 shows average
mixing heights in summer (usually the season of
highest photochemical oxidant concentrations)
that occur a few hours after sunrise. As surface
heating continues, the low-level inversion is
usually eliminated completely and, as depicted in
Figure 4-4,65 the mixing height increases to a
maximum in the afternoon. Notice that these
average afternoon mixing heights range from 600
m (1970 ft) along the California coast to more than
4000m (13,000 ft) within the southern Rocky
Mountain area. Over inland areas, the afternoon
mixing heights are greatest in summer and least in
winter (about half of the summer values). How-
ever, seasonal variations in coastal regions are
usually small. To a large extent, the height of the
afternoon mixing layer has a significant impact on
the day's accumulation of oxidant/ozone.
The rate at which relatively unpolluted air moves
into a region is related to wind speed. Wind speed
and direction are highly variable in time and space
(Figures 4-5 and 4-6).65 The depicted wind speeds
are averages of those speeds within the respective
mixing layers. These averages tend to smooth out
the greater variations in spaceandtime that would
occur it only surface winds had been considered.
During summer mornings, the slowest average
speeds are less than 2 m sec"1 and occur in the Far
West. However, in the morning, average speeds of
3 m sec^1 or less occur over most of California and
Oregon as well as part of Nevada; in the East they
occur over the Central Appalachians and
Mississippi. On summer afternoons, the average
wind speed pattern is similar to that in the
morning, except that the average speeds in the
afternoon tend to be 1 to 2 m sec' higher than in
the morning. This diurnal variation is generally
true for all seasons.
Using daily values of morning and afternoon
mixing height and wind speed, the occurrence of
limited mixing episodes was determined for 62
weather stations in the contiguous United
States.65 The most limiting conditions were mixing
heights of 500 m (1 640 ft) or less, wind speeds of 2
37
-------
6 5
Figure 4-3. Isopleths (m * 102) of mean summer morning mixing heights.65
Figure 4-4. Isopleths (m * 102) of mean summer afternoon mixing heights.65
38
-------
Figure 4-5, Isopleths (m sec ') of
mean summer wind speed averaged through the aftwnoon misting layer,"*
Figure 4-6, Isopleths (m sec"') of mean summer wind speea averaged through the morning mixing layer,66
39
-------
10
DATA BASED ON FORECASTS ISSUED:
1 AUGUST 1960 TO 3 APRIL 1970 FOR EASTERN PART OF THE UNITED STATES
1 OCTOBER 1963 TO 3 APRIL 1970 FOR WESTERN PART OF THE UNITED STATES
Figure 4-7. Isopleths of total number of forecast days of high meteorological potential for air pollution.65
m sec 1 or less, and no precipitation during at least
5 consecutive days. Only six such epidoses
occurred with a total of 39 episode-days at two
stations (Medford, Oregon, and Lander, Wyoming)
over a 5-year period. On the other hand, at least
one episode of 2 days' duration, with mixing
heights of 1000m (3300 ft) or less and wind
speeds of 6 m sec ' or less, occurred at each of the
62 stations. In general, limited mixing episodes are
most common in the Far West and within the Rocky
Mountain region, least common over the Plains
States, and of intermediate frequency east of the
Mississippi (Figure 4-7).66
Korshover76 compiled statistics on stagnation
episodes occurring during the various months of
the year east of the Mississippi. Such statistics are
illustrated by graphs in Figures 4-8 and 4-9.
Korshover's data, however, suffer from two
limitations First, stagnation episodes were
determined indirectly, using sea level pressure
gradient data, rather than the more direct indices
of mixing height and wind speed; second, the
requisite pressure gradient data could be obtained
only for the relatively low and flat terrain east of the
Rocky Mountains. A more detailed discussion of
the Korshover statistics is included in the chapter
dealing with the natural sources of oxidant/ozone.
In conclusion, the atmospheric mixing
parameters have a strong effect on both the
accumulation of oxidant-precursor pollutants and
the photochemical oxidant-forming process. The
effect on the latter process, however, although
certain to exist, has only recently been studied
directly and comprehensively;73'118 changes in
these mixing parameters cannot be simulated
easily in smog chambers. Nevertheless, the
information presented here is useful in that it
suggests the areas and seasons in which
atmospheric mixing conditions are relatively more
conducive to oxidant formation and accumulation.
SUNLIGHT
The significance of sunlight is related to the
intensity of sunlight and its spectral distribution,
40
-------
105° 100° 95° 90° 85° 80° 75° 70° 65° 60°
50'
45'
35°
30'
25°
Figure 4-8. Monthly distribution of number of cases of 4 or
more days of atmospheric stagnation, 1936-75
(September).76
45
40°
35
30°
25°
20°
105° 100C
95"
90°
95°
80°
75
105 100° 95° 90° 85° 80° 75° 70° 65° 60°
I I
50
45
40'
35
30r
25" —
'Figure 4-9. Monthly distribution of number of cases of 4 or
more days of atmospheric stagnation, 1936-75 (August).76
I _ [ _ i _ | _ |
45°
40°
35'
30
25°
20°
105° 100°
95°
86°
80°
75°
41
-------
both of which have direct effects on the specific
chemical reaction steps that initiate and sustain
oxidant formation. Steps of varying significance
include the following (and possibly others also):
NO2 + hv - 0 + NO
MONO + hv/- OH + NO
H2CO + hv/ - H + HCO
RCHO + hv/- CH3 + HCO
03 + hv/- 0 ('D) + 02
H202 + hv - 20H
HN03 + hv/ - OH + N02
(4-19)
(4-20)
(4-21)
(4-22)
(4-23)
(4-24)
(4-25)
Whereas the effect of sunlight intensity is direct
and amply demonstrated,78 the effect of
wavelength distribution on the overall oxidant
formation process is a subtle one Experimental
studies have shown the photolysis of aldehydes to
be strongly dependent on radiation wavelength in
the near-UV region.28 Since aldehydes are major
products in the atmospheric photo-oxidation of
HC-NOx mixtures, it is inferred that the radiation
wavelength should have an effect on the overall
photo-oxidation process. This inference was
directly verified, at least for the propylene/NO, and
/i-butane/NOx chemical systems, in recent smog
chamber studies.70 In the ambient atmosphere,
some variation in the wavelength distribution of
sunlight does occur as a result of variations in
stratospheric ozone, ambient aerosol,'31 and cloud
cover. Obviously, such variation should be
recognized and properly considered in conducting
and interpreting outdoor experimentation. Beyond
this consequence, however, the wavelength
distribution factor has no other practical
significance.
Sunlight intensity vanes with season and
geographical latitude, as shown in Figure 4-10.78
The latitude effect is strong, but only during the
winter months. During summer, throughout the
contiguous United States, the maximum light
intensity is fairly constant, and only theduration of
the solar day varies to a small degree with latitude.
Some variations in light intensity also occur with
longitude during the summer months, with the
highest intensities occurring in the western United
States.
Absolute levels of sunlight intensities were
calculated by Leighton nearly two decades ago.78
Since then, several investigators have re-
calculated actimcfluxes and measured intensities
by more direct methods. Thus recent
measurements and recomputations by EPA
investigators gave results somewhat different
from those reported by Leighton. These differences
VALUES OF SOLAfi SPECTRUM
ARE FOP THE 100 /\ INTERVAL
CENTERED AT 3700 \
SUMMER SOLSTICE
I 20 N. LAT
II 34 N. LAT
lit 50 N. LAT.
WINTER SOLSTICE
" IV 20 N. LAT
V 36 N. LAT.
VI 50 N LAT
a
t-
o
I
Figure 4-10. Diurnal variations in actinic irradiance.78
are summarized in Table 4-6.)0° The recomputed
values are lower than Leighton's in the 295- to
395-nm wavelength interval, slightly higher in the
395- to 450-nm interval, and considerably higher
in the 450- to 700-nm interval. The same EPA
investigators expressed their recomputed light
intensities also in terms of values for the NO2
photodissociation constant, k, (min ').101 Resulting
values are listed in Table 4-7. These values are
higher than the earlier values used in laboratory
simulations of polluted atmospheres. Jackson et
al.,69 Zafonte,147 and Harvey et al.56 measured k,
directly; their values also were higher than those
based on Leighton's calculations or on data
obtained by standard radiometric techniques.
TABLE 4-6. PERCENTAGE DIFFERENCE BETWEEN THE
PETERSON100 AND THE LEIGHTON VALUES AT THE
EARTH'S SURFACE OVER SELECTED WAVELENGTH
INTERVALS AND SOLAR ZENITH ANGLES
nm
295-395
395-450
450-700
0
-6.1
+2.4
+ 18.8
Zenith angle (c>
20 40
-64 -55
+2 1 +3.4
+ 18.4 +17.0
60
-43
+5.9
+ 15.7
42
-------
TABLE 4-7. CALCULATED VALUES OF NO2 PHOTODISSOCIATION RATE CONSTANT3 AT THE EARTH'S
SURFACE AND PERCENTAGE INCREASE OF THE RATE CONSTANT FROM THE SURFACE TO VARIOUS HEIGHTS101
Zenith angle (°)
Height
Surface
0 15 km
036 km
064 km
098 km
1 84 km
2.91 km
0
0.579
60
11 4
16.6
21.4
29.9
364
10
0574
6.1
11 5
16 9
220
303
37 1
20
0.560
6.4
12 1
17 7
22.9
31 6
35.8
30
0535
69
13.1
19 1
24.7
340
41 7
40
0496
77
145
21 2
272
377
462
50
0432
89
16.9
244
31 5
434
537
60
0352
11 1
21 0
304
392
53.7
666
70
0231
143
277
407
52.3
732
91 3
78
0 114
167
342
52.6
702
101 0
132.0
86
0025
80
200
350
48.0
720
1000
aln units of mm"1
The light intensity factor has been studied in the
laboratory both with respect to its effect on
individual photolytic reaction steps and with
respect to its effect on the overall process of
oxidant formation.94 All of the early studies,
however, employed constant light intensity
conditions, in contrast to the diurnally varying
intensity in the ambient atmosphere. Only within
the last 3 years has the diurnal variation of light
intensity been recognized and studied as a factor.
Such studies have shown this factor to have a
varying (with initial reactant concentration
conditions), rather unpredictable, but somewhat
significant effect.72
TEMPERATURE AND RELATIVE HUMIDITY
Effects of temperature and relative humidity on
photochemical oxidant formation were suggested
by early laboratory studies discussed in the
predecessor of this criteria document.92 Additional
laboratory and field studies in more recent years
verified a significant, positive temperature
effect.1073Thus outdoor smog chamber studies of
synthetic HC-NOx mixtures showed that in winter,
with maximum daily temperatures at 50° to 60°F
(10.0° to 15.6°C), oxidant yields were lower than
those in the summer (80° to 90°F, or 26.7° to
32.2°C) by 70 to90percent, depending on reactant
composition.73 Also, an RTI analysis of field data
showed (1) good correlation between daily
maximum 1 -hr oxidant levels and daily maximum
temperature; and (2) no oxidant concentrations
greater than 0.08 ppm on days when temperatures
were below 62°F (16.7°C). The evidence10'73
suggests that below about 55° to 60°F (13° to
16°C), photochemical oxidant concentrations are
not likely to exceed the 0.08 ppm standard. Also,
information on the temperature factor indicates
that in the summer, temperature conditions are
conducive to ambient oxidant formation through-
out the United States as far north as Alaska.
Unlike the temperature factor, the effect of
relative humidity remains somewhat uncertain.
Reveiws by Quon and Wadden103 and by Altshuller
and Bufalini4 have pointed out inconsistencies in
the available evidence. Model simulations of
irradiated HC-NOX predict only a small humidity
effect;94 however, these predictions have not been
verified by consistent experimental evidence.
TRANSPORT PHENOMENA
Relative contributions of natural and an-
thropogenic sources in a given region or area are a
major question relevant to the ambient
oxidant/ozone problem. This question has not
been answered unequivocally and quantitatively
chiefly because anthropogenic pollutant transport
makes it difficult to assess the strength of the
natural sources. The question has been answered
qualitatively. The consensus is that oxidant/ozone
and/or oxidant precursor transport does occur and
contributes to oxidant/ozone buildup in areas far
from the sources. The magnitude of this con-
tribution and the magnitude of the contributions
from the stratosphere and from other natural
sources have been viewed as unresolved issues.43
Aside from its connection with the natural-
versus-anthropogenic-sources question, the
phenomenon of oxidant transport is important for
another reason. The phenomena of urban
oxidant/ozone plume formation and movement,
rural oxidant/ozone occurrence (at problem
levels), Sunday-weekday effect, and nighttime
oxidant/ozone occurrence were previously either
unnoticed or thought to be odd. They are now
believed to be true manifestations of an extremely
complex emission/pollutant dispersion process.
Such complexities may be introduced by horizontal
and/or vertical transport of oxidant/ozone and/or
of precursor mixtures for long distances without
excessive dilution.
43
-------
Consequences of pollutant transport are
discussed here. Several studies concerned with
oxidant transport have been conducted in the last 5
years. Such studies have resulted in a wealth of
information. This information and its implications
are presented and discussed in the remainder of
this section.
The first item of interest is evidence on the
occurrence and range of oxidant transport. It
should be clarified that the term "oxidant
transport" is commonly used to refer to transport
of either oxidant/ozone or oxidant precursors. The
distinction between oxidant/ozone and oxidant
precursors should be kept in mind, as it is essential
to the understanding of the nature, mechanisms,
and implications of the oxidant pollution problem.
Oxidant transport has been recognized to occur
on three scales in terms of geographical distance
or area subjected to the transport effects. These
scales are.
1. The urban-scale transport, as a result of
which the peak oxidant concentrations
develop in the suburbs some miles
downwind from the city core area where
the oxidant/ozone and their precursors
originated;
2. The mesoscaletransportthatencompasses
land- and sea-breeze circulation, and the
formation of urban oxidant plumes that
create oxidant problems asfaras 100 miles
or more downwind from the source city;
and
3. The synoptic-scale transport, a much
longer and broader range of transport
associated with high pressure systems
Urban-scale transport has been observed in Los
Angeles,134 New York, Houston, Phoenix, and
several other urban centers.327 All of the evidence
available points to the conclusion that conditions
at the center of source-intensive areas are not the
most conducive conditions for oxidant accumu-
lation, mainly because of the strong scavenging
effect of oxidant precursors, especially nitric oxide.
At higher elevations or at horizontal distances
downwind from the sources, where the precursor
scavenging effect is less important, oxidant
concentrations are, in general, greater, and their
levels are determined primarily by the intensity of
the photochemical activityandbyambientdilution.
Measurements in several cities during days with
various wind speeds showed the peak oxidant
concentrations to occur at distances 8 to 1 36 km (5
to 85 miles) downwind from the city center.87
During days of stagnation when the highest levels
of oxidant/ozone occur, this peak-concentration
distance was estimated by EPA to be between 1 5
and 25 km (9 and 1 6 miles),53 and more recently, to
be 15 to 30 km (9 to 19 miles).
This urban-scale transport of oxidant has two
important implications. First, it suggests an
obvious guideline to be used in siting stations for
oxidant/ozone monitoring. The second im-
plication, suggested by experimental data on the
oxidant-to-precursor dependencies,42 is that
urban-scale transport may obscure the real impact
of emission control on oxidant air quality. This
latter implication may be clarified as follows: small
or even moderate decreases in emission rates for
HC and NOX are expected to affect mainly the time
at which the oxidant concentration will reach the
day's peak levels; the effects on the peak level itself
are of smaller magnitude. This time lag effect, in
turn, translates into an increase in the distance
from the source area where peak oxidant
concentrations develop (i.e., the location of the
peak concentration relative to the location of the
core of the source area). Thus, for example, the
actual effect of a modest HC emission reduction
will be to shift the peak oxidant concentration
farther downwind, rather than to effect a
proportional reduction locally. Therefore, the
impact of emission control will appear to be either
beneficial or detrimental, depending on the
location of the oxidant monitoring station.
Evidence on occurrence of mesoscale and
synoptic-scale transport has been obtained in
numerous extensive and conclusive studies. These
studies have been reported individually 10~
1 2,25,26,28,30,38,39,44,51,61,66,68,85,86,11 5,11 6,129,130,1 36,138,1 39,-
145 and collectively in reviews.85'9712° The evidence
shows that many cities produce urban oxidant
plumesthat cause elevated oxidant concentrations
in downwind areas as far as 300 km (190 miles) or
more from the source City.^5,6i,74,i28,i36,i43
Transport associated with synoptic-scale high
pressure systems, unlike such mesoscale
transport, is not characterized by well defined
urban oxidant plumes and can extend to distances
several times greater than 300 km (190 miles).
38,39,97,115,116,120,136,145 |p ^^ ^^ ^ sigmf)cam
implication is that urban emissions may be
creating oxidant problems not only within and near
their source areas, but also in other downwind
urban areas, as well as oxidant problems in rural
and even in remote areas. It should be stressed
that the evidence on horizontal transport of
oxidant/ozone suggests but does not prove that
most of the oxidant observed in rural and remote
44
-------
areas is transported from the cities. Also,
horizontal transport of oxidant over long distances
can only occur in air layers aloft; ozone at ground
level is rapidly destroyed on surfaces and in
reactions with NO and NC. *>.w».™."w» For
oxidant/ozone to reach the surface, there must be
vertical air movement, a movement that may also
bring to the surface ozone from the stratosphere or
from other natural sources. In general, because of
the multiplicity of the sources that potentially
contribute to surface oxidant/ozone buildup, the
rural oxidant/ozone problem cannot automatically
be attributed entirely to urban oxidant/ozone
transport. The indications are that direct urban
transport is a relatively important source of
oxidants in many locales but not necessarily
everywhere.
Because of their elevated ambient ox-
idant/ozone levels, the State of California, the Gulf
Coast and Texas, the upper Midwest, and the
Northeast are the areas of the United States that
have been studied most. The explanation for
oxidant/ozone levels in California is relatively
simple: Ozone levels in that area can be explained
by local photochemistry and oxidant/ozone
transport; nonanthropogenic explanations are not
required. The most striking and clearly explained
oxidant transport case reported for that area is the
one involving overnight transport over water of Los
Angeles air containing as much as 590 /jg/m3 (0.3
ppm) oxidant to San Diego.12'120
Unlike California, the situation in the Gulf
Coast/ Texas area is considerably more complex
and requires more careful examination. Early
studies in Texas led investigators to conclude that
widespread ozone concentrations were not of
anthropogenic origin. The basis of that early
conclusion was the observation that ozone was
frequently high when there were onshore winds at
coastal stations.120 However, recent studies by
Price'02 and by Decker et al.38 at the Research
Triangle Institute (RTI) concluded that natural
tropospheric ozone was not a factor in the high
ozone events in Texas. The RTI study, in particular,
involved extensive aerial and ground meas-
urements of ozone; HC, and NOX, and wind
trajectory analysis. Results from that study
showed the high ozone concentrations over the
Gulf to be associated with recent passage of the air
stream over land areas and often over source-
intensive areas. This finding, similar to that by Bell
for the Los Angeles/San Diego case,12 explains
more rationally the observations made in the early
studies of Texas and implies that the Texas ozone
problem is at least partly of anthropogenic origin.
The RTI study also indicated an association of high
ozone concentration with the passage of high
pressure systems, an association which had been
recognized for years and which was more firmly
established in the midwestern part of the United
States.
Transport of ozone for an unusually long dis-
tance was reported in an episode of widespread
haze over southern Florida from May 21 to 27,
1972.'23 The initial report concerned metropolitan
Dade County and was supplemented by concur-
rent data obtained at the site of a proposed jetport
65 km (41 miles) west of Miami. At this rural site,
the ozone concentration equaled or exceeded 160
fjg/m3 (0.08 ppm) on 5 of those days, and the air
quality standard was exceeded for 14 consecutive
hr on both May 22 and 23. The analysis of the
transporting trajectory that brought the haze and
accompanying ozone was verified by the EPA
Meteorological Laboratory. The source region
appeared to be the industrial area among the
states south of the Great Lakes, making a transport
distance of over 1000 miles (1600 km), of which
400 miles (640 km) were over the Gulf of Mexico.
Both the haze and the ozone of its precursors may
have been augmented along the route.46
Most of the oxidant transport studies in the
upper Midwest dealt with the role of high pressure
meteorology in wide area oxidant problems. The
connection between stagnating anticyclonic
conditions and high oxidant concentrations over
large areas was studied by several investigators
and is now believed to be relatively well
understood.3ei39-115'116'1zo'1M'130-136'138-139'l45This
belief is based on (1) data from direct monitoring
with an aircraft of the ozone concentrations in the
path of a moving high pressure system, and (2)
presence of statistically supported correlation
between high pressure systems and elevated
ozone concentrations during days with such-
conditions. This correlation has been verified by
numerous studies. The reverse has not been
established; that is, elevated ozone levels do not
occur only during stagnating anticyclonic epi-
sodes.
Within the moving and clockwise rotating
anticyclone, the highest ozone concentrations are
observed in the trailing or western portion of the
cell where the air parcels have the longest
residence time in the anticyclonic regime. RTI
investigators made some detailed model
calculations to show that air parcels initially in the
northeast quadrant of the cell had a potential
45
-------
residence time as long as 6 days, and that air
parcels on the western side of the circulation
pattern had been in the cell the longest time. This
latter conclusion is in general agreement with
ozone measurements in source areas such as have
been described by Westberg et al.138 Under
conditions of the high intensity sunlight and
stagnation or undiluted flow that typically prevail
within the anticyclone, oxidant/ozone formation is
expected to be accelerated and depends on the
emission density within the area covered by the
high pressure system. This latter dependence
seems to be manifested by the data obtained in the
paths of high pressure systems moving from less to
more densely populated areas.38
A significant percentage of the high oxi-
dant/ozone episodes observed in the United
States is associated with passage of high pressure
systems. Statistics of high pressure episode
occurrence have been obtained by Korshover76 for
the eastern part of the United States for all months
of the year. The study could not be extended in the
western United States because the extreme
terrain irregularities result in unrepresented
pressure distributions. Such statistics for the
months'of August and September are shown by
the plots in Figures 4-8 and 4-9.
Oxidant transport in the Northeast received
considerable attention because high oxi-
dant/ozone concentrations have been measured
almost routinely in rural and distant suburban
areas for years. Careful analysis by Bell
Laboratories and other investigators of aerometric
data obtained in northern New Jersey, eastern
New York, Connecticut, and Massachusetts
provided strong evidence of extensive mesoscale
transport, especially from the New York City/New
Jersey area to the Northeast into Connecticut and
Massachusetts.120'146 Such transport was char-
acterized (1) by low oxidant/ozone levels within
urban core areas because of scavenging by
precursors, (2) by appearance of highest
oxidant/ozone concentrations downwind from the
main urban center at a distance equivalent to 1 to 2
hr of midday wind travel, and (3) by shifting of peak
oxidant/ozone concentrations with time and
distance, refleri.ng the regional movement of the
main urban area plume along the wind trajectory.
The results from Bell Laboratories seem to be in
agreement with subsequent studies by other
groups. Interestingly, the evidence descriptive of
the Northeast situation suggests that weather
patterns seem to play a minor role in establishing
oxidant/ozone levels in that part of the United
States. For example, oxidant/ozone con-
centrations seem to be less sensitive to the
location or intensity of an anticyclonic system.120
This could be a result of the high population density
in that area and the relative dominance of urban
plumes over the nonanthropogenic background
levels in the air mass. Finally, local circulations in
the surface layers were found to play a dominant
role in the coastal areas in the Northeast.120'139
Such circulation systems, subjecting areas to next-
day effects from their own emissions, were also
observed in the Gulf Coast38 and Midwest86 areas
but were not as dominant.
The evidence on transport has had another
important impact on the understanding of the
oxidant/ozone problem. It provided one plausible
explanation to some phenomena previously
thought to be paradoxical (e.g., the occurrence of
elevated oxidant/ozone concentrations in rural
areas and at night115 and the Sunday-weekday
effect13'24'45'80'135). Thus downward intrusion, from
layers aloft, of transported anthropogenic
oxidant/ozone explains convincingly the occur-
rence of elevated oxidant/ozone concentrations in
rural areas and during the night in the summer-fall
months. The Sunday-weekday effect could also be
explained in part as resulting from anthropogenic
oxidant/ozone present in layers aloft. Although
these explanations have not been supported by
documentation in all cases, they at least have
served to clarify these phenomena and to increase
confidence in the current knowledge and
understanding of photochemical processes.
Identification and understanding of ox-
idant/ozone transport represents a significant
advance in understanding photochemical pol-
lution. However, it has also raised several
questions that need to be addressed. For instance,
the impact of pollutant transport on the oxidant
concentration observed in a locality cannot be
quantified at this time.43 The difficulty here is not
merely the absence of sufficient data (e.g., data on
pollutant composition in transported, aged air
masses), but also the absence of valid investigative
methods. Clearly the need is for both laboratory
and field studies that would provide a better
understanding of the quantitative aspects of the
interaction of transported, aged pollutant mixtures
with locally emitted fresh pollutants. More
specifically, a method should be developed for
estimating the proportion of local oxidant/ozone
concentrations resulting from pollutant trans-
port.43 Evidence must also be obtained on the
46
-------
oxidant-to-precursor dependencies for trans-
ported oxidant/ozone.43
NATURAL SOURCES OF OXIDANTS
Introduction
The occurrence of ozone in remote areas and
during the winter months constitutes strong
evidence that natural sources of ozone must exist,
providing an atmospheric background on which
the anthropogenically generated ozone is
superimposed. Promulgation of the 160-A
-------
It is well established that air interchange
between stratosphere and troposphere occurs via
the following four mechanisms:111 (1) mean
meridional circulation, (2) large-scale eddy
transports (jet streams), (3) seasonal adjustment of
tropopause level, and (4) mesoscale and small-
scale eddy transport. Each of these mechanisms is
characterized by short-term, seasonal, and long-
term fluctuations, and their impact in terms of
surface concentration of stratospheric ozone
varies with geographical latitude.
The mean meridional circulation (MMC)
mechanism involves upward flux of tropospheric
air mainly in the low latitudes and downward flux
of stratospheric air in the middle latitudes. It varies
somewhat in intensity from season to season, the
most intensive period being in winter (December-
February). Reiter111 estimates the annual amount
of stratospheric-MMC air mass transferred into
the troposphere to be about 43 percent of the total
mass of one stratospheric hemisphere, or about
1.84 x 1020 g of air. Reiter further estimates that
the upper limit of average ozone concentration
shortly above the tropopause is0.05/yg/g. These
estimates of mass exchange and the mean ozone
concentration result in a calculated average ozone
exchange of 9.22 x 1013g of ozone due to the MMC
process. Furthermore, assuming that the MMC
ozone flux exchange is concentrated over half the
area of one hemisphere, Reiter calculates a yearly
mean ozone flux of 0.23 x 1CT7 g/mz-sec. During
the winter period, however, when the strongest
cross-tropopause fluxes occur, this ozone flux
could be 0.4 x 10~T g/m2-sec. From this
stratospheric mass flux (0.4 x 10~7 g/m2- sec) and
estimates of the mean tropospheric vertical
velocity, Reiter calculated the mean background of
stratospheric-MMC ozone in the springtime and in
themidlatitudestobe14,6/yg/m3(0.007ppm).The
yearly mean from Reiter's flux estimate would be
about 8, fjQ/m3 (0.004 ppm). Junge, as cited in
Reiter,111 using a somewhat higher flux value and
a lower vertical velocity value, estimated such a
mean background ozone concentration to be 35,6
/yg/m3 (0.017 ppm).111
Superimposed on the background concentration
caused by the MMC process, there is an ozone
increment of stratospheric origin caused by large-
scale eddy mixing occurring in the jet stream
regions through a tropopause folding (TF) process.
The intensity of this process relative to the MMC
process Is somewhat uncertain; however, its
impact per unit of ozone mass transported on the
lower troposphere is unquestionably stronger and
more localized. This is because relative to the MMC
mechanism, stratospheric air during TF intrusion
moves downward much faster and therefore is
subject to much less tropospheric dispersion.
Stratospheric ozone intrusions of the TF type occur
mostly along the polar-front jet stream and yield,
on the average, maximum stratospheric ozone
accumulations at ground level in midlatitudes
during March and April.
Using the studies of Danielsen, Mohnen90 has
examined the intensity of the TF process, as has
Reiter,111 using data from Mahlman. From case
studies of TF events, Danielsen and Reiter
estimated that the stratospheric air mass
transported per TF event was 4 to 6 x 1017 g. From
the average number of TF events observed during
1963-64, Reiter calculated the annual air mass
transport to be equivalent to approximately 20
percent of the air mass in the northern
stratospheric hemisphere. In contrast, Danielsen's
estimate was 90 percent.
To estimate the impact of the TF-type
stratosphere-troposphere interchange in terms of
surface ozone concentration, it is necessary to
know (a) the ozone concentration in the
transported stratospheric air and (b) the
(stratospheric) ozone decay process occurring
within the troposphere. Indirect estimates of these
have been made based on radioactive 90Sr tracer
methods and on isentropic trajectory analysis
techniques. From measurements of the 90Sr-to-
ozone ratio within the lower stratosphere and from
90Sr-related radioactivity data at ground level,
Reiter re-estimated recently the 24-hr average
surface concentration of stratospheric-TF ozone to
be as high as 60 /yg/m3 (0,033 ppm) in 1963 and
120 /yg/m3 (0.066 ppm) in 1964.113 Reiter's
estimate must be considered an upper limit, since
it assumes that the 90Sr-to-ozone ratio is
conserved during transport from the lower
stratosphere to ground level. From use of
isentropic trajectory analysis techniques, it was
deduced that exceptionally strong TF-type
intrusions occur only occasionally, but can cause
surface ozone concentrations of 160/yg/m3 (0.08
ppm) or higher. Such strong intrusions are
expected to occur about once a year, usually in the
southern and eastern United States. Reiter
estimated the probability of such occurrence to be
0.2 percent, measured in days of observations on
an annual basis at a given location.111
Direct evidence of the impact of stratospheric
ozone on ground-level oxidant during TF events
has been reported by Lamb,77 He oerformfid a
48
-------
detailed study of an incidentthat occurred in Santa
Rosa, Calif., in which hourly averaged ground-level
oxidant concentrations exceeded 1 60 jug/m3 (0.08
ppm) for 5 consecutive hours just before dawn on
November 19, 1972. The largest of the five hourly
averaged concentrations was 450 jug/m3 (0.25
ppm). All available evidence indicated that the
ozone responsible for this episode originated in the
stratosphere during a TF event off the south coast
of Alaska 2 days earlier. Chatfield and Harrison22
found a positive correlation between increases in
ozone concentration in the remote Olympic
Mountains of Washington and the passage of low
pressure systems originating in the same region
that the TF event responsible for the Santa Rosa
episode occurred. Attmannspacher and Hart-
mannsgruber9 have also reported ozone on an
anomalous 1000-m (3300-ft) mountain peak in
northern Germany attributable to stratospheric
intrusion. On three different occasions during the
winter of 1971, they observed fluctuations in
ozone concentrations between 490 jug/m3 (0.025
ppm) and 980 jug/m3 (0.50 ppm) that lasted longer
than 10 min. Each occurrence was during strong
snow showers associated with passing cold fronts.
(The Santa Rosa episode was initiated by a brief
rain shower shortly after the passage of a cold
front.) On three occasions, balloons were launched
to measure ozon 3 concentrations aloft, and in all
three cases, a strong, secondary ozone
concentration maximum was observed just above
the tropopause.
Besides the .VIMC and TF processes, strat-
osphere-troposphere interchange occurs ajso the
seasonal tropop luse adjustment (STA) and the
small-scale eddy transport(SSET)mechanisms. Of
these, the STA mechanism has been estimated to
cause a stratospheric air mass flux equivalent to
10 percent of the entire stratospheric hemisphere.
90'111 Such interchange occurs mainly in the same
season and latitudes as the TF process. Therefore,
its impact is included in the estimates made for the
TF process. The SSET process is of much less
importance and contributes to the tropospheric
ozone problem only at the noise level. 90'111
In conclusion, based on Reiter's estimates of
stratosphere-troposphere interchange and on 90Sr
data, the annual average total of stratospheric
ozone accumulation expected at groundlevel
amounts to 20 to 30 /yg/m3 (0.01 to 0.015 ppm).
84'113 If one considers the fact that some ozone can
be destroyed during transport to ground level, the
preferred mean ozone concentration calculated by
Reiter would be closer to the lower limit of 20
(0.08 ppm). Occasional excursions to about
160 Afg/m3 (0.080 ppm) can be expected, however.
Danielsen's estimates of stratospheric ozone flux
are a factor of two to three higher than Reiter's
estimates and would lead to correspondingly
higher yearly mean ozone concentrations.
Danielsen estimates an ozone flux rate.of 8 x 1010
molecules/cm2- sec. Singh et al.127 suggests that
this flux rate would be compatible with ground
level ozone measurements at remote locations
(yearly mean of 60 yug/m3, or 0.030 ppm) if an
ozone lifetime of about 4 months is assumed in the
troposphere. The lifetime of ozone is uncertain,
and all estimates of ground level ozone
determined from stratospheric ozone intrusions
suffer from this uncertainty. Both of Reiter's and
Danielsen's estimates apply to situations where
stratospheric ozone is brought to ground level
without the assistance of precipitation-driven
downdrafts. The case studies cited above by
Lamb77 and Attmannspacher and Hartmanns-
gruber9 (and others) indicate that stratospheric
ozone accompanying TF events can reach ground
level in concentrations exceeding 390/yg/m3 (0.20
ppm) when it is transported part of the way in
downdrafts caused by precipitation. The frequency
of this type of event is not yet known, but the mean
ozone contribution resulting from such events is
expected to be minimal.
Mohnen's estimate of the annual average
stratospheric ozone at ground level is 40 to 70
/yg/m3 (0.022 to 0.035 ppm).90 Such background
levels vary with season and latitude because the
stratospheric ozone reservoir also varies with
season and because the various stratosphere-
troposphere interchange mechanisms vary in
intensity with season and latitude. Considering all
these variables, there seems to be a consensus of
opinion that at locations in the midlatitudes (e.g.,
the United States), the ground level ozone
concentrations of stratospheric origin peak in
winter and spring.
The preceding paragraphs dealt mostly with the
evidence and conclusions obtained from analysis
of global circulation patterns. As stated earlier in
this section, relevant evidence was obtained also
from analysis of tropospheric ozone concentration
data in remote areas. Such evidence, although
seemingly more valid as being more direct, should
be examined carefully, for it is often distorted or
obscured by hidden interfering factors. Examples
of such factors are destruction of ozone on
surfaces or in reaction with HC and NO, and
photochemical formation of ozone.106
49
-------
Mohnen90 examined the evidence both on
stratosphere-troposphere exchange and surface
ozone concentration and concentration variation,
but, unlike Reiter,111 he elected to base his
conclusions on stratospheric ozone intrusion
mainly on tropospheric ozone data. Utilizing the
analytical method introduced by Junge90 and data
from monitoring stations in remote areas
presumably free of local or regional anthropogenic
emissions, Mohnen concluded that a range of
annual mean concentrations equal to 40 to 70
fjg/m3 (0.022 to 0.035 ppm) constitutes a
representative tropospheric ozone level (of
stratospheric origin) for 35° to 50° N latitudes.
Mohnen also concluded that TF events may lead to
ozone concentrations at ground level as high as 30
£»g/m3(0,15 ppm) and lasting from 2 or 3 hr up to 1
to 3 days.
Singh et al.1Z8 examined aerometric data on
ozone, HC, and NO, obtained at 10 stations
selected to be as remote as possible and with air
quality records of at least 2 years' duration. Seven
of these stations were in the western continental
United States, one in Hawaii, one on the summit of
Whiteface Mountain in New York, and one on the
Zugspitze in West Germany at a 3000-rn (9800-ft)
elevation. The ozone variation patterns observed at
most of these stations are exemplified by the
patterns observed at the Quillayute, Washington,
andMaunaLoa, Hawaii, stations (Figures4-11 and
4-12), The striking features of these patterns are
the maximum ozone concentrations observed in
April and the 160-^rg/m3 (0.08-ppm) or higher
ozone concentrations of such maxima.
Considering the low sunlight intensity and
temperature conditions prevailing in April at
Quiliayute, it can be deduced that these high ozone
concentrations must be primarily of natural,
stratospheric origin. Such a deduction is further
supported by the lack of the familiar diurnal ozone
variation pattern that characterizes ozone
formation from local photochemistry. Signi-
ficantly, this stratospheric ozone declines as
summer months approach, indicating that the
effects of stratospheric intrusion are at a low ebb
during the smog season. Singh et al.126 did observe
elevated ozone concentrations during the summer
months in some stations. Such maxima, however,
were explained in terms of photochemical ozone,
either locally produced or transported to the
stations from upwind areas.
0,120
I
a
o
N
o
0.100 —
0,080
0.060
0.040
0.020
I
I
STATION NO. 1
(QUILLAYUTE)
] T I j I I
— —HIGHEST 1-hr MONTHLY 03 MAX
"MONTHLY AVERAGE OF DAILY 1-hr O3 MAX
—— MONTHLY O3 AVERAGE OF ALL 1-hr Oa VALUES
—— LOWEST 1-hr MONTHLY O3 MAX
DEC.
<1973>
APRIL
AUG.
DEC.
11974)
APRIL
(1975)
MONTHS
Figure 4-11. Long-term ozone variations at Quillayute,1
50
-------
E
a.
a
O
N
O
0.120
0.100 —
0.080
0.060
0.040
0.020
STATION NO, 9
(MAUNA LOA, HAWAII)
HIGHEST 1-hr MONTHLY O3 MAX
MONTHLY AVERAGE OF DAILY 1-hr O3 MAX
MONTHLY O3 AVERAGE OF ALL 1-hr O3 VALUES
LOWEST 1-hr MONTHLY O3 MAX
DEC.
(1973)
MAY
OCT.
MAR.
AUG.
Figure 4-12. Monthly ozone variations at Mauna Loa, Hawaii.1
JAN.
(1976)
Data collected by Singh et al.
126,127
suggest
yearly mean ozone levels of about 60//g/m3(0.03
ppm) between latitudes 18° N and 48° N, The
springtime mean values are found to be about 80
fjQ/m3 (0.04 ppm), and the fall mean is about 40
fjg/m3 (0.02 ppm). The yearly 1 -hr maximum ozone
concentrations also show large year-to-year
variation. For example, at Mauna Loa (Figure 4-
12), the 1 -hr ozone maximum of 160 fig/m3 (0.08
ppm) was exceeded on 1.6 percent of the days in
1975 and on zero percent of the days in 1974.
The Whiteface Mountain data were examined
also by Coffey et al.28 along with data from other
remote areas and from urban locations in New
York State. The observations made by Coffey et
al.2B from these data are: (1) Ozone concentrations
at Whiteface Mountain often exceed the 160-
fjg/m3 (0.08-ppm) level; (2) a background ozone
blanket covers the entire state; and (3) there is
monthly variation of ozone with a strongly defined
maximum in August (Figure 4-1 3). Presence of the
ozone blanket suggests that the ozone source lies
aloft and cannot be due to local photochemistry.
However, occurrence of the peak concentrations in
the summer (August), coupled with evidence on
presence of manmade pollutants (halocarbons)106
can only be interpreted to mean that the Whiteface
Mountain data did reflect anthropogenic in-
fluences, a conclusion pointed out also by Reiter111
and by Singh et aI.126Thustransport, in paths aloft,
of ozone and/or ozone precursors from upwind
anthropogenic sources provides a more plausible
explanation of the Whiteface Mountain data
patterns, at least for 1973.
More recently, Husain et al.67 obtained and
analyzed ozone and 7Be data in Whiteface
Mountain during the smog season. In their
analysis, these investigators used data on 7Be and
air trajectory analysis to establish stratospheric
origin, and NO, and aerosol concentrations and Os-
7Be correlation patterns to verify absence of
photochemical ozone. Interpretation of results
suggested an upper 24-hr concentration limit of 75
fjg/m3 (0.037 ppm) stratospheric ozone at
Whiteface Mountain during July 1975.
Measurements of ozone on the Zugspitze, West
Germany, were analyzed both by Reiter111 and by
Singh et al.128 Reiter's frequency distribution
analysis of data obtained during August 1973 to
October 1 975 indicates that (a) 0.2 percent of the
hourly concentrations exceeded the 160-^g/m3
(0.08-ppm) level, and (b) relatively high
concentrations occurred during summer rather
than spring. The latter indication again suggests
an anthropogenic-influences explanation. How-
ever, in this case, such an explanation alone is not
51
-------
0.070
0.060 -
0.050 -
0.040 -
g 0.030 -
0.020 -
0.010 -
JFMAMJ JASONDJFMAM
-1973
1974
Figure 4-13. Average monthly ozone concentrations record-
ed at summit of Mount Whiteface.128
convincing because the Zugspitze station, at the
altitude of 3000 m (9800 ft), is above the planetary
boundary layer where anthropogenic pollutant
transport occurs. Reiter suggests that local
circulation along the mountain slopes may be
moving anthropogenicafly contaminated air from
lower altitudes to the higher altitude of the station
site.
Finally, review of 1477 ozonesonde obser-
vations between December 1962 and December
1965 showed that 2 percent of the ozone
concentration values exceeded 150/yg/m3 (0.075
ppm).112 However, of the 31 cases with such
elevated concentrations, only three qualified as
being unaffected by tropospheric source
interferences. Thus only about 0.2 percent of the
sample exceeded the 150-/yg/m3 (0.075-ppm)
level.
The tropospheric ozone data appear to agree
with the theoretical deductions both on the
intensity of the stratospheric ozone impact and on
the seasonal variation of such impact. Both bases
of evidence suffer from uncertainties, and
additional research studies are needed for a fuller
and more confident assessment of such impact.
Specific research needs include:
1. More definitive information on the
frequency of occurrence and intensity of
cyclogenetic events resulting in strat-
ospheric ozone intrusion.
2. More definitive information on the ozone
decay (and formation, if any) attending the
subsidence of stratospheric ozone within
the lower troposphere, especially within
the planetary boundary layer.
3. Development of techniques for forecasting
stratospheric ozone intrusions.
4. Further analysis of tropospheric ozone
measurements to delineate the strato-
spheric and anthropogenic contributions to
such ozone. The need here includes also
the measurement of trace contaminants
characteristic of anthropogenic sources.
The information and analysis presented in the
preceding pages, although they do not answer all
of the questions on the contribution of
stratospheric ozone to ground level ozone
concentrations, do provide important clarification.
For example, it is now clear that violations of the
oxidant/ozone NAAQS observed outside the smog
season, particularly during the spring months,
should not be attributed exclusively to
anthropogenic causes. However, it is equally
important that significant stratospheric ozone
concentrations at ground level, though they may
occur at any time, are less frequent during the
summer months (that is, during the smog season).
Within the summer months, such ozone is
expected to occur at a background level somewhat
lower than that of the annual average. Since the
annual average ozone concentration has been
estimated to be about 20 /yg/m3 (0.01 ppm) by
Reiter113 and about 60 /yg/m3 (0.03 ppm) by
Mohnen90 and Singh et al.,126 it follows that the
summer average could be 14 to 40/yg/m3 (0.007 to
0.020 ppm) because of the spring-summer
gradient. Data presented by Singh et al.126 indicate
that the 1-hr ozone concentrations in any given
season in a remote atmosphere are typically about
twice the ozone mean values. Thus an hourly
ozone concentration range of 30 to 80/yg/m3 (0.15
to 0.04 ppm) resulting from the stratospheric
source alone can be expected in the summer
season. The lower limit of 30 /yg/m3 (0.01 5 ppm)
would thus be suggested by Reiter's analysis,111
whereas Mohnen90 and Singh et al.126 would
support the upper limit of 80 /yg/m3 (0.040 ppm).
Photochemistry of Natural Organics/NO*
As in the case of stratospheric ozone, the role
and importance of naturally emitted organics and
NOx in the formation of oxidant/ozone hasceenan
issue.43 The issue was originally raised as <" result
of early reports that the rates of organic emissions
from vegetation are considerably higher, on a
global basis, than those from manmade sources.107
52
-------
Although natural emissions and anthropogenic
emissions are, for the most part, geographically
segregated, it is nevertheless reasonable to
suspect that the much more abundant natural
organics, emitted either within an urban or
nonurban area, or brought in through transport,
could contribute to ambient oxidant/ozone
concentrations to an important degree. These
suspicions became considerably stronger as a
result of two recent findings: (a) The occurrence of
a pervasive rural oxidant/ozone problem,118 and (b)
the high reactivity of the terpenes.48'137 In either
case, the finding could be interpreted to mean that
natural organics may constitute a significant
source of oxidant/ozone. As in the case of
stratospheric ozone, the question regarding the
role and importance of natural emissions as an
oxidant/ozone source needs to be answered.
From a cursory examination of the current
evidence, it becomes immediately apparent that
certain questions regarding natural emissions
have obvious answers or have been resolved by
scientific evidence, whereas other questions
appear unresolved. For example, it is certain that
vegetation emits organic vapors and that some of
these vapors (terpenes) play the dual role of ozone
precursor and ozone scavenger. Questions that do
not have obvious answers are; (1) What is the net
effect on ozone concentrations of the atmospheric
reactions of terpenes and of other natural
organics? (2) What reactive organics, other than
terpenes, are emitted by natural sources and at
what rates?
Evidence pertinent to these questions is
available and is fairly complete and definitive in
some respects, but it is incomplete and ambiguous
in others. The main source of ambiguity is that
most researchers are, in general, relatively
unfamiliar with natural organic emissions. Thus,
unless convincing documentation is provided,
results and conclusions on chemical identity,
ambient concentrations, and emission and sink
processes of natural organics are viewed with
considerable skepticism. In the remainder of this
section, a brief review is given of the relevant
evide'nce available, and an attempt is made to
interpret such evidence so as to provide the best
possible judgment regarding the importance of
natural emissions as an oxidant/ozone source.
There are two questions regarding the net
impact of terpenes on ambient oxidant/ozone
concentrations. First, what are the terpene
concentrations actually observed in atmospheres
with high oxidant/ozone levels? And second, what
is the net effect of terpenes at such concentrations:
Is it to produce or to destroy oxidant/ozone? The
evidence available appears sufficiently complete
and definitive137 to resolve these questions.
Most of the qualitative analysis on organic
emissions from vegetation has been done by
Rasmussen and his coworkers. Chromatographic
analysis by Rasmussen and Holdren108 of ambient
organics collected at remote sites showed the
presence of 10 to 60 organic components in the C5-
C10 range; individual component concentrations
were usually below 2 /yg/m3 (0.001 ppm).
Hydrocarbon concentrations in rural areas have
also been reported formally by Whitby et al.141 and
by Whitehead and Severs,144 and informally by
others.89'117 However, individual organics were not
identified in these studies. Because of this
deficiency, and in view of the well documented
occurrence of extensive urban pollutant transport,
a natural origin for the organic compounds
measured by these investigators cannot be
supported. Nevertheless, their evidence leaves the
question somewhat open. Some better
documentation on natural organic compounds in
ambient air was obtained by Westberg and
Holdren.137'140 Using a gas chromatograph linked
to a mass spectrometer, these investigators
identified in a forested area in Idaho, a-pinene, /J-
pinene, A-carene, and limonene present at
concentrations from a few ppt for limonene to 730
ppt for /3-pinene. Westberg and Holdren also
measured rates of emission of these terpenes from
the forest and found such rates to be consistent
with their measurements of the ambient
concentrations. The most significant finding from
this study, however, was that air samples taken
outside the forest canopy showed no measurable
terpenes, suggesting an extremely short life for
those organic compounds. Measurements by
Lonneman et al.82'83 in a forested area near
Durham, North Carolina, showed the presence of a
a-pinene, /3-pinene, myrcene, and A-carene at
concentrations from 0.3 ppb C (myrcene) to 90 ppb
C (/3-pinene)—roughly comparable to those
reported by Westberg and Holdren. Other
measurements in this same area showed the
terpene concentrations to be at lower levels when
the ambient ozone concentration was at higher
levels. Finally, in addition to the monoterpenes
already mentioned, isoprene was also found to be
emitted by vegetation and to occur in ambient
air.104
In conclusion, there is sufficient evidence to
show that monoterpenes do occur in the ambient
53
-------
air but only within forested areas; in such areas,
total terpene concentrations average 10 to 50 ppb
C.137 The evidence further shows that terpenes
have extremely short lifetimes, such that, with the
exception of isoprene, they are not transported
downwind from their sources in appreciable
quantities. Within urban areas, terpenes have not
been detected. In their numerous efforts to search
for and identify terpenes in urban atmospheres,
EPA investigators only occasionally found
isoprene, at ppb C concentrations. Thus the
evidence does not support an important direct
contribution of terpenes to urban oxidant/ozone
concentrations; terpenes may contribute tc the
urban problem indirectly, however, through
terpene-induced oxidant formation in rural areas.
This possibility will be examined next.
Evidence of the potential of terpenes to form
oxidant/ozone in rural areas has been observed
both in laboratory and field studies. Smog chamber
testing of terpene-NO* mixtures showed clearly
that terpenes are potent oxidant precursors and,
more important, that relative to the less reactive
hydrocarbons, their oxidant-forming potential is at
a maximum at low hydrocarbon-to-NO» ratios,
typically at 2:1 to 2.5:1 molar concentrations. 4fM4°
The latter finding is significant because it means
that in rural atmospheres with hydrocarbon -to-
NOx ratios typically much higher than 2.5:1, the
terpenes could be effective as oxidant producers
only partially, if at all. In fact, in the absence of
sufficient NO*, terpenes would act as ozone
scavengers.52 Under optimum hydrocarbon-to-
NO» ratio conditions, terpenes produced in smog
chambers a maximum of 470/yg/m3 (0.24 ppm) of
oxidant/ozone for 5 ppm C of reactant terpene.
Assuming a proportionality dependence of oxidant
on organic reactant, it follows that at their typical
ambient concentrations of 10 to 50 ppb C in the
rural air, terpenes can produce no more than 2to4
(0.001 to 0.002 ppm) of oxidant/ozone.
137,140
The potential of rural-atmosphere organic
compounds to produce oxidant was also explored
in studies in which rural air samples were captured
in plastic bags, spiked with various amounts of
NO,, and exposed to sunlight. Results by one
investigator were interpreted to mean that in some
rural areas, the ambient organic compounds have
the potential to produce 40 to 1 20 /yg/m3 (0.02 to
0.06 ppm) of oxidant/ozone.105 A recent study in
rural Wisconsin showed that irradiation of
captured air samples spiked with NO did not
produce significantly more oxidant/ozone than the
54
unspiked samples.2 Finally, Singh etal.126 analyzed
ambient ozone monitoring data from a station at
White River, Utah, and concluded that elevated
levels of ozone in the summer occurred only when
ambient NOx was present at concentrations
sufficiently high to act as photochemical ozone
precursors (Figure 4-14). This evidence is
suggestive but far from being conclusive insofar as
the role and importance of natural organic
compounds are concerned. First, it is extremely
difficult to establish with confidence that
oxidant/ozone formation in the captured rural air
samples results solely from natural organic
compounds; the presence of anthropogenic
pollutants and of experimental artifacts cannot be
ruled out.137 Also, relative to free ambient air,
captured air samples probably yield higher
oxidant/ozone concentrations. Singh et al.126also
agree that anthropogenic sources cannot be ruled
out in explaining the elevated ozone con-
centrations observed in White River, Utah, in the
summer (Figure 4-14).126 Finally, substantial
oxidant/ozone formation from natural organic
compounds is not consistent with the laboratory
evidence on terpenes, unless, of course, there are
other yet unknown natural organic compounds
with photochemical reactivities different from tha*
of the terpenes. This latter possibility remains
largely unexplored except for the case of methane,
which is somewhat uncertain.
The modeling of the photochemical methane
oxidation cycle in the lower troposphere has been a
subject of considerable interest and controversy in
recent years.
21,29,35,36,81
The importance of
understanding the methane photo-oxidation cycle
Js not only relevant in its potential role as a source
for natural background ozone concentration, but
also in elucidating the half-lives and budgets for
chemical species present in the atmosphere. Much
of the debate has centered on whether or not the
methane photo-oxidation cycle in the unpolluted
troposphere represents a net ozone production or
destruction process. This in turn reflects on the
classical concept of ozone intrusion from the
stratosphere as being the dominant source of
surface ozone in the unpolluted troposphere.90
Current models of the uncontaminated
troposphere2'81 are sufficiently sensitive to
uncertainties in initial and boundary conditions,
rate constants, transport parameters, and
heterogeneous removal rates that quantitative
statements regarding net ozone production or
destruction are still tenuous. In the final analysis,
the importance of modeling the clean troposphere
-------
X
O
.0
a
a
O
N
O
120
100
80
60
40
20
T
T
T
T
— HIGHEST 1-hr MONTHLY O3 MAX
IMONTHLY AVERAGE OF DAILY 1-hr O3 MAX
" MONTHLY 03 AVERAGE OF ALL 1-hr O3 VALUES
— — LOWEST 1-hr MONTHLY O3 MAX
_.._ MONTHLY AVERAGE OF ALL 1-hr NOX VALUES
AVG. O3
\
• AVG. NO
\
I
PROJECTED O3
PROFILE WITHOUT
NO ADDITION
\
-\
NOV.
(1974)
APR.
SEPT.
MONTHS
FEB.
JULY
(1976)
Figure 4-14. Effect of NOx concentration increase on ozone formation at White River, Utah.1
resides in a clearer understanding of the physical
and chemical processes in operation. This in turn
will provide a better understanding of the effects of
low-level anthropogenic contributions to rural
oxidant/ozone contributions.
One may conclude that natural organic
emissions (at least terpenes) do not directly affect
oxidant/ozone formation within urban areas. Their
indirect effects, however, are not as well
understood. Natural organic compounds do occur
in rural atmospheres and could possibly contribute
to the large-area oxidant/ozone concentrations
associated with high pressure episodes.
Furthermore, products from the atmospheric
degradation of terpenes could also contribute to
those concentrations. The evidence available,
however, suggests that within rural areas where
ambient HC-to-NOx ratios are high, terpenes act as
scavengers rather than producers of ozone.
Furthermore, terpenes decay too fast to survive
transport into urbanized areas where the HC-to-
NOx ratios are more conducive to oxidant formation
from terpenes. The limited evidence available
suggests that chemical degradation products of
terpenes are mainly particulates. Nevertheless,
small amounts of gaseous products, including
aldehydes, ketones, and organic acids are formed
and could conceivably be transported into urban
areas.48 Considering all the evidence, there is no
convincing reason to believe that natural organic
emissions have an important impact on the
oxidant/ozone-related air quality. Such a view is
supported by some, investigators90 but not by
others.29
SINKS OF OXIDANTS AND OF OXIDANT
PRECURSORS
Introduction
The preceding sections identified and described
the various sources of ambient oxidant/ozone and
discussed their absolute and relative contributions
to surface accumulation of oxidant/ozone. Implicit
in those discussions was the assumption that
observed concentrations of oxidant/ozone do not
reflect source emission levels alone; rather, they
reflect the composite impact of the source and sink
processes. Not all sink processes have the same
significance and role. For example, removal of
ambient N02 in reactions forming PAN, nitrates,
etc., is a sink process with entirely different
significance than the removal of NOg on surfaces.
Sink processes relevant to oxidant/ozone are
those that directly or indirectly lessen the impact of
55
-------
the oxidant/ozone sources, thus reducing
concentrations of ambient oxidant/ozone.
Obviously, such sink processes include not only
those for oxidant/ozone, but also those for
oxidant/ozone precursors. For this reason, sink
processes for hydrocarbons and NO* will be
discussed here rather than in the subsequent
chapter on oxidant precursors.
Reviews of the various mechanisms by which
pollutants are removedf rom ambient air have been
made by Robinson and Bobbins,121 by Hidy,63 and
by Rasmussen et al.109'110 The important
mechanisms appear to be: (1) precipitation
scavenging, including rainout (i.e., absorption in
the cloud) and washout (i.e., capture by falling
raindrops), (2) chemical reactions resulting in
products that are not oxidants or that are rapidly
removed from ambient air, (3) dry deposition (i.e.,
adsorption on aerosols and subsequent deposition
on surfaces), and (4) adsorption on ground
surfaces (soil, vegetation, etc.). Information on the
occurrence of such processes for various
pollutants exists, but it is mostly qualitative. For
such information to be useful, it must be
quantitative—that is, expressed in terms of data
compatible with the available mathematical air
quality simulation models. Such data, for example,
are reaction rate constant and kinetic order values
for the various chemical processes, and deposition
rates for the various physical sink processes.
Sinks for Oxidants and Oxidant Precursors
Ozone is relatively insoluble m water. This,
despite evidence that the oceans remove some
ozone,1 suggests that rainout and washout cannot
be important sink processes for ozone.
Reaction of ozone with NO and with HC is an
important sink process accounting for the
complete removal of ozone from the surface air
layers in urban areas at night. On-surface
adsorption/destruction, and some destruction
from locally emitted NO and HC account for the
nighttime absence of ozone from the rural
atmospheres at grou nd level. Thus, under ground-
level conditions and at night, ozone appears to
have a lifetime of only a few hours at most.
In air layers aloft, relatively free of fresh
precursor and surface influences, ozone has been
observed to remain extremely stable, at least
during the night.39 During the day, ozone is
photoJyzed rapidly by sunlight, but the resultant
oxygen atoms react rapidly with molecular oxygen
to regenerate the ozone. Thus sunlight has very
little net effect on ambient ozone in the absence of
other pollutants. In the presence of other
pollutants at their global background levels, ozone
participates in a photochemical destruction cycle
resulting in an ozone half-life estimated to be 7 to 8
hr.15 In the presence of other pollutants at or near
urban concentrations, ozone is both produced and
destroyed in such a way that the initial ozone loses
its identity and significance. The ozone
concentration that develops is the net result of the
formation and destruction processes. The
concentration and lifetime of such ozone depends
on the composition and concentrations of the other
reactants. From the evidence on long-range
oxidant transport, it may be deduced that such
transported ozone could have a lifetime of several
days.
Chemical sink processes for ozone are
qualitatively well established. Quantitatively, the
processes occurring at ground level have not been
defined from direct, real-atmosphere data because
of the difficulty in isolating the chemical process
from the on-surface physical processes.
Nevertheless, reasonably accurate quantification
has been achieved since the main ozone
destruction/reaction steps are well defined
kinetically and are included in the mathematical
models relating air quality to emissions.
The dry deposition sink process for ozone is not
well defined. Surfaces are known to act as ozone
scavengers, but it is not certain that the
composition and surface area of ambient aerosols
are such that they cause substantial ozone
destruction. The stability of ozone in polluted air
layers aloft and in aerosol-containing smog
chamber atmospheres suggests that removal of
ambient ozone by aerosols cannot be an important
sink process.
Absorption and/or destruction of ozone on
ground surfaces (soil, vegetation, water, etc.)
constitute a major sink for this gas on a global
basis.109 Measurements of uptake rates of
different pollutants by certain forms of vegetation
(e.g., alfalfa) showed the ozone rate to be
comparable to that for NQa and somewhat lower
than that for SQa.64 Recently, Garland and
Penketf7 reported ozone deposition velocities of
0.5 cm sec"1 for grass, 1.6 cm sec"1 for soil, and 0.4
cm sec ~1 for sea water (compared to 2 8 cm sec
and 0.2 to 0.7 cm sec"1 for S02 for alfalfa and soils,
respectively). These values are about three times
lower than those reported by Hill,64 but they are
claimed to be more representative of average field
conditions. At any rate, even the low values of
56
-------
Garland and Penkett indicate a substantial
removal of ozone by ground surfaces.
The evidence indicates that of the various
possible sink processes for ozone, chemical
destruction, chemical reaction, and on-surface
destruction are probably the most important. The
importance of the surface sinks, however, can be
assessed only through use of models. One air
quality 'Simulation model has been used to
compute ozone concentrations in a moving
polluted air parcel with and without the
incorporation of surface losses.75 Results showed
that with representative surface loss rates for
ozone and N02 included, the air parcel at the end of
an 8-hr trajectory was computed to have an
average ozone concentration 25 percent lower
than that with the surface losses excluded.
Information on sink processes for oxidants other
than ozone is limited to PAN and NO2. Judging
from PAN solubility in water, rainout and washout
are not expected to be significant sinks for PAN.
The reaction of PAN with NO (0.16 ppm mirf1) is
some two orders of magnitude slower than the
reaction of ozone with NO; nevertheless, this sink
process could be important in urban areas during
evening hours when NO accumulates to high
concentrations. PAN is not destroyed at a
significant rate by sunlight photolysis (k = 5 x 10~3
hr'V but is subject to reversible thermal
decomposition into NO259 (see also section on
atmospheric reaction mechanisms). Because of
this latter process, any removal process for NO2
will also constitute an indirect sink for PAN. The
low vapor pressure of PAN suggests that dry
deposition could be an important sink for this
oxidant; however, quantitative data are not
available. In contrast, data for on-surface
adsorption do exist and show PAN to be somewhat
less adsorbable than ozone, with a deposition
velocity of about 0.25 cm sec"1 for grass and soil
surfaces.47 Such a value suggests that removal on
surfaces is probably one of the major sinks for PAN
in the lower troposphere. In air layers aloft, where
the NO-reaction and surface sinks are ineffective,
PAN can travel long distances.
Nitric oxide and nitrogen dioxide can be removed
to varying degrees by all four sink mechanisms.
The primary sink for NO is the reaction with
oxidizing radicals to produce N02. Rainout and
washout removal of NO are minimal because of its
low solubility. Adsorption of NO by vegetation and
by soil occurs very slowly, at deposition velocities
around 0.1 cm sec"1.109 No2 is removed chiefly by
oxidation into nitrates and peroxyacylnitrates and
subsequent removal through precipitation, dry
deposition, and surface adsorption. Removal of
N02 on alfalfa surfaces was measured and found
by Hill84 to occur with a deposition velocity of 2 cm
sec"1. The impact of surface sinks for N02 on
ambient ozone and N02 was computed by Killus
and Jerskey,75
Information on sinks for HC is far more
incomplete than for the other oxidant-related
pollutants, probably because of the large variety of
organic compounds (hydrocarbon and non-
hydrocarbon) in ambient air. Hydrocarbons are, in
general, insoluble in water and therefore cannot
be removed significantly by precipitation.
Chemical reactions converting the reactive HC to
soluble and/or condensable products constitute
the most important sink processes. Limited data
reported by Hidy83 indicate that 1 to 10 percent (by
weight) of reactive HC emissions are converted to
removable aerosol. The remaining HC are
eventually oxidized to CO2 and H2O, but with
parallel formation of oxidant/ozone. Such a
conversion does not, therefore, constitute a sink.
Removal by soil apparently does occur, but data
exist for a few hydrocarbons only (e.g., ethylene
and acetylene).109 The sink processes for
unreactive hydrocarbons are the same, in general,
as for the reactive ones, but they occur more
slowly.
SUMMARY
Evidence shows that in and around urban
centers with high oxidant/ozone concentrations,
photochemical oxidant and ozone are mainly
formed from anthropogenic organic and NO*
emissions. In the years since publication of the
predecessor to this criteria document, the
mechanisms of the atmospheric oxidant/ozone
formation process have been studied intensively
and are now better understood. Most noteworthy
are recent findings pertaining to the roles of
hydroxyl (OH-) and (HO-2), The reaction with OH
has been established to be a major HC-consuming
process, and OH and H02 have been identified as
having major roles in the atmospheric oxidation of
NO to NO2. Aldehydes and PAN were also found to
play important roles in the atmospheric reaction.
For olefins and paraffins, at least, these reactions
are now understood to the extent that the kinetics
of photochemical hydrocarbon-NO, reaction
systems, as observed in the laboratory, can be
described with reasonable accuracy. Additional
research is needed to understand the atmospheric
57
-------
reactions of aromatic hydrocarbons and to further
clarify the differences between laboratory and
ambient atmospheric chemical systems.
The photochemical formation of oxidant/ozone
is the result of two coupled processes: a physical
process involving dispersion and transport of the
oxidant precursor emissions (e.g., HCand N0x)and
the photochemical reaction process. Both
processes are strongly influenced by mete-
orological factors such as dispersion, solar
radiation, temperature, and humidity. Recently
compiled statistics on wind velocity and mixing
height showed that episodes of limited dispersion
are most common in the Far West and within the
Rocky Mountain region, least common over the
Plains States, and of intermediate frequency east
of the Mississippi. New measurements of solar
radiation lead to results somewhat different from
those reported earlier. Recent field and laboratory
studies suggest that at temperatures below
approximately 55° to 60°F (13° to 16°C), con-
centrations of photochemical ozone are unlikely to
exceed the national 1-hr 160-A/g/m3 (0.08-ppm)
standard.
Of all recent developments regarding the
understanding of the sources of photochemical
oxidant/ozone, perhaps the most important ones
were in the area of photochemical oxidant/ozone
transport. Short-range (urban-scale) transport has
been shown to occur and to cause the highest
ozone concentrations some distance downwind
from the core area (region of highest emission) of
urban center Intermediate-range (mesoscale)
transport has been identified in the form of urban
oxidant/ozone plumes extending as far as 100
miles (160 km) or more downwind, and also
involved in land-sea-breeze circulation. Finally,
synoptic-scale transport (over a range of several
hundred miles) associated with high pressure
systems has been found to occur extensively.
These findings have significant implications with
respect to the location of oxidant/ozone monitors.
Conditions under long-range transport are such
that ozone production per unit of precursor is
enhanced. Also, many organics previouslythought
to be unreactive are now believed to have
significant ozone-producing potential.
In addition to the troposphenc photochemistry of
anthropogenic emissions, potential sources of
troposphenc oxidant/ozone are the stratosphere
and the photochemistry of natural organic and NO*
emissions. Estimates of stratospheric ozone in-
trusions at ground level are based on two types of
evidence. (1) global circulation patterns, namely.
patterns in air interchange between stratosphere
and tropsphere, and (2) data on variations of ozone
concentrations in remote rural areas. Based on the
evidence of stratosphere-troposphere inter-
change, the annual average stratospheric
contribution to ozone concentrations at ground
level is estimated to be 0.022 to 0.05 ppm. The
highest concentrations, at or above 0.08 ppm, from
that source are expected to occur mostly during
April and May. Occurrence of such major
intrusions of stratospheric ozone concentration
during the spring months in midlatitudes was
indicated also by analysis of the ozone data for
remote rural areas. The evidence suggests that the
probable concentration of stratospheric ozone
reaching ground level during the smog season
(usually late summer or early fall) is about 60
/Kj/m3 (0.03 ppm). There are also more recentdata
obtained at Whiteface Mountain, New York,
suggesting a maximum 24-hr concentration of 72
/Kj/m3 (0.037 ppm) stratospheric ozone in July.
Certain organic emissions from vegetation
(terpenes) were found to play the dual role of ozone
precursor and destroyer. Despite the substantial
rates at which they are emitted in forested areas,
the ambient concentrations of such organics,
because of their reactivity and the areal dispersion
of their sources, seldom exceed a few parts per
billion. At these ambient concentrations, the direct
potential of terpenes for photochemical ozone
formation is estimated to be negligible. It is
conceivable, however, that the products of
atmospheric reactions involving large amounts of
terpenes do have a significant impact on
oxidant/ozone-related air quality.
REFERENCES FOR CHAPTER 4
1 Aldaz, L Flux measurements of atmospheric ozone over
land and water J Geophys Res 746943-6946,1969
2 Allwme, K J , and H Westberg Potential Impact of
Coal-Fired Power Plants on Ground Level Ozone
Concentrations in Wisconsin Final Report—Wisconsin
Public Service Commission Contract No 8110,
Washington State University, Pullman, Wash , 1977
3 Altshuller, A P Evaluation of oxidant results at CAMP
sites in the United States J Air Pollut Control Assoc
25 19-24, 1975
4 Altshuller, A P, and J J Bufalmi Photochemical
aspects of air pollution A review Environ Sci Technol
539-64, 1971
5 Altshuller, A P, and J J Bufalmi Photochemical
aspects of air pollution A review Photochem
Photobiol. 4 97-146, 1965
6 Altshuller, A P , and S P McPherson Spec-
trophotometnc analysis of aldehydes in the Los Angeles
atmosphere J Air Pollut Control Assoc 73109-111,
1963
58
-------
7. Atkinson, R., B. J. Finlayson, and J. N. Pitts, Jr.
Photoionization mass spectrometer studies of gas
phase ozone-olefin reactions. J Am. Chem. Soc.
95:7592-7599, 1973.
8. Atkinson, R., D. A. Hansen, and J. N. Pitts, Jr. Rate
constants for the reaction of OH radicals with Hz and NO
(M = AR and N2). J. Chem Phys. 62:3284-3288, 1975.
9. Attmannspacher, W., and R. Hartmannsgruber. On
extremely high values of ozone near the ground. Pure
Appl. Geophys. 106-108,}091 -1096, 1973.
10. Bach, W. D., Jr. Investigation of Ozone and Ozone
Precursor Concentrations in Nonurban Locations in the
Eastern United States. Phase II, Meteorological
Analyses. EPA-450/3-74-034a, U.S. Environmental
Protection Agency, Research Triangle Park, N.C.,
February 1975.
11. Bach, W. D., Jr., J. E. Sickles, II, R. Denyszyn, and W. C.
Eaton. Chemical and meteorological analysis of the
mesoscale variability of ozone concentrations over a six-
day period. In'. International Conference on
Photochemical Oxidant Pollution and Its Control.
Proceedings. Volume I. B Dimitnades, ed EPA-600/3-
77-001 a, U.S. Environmental Protection Agency,
Research Triangle Park, N.C., January 1977. pp. 197-
210.
12. Bell, G. B. Meteorological Conditions During Oxidant
Episodes in Coastal San Diego County in October and
November, 1959. State of California, Department of
Public Health, Sacramento, Calif., 1960.
13. Bruntz, S. M., W. S. Cleveland, T. E. Graedel, B. Kleiner,
and J. L. Warner. Ozone concentrations in New Jersey
and New York: Statistical association with related
variables. Science 756:257-259, 1974.
14. Bufalini, J. J., B. W. Gay, Jr., and K. L. Brubaker.
Hydrogen peroxide formation from formaldehyde
photooxidation and its presence in urban atmospheres.
Environ. Sci. Technol. 6:816-821, 1972.
15. Bufalini, J. J , T. A. Walter, and M. Bufalini. Ozone
formation potentials of organic compounds. Environ.
Sci. Technol. 70:908-912, 1976.
16 Calvert, J G. Modes of formation of the salts of sulfur
and nitrogen in an NOx-SOz-hydrocarbon-polluted
atmosphere. In: Conference on Atmospheric Salts and
Gases of Sulfur and Nitrogen in Association with
Photochemical Oxidant. Volume II. Reference
Documents. T. T. Crocker, ed. California Air Resources
Board Contract No 3-197, University of California,
Irvine, Calif., January 1974. Chapter VIII.
17 Calvert, J. G Review and analysis. In: International
Conference on Oxidants, 1976—Analysis of Evidence
and Viewpoints. Part II. The Issue of Reactivity. EPA-
600/3-77-114, U.S. Environmental Protection Agency,
Research Triangle Park, N.C., October 1977. pp. 5-34.
18. Calvert, J. G. Test of the theory of ozone generation in
Los Angeles atmosphere. Environ. Sci. Technol.
70.248-256, 1976.
19. Calvert, J. G., and R. D. McQuigg. The computer
simulation of the rates and mechanisms of
photochemical smog formation. In: Proceedings of the
Symposium on Chemical Kinetics Data for the Upper
and Lower Atmosphere. Int. J. Chem. Kmet. Symp.
(1).113-154, 1975.
20. Castleman.A W., Jr., R. E. Davis, H. R. Munkelwitz, I.N.
Tang, and W. P. Wood. Kinetics of association reactions
pertaining to HSSO4 aerosol formation. In: Proceedings
of the Symposium on Chemical Kinetics Data for the
Upper and Lower Atmosphere. Int. J. Chem. Kinet.
Symp. (1):629-640, 1975.
21. Chameides, W., and J. C. G. Walker. A photochemical
theory of tropospheric ozone. J. Geophys. Res. 75:8751 -
8760, 1973.
22. Chatfield, R., and H. Harrison. Ozone in the remote
troposphere: Higher levels? In: Specialty Conference
On: Ozone/Oxidants—Interactions with the Total
Environment. Air Pollution Control Association,
Pittsburgh, Pa., 1976. pp. 77-83.
23. Chatfield, R and R. A. Rasmbssen. An assessment of
the continental lower tropospheric ozone budget. In:
International'Conference on Photochemical Oxidant
Pollution and Its Control. Proceedings. Volume I. B.
Dimitriades, ed. EPA-600/3-77-001 a, U.S.
Environmental Protection Agency, Research Triangle
Park, N.C , January 1977. pp. 121-136.
24. Cleveland, W. S., T. E. Graedel, B. Kleiner, and J. L.
Warner. Sunday and workday variations in
photochemical air pollutants in New Jersey and New
York. Science 7S6.1037-1O38, 1974.
25. Cleveland, W. S., R. Guarino, B. Kleiner, J. E. McRae,
and J. L. Warner. The analysis of the ozone problem in
the northeast United States. In: Specialty Conference
On: Ozone/Oxidants—Interactions with the Total
Environment. Air. Pollut. Control Association,
Pittsburgh, Pa., 1976. pp 109-120.
26. Coffey, P. E., and W. N. Stasiuk. Evidence of atmospheric
transport of ozone into urban areas. Environ. Sci.
Technol. 9:59-62, 1975.
27. Coffey, P. E., and W. N. Stasiuk. Urban ozone: Its local
and extraregional components. Presented at the 79th
Annual National Meeting, American Institute of
Chemical Engineers, Houston, Texas, March 1975.
Paper No.60C.
28. Coffey, P. E., W. N. Stasiuk, and V. Mohnen. Ozone in
rural and urban areas of New York State. In:
International Conference on Photochemical Oxidant
Pollution and Its Control. Proceedings. Volume I. B.
Dimitriades, ed. EPA-600/3-77-001 a, US.
Environmental Protection Agency, Research Triangle
Park, N.C , January 1977. pp. 89-96.
29. Coffey, P. E., and H.Westberg International Conference
on Oxidants, 1976—Analysis of Evidence and
Viewpoints. Part IV. The Issue of Natural Organic
Emissions. EPA-600/3-77-116, U.S. Environmental
Protection Agency, Research Triangle Park, N.C.,
October 1977.
30. Cox, R. A , A E. J. Eggleton, R G Derwent, J. E.
Lovelock, and D. H. Pack Long-range transport of
photochemical oxidant in northwestern Europe. Nature
(London) 225(55404).118-121, 1975.
31 Cox, R. A., and R. G. Derwent Kinetics of the reaction of
HOz with nitric oxide and nitrogen dioxide. J.
Photochem. 4.139-153, 1975.
32. Cox, R. A., R. G. Derwent, and P M. Holt Relative rate
constants for the reactions of OH radicals with Hz, CH4,
CO, NO, and HONO at atmospheric pressure and 296K.
J. Chem. Soc. Faraday Trans. 1. 72:2031-2043, 1976.
33. Cox, R. A , R. G. Derwent, and A. J L. Hutton. The
significance of peroxynitric acid in atmospheric
chemistry of the nitrogen oxide's. Nature (London)
270:328-329, 1977.
59
-------
34, Cox, R. A., and M. J. Roffey. Thermal decolnposition of
peroxyacetyl nitrate in the presence of nitric oxide.
Environ Sci. Technol. 77:900-906, 1977.
35. Crutzen, P. J. Photochemical reaction initiated by and
influencing ozone in unpolluledtroposphericair Tellus.
26:47-56, 1974,
36. Crutzen, P. J., and J. Fishman. Average concentrations
of hydroxyl radicals in the troposphere and the budgets
of methane, carbon monoxide, molecular hydrogen, and
1,1,1-trichloroethane. Geophys, Res Lett. 4321-324,
1977
37. Darnali, K, R,, W. P. L. Carter, A. M. Winer, A. C. Lloyd,
and J. N. Pitts, Jr. Importanceof ROa + NO in alkyl nitrate
formation from C^-Cs alkane photooxidations under
simulated atmospheric conditions J. Phys. Chem.
80:1.948-1950, 1976.
38. Decker, C, E., L. A. Ripperton, J J. B. Worth, F. M.
Vukovich, W, D Bach, J B. Tommerdahl, F. Smith, and
D. E. Wagoner. Formation and Transport of Oxidants
Along the Gulf Coast and in Northern U.S. EPA-450/3-
76-033, U.S. Environmental Protection Agency,
Research Triangle Park, N.C., August 1976.
39. Decker, C E., J J, B. Worth, L. A. Ripperton, and W. D.
Bach. Ambient Monitoring Aloft of Ozone and
Precursors in the Vicinity and Downwind of a Major City.
EPA-450/3-77-009, U.S Environmental Protection
Agency, Research Triangle Park, N.C., January 1977
40. Demerjian, K. L , J. A Kerr, and J. G. Calvert. The
mechanisms of photochemical smog formation.
Environ. Sci Technol 4:1-262, 1974.
41. Dimitriades," B. Oxidant control strategy: ~ Recent
developments' In; International Conference on
Photochemical Oxidant Pollution and Its Control,
Proceedings Volume II. B. Dimitriades, ed. EPA-600/3-
77-001 b, U.S, Environmental Protection Agency.
Research Triangle Park, N.C., January 1977. pp. 1143-
1154,
42. Dimitriades, B. Oxidant/Os air quality benefits from
"emission substitution." In. Specialty Conference On:
Hydrocarbon Control Feasibility Its Impact on Air
Quality. Air Pollution Control Association, Pittsburgh,
Pa., 1977. pp. 111-121.
43 Dimitriades, B,, and A. P Altshuller. International
conference on oxidant problems: Analysis of the
evidence and viewpoints presented. Part I Definition of
key issues. J. Air Pollut. Control Assoc. 27:299-307,
1977.
44. Dobbins, R, A., J. L Nolan, J. P Okolowicz, and A J.
Gilbert, Meteorological factors controlling photo-
chemical pollutants in southeastern New England. !n;
International Conference on Photochemical Oxidant
Pollution and Its Control, Proceedings Volume I. B,
Dimitriades, ed. EPA-600/3-77-001 a, U.S,
Environmental Protection Agency, Research Triangle
Park, N.C,, January 1977. pp. 109-118
45. Elkus, B., and K. R. Wilson, Photochemical oxidant:
Weekday-weekend differences. Draft. 1976.
46. Fankhauser, R K. Ozone levels in the vicinity of 33
cities. J. Air Pollut, Control Assoc. 26771-777, 1976.
47. Garland, J A , and S A. Penkett. Absorption of peroxyl
acetyl nitrate and ozone by natural surfaces. Atmos.
Environ. W 1127-1131, 1976
48. Gay, B. W., Jr., and R. R. Arnts. The chemistry of
naturally emitted hydrocarbons. In. International
Conference on Photochemical Oxidant Pollution and Its
Control. Proceedings Volume I. B. Dimitriades, ed. EPA-
600/3-77-001 b, U.S Environmental Protection
Agency, Research Triangle Park, N C , January 1977.
pp 745-751
49. Graedel, T. E , L. A. Farrow, and T. A. Weber. The
influence of aerosols on the chemistry of the
troposphere. In, Proceedings of the Symposium on
Chemical Kinetics Data for the Upper and Lower
Atmosphere. Int J. Chem. Kmet. Symp. (1 (.581-594,
1975.
50. Graham, R. A., A. M. Winer, and J. N. Pitts, Jr,
Temperature dependence of the unimolecular
decomposition of pernitric acid and its atmospheric
implications. Chem. Phys, Lett 57.215-220, 1977
51. Greenfelt, P. Ozone episodes on the Swedish west
coast, In- International Conference on Photochemical
Oxidant Pollution and Its Control. Proceedings Volume
I. B. Dimitriades, ed. EPA-600/3-77-001 a, U.S.
Environmental Protection Agency, Research Triangle
Park, N.C., January 1977 pp. 329-337
52. Grimsrud, E. P., H. H. Westberg, and R. A, Rasmussen.
Atmospheric reactivity of monoterpene hydrocarbons,
NO«-photooxidation, and ozonalysis. In. Proceedings of
the Symposium on Chemical Kinetics Data for the Upper
and Lower Atmosphere, Int. J, Chem Kinet. Symp.
(1):183-195, 1975.
53. Guidance for Air Quality Monitoring Network Design
and Instrument Siting. Guideline No 1 2-012, U.S
Environmental Protection Agency, Research Triangle
Park, N.C., 1975.
54. Haagen-Smit, A J. Chemistry and physiology of Los
Angeles smog. Ind Eng. Chem. 44.1342-1346, 1952.
55. Hanst, P. L, W. E. Wilson, Jr., R, K. Patterson, B W Gay,
Jr., L. W. Chaney, and C. S. Burton A spectroscopic
study of Pasadena smog. In: Abstracts of Papers: 167th
National Meeting American Chemical Society.
American Chemical Society, Washington, D C., 1974.
Abstract GNVT 056.
56. Harvey, R. B., D. H Stedman, and W Chameides.
Determination of the absolute rate of solar photolysis of
NOs J. Air Pollut. Control Assoc, 27'663-666, 1977
57. Hathorn, J. W., Ill, and H. M Walker. A "Texas size"
ozone episode tracked to its source. In: International
Conference on Photochemical Oxidant Pollution and Its
Control.Proceedings Volume I. B. Dimitriades, ed. EPA-
600/3-77-001 a, U.S. Environmental Protection
Agency, Research Triangle Park, N.C., January 1977,
pp. 353-380.
58. Heicklen, J.,K. Westberg, and N Cohen, The conversion
of NO to NOz in Polluted Atmospheres. CAES
Publication No. 115-69, Pennsylvania State University,
Center for Air Environment Studies, University Park,
Pa., 1969,
59. Hendry, D. G , and R. A. Kenley Gas-phase free radical
reactions of peroxyacetyl nitrate. Presented at the 173rd
National Meeting, American Chemical Society, New
Orleans, Louisianna, March 21-25, 1977
60. Hendry, D. G., and R. A. Kenley. Generation of peroxy
radicals from peroxynitrates (ROzNO2). Decomposition
of peracyl nitrates. J Am. Chem. Soc 55:3198-3199,
1977.
61 Hester, N. E., R. B. Evans, F. G Johnson, and E. L.
Martinez, Airborne measurements of primary and
60
-------
secondary pollution concentrations in the St. Louis
urban plume, ln~ International Conference on
Photochemical Oxidant Pollution and Its Control.
Proceedings. Volume i. B, Dirrtitrtades, ed, EPA-600/3-
77-001 a, U.S Environmental Protection Agency,
Research Triangle Park, N.C., January 1977. pp. 259-
274.
62. Heuss, J. M,, and W. A, Glasson. Hydrocarbon reactivity
and eye irrritation. Environ. Sci.Teohnol. 2:1109-1116,
1968,
63, Hidy, G M. Removal processes of gaseous and
paniculate pollutants. In: Chemistry in the Lower
Atmosphere S. I. Rasool, ed. Plenum Press, New York,
1973.
64. Httl. A. C. Vegetation, A sink, for atmospheric pollutants,
J Air Pollut. Control Assoc. 2/:341-346, 1971.
65, Holzworth, G. C. Mixing Heights, Wind Speeds, and
Potential for Urban Air Pollution Throughout the
Contiguous United States, Office of Air Programs
Publication No, AP-101, U.S. Environmental Protection
Agency, Research Triangle Park, N.C., January 1972.
66, Huffman, G. D., G, W. Haering, R. C, Bourke, P. P. Cook,
and M. P. Sillars. Ozone observations in and around a
midwestern metropolitan area. In, International
Conference on Photochemical Oxidant Pollution and Its
Control. Proceedings. Volume I. B. Dimitriades, ed, EPA-
600/3-77-0018, U.S, Environmental Protection
Agency, Research Triangle Park, N.C., January 1977,
pp. 341 -352.
67. Husam, L., P. E. Coffey, R. E, Meyers, and R. T.
Cederwatl. Ozone transport from stratosphere to
troposphere Geophys. Res. Lett. 4:363-365, 1977.
68. Husar, R. B., D. E, Patterson, C. In: International
Conference on Photochemical Oxidant Pollution and Its
Control. Proceedings. Volume I, B, Dimitriades, ed.EPA-
600/3-77-001 a, U.S. Environmental Protection
Agency, Research Triangle Park, N.C., January 1977,
pp. 275-282.
69. Jackson, J. O,, D, H. Stedman, R. G.Smith, J, H. Hecker,
and P 0. Warner. Direct NOs photolysis rate monitor.
Rev, Sci. Instrum. 46:376-378, 1975,
70, Jaffe, R. J., F, C, Smith, Jr, and K, W. Last. Study of
Factors Affecting Reactions in Environmental
Chambers. Phase III. EPA-65O/3-74-004b, U.S.
Environmental Protection Agency, Research Triangle
Park, N.C., November 1974.
71. Jeffries, H, E, Review and analysis. In: International
Conference on Oxidants, 1976—Analysis of Evidence
and Viewpoints. Part II. The Issue of Reactivity. EPA-
600/3-77-114, U.S. Environmental Protection Agency,
Research Triangle Park, N.C., October 1977. pp 35-70
72 Jeffries, H., D, Fox, and R. Kamens, Outdoor Smog
Chamber Studies. Effect of Hydrocarbon Reduction on
Nitrogen Dioxide, EPA-650/3-75-01 1, U.S.
Environmental Protection Agency, Research Triangle
Park, N.C,, June 1975.
73. Jeffries, H, E., R. M. Kamens, and C. E. Feigley. Outdoor
Simulations of Air Pollution Control Strategies.
Progress Report, 1975-1976, University of North
Carolina, School of Public Health, Chapel Hill, N.C,,
January 1977.
74. Jerskey,T.N.,T B. Smith, S. L. Marsh, andW.H,White,
The effect of ozone layers aloft on surface
concentrations, In: International Conference on
Photochemical Oxidant Pollution and Its Control
Proceedings. Volume I. B Dimitriades, ed. EPA-600/3-
77-001 a, U.S, Environmental Protection Agency,
Research Triangle Park, N. C., January 1977. pp. 145-
153.
75. Killus, J. P., and T. N, Jerskey Surface sinks. In:
Continued Research in Mesoscale Air Pollution
Simulation Modeling: Volume V—Refinements in
Numerical Analysis, Transport, Chemistry, and
Pollutant Removal. Draft Final Report. U.S.
Environmental Protection Agency Contract No. 68-02-
2216, Systems Applications, Inc., San Rafael, Calif.,
March 1977. pp. 67-120,
76. Korshover, J. Climatology of stagnating anticyclones
east of the Rocky Mountains, 1936-1975. NOAA
Technical Memorandum ERL ARL-55, U.S. Department
of Commerce, National Oceanic and Atmospheric
Administration, Washington, D.C., July 1975.
77, Lamb, R. G. A case study of stratospheric ozone
affecting ground-level oxidant concentrations, J. Appl.
Meteorol. 16:780-794, 1977.
78 Leighton, P. A. Photochemistry of Air Pollution. Aca-
demic Press, Inc., New York, 1961.
79. Levine, S. Z., W M. Uselman,W. H. Chan, J. G. Calvert,
and J, H, Shaw. The kinetics and mechanism of the HOa-
NOz reactions: The significance of peroxynitric acid
formation in photochemical smog. Chem. Phys, Lett.
45:528-535, 1977.
80. Levitt, S. B, and D, P. Chock, Weekday-weekend
pollutant studies of the Los Angeles Basin. J Air Pollut.
Controi Assoc. 26:1091-1092, 1976.
81. Levy, H., II. Photochemistry of the lower troposphere,
Planet. Space Sci. 20:919-935, 1972.
82. Lonneman, W. A. Ozone and hydrocarbon meas-
urements in recent oxidant transport studies. In
International Conference on Photochemical Oxidant
Pollution and Its Control. Proceedings. Volume I. B.
Dimitriades, ed. EPA-600/3-77-001a, U.S, Envi-
ronmental Protection Agency, Research Triangle Park,
N.C., January 1977. pp, 211-223.
83. Lonneman, W. A.. R. L. Seila, and S. A. Meeks.
Preliminary results of hydrocarbon and other pollutant
measurements taken during the 1975 northeast oxidant
transport study. In: Proceedings of Symposium on 1975
Northeast Oxidant Transport Study. EPA-600/3-77-
017, U.S. Environmental Protection Agency, Research
Triangle Park, N.C., February 1977. pp. 40-53.
84. Ludwig, F. L., E. Reiter, E, Shelar, and W. B. Johnson,
The Relation of Oxidant Levels to Precursor Emissions
and Meteorological Features Vo). 1 Analysis and
Findings. EPA-450/3-77-022a, U.S. Environmental
Protection Agency, Research Triangle Park, N.C,,
September 1977.
85. Lyons, W. A., and H. S. Cole, Photochemical oxidant
transport: Mesoscale lake breeze and synoptic-scale
aspects. J Appl. Meteorol. 75:733-743, 1976.
86. Marks, J. E., and J. L. Herman. Mesoscale trans-
portation of ozone: An episode, August 18, 1974. In,
Specialty Conference On: Ozone/Oxidants—
Interactions with the Total Environment. Air Pollution,
Control Association, Pittsburgh, Pa., 1976.pp 293-302.
87, Martinez, E, L., and E. L. Meyer, Jr, Urban-nonurban
ozone gradients and their significance. In: Special
.. Conference on: Ozone/Oxidants—Interactions with the
61
-------
Total Environment. Air Pollution Control Association,
Pittsburgh, Pa., 1976. pp. 221-133.
88. McAfee, J. M , A. M. Winer, and J. N Pitts, Jr. Infrared
product analysis of ozone-olefin reactions by Fourier
interferometry. Presented at the Symposium on
Chemical Kinetics Data for the Upper and Lower
Atmosphere, Stanford Research Institute and National
Bureau of Standards, Warrenton, Va., September 15-
18, 1974.
89. McKee, H. C Atmospheric Measurements at Seadock
Site Southwest Research Institute, Houston, Texas,
April 1974.
90 Mohnen, V A. Review and analysis In International
Conference on Oxidants, 1976—Analysis of Evidence
and Viewpoints. Part III The Issue of Stratospheric
Ozone and Intrusion. EPA-600/3-77-115, U.S
Environmental Protection Agency, Research Triangle
Park, N C , December 1977 pp. 7-66.
91. Morris, E, D , Jr., and H Niki Reaction of dinitrogen
pentoxide with water J. Phys, Chem. 77.1929-1932,
1973.
92, National Air Pollution Control Administration. Air
Quality Criteria for Photochemical Oxidants. NAPCA
Publication No AP-63, U.S Department of Health,
Education, and Welfare, Public Health Service,
Washington, D C., March 1970.
93. National Research Council. Ozone and Other
Photochemical Oxidants National Academy of
Sciences, Washington, D.C., 1977.
94. Niki, H., E. E. Daby, and B. Wemstock. Mechanisms of
smog reactions Adv. Chem Ser. 7/3:16-57, 1972.
95, Niki, H., P D. Maker, C M. Savage, and L. P
Breitenbach Fourier transform IR spectroscopic
observation of pernitnc acid formed via HOO + NOs T
HOONO2. Chem. Phys. Lett 45 564-566, 1977
96 Office of Air Quality Planning and Standards Position
Paper on Regulation of Atmospheric Sulfates. EPA-
450/2-75-007, U.S Environmental Protection Agency,
Research Triangle Park, N.C , September 1975.
97. Pack, D H. Review and analysis. In International
Conference on Oxidants—1976 Analysis of Evidence
and Viewpoints Part V. The Issue of Oxidant Transport.
EPA-600/3-77-1 17, US Environmental Protection
Agency, Research Triangle Park, N.C., November 1977
p. 9-20
98. Pate,C.T.R.Atkinson,andJ.N Pitts.Jr Rateconstants
for the gas phase reaction of peroxyacetyl nitrate with
selected atmospheric constituents, J Environ Sci.
Health Environ Sci Eng. A17(1) 19-31, 1976
99 Payne, W. A., L J. Stief, and D.D.Davis. A kmeticsof the
reaction of HO2 with SOs and NO. J. Am. Chem. Soc.
95.7614-7619, 1973.
100. Peterson, J T Calculated Actinic Fluxes (290-700 nm)
for Air Pollution Photochdmstry Applications. EPA-
600/4-76-025, U.S. Environmental Protection Agency,
Research Triangle Park, N C., June 1976.
101. Peterson, J T.,K L Demerjian, and K. L Schere Actinic
solar flux and photolytic rate constants in the
troposphere In' International Conference on
Photochemical Oxidant Pollution and Its Control.
Proceedings. Volume I. B Dimitnades, ed EPA-600/3-
77-001 a, U.S. Environmental Protection Agency,
Research Triangle Park, N C, January 1977. pp. 763-
773.
102 Price J. H A study of factors associated with high urban
ozone concentrations in Texas. In Specialty Conference
on1 Ozone/Oxidarrts—Interactions with the Total
Environment. Air Pollution Control Association,
Pittsburgh, Pa., 1976. pp. 282-292,
103 Quon, J, E., and R A. Wadden. Oxidants in the Urban
Atmosphere. IIEQ Publication No. 75-04, State of
Illinois, Institute for Environmental Quality, Chicago, III.,
January 1975
104 Rasmussen, R, A. Isoprene; Identified as a forest-type
emission to the atmosphere. Environ. Sci. Techno!
4:667-671, 1970
105. Rasmussen, R A Recent field studies. In, Scientific
Seminar on Automotive Pollutants. EPA-600/9-75-
003, U.S. Environmental Protection Agency, Research
Triangle Park, N.C., February 1975. Sect. 5, Subset J
106. Rasmussen, R A, Surface ozone observations in rural
and remote areas. J Occup. Med. 18.346-350, 1976
107. Rasmussen, R A. What do the hydrocarbons from trees
contribute to air pollution? J Air Pollut Control Assoc.
22537-543, 1972
108. Rasmussen, R. A., and M. W. Holdren. Analysis of C5 to
Cto hydrocarbons in rural atmospheres. Presented at
the 65th Annual Meeting Air Pollution Control
Association, June 18-22, 1972
109 Rasmussen, K. H., M. Taheri, and R. L. Kabel. Global
emissions and natural processes for removalof gaseous
pollutants Water Air Soil Pollut 4 33-64, 1975.
1 10. Rasmussen, K. H , M. Taheri, and R L. Kabei. Sources
and Natural Removal Processes for Some Atmosphere
Pollutants EPA-650/4-74-032, U.S Environmental
Protection Agency, Research Triangle Park, N C. 1974
111. Reiter, E. R In, International Conference on Oxidants,
1976—Analysis of Evidence and Viewpoints. Part III.
The issue of Stratospheric Ozone Intrusion. EPA-
600/3-77-1 15, U.S. Environmental Protection Agency,
Research Triangle Park, N C., December 1977
112. Reiter, E. R. The role of stratospheric import on
troposphenc drone concentrations. In: International
Conference on Photochemical Oxidant Pollution and Its
Control. Proceedings. Volume I. B. Dimitnades, ed. EPA-
600/3-77-001 a, US. Environmental Protection
Agency, Research Triangle Park, N C., January 1977.
pp 393-410
113. Reiter, E. R The transport of radioactive debris and
ozone from the stratosphere to the ground. Arch
Meteorol. Geophys. Bioklimatol. Ser. A 26'323-340,
1977.
114 Renzetti, N A., and R. J Bryan. Atmospheric sampling
for aldehydes and eye irritation m Los Angeles smog—
1960. J Air Poilut. Control Assoc. 77 421-424, 1961.
115. Research Triangle Institute Investigation of Ozone and
Ozone Precursor Concentrations at Non-Urban
Locations in the Eastern United States. EPA-450/3-74-
034, U.S. Environmental Protection Agency, Research
Triangle Park, N.C., May 1974.
116 Research Triangle Institute. Investigation of Rural
Oxidant Levels as Related to Urban Hydrocarbon Control
Strategies. EPA-450/3-75-035, U.S Environmental
Protection Agency, Research Triangle Park, N.C., March
1975
117 Richardson, R, L, D. Johnson, and R. R. Wallis
Background Concentrations of Hydrocarbons in the
Atmosphere of the Northwest Gulf of Mexico. Texas Air
Control Board, Austin, Tex., September 20, 1973.
62
-------
118. Ripperton, L A., J. J. B. Worth, F. M. Vukovich, andC. E.
Decker. Research Triangle Institute studies of high
ozone concentration in nonurban areas, In'
International Conference on Photochemical Oxidant
Pollution and Its Control. Proceedings. Volume I. B.
Dimitnades, ed. EPA-600/3-77-001 a, U.S
Environmental Protection Agency, Research Triangle
Park, N.C , January 1977.
119. Roberts, P. T., and S. K. Fnedlander. Photochemical
aerosol formation of SO2, 1-heptene, and NO» in
ambient air. Environ. Sci. Technol. 10.573-580, 1976.
120 Robinson, E. Review and analysis. In' International
Conference on Oxidants—1976 Analysis of Evidence
and Viewpoints. Part V. The Issue of Oxidant Transport.
EPA-600/3-77-117, U.S. Environmental Protection
Agency, Research Triangle Park, N.C., November 1977.
pp 21-38
121. Robinson, E., and Robbms, R C Sources, Abundance
and Fate of Gaseous Atmospheric Pollutants. SRI
Project PR-6755, Stanford Research Institute, Menlo
Park, Calif , 1963.
122 Sander, S. P., and J H. Seinfeld. Chemical kinetics of
homogeneous atmospheric oxidation of sulfur dioxide.
Environ. Sci Technol. 701114-1123,1976.
123. Schmitz, H. J., and D Cox. Local Air Pollution Episode
from May 16 to May 27, 1972. Metropolitan Dade
County Pollution Control, Miami, Fla., June 1972
124. Seinfeld, J. H. Letter to the Editor. Environ. Sci Technol
7/1218-1219, 1977.
125 Simonaitis, R., and J Heicklen. The Temperature
Dependence of the Reactions of HO2 with NO and NO2.
PSU-IRL-SCI-454, Pennsylvania State University,
University Park, Pa , March 1977.
1 26. Singh, H. B , F. L. Ludwig, and W B Johnson Ozone in
Clean Remote Atmospheres Concentrations and
Variabilities. SRI Project 5661, Stanford Research
Institute, Menlo Park, Calif., June 1977.
127. Singh, H. B , F L Ludwig, and W. B Johnson.
Tropospheric ozone: Concentrations and variabilities m
clean remote atmospheres. Atmos. Environ. /2.2185-
2196, 1978.
128. Siple, G. W., C. K. Fitzsimmons, K F. Zeller, and R. B.
Evans Long range airborne measurements of ozone off
the coast of the northeastern United States In:
International Conference on Photochemical Oxidant
Pollution and Its Control Proceedings. Volume I. B.
Dimitnades, ed. EPA-600/3-77-001 a, US
Environmental Protection Agency, Research Triangle
Park, N.C., January 1977. pp 249-258
129 Spicer, C W,J.L Gemma, D. W. Joseph, P R Sticksel,
and G. F. Ward. The Transport of Oxidant Beyond Urban
Areas. EPA-600/3-76-018a, U.S. Environmental
Protection Agency, Research Triangle Park, N.C.,
February 1976
130. Spicer, C. W., J. L. Gemma, and P. R. Sticksel. The
Transport of Oxidant Beyond Urban Areas Data
Analysis and Predictive Models for the Southern New
England Study, 1975 EPA-600/3-77-041, U.S.
Environmental Protection Agency, Research Triangle
Park, N.C., May 1977.
131. Stair, R The spectral radiant energy from the sun
through varying degrees of smog at Los Angeles. In
Proceedings of the Third National Air Pollution
Symposium. National Air Pollution Symposium, Los
Angeles, Calif., 1955. pp 48-55.
132 Stedman, D. H., and J. O. Jackson. The photostationary
state m photochemical smog. In: Proceedings of the
Symposium on Chemical Kinetics Data for the Upper
and Lower Atmosphere. Int J. Chem. Kinet. Symp.
(1)-493-501, 1975.
133 Stedman, D. H., F. D. Norvis, E E. Daley, H. Niki, and B.
Weinstock The role of OH radicals in photochemical
smog reactions. Paper presented at 160th National
Meeting of American Chemical Society, Chicago,
September 1970.
134. Tiao, G. C , G. E. P. Boxand W J. Hamming Analysisof
Los Angeles photochemical smog data: A statistical
overview. J Air Pollut. Control Assoc. 25-260-268,
1975.
135. Tiao, G. C , M. S Phadke, and G. E. P Box. Some
empirical models for the Los Angeles photochemical
smog data Presented at 68th Annual Meeting, Air
Pollution Control Association, Boston, Mass , June 15-
20, 1975.
136 Vukovich, F Review and analysis. In. international
Conference on Oxidants, 1976—Analysis of Evidence
and Viewpoints. Part V. The Issue of Oxidant Transport.
EPA-600/3-77-117, US. Environmental Protection
Agency, Research Triangle Park, N C , November 1977.
pp. 39-56
137 Westberg, H Review and analysis In: International
Conference on Oxidants, 1976—Analysis of Evidence
and Viewpoints. Part IV. The Issue of national Organic
Emissions EPA-600/3-77-116, U.S. Environmental
Protection Agency, Research Triangle Park, N.C ,
October 1977.
1 38 Westberg, H., K. J. Allwine, and D. Ehas. Vertical ozone
distribution above several urban and adjacent rural
areas across the United States. In: Specialty Conference
on- Ozone/Oxidants—Interactions with the Total
Environment Air Pollution Control Association,
Pittsburgh, Pa., 1976. pp. 84-95.
139 Westberg, H , K. J. Allwine, and E Robinson The
Transport of Oxidant Beyond Urban Areas. Light
Hydrocarbon and Oxidant Data, New England Study
1975 EPA-600/3-78-006, U.S Environmental
Protection Agency, Research Triangle Park, N C.,
January 1978.
140. Westberg, H , and M Holdren Quarterly Report US.
Environmental Protection Agency, Research Triangle
Park, NC, 1976
141. Whitby, R, L. Roland, V. Mohnen, and P. E Coffey.
Measurement of background hydrocarbons in remote
areas. Presented at the Symposium on the Non-Urban
Tropospheric Composition American Geophysical
Union and American Meteorological Society,
Hollywood, Fla , November 10-12, 1976.
142. White, W H, J. A. Anderson, W R Knuth, D. L.
Blumenthal, J. C Hsiung, and R. B. Husar. Midwest
Interstate Sulfur Transformation and Transport Project:
Aerial Measurements of Urban and Power Plant
Plumes, Summer 1974 EPA-600/3-76-110, U.S.
Environmental Protection Agency, Research Triangle
Park, N.C , November 1976
143 White, W. H , D. L. Blumenthal, J. A. Anderson, R. B.
Husar, and W E. Wilson, Jr. Ozone formation m the St.
Louis urban plume. In' International Conference on
Photochemical Oxidant Pollution and Its Control
Proceedings Volume I. B. Dimitriades, ed. EPA-600/3-
63
-------
77-001 a, U.S. Environmental Protection Agency,
Research Triangle Park, N.C., January 1977. pp 237-
247.
144. Whitehead, L W, and R. Severs. Background
hydrocarbon levels in east Texas. Presented at the 83rd
Annual National Meeting, American Institute of
Chemical Engineers, Houston, Texas, March 1977.
Paper no. 21 A.
145. Wolff, G. T , P. J. Lioy, G. D. Wight, R E Meyers, and R.
T. Cederwall. An investigation of long-range transport of
ozone across the midwestern and northeastern U.S. In:
International Conference on Photochemical Oxidant
Pollution and Its Control. Proceedings. Volume I B.
Dimitriades, ed. EPA-600/3-77-001 a, U.S
Environmental Protection Agency, Research Triangle
Park, N.C., January 1977 pp. 307-317
146. Wolff, G. T., P. J Lioy, G. D. Wight, and R. E. Pasceri
Aerial investigation of the ozoneplume phenomenon. J.
Air Pollut. Control Assoc. 27-460-463, 1977.
147. Zafonte, L., P. L. Rieger, and J. R. Holmes. Nitrogen
dioxide photolysis in the Los Angeles atmosphere.
Environ. Sci. Technol 77.483-487, 1977.
64
-------
5. OXIDANT PRECURSORS
INTRODUCTION
As has been established, most of the oxidants
found in urban atmospheres are secondary
pollutants, and the primary pollutants that act as
oxidant precursors are the hydrocarbons and the
nitrogen oxides (i.e., nitric oxide and nitrogen
dioxide, NO and N02>. Not all atmospheric
hydrocarbons are significant oxidant producers.
Methane is the most important exception. In
addition, there are nonhydrocarbon organics (e.g.,
aldehydes) that are also capable of forming
oxidants. However, the term "hydrocarbons," or,
specifically, "nonmethane hydrocarbons," best
describes the organic component of the oxidant
precursor mixture in the atmosphere.
Data on the nature, sources, concentration, and
the concentration variation patterns for the
oxidart precursors are discussed in this chapter.
Emphasis is placed on the organic precursors, as
the KOX data are presented in detail in the
litera ure.22 The quantitative relationships
betwesn oxidant and oxidant precursors are
treatt J in a separate chapter.
AMBIENT LEVELS AND VARIATIONS OF OXI-
/OZONE PRECURSORS
Orgar ic Compounds in Urban Atmospheres
Information on the nature and relative amounts
of the various organic compounds present in
polluted atmospheres is needed to identify
pollutants that have direct, adverse effects on
human health and/or welfare. It is also needed to
provide direct estimates, however rough they may
be, of the levels of controllable and
noncontrollable emissions. Data on the
composition of polluted atmospheres are also
necessary for characterizing in terms of their
oxidant-producing potential the emissions
discharged from various sources. A large part of
the information needed can be obtained through
analysis of the ambient air. The composition of
emissions can also be determined from
measurements at the sources. It is only from the
combined use of such ambient air and source
emission datathatthe impact of organicemissions
on air quality can reliably and comprehensively be
assessed.
Most of the detailed stupes of organic
compounds in the urban atmosphere were
conducted during the 1960's. The results were
summarized and discussed in the criteria
document for hydrocarbons.60 Those studies and
other more recent ones identified a large number
of hydrocarbons (Table 5-1 )4'62'78 and some
aldehydes in the urban atmosphere.8 It is almost
certain that such identification is far from
complete because of the analytical problems
created by the enormous complexity of the
ambient organic mixtures and the trace
concentrations of the various components. The
composition of organics in urban atmospheres
differs from city to city and reflects emission
composition and the influences of chemical
reaction. For example, industrial emissions and
chemical reactions cause higher relative levels of
paraffins and of other nonreactive organics in the
ambient air.45'51 Automobile emissions usually
dominate the composition of organics in urban
ambient atmospheres, especially during the peak
traffic hours.45 Some stationary emission sources,
however, can be important also. More detailed but
indirect information on the composition of
organics likely to be present in urban ambient
atmospheres has been provided by analysis of
source emissions. Analyses of the exhaust and
evaporative emissions from gasoline-fueled
automobiles have identified some 200
hydrocarbons30 and several oxygenated
hydrocarbons,76 many of which could not be
detected in the ambient air, although they were
certain to be present. Data are also available in
varying degrees of detail and specificity on the
composition of organic emissions from diesel-
powered automobiles (Figure 5-1 ),39 from aircraft
(Figure 5-2),21 from natural gas lines, and from
solvent evaporation. The presence of these latter
organic emissions in the ambient air and, in fact,
65
-------
their quantitative contributions to the ambient
organic load have been explored by inves-,
tigators53'79 with some definitive results.
Contributions, for example, have been estimated
by Mayrsohn and Crabtree in their source study
(reconciling atmospheric hydrocarbon data with
known hydrocarbon sources and their emissions)
of the Los Angeles atmosphere.53 Mayrsohn's
conclusions were that (1) automotive exhaust
accounts for somewhat less than 50 percent of the
nonmethane hydrocarbons; (2) gasoline and
gasoline vapor together constitute the second
largest source of atmospheric hydrocarbons,
totaling 30 to 35 percent by weight; (3) commercial
and geogenic natural gas makes up about 20
percent of the total by weight; and (4) other
emissions, such as those from solvent evap-
oration, diesel and aircraft emissions, and
refrigerant and propellent usage are also present
so that the percentage contributions attributed to
the preceding three sources should be regarded as
upper limits.
Probably not all organics emitted by the various
sources will be found intact in the ambient air.
Many may be removed through sink processes
(e.g., deposition on aerosol and ground surfaces
and dissolution in water) before they can
participate substantially in atmospheric reaction
processes. Considering this latter uncertainty and
the limited compositional data available for most
source emissions (e.g., for the exhausts from
gasoline- and diesel-powered automobiles and
from aircraft), one can only assume that current
knowledge of the composition of the organics in
the urban atmosphere is incomplete and,
therefore, that the search for and identification of
unknown organic air pollutants should continue.
TABLE 5-1. HYDROCARBONS IDENTIFIED IN AMBIENT AIR
Carbon
number
Compound
Methane
Ethane
Ethylene
Acetylene
Propane
Propylene
Propadiene
Methylacetylene
Butane
Isobutane
1-Butene
rrans-2-Butene
Isobutene
1,3-Butadiene
Carbon
number
6 (com.;
Compound
3-Methylpentane
2,2-Dimethylbutane
2,3-Dimethylbutane
c/s-2-Hexane
trans-2-Hexane
c/s-3-Hexane
rrans-3-Hexane
2-methyl-1-Pentene
4-methyM-Pentene
4-methyl-2-Pentene
Benzene
Cyclohexane
Methylcyclopentane
2-Methylhexane
3-Metnylhexane
2,3-DimethyIpentane
2,4-DimethyIpentane
Toluene
Pentane
Isopentane
1-Pentene
c's-2-Pentene
frans-2-Pentene
2-methyl-1-Butene
2-methyl-2-Butene
3-methyl-1-Butene
2-methyl-1,3-Butadiene
Cyclopentane
Cyclopentene
Hexane
2-Methylpentane
10
2,2,4-Trimethylpentane
o-Xylene
m-Xylene
p-Xylene
m-Ethyltoluene
p-Ethyltoluene
1,2,4-Tnmethylbenzene
1,3,5-Tnmethylbenzene
sec-Butylbenzene
66
-------
J^
28
24
20
, c 16
Z o
o-e 12
1- S 8
< 8 °
OC -5 4
££ 0
1 2-CYCLE ENGINE
-_ AT IDLE
~
„
„
~I I
-I 1
-I .1
ErL . ill
•
1,
ll
Him
L.
I_
—
_
-I
1
1
3
4-CYCLE ENGINE
AT IDLE
I
i 1
III .
ll. . Illllllllllll...
ga
l£ 32
CJ o 28
31 24
0-70
S « zu
a 16
12
8
4
n
1 2-CYCLE ENGINE
~_ FULL LOAD
"
-
- • •
-I II.
ih.. . ..lllllllin...
__
m.
—
—
I_
1
4-CYCLE ENGINE
FULL LOAD
I
1. . . llllllllin,.
8 12 16 20 24
14 8 12 16 20 24
NUMBER OF CARBON ATOMS PER MOLECULE
Figure 5-1. Distribution of hydrocarbons in diesel exhaust gas.'
Data on total levels of organic compounds in
urban atmospheres are relatively abundant, but
most of these data suffer from considerable
measurement error (see Chapter 7). Some statis-
tics on such data are shown for six urban centers
in Table 5-2,24 Total organic concentrations in
ambient air are usually reported as totals of
nonmethane hydrocarbons and as averages over
the 6- to 9-a,m. period. Methane is excluded
because it is nonreactive and chiefly of natural
origin, and because its concentration usually
overwhelms that of ail other organics combined.
The 6- to 9-a.m. period has been chosen and
adopted in the early efforts by the Federal
government to develop a model relating air quality
to emissions (observational model).22 The
rationale behind that choice was that ambient
concentrations during the early morning hours are
a measure of the strength of the emissions only, as
they are virtually unaffected by the chemical sinks.
Specifically, the 6-to 9-a.m. concentratrons reflect
mainly the vehicular emissions, and such
concentrations are relatively unaffected by
changes in emissions from stationary sources.
Some recent evidence suggests that a given
amount of precursors, when injected into the
reaction system, gradually, throughout the
morning and afternoon, produce essentially the
same ozone concentration as the same amount of
precursors injected during the 6- to 9-a.m.
period.43 Such evidence was obtained from smog
chamber studies and should be verified
The concentration levels listed in Table 5-2 are
typical of those in ordinary urban atmospheres, but
they can be considerably higher. In Los Angeles,
for example, total hydrocarbon concentrations
have been as high as 30 ppm. This level is
equivalent to a nonmethane hydrocarbon total of
greater than 10 ppm.16 In addition to the
nonmethane organics upon which local
contributions are superimposed methane also is
present at a global background concentration of
about 1.4 ppm.10
Seasonal and diurnal patterns of variation in
nonmethane total organic levels have been
discussed in the literature.23'60 Briefly, seasonal
variations have not been well established for lack
of adequate data. In 14 of the 17 California cities
for which adequate data exist, the highest
hydrocarbon concentrations occurred in October
or November. Such consistency is presumably a
consequence of the generally similar meteoro-
logical conditions from year to year along the
California coast. Cities elsewhere would be
expected to show other patterns, depending on
their particular meteorology. Diurnat patterns
67
-------
observed in many cities show two distinct daily peak in the afternoon hours. Both peaks pre-
peaks, one during 6- to9-a.m.,andanotherbroader sumably reflect the automobile traffic and local
meteorological dispersion patterns.
30
20
10
0
FUEL-JP4
,l
III....
12
16 19+
30
20
10
EXHAUST
IDLE
(TOTAL BY FID 686 ppm C)
I I I I I I I I I
30
CO
Z 20
O
ffi
10
u
O
§
I
12
16 19+
APPROACH
(TOTAL BY FID <10 ppm C)
... I I I I I I I I I I I I
12
16 19+
30
20
10
n
[" CRUISE
*~
_
f . . . 1 . .
(TOTAL BY FID
|
| ,
< ppm C)
30
20
10
0
TAKE-OFF
(TOTAL BY FID <10 ppm C)
8 12
CARBON NUMBER
16
19+
Figure 5-2. Distribution of hydrocarbons in jet-aircraft engine exhaust (determined by gas
chromatograph with flame ionization detector).21
68
-------
TABLE 5-2. FREQUENCY DISTRIBUTIONS FOR 6-TO 9-a.m. NONMETHANE HYDROCARBON CONCENTRATIONS
AT CAMP SITES. 1967-72"
Ippm C)
Site
Number
of
samples Mm
PBrcemile"
10
20
30
40
50
60
70
8O
90 Max
Arithmetic
Sid
Mean dev.
Geometric
Mean
Sid
dav
Denver Colo
(060680002A1 0)
67b
68"
69"
70b
71b
72
29 00
161 00
219 0,0
231 0.0
178 0.0
282 0.0
00 0.3
0.2 0.4
0.0 0.1
0.0 0.2
0.3 0.6
0.1 0.4
0.4 0 6
0.6 08
0.3 0.5
0.3 0.5
0.7 0.9
0.5 0.7
09 11
1 0 1.1
0.7 0.8
0.6 0.8
1.1 1.3
0.9 1.1
1 8
1 4
1.0
1.0
1 5
1.3
13 27
.7 27
.3 1.7
.1 1.6
.8 2.4
.6 2.0
52 1.3
5.3 1.2
4.9 0.8
5.9 0.8
5.7 1 .3
3.9 1 .0
1.2
1.0
0.8
0.7
0.9
07
0.74
0.87
0.50
0.52
0.99
0.74
3.69
3.62
3.26
2.96
2.29
2.73
Washington, D.C.
(090020002A1 0)
Chicago, III.
(141220002A10)
Philadelphia, Pa.
(397140002A10)
Cincinnati, Ohio
St. Louis, Mo.
(264280002A10)
66b
68b
69"
70b
71
72"
68"
69
70"
71"
66"
67"
68b
69b
70"
71"
72"
68"
69b
70"
71B
72b
68b
69
70"
71
72"
250
244
231
267
279
187
146
300
220
269
142
260
154
192
264
234
59
109
188
99
104
238
198
292
232
314
235
0.0
0.0
on
0.0
on
on
on
on
on
on
0.0
0.0
on
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
00
0.0
nn
00
n 1
0.0
00
00
0.0
0.0
nn
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.1
0.1
0.0
0.0
0.0
0.0
0.0
0.1
0.1
0.0
0.0
0.0
0.0
0.2
0.1
0.0
0.2
0.3
0.1
0.0
0.1
0.1
0.0
0.0
0.0
0.0
0.0
0.2
0.2
0.0
0.1
0.0
0.0
0.0
0.2
0.1
nn
0.0
nn
nn
n4
n.i
0.0
03
0.5
0.2
n i
0.2
0.2
0.1
0.1
0.1
0.2
0.0
0.3
0.3
0.2
0.2
0.1
0.0
0.0
0.3
0.3
00
0.0
0 1
0 1
0.6
0.5
00
04
0.6
0.3
0 2
0.3
0.3
0.2
0.3
0.2
0.2
0.0
0.4
0.3
0.4
0.3
0.2
0.0
0,1
0.4
0.4
00
0.0
01
0 1
0.8
07
00
05
0.7
0.5
03
0.5
0.4
0.2
0.4
0.3
0.4
0.0
0.5
0.4
0.6
0.4
0.3
0.1
0.2
0.5
0.5
0 1
0.1
0.2
03
1.0
08
0 2
06
0.8
0.6
05
0.6
0.6
0.4
0.4
0.4
0.4
0.0
0.6
0.4
0.8
0.5
0.4
0.2
0.2
0.6
0.6
0.3
0.3
0.3
05
1.1
1.1
05
08
1.1
0.8
0.6
0.7
0.7
0.4
0.6
0.6
0.5
0.0
0.7
0.5
0.9
0.7
0.6
0.3
0.3
0.7
0.8
03 (
0.5 (
0.5 (
0 6
1 4
1 4
08
1 0
1.4
1.1
09
0,9
1.0
0.6 (
0.7
0.8
0.7
0.1 (
0.9
0.7
1.1
1.0
0.7
0.4 (
0.4 (
.0 1.8
.1 2.6
35 23
).8 2.0
3.8 3,3
1 20
.8 2.8
8 36
4 4 1
4 27
.7 4.2
.6 5.9
6 37
.3 3.7
.3 2.3
3.9 3.0
.1 2.4
.4 3.4
.1 3.1
3.4 2.3
.1 2.0
.0 3.4
.6 4.4
.4 3.4
.1 2.9
3.7 2.5
3.7 1.8
0.5
0.5
0,2
0,3
0.3
0.4
0.8
0.8
05
0.6
0,9
0.7
06
0.6
0.5
0.4
0.5
0.5
0.5
0.1
0.6
0.5
0.7
0.5
0.4
0.2
0.3
0.3
0.4
0.2
0,3
0.3
04
0.6
07
07
05
0.7
0.7
0 7
0.5
0.4
0.4
0.4
0.6
0.5
0.3
0.4
0.4
0.7
0.5
0.4
0.3
0,2
0.35
0.35
n 16
0.19
020
n 25
n64
n54
n25
046
0.64
0.42
n 34
0.42
0.37
0.27
0.31
0.31
0.32
0.23
0.45
0.40
0.44
0.37
0.31
0.19
0.21
2
2
2
2
2
2
2
2
3
2
2
3
3
2
2
2
3
3
2
2
2
2
3
2
2
2
2
56
73
27
62
64
82
58
99
20
58
70
11
17
74
90
74
02
18
79
22
37
09
25
83
72
54
42
"Concentrations greater than or equal to specified value in indicated percentage of samples
Nearly standard exceeded
Since hydrocarbon emission controls went into
effect in the 1960's, ambient air hydrocarbon
concentrations have been observed to trend
downward with time in some urban areas82 but not
in others.80 Figure 5-3 shows such a downward
trend for the Los Angeles atmosphere.40 The trend
is probably a real one, even though the
nonmethane hydrocarbon data used to construct
these trend plots were obtained indirectly (that is,
they were computed from total hydrocarbon data
and data on established relationships between
total and nonmethane hydrocarbon for the Los
Angeles atmosphere). Such data, as already
stated, reflect control of automotive emissions
only. In areas where substantial controls have
been applied to stationary sources also, the impact
of these latter controls on daily maximum or 6- to
9-a.m. hydrocarbon concentrations will not be
detectable; however, a detectable impact on the
24-hr average hydrocarbon concentrations may be
observed. In general, the analysis and inter-
pretation of ambient air hydrocarbon data to
assess the impact of emission controls should be
done with great care to avoid biases and, hence,
misleading results. Such biases can be introduced,
for example, by ambient air measurement errors,
by inappropriate definition of the ambient
concentration, and by the disproportionate in-
fluence of a single emission source on meas-
urements made at an improperly located
measurement station.
69
-------
go
8*
5s
is
o
MAX-HOUR AVERAGES - ALL DAYS OF YEAR
MAX-HOUR AVERAGES - JULY, AUG., SEPT.
6-to-9-a.m. AVERAGES - JULY, AUG., SEPT.
1963
1964
1965
1966
1967
YEAR
1968
1969
1970
1971
1972
Figure 5-3. Nonmethane hydrocarbon trends in Los Angeles, 1963-1972."
Organic Compounds in Rural and Remote Areas
Occurrences of ambient organic compounds in
rural and remote areas are of interest only to the
extent that they constitute evidence on the role of
natural organics in the ambient air oxidant
problem.28 Since such evidence has already been
discussed (in the chapter on the natural sources of
oxidant/ozone) and the conclusion was reached
that the terpenes and methane have no important
role, the information to be presented in this section
will serve merely to provide additional detail and
documentation for that previous discussion and
conclusion. Specifically, more complete and
detailed evidence will be presented here on the
occurrence of naturally emitted organics in rural
and remote areas, but with the clear under-
standing that such evidence relates to occurrence
alone and not to the impact of such organics on
oxidant air quality.
Besides the ubiquitous anthropogenic pol-
lutants, rural atmospheres also should contain
natural organics such as methane and organics
emitted by vegetation. Several of these latter
organics have been identified either directly,
through ambient air analysis, or indirectly, through
analysis of the material emitted by vegetation
samples processed in the laboratory.69 Organic
compounds found by Rasmussen to be emitted by
vegetation are listed in Table 5-3.68 This list is by no
means complete, since not all types of vegetation
were examined. Studies reported in the Russian
literature indicate that a considerable number of
plant species release low molecular weight
hydrocarbons ar^d aldehydes and a wide variety of
essential oil components.41
TABLE 5-3. VOLATILE PLANT PRODUCTS IDENTIFIED
BY RASMUSSEN68
Natural organic compounds
cf-Pmene
/5-PinenB
Myrcene
D-Limonene
Santene
Camphene
n-Heptane
Isoprene
a-lonone
0-lonone
cr-lrone
Aside from the organics clearly associated with
emissions from vegetation, measurements ,in
remote areas have also shown the presence of
organics that are commonly, but apparently not
exclusively, associated with anthropogenic
activities. For example, at Point Barrow, Alaska,
Cavanagh et al. found benzene, pentane, butane,
ethane, ethylene, acetaldehyde, and acetone at
sub-ppb levels (acetone at about 100 ppb) and
attributed their origin to biological processes.19
70
-------
A few data on total organic concentrations in
rural and remote atmospheres are available, but
here again, as in the case of the urban
atmospheres, most of the existing data are
unreliable. One cause of unreliability is the
relatively large error associated with the
measurement of total non-methane organics. An
additional problem, however, with the rural and
remote atmospheric data arises from the lack of
evidence on the composition of the organic
pollutant mixture and, specifically, on the relative
levels of vegetation-related and anthropogenic
pollutants. This lack in compositional information
has two consequences. First, it does not permit
proper calibration of the ionization detector of the
hydrocarbon analyzer, resulting in additional
measurement error. Second, it attributes
erroneously high concentration to the natural
organics, since corrections for the anthropogenic
components are not possible. Probably it is partly
because of such errors that some of the reported
values for total natural organic concentrations
have been as high as 5 to 10 ppm C.85 Extensive
measurements in rural areas by EPA and other
investigators using sampling and analysis
procedures carefully designed to minimize such
errors did not show the total of natural organics to
be more than 0.1 ppm C.69'84 Whenever and
wherever concentrations were higher, there was
invariably evidence of anthropogenic con-
tamination.50'72 Finally, though it is probable that
there are natural organic pollutants as yet
unidentified, there are no strong indications that
such unknowns could drastically change the
current view regarding the role of natural organics
in the oxidant/ozone problem. Such a role does not
appear to be an important one, as was concluded
also from the discussions in the preceding chapter
dealing with the natural sources of ambient air
oxidants.
Nitrogen Oxides
Information on the sources, concentrations, and
variations in concentrations of nitrogen oxides has
been reported elsewhere in detail.2'26 The
information presented here is limited to data on
ambient concentrations in urban and rural/remote
atmospheres and is intended merely to
complement the information on oxidant and
oxidant precursor concentrations.
Tables 5-4 and 5-5 include data on ambient NO
and NOa concentrations in seven urban centers
outside California.24 Data from selected California
sites, where concentrations levels are generally
higher, are shown in Tables 5-6 and 5-7.2 Such
data show a slight upward trend with time, as
illustrated in Figure 5-4.16
Concentrations of NO and NO2 in rural and
remote areas are extremely low, the average levels
being at or below the limit of detectability for
current monitoring instruments. Thus, early
measurements of the NOa background con-
centration showed levels generally below 5
ppb.43'49 More recent measurements of NO and
NO2 in rural areas in Ohio, Maryland, and
Pennsylvania indicated NO and NO2 levels around
5 ppb, with a few excursions to as high as 12
ppb.71'72 Furthermore, it appears that the
concentration of NOX in an urban air mass decays
rapidly as the air mass moves away from the city.
This is illustrated by ambient NOX data taken within
and outside Dayton, Ohio. Such data showed the
city core levels of NOX to range from 75 ppb to 450
ppb, whereas the concentrations at a site 18 miles
from the downtown area ranged only from 8 to 80
ppb.77
The extremely low levels of NO in the rural and
remote areas or, alternatively, the lack of
adequately sensitive analytical instruments, is a
problem in that it limits the conclusiveness of the
field studies on oxidant. Thus at present, it is not
possible to establish reliably the levels of natural
NOX and hence to estimate the relative
contributions of the natural and manmade sources
to rural NOX. This uncertainty, in view of the
important role of the NO» factor in rural oxidant
formation (see Chapter 4) makes it difficult to
delineate completely and clearly the relative roles
of anthropogenic hydrocarbon and NOX as
precursors in the atmospheric oxidant/ozone
formation process.
SOURCES OF OXIDANT PRECURSORS
Introduction
Hydrocarbons and nitrogen oxides are emitted
into the atmosphere from both natural and
manmade sources, with the contribution of natural
sources being the greater. However, the larger
contribution by the natural sources does not
significantly influence the oxidant problem, since
the natural and anthropogenic emissions are
geographically segregated, with the anthro-
pogenic ones concentrated in the populated areas.
Natural hydrocarbon emissions arise mostly from
biological processes, trees, and localized sources
such as seepage from natural gas and oil fields. Of
the anthropogenic sources, combustion of fuels is
71
-------
TABLE 5-4. FREQUENCY DISTRIBUTION DATA FOR 6-TO 9-a.m. NITRIC OXIDE CONCENTRATIONS AT CAMP SITES,
1962-72"
Stte
Ssn Francisco,
Calif.
(056860002A10)
Denver, Colo.
(06058000 2A10)
Washington, D.C.
(090020002 A 10}
Washington, D.C.
(090020003A10)
Chicago, III.
(141220002A10)
New Orleans, La,
(192020OO4A10)
bi. LOUIS, Mo
(264280002A10)
Cincinnati, Ohio
(361220003A10)
Year
62b
63"
64"
65s
66
67B
68"
69
70"
71"
72"
62
63
64
65
66
67
68"
69"
70
71
72"
62
63
64
65
66
67
68
69
70
71b
72
62"
63
64"
65
66
67
68"
69
70b
71
72"
62
63"
64
65
66"
67"
68"
69"
70"
71b
72"
Number
cf
samples
112
179
13
223
297
191
192
285
252
184
241
289
279
310
274
310
304
270
269
299
308
219
283
315
318
317
286
311
319
318
320
26S
321
259
325
259
333
294
326
256
322
228
311
234
286
273
334
310
206
188
136
196
129
215
251
Mm. 10
0.00 0.00
0.00 0.00
0.01 0.01
0.00 0.01
0.00 0.01
0.00 0.01
boo 001
0.00 0.01
0.00 0.02
0.00 0.02
0.01 0,02
0.00 0.01
0.00 0.00
000 0.00
0.00 0.00
0.00 0.01
0.00 0.01
0.00 0.01
0.00 0.01
0.00 0.01
0.00 0.01
0.00 0,00
0.00 0.04
0.00 0,04
0.00 0.04
0.00 0.04
0.00 0,04
0.00 0.03
0.00 0.03
0.01 0.04
0.00 0.04
0.00 0.04
0.01 0.04
000 001
0.00 0.00
0.00 0.01
0.00 0.00
0.00 0,00
0.00 0.01
0.00 0.01
0.00 0.00
0,00 0.01
0.00 0.01
0.00 0.01
0.00 0.01
0.00 0.01
0.00 0,01
0.00 0.00
0 00 0.01
0.00 0.01
0.00 0.01
0.00 0.00
0.00 0,00
0.00 0.01
0.00 0.01
20 30
0.00 0.01
0.00 001
0.02 0.02
002 0.03
0.02 0.03
0.02 0.03
0 02 0 02
0.02 0.03
0.03 0.04
0.03 0.04
0.04 0.05
0.01 0.02
0.01 0.02
001 0,01
0.01 0,02
0.01 0.02
0.02 0,03
0.02 0.03
0,02 0.03
0,02 0.03
002 003
0.02 0,03
0.05 0.07
0,06 0,08
0.06 0.07
0.06 0.08
0 07 0 09
0.05 0,07
0.04 0.06
0.06 0.09
0.08 0.10
0,07 0.09
0,07 0,10
0 01 0 02
0.00 0.01
0.02 0.04
0.01 0.01
0.01 0.02
0.01 0.02
0.01 0.02
0.01 0,02
0.02 0.02
0.02 0.03
0.02 0.03
0.01 0,02
0.01 0,02
0.01 0.01
0.01 0.01
0.01 0.12
0.02 0.02
0,03 0.04
0.01 0.02
0.01 0.02
002 0.03
0,02 0.02
Percentll
40 60
0.01 002
0.02 002
0,03 0.04
0.03 004
0,04 0.05
0.04 0,05
0 04 005
0 04 005
0,06 0.07
0 05 0 07
0.07 0,08
0.02 0.03
0.02 0.04
002 003
0.02 0.03
0.03 0.04
0.04 005
0.03 004
0.04 0,05
0,04 0,06
0 04 004
0.04 0.06
0.09 0.11
0.09 0.11
0.09 0.11
0.10 0.12
011 013
0.08 0.10
0.07 0.09
0.10 0.13
0.13 0.15
0.11 0.13
0.13 0.15
003 004
0.01 0,01
0,04 0.06
0.02 0.03
0.03 004
0.03 0.04
0.03 0.04
0.03 0.04
0.04 0.06
0.05 0,16
0.05 0.07
0.02 0.03
0.02 0.03
0.02 0.03
0.02 0.02
0.03 0.04
0.02 0.03
0.05 0.07
0.03 0.04
0.03 0.03
0.03 0.04
0.03 0.04
Arithmetic
«'
60 70
0 02 0 03
003 005
0.04 0.05
0 05 006
0.06 0.08
0.06 0.07
006 007
007 009
0,09 0.11
008 0 10
0.10 0.12
0.04 0.05
0.05 0.06
0 04 005
0.04 0.06
0.04 0.05
0.06 008
0.05 0.06
0.06 0.07
0.07 0,09
0 05 0 07
0,08 0.1 1
0.13 0.16
0.12 0.13
0,13 0.15
0.13 0.15
015 0.1 7
0.12 0.13
0.10 0.12
0.15 0.18
0.19 0.24
0,16 0.20
0.18 0.21
0 05 0 07
0.02 0.02
0.07 0.08
0.04 0,05
0.05 006
0.05 0.07
0.05 0.06
0.05 0.07
0.08 0.1 1
0.08 0.10
0.10 0,12
0,04 0.06
0.04 0.05
0.04 0.05
0.03 0.05
0.05 0.06
0.04 0.05
0.10 0.14
0,05 0.07
0.04 0.05
0.05 0.07
0.05 0.07
80 90
0 04 0 06
0.06 008
0.06 0.20
0 08 011
0.11 0.15
0.09 0.13
0 10 004
012 016
0.13 0.16
012 016
0.15 0.22
0.06 0.10
0.08 0.13
0 08 014
0.08 0.12
0,07 0,12
0.10 0.16
0.08 0.13
0.08 0.14
0.11 0.06
0 09 013
0.13 0.15
0.19 0.21
0.16 0.20
0,17 0.24
0.17 0,21
0.19 023
0.15 0.18
0.15 0.19
0.21 0.26
0.28 0.33
0.22 0.27
0.24 0.29
0 09 013
0.03 0.04
0.09 0,12
0.06 008
0.08 0.12
0.08 0.1 1
0.08 0.1 1
0.09 0 1 2
0.14 0.17
0.12 0.15
0.14 0.19
0.08 0.12
0.0& 0.1 2
0.09 0.13
0.07 0.11
0,09 0.14
0.08 0.13
018 0.29
0.10 0.18
0.08 0.13
0.10 0.20
0.09 0.17
Max Mean
0 13 0.02
0 32 0.03
0.29 0.07
031 005
0,30 0.06
0.32 0.06
027 006
045 0.07
0.42 0.08
0 40 0.08
0.44 0.10
0.32 0.04
0.40 0.06
0 66 0.05
0.38 0.05
0.53 0.05
0 74 0.07
0.49 0.06
0,62 0.07
0.54 0.08
0 54 0.06
0.41 0.08
0.56 0.12
0.38 0.11
0.71 0.13
0.43 0.12
0 51 0.13
0.43 0.10
0.37 0.10
1.02 0.15
0.65 0.18
0.51 0.15
053 0.16
0 36 0.06
0.24 0.02
0.44 0.06
0.19 0.04
0.49 0.05
0.26 0.05
0.28 0.05
0.51 0.05
047 0.08
0.36 0.07
0.40 0.09
0.42 0.05
0.26 0.05
0.45 0,05
0.43 0.04
0.67 0.06
0.53 0.06
0,69 0.12
0.56 0.07
0.52 0,05
0.48 0.07
0.47 0.06
Sid
dev
0.02
0.03
0.08
004
0.06
0.05
005
0.07
0.06
0.07
0.08
0.04
0.06
0.07
0.06
0.06
0.08
0.06
0.08
0.08
0.06
0.08
0.08
006
0.09
0.07
0.08
0.06
0,06
0,11
0.11
0.09
0.09
005
0.12
0.04
0.03
0.05
0.04
0.04
0.06
0.07
0.06
0.07
0.05
0.04
0.07
0,05
0.07
0.08
0.14
0.09
0,07
0.07
0.07
Geometric
Mean
0.02
0.02
0.04
004
0.05
0.05
004
0.05
0.06
006
0.08
0.03
0,04
0.03
0.03
0.04
0.05
0.04
0.05
0.05
0.04
0.05
0.10
0.10
0.10
0.11
0,12
0.09
0.08
0,12
0.14
0,12
0.13
0.04
0,02
0,05
0.03
0.04
0.04
0.04
0.03
0.05
0.06
0.06
003
0,03
0.03
0.03
0.04
0.04
0.07
0.04
0.03
0.05
0,04
Std
dev
? 51
? 98
2.63
? 14
2,61
2.31
? 71
? 76
2,52
? 38
2.32
2.53
2.82
3 11
2,83
2.66
2.74
2.56
2.73
2.80
? 52
3.10
2.14
1,99
2.06
1.98
1 99
2.16
2.13
2.03
2,31
2.08
2,10
? 77
2.45
2.22
2.70
2.90
2.68
2.80
3.10
2.92
2.47
2.48
2.74
2.52
2.90
2.86
2.73
2.62
3.03
3.04
2.96
2.64
2.64
72
-------
TABLE 5-4 FREQUENCY DISTRIBUTION DATA FOR 6-TO 9-a.m. NITRIC OXIDE CONCENTRATIONS AT CAMP SITES,
1962-72" (cont'd).
(ppm)
Arithmetic
Sue
Philadelphia, Pa.
Year
62b
63b
64
65
66
67
68"
69
70
71
Number
of
samples
239
233
302
290
336
296
266
276
305
278
Mm 10
0.00 0 00
0.00 0.00
0.00 0.01
0.00 0.01
0.00 0.01
0.00 0.02
000 001
0.00 0.01
0.00 0.01
0.00 0.01
20 30
0 00 0.01
0.01 0.02
0.02 0.03
0.02 0.02
0.02 0.03
0.03 0.04
002 0.03
0.12 0.03
0.03 0.04
0.02 0.02
Percent!
40 50
0.01 0.01
0.03 0.04
0.03 0.04
0.04 0.04
0.04 O.OB
0.05 0.07
004 005
0.04 0.05
0.05 0.06
0.03 0.04
le"
60 70
0.02 0.02
0.05 0.07
006 0.07
0.06 0.07
0.06 0.08
0.08 0.10
0.07 0.08
0.06 007
0.08 0.10
0.05 007
80 90
0.03 0.06
0.10 0.15
0.10 0.14
0.10 0.16
0.10 016
0.13 0.17
012 017
0.09 0.12
0.12 0.16
0.09 0.13
Max Mean
0 23 0 02
0.79 0.07
0.42 0.06
0.64 0.07
0.79 0.07
0.71 0.09
0 55 0.08
0.38 0.06
0.81 0.08
0.42 0.06
Std
dev
0.02
0.10
0.06
0.07
0.08
0.08
0.08
0.05
0.09
0.06
Geometric
Mean
0.02
0.04
0.05
0.05
0.05
0.07
0.05
0.05
006
0.04
Std
dev
2.54
3.30
2.61
2.65
2.50
238
2.60
2.48
2.47
2.78
'Concentrations greater than or equal to specified value in indicated percentage of samples
"Yearly standard exceeded
TABLE 55. FREQUENCY DISTRIBUTION DATA FOR 6-TO 9-a.m. NITROGEN DIOXIDE CONCENTRATIONS
AT CAMP SITES, 1962-7224
(ppm)
Site
San Francisco,
Calif.
(056860002A10)
Denver, Colo.
(060580002A10)
Washington, D.C.
(090020002A10)
Washington, D.C.
(090020003A10)
Chicago, III.
(141220002A10)
New Orleans, La.
(192020004A10)
Year
62b
63"
64"
65"
66
67"
68"
69
70
71"
72"
62
63
64
65
66
67
6ft"
69"
70
71
72"
62
63
64
65
66
67
68
69
70
71
72
62
63
Number
of
samples
142
155
10
247
324
180
226
283
282
171
253
313
312
308
336
275
305
237
254
298
318
214
293
318
303
330
304
290
312
330
329
285
336
288
289
Arrthmetfc
Geometric
Percentile"
MID 10
0.00 0.00
0.00 0.01
0.01 0.01
0.01 0.02
0.00 0.02
0.00 0.02
0.00 0.02
0.00 0.01
0,00 0.02
0.00 0.02
0.00 0.01
0.00 0.01
0.00 0.01
0.01 0.02
0.00 0.02
0.00 0.02
0.00 0.02
0.00 0.02
0.00 0.02
0.00 0.02
0.01 0.02
0.00 0.00
0.00 0.02
0.00 0.02
0.00 0.02
0.01 0.02
0.00 0.03
0.00 0.02
0.01 0.03
0.00 0.03
0.00 0.03
0.02 0.03
0.00 0,03
0.00 0.02
0.00 0.00
20 30
0.00 0.01
0.02 0.02
0.02 0.03
0.03 0.03
003 003
0.02 0.03
0.02 0.03
0.02 003
0.02 0.03
0.02 0.03
0.02 0.03
0.02 0.02
0.02 0.02
0.02 0.03
0.02 003
0.02 0.02
0.02 0.03
0.03 003
0.02 0,03
0.03 0.04
0.02 0.03
0.02 003
0.02 0.03
0.03 0,03
0.03 0.03
0.03 O.O3
0.03 O.O4
0.03 0.03
0.03 0.03
0.03 0.03
0.04 0.04
0.03 0,04
0.03 0.04
0.02 0.03
0.01 0.01
40
0.01
0.03
0.04
0.03
003
0.03
0.03
0.03
0.03
0.03
0.03
0.02
0.03
0.03
0.03
0.03
0.03
0.04
0.03
0.04
0.03
0.03
0.03
003
0.03
0.03
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.03
0.01
50 60
0.01 0.02
0.03 0.04
0.04 0.04
0.04 0.04
004 004
004 0.04
0.04 0.04
003 0.04
0.04 0.04
003 0.04
0.04 0.04
0.03 0.03
0.03 0.03
0.03 0.04
0.03 0.03
0.03 0.04
0.04 0.04
0.04 0.05
0.04 0.04
0.05 0.06
0.03 0.04
0.04 0.04
0.03 0.04
0.03 0.04
0.04 0.04
0,04 0.04
0.05 0.05
0.04 0.05
0.04 0.05
0.04 0.05
0,05 0.05
0.04 0.05
0.04 0.05
0.03 0.04
0.01 0.02
70 80
0.03 0.03
0.04 0.04
0.04 0.05
005 0.05
004 005
0.04 0.05
005 0.05
004 005
005 0.05
0,04 0.05
005 0.06
0,03 0.04
0.04 0.04
0,04 0.05
0.04 0.05
0.04 0.05
0.05 0.05
0.06 0.06
O.O4 0.05
0.06 0.07
0.04 0.04
0.05 006
0.04 0.05
0.04 0.05
0.05 0.06
0.04 0.05
0.06 0.07
0.05 0.06
0.05 0.05
0,05 0.06
0.06 0.07
0.06 0.07
0.06 0.06
0.04 0,05
0.02 0,02
90 Max
0,04 0.09
0.06 0,19
0.05 0.06
0.06 0.09
005 010
0.06 0.09
0.06 0.1 1
005 015
0.07 0.13
0.06 0.10
0.07 0,12
0.05 0.16
0.06 0.13
0.06 0.15
0.05 0.18
0.05 0.13
0.06 0.14
0.08 0.18
0.06 0.11
0.09 0.16
0,05 0.1 1
0.09 0,40
0,07 0.15
0.06 0.11
0.07 0.16
0.06 0,11
0.09 0.23
0.07 0.17
0.06 0.10
0.07 0.17
0.08 0.14
0.08 0.39
0.08 0.16
0.07 0.14
0.03 0.05
Std
Mean dev
0.02 0.01
0.03 002
0.04 0.01
0.04 001
004 001
0.04 0.01
0.04 0.01
003 0.01
0.04 0.02
0.03 0.01
0.04 0.02
0.03 0.01
0.03 0.02
0.03 0.01
0.03 0.01
0.03 0.01
0.04 0.02
0.05 0 02
0.04 0.01
0.05 0.02
003 0.01
0.05 0,06
0.04 0.02
0.04 0.01
0.04 0.02
0.04 0.01
0.05 0.02
0.04 0.02
0.04 0,01
0,04 0,02
0.05 0.02
0.06 0.05
0.05 0.02
0.04 0.02
0.01 0,00
Std
Mean dev
0.02 2.23
0.03 1 .80
0.04 1 .54
0.04 1 49
004 1 47
0.04 1.57
0.04
0.03
0.04
0.03
0.04
0.03
0.03
004
0.03
003
0.04
0.05
0.04
0.05
0.04
64
.63
.76
.66
.78
.76
.75
.63
.55
.61
,65
56
.56
,84
.50
0.04 2.55
0.04 1 77
0.04 1 .49
0.04 1 .57
0.04 1 .43
0.05 1 .56
0.04 .62
0.04 .45
0.05 .51
0.05 .40
0.05 .67
0.05 .60
0,04 1 .79
0.02 1 ,86
73
-------
TABLE 5-5. FREQUENCY DISTRIBUTION DATA FOR 6-TO 9-a.m. NITROGEN DIOXIDE CONCENTRATIONS
AT CAMP SITES, 1962-7224 (cont'd).
(ppm)
Arithmetic
Geometric
Site Year
St. Louis, Mo. 64"
(264280002A10) 65
66
67
68
69°
70"
71
72"
Cincinnati, Ohio 62
(36122000A10) 63
64
65
66"
67b
68
69b
70"
71"
72"
Philadelphia, Pa 62"
(397140002A10) 63
64
65
66
Number
of
samples
255
348
322
352
322
258
233
309
231
292
304
336
313
238
223
280
231
129
236
227
211
286
281
286
340
Mm
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.01
000
0.00
0.00
0.01
0.00
0.00
onn
0.00
0.01
0.00
000
0,00
0,00
0.00
0.00
0.00
10 20
0.01 002
0.01 0.01
0.01 0.02
0.01 0.01
0.00 0.01
0 00 0.01
0.01 0.02
0,02 0,02
0.02 002
0.01 0.02
0.01 0.02
0.01 0.02
002 0.02
002 0.02
0.01 0.02
0.02 0.02
0.02 0 02
0.02 0.03
0,01 0.02
0.02 0.03
0.00 0.01
0.01 0.02
0.02 0.02
0.02 002
0.02 0.02
30
0.02
0.01
0.02
0.01
0.01
0.02
0.02
0.02
0.03
0.02
0.02
0.02
0.02
0.02
0,02
no?
0.02
0.03
0.02
0.03
0.01
0.02
0.03
0.02
0.02
Percent*!
40 50
0.03 0.03
0.02 0.02
0.02 0.03
0,02 0.02
0.02 0.02
0.02 0.02
0.02 0,02
0.02 0.03
0.03 0,04
0 02 0.03
0.02 0.03
0.02 0.03
0.03 0.03
0.03 0,03
0.02 0.03
0.03 0.03
0.02 0.03
0.03 0.03
0.02 0.03
0.04 0.04
0.02 0.02
0.03 0.03
0.03 0.03
0.03 0.03
0.03 003
e*
60 70
004 0,04
0.02 0.02
0.03 0.04
0.02 0.03
0.02 0.02
003 0.03
003 0.03
003 0.04
0,05 006
0.03 0.03
0.03 0.03
0,03 0.03
0.03 0.04
0.04 0.04
003 0.03
003 004
0.03 0.03
0.04 0.04
0.04 0.04
0.04 0.05
0.02 0.03
0.04 0.04
0.04 0.04
0.03 0.04
0.04 0.04
80 90
0.05 0.06
0.03 0.04
0.04 0.06
0.03 0.03
0.03 0.04
0.04 0.05
0.04 0.05
005 0.06
0.07 0.08
0.04 0.05
0.04 0.05
0.04 0.06
0.04 0.05
0.05 0.06
0.03 0.04
004 005
0.04 0.05
0.05 0.07
0.05 0.07
0,05 0.06
0.03 0.04
0.05 0.07
0.05 0.07
0,05 0.05
0.05 0.06
Max Mean
0.16 0.03
0.08 0.02
0.12 0.03
0.06 0.02
0.11 0.02
0.11 0.02
0.10 0.02
0.11 0.03
0.16 0.04
0.13 0.03
0.13 0.03
0.15 0.03
0.11 0.03
0.10 0.03
0.09 0.02
0 09 0 03
0.10 0.03
0.12 0.04
0.11 0.03
0.13 0.04
0,10 0.02
0,15 0.03
0.20 0.04
0.12 0.03
0.11 0.03
Std
dev
0.02
0.01
0.01
0.01
0.01
0.01
0.01
0.01
002
001
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.02
0.02
0.01
0.01
0.02
0.02
0.01
0.01
Mean
0,03
0.02
0.03
0.02
0.02
002
0.03
0.03
004
0.03
0.03
0.03
0.03
0.03
0.03
003
0.03
0.04
0.03
0.04
0.02
0.03
0.04
0.03
0.03
Std
dev
1.85
1.80
1.80
1.83
1.94
2.11
1.66
1.61
1.66
1.82
1.71
1.77
1.45
.69
.52
.59
.56
.53
2.00
1.64
1.56
1.97
1.72
1.58
1.66
"Concentrations greater than or equal to specified value in indicated percentage of samples
Nearly standard exceeded
by far the most important source of hydrocarbon
and nitrogen oxide emissions. In addition,
hydrocarbon and nonhydrocarbon organic
emissions also arise from the use of such organics
in the processing of raw materials.
From a pollution control standpoint, the
distinction made between natural and manmade
sources and the use that is made of annual
emission rate data are not entirely satisfactory. For
example, anthropogenic emissions such as those
from leaking fuel lines, home appliances, etc.
cannot be subjected to systematic control because
of their accidental nature. Thus from a control
standpoint, such emissions may be considered to
be natural. To generalize, classification of sources
and emissions as controllable and uncontrollable
may be more appropriate than the distinction now
in use. Also, emission rate data reflecting annual
or even daily averages are inadequate in that they
mask seasonal and diurnal variations in emission
rates. Considering that only the morning-to-noon
emissions in the summer-to-fall season are of
main photochemical consequence, it is evident
that ignoring the seasonal and diurnal variations in
emission rates does not permit equitable
assessment of the various emission sources.
The sections that follow deal with the main
anthropogenic and natural emissions; emission
rate data are presented in the form in which they
are available—mostly as daily and yearly averages.
74
-------
TABLE 5-6. NITRIC OXIDE CONCENTRATION" IN CALIFORNIA BY AVERAGING TIME AND FREQUENCY.
1963-67*
(ppm)
Place, sue no .
averaging
time
Anaheim-176:
1 hr
8 hr
1 day
1 mo
1 yr
Oakland-327;
1 hr
8 hr
1 day
1 mo
1 Vr
Riverside-126:
1 hr
8 hr
1 day
1 yr
Sacramento-276:
1 hr
8 hr
1 day
1 mo
1 Yr
San Bernardino-151:
1 hr
8 hr
1 tlay
1 mo
1 yr
San Diego-101:
1 hr
8 hr
1 day
1 mo
1 yr
Stockton-252;
1 hr
8 hr
1 day
1 mo
1 yr
Maximum for year
63
0.29
0.18
0.11
0.05
0.02
0.93
0.57
033
0.02
—
—
—
—
—
1.08
0.60
0,35
0.14
0.04
0.25
0.12
0.06
0.03
0.01
0,74
0.45
0.24
0.14
0.04
0.38
0.18
0.14
0.04
—
64
0.30
0.19
0.09
0.04
—
0.93
0.60
0.35
0.14
0,07
1.10
0.59
0.11
—
1.08
0.53
0.26
0.09
0.04
0.25
0.12
0.06
0.02
—
1.10
0.63
0.21
0.08
0.02
0.50
0.29
0.15
0.06
0.03
85
0.70
0.29
0.17
0.05
—
0.66
0.34
0.26
0.11
0.05
0.57
0.36
0.10
0.04
0.97
0.62
0.28
0.10
0.03
0.34
0.16
0.10
0.04
—
0.90
0.43
0.23
0.09
0.03
0.48
0.33
0.23
0.11
0.03
66
0.40
0.24
0.15
0.04
0.02
0.68
0.34
0.26
0.13
0.05
0.43
0.26
0.08
0.04
0.75
0.49
0.23
0.07
0.04
0.50
0.26
0.20
0.07
—
1.20
0.42
0.26
0.12
0.05
0.87
0.50
0.26
0.11
0.03
67
0.66
0.41
0.18
0.07
0.04
0.91
0.52
0.30
0.11
0.05
0.52
0.26
0.08
0.04
0.90
0.58
0.33
0.07
0.03
0.36
0.15
0.11
005
0.03
0.80
0.34
0.22
0.10
0.04
0.36
0.16
0.09
0.02
0.01
001 01
0.62 0,40
— 0.28
— _
— —
— —
0.92 0.70
— 0.55
— —
— —
— —
0.74 0.47
— 0.38
— —
— —
0.97 0.70
— 0.54
— —
— —
— —
0.47 0.32
— 0.25
— —
— —
— —
1.OO 0.65
— 0.42
— —
— —
— —
0.68 0.47
— 0.39
— —
— —
— —
1
0.22
0.18
0.12
—
—
0.41
0.32
0.26
—
—
0.29
0.21
0.18
—
0.37
0.29
0.22
—
—
0.15
0.12
0.09
—
—
0.38
0.26
020
—
—
0.27
0.23
0.19
—
—
Percentile"
10
0.07
0.07
0.06
0.06
—
0.15
0.14
0.13
0.11
—
0.11
0.10
0.09
—
0.09
009
0.09
0.07
—
0.05
0.05
0.05
•0.04
—
0.10
0.10
0.10
009
—
0.08
0.07
0.07
0.06
—
30
003
0.03
0.03
0.04
—
005
0.05
0.06
0.07
—
0.04
0.04
005
—
0.03
0.03
0.04
0.05
—
0.02
0.02
0.03
0.03
—
0.02
0.03
0.04
0.05
—
0,02
0.02
0.03
0.03
—
50
0,01
0.02
0.02
0.02
0.02
0.02
0.03
0.03
0.04
0.05
0.02
0.02
0.03
0.04
0.02
0.02
0.02
0.03
0.04
0.01
0.01
0.02
0.02
0.03
0.01
0.01
0.01
0.02
0.04
0.01
0.01
0.01
0.02
0.03
70
0.00
0.01
0.01
0.02
—
0.01
0.01
0.02
0.02
—
0.01
0.01
0.01
—
0.00
0.01
0.01
0.01
—
0,00
0.01
0.01
0.01
—
0.00
0.00
0.00
0.01
—
0.00
0.00
0.01
0.01
—
90
0.00
0.00
0.00
0.01
—
0.00
0.00
0.01
0.02
—
0.00
0.00
0.00
—
0.00
0.00
0.00
0.01
—
0.00
0.00
0.00
0.00
—
0.00
0.00
0.00
0.00
—
0.00
0.00
0.00
0.00
—
'Determined by continuous Gness-Saltzman method
"Concentrations greater than or equal to specified value in indicated percentage of samples
75
-------
TABLE 5-7, NITROGEN DIOXIDE CONCENTRATION8 IN CALIFORNIA BY AVERAGING TIME AND FREQUENCY,
1963-67
(ppmj
Place, sue no ,
averaging
time
Anahaim-176:
1 hr
8 hr
1 day
1 mo
1 yr
Oakiand-327'
1 hr
8 hr
1 day
1 mo
1 vr
Riverside-1 26
1 hr
8 hr
1 day
1 mo
1 vr
Sacramento-276.
1 hr
8 hr
1 day
1 mo
1 yr
San Bernardino-151.
1 hr
8 hr
1 day
1 mo
1 yr
San Diego-IOV.
1 hr
8 hr
1 day
1 mo
1 yr
Stockton-252'
1 hr
8 hr
1 day
1 mo
1 yr
Maximum for year
63
0.20
0.15
0.11
0.06
0.03
0,28
0,19
0.10
0,05
0.03
0.27
0,18
0.14
0.06
—
0,29
0.17
0,13
007
0.04
0.14
0.12
0,06
0.03
0.01
0.33
018
012
006
0.03
0.13
0.08
0.06
0.02
— •
64
0.22
0.13
0,11
0.04
—
0.41
0.25
0.16
0,05
003
0.56
026
0.19
005
0.04
0.32
0,21
013
0.07
0.04
0.06
0.03
0.02
—
—
0.35
0 15
0.09
0.06
0,02
0.22
0.15
0.08
0.05
0.03
65
0,23
0.16
0.13
0.05
—
0.23
0.15
0.10
0,06
0,03
0.49
0.31
0.25
0.08
0.05
0.30
020
0,13
0.07
0.04
011
007
0.05
0.03
—
0.52
0.22
0,12
0,04
0,02
0.14
0.11
008
0.03
0.02
66
0.27
0.15
0.12
0.05
0.04
0.29
0.18
0.13
0.05
0.03
025
0,18
0.13
0.05
0.04
0.27
0.21
0,15
0.05
*0.03
0.25
0.14
0.11
0.06
0.04
0.40
0,19
0.12
0.06
0.03
0,16
0.11
0.07
0.03
0.02
67 001
0.27 0.25
0,19 —
0.13 -
0,07 —
0.04 —
0.33 0.33
023 —
0.15 —
0.06 —
0.04 _
0.31 0.49
0.23 —
017 —
0.05 —
0.04 —
0.30 0.30
0.20 —
0.13 —
005 —
0.04 —
0.22 0.23
0.17 —
0.13 —
0.07 —
0.05 —
0,34 0.35
017 —
008 —
0.03 —
0.02 —
0.18 0.20
010 —
0.06 —
0.03 —
0.02 —
01
0.21
0.18
—
—
—
0.23
0.20
—
—
—
0.32
0,29
—
—
—
0.22
0.19
_
—
—
0.19
0.16
_
—
—
0.22
~0.18
—
—
—
0.16
0.12
—
—
—
1
0.15
0.13
0,10
—
—
0,14
0.13
0.10
—
—
0.18
0.15
0.14
—
—
0.14
0.12
0,11
—
—
0.13
0.12
0.10
—
—
0.12
0.11
0.09
—
—
0.09
0.07
0.06
—
—
Perce
10
0.07
0.07
0.06
0.06
—
0.07
0.06
0.06
0,05
—
0.09
0.08
0.07
0,05
—
0.07
0.06
0,06
0.05
—
0.07
0.06
0.06
0.05
—
0.06
0.06
0.05
0,04
—
0.04
0.04
0.04
0.03
—
intile"
30
0.05
004
0,04
0.05
—
0.04
0.04
004
0.04
—
0.05
005
0.05
005
—
0.04
0.04
0.04
0.04
—
004
0.04
0.04
0.04
—
0.03
0.03
0.03
0.03
—
0,02
0.03
0.03
0.02
—
50
0.03
0.03
0.03
0.04
0.04
0.03
0.03
0,03
0.03
0.03
0.03
0.04
0.04
0,04
0.04
0.03
0.03
0.03
0.03
0.04
002
002
003
0,03
004
0.01
0.02
0.02
0.02
0,02
0.02
0.02
0.02
0,02
0.02
70
0.02
002
0.02
0.03
—
0.02
0.02
0.02
0.03
—
0.02
0.02
0.03
0.04
—
0.02
0.02
0.02
0,03
—
0.01
0.01
0.01
0.01
—
0.00
0.01
0.01
0,01
—
0.01
001
0,02
0.02
—
90
001
0.01
0.01
002
—
0.01
0.01
0.02
0.02
—
0.01
001
0.02
0.03
—
0.01
0.01
0.02
0.02
—
0.00
0.00
000
000
—
0.00
0.00
0.00
0.01
—
001
0.01
0.01
0,01
—
"Determined by continuous Gness-Saitzman method
Concentration greater than or equal to specified value in indicated percentage of samples
76
-------
s
X
o
50
40
30
20
10
MAX-HOUR AVERAGES • ALL DAYS OF YEAR
MAX-HOUR AVERAGES - JULY, AUG., SEPT.
6-to-9-a.m. AVERAGES - JULY, AUG., SEPT.
1963 1964 1965 1966 1967 1968 1969 1970 1971 1972
YEAR
Figure 5-4. Oxides of nitrogen trends in Los-Angeles, 1963-1972, 6-to-9-a.m. and maximum 1-hr average
concentrations.16
Summary of Organic Emissions Data
A summary of several estimates of the total
hydrocarbon emissions from mobile, stationary,
and natural sources within the United States is
presented in Tables 5-8 and 5-9.84 The differences
between estimates are due primarily to differences
in calculation techniques, data sources, and
underlying assumptions. Most investigators agree
that emission estimates are improving as more is
learned about sources and measurement
techniques, and that the more recent estimates in
Tables 5-8 and 5-9 are probably more reliable than
earlier ones.
TABLE 5-9. TOTAL YEARLY HYDROCARBON EMISSION
RATE FOR THE CONTINENTAL UNITED STATES, BASED
ON LEAF BIOMASS8*
Item Emissions, 106 tons/yr
Summer (5 months) 52.4
Vegetation emissions 48 2
Leaf litter emissions 4.2
Winter (7 months) 34.9
Vegetation emissions 34.9
Leaf litter emissions ~0
Total yearly emissions from
vegetation and leaf litter 87.3
Total year emissions from
anthropogenic sources 21.2
Natural emissions as percent of total
emissions from all sources 80%
TABLE 5-8. ESTIMATES OF TOTAL HYDROCARBON
EMISSIONS FROM MANMADE AND
NATURAL SOURCES
Estimate, 10° metric lons/yr
Source type
1972 19?2MSA" 1972'3 1972«= 1973" 1974"
Rasmussen Research Corp. NEDS NADB NADB NADB
Mobile
Stationary
Natural
ND
NO
2.3'
ND
23.1
ND
14.8
10.2
ND
12.8
15.1
ND
12.5
15.5
ND
11.6
15.5
ND
'Emissions from tree foliage only (see J Air Poll. Contr Assoc 22 537-543.1972)
More detailed hydrocarbon emission inventories
for the United States and hydrocarbon emission
trends in 1970-75 are presented in Tables 5-10
and 5-11,63 Information presented in Tables 5-10
and 5-11 was taken from the EPA Data File of
Nationwide Emissions63'66 and has been validated
recently. The data in Tables 5-10 and 5-11 should
not be compared with previously published data on
1970-75 emissions.
77
-------
TABLE 5-10. 1974 NATIONWIDE ESTIMATES OF TOTAL
HYDROCARBON SOURCES AND EMISSIONS66
Source category
Transportation (total)
Highway
Non-highway
Stationary fuel combustion (total)
Electric utilities
Other
Industrial processes (total)
Chemicals
Petroleum refining
Metals
Others
Solid waste (total)
Miscellaneous (total)
Forest wildfires
Forest-managed burning
Agricultural burning
Coal refuse burning
Structural fires
Organic solvents
Oil and gas production and
marketing
Total
Emissions,
1974
(11 3)
9.8
1.5
(1.6)
0.1
1.5
(3.3)
1.6
0.8
0.2
07
. (.9)
(127)
0.5
0.2
0.1
01
0.0
8.1
3.7
29.8
106 metric tons/yr
1975 (preliminary)
(106)
9 1
1 5
(1 3)
01
1 2
(32)
1 5
08
02
07
(8)
(122)
05
02
0.1
0 1
<0 1
7.5
38
280
TABLE 5-11. NATIONWIDE TOTAL HYDROCARBON
EMISSION TRENDS. 1970-75"
Year
1970
1971
1972
1973
1974
1975
Emission, 106 metric tons/yr
30.7
30.2
30.9
30.8
298
28.0
Table 5-12 presents a more detailed breakdown
of hydrocarbon emissions from mobile sources
than is given in either Table 5-10 or 5-11. Data for
this table were obtained from the EPA Data File of
Nationwide Emissions.66 The table indicates the
relatively large contribution that gasoline-fueled
light vehicles make to total hydrocarbon emissions
from mobile sources. Of course, it must be
recognized that hydrocarbon emissions from light-
duty, gasoline-fueled vehicles have been reduced
since 1974, and the exhaust emission rate has
decreased oy about 70 percent.65 Evaporative and
crankcase emissions have remained essentially
unchanged during this period. It is difficult to
estimate national emission levels because of the
uncertainties associated with vehicle populations,
miles driven, and deterioration of emission control
systems. But from the model developed by
Heywood and Martin, a net decrease of about 28
percent has been projected for aggregate
hydrocarbon emissions from automobiles during
the 1972-77 period.38 Tables 5-13 and 5-14 give
organic emission and reactive-organic emission
inventories for the Los Angeles Air Quality Control
Region.81
TABLE 5-12. HYDROCARBON EMISSIONS FROM
MOBILE SOURCES"
Source category
Land vehicles (total):
Gasoline-fueled (total)
Light vehicles
Heavy vehicles
Off-highway
Diesel-fueled (total)
Heavy vehicles
Off-highway
Rail
Aircraft (total)
Military
Civil
Commercial
Vessels (total)
Coal
Diesel
Gasoline
Total
Emissions,
106 metric tons/yr
(10.11)
8.41
1.26
0.44
(0.46)
0.15
0.13
0.18
(033)
0.22
0.04
0.07
(0.43)
<0.01
0.02
0.41
11.33
% of total
(89.23)
74.23
11.12
3.88
(4.06)
1.32
1.15
1 59
(2.91)
1.94
0.35
0.62
(3.79)
<0.01
0.18
3.61
—
Hydrocarbon Emissions from Natural Sources
The major natural sources that have been
identified and for which quantitative estimates are
available are the biologic decomposition of organic
matter, seepage from natural gas and oil fields, and
emission of volatile compounds from plants.
However, there is information in the literature that
indicates that there are many other natural
sources of hydrocarbons and oxygenates that have
not been considered heretofore. This section
discusses the major natural sources for which
quantitative estimates are available and indicates
some of the minor sources that have been
identified.
Methane is produced in the anaerobic bacterial
decomposition of organic matter in swamps, lakes,
marshes, paddy fields, etc. Koyama46 has
estimated the global production of methane as 2.7
x 1014 g/year (300 million tons/year) on the basis
of his measurements of methane fermentation in
various soils and lake sediments under controlled
experimental conditions. Robinson and Bobbins74
have estimated the production from swamps and
humid tropical areas and added it to Koyama's
figure to derive an estimate of methane production
of 14.5 x 1014 g/year (1.6 billion tons/year). Other
sources of methane are gas, oil, and coal fields.5
The seepage of natural gas from these areas has
78
-------
been determined by Ehhalt,33 through carbon-14
measurements, to contribute as much as 25
percent to the total atmospheric methane. The
remaining 75 percent is of recent biogenic origin.
Our understanding of the natural sources of
methane is still highly limited. As a result, current
estimates of global production remain speculative.
Plants release a variety of volatile organic
substances, including ethylene, isoprene, a-
pinene, and a variety of other terpenes.
Rasmussen68 and Rasmussen and Went70
measured and reported ambient concentrations of
volatile plant organics (such as isoprene, a- and &-
pinene, limonene, and myrcene) in air at remote
sites. From the average concentration of 10 ppb
that they measured, a global production rate of
volatile plant organics was estimated to be 438
million tons/year. From the available estimates of
methane and terpene emissions, it is estimated
that worldwide natural hydrocarbon emissions are
TABLE 5-13 ORGANIC EMISSION INVENTORY FOR THE METROPOLITAN LOS ANGELES
AIR QUALITY CONTROL REGION, 1972"
Source category
Stationary sources: Organic fuels and combustion
Petroleum production and refining:
Petroleum production
Petroieum refining
Gasoline marketing:
Underground service station tanks
Auto tank filling
Fuel combustion
Waste burning and fires
Stationary sources: Organic chemicals
Surface coating'
Heat treated
Air dried
Dry cleaning;
Petroleum- based solvent
Synthetic solvent
-------
about 2 billion tons/year. However, there is
evidence that many other volatile organics are
emitted from natural sources into the atmosphere.
These may be only trace amounts emitted on a
local scale, but the total worldwide production can
be very large.
Annual hydrocarbon emissions from natural
sources in the United States can be estimated from
the preceding information. If the natural emission
of methane is uniform in all land areas, Robinson
and Bobbins'74 estimate of the worldwide emission
of methane (1,6 billion tons/year) can be used to
calculate a U.S. emission of 100 million tons/year.
The assumption that natural methane is emitted
TABLE 5-14. REACTIVE EMISSION INVENTORIES FOR THE METROPOLITAN LOS ANGELES
AIR QUALITY CONTROL REGION81
Total
emissions
Source category
Stationary sources. Organic fuels
and combustion'
Petroleum production and refining:
Petroleum production
Petroleum refining
Gasoline marketing:
Underground service
station tanks
Auto tank filling
Fuel combustion
Waste burning and fires
Stationary sources Organic chemicals:
Surface coating;
Heat treated
Air dried
Dry cleaning:
Petroleum based solvent
Synthetic solvent (PCE)
Degreasmg.
TCE solvent
1,1, 1-T solvent
Printing:
Rotogravure
Flexigraphic
Industrial process sources:
Rubber and plastic manufacturing
Pharmaceutical manufacturing
Miscellaneous operations
Mobile sources
Gasoline-powered vehicles:
Light duty vehicles:
Exhaust emissions
Evaporative emissions
Heavy duty vehicles'
Exhaust emissions
Evaporative emissions
Other gasoline-powered equipment.
Exhaust emissions
Evaporative emissions
Diesel-powered motor vehicles
Aircraft
Jet
Piston
Total
Tons/
day
62
50
48
104
23
41
14
129
16
25
11.
95
31
15
42
16
83
780
481
285
67
110
22
12
20
22
2604
%o(
total
2.3
1.9
1,8
4,0
0.9
1.6
0.5
5.0
0,6
1.0
0.4
3.6
1.2
0.6
1 6
06
3.2
30.0
18.5
10.9
2.6
4.2
0.8
0.5
0.8
0.9
100
Reactive emissions
Reactive tons/day"
2-groupb
scheme
24
33
47
94
6
22
9
88
9
0
6
0
21
15
33
10
40
562
362
205
48
79
16
8
10
18
1749
5-groupb
scheme
28
27
40
76
13
32
8
71
6
1
5
5
16
14
39
9
38
562
293
205
41
79
13
9
10
20
1660
6-groupb
scheme
18
27
40
77
8
27
8
71
6
1
5
5
16
14
39
9
38
562
293
205
41
79
13
9
10
20
1641
2-group
scheme
1.4
1.9
2.7
5.4
0,3
1,3
0.5
5.0
0.5
0.0
0.3
0.0
1.2
0.8
1.9
0.6
2.3
32.1
19.8
11.7
2.7
4.5
0.9
0.5
0.6
1.0
100
% of total
5~group
scheme
1.7
1.6
2.4
4.6
0.8
1.9
0.5
4.3
0.4
0.1
0,3
0.3
1.0
0.8
2.3
0.5
2.5
33.9
17.7
12.3
2.5
4.8
0.8
0.5
0.6
1.2
100
6-group
scheme
1.1
1.6
2.4
4.7
05
1.6
0.5
4.3
0.4
0.1
0.3
0.3
1.0
08
2.4
0.5
2.3
34.2
17.9
12.5
2.5
4.8
0.8
0.5
0.6
1.2
100
"To convert to reactive tan moles per day, multiply by 0 0145
bFor reactivity definition, see reference 42
80
-------
uniformly from all land areas is probably in error,
since emission rates are higher in tropical than in
nontropical areas. More realistically, assuming
that the U.S. methane emission rate per unit of
land area is half the world average, the U.S. natural
methane emission would be 50 million tons/year.
Similarly, if the estimate of Rasmussen and Went70
of worldwide terpene emission (438 million
tons/year) is based on uniform distribution over
the forested areas of the world, the U.S. emission
would be 22 million tons/year. Emission of
ethylene from plants in the United States has been
estimated at 20,000 tons/year.1 By combining
these estimates, a natural hydrocarbon emission
in the United States of about 72 million tons/year
is obtained.
Hydrocarbon Emissions from Anthropogenic
Sources
MOBILE SOURCES
As indicated in Table 5-13, light-duty, gasoline-
fueled motor vehicles account for a vast majority of
the hydrocarbons emitted into the environment
from mobile sources. There are three primary
sources of hydrocarbon emissions from the motor
vehicle: crankcase ventilation, gasoline
evaporation, and combustion exhaust. Crankcase
blowby emissir ns were essentially eliminated by
means of the positive crankcase ventilation system
introduced in "'963; evaporative emissions were
reduced by ?3out 30 percent by means of
adsorption-regeneration systems introduced in
1971.65 Exha st hydrocarbon emissions have
been reduced by about 90 percent in new vehicles
since 1968 wil n a variety of engine modifications,
and most recently with the application of
oxidation-cata'/st systems. Although the true
reductions in emissions are somewhat less than
these figures because of control device
deterioration, it is reasonable, nevertheless, to
believe that the current and the more advanced
controls of the futu re will effect a further reduction
in motor vehicle aggregate hydrocarbon emissions
through the middle to late 1980's.3a Vehicle
population growth and miles traveled will become
the controlling factors subsequent to that time.
The patterns of hydrocarbon emissions from
motor vehicles, as well as their mass, are
dependent on a large number of factors. Driving
patterns, emission"control system deterioration,
ambient temperature, pressure, humidity, and the
type of fuel used are all significant factors.11>5B'87
The relative contributions of evaporative gasoline
emissions and combustion exhaust in the
aggregate are also important. Therefore, it is very
difficult to speak in absolute terms about the
relative abundances of hydrocarbon types in
vehicle emissions. However, trends in emissions
can be discussed with reasonable authority.
Historically, combustion exhaust dominated the
hydrocarbon emissions from automobiles, and its
pattern was typical of atmospheric hydrocarbons
from motor vehicle sources. The advances in
emission controls, however, have changed this
situation. The automobile manufacturers have
been very successful in controlling exhaust
hydrocarbons. Since the 1975 model year,
evaporative hydrocarbons have constituted the
major fraction of the hydrocarbon aggregate from
new automobiles.65 It has been estimated that for a
typical 1975 passenger car, evaporative gasoline
accounts for about 60 percent of the total
aggregate emissions, and combustion exhaust, 40
percent.66
The pattern of exhaust hydrocarbons has
changed with the use of oxidation catalysts. Most
catalytic control systems show greater activity for
unsaturated hydrocarbons than saturated. They
show very little activity for methane.13'14 Thus the
relative abundance of paraffinic hydrocarbons in
the exhaust has increased with recent generation
automobiles. Typically, exhaust from a catalyst-
equipped automobile would contain about 62
percent paraffinic hydrocarbon, 17 percent
aromatic hydrocarbon, 18 percent olefinic
hydrocarbon, and 3 percent acetylenic
hydrocarbon, as compared with 40, 24, 26 and 11
percent, respectively, for noncatalytically equipped
automobiles. The methane levels generally range
from about 10 to 30 percent. Evaporative gasoline
is also dominated by paraffinic hydrocarbons, with
the C4-C6 paraffins typically accounting for about
70 percent of the total.55'83
Thus, the trend in hydrocarbon emissions from
passenger cars is toward an increased relative
abundance of paraffinic hydrocarbons.12 The
oxidation-catalyst exhaust emission control device
has generally resulted in an increased relative
abundance of methane and a decreased relative
abundance of unsaturated hydrocarbons,
particularly notable in the olefinic and acetylenic
hydrocarbons. Evaporative gasoline has become a
more significant fraction of the total aggregate of
motor vehicle hydrocarbon emissions.
Besides hydrocarbons, exhaust gases contain
oxygenated hydrocarbon compounds such as
aldehydes, ketones, alcohols, ethers, esters, acids,
and phenols. The total oxygenate concentration is
81
-------
about 5 to 10 percent of the total hydrocarbon
concentration. Aldehydes are generally believed to
be the most important class of exhaust oxygenates.
A reasonably complete quantitative analysis of
exhaust aldehydes is possible with gaschro-
matographic techniques. Formaldehyde is by far
the predominant aldehyde, constituting about 60
to 70 percent of the total (on a volume basis);
acetaldehyde is next, at about 10 percent; and pro-
pionaldehyde, acrolein, benzaldehyde, and the
tolualdehydes are all found in appreciable
amounts. As might be expected, the nature of the
gasoline influences the aldehydes formed.12'34'64'
86,89
There is little published information on
noncarbonyl oxygenates such as ethers, alcohols,
epoxides, and peroxides, but Seizinger and
Dimitriades75'76 measured 10 aldehydes, 6
ketones, and 16 noncarbonyl oxygenates in
exhaust from 22 different simple fuels, each
containing 1, 2, or 3 hydrocarbons. They developed
formulas from their data that can be used to
calculate the estimated concentrations of
oxygenate in gasoline exhaust.
The composition of organic emissions from the
other mobile sources (diesels, aircraft, etc.) is not
well defined. Diesel exhaust organics consist of a
light fraction, in the Ci-C4 range, and a heavy
fraction (C8-C22), with the relative amounts of the
two fractions varying with engine load and speed
(see Figure 5-1 ).39 Several diesel exhaust organics
have been identified in the course of researching
diesel exhaust odor.18 Such organics include
indans, indenes, naphthalenes, and tetralins, as
well as high molecular weight carbonyl and other
oxygenated compounds. Exhaust organics from
aircraft consist primarily of organics in the C8-C22
range (see Figure 5-2).79
There is little doubt that organic emissions from
gasoline-powered mobile sources are oxidant
producers almost in their totality. Emissions from
diesels and aircraft will participate in gas-phase
reactions only partly. A substantial part of such
emissions is expected to condense on surfaces,
thus contributing significantly, perhaps, to
visibility reduction.
STATIONARY SOURCES
A detailed description of the various stationary
sources of organic emissions, including emission
rate data, can be found in a recent National
Academy of Sciences report.61 Major constituents
of the emissions associated with fuel combustion
are organic acids, followed by hydrocarbons and
aldehydes.61 Such composition isentirelydifferent
from the combustion-related emissions from
mobile sources in which hydrocarbons are the
major constituent and organic acids are negligible.
Industrial processes discharge a variety of organic
compounds in the atmosphere, including C4-Ce
hydrocarbons from refineries and aromatic
hydrocarbons and acid derivatives, aldehydes,
alcohols, and phenols from chemical processing
operations. Solvent evaporation from painting,
coating, drycleaning, printing, etc., result in
substantial levels of emissions consisting mainly
of petroleum naphtha. Forest fires, structural fires,
and agricultural burning result in emissions that
are difficult to characterize and nearly impossible
to measure. Finally, gasoline marketing
operations, (i.e. storage, transportation, and
service station handling of gasoline) result in
emissions consisting primarily of C4-Ce
hydrocarbons and secondarily of whole gasoline
vapors. The relative importance of the stationary
source emissions, in terms both of amounts and
oxidant-producing potential, is illustrated by the
Los Angeles data shown in Tables 5-13 and 5-14.81
Emissions of Nitrogen Oxides
The distribution of nitrogen oxide (NO*)
emissions by major source categories is indicated
in Table 5-15.63 Data shown in this table have been
validated recently and are therefore more reliable
than previously reported data on 1974 NOX
emissions. Fuel combustion is the major cause of
technology-associated emissions. In 1974,
combustion of coal, oil, natural gas, and motor-
vehicle fuel accounted for more than 22 of an
estimated 23 million tons of manmade NO* in the
United States. An estimated 9.6 million tons was
emitted from transportation sources, 7.4 million
tons of which was from motor vehicles. Industrial
processes, solid waste disposal, and other
miscellaneous sources accounted for about 1
million tons of NOx.
Relatively small quantities of NOX are emitted
from noncombustion industrial processes, mainly
the manufacturing and use of nitric acid.52 Even
though total quantities may be small, high
concentrations of N0xcan be emitted from some of
these chemical processes. Electroplating,
engraving, welding, metal cleaning, and explosive
detonation also can be responsible for industrial
NO, emissions, as can the maufacture and use of
liquid-N02-based rocket propellants.
82
-------
TABLE 5-15. 1974 NATIONWIDE ESTIMATES OF
NITROGEN OXIDE SOURCES AND EMISSIONS"
Source category
Emission, 106 metric tons/yr
Transportation (total)
Highway
Nonhighway
Stationary fuel combustion (total)
Electric utilities
Other
Industrial processes (total)
Chemicals
Petroleum refining
Other
Solid waste (total)
Miscellaneous
Forest wildfires
Forest-managed burning
Agricultural burning
Coal refuse burning
Structural fires
Total
(9.6)
7.4
2.2
(12.1)
6.3
5.8
(0.6)
0.3
0.3
<0.1
(0.2)
(0.2)
0.1
01
22.7
Natural sources of nitrogen oxide emissions
include biological processes in soil, atmospheric
oxidation of ammonia (NHa), and, possibly,
photolysis of N02. Estimates of N0« emission rates
from natural sources vary considerably among
investigators. Robinson and Bobbins73 estimated
the annual N0« emissions from biological
processes to be 770 million metric tons (as N02>,
compared to 52 million metric tons from manmade
sources. Other investigators question the
importance of the biological processes, and they
offer the oxidation of NH3 resulting in 230 million
tons of NO* per year, as the main source.54 As in
the case of the hydrocarbon emissions, the natural
sources of NOX seem to dominate the anthro-
pogenic ones. But again, as in the hydrocarbon
case, this dominance has no significant relevance
to the ambient oxidant problem or even to the
ambient N02 problem, since the natural and
anthropogenic emissions are, for the most part,
segregated geographically, with the anthro-
pogenic ones concentrated in the populated areas.
REACTIVITY OF ORGANIC EMISSIONS
From an air pollution standpoint, the
photochemical reactivity of an organic pollutant
denotes the intrinsic ability of the pollutant to
participate in atmospheric chemical reactions that
result in photochemical smog formation. The
concept of hydrocarbon reactivity (the term
"hydrocarbon" is meant here to encompass all
organic substances) was developed when
laboratory research showed that different organic
substances, when exposed to atmospheric
conditions, did not react similarly. Specifically,
when traces of an individual organic and NO in air
were irradiated with artificial sunlight, the
resulting levels of smog, in terms of oxidant/ozone
yield, eye irritation, plant damage, visibility
reduction, etc., were found to vary widely with the
chemical structure of the organic reactant. As a
result of these studies, the concept of hydrocarbon
reactivity has evolved to include several reactivity
types, each type corresponding to a specific
chemical-biological manifestation of photo-
chemical smog.6The reactivity type of interest here
is the one associated with the oxidant/ozone yield.
The fact that organic substances differ greatly in
reactivity is extremely significant in understanding
the photochemical oxidant/ozone-forming
process. Such reactivity data are presently
available. The data and the methods used for
obtaining them are discussed next.
The procedure commonly used for reactivity
measurements is to irradiate clean air mixed with
the organic vapor, NO, and N02 at prescribed
concentrations in a smog chamber and to measure
the formation of oxidant/ozone. Experimental
conditions for such measurements are, to the
extent feasible and practical, similar to the
conditions typically present in polluted ambient
atmospheres. Thus the reactant concentrations,
the intensity and spectral characteristics of
radiation, the temperature, and often the relative
humidity are comparable to actual conditions in
the real atmosphere during the summer. Because
the concern traditionally has been for the oxidant
problems observed in urban areas, nearly all of the
reactivity data presently available were obtained
under experimental conditions simulating urban
atmospheres, and, more specifically, the Los
Angeles atmosphere. Recent developments,
however, led to recognition of two distinct
atmospheric situations (to be referred to here as
the urban, or no-transport, situation and the rural,
or transport, situation) for which an organic may
manifest different reactivities.27 The terms
"urban" or "no-transport" are used to designate
the situation in which emissions react for a few
hours and cause oxidant problems in the vicinity of
their sources. Conversely, in the rural or transport
situation, emissions undergo extensive transport
and prolonged photochemical reaction and cause
oxidant/ozone buildup in distant downwind areas,
which may be rural or urban. Again, nearly all of
the reactivity data currently available are
applicable to the no-transport situation only; there
83
-------
are very few measurements of reactivity under
simulated transport conditions.
The scientific evidence detailing the impact that
the reactivity of the organic emissions has on air
quality within the area of the sources and in the
areas far downwind is examined in the following
discussion. Scientific evidence pertinent to these
questions does exist25 and is presented next, first
for the urban or no-transport situation, and then
for the rural or transport case.
Reactivity data obtained through 1965, all
applicable to the urban situation, were compiled by
Altshuller3 and presented in terms of a reactivity
classification of hydrocarbons and aldehydes, as
shown in Table 5-16. Since 1965, several re-
activity studies have been reported, including
detailed reactivity data on hydrocarbon and
nonhydrocarbon organics. These studies and the
types of reactivity data reported are listed in Table
5-17.
TABLE 5-16. COMPARISON OF REACTIVITIES OF DIFFERENT TYPES OF ORGANICS3
Reactivity on 0 to 10 scale
Substances or
subclass
Ci-C3 paraffins
Acetylene
Benzene
C4+ paraffins'
Toluene (and other monoalkylbenzenes)
Ethylene
1 -alkenesd
Diolefms
Dialkyl- and trialkyl-benzenes
Internally double-bonded olefins
Aliphatic aldehydes
Ozone or
oxidant
0
0
0
0-4
4
6
6-10
6-8
6-10
5-10
5-10
Peroxy-
acvl-
nitrate
0
0
0
ob
NDC
0
4-6
0-2
5-10
8-10
+•
Formal-
dehyde
0
0
0
0"
2
6
7-10
8-10
2-4
4-6
+e
Aerosol
0
0
0
0
2
1-2
4-8
10
+•
6-10
NDC
Eve
irnta-
•ion
0
0
0
0"
4
5
4-8
10
4-8
4-8
+e
Plant
damage
0
0
0
0
0-3
+"
6-8
Ob
5-10
10
+•
Overall
.eactivity
0
0
0
1
3
4
7
6
6
8
-
"Averaged over straight-chain and branched-cham paraffins
Very small yields or effects may occur after long irradiations
'No experimental data available
Includes measurements on propylene through 1 -hexene, 3-ethyl-l-butene and 2,4,4-tnmethyl-1-1 -pentene.
"Effect noted experimentally, but data insufficient to quantify
TABLE 5-17. REACTIVITY DATA
Investigator
Data reported
Altshuller and Bufalini7
Brunelle et al16
Dimitnades et al.31
Dimitnades et al.31
Dimitriades and Wesson32
(Bureau of Mines)
Heuss, J.36
Heuss and Glasson37
(General Motors)
Laity et al.47
(Shell)
Levy and Miller48
(Battelle)
McReynolds et al.66
Miller et al.57
Wilson and Doyle88
(Stanford Research Inst.)
Yanagihara et al.90
Reactivity data on hydrocarbons and aldehydes.
Reactivity data for solvent organics.
Reactivity data for hydrocarbons and aldehydes.
Rate of NOz formation and product yield reactivity data for hydrocarbons.
Rate of NO2 formation and product yield reactivity data for aldehydes.
Reactivity data on hydrocarbons.
Eye irritation, rate of NO: formation, and product yield reactivity data
for hydrocarbons.
Reactivity data for solvent organics.
Reactivity data for solvent organics.
Reactivity data for hydrocarbon disappearance.
Aerosol activity data for hydrocarbons.
Reactivity data for solvent organics.
Reactivity data for solvent organics.
Of these studies, however, only a few included
systematic testing of a variety of organic
compounds. The data from these main studies,
summarized in Table 5-18, were used by EPA to
classify organics into three reactivity classes,
shown in Tables 5-18 and 5-19. The data in Table
5-18 show that organics do differ widely in
reactivity.
84
-------
TABLE 5-18. SUMMARY OF DATA FROM STUDIES ON REACTIVITIES (TOLUENE EQUIVALENTS) AND
CLASSIFICATION OF ORGANICS
Organic
Paraffinic hydrocarbons:
CvCa paraffins
Ci. paraffins
Cycloparaffins
Olefinic hydrocarbons
Aromatic hydrocarbons:
Benzene
Primary and secondary mono-
alkylbenzenes
Tert-monoalkylbenzenes
Dialkylbenzenes
Tri-, tetraalkylbenzenes
Styrene
Me-styrenec
Aldehydes:
Aliphatic aldehydes
Benzaldehyde
o-Tolualdehyde
mf>- Tolualdehyde
Ketones:
Acetone
Me-et-ketonec
n Alkylketones
Branched alkylketones
Cyclic ketones
Unsaturated.ketones
Alcohols:
Methanol
Ethanol
Isopropanol
Primary and secondary
Ca.u alcohols
Tertiary alky) alcohols
Diacetone alcohol
Ethers:
Diethyl ether
Tetrahydrofuran
Ethyl cellosolves
Esters:
Methyl acetate
Primary, secondary alkyl
C2+ acetates
Tertiary alkyl acetates
Phenyl acetate
Methyl benzoate
Amines:
Ethyl amines
N-Me-pyrrolidone
N,N-dimethyl-formarriide
N,N-dimethyl-acetamide
Nitroalkanes:
2-Nitropropane
Halocarbons;
Perhalogenated hydocarbons
Partially halogenated paraffins
Partially halogenated olefins
Halogenated benzenes
"Japan EA - Japan Environment Agency, BOM
1 Ahtf*ratnriaG~rnlumhiifi QKali = QhaEl ("111 C.n
Japan EA' BOM' CM"
0-0.1 -- 0-0.2
~..
0.1-0.7 0.1 o.B-o.eio.oe)"
1.6-2.3 1.6-2.5 0.9-2.0
0 0 0.2
0.9 0.8-1.0 0.6-1.0
0.4
1.4-2.0 1.4-1.5 0.9-1.3
1.7 1.5
--
..
1.5-2.0
0.1
0.4
0.2
0
0.1
—
1.4
0.35
--
0
0
0.1
0.7-1.0
-.
—
..
,,
1.9.
0
0.1
,.
..
—
..
..
.,
..
-.
0.1
..
0.5
0-0.2
= Bureau of Mines, U S Dept o( interior, QM * General Motors Corp ,
Battells
-
--
0.4-0.
1.3-1
0
0.9-1
0.6
1 .0-1
1.5
0.7
1.5
-
-
--
--
0
0.6
0.5-0
1.0-1
0.2
1.5-1
--
.-
0.2
--
„
1.4
—
1.9
1.5
--
0.2
—
0
0
0.1-0
0.7
„
..
0.2
--
--
—
-
>' Shell'
--
--
,6 0.8-1.0
,5 1.8-3.1
0.2
.0 1.0-1.2
05
.2 1.3-1.7
3.2
--
--
--
--
,.
—
0.1
0.9
.8 1.4
.8 1.3
0.5-0.6
.7
--
1.0
0.6
—
0.3
„
2.5
1.4
--
--
0.8-1.0
0.5
--
--
2
—
0.2
0.9
0.7
..
--
--
Class
1
II
III
1
III
II
III
III
III
III
III
1
11
II
I
1
II
III
II
111
1
II
1
III
1
III
III
III
III
II
II
1
1
1
III
1
II
II
1
1
III
1
Battelle = Battelle Memorial Institute
Laboratories-Columbus, Shell = Shell Oil Co
Values were obtained from two different GM studies in which last conditions ware different.
cMe = methyl, et - ethyl
85
-------
TABLE 5-19. CLASSIFICATION OF ORGANICS WITH RESPECT TO THEIR OXIDANT-RELATED
REACTIVITY IN URBAN ATMOSPHERES
Class I
(low reactivity)
Class II
(moderate reactivity)
Class III
(high reactivity)
Ci-Cs paraffins"
Acetylene"
Benzene
Benzaldehyde"
Acetone"
Methanol
Isopropanol
Tert-alkyl alcohols"
Methyl acetate"
Methyl benzoate
Ethyl amines"
N, N-dimethyl
formamide"
Perhalogenated
hydrocarbons
Partially halogenated
paraffins
Mono, dichlorobenzenes
Methyl-ethyl-ketone
Tert-monoalkyl benzenes
Cyclic ketones
Tolualdehydes
Tert-alkyl acetates"
2-Nitropropane"
C<+ paraffins, cyclo-
paraffms"
Ethanol
Prim , sec C2+ alkyl
N, N-dimethyl acetamide"
n-alkyl C5+-ketones"
Prim., sec. monoalkyl benzenes
Dialkyl benzenes
Styrene
N-Methyl pyrrolidone
Partially halogenated
olefins
Aliphatic olefins
Tri-, tetra-alkyl
benzene
Methyl styrene
Branched alkyl ketones
Unsaturated ketones
Aliphatic aldehydes"
Diacetone alcohol
Ethers'
2-Ethoxy-ethanol
'Currently classified as not photochemically reactive under Los Angeles County Rule 66 and simihar regulations.
For a practical application of these differing
reactivities, it is necessary that the photochemical
behavior of an organic pollutant mixture be
consistent with the behavior of the individual
components. This requirement was explored in
smog chamber studies, and overall results,
although not always conclusive, were positive. For
example, for an ethane/butane/propylene/NOx
mixture at the organic-to-NOx ratio of 10:1,
removal of the extremely reactive propylene
resulted in less smog, which is consistent with the
individual component reactivities. At a ratio of
20:1, however, removal of propylene had no
effect.9'44 Smog chamber data on solvent mixtures
also showed that less smog was formed when less
reactive solvents were substituted for more
reactive ones.29 For the more complex automotive
exhaust mixtures, the values for the maximum 1 -
hr oxidant observed in the smog chamber did not
correlate well with the reactivity values computed
from exhaust composition and individual
component reactivity data, as shown in Figure 5-
5.31 Such lack of correlation, however, was
attributed primarily^to misidentification of exhaust
components and to the inappropriateness of the
linear simulation method for calculating mixture
reactivities.31
Overall, the smog chamber data suggest that the
relatively low organic-to-NOx ratios encountered
in typical no-transport atmospheres will probably
result in lower reactivity. In the real atmosphere,
oxidant formation is expected to be different from
that suggested by the chamber data because the
main effect of a reduction in reactivity of the
organic reactant is a delay in oxidant formation.
This delay, in turn, means that the daily oxidant
maximum will not be eliminated but, will rather be
shifted some distance downwind. Thus the main
result to be expected from reduced reactivity is a
reduction in the peak oxidant concentration that
results from the additional dispersion associated
with the time delay. Of course, if the oxidant
accumulation is delayed sufficiently, (i.e., until late
afternoon), then such accumulation will not occur
for lack of radiation. Note, however, that in this
latter case, a greater part of the organic precursor
will escape the photochemical process during the
first solar day and will be transported downwind
where, if sufficient NOx is also present, it will react
to form oxidant/ozone. This case is discussed
further in a subsequent section of this chapter.
The preceding discussion dealt with the urban or
no-transport situation. Analogous evidence on
reactivities of organics and on mixture behavior for
the rural or transport situation is very scant;
therefore, possible answers presented here will
have to be surmised for the most part.
Reactivities of organics under simulated
transport conditions have not been measured
systemically as they have in the case of no-
transport. Nevertheless, from the limited direct
data available29'35 and from current knowledge of
the oxidant formation mechanism, it has been
established that under transport conditions, the
effective range of reactivities is more narrow than
for the no-transport case. Therefore, and on this
86
-------
v> 1.0
UJ
<
| 0.9
a
UJ
UJ
Z
UJ 0.8
>
a.
u.
_• 0.7
0.6
O
D 0.5
2
cc
3
m
£0.4
t
<
ui
CC
M
o
I
X
0.2
a o.i
UJ
U
-i
2
LEAST SQUARES CORRELATION LINE
CORRELATION COEFFICIENT - 0.34
I I J 1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
OBSERVED MAXIMUM O3 REACTIVITY, PROPYLENE EQUIVALENTS
Figure 5-5. Correlation of observed and calculated maximum On reactivities.3
basis alone, the impact of emission substitution on
air quality well downwind from the sources should
be less than the impact within the source area.
The evidence and reasoning discussed in the
preceding paragraphs do not preclude the
existence of organics that are so slightly reactive
that they could be of almost no concern at all
insofar as the oxidant problem is concerned.
Evidence directly addressed to this possibility is
neither adequate nor easy to obtain. One approach
to identifying such organics has been offered by
EPA recently29 and is based on the use of smog
chambers. Briefly, the method entails smog
chamber testing of organics under a variety of
reactant concentrations (but within a realistic
range), organic-to-NO* ratio, and prolonged
irradiation conditions. Truly nonreactive organics
are proposed to be defined as those that, when
tested in an appropriate smog chamber do not yield
more'than 160/ug/m3(0.08 ppm) ozone under any
set of test conditions. An alternative method
proposed by Chang and Weinstock20 is based on
the use of a photochemical model to predict the
role and impact of less reactive organics in
transport atmospheres. The two methods were
critically examined by Calvert and Jeffries,17 and
their relative merits and limitations were
discussed. One problem with the chamber method
87
-------
arises from the presence of chamber wall effects.
Such effects create problems; first, because they
cause reactivity manifestations that, when
extremely unreactive organics are tested, are
comparable to or even stronger than those caused
by the test reactant mixture; and second, because
they are not well understood. Thus, though cham-
ber reactivities of the various organics are probably
reliable in a relative sense, their absolute values
are somewhat questionable. This latter weakness
also affects the chamber-determined reactivity
level that separates reactive organics from those of
no concern insofar as the oxidant problem is
concerned.
The modeling method of Chang andWeinstock20
also has weaknesses because of the questionable
chemical mechanism and kinetic data used by the
authors. Furthermore, because such models are
derived theoretically, and because their validation
is based mainly on comparison with smog chamber
data, their predictions cannot be much more valid
than those obtained by the smog chamber method.
Finally, this modeling method has very limited
utility, since photochemical models exist only for
those very few organics for which the atmospheric
photo-oxidation mechanism has been elucidated.
The key problem here is to identify the boundary
organic that separates the reactive organics from
the nonreactive ones. Solving this problem would
probably require application of both the chamber
method and modeling techniques. The chamber
method would be used to determine the relative
reactivities of organics, and photochemical
modeling would be used to predict the absolute
reactivity of the boundary organic. Because the
chamber data suggest that propane may be a
reasonable candidate for the boundary position,28
and because the mechanism of the atmospheric
photo-oxidation of propane is relatively well
established, it might be advisable to estimate the
oxidant-forming potential of propane using the
modeling method.
As an alternative to smog chamber measure-
ment, Pitts etal.67 suggested that the rate constant
for the reaction between the OH-radical and an
organic be used as a measure of the reactivity of
the organic. Such rate constants have been
measured for over 100 hydrocarbon and non-
hydrocarbon organics, and with exceptions, their
relative values roughly parallel the relative
reactivity values obtained by the smog chamber
method. Relative to the smog chamber reactivities,
these rate constant reactivities are superior in
some respects but inferior in others. For example,
they are more reliable measures of reactivity in the
cases of extremely unreactive organics, and
therefore they provide better bases for identifying
such organics. On the other hand, this rate
constant is not a direct measure of the organic's
ability to produce oxidant/ozone; therefore, it has
only limited validity.17
In summary, the use of the concept of reactivity
is-generally accepted as sound "because organic
emissions do differ widely in reactivity. Additional
research should be done to provide more reliable
bases for identifying those organic emissions that
are of no concern insofar as the oxidant problem is
concerned,
SUMMARY
Organic pollutants in urban atmospheres
consist mainly of hydrocarbons emitted by
automobiles and from fuel evaporation, and of
oxygenated hydrocarbons from manufacturing and
the use of organic chemicals. In u rban atmospheres,
the ambient total organic concentrations, reported
usually as 6- to 9-a.m. averages of total non-
methane hydrocarbons (NMHC), range typically
around 1 ppm and can be as high as 10 ppm or even
higher (e.g., in Los Angeles). In rural and remote
atmospheres, the composition and concentrations
of organic pollutants are considerably more
uncertain, mainly because of deficiencies in the
analytical methods available for the con-
"centratipns involved. Overall, the evidence
suggests that natural NMHC levels are generally
less than 0.1 ppm, a fraction of which consists of
vegetation-related terpenes.
Concentrations of NO* in urban atmospheres
vary within a wide range, with highest values
exceeding 1 ppm. Such concentrations appear to
decline rapidly as the urban air moves away from
the city. In rucal and remote areas, ambient
concentrations do not exceed a few ppb and
therefore are often below the 5-ppb detection limit
of current commercial N0» analyzers.
Hydrocarbons and N0« are emitted into the
atmosphere from both natural and manmade
sources, with the natural contribution being the
greater. This greater contribution, however, does
not influence oxidant/ozone formation, because
the sources of natural and anthropogenic emis-
sions are spatially segregated, with anthropo-
genic ones concentrated in the populated areas. At
present, mobile sources account for a major part of
the organic emissions in most urban centers.
88
-------
Ongoing controls of mobile source emissions have
reduced levels and changed the composition of
emissions in favor of the paraffinic component.
Organic emissions differ widely in reactivity (i.e.,
in their ability to produce photochemical ozone and
other oxidants). Thus some organics probably do
not contribute to photochemical smog formation.
More research is needed to provide reliable bases
for identifying those organics that have no bearing
on the oxidant/ozone problem.
REFERENCES FOR CHAPTER S
1. Abeles, F. B,, L. E. Craker, L. E. Forrence, and G, R.
Leather. Fate of air pollutants: Removal of ethylene,
sulfur dioxide, and nitrogen dioxide by soil. Science.
773:914-916,1971.
2. Air Pollution Control Office. Air Quality Criteria for
Nitrogen Oxides. APCO Publication No. AP-84, U.S.
Environmental Protection Agency, Washington, O.C.,
January 1971.
3. Altshuller, A. P. An evaluation of techniques for the
determination of the photochemical reactivity of organic
emissions, J. Air Pollut. Control Assoc. 16.257-260,
1966.
4. Altshuller, A, P Hydrocarbons and carbon oxides. In: AIT
Pollution, Volume III. Measuring, Monitoring, and
Surveillance of Air Pollution. A, C. Stern, ed. Academic
Press, Inc., New York, 1976. pp. 183-211
5. Altshuller, A. P. Natural sources of gaseous pollutants in
the atmosphere. Tellus 70:479-492, 1958,
6. Altshuller, A. P. Reactivity of organic substances in
atmospheric photo-oxidation reactions. Int. J. Air Water
Pollut. 70.7,13-733, 1966.
7. Altshuller, A. P., and J. J. Bufalmi. Photochemical
aspects of air pollution: A review. Environ. Sci. Technol.
5:39-64, 1971.
8. Altshuller, A. P., and S. P. McPherson. Spec-
trophotometric analysis of aldehydes in the Los Angeles
atmosphere. J, Air Pollut. Control Assoc. 73:109-111,
1963.
9. Altshuller, A, P., S. L. Kopczynski, D. Wilson, W. A,
Lonneman, and F. D. Sutterfield. Photochemical
reactivities of n-butane and other paraffinic
hydrocarbons. J. Air Pollut. Control Assoc. 73.-787-790,
1969.
10, Altshuller, A. P., G. C. Ortman, and B. E. Saltzman.
Continuous monitoring of methane and other
hydrocarbons in urban atmospheres J. Air Pollut. Control
Assoc. 76:87-91, 1966.
11, Ashby, H. A., R. C, Stahman, B. H. Eccleston, and R, W,
Hum. Vehicle emissions—summer to winter. SAE Tech.
Pap. No. 741053, Society of Automotive Engineers, Inc.,
Warrendale, Pa., 1974.
12. Black, F. M. The impact of emissions control technology
on passenger car hydrocarbon emission rates and
patterns. In: International Conference on Photochemical
Oxidant Pollution and Its Control Proceedings. Vol. II. B.
Dimitriades, ed. EPA-600/3-77-001 b, U.S.
Environmental Protection Agency, Research Triangle
Park, N.C., January 1977. pp. 1053-1067.
13. Black, F. M., and R. L. Bradow. Patterns of hydrocarbon
emissions from 1975 production cars. SAETech. Pap. N o.
750681, Society of Automotive Engineers, Inc.,
Warrendale, Pa., 1975.
14, Black, F. M., and L. Hish, Automotive hydrocarbon
emission patterns and the measurement of nonmethane
hydrocarbon emission rates. SAE Tech. Pap No. 770144,
Society of Automotive Engineers, Inc., Warrendale, Pa.,
1977.
15. Brunelle, M. F., J. E, Dickinson, and W. J. Hamming.
Effectiveness of Organic Solvents in Photochemical
Smog Formation. County of Los Angeles, Air Pollution
Control District, Los Angeles, Calif., July 1966
16. California Air Resources Board. Ten Year Summary of
California Air Quality Data, 1963-1972. State of
California, Air Resources Board, Sacramento, Calif,
January 1974.
17. Calvert, J G., and H.E Jeffries. International Conferervce
on Oxidants, 1976—Analysis of Evidence and
Viewpoints. Part II. The Issue of Reactivity. EPA-600/3-
77-114, U.S. Environmental Protection Agency,
Research Triangle Park, N.C., October 1977,
18. Caragay, A, B., J. T. Funkhauser, D. A Dendall, G.
Leonardos, and P. L Levins. Identification of diesel
exhaust odor species, 1. Hydrocarbons-. Oily-kerosene
odor complex. Anal. Chem, 1978, (In press).
19. Cavanagh, L. A., C. F. Schadt, and E Robinson
Atmospheric hydrocarbons and carbon monoxide
measurements at Point Barrow, Alaska, Environ Sci
Technol. 3:251-257, 1969
20. Chang, T. Y., and B. Wemstock. Net ozone formation in
rural atmospheres. In: International Conference on
Photochemical Oxidant Pollution and Its Control.
Proceedings. Vol. I. B. Dimitriades, ed. EPA-600/3-77-
001 a, U.S. Environmental Protection Agency, Research
Triangle Park, N.C., January 1977. pp. 451-465
21. Chase, J. 0., and R, W Hum. Measuring gaseous
emissions from an aircraft turbine engine SAE Tech.
Pap. No, 700249, Society of Automotive Engineers, Inc.,
Warrendale, Pa., 1970.
22. Committee on Challenges of Modern Society. Air Quality
Criteria for Nitrogen Oxides. NATO/CCMS Document No
15, North Atlantic Treaty Organization, Brussels,
Belgium, June 1973.
23. Committee on Challenges of Modern Society. Air Quality
Criteria for Photochemical Oxidants and Related
Hydrocarbons. NATO/CCMS Publication No 29, North
Atlantic Treaty Organization, Brussels, Belgium,
February 1974.
24. Community Air Monitoring Program Data National
Aerometrtc Data Bank, U.S. Environmental Protection
Agency, Research Triangle Park, N.C
25. Dimitriades, B. Oxidant/Oa quality benefits from
"emission substitution." In Specialty Conference On:
Hydrocarbon Control Feasibility: Itslmpacton AirQuality.
Air Pollution Control Association, Pittsburgh, Pa., 1977.
pp. 111-121.
26. Dimitriades, B. Photochemical Oxidants in the Ambient
Air of the United States. EPA-600/3-76-017, U.S.
Environmental Protection Agency, Research Triangle
Park, N.C., February 1976.
27. Dimitriades, B. Proceedings of the Solvent Reactivity
Conference. EPA-650/3-75-010, US. Environmental
Protection Agency, Research Triangle Park, N.C.,
November 1974.
28. Dimitriades, B., and A. P. Altshuller. In: International
89
-------
Conference on Oxidant Problems. Analysis of the
Evidence and Viewpoints Presented Part I. Definition of
Key Issues J Air Pollut Control Assoc 27-299-307,
1977
29 Dimitnades, B , and S B Joshi Application of reactivity
criteria in oxidant-related emission control in the USA. In:
International Conference on Photochemical Oxidant
Pollution and Its Control Proceedings Vol II. B,
Dimitriades, ed EPA-600/3-77-001 b, U.S
Environmental Protection Agency, Research Triangle
Park, N.C, 1977 pp 705-711
30. Dimitnades, B., C J Raible, and C. A. Wilson.
Interpretation of Gas Chromatographic Spectra in
Routine Analysis of Exhaust Hydrocarbons Report of
Investigation HI 7700, U S Department of the Interior,
Bureau of Mines, Washington, D.C , 1972.
31 Dimitnades, B., G. P Sturm, Jr, T C Wesson, and F D.
Sutterfield. Development and Utility of Reactivity Scales
from Smog Chamber Data Report of Investigation Rl
8023, U.S. Department of the Interior, Bureau of Mines,
Washington, D.C , March 1975
32 Dimitnades, B , and! C Wesson Reactivities of Exhaust
Aldehydes Report of Investigation Rl 7527, U S
Department of the Interior, Bureau of Mines,
Washington, D C , May 1971.
33. Ehhalt, D H. Methane in the atmosphere J. Air Pollut.
Control Assoc. 7 7:518-519, 1967
34 Fracchia, M F, F J. Schuette, and PK Mueller. A
method for sampling and determination of organic
carbonyl compounds in automobile exhaust, Environ Sci.
Technol /:91 5-922, 1967
35. Glasson, W A, and P H Wendschuh Multi-day
irradiation of N0,-organic mixtures. In: International
Conference on Photochemical Oxidant Pollution and Its
Control Proceedings Vol II. B. Dimitriades, ed. EPA-
600/3-77-0016, U.S. Environmental Protection Agency,
Research Triangle Park, N C, January 1977. pp 677-
685
36. Heuss, J. Smog chamber simulation of the Los Angeles
atmosphere. In Scientific Seminar on Automotive
Pollutants. U S Environmental Protection Agency,
Research Triangle Park, N C , 1975.
37 Heuss, J. M , and W. A Glasson. Hydrocarbon reactivity
and eye irritation. Environ. Sci. Technol. 2:1109-1116,
1968.
38. Heywood, J B , and M. K. Martin Aggregate emissions
from the automotive population. SAE Tech. Pap. No,
740536 Society of Automotive Engineers, Inc.,
Warrendale, Pa , 1974
39. Hum, R W, and W F Marshall Techniques for diesel
emissions measurements SAE Trans. 77 (sect,2) 1492-
1501, 1971
40 KmosiantJ R., and J Paskind Hydrocarbon, Oxides of
Nitrogen and Oxidant Trends in the South Coast Air Basin
1963-1972, State of California, Air Resources Board,
Sacramento, Calif., June 1973
41 Ivanov, V P, and G A Yakobson Exchange of
metabolites in plants via aerial organics Soviet Plant
Physiol (Engl Transl ) 72,351-356, 1965.
42. Jeffries, H. £„ R, Kamens, D. L. Fox, and B. Dimitriades,
Outdoor smog chamber studies. Effect of diurnal light,
dilution, and continuous emission on oxidant precursor
relationships In. International Conference on
Photochemical Oxidant Pollution and Its Control.
Proceedings Vol I B Dimitriades, ed., EPA-600/3-77-
001 b, U.S. Environmental Protection Agency, Research
Triangle Park, N.C..January 1977.
43 Junge, C. E. Recent investigations in air chemistry
Tellus. 8:127-139, 1956,
44. Kopczynski, S. L., R L Kuntz, and J J Bufalmi
Reactivities of complex hydrocarbon mixtures. Environ
Sci. Technol. 9.648-653, 1975.
45. Kopczynski, S. L, W. A, Lonneman, T. Winfield, and R
Seila. Gaseous pollutants in St. Louis and other cities. J
Air Pollut Control Assoc. 25:251-255, 1975.
46. Koyama, T. Gaseous metabolism in lake sediments and
paddy soils and the production of atmospheric methane
and hydrogen. J. Geophys. Res. 65:3971-3973, 1963
47. Laity, J L , I. G. Burstam, and B. R. Appel Photochemical
smog and the atmospheric reactions of solvents. Shell Oil
Company Paper No. P-1957, Presented at the National
American Chemical Society Meeting, Washington, D.C.,
September 12-17, 1971,
48 Levy, A., and S. E. Miller Paint Industry Smog Chambers.
Final Technical Report on the Role of Solvents in
Photochemical Smog Formation National Paint, Varnish,
and Lacquer Association, Inc., Washington, D C , 1970,
49, Lodge, J. P., Jr, and J. B. Pate. Atmospheric gases and
particulates in Panama. Science 753:408-410, 1966
50. Lonneman, W.A. Ozone and hydrocarbon measurements
in recent oxidant transport studies In: International
Conference on Photochemical Oxidant Pollution and Its
Control. Proceedings. Vol. I. B. Dimitriades, ed EPA-
600/3-77-001 a, U.S Environmental Protection Agency,
Research Triangle Park, N.C., January 1977 pp 211-
223.
51. Lonneman, W. A., S L Kopczynski, P. E. Darley, and F D
Sutterfield. Hydrocarbon composition of urban air
pollution. Environ Sci. Technol 8:229-236,1974
52. Manufacturing Chemists Association, Inc., and Public
Health Service. Atmospheric Emissions from Nitric Acid
Manufacturing Processes. Public Health Service
Publication No. 999-AP-27, U.S. Department of Health,
Education and Welfare, Public Health Service, National
Center for Air Pollution Control, Cincinnati, Ohio, 1966.
53. Mayrsohn, H., and J H Crabtree. Source Reconciliation
of Atmospheric Hydrocarbons State of California, Air
Resources Board, Sacramento, Calif March 1975.
54 McConneli, J. C , and M B. McElroy. Odd nitrogen in the
atmosphere. J Atmos. Sci. 30:1465-1480, 1973.
55 McEwen, D. J The analysis of gasoline vapors from
automotive fuel tanks Paper presented at the 153rd
National Meeting of the American Chemical Society,
Miami Beach, Florida, April 1 967
56. McReynolds, L A., H. E. Alquist, and D B. Wimmer
Hydrocarbon emissions and reactivity as functions of fuel
and engine variables. SAE Trans, 74:902-911, 1966.
57. Miller, D. F., A. Levy, and W. E. Wilson, Jr A Study of
Motor Fuel Composition Effects on Aerosol Formation
Part II. Aerosol Reactivity Study of Hydrocarbons. API
Project EF-2, Battelle Memorial Institute Laboratories,
Columbus, Ohio, February 21, 1972.
58. Morris, W. E , and K. T. Dishart. Influence of vehicle
emission control systems in the relationship between
gasoline and vehicle exhaust hyarocarbon composition
In. Effect of Automotive Emission Requirements on
Gasoline Characteristics. ASTM Special Technical Publ.
90
-------
No 487, American Society for Testing and Materials,
Philadelphia, Pa, 1971. pp. 63-101.
59. MSA Research Corp. Hydrocarbon Pollutant System
Study Volume I Publication No APTD-1499, U.S
Environmental Protection Agency, Research Triangle
Park, N.C, 1972.
60 National Air Pollution Control Administration. Air Quality
Criteria for Hydrocarbons. NAPCAPublicationNo AP-64,
U S Department of Health, Education and Welfare,
Public Health Service, Washington, D.C., March 1970
61 National Research Council Vapor-Phase Organic
Pollutants National Academy of Sciences, Washington,
D C, 1975
62. Neligan, R. E Hydrocarbons in the Los Angeles
atmosphere A comparison between the hydrocarbons in
automobile exhaust and those found in the Los Angeles
atmosphere Arch Environ. Health 5-581-591, 1962.
63. OAQPS Data File of Nationwide Emissions U.S,
Environmental Protection Agency, Office of Air Quality
Planning and Standards, Research Triangle Park, N C.,
August 1976. (These reports are standard computer
reports available from the National Air Data Branch,
Monitoring and Data Analysis Division, OAQPS )
64. Oberdofer, P E The determination of aldehydes in
automobile exhaust gases SAE Tech. Pap No 670123,
Society of Automotive Engineers, Inc., New York, 1967
65 Office of Air Quality Planning and Standards
Compilation of Air Pollutant Emission Factors. Part B.
Third Edition AP-42, U S Environmental Protection
Agency, Research Triangle Park, N C., February 1976
66 Officeof AirQualityPtannmgandStandards NationalAir
Quality and Emissions Trends Report, 1 975 EPA-450/1-
76-002, U S Environmental Protection Agency,
Research Triangle Park, N C., November 1976
67 Pitts, J N,Jr,A M Winer, K R Darnall, A C.Lloyd, and
G J Doyle Hydrocarbon reactivity and the role of
hydrocarbons, oxides of nitrogen, and aged smog in the
production of photochemical oxidants In International
Conference on Photochemical Oxidant Pollution and Its
Control. Proceedings Vol II B Dimitnades, ed EPA-
600/3-77-001 b, U.S Environmental Protection Agency,
Research Triangle Park, N.C, January 1977 pp 687-
704
68 Rasmussen, R A Terpenes Their Analysis and Fate in
the Atmosphere Ph.D Thesis, Washington University,
St Louis, Mo , 1964
69 Rasmussen, R A , R B Chatfield, M W Holdren, and E
Robinson Technical Report on Hydrocarbon Levels in a
Midwest Open-Forested Area Submittedto Coordination
Research Council Washington State University, College
of Engineering, Pullman, Wash , October 1 976
70 Rasmussen, R A., and F W Went. Volatile organic
material of plant origin m the atmosphere Proc Natl
Acad Sci. U S.A 53 215-220, 1965
71 Research Triangle Institute. Investigation of High Ozone
Concentration in the Vicinity of Garrett County, Maryland
and Preston County, West Virginia EPA-R4-73-019, U.S.
Environmental Protection Agency, Research Triangle
Park, N C., January 1973
72 Research Triangle Institute Investigation of Rural
Oxidant Levels as Related to Urban Hydrocarbon Control
Strategies EPA-450/3-75-036, U.S Environmental
Protection Agency, Research Triangle Park, N.C , March
1975
73. Robinson, E., and R. C Robbms. Gaseous nitrogen
compound pollutants from urban and natural sources J.
Air Poltut Control Assoc. 20'303~306, 1970.
74. Robinson, E, and R C Robbms. Sources, Abundance,
and Fate of Gaseous Atmospheric Pollutants SRI Project
PR-6755, Stanford Research Institute, Menlo Park, Calif,
1968.
75. Seizmger, D. E , and B, Dimitnades. Oxygenates in
Automotive Exhausts. Effect of an Oxidation Catalyst
Report of Investigations Rl 7837. U.S. Department of the
Interior, Bureau of Mines, Washington, D.C , 1973.
76 Seizmger, D E, and B Dimitnades Oxygenates in
exhaust from simple hydrocarbon fuels J. Air Pollut.
Control Assoc. 22.47-51, 1972.
77 Spicer, C W,J L Gemma, D.W Joseph, P R. Sticksel,
and G F Ward. The Transport of Oxidant Beyond Urban
Areas. EPA-600/3-76-01fla, U S, Environmental Pro-
tection Agency, Research Triangle Park, N.C., February
1976
78 Stephens, E R., and R F. Burleson. Distribution of light
hydrocarbons m ambient air. Presented at 62nd Annual
Meeting, Air Pollution Control Association, New York,
June 22-26, 1969.
79 Stephens, E R , E. F Darley, and F. R. Burleson Sources
and reactivity of light hydrocarbons in ambient air Proc.
Div Refm Am Pet Inst 47466-483, 1967
80. Tannahill, G K The hydrocarbon/ozone relationship in
Texas In: Specialty Conference On: Ozone/Oxidants—
Interactions with the Total Environment. Air Pollution
Control Association, Pittsburgh, Pa , 1976 pp. 26-37
81 Tnjonis, J. C , and K W Arledge Utility of Reactivity
Criteria in Organic Emission Control Strategies.
Application to the Los Angeles Atmosphere EPA-600/3-
76-091, US Environmental Protection Agency,
Research Triangle Park, N C , August 1976
82. Tnjonis, J , T Peng, G. McRae, and L Lees. Oxidant and
precursor trends in the metropolitan Los Angeles region.
In International Conference on Photochemical Oxidant
Pollution and Its Control. Proceedings.Vol.il B. Dimitriades,
ed EPA - 600 / 3-77-001 b, U.S. Environmental Protection
Agency, Research Triangle Park, N. C., January, 1 977. pp.
1077- 1093.
83 Wade, D T. Factors influencing vehicle evaporative
emissions SAE Trans 76 (sect 2)'811-823, 1968
84 Westberg, H Review and analysis. In. International
Conference on Oxidants, 1976—Analysis of Evidence
and Viewpoints Part IV The Issue of Natural Organic
Emissions EPA-600/3-77-116, US Environmental
Protection Agency, Research Triangle Park, N.C , October
1977
85 Whitehead, L W , and R Severs Background
hydrocarbon levels in east Texas. Presented at the 83rd
Annual National Meeting, American Institute of
Chemical Engineers, Houston, Texas, March 1977 Paper
No 21A
86. Wigg, E.E.R J Campion, and W L Petersen. The effect
of fuel hydrocarbon composition on exhaust hydrocarbon
and oxygenate emissions SAE Tech Pap No 720251,
Society of Automotive Engineers, Automotive Engineers,
Inc., Warrendale, Pa , 1972
87 Williams, M. E , J T. White, L A Platte, and C J Domke
Automobile Exhaust Emission Surveillance—Analysis of
the FY 72 Program EPA-460/2-74-001, US.
91
-------
Environmental Protection Agency, Ann Arbor, Mich ,
February 1974.
88 Wilson, K. W , and G. J Doyle Investigation of
Photochemical Reactivities of Organic Solvents APTD-
1232, U.S. Department of Health, Education and Welfare,
Public Health Service, Cincinnati, Ohio, September 1 970.
89. Wodkowski, C S.,andE E Weaver. The effects of engine
parameters, fuel composition, and control devices on
aldehyde exhaust emissions Presented at the Meeting of
the West Coast Section, Air Pollution Control
Association, San Francisco, Calif, October 1970.
90. Yanagihara, S., I. Shimada, E. Shinoyama, F. Chisaka,
K.Saito, and T.lshii. Photochemical Reactivities of Organic
Solvents. Environment Agency, Tokyo, Japan. Undated.
92
-------
6. RELATIONSHIPS BETWEEN AMBIENT OXIDANTS AND
PRECURSOR EMISSIONS
INTRODUCTION
This chapter examines the functional rela-
tionships observed or theoretically derived
between photochemical oxidants (Ox), including
ozone (Os), and the two classes of oxidant
precursors, hydrocarbons (HC) and nitrogen oxides
(NOX). These Ox/HC/NOx relationships are
extremely important in that they constitute the
basis of the methods for predicting impacts of
emissions on air quality.
Since it is now well established that the impact
of emissions is not confined solely and entirely
within the emission source area, it is necessary to
consider emission distributions and related
processes that affect the requisite 0X/HC/NOX
relationship. One concept used to clarify this
geographic distinction of the emissions effects is
the source-receptor relationship, that is, the
spatial and temporal relationship between the area
and time of emission discharge and the
corresponding area and time of oxidant oc--
currence. Thus the requisite Ox/HC/NO*
relationship(s) must be applicable to the source-
receptor relationship encountered in the real
atmosphere.
Additional demands on the 0,/HC/NOX
relationships are posed by considerations of utility
rather than validity. For example, it would be useful
to have valid Ox/HC/NOx relationships for small-
scale applications such as the prediction of
localized impact from addition or deletion of an
emission source within an urban area.
All of these requirements must be met by the
0X/HC/NOX relationships if such relationships are
to serve their purposes ideally. In reality, however,
such ideal validity and utility are unattainable for
various reasons. Two of the most important
reasons are (1) it is extremely difficult to pinpoint
the areas of the specific emission sources that are
responsible for the oxidant problem observed in a
given locality, and (2) the proposed 0X/HC/NOX
relationships cannot be validated ideally or directly
(i.e., by using real atmosphere data without
substantial uncertainty). These reasons led to the
adoption of the following two premises: first,
perfectly accurate relationships will never be
obtained; and second, the development of usable
0X/HC/NOX relationships will have to be pursued
in the form of an iterative process in which
judgment developed from and during the use of
less accurate relationships is fed back into the
development of improved ones.
Consistent with these premises, relationships
between oxidant/ozone-related air quality and
emissions have been pursued following three
distinct approaches that differ mainly in degree of
empiricism. In order of decreasing empiricism,
these approaches are as follows:
1. Empirical Approach. This approach entails
statistically or nonstatistically associating
ambient oxidant-related air quality data
either with ambient concentrations of
precursors or with precursor emission
rates. These associations are clearly not
cause-effect in nature, and their intended
use is not to predict absolute air quality;
rather, it is to estimate changes in air
quality resulting from changes in emission
rates.
2. Mechanistic Models of OX/HC/NOX. This
approach entails deriving cause-effect
relationships between oxidant and pre-
cursors through laboratory testing and
chemical mechanistic simulations. As in
the preceding case, this approach is
intended to predict only changes in air
quality resulting from changes in emission
rates.
3. Air Quality Simulation Model (AQSM)
Approach. This approach entails deriving
the requisite air quality-emission rela-
tionships through mathematical repre-
sentation of the transport, dispersion,
transformation, and deposition processes.
93
-------
Its intended use is to predict absolute levels
of air quality from a given emission rate and
meteorological data.
The oxidant-emissions relationships or models
developed to date through all these approaches are
applicable only to the urban oxidant problem, and
more specifically, to situations in which the
geographical dimension of the source-receptor
relationship is comparable to that of the urban
area. Models relating local emissions and oxidant
concentrations (caused by these emissions) in
distant downwind areas are currently under
development.
Currently available models have received only
limited testing; their validity, therefore, is only
limited. Furthermore, assessments made of these
models do not reflect a consensus of opinion. The
problems in validating air quality models are
considered to be prohibitive by some investigators,
but not insurmountable by others.14
The following sections will cover the subject of
the oxidant/ozone-emissions relationships by
describing and discussing first the simpler
empirical models and then the more sophisticated
ones. Consistent with the previously stated
premise that such models will never be perfect, the
emphasis in the following discussions will be
placed on the relative merits and drawbacks of the
various models rather than on their absolute
validity and utility. For a more comprehensive
coverage of models for predicting air quality, the
reader is referred to a recent report by the National
Academy of Sciences.38
It should be noted that of the Ox/HC/NOx
relationships currently derived and reported, some
pertain to oxidant, as measured by the potassium
iodide method (see Chapter 7), and some pertain to
ozone. Relationships derived from the chemical
mechanism of the atmospheric photochemical
process pertain to ozone.
MODELS BASED ON EMPIRICAL
RELATIONSHIPS
Rollback Model
To relate precursor emission rates to oxidant or
ozone air quality, it is necessary that two main
processes bequantitated: (a)thecombined process
of dispersion and sink-removal of precursor
emissions, and (b) the combined process of
photochemical formation and sink-removal of
oxidant/ozone. One proposed method for
quantitating the dispersion-removal process is by
assuming a simple linear rollback model expressed
by the following equation:7
c, = b + ke
(6-1]
where c, is the concentration of a pollutant at point
i in the ambient air, b is the background con-
centration of the pollutant, e is the pollutant
emission rate, and k is a constant dependent on
meteorology, location of sources relative to point i,
and on other factors (e.g., sinks) affecting the
impact of the sources at point i. Equation (6-1),
after processing and simplification,7 yields
equation (6-2):
(gf) (PAQ) - b (6-l2)
where R is the percentage reduction needed to
achieve the standard; PAQ and Std denote present
and desired air quality (standard), respectively, in
terms of pollutant concentration; and gf is the
emission growth factor.7 In the case of oxidant or
ozone, the values for PAQ and Std are the second
highest oxidant or ozone concentration observed in
the base year and 160 fjg/m3 (0.08 ppm),
respectively. By inserting those values and values
for gf and b, calculation can be made of the percent
reduction of oxidant/ozone needed to achieve the
air quality standard of 160 fjg/m3 (0.08 ppm) of
ozone. If ambient oxidant/ozone concentration is
assumed to be proportional to the reactive organic
emission rate, then the calculated oxidant/ozone
control requirement equals numerically the con-
trol requirement for reactive organic emissions.
Reactive organics are usually meant to include all
hydrocarbons except methane; however, other
reactivity definitions have been used in which
other organics as well as methane were exempted
as being nonreactive.60
The preceding method for relating emissions to
air quality is known as the simple or linear rollback
model. It has several limitations, the main ones
arising from the the assumptions used in the
quantification of both the dispersion process and
the photochemical process.7 Thus the dispersion-
related equations (6-1) and (6-2) cannot be
validated experimentally, since such validation
would require that emission rates for each and
every source be changed (reduced) identically, a
requirement that obviously cannot be met in real
situations. Also, the assumption of proportionality
between oxidant and reactive organic emission
rates is of questionable validity, as it is not
supported by either smog chamber11 or theo-
retical evidence.37'47 Finally, the method ignores
the role of the NO* precursors, which have been
94
-------
established as exerting an important role. Despite
these and other limitations, the simple rollback
method has received attention56'57 mainly because
of its computational simplicity and its relatively
small demands for input information. In terms of
accuracy and validity, this model is clearly the
crudest of those currently available.
Modified Rollback: Observational Model
The main difference between simple and
modified rollback models is in the quantification of
the photochemical process. Thus, although in the
simple rollback the percent control needed for HC
emissions is taken to be equal to the percent
control needed for oxidant/ozone, the modified
rollback model utilizes an oxidant-to-precursor
dependency derived from aerometric data. This
latter derivation and resulting oxidant-to-
hydrocarbon dependency is often referred to as the
observational model and is the quantitative basis
of EPA's first proposed method, the Appendix-J
method, for calculating oxidant-related control
requirements.59 The observational model has been
described and discussed in detail elsewhere;1'50
therefore, only a brief description will be included
here.
The observational model is based on the
assumption that early morning precursor
concentrations are indicators of the oxidant levels
that will occur later in the day. Consistent with this
assumption, aerometric data taken in several
urban centers were used to draw the upper limit
curve shown in Figure 6-I. Specifically, this curve
was constructed by plotting daily maximum l-hr
oxidant concentrations against 6- to 9-a.m.
average nonmethane hydrocarbon (NMHC)
concentrations. The curve was drawn through the
uppermost points. Note that both the oxidant and
NMHC data were taken at the same monitoring
site—Continuous Air Monitoring Program (CAMP)
site—invariably located in the downtown area of
the respective urban center. Thus the upper limit
curve is taken to depict the relationship between
the hydrocarbon precursor and the oxidant formed
within source-intensive areas under those
meteorological conditions that are most conducive
to oxidant formation (i.e., clear skies, high tem-
perature, etc.).
Accepting the upper limit curve as depicting the
quantification of the photochemical process,
control requirements for oxidant reduction can
then be calculated using the rollback equation (6-
2) as follows: From the curve in Figure 6-1, onecan
read off the NMHC value corresponding to PAQ
(that is, to the second highest oxidant
concentration observed in a location) as well as the
NMHC corresponding to the standard (that is, to
160 fJQ/m3, or 0.08 ppm). This latter NMHC was
determined to be 0.24 ppm C, and it is the air
quality standard for hydrocarbon (to be used as a
guide only and not as a true air quality standard).58
By using these NMHC values and assuming b = o
(i.e., no background oxidant or NMHC),58 the
percentage control required for NMHC emissions
was calculated as a function of oxidant
concentration, and the resultant function was
published by EPA in the form of the Appendix-J
curve.53 A somewhat different approach was used
by Schuck and Papetti to construct an upper limit
curve specific for Los Angeles.60 By this latter
approach, the NMHC values used were averages of
data taken at several sites in the Los Angeles
Basin; whereas, the oxidant values used were
those for the highest oxidant concentration
observed anywhere in the Basin. This Los Angeles
curve is shown in Figure 6-2.
This modified rollback, or Appendix-J, model has
been critiqued extensively.
9,13,24
Briefly, the model
has limitations and advantages related (1) to the
rollback equation (6-2), (2) to the assumption of
zero background oxidant or NMHC, and (3) to the
upper limit curve. Limitations and justification of
the rollback equation are discussed in the
preceding section. The assumption that there is no
background oxidant or NMHC is clearly incorrect,
but it has been adopted in the interest of simplicity
and because reliable data on the level of such
background are lacking. The limitations and
advantages of the upper limit curve are
summarized as follows:
1. The curve probably depicts the dependence
of oxidant on the dispersion factor rather
than the dependence on the hydrocarbon
reactant factor.13
2. The curve depicts a purely empirical
relation, not a cause-effect one, and it,
therefore, cannot automatically be as-
sumed to have predictive value; (it would
have a more cause-effect nature if oxidant
were measured within the same air mass in
which the HC and NO, measurements were
made).
3. The curve disregards the NOX factor.
4. Experimental error makes the low end of
the curve (air quality standard for NMHC)
highly uncertain.
95
-------
5. Scarcity of data points makes the upper part
of the curve highly uncertain.
6. The curve is not necessarily valid in
locations other than those from which it
was derived.
7. The curve disregards the oxidant transport
phenomena.
8. The curve has no statistical nature.
Justification of the upper limit curve is mainly
based on the following.
3.
The curve is derived from real atmospheric
data, which is a more realistic alternative to
laboratory smog chamber data, for
example.
The curve is in qualitative agreement with
the smog chamber data, at least insofar as
the oxidant-to-hydrocarbon dependence is
concerned (Figure 6-3).
The curve can be improved with acquisition
of additional data.
0.30
0.25
0.20
a
a
a
x
o
0.15
0.10
0.05
APPROXIMATE UPPER LIMIT
OBSERVED OXIDANT
/. . V .*
•••*•••• •
4*. • • • v
••• *. ;..:•: ••••
0.5
1.0
1.5
2.0
2.5
NOMMETHANE HYDROCARBONS, ppm C
Figure 6-1. Maximum daily 1-hr average oxidants as a function of 6-to-9-a.m average of
nonmethane hydrocarbon (CAMP data from four U.S. cities).150
96
-------
O
<
EC
60
50
40
2 30
O M
£
E
X
20
10
i r
AVG. NONMETHANE HYDROCARBONS
AT EIGHT STATIONS (1971), ppm C
Figure 6-2 Upper limit oxidant values in the Los Angeles south coast air basin as a function of
average 6-to-9-a.m, hydrocarbon concentrations.60
0.90
0.80
0.70
0,60
9 0.50
8
5 0.40
s
E
X 0.30
0.20
0.10
SMOG CHAMBER DATA HC/NOx = 12
SMOG CHAMBER DATA NO = 0.4 ppm
AEROMETRIC DATA (UPPER LIMIT CURVE)
SMOG CHAMBER DATA HC/NO = 3.5
0 1.0 2.0 3.0 4.0
NONMETHANE HYDROCARBON, ppm C
Figure 6-3. Oxidant-hydrocarbon relationships from smog chamber and aerometric data.13
5.0
97
-------
Overall, the Appendix-J method is judged to be
no more valid than the simple rollback model
because it is based on relatively few data taken
from several cities. The method has more validity
when it is derived from and used in one and the
same locality, as, for example, in the case of the
Schuck-Papetti curve for Los Angeles. This,
however, requires an abundance of data, provided
by multiyear rather than multistation meas-
urements. Such data are currently available only
for a very few cities.1 Lack of adequate data,
ambient measurement error, and influences
associated with siting of the monitoring stations
are probably the main reasons for the rather large
differences observed among individual location
upper limit curves.61
Statistical-Empirical Models
Statistical-empirical models are based on
statistically defined associations among ambient
air quality indices, meteorological parameters, and
emissions rates. A considerable amount of work
has been done in establishing such associations.
However, much of that work was addressed to the
air-quality-to-meteorology dependency and the
development of.predictive models for short-term
episode forecasting. Such models are not relevant
to the subject of this document. Relevant models
are those addressed to the air-quahty-to-emission
dependency—that is, models intended to predict
long-term impact of emission controls on oxidant
air quality and to calculate control requirements
for oxidant reduction. The most notable of these
models were recently reviewed by Myrabo et al.35
The following discussion of such models is based
partly on the Myrabo review.
With one exception, the statistical-empirical
models relating ambient oxidant to precursors
neglect the spatial and temporal distribution of the
precursor emissions. Accordingly, their utility is
limited to relating changes in total levels of
emissions within a region to changes in region-
wide air quality.
Merz et al.33 used regression analysis to relate
daily 1-hr maximum oxidant to 6- to 9-a.m.
concentrations of NOX and NMHC, with the latter
taken to be 50 percent of total hydrocarbon. The
data used were for downtown Los Angeles, and
only for the months of August, September, and
October. Results are shown in Figure 6-4. The
diagrams of Figure 6-4 indicate that NO* control
would have a slight but beneficial impact on
oxidant air quality.
Kmosian and Paskmd26 also used July-
September data from Los Angeles to relate,
through regression analysis, daily maximum 1 -hr
oxidants to 6- to 9-a.m. NO, for constant NMHC.
The NMHC data, again, were not obtained directly;
they were computed from total hydrocarbon (HC)
data using establishedHC-to-NMHCrelationships.
Results from four Los Angeles sites are shown in
Figures 6-5 through 6-8. These results, unlike the
Merz et al. model,33 indicate that NO, control,
unless drastic, is detrimental to oxidant air quality.
Finally, Trijonis55 used a stochastic model to
relate emission rates in downtown Los Angeles
(DOLA) to 7:30-9;30-a.m. ambient concentrations
of hydrocarbon and NO*, which were then related
to the oxidant concentrations observed in the
sections of the Los Angeles Basin lying downwind
from DOLA. Specifically, Trijonis used average
values of oxidant concentrations measured at
DOLA, Pasadena, and Burbank, weighted
according to direction and speed of the morning-
to-noon wind. Trijonis expressed his results in
terms of annual exceedance of the California
oxidant standard (0.1 ppm as ozone) as a function
of NO, and reactive hydrocarbon emission levels in
DOLA (Figure 6-9).55 These results again show NO,
control to have a detrimental impact on oxidant air
quality, at least for Los Angeles.
All three models discussed in the preceding
paragraphs are not capable of treating spatial
resolution of emissions. Thus, by those models, a
region or urban center is treated as a single point
source, a simplification that is obviously helpful,
but that is also penalizing in that it makes it
impossible to deal with questions of localized
impact from addition or deletion of emission
sources within the urban area. To lessen this latter
penalty, investigators from the University of
California, San Diego, offered a statistical-
empirical model with some spatial resolution
incorporated.5'34 Specifically, their model related
oxidant air quality (expressed in terms of number of
hourly violations of the Federal oxidant standard of
0.08 ppm as ozone) to reactive hydrocarbon
emission levels integrated over the prevailing
windstream corridor leading to the oxidant
monitoring station. Spatial resolution was
achieved by defining the windstream corridor
width to be approximately one-tenth of the
maximum dimension of the urban area. Using data
from 1 7 stations in San Diego and Los Angeles, the
investigators generated a hydrocarbon-emission-
to-air-quality relationship that they then proposed
98
-------
1.0
a.
a.
*x
O
QJ>
6
10
01
DATA BOUNDARIES
t
t
0.25
\
0.20
\
0.15
MAX 1-hrOXIDANT
(ppm)
0,10
[ t 0.05 \ 0.08
0.1
1.0
10
6 to-9-a.m. NONMETHANE HYDROCARBONS, ppm C
Figure 6-4, Merz, Painter, and Ryason's relation of NO, and NMHC assumed as 50 percent of
total HC and oxidant for downtown Los Angeles.33
a
a.
z
O
a.
O
u
I-
i
X
O
X
I
25
20
15
10
6(3.0)
7(3.6!
10
20
30
40
50
60
70
6-to-9-a.m. OXIDES OF NITROGEN CONCENTRATION, pphm
Figure 6-5. California Air Resources Board aerometric results, relation between 6 to-9-a.m.
NO,, 6-to 9 a m HC, and maximum hourly oxidant concentrations in downtown Los Angeles.
(Individual curves show total and NMHC concentrations in ppm.)76
99
-------
for use for predictive purposes. The NO* factor,
obviously, is ignored by this model.
All the statistical-empirical models described in
the preceding paragraphs, and other similar
ones,2'38 are relatively simple in that they treat the
conversion of emissions into ambient oxidant as
one process. Thus distinct components of this
process such as the vertical dispersion of
emissions within a definable atmospheric layer
(mixing layer), local (intracity) transport by wind,
and chemical reactions are not delineated and
treated separately. Because of this simplification,
these models lose much of their validity when
applied to areas with characteristics widely
different from those of the area for which they
were developed (for example, mixing heights, wind
patterns, or photochemical conditions). To reduce
this problem, Tiao et al." developed a model that
treats the dispersion, advection, and reaction
processes separately, using separate terms in
mass conservation equations for NO, NOa, and
ozone. The coefficients of these terms were
determined statistically by fitting the equations to
the observations available. Because of its
mechanistic detail, the Tiao et al. model54 is
inherently more valid than the preceding models in
treating diverse situations. However, because it
does not include spatial resolution, this model, like
the preceding ones, is not applicable in areas
where a major fraction of the local oxidant is
caused by extraneous sources.
10
20
40
10
20
30
40
6-to-9-a.m. OXIDES OF NITROGEN
CONCENTRATION , pphm
6-to 9 a.m. OXIDES OF NITROGEN
CONCENTRATION, pphm
Figure 6-6 California Air Resources Board aerometric
results, relation between 6-to-9-a.m. ISIO». 6-to-9-a.m. HC,
and maximum hourly oxidant concentrations in Azusa.
(Individual curves show total and NMHC concentrations in
ppm )26
Figure 6-7. California Air Resources Board aerometric
results, relation between 6-to-9-a.m. NO», 6-to-9-a.m. HC,
and maximum hourly oxidant concentrations in San
Bernardino. (Individual curves show total and NMHC
concentration in ppm.)2B
100
-------
20
30
40
6-to-9-a.m. OXIDES OF NITROGEN
CONCENTRATION, pphm
Figure 6-8. California Air Resources Board aerometric
results, relation between 6-to-9-a.m. NO,, 6-to-9-a.m. HC.
and maximum hourly oxidant concentrations in Anaheim.
(Individual curves show total and NMHC concentrations in
ppm.)26
The statistical-empirical models have ad-
vantages arising from the fact that the models are
constructed from real atmospheric data and are
statistical in nature. Furthermore, such models are
computationally simple and inexpensive, and
unlike the rollback models, most of them consider
both precursor factors, HC and NO*.
Limitations of the statistical-empirical models
arise mainly from the stochastic nature of such
models. Because they are based on associative
rather than cause-effect relationships, such
models have limited predictive validity, especially
when predictions call for extrapolating the model
beyond the range of data from which it was
derived. Furthermore, the models that relate
oxidant to ambient concentrations rather than to
the emissions of precursors are subject to the
same validity limitation astheAppendix-J method,
namely, that the model does not directly relate air
quality to precursors. Instead, the direct rela-
tionship is between oxidant and the atmospheric
dispersion factor. In general, the statistical-
empirical models have less validity when they are
derived from and are applied to a variety of areas
with significantly different meteorology and
emission characteristics.
In conclusion, when all the strengths and
limitations of the statistical-empirical models are
evaluated, such models are judged to be superior
to the rollback models, especially when^they are
designed for local use.
MECHANISTIC MODELS OF Ox/HC/NOx
RELATIONSHIPS
The development of reaction mechanisms
describing the photo-oxidation processes of
hydrocarbons and oxides of nitrogen has evolved
extensively through the use of experimental data
from smog chamber studies.10 A basic
understanding of the 0X/HC/NOX chemical
relationship is a prerequisite to developing
techniques for relating emission to ambient
oxidant levels. It is important, therefore, that smog
chamber studies be performed under conditions as
similar as possible to those of real polluted
atmospheres. Thus, considerations such as light
intensity and its spectral distribution, temperature,
humidity, and composition and concentration of
reactants should be comparable to those in typical
polluted atmospheres. The utility of one such smog
chamber data set exhibiting many of the above-
mentioned attributes has been demonstrated via
the development of a chemical kinetic model for
predicting impacts of emission change on air
quality and for calculating ozone-related control
requirements.8 This technique, given the acronym
EKMA for Empirical Kinetic Modeling Approach, is
one of several discussed by EPA in their report,
Uses, Limitations and Technical Basis of Pro-
cedures for Quantifying Relationships Between
Photochemical Oxidants and Precursors.39 Since
application of the technique for air quality control
is covered in detail in this work, only the scientific
basis of the technique will be considered here.
EKMA is based on the ozone-to-precursor
relationship <03/HC/NOX) derived from a smog
chamber study of automotive exhaust mixtures.11
The selection of this study over several other smog
chamber data sets in existence 11'40 was primarily
based on the rather representative nature of the
experimental conditions to that of urban polluted
atmospheres.10 Although the quality of the
101
-------
selected data set is considered high, deficiencies
nevertheless still exist, as follows:
1. The selected data are limited to those for
hydrocarbon-to-NOx reactant ratios
ranging from 1:1 to 1 2:1, a range that does
not extend to sufficiently high values.
2. The ozone yield data are erroneously high
for reactant concentrations close to zero,
because at such low concentrations, the
chamber background reactivity becomes
important relative to the reactivity of the
test mixture.
-3. The selected data were obtained under
fixed radiation intensity conditions in
contrast to the diurnally varying sunlight
intensity in the real atmosphere.
Consideration of these and other minor
deficiencies as well as practical constraints on the
number of parameter variations that can be
studied in smog chambers lead to the utilization of
a photochemical kinetic mechanism to simulate
and project the experimental data. The detailed
photochemical mechanism wasfirstfittothe smog
chamber data on automotive exhaust mixtures and
then used to extrapolate ozone yields for varying
initial hydrocarbon and NOX concentration and
irradiation conditions. Resulting Os/HC/NO*
relationships for conditions similar to those in the
Los Angeles atmosphere during the smog season
are depicted in Figure 6-10 in terms of a family of
ozone isopleths.
•o
o
g
55
in
I
LJ
z
o
CD
cc
<
o
o
in
in
a:
z
o
u
1200 -
1000 -
800 -
600 -
400 -
200 -
400
800
1200
1600
L. A. COUNTY NOX EMISSIONS, tons/day
Figure 6-9. Expected number of days per year exceeding 0.10 ppm versus NO. and reactive
hydrocarbon emission levels for central Los Angeles.55
102
-------
OXIDANT/O3, ppm
0.08 0.20 0.30 0.40 0.50 0.55 0.60 0.65
i
a.
i
x
O
z
0.2 —
0.1 _
1.0
2,0
NMHC, ppm C
Figure 6-10. Oxidant/Oa isopleths derived from combined use of smog chamber and
photochemical modeling techniques.16
The chemical kinetic modeling technique16 used
to derive the isopleths of Figure 6-10 is based on a
75-reaction-step mechanism for a propylene and
n-butane hydrocarbon mixture. The two-
component hydrocarbon system is a surrogate for,
the automotive exhaust hydrocarbons, which in
turn is a surrogate for atmospheric organic
pollutant mixtures. Seventeen smog chamber
experiments were modeled in which the
agreement between observed and model-
predicted ozone yields was maximized by
optimizing the butane-to-propylene ratio in the
model. Chamber background reactivity was
simulated in the model by assuming a source of
propylene resulting from desorption from the
chamber walls. After optimization, the model was
extended to consider conditions more realistic to
urban polluted atmospheres. Major changes in the
photochemical kinetic model included removal of
the chamber background reactivity steps,
consideration of diurnally varying photolytic
reaction rate constants, simulations over 9-hr
irradiation periods corresponding to the hours
between 7 a.m. and 4 p.m., and consideration of
dilution rate more appropriate to that of an urban
atmosphere.
With these adjustments, the model-generated
isopleths shown in Figure 6-10 are taken to
represent roughly the atmospheric situation in the
Los Angeles Basin. The approximate nature of the
representation arises since the isopleths, despite
the model adjustment, reflect biases caused by the
simplifications and possible perturbations
inherent to the smog chamber experiment. Thus,
for example, the absence of continuous fresh
emission influx in the smog chamber simulation is
one simplification that may have introduced a bias.
The isopleths may also reflect biases introduced by
unknown effects of the chamber walls, by use of
unrealistic dilution and reactant mixing conditions
in the smog chamber test, and possibly by other
factors as well.
In light of these uncertainties, the ozone
isopleths derived from this model (Figure 6-10) are
not assumed to have absolute validity and
therefore should not be used to predict ambient
ozone levels from given ambient 6- to 9-a.m.
concentrations of NMHC and NO, or vice versa.
Application of the isopleths is considered, in a
relative sense, as illustrated in the following
example calculation for emission control
requirements. The detailed reasoning behind this
103
-------
calculation procedure and its advantages are given
elsewhere.11
Control requirements for achievement of the
ozone and N02 standards are calculated for a
region from the second highest 1-hr
oxidant/ozone concentration observed in the
region's atmosphere, from the NMHC-to-NO* ratio
during 6 to 9 a.m., and from the annual N02 mean
concentration. For example, if the ozone
concentration, the NMHC-to-NO* ratio, and the
annual N02 mean values are 0.4 ppm, 8.0 ± 4.0,
and 0.06 ppm, respectively, then the isopleth
counterparts of the ambient NMHC and NO,
concentrations causing the 0.4 ppm of ozone are,
on the average, those defined by the point e; that is,
the intercept of the 0.4 ppm ozone isopleth with the
constant slope line ab corresponding to a NMHC-
to-NOx ratio equal to 8.0 (Figure 6-10). Next, the
NOX reduction needed for achieving the national
ambient air quality standard (NAAQS) for N02
(annual mean of 0.055 ppm) is calculated to be 10
percent, assuming proportionality between
ambient NO* reduction and ambient N02
reduction. Reducing NO* by 10 percent defines
point e. For achievement of the NAAQS for ozone,
the NMHC must then be reduced along the line ef
to the point of being the intercept of line ef with the
0.08 ppm ozone isopleth. Such a NMHC reduction
is calculated to be 81 percent. Similar calculations
for the upper and lower limits of the NMHC-to-NO*
ratio range give NMHC control requirements equal
to 62 percent and 87 percent, the latter control
being the one that will achieve the ozone standard
regardless of NMHC-to-NO* ratio.
Note that the EKMA is applicable only in those
situations in which the source-receptor
relationship is known and temporally consistent
with that implied in the model. Thus since the
model predictions are for 9-hr irradiation of
emissions and the computed ozone peak appears
usually after 5 to 6 hours of irradiation, the model-
predicted control requirements are only for
emission sources lying within an upwind-distance
equivalent to 5 to 6 hr (at the most, 9 hr) of wind
transport. Situations in which a major fraction of
the observed ozone appears to be caused by
emissions from more distant upwind sources or
from transported ozone aloft (anthropogenic or
stratospheric in origin) cannot be appropriately
treated by this model.
The possibility of separating out the ozone
contribution of the extraneous sources and of
defining the local ozone component on which the
isopleths are applicable has been explored, but
without much success. At this time, the best
judgment that can be made on this question is only
qualitative. One of many conceivable scenarios
describing the interaction between extraneous
ozone and local ozone formation is as follows.
Local emissions are discharged in the early
morning into a relatively thin, stable layer (100 to
200 m) of ozone-free air. As the day progresses,
this inversion is eroded, and the reacting
emissions disperse into the subsidence inversion
layer that contains the extraneous ozone. By early
afternoon, vertical mixing is complete, and
pollutants (local and extraneous) are confined in a
single layer under the subsidence inversion. For
such a scenario, the extraneous ozone can be
assumed to have undergone the following
reduction processes:
1. Dilution resulting from mixing with the
ozone-free air in the radiation inversion
layer.
2. Destruction on surfaces.
3. Destruction in reaction with precursors and
other reaction mixture constituents.
In contrast, this same extraneous ozone also
enhances local oxidant formation processes. The
quantification of the net impact of the extraneous
ozone on ground-level ozone concentrations is
beyond the capabilities of the EKMA approach and
lies in the realm of the air quality simulation model.
Nevertheless, there are some indications that the
destructive processes outweigh those that
enhance ozone formation. Thus, for example,
Seinfeld performed modeling tests51 that showed
that for an urban scenario similar to the one
described here, extraneous ozone had a positive
effect on ground-level ozone concentration, but
that such an effect was less than additive.
Therefore, subtracting the entire extraneous ozone
concentration (measured upwind or aloft) from the
maximum ozone concentration observed in an
urban center may introduce substantial error in
estimating the local ozone component for which
the isopleth method is applicable. Conservative
first-order approximation estimates can be made
in areas subject to extraneous ozone by assuming
negligible impact from the extraneous source.
Further guidance regarding such calculations is
provided elsewhere.41
From the preceding discussion, it is clear that the
EKMA is not ideally applicable to situations in
which the oxidant problem is largely of extraneous
origin. The question that needs to be examined
104
-------
next is whether, with the exception of these
situations, the same set of isopleths can be used
everywhere within the continental United States.
Possible reasons for concern are that the
meteorology- and emission-related factors that
affect local ozone formation (and hence the ozone-
to-precursor relationships) differ from location to
location. Thus prevailing sunlight intensity,
temperature, and dispersion conditions in Los
Angeles are different from those in Chicago, for
example. Likewise, the organic emission
composition, organic-to-NO« ratio, and emission
discharge patterns may be different for various
urban areas. To explore the sensitivity of the EKMA
to different urban areas with widely diverse
conditions, the method was subjected to a
sensitivity test,'7 as follows. First, new sets of
isopleths were generated by changing the model
parameters related to light intensity, dilution rate,
hydrocarbon composition (i.e., propylene-to-
butane ratio), and emission discharge after 9 a.m.
These new sets of isopleths were then used to
compute emission control requirements for given
ozone reduction strategies. Results showed that
although the various sets of isopleths differed in
isopleth shape and spacing, the control
requirement estimates were nearly identical for all
of them. Thus the same set of isopleths could be
used in different locations without introducing a
substantial error in the calculation of control
requirements. However, given that such error was
substantial, it could be reduced by adjusting the.
appropriate model parameters and generating a
new set of isopleths that would more closely reflect
the prevailing conditions in the region.
Note that the sensitivity to parameter variations
is that of the model and may not necessarily reflect
responses of the real atmosphere. Therefore, this
technique, like all others, must be verified as to its
predictive accuracy.
To complete the discussion on the modeling
Oa/HC/NO, relationships, it should be added that
a model has been proposed by Los Angeles County
investigators12 that combines smog chamber and
aerometric data. The method has been discussed
in detail by Souten et al.,52 who have pointed out
several weaknesses in the approach. Because of
these weaknesses, the model has not received
much attention.
In concluding this discussion on the mechanistic
Oa/HC/NO, models, it is stressed that such
models, although clearly superior to the rollback
and statistical models, are also acknowledged to
have several imperfections, as discussed
elsewhere.8'11'16 Briefly, the specific model
described here, (EKMA), has a conceptual
limitation arising from the fact that the ambient
atmosphere, with all its complexity, cannot be
duplicated in the smog chamber. Furthermore, the
model does not predict frequency of occurrence of
ozone concentrations. The use of photochemical
modeling techniques in deriving the EKMA
Oa/HC/NO, relationships (Figure 6-10) has
increased the utility of these relationships, but the
photochemical model used also has its
imperfections. The EKMA model is also limited in
that it requires input information, someof which is
not available. Thus the required NMHC-to-NOx
ratio data are not commonly available, mainly
because measurement of ambient NMHC is not
required by law. Finally, the model does not have
the spatial and temporal resolution needed to treat
distribution of ozone within an urban area.
All of these imperfections are real and of
consequence, but certainly they do not altogether
prohibit the use of this model. Viewing these
imperfections in the context of the premise that a
model shall be judged mainly in relation to other
existing models, it would appear that the model
described here is somewhat superior to the others
presently ready for use. Thus many of the model's
imperfections are also shared by the rollback and
empirical-statistical models, whereas some of its
strengths are not. For example, the strongest
points of this model are the cause-effect nature of
the model's ozone-to-precursor relationships, and
the quantitative consideration of the NO,-
precursor factor. The only models that have both
these features are the air quality simulation
models. Imperfections of the photochemical model
used to supplement and adjust the smog chamber
data are not critical since the photochemical model
is validated against laboratory data. Thus, future
changes in the chemical mechanism of the
photochemical model can have only a small effect
on the isopleths and an even smaller effect on
control requirement estimates computed from the
isopleths.
AIR QUALITY SIMULATION MODELS (AQSM)
Air quality simulation models representthe most
fundamental approach to relating primary
pollutant emissions to secondary pollutant
concentrations. This approach was first conceived
and introduced by Friedlander and Seinfeld.20
Such models are based on a mathematical
105
-------
description of the physical and chemicaFprocesses
involved in the atmospheric behavior of air
pollutants. The basis of all AQSM's is the equation
of conservation of mass, which with the help of
certain simplifying assumptions is reduced to
equation 6-3, commonly known as the atmos-
pheric diffusion equation:
3d
3t
3c,
3x
3c,
w
3d
= 9_ KH^L
3x 3x
3y
3y
3z
3y
+ R, (C1 ..... Cn.T)
(6-3)
3z 3z
where c, is the mean concentration of species i;
x,y,z are the Cartesian coordinates; u,v,w are the
mean velocity components; KH,KV are the
horizontal and vertical eddy diffusivities; R, is the
rate of production (or negative of the rate of
consumption); T is temperature; and t is time.
The solution of the set of equations 6-3 yields the
theoretical mean concentrations as a function of
location and time. The solution of equation 6-3
requires, as input information, initial and boundary
concentrations of each of the n species, the wind
field, turbulent diffusivities, source emissions as a
function of location and time (entering as boundary
conditions), and a chemical reaction mechanism.
All air quality models are mathematical
descriptions of the physical processes that are
known to occur in the atmosphere. However, no
model is a perfect descriptor; it is only as true a
simulator of actual processes as the assumptions
on which it is based are valid. Since models are
known to be inaccurate, it is of interest to ascertain
how inaccurate their predictions might be.
Unfortunately, quantitative determination of the
expected inaccuracy of a model is extremely
difficult, if not impossible, to achieve. There are
several reasons for this.
First, meteorological processes are random in
nature, whereas all of the current AQSM's are
deterministic (as contrasted with stochastic).
Outputs (and observations) are thus meaningfully
expressed in terms of random variables. Predictive
results of deterministic models are not
commensurate with this requirement. Thus
comparison of actual observations with model
predictions is not entirely appropriate. The second
and related reason for the difficulty of quantitative
determination is that the atmosphere is highly
variable in terms of all measures used by scientists
to describe it. Spatial and temporal variations of
virtually all parameters that serve as inputs to
models are usually quite significant and, in fact,
often rather large. Third, the data base one might
use both for input to the model and for
comparisons with predictions is virtually always
sparse and sometimes inaccurate. Finally, it is not
possible to derive a totally theoretically based
estimate of expected errors associated with a
particular formulation, because it is not possible to
formulate a truly accurate representation of the
natural system with which to compare.
If equation 6-3 could be demonstrated to be an
accurate representation of atmospheric processes,
and if the necessary input information were known
with sufficient accuracy, the AQSM would clearly
be the preferred approach for relating primary
emissions to secondary (e.g., oxidant) air quality.
Commonly, however, there are inaccuracies
associated with the use of equation 6-3. These
inaccuracies can be categorized either as
fundamental inaccuracies or input-related
inaccuracies. The fundamental inaccuracies are
those associated with the assumptions made in
the derivation of equation 6-3. Such inaccuracies,
for example, include those related to the
assumptions made in modeling atmospheric
turbulent flow. Input-related inaccuracies are
those that result from uncertainties in the input
information such as the source emissions, wind
velocities, and the chemical reaction mechanism.
A measure of the error resulting from all sources
can be obtained through comparison of the
concentrations predicted by the model with those
measured. Ideally, such comparisons should be
made (for a particular region) for a variety of
meteorological and emission conditions. One
discrepancy between predictions and data will
arise because equation 6-3 predicts the mean
concentrations, and ambient data reflect
instantaneous conditions. Also, equation 6-3 must
generally be solved numerically on a grid, so that
the concentration, c,, implicitly involves some
degree of spatial averaging; but the data usually
consist of concentrations at a point. If a sufficient
data base is available, extensive model evaluation
over varying meteorological and emission
conditions will provide insight into the accuracy of
the model, thereby establishing the degree to
which the model can be extrapolated to conditions
beyond those in the domain of the evaluation.
106
-------
Unfortunately, a data base that is truly suitable for
model evaluation has only recently become
available (see subsequent discussion), and thus
conclusive model evaluation studies for AQSM for
photochemical oxidant have yet to be carried out.
The subject of this section is a brief review of the
capabilities and limitations of AQSM for
photochemical oxidant. More extensive
discussions of AQSM's can be found in several
recent reviews.15'38'51 AQSM's have the capability
to predict oxidant (and other secondary pollutant)
concentrations as a function of location and time
over a region for any specified meteorological and
emission conditions. Such prediction does require
extensive data input requirements, such as a
spatially and temporally resolved emission
inventory and wind field, and a chemical reaction
mechanism. The principal limitations associated
with AQSM's result from the need for extensive
input data and from inaccuracies in the input
information. Because AQSM's have not yet been
evaluated to the extent ultimately desired, it is
difficult to provide precise indications of their
ability to simulate oxidant formation accurately.
An open issue is whether or not AQSM's
presently exist that are sufficiently validated and
appropriate for application in designing oxidant
control strategies.14 This section outlines the key
aspects of this question. First, the basic types of
AQSM's for photochemical oxidant are discussed,
and several specific models that are available are
summarized. Then follows a discussion of the
general issue of the level of detail of treatments of
chemistry and meteorology in AQSM's. Finally, an
attempt is made to assess the level of accuracy of
AQSM's for photochemical oxidant. Most of the
analysis presented here is from a recent review by
Seinfeld.51
A synopsis of the types of AQSM's is presented
in Table 6-1, and several available AQSM's for
photochemical oxidant are summarized inTableS-
2. The AQSM's can be categorized according tothe
level of detail of the treatment of both mete-
orological and chemical processes. From a mete-
orological point of view, the basic distinction lies in
the level of spatial resolution of the model. So-
called grid models are based on the solution of
equation 6-3 on a three-dimensional grid repre-
senting the region of interest. An example of a
model of this type is the SAI model summarized in
Table 6-2. There is also a class of grid models
wherein explicit computations of vertical transport
are not carried out; the LIRAQ model exemplifies
this type. So-called trajectory models represent the
chemical and vertical transport processes taking
place in an advecting air column. Table 6-3
summarizes the meteorological treatments of each
of the four AQSM's described in Table 6-2.
TABLE 6-1. FORMS OF AIR QUALITY
SIMULATION MODELS
Form
Distinguishing features
Grid model Model is based on numerical solution of
the coupled atmospheric diffusion
equations in three spatial dimensions
on a grid over the region of interest.
Trajectory model Model is based on simulating chemistry
and vertical transport in air column
advecting with the local mean wind
velocity
Box model Model is based on simulating chemical
processes in a well-mixed region in
which no spatial inhomogeneitiesare
assumed to exist and within which
emissions are mixed instantaneously
throughout the region
With respect to chemistry, there are two types of
chemical kinetic mechanisms employed in
AQSM's: (1) lumped mechanisms, in which the
various organic precursor species are grouped,
based either on molecular structure or on
reactivity, and (2) surrogate mechanisms, in which
the organic species in a particular class, e.g.,
olefins, are represented by a single member of that
class (e.g., propylene). The high degree of chemical
detail in the surrogate mechanisms creates
computational demands that prohibit detailed
treatment of the dispersion processes. For this
reason, surrogate mechanisms cannot be
accommodated along with a detailed treatment of
the transport process. Most AQSM's use lumped
chemical mechanisms (SAI, LIRAQ, DIFKIN),
whereas the Bell Laboratories model employs a
surrogate mechanism. Which of these two
approaches is more accurate depends on the
relative errors that result from simplification of the
chemical process versus the dispersion process. In
his analysis of current AQSM's, Seinfeld51
suggested that the appropriate combination of
organic and free radical species in the Bell
Laboratories mechanism leads to a lumped
mechanism essentially similar to the lumped
mechanisms in the SAI and LIRAQ models. He
notes that there is virtually noadvantage gained as
a result of the extensive chemical detail of the Bell
model at the expense of a lack of treatment of
transport and dispersion in that model. Thus,
models that achieve a proper balance between
107
-------
chemical and meteorological detail are the most
appropriate for use in evaluating oxidant control
strategies.
Air quality simulation models represent the
preferred approach for oxidant prediction if the
necessary input information for exercise of the
model is available and accurate. Because the
necessary input information is not always
available (and may be costly to generate), and
because even when it is available it may not be of
an accuracy that justifies use of the model, it has
been necessary to utilize empirical and semi-
empirical methods for relating precursor
emissions to oxidant concentrations. These
methods have already been discussed earlier in
this chapter. (Of course, there may be certain uses
for which empirical or semi-empirical methods
may be preferred over AQSM even if accurate
input information for the AQSM were available.
These uses would generally be those for which the
cost of employing an AQSM might not be justified.)
Though considerable time and money have been
devoted to the development of AQSM's for photo-
chemical oxidant, it has generally been stated that
existing models do not possess sufficient accuracy
to be judged reliable. Unfortunately (but for good
reason), measures of the expected inaccuracies of
model predictions are rarely available. As a
consequence, the utility of AQSM has been
debated on a subjective rather than an objective
basis. The object of the remainder of this section is,
where possible, to assess the major uncertainties
in AQSM and estimate the level of accuracy of
oxidant predictions. (Whether, given an estima-
ted level of uncertainty in oxidant predictions for
AQSM's, the AQSM is to be preferred over another
method for evaluating oxidant control strategies
given an estimated level of uncertainty in oxidant
predictions for AQSM depends on the levels of
uncertainty of the empirical and semi-empirical
methods that are alternatives.)
In deriving the atmospheric diffusion equation 6-
3, the assumption of greatest concern isthatof the
turbulent mass flux [terms of the form u'cjV where (')
indicates the fluctuating component of velocity and
concentration] being set equal to the product of a
turbulent eddy diffusivity, K, and a concentration
gradient. The problem is that K is not constant;
rather it is a function of wind velocity, wind shear,
the local vertical temperature gradient, and other
variables. Thus K is a variable in time and space
and is a complex function of several other
TABLE 6-2. CURRENTLY AVAILABLE AIR QUALITY SIMULATION MODELS FOR PHOTOCHEMICAL OXIDANT52
AQSM
Description of AOSM
gA)a,22,29,43,44,46,4e,43,62 Th|s three_dimensjona| grid model is based on numerical solution of the atmospheric diffusion
equation The three-dimensional wind field is derived from ground-level measurements.
Pollutants emitted from ground-level sources are injected into the bottom layer of grid cells;
emissions from stacks are distributed among the grid cells aloft. A 36-step kinetic mechanism
derived from the Hecht/Semfeld/Dodge (1974) mechanism"'23 is used. Numerical solution is by
the method of fractional steps with advection treated by the SHASTA algorithm (Boris and Book,
1973),3 vertical diffusion and chemistry are by the Crank-Nicholson method
LIRAQ"'31 This model predicts the temporal variation of pollutant concentrations in a two-dimensional array
of grid cells Each cell is bounded on the bottom and top by the terrain and inversion base,
respectively For computational purposes, the pollutants ate assumed to be well mixed in each
cell. An empirical algorithm is used to relate the csll-averaged concentration to the
predicted ground-level value. A two-dimensional wind field is used. Pollutants emitted at ground
level and aloft are injected uniformly into the appropriate well-mixed cell A 48-step chemical
reaction mechanism, similar in nature to the Hecht/Semfeld/Dodge mechanism, is used °'23
The governing equations are solved using a modified version of Gear's method (Hindmarsh,
1974) 25
DIFKIN"'19'32 DIFKIN is a trajectory model based on a moving column of air in which vertical diffusion and
chemical reactions take place Pollutants are emitted into the appropriate vertical cell. The
column of air follows a surface trajectory interpolated from surface wind data A 1 2-step kinetic
mechanism is used. Governing equations are written in the form of ordinary differential
equations and solved using a Fade's approximation method.
Bell Laboratory This model uses three well-mixed cells in series. Wind is represented by volumetric air flow from
model"21 cell to cell Emissions are instantaneously mixed in cell A 143-step kinetic mechanism based
on detailed chemistry of propylene, formaldehyde, acetaldehyde, and propionaldehyde" \s used.
(Interaction between free radicals and aerosols is included.) Governing ordinary differential
equations are solved by Gear's method (Edelson, 1976).18
aSAI - Systems Applications, Inc , mode., LIRAQ = Livermore regional air quality model, DIFKIN = diffusion kinetics model
'The 36-step mechanism includes five reactions describing S02 oxidation
cThe 48-step LIRAQ mechanism does not include S02 oxidation
*The 143-step mechanism includes 19 reactions describing S02 oxidation
108
-------
variables. As a consequence, it is frequently quite
difficult to estimate.
Solution of equation 6-3 requires specification
of a wind field, (u,v,w). As the wind velocity varies
in space and time, its specification, whether based
on the solution of the fluid dynamics equations or
on the interpolation of a set of measurements, is
very difficult, and as a consequence, it is frequently
one of the greatest sources of inaccuracy in model
predictions. If one were to select two arbitrary
locations separated by a mile or two and monitor
the wind speed and direction on an hourly basis for
a day, the probability is considerable that one
would find significantly different readings, both in
TABLE 6-3. TREATMENT OF METEOROLOGICAL VARIABLES IN AIR QUALITY SIMULATION MODELS
Model
Mining depth
Turbulent diffusion
SAI"
LIRAO"
Bell9
Model requires a three dimen-
sional wind field. Hourly
averaged surface measurements
are interpolated to a fine
mesh. The surface field
together with the lower bound-
ary condition, w=0, is used to
derive the vertical velocities.
The model employs upper level
wind measurements and objec-
tive analysis procedures to render
the wind-field mass consistent.
Theoretical wind-shear
relationships derived by Lamb27
using the predictions of a
planetary boundary layer model
developed by Oeardorff6 are
employed. These relationships
are useful in instances when wind
measurements aloft are not
available. For situations in which
wind measurements are available
both at the surface and aloft, an
objective technique for preparing
appropriate three-dimensional
wind inputs to the model has been
developed by Reynolds."
An approximate two-dimensional
wind field is constructed using
a Gaussian weighting function
to interpolate sparse, 3-hr
average wind measurements
to grid cells. An iterative
vanational procedure is used
to refine the interpolated
values so as to render the
field mass consistent. A feature
of the technique is that it allows
a parameterized treatment of
flow through the inversion
base and around topographic
features
The mixing depth field is
developed using area wide
interpolation. Temperature
profiles from radiosondes
and acoustic sounder
measurements are the basic
inputs used to determine
the depth of the mixed layer.
Mixing depths over the
region are inferred from
sparse surface measure-
ments of temperature, ele-
vated profile data, topo-
graphy, and the temperature
advection from prevailing
flows. These data are used
to construct an interpolated
field The mixing depth is an
important input parameter for
the mass consistent wind-field
calculation
A bulk wind flow along the
axis of the cells is assumed; no
detailed vertical or horizontal
resolution is employed in the
model. Wind speed variations
throughout the day are defined
from the medians of seasonal
observation These data are
smoothed to provide a continuous
temporal wind variation.
Vertical eddy diffusivity coefficients
Kv have been derived through
use of a methodology developed
by Lamb et al.27'28 In this
procedure, which employs
flow fields predicted by
the model of Deardorff,6 particles
are released from a point and
followed as they are transported
downwind. From the particle
trajectories, it is possible to
calculate the pollutant
concentration field downwind of the
release point. Given the
concentration and mean flow fields,
the diffusivity profile is obtained
through use of optimal control
theory techniques. Horizontal
eddy diffusivity coefficient KH
specified is an input parameter.
The mixing depth is averaged
over each cell, and only three
spatial values are required These
data were determined from lidar
measurements of vertical
eerosol extent in combination
with mixing height data obtained
from other sources.
Spatial and temporal variations
of KH calculated are based on
similarity theory. Inputs are
root-mean-square dispersion,
grid cell size, energy
dissipation rate, mixing depth, and
reference height wind velocities.
Although LIRAQ is a dimensional,
vertically integrated model,
a value of Kv is required to
establish boundary conditions for
the vertical concentration profile.
Kv is determined from a power-law
profile for wind speed and the
assumption that the friction
velocity is 0.1 x the horizontal
velocity at a height of 1 m
Diffusion in a cell is instantaneous.
"SA! - Systems Applications, Inc. model, LIRAQ - Livermore regional air quality mode!. Bell - Be!! Laboratories mode!
109
-------
magnitude and direction, between the locations.
Thus an accurate model must allow for the
possibility of a laterally varying wind field. In
addition, variations of wind speed with height can
often be substantial, so it is necessary that a model
include wind shear. As noted, the problems
associated with determining the spatially and
temporally varying wind field for a particular time
period in an area are substantial. Measurements of
winds aloft are typically unavailable, and except for
special circumstances, it is very difficult to predict
wind speed and direction at elevations much
removed from the surface, with only ground
observations available. In addition, the density of
ground-level wind stations is generally less than
that required to interpolate a ground-level wind
field with a resolution comparable to that of the
AQSM,
The chemical mechanism is that element of the
AQSM that has undergone the most change over
the period of time that models have been
developed. This change has been the result of
continuous investigation of the chemistry of
photochemical smog and includes revised rate
constants and new products for some of the
individual reactions. Whereas the 'chemical
mechanisms employed in various models differ
(see Table 6-2), an increasing degree of uniformity
in the treatment of the chemical processes is
beginning to emerge as new chemical information
is incorporated into the mechanisms in the
AQSM's, The most direct evidence of the adequacy
of a chemical mechanism is that obtained through
simulation of smog chamber data.
In principle, every reaction appearing in a
photochemical smog mechanism is subject to
some degree of uncertainty, whether in the rate
constant or the nature and quantity of the
products. In validating a mechanism, the accepted
procedure is to compare the results of smog
chamber experiments, usually in the form of
concentration-time profiles, with simulations of
the same experiments using the proposed
mechanism. A sufficient number of experimental
unknowns exist in all such mechanisms that a
certain degree of adjustment of rate constants (and
perhaps products) is possible. The inherent validity
or accuracy of any mechanism should be judged on
the basis of how realisticthe parameter variations
are.
Uncertainties in the kinetic mechanism are
related to inaccurately known rate constants or
products for reactions in the mechanism.
Uncertainties associated with comparison of the
predictions of the mechanism to experimental
smog chamber data arise, in addition, because the
properties of the photochemical reactor, asso-
ciated equipment, and experimental procedures
are not completely known. The accuracy of any
method based on smog chamber data for relating
precursor emissions or concentrations to oxidant
levels, (whether it is based directly on ozone
isopleths as a function of initial precursor
concentrations or on a kinetic mechanism
validated with smog chamber data) depends on the
extent to which the influence of the smog chamber
on the homogeneous kinetics is understood. Some
of the specific chamber effects that must be
considered are the spectral distribution and
absolute intensity of the photolyzing lamps; the
adsorption, desorption, and chemical reaction of
species on the walls; the initial loading of impurity
species in the chamber air or on the walls; and the
effects of leakage, sampling, and possible
temperature variations during the run. Of these
effects, probably the most important are the effects
of the photolyzing lamps and of species absorbed
on the walls. Photolytic rates of absorbing species
cannot be predicted with accuracy if the incident
light intensity distribution is not known with
accuracy. This information must be coupled with
the absolute rate of photolysis of at least one
species, such as NOz, in order to compute the
appropriate photolytic rate constants.
Characterization of the initial contaminant loading
in the gas and on the walls is important in
simulating the proper initial rates of conversion.
In summary, the three classes of phenomena
that often require treatment as parameters in
kinetic mechanisms in simulating smog chamber
data are; (1) unknown rate constants, (2) the
photolytic properties of the reactor, and (3) wall
absorption, desorption, and heterogeneous
chemistry. Tuning the mechanism to account for
unknown or uncertain chemical and physical
effects is a legitimate procedure provided that the
exact steps are spelled out in detail and lie within
physically realistic bounds. Recent experimental
elucidation of NOx/HOx chemistry (reactions of OH
andHO2with NO and NOz) and PAN chemistry has
significantly improved the accuracy of postulated
mechanisms for photochemical air pollution.
There still exist, however, important uncertainties
m ROx/NOx chemistry (reactions of RO and ROa
with NO and NOZ). Nevertheless, a properly tuned
mechanism is capable of predicting the
110
-------
concentration-time profiles of species such as NO,
N02, ozone, and hydrocarbons within 20 percent
over a wide range of initial conditions. In theory,
such a mechanism, minus the steps included to
account exclusively for chamber effects, should be
capable of predicting atmospheric concentrations
with the same accuracy.
Ozone prediction uncertainties due to
uncertainties in the meteorological and emission
inventory input data have been explored in a
comprehensive sensitivity study by Liu etal.30The
uncertainties in predicted ozone concentrations
from all mechanisms arid input uncertainties were
estimated by Seinfeld51 (Table 6-4). These estimates
from the evaluation studies illustrate the magnitudes
of the possible errors associated with the various
components of an AQSM. The total error, of course,
cannot be determined from such data alone. Seinfeld
did estimate an overall uncertainty value of ± 50 per-
cent.
TABLE 6-4. UNCERTAINTIES IN PREDICTED OZONE
CONCENTRATIONS FROM ALL MECHANISMS AND
INPUT UNCERTAINTIES
Source
Uncertainty in predicted
absolute ozone concentrations,
Chemical mechanism
Meteorology
Wind speed and direction
Mixing depth
Light intensity
Initial and boundary conditions
Emission inventories.
NO,
Hydrocarbons
+20
+20
+25
+20
+50
±20
+30
Evaluation studies using some of the current
AQSM's are summarized in Table 6-5. As shown,
the variety of regions and conditions for which the
AQSM's have been evaluated is limited. In
particular, there is a lack of evaluations under
widely varying emission conditions. Most
important, however, evaluation periods have been
limited to 1 to 3 days, or 6 days atthe most. Overall,
the past evaluations are inadequate in several
respects. First, because of deficiencies in the data
bases used, it is difficult to discern whether or not
the disagreements between prediction and
observations are the result of errors in the model,
inaccuracies in input parameters, or deviations
arising from comparison of point data and volume-
average prediction. Second, the variety of
conditions used in the verification tests was
limited, as already mentioned. Third, and perhaps
most important, the amount of verification m terms
of number of test days is limited to the extent that
the resulting assessment of the model's accuracy
is qualitative, at best; certainly, iris not sufficient
for statistically supported quantitative estimates.
And finally, another problem in interpreting results
from current verification studies is that of tuning of
model parameters or inputs (adjustment of some
influential parameters within their uncertainty
limits to maximize agreement between predictions
and observations). Such tuning varies in extent
from study to study. However, even in the most
serious studies, there are required data (e.g., initial
and boundary concentrations aloft) that are simply
not available and must be estimated; such input
data are often adjusted to obtain the best fit. Thus
the value of the model predictions is lessened to an
extent, depending on the amount of tuning
performed and, more important, on the sensitivity
of the model to the parameters tuned.
A major verification study involving several of
the current AQSM's is now close to completion
and has been reported extensively.15 Known as the
St. Louis Regional Air Pollution Study (RAPS), this
study is a 5-year field program sponsored by EPA.
The major objective is the development, eval-
uation, and verification of AQSM's. The magni-
tude of the study is such that the amount of data to
be obtained will be adequate for a statistical
assessment of AQSM performance. Thus some 50
simulation days will be considered for statistical
evaluation of each model. The statistical
procedures to be used will be based on methods
such as those reported by Brier,4 Nappo,36 and Liu
,et al.30 Following completion of this major effort,
and based on AQSM performance standards to be
defined, EPA will make judgments on the specific
AQSM's to be recommended for use in the
development of oxidant control strategies.
SUMMARY
Quantitative relationships between ambient
oxidant/ozone and precursor emissions are
needed for predicting the impact of emissions on
air quality. Such relationships represent, with
varying degrees of complexity, the physical and
chemical processes taking place in the atmos-
phere.
The most fundamental approach to relating
precursor emissions to oxidant air quality is that
based on the equation of conservation of mass,
which with the help of certain simplifying as-
sumptions, is reduced to the atmospheric diffusion
equation (equation 6-3). This approach, which
uses AQSM's, requires knowledge of the wind
111
-------
field and boundary conditions of the region, the
source emissions as a function of location and
time, and a chemical reaction mechanism. The
complexity of air quality simulation models
(AQSM) is a source of both strength and weakness.
The high degree of spatial and temporal detail
makes the models suitable for a wide variety of
applications, but at the expense of costly
computations. No AQSM is a perfect description of
the atmosphere; inaccuracies arise from
assumptions needed to represent atmospheric
turbulent processes in a manageable form and
from uncertainties in the input information, such
as chemical reaction rate constants. The largest
sources of uncertainty in model predictions are
caused by uncertainties in the initial and boundary
conditions and in the chemical mechanism.
Uncertainties are introduced also by errors in the
wind, mixing height, and light intensity, and in the
HC and NO, emission inventory data. To date,
verification of existing AQSM's has been
inadequate in at least two respects. First, because
of deficiencies in the data base used, it has been
difficult to discern whether the disagreements
between observations and model predictions are
the result of errors in the model, inaccuracies in
input parameters, or deviations arising from
comparison of point data and volume-average
prediction. Second, the amount of validation in
terms of number of test days and the variety of
conditions used in such validation have been
limited to the extent that resulting assessments of
model accuracy have been, at best, qualitative.
Because AQSM's have not been developed to
the point that their degree of accuracy is well-
quantified, it has been necessary to employ
empirical and semi-empirical methods for relating
precursor emissions to oxidant air quality. Of the
empirical methods for relating precursor
emissions to oxidant air quality, linear rollback is
the simplest; it is based on assumed
proportionality between ambient oxidant/ozone
concentration and the reactive organic emission
rate. The method is of highly questionable validity
and limited utility, since most information
indicates that peak oxidant concentrations do not
decrease linearly with reductions in hydrocarbon
emissions.
TABLE 6-5. PRIOR VALIDATION STUDIES OF AIR QUALITY SIMULATION MODELS
Model
SAl"
Time periods
Pollutants compared
LIRAQ"
DIFKIN8
Belt8
Portions of south coast air basin,
Both 50 x 50 mi and 80 x 100 mi
regions. Denver, Colo fSAI, 1977)
San Francisco Bay area.
170 x 210 km region. A variety
of subregions and grid sizes
(1-5 km) were employed.
Trajectories in south
coast air basin
Morris, Essex, Hudson
counties. New Jersey
(Comparisons shown only for
Hudson County).
6 days in 1969
26 June 197441"5
26, 27 July 19733'
20 August 197331
26-28 September 197331
6 days in 1969
LARPP data «
May-September 1972-1974. Cloud-
less summer weekdays with
normal con>'ective mixing and
westerly winds Only days when
the integrated 0600-1300 hr solar
flux was greater than 200
Langleys, average 0400-1300 hr
wind velocity 6 7 mi/hr < v <
10 1 mi/hr, and 0400-1300 hr
wind direction was from 80°
sector encompassing Essex County
to the west-northwest J1
Validation studies performed
with both the 1 5- and 31 - step
kinetic mechanisms. In both versions,
pollutants compared were NO, NOz,
Oa, CO, reactive and unreactive
hydrocarbons.
Pollutants compared were NO,
NOa, On, CO, reactive and unreactive
hydrocarbons.
Pollutants compared were NO,
NOz, On, CO, reactive and unreactive
hydrocarbons
Oj, NO, NOz, SOi and computed
Os behavior matches well the median
of data. Computed concentration
levels of secondary species agree
qualitatively with levels measured in
New Jersey or elsewhere (acrolein,
acetaldehyde, formaldehyde, PAN,
H2O2, HNO2, HNO3) Levels of HO2
radicals estimated on the basis of
predicted HjO2 levels. Levels of OH
radicals estimated on the basis of
predicted HNOs levels
"SAI - Systems Applications, Inc. LIRAQ ~ Livermore regional air quality, DIFKIN - diffusion kinetics. Bell = Bell Laboratories
112
-------
The "Appendix-J method" is a modified version
of linear rollback in which ambient oxidant/ozone
is related to hydrocarbon precursors in the form of
an upper limit curve derived from aerometric data
from several cities on 6- to 9-a.m. NMHC and
maximum oxidant concentration. The method is no
more valid than linear rollback, except perhaps
when used for one locality on the basis of data
taken in that locality. Statistical methods that
relate ambient oxidant/ozone to both HC and N0»
precursors are superior to linear and modified
rollback methods that are based only on
hydrocarbon precursors. The methods, however,
should be employed only for the locality from
which the data on which they are based have
come.
An attempt to base the precursor/ozone
relationship on chemical reality while avoiding the
complex computational requirements of AQSM led
to the development by Dimitriades of peak oxidant
isopleths as a function of initial hydrocarbon and
NO, concentrations derived from smog chamber
data. The advantage of the approach is that it is
based on actual smog chamber data and also on
the results of simulations with a chemical reaction
mechanism validated with smog chamber data.
The limit Jtions of the method arise from lack of
spatial and temporal resolution and from the
question- ble comparability of the atmosphere of a
smog ch mber with the real atmosphere.
In sunmary, air quality simulation models
represen the most fundamental approach to
relating precursor emissions to oxidant air quality.
Although AQSM's have now evolved to the point
where .hey can be employed with some
confidence, they are still characterized by
uncertainties that arise from our lack of ability to
represent truly atmospheric processes. Based on
extensive evaluation of several AQSM's with the
St. Louis RAPS data, EPA will make judgments on
specific AQSM's to be recommended. Because of
their extensive computational requirements, even
where fully validated, the AQSM approach may not
be suitable for all applications involving oxidant
control planning.
REFERENCES FOR CHAPTER 6
1 Air Pollution Control Office Air Quality Criteria for
Nitrogen Oxides APCO Publication No. AP-84, U S
Environmental Protection Agency, Washington, D.C
2 Bailey, B S Oxidant-HC-NO, relationships from
aerometric data - L A. studies. In: Scientific Seminar on
Automotive Pollutants EPA-600/9-75-003, U S
Environmental Protection Agency, Research Triangle
Park, N C , February 1975 Sec 5, subsect C
3 Boris. J. P,, and D. L Book Flux corrected transport. I
Shasta, a fluid transport algorithm that works. J.Comput
Phys. ??;38-69, 1973.
4. Brier, G W. Statistical Questions Relating to the
Validation of Air Quality Simulation Models. EPA-650/4-
75-010, US. Environmental Protection Agency,
Research Triangle Park, N C., March 1975.
5. Caporaletti, J. M., L. N. Myrabo, P. Schleifer, A. Stanonik,
and K R. Wilson. Statistical oxidant air quality prediction
for land useandtransportationplannmg.Atmos. Environ.
77:449-458, 1977.
6. Deardorff, J W A three-dimensional numerical
investigation of the idealized planetary boundary layer.
Geophys. Astrophys Fluid Dyn. 7 377-410, 1970.
7. DeNevers, N , and J. R Morris Rollbac modeling Basic
and modified. J. Air Pollut. Control Assoc 25.943-947,
1975
8 Dimitriades, B. An alternative to the Appendix-J method
for calculating oxidant- and NO2-related control
requirements. In: International Conference on
Photochemical Oxidant Pollution and Its Control
Proceedings Vol. II. B. Dimitnades, ed, EPA-600/3-77-
001 b, U.S. Environmental Protection Agency, Research
Triangle Park, N.C.January 1977 pp. 871-879
9. Dimitriades, B Chemistry In: Assessing Transportation-
Related Air Quality Impacts. Special Report 1 67. National
Academy of Sciences, Washington, D C., 1976. pp 8-20.
10. Dimitriades, B. On the Function of Hydrocarbon and
Nitrogen Oxides in Photochemical Smog Formation
Report of Investigations Rl 7433. U S Department of the
Interior, Bureau of Mines, Washington, D C,, September
1970
11 Dimitriades, B. Oxidant control strategies Part I. Urban
control strategy derived from existing smog chamber
data. Environ. Sci. Technol 7 7:80-88, 1977
12. Dimitriades, B, chairman Smog Chamber Conference
Proceedings U.S Environmental Protection Agency,
Research Triangle Park, N C , April 1976
13. Dimitriades, B The use of smog chamber data in
formulating oxidant control strategies In Report on UC-
ARB Conference. Technical Bases for Control Strategies
of Photochemical Oxidant. Current Status and Priorities
in Research Univ of California Statewide Air Pollution
Research Center Riverside, Calif, May 1976.
14 Dimitriades, B, The use of smog chamber data in
formulating oxida'nt control strategies In. Report on UC-
ARB Conference "Technical Bases for Control Strategies
of Photochemical Oxidant Current Status and Priorities
in Research." University of California, Statewide Air
Pollution Research Center, Riverside, Calif, May 1976
pp 76-95
15 Dmerjian, K L Photochemical air quality simulation
modeling, current status and future prospects In:
International Conference on Photochemical Oxidant
Pollution and Its Control Proceedings Vol II B
Dimitriades, ed EPA-600/3-77-001 b, US
Environmental Protection Agency, Research Triangle
Park, N C., January 1977 pp 777-794
16 Dodge, M C. Combined use of modeling techniques and
smog chamber data to derive ozone-precursor
relationships In: International Conference on
Photochemical Oxidant Pollution and Its Control
Proceedings. Vol II B Dimitriades, ed EPA-600/3-77-
001 b, US Environmental Protection Agency, Research
113
-------
Triangle Park, N.C , January 1977 pp., 881 -889
17. Dodge, M. C Effect of Selected Parameters on Prediction
of a Photochemical Model EPA-600/3-77-048, U S
Environmental Protection Agency, Research Triangle
Park, N.C , June 1977.
18 Edelson, D A simulation language and compiler to aid
computer solution of chemical kinetic problems Comput
Chem. 7:29-33, 1976
19 Eschenroeder, A Q, and J R Martinez. Concepts and
applications of photochemical smog models Adv Chem
Ser.l113): 101-168, 1972.
20 Fnedlander, S K.andJ H Seinfeld. A dynamic model of
photochemical smog Environ. Sci. Technol 31175-
1181, 1969
21 Graedel, T E , L A Farrow, and T A Weber Kinetic
studies of the photochemistry of the urban troposphere.
Atmos. Environ 70.1095-1116, 1976
22, Hecht, T. A Urban Air Shed Photochemical Simulation
Model Study Voi I Development and Evaluation
Appendix B. Generalized Mechanism for Describing
Atmospheric Photochemical Reactions EPA-R4-73-
030c, U S Environmental Protection Agency, Research
Triangle Park, N C , July 1973
23 Hecht, T A, J H Seinfeld, and M. C Dodge Further
development of generalized kinetic mechanism for
photochemical smog Environ Set Technol 8 327-339,
1974
24 Heuss, J M,G J Nebel, andj M Colucci National air
quality standards for automotive pollutants - A critical
review J Air Pollut Control Assoc. 27 535-544, 1971
25 Hindmarsh, A C GEARS Solution of Ordinary
Differential Equations Having Banded~ Jacobian
University ofCaliforma, Lawrence Livermore Laboratory,
Livermore, Calif UCID-30059, Rev. 1, 1974
26 Kmosian.J R,, a.ndJ J Paskmd Hydrocarbons, Oxides of
Nitrogen, and Oxidant Trends in the South Coast Air
Basin, 1963-1972 State of California, Air Resources
Board, Sacramento, Calif, June 1973
27. Lamb, R G Continued Research in Mesoscale Air
Pollution Modeling Vol III Modeling of Microscale
Phenomena EPA-600/4-76-01 6c, US. Environmental
Protection Agency, Research Triangle Park, N C , May
1976
28 Lamb, R G,W H Chen, and J H Seinfeld Numerico-
empincal analysis of atmospheric diffusion theories J
Atmos Sci. 321784-1807, 1975
29, Liu,M.K.,andP. M Roth Urban Air ShedPhotochemical
Simulation Model Study Vol. I Development and
Evaluation. Appendix C Microscale Model of Local
Vehicular Source Contributions to Measured Pollutant
Concentrations EPA-R4-73-030d, U S Environmental
Protection Agency, Research Triangle Park, N.C , July
1973.
30 Liu, M K., D C Whitney, J H Seinfeld, and P M Roth
Continued Research in Mesoscale Air Pollution
Simulation Modeling Vol I Assessment of Prior Model
Evaluation Studies and Analysis of Model Validity and
Sensitivity. EPA-600/4-76-016a, U.S. Environmental
Protection Agency, Research Triangle Park, N C , May
1976.
31. MacCracken, M C , and G D Sauter, eds Development
of an Air Pollution Model for the San Francisco Bay Area
Final Report to the National Science Foundation Vols. 1
and 2, UCRL-51920, University of California, Lawrence
Livermore Laboratory, Livermore, Calif, October 1975
32 Martinez, J R User's Guide to Diffusion/Kinetics
(DIFKIN) Code EPA-R4-73-012b, U S Environmental
Protection Agency, Research Triangle Park, N C , October
1972.
33 Merz, P H , L J Painter, and P R Ryason Aerometric
data analysis-time series analysis and forecast and an
atmospheric smog diagram Atmos Environ 6319-342,
1972
34. Myrabo, L N , P Schleifer, and K. R. Wilson Oxidant
Prediction Model for Land Use and Transportation
Planning, California Air Environment 4 3, 1974
35 Myrabo, L N,K R Wilson, and J C Trijonis Survey of
statistical models for oxidant air quality prediction. In.
Assessing Transportation-Related Air Quality Impacts
Special Report 167, National Academy of Sciences,
Washington, D.C , 1976 pp 46-62
36 Nappo, C J , Jr A method for evaluating the accuracy of
air pollution prediction models Symposium on
Atmospheric Diffusion and Air Pollution American
Meteorological Society, Boston, Mass, 1974 pp 325-
329
37 National Academy of Sciences Air Quality and
Automobile Emission Control A Report by the
Coordinating Committee on Air Quality Studies Vol 3
The Relationship of Emissions to Ambient Air Quality
Studies prepared for the Committee on Public Works,
U S Senate U.S Government Printing Office,
Washington, D C., 1974.
38. National Research Council Ozone and Other
Photochemical Oxidants. National Academy of Sciences,
Washington, D C , 1976.
39 Office of Air Quality Planning and Standards Uses,
Limitations and Technical Basis of Procedures for
Quantifying Relationships Between Photochemical
Oxidants and Precursors. EPA-450/2-77-021a, U S
Environmental Protection Agency, Research Triangle
Park, N C, November 1977
40 Pitts, J N , Jr., A. M.Winer, K. R Darnall, G.J Doyle, and
J M McAfee Chemical Consequences of Air Quality
Standards and of Control Implementation Programs
Roles of Hydrocarbons, Oxides of Nitrogen and Aged
Smog in the Production of Photochemical Oxidant
California Air Resources Board Contract No 4-214.
University of California, Statewide Air Pollution Research
Center, Riverside, Calif., May 1576
41 Ranzien, A., California Air Resources Board,
Sacramento, Calif. Personal Communication to Prof J H
Seinfeld, Calif Inst Technology, Pasadena, Calif, 1977
42 Reynolds, S D , J Ames, T A Hecht, J P Meyer, D C.
Whitney, and M A Yocke Continued Research in
Mesoscale Air Pollution Simulation Modeling Vol II
Refinements in the Treatment of Chemistry,
Meteorology, and Numerical Integration Procedures.
EPA-600/4-76-016b, U.S. Environmental Protection
Agency, Research Triangle Park, N C , May 1976
43, Reynolds, S D Urban Air Shed Photochemical
Simulation Model Study Vol I Development and
Evaluation. Appendix D. Numerical Integration of
Continuity Equations. EPA-R4-73-030e, U S
Environmental Protection Agency, Research Triangle
Park, NC, July 1973
114
-------
44. Reynolds, S D., M.K Lm,T A. Hecht.P.M Roth, and J.H.
Seinfeld. Urban Air Shed Photochemical Simulation
Study Vol I Development and Evaluation. EPA-R4-73-
030a, U S. Environmental Protection Agency, Research
Triangle Park, N.C., July 1973.
45 Reynolds, S D,M K. Liu.T A.Hecht.P M Roth, and J.H
Seinfeld Mathematical modeling of photochemical air
pollution III Evaluation of the model. Atmos. Environ.
8.563-596, 1974
46. Reynolds, S. D , P. M Roth, and J H. Seinfeld.
Mathematical modeling of photochemical air pollution. I.
Formulation of the model Atmos. Environ. 7/1033-1061,
1973
47 Reynolds, S D , and J. H Seinfeld. Interim evaluation of
strategies for meeting ambient air quality standard for
photochemical oxidant Environ. Sci Technol 9/433-
447, 1975.
48. Roberts, P. J., M. K Liu, S D. Reynolds, and P. M Roth.
Urban Air Shed Photochemical Simulation Model Study.
Vol I Appendix A Contaminant Emissions Model and
Inventory for Los Angeles. EPA-R4-73-030b, U.S.
Environmental Protection Agency, Research Triangle
Park, N C., July 1973.
49 Roth, P. M ,P J.W.Roberts, M.K.Liu.S.D Reynolds, and
J H. Seinfeld. Mathematical modeling of photochemical
air pollution II. A model and inventory of pollutant
emissions. Atmos. Environ. 5/97-130, 1974.
50. Schuck, E. A., A. P Altshuller, D. S. Barth, and G. B
Morgan. Relationship of hydrocarbons to oxidants in
ambient atmospheres J. Air Pollut. Control Assoc.
20.297-302, 1970
51. Seinfeld, J. H. International Conference on Oxidants,
1976—Analysis of Evidence and Viewpoints. Part VI. The
Issue of Air Quality Simulation Model Utility. EPA-
600/3-77-118, U.S. Environmental Protection Agency,
Research Triangle Park, N.C., November 1977.
52. Souten, D. R., C. J. Hopper, and R. L Mueller A critical
review of the Los Angeles County APCD method for
simulating atmospheric oxidant based on smog chamber
irradiation experiments. Presented at the 68th Annual
Meeting, Air Pollution Control Association, Boston,
Mass., June 15-20, 1975.
53. Taylor, G. H., and A Q. Eschenroeder.TestsoftheDIFKIN
photochemical/diffusion model using Los Angeles
reactive pollutant data In: International Conference on
Photochemical Oxidant Pollution and Its Control.
Proceedings Vol. II. B Dimitriades, ed. EPA-600/3-77-
001 b, U.S Environmental Protection Agency, Research
Triangle Park, N.C , January 1977. pp. 817-825.
54. Tiao, G.C..M.S Phadke,andG. E. P. Box Someempincal
models for the Los Angeles photochemical smog data.
Presented at the 68th Annual Meeting, Air Pollution
Control Association, Boston, Mass., June 1975.
55 Trijoms, J. C. Economic air pollution control model for Los
Angeles County in 1975 General least cost air quality
model Environ. Sci. Technol 8.811-826,1974.
56. U S. Environmental Protection Agency. Approval and
promulgation of implementation plans Promulgation of
Texas transportation control plan. Fed. Regist. 38/30633-
30650, November 6, 1973
57. U S Environmental Protection Agency Approval and
promulgation of implementation plans. Boston, Mass :
Promulgation of transportation control plan. Fed. Regist.
38/30960-30971, Novembers, 1973
58. U.S. Environmental Protection Agency. National primary
and secondary air quality standards Fed Regist
36/8186-8201, April 30, 1971
59 U.S. Environmental Protection Agency. Requirementsfor
preparation, adoption, and submittal of implementation
plans Fed Regist. 36/15486-15506, August 14, 1971.
60. U.S. Environmental Protection Agency. Technical
Support Document for the Metropolitan Los Angeles
Intrastate Air Quality Control Region Transportation
Control Plan Final Promulgation Published in November
12, 1973 Federal Register. U.S. Environmental
Protection Agency, Region IX, San Francisco, Calif.,
October 30, 1973
61. Walker, H M A "J" relationship for Texas. In.
International Conference on Photochemical Air Pollution
and Its Control Proceedings Vol. I B Dimitriades, ed.
EPA-600/3-77-001b, US. Environmental Protection
Agency, Research Triangle Park, N.C , January 1977. pp.
851-869.
62 Whitten, G. Z., and H. Hogo Mathematical Modeling of
Simulated Photochemical Smog EPA-600/3-77-011,
U.S. Environmental Protection Agency, Research
Triangle Park, N C , January 1977.
115
-------
7. MEASUREMENT METHODS FOR OZONE, OXIDANTS, AND
THEIR PRECURSORS
INTRODUCTION
Since the publication of the air quality criteria
documents for photochemical oxidants,60 hydro-
carbons,61 and nitrogen oxides,1 there have been
significant advances in the measurement of these
pollutants in ambient air. The chemiluminescent
reaction of 03 with ethylene has been successfully
exploited to produce instrumentation whose
response is specific for O3 and is linear with O3
concentrations over the range found in ambient
air Advances in electronics technology have
allowed development of photometers of adequate
precision to measure 03 by ultraviolet absorption
photometry. Continuous analyzers based on the
chemiluminescent reaction of NO with 03 have
been successfully developed for the measurement
of NO, nitrogen dioxide (NO2), and oxides of
nitrogen (NO*). Improved manual (integrated)
methods for measuring N02 have also been
developed. Lesser advances have been reported in
the measurement of ambient hydrocarbons and
peroxyaclynitrates.
The emphasis in this chapter will be placed
primarily on methods for measuring 03, on the O3/
oxidant relationship, and on methods for
measuring hydrocarbons and NO, N02, and NOX.
Continuous methods for measuring total oxidants
based on the reaction of ambient air oxidants with
potassium iodide(Kl) will also be discussed briefly.
A more detailed discussion of total oxidant
methodology is given in the earlier criteria
document
60
SAMPLING FACTORS IN AMBIENT AIR
MONITORING
In addition to analytical principles and
measurement procedures, sampling factors also
have a crucial effect on the quality of ambient and
experimental atmosphere measurements. In
sampling the ambient atmosphere, it is extremely
important that the sampling be performed in a
manner that (1) is consistent with the specific
purpose of the measurement, and (2) preserves the
integrity of the pollutant mixture in the ambientair
sample. These sampling factors and their
significance in 03 or oxidant-related air quality
monitoring will be discussed here briefly. For more
detailed discussions of this subject, the reader is
referred to other EPA documents and to reports
prepared for EPA by the National Academy of
Sciences.62'63'67'72
Air monitoring data relevanttotheambientOsor
oxidant problem are collected for a diversity of
specific purposes, including serving as an indicator
of progress in attainment of the national ambient
air quality standard for 03 in the development of 03
control strategies, in the development and vali-
dation of ozone-related air quality simulation
models, in the investigation of causes of ozone
problems, etc. Each of these specific purposes or
needs requires special considerations in designing
an air sampling strategy responsive to the
respective need. For example, statistical
considerations enter into the design of a moni-
toring network capable of detecting air quality
changes over a given period of time. Other
considerations arising from the chemical, physical,
and meteorological aspects of the ambient ozone
problem are explained briefly as follows.
Ozone is a product of photochemical reactions
that involve sunlight, hydrocarbon, and NOX re-
actants, which are heavily discharged into urban
atmospheres by automobiles during the morning
peak traffic hours. This photochemical process
occurs at a rate such that the 03 concentration
reaches its daily peak level some time in midday at
locations downwind from the source-intensive
center-city area. Thus if peak O3 concentrations
are to be measured, monitoring stations should be
located downwind from city centers at distances
that have been determined by EPA to be 1 5 to 30
km (9 to 19 miles), depending on the area's
predominant wind patterns.67 Monitoring stations
should also be located within the source-intensive
116
-------
area for measurement of peak concentrations of
oxidant precursors. Once a station is located,
additional considerations arise because of the
chemical reactivity and instability of the O3
molecule. Ozone reacts extremely rapidly with NO
and with some hydrocarbon compounds, including
most of those emitted by vegetation. Also, O3
decomposes readily on contact with the surface of
many materials. Consideration of these effects led
to the development of specific criteria for locating
an O3 monitoring station.62'63'67'72 Briefly, the inlet
of the Os analyzer's sampling probe should be
positioned 3 to 15 m (10 to 49 ft) above ground, at
least 4 m 113 ft)from large trees, and 120 m (394 ft)
from heavy automobile traffic. Sampling probes
should be designed to minimize O3 destruction by
surface reaction or by reaction with NO.
Air monitoring data, as commonly obtained,
have only limited validity as measures of absolute
air quality. The reason is that at ground level, the
ambient atmosphere is inhomogeneous as a result
of a continuous influx of fresh emissions,
incomplete mixing, and destruction of 03 by fresh
and unreacted emissions. In view of such
inhomogeneity, monitoring data from a fixed
network provide measures of air quality at a
discrete number of locations but may not detect
possibly existing hot-spots. This problem can be
alleviated by use of a greater density of monitoring
stations or, perhaps at a lower cost, by use of an air
quality model. Such models (Chapter 6) are
capable of quantifying the emission dispersion and
chemical reactions processes, and their outputs
can provide data on the distribution of air quality
concentrations between widely spaced ambient
monitors.
MEASUREMENT OF OZONE
Gas-Phase Chemiluminescence (EPA Reference
Methods)
The U.S. Environmental Protection Agency
promulgated National Ambient Air Quality
Standards (NAAQS) for six pollutants on April 30,
1971.104 The standards are now codified at Title
40, Code of the Federal Register (40 CFR), Part 50.
At the same time, EPA published reference
methods (presently described in the appendices to
Part 50) to be used by EPA and by state and local
agencies in measuring ambient concentrations of
the six pollutants. Appendix D of 40 CFR Part 50for
the measurement of photochemical oxidants
describes the measurement principle for an
automated method (analyzer) based on the gas-
phase chemiluminescent reaction of 03 with
ethylene.65'99 Measurements made with analyzers
based on this principle are therefore ozone-
specific and differ from measurements made with
analyzers designed to measure total oxidant.
These analyzers are designed such that ambient
air and ethylene are delivered simultaneously to a
reaction cell where the Os in the air reacts with the
ethylene. The reaction produces a small fraction of
an energetically excited species (thought to be an
electronically excited formaldehyde molecule) that
decays to the ground state with the emission of
light. The intensity of the emitted light
(Chemiluminescence), which is detected by a
photomuitiplier tube, is proportional to the 03
concentration over the range of 4 to at least 5000
//g/m3 (0.002 to at least 2.5 ppm). The quantitative
relationship between the intensity of the
Chemiluminescence and the 03 concentration
must be established for each analyzer, using
atmospheres containing known concentrations of
03.
Analyzers utilizing the gas-phase Chemilu-
minescence measurement principle have been
evaluated under a variety of conditions.718'22'98
These studies indicate that the performance of
commercially available analyzers is generally
satisfactory. However, the results of a
collaborative test by McKee et al. indicated
considerable variability and a negative bias of 16 to
37 percent.58 The reason for the variability is no
doubt, m part, the variability in the 1 percent
neutral buffered potassium iodide (NBKI)
calibration procedure, which is discussed in detail
in a subsequent section on 03 calibration
procedures. The reported negative bias is in
conflict with subsequent studies of the 1 percent
NBKi procedure, which generally indicate that this
procedure produces a positive bias of 5 to 30
percent (see section on O3 calibration procedures).
Under the provisions of the Ambient Air
Monitoring Reference and Equivalent Methods
Regulations102 promulgated February 18,1975 (40
CFR Part 53), several commercial analyzers have
been designated as reference methods for
determining compliance with the NAAQS for
photochemical oxidants. These analyzers have
been subjected to the required testing and have
met the EPA performance specifications for
automated methods. These specifications are
given in Table 7-1, and a list of the ozone analyzers
(designated as of March 3, 1978) is given in Table
117
-------
7-2. Information concerning the applications
supporting the designation of these analyzers as
reference or equivalent methods may be obtained
by writing the Environmental Monitoring and Sup-
port Laboratory, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina
27711.
TABLE 7-1. PERFORMANCE SPECIFICATIONS FOR
AUTOMATED METHODS
Performance parameter
Ozone
Range, ppm
Noise, ppm
Lower detectable limit, ppm
Interference equivalent:
Each mterferant, ppm
Total mterferant, ppm
Zero drift, 12- and 24-hr, ppm
Span drift, 24-hr
20% of upper range limit, ppm
80% of upper range limit, ppm
Lag time, min
Rise time, min
Fall time, mm
Precision:
20% of upper range limit, ppm
80% of upper range limit, ppm
0-0.5
0005
001
±0.02
006
±002
±200
±50
20
15
15
0.01
0.01
A review of the performance data submitted in
support of the designations listed in Table 7-2
indicates that these analyzers exhibit performance
better than that specified in Table 7-1. For the
analyzers tested, the zero drift results (12- and 24-
hr) were all less than 5 ppb and typically less than 3
ppb. The span drift results (at 20 and 80 percent of
the full-scale range of 0 to 0.5 ppm) were all less
than 5 percent, and typically 2 to 3 percent. The
precision results (at 20 and 80 percent of the full-
scale range of 0 to 0.5 ppm) indicate a typical
precision of 1 ppb. The response times (lag, rise,
and fall) were all less than 2 min, and typically less
than 1 min. The interference equivalent results
(tests for carbon dioxide, hydrogen sulfide, and
water vapor) were typically less than 1 ppb.
However, under 40 CFR Part 53, the test procedure
for water interference specifies that the test be
conducted in the absence of 03. There have been a
number of reports of a positive water interference
of 3 to 12 percent under conditions of high
humidity in the presence of O3.51'53'73
Gas-Solid Chemiluminescence
The measurement of O3 by a gas-solid chemilu-
minescence technique was first reported by
Regener.77'78 In Regener's approach, air containing
O3 is passed across a surface of Rhodamine B and
absorbed on silica gel, resulting in the emission of
light (chemiluminescence). The intensity of the
emitted light is measured with a photomultiplier
tube and is proportional to the concentration of O3
in the air sample. The method is highly specific and
extremely sensitive (lower detection limit less than
0.001 ppm). Because the sensitivity of the
chemiluminescence surface gradually decays with
time, automated analyzers based on this
measurement principle are designed to
incorporate frequent internal calibration cycles.
Prototype Regener O3 analyzers have been
evaluated under field conditions and have given
excellent performance.6
A commercial analyzer, based on a slight
modification of the gas-solid chemiluminescence
principle, has been designated as an equivalent
method under the EPA 40 CFR Part 50 regulations
(Table 7-3). This analyzer, the Philips Model
PW9771 O3 Monitor, uses a detector disc
containing Rhodamine B and an intermediate
reagent, gallic acid. The O3 in the air sample reacts
with the gallic acid to produce oxygen and a gallic
acid derivative, which in turn reacts with the
Rhodamine B to produce the chemiluminescence.
Ultraviolet Photometry
Older ozone monitors, based on the absorption
of ultraviolet (UV) light, and their attendent
problems are described in the earlier criteria
document for photochemical oxidants.60 Modern
electronic techniques have permitted the
successful development of an ambient O3 analyzer
based on ultraviolet absorption.12 This analyzer,
the Dasibi Model 1003-AH Ozone Analyzer, has
been designated as an equivalent method under
the EPA 40 CFR Part 53 regulations (Table 7-3).
The absorption of 254 nm ultraviolet light by O3 in
an ambient air sample is measured, and a separate
measurement is made on a similarairsamplefrom
which the O3 has been removed by a manganese
dioxide (MnO2) scrubber. These two
measurements are processed electronically to
produce a digital readout of the O3 concentration in
the air sample. The O3 determination is based on
the well established absorption coefficient of O3 for
254 nm light.23-32-36-43-100
Although not in wide use, this method is used
extensively in California. Its main advantage is that
it does not require the use of support gases such as
ethylene (as is the case with the gas-phase
chemiluminescence methods). In addition, the O3
measurement can be referenced to the absorption
118
-------
coefficient of Oa without reference to a dynamic
calibration standard. In practice, however,
analyzers may require dynamic calibration with 63
standards because of problems associated with
sample integrity. While potential interferences are
few, there are some ambient pollutants, including
sulfur dioxide and benzene, that absorb light at 254
nm. Presumably, these potential interferants are
unaffected by the Mn02 scrubber and therefore do
not interfere with the ozone measurement.
However, if the MnOz scrubber eliminates or
reduces such components, the analyzer gives
erroneous readings. Field use of this analyzer is
discussed by Zafonte et al.104
MEASUREMENT OF TOTAL OXIDANTS
The methods mostfrequently used in the past for
measurement of ambient total oxidants were
based on the oxidation of potassium iodide (Kl) in
solution and the electrochemical or colorimetric
detection of the iodinelb) produced. The Kl reagent
produces fe by reaction with O3, NO2, peroxyacyl-
nitrates, chlorine (Cb), bromine (Br2>, and pero-
xides.40'85'101 Reducing agents such as SOa are
negative interferants because they react with the I2
formed by the oxidizing species. Terms such as
"corrected oxidant" or "adjusted oxidant" have
often been used to indicate that the oxidant
measurement has been corrected for
interferences from NO* and/or SOa. The corrected
or adjusted measurement results are then taken to
represent an oxidant mixture that is predominantly
Oa with small amounts of other oxidizing com-
pounds.
TABLE 7-2. LIST OF DESIGNATED REFERENCE METHODS
Designation
number
Fed Register Notice
Identification and source
RFOA-1075-003
RFOA-1075-004
RFOA-1076-007
RFOA-1076-014
RFOA-1076-015
RFOA-1076-016
RFOA-1176-01 7
RFOA-0577
Meloy Model OA 325-2R Ozone Analyzer
Meloy Laboratories, Inc
6715 Electronic Drive
Springfield, Virginia 22151
Meloy Model OA 350-2R Ozone Analyzer
Meloy Laboratories, Inc
671 5 Electronic Drive
Springfield, Virginia 22151
Bendix Model 8002 Ozone Analyzer
The Bendix Corporation
Post Office Drawer 831
Lewisburg, West Virginia 24901
MEC Model 1100-1 Ozone Meter
MEC Model 1100-2 Ozone Meter
MEC Model 1 100-3 Ozone Meter
Columbia Scientific Industries
11950 Jollyvifte Road
PO Box 9908
Austin, Texas 78766
Monitor Labs Model 841 OE
Ozone Analyzer
Monitor Labs, Inc
4002 Sorrento Valley Blvd
San Diego, California, 92121
Beckman Model 950A Ozone Analyzer
2500 Harbor Boulevard
Fullerton, California 92634
_Vol_
40
40
41
41
41
42
Page Dale
54856 11/26/75
54856 11/26/75
5145 2/4/76
466747 10/22/76
30235 6/13/77
53684 12/8/76
28571 6/3/77
TABLE 7-3. LIST OF DESIGNATED EQUIVALENT METHODS
Designation
number
EQOA-0777-023
EQOA-0577-019
Identification and source
Philips PW9771 O3 Analyzer,
Philips Electronic Instruments. Inc
85 McKee Drive
Mahwah, New Jersey 07430
Dasibi Model 1003-AH
Analyzer
Dasibi Environmental Corp
616 E Colorado St
Glendale, California 91205
42
42
Fed Register Nonce
Page Date
38931 8/1/77
28571
6/3/77
119
-------
A variety of Kl methods have been used for the
continuous measurement of ambient total
oxidants, Amperometric85'101 (often called coulo-
metric) analyzers draw the air sample into a sensor
cell where the oxidant reacts with the electrolyte
(neutral buffered KI:KBr solution) to release I2. The
\2 is in turn reduced back to iodide, either
electrolytically with a bias voltage across the cell,
as used by mast, or galvanically, with different cell
electrode materials according to Hersh and
Deuringer.37 Nitrogen dioxide is a positive
interferant that causes a response equivalent to 3
to 20 percent of that of Oa, depending on a detector
design, operating conditions, and other unknown
factors. Sulfur dioxide results in a negative
interference if la is present in the electrolyte. When
oxidant is present in excess, a given concentration
of SC>2 reduces the oxidant response on a mole
Oa/mole SOa basis.
Colorimetric analyzers utilize either a 10- or a
20-percent neutral buffered Kl40'60'101 solution.
Oxidants react with Kl to produce Iz and the tri-
iodide ion (T3), which is formed in the presence of a
large excess of iodide ion and is measured in a
continuous colorimeter. Sulfur dioxide results in a
negative response similar to the case for the
amperometric analyzer. Nitrogen dioxide causes a
positive response equivalent to 21 and 30 percent
of that of 03 for the 10- and 20-percent Kl
analyzers, respectively.
To eliminate errors resulting from SO2
interference with these Kl methods, it is necessary
either to measure S02 and adjust the total oxidant
measurement appropriately, or to remove SO2 by
passing the air sample through a chromium
trioxide (CrO3) scrubber.40 In the first case,
adjustments can be made only if the oxidant levels
are in excess of the SOa levels. In the latter case,
the Cr03 scrubber may cause new interferences by
partially oxidizing the nitric oxide (NO) and
hydrogen sulfide (H2S) possibly present in the
sample to NO2 and SOa, respectively, and by
destroying part of the oxidant. Because of these
potential problems, this method of correcting for
SOa interference can be used reliably only in
certain situations and only by skilled an-
alysts.63'85'101
OZONE CALIBRATION PROCEDURES
Calibration of 03 analyzers is complicated by the
lack of Standard Reference Material (SRM) for O3
analogous to those available from and certified by
the National Bureau of Standards (NBS) for SOZ,
NO, and carbon monoxide (CO). The instability of
O3 prohibits the storage of O3 standards for any
practical length of time. Therefore, standard
samples in air for calibration of Os analyzers must
be generated and analyzed at the time and place of
use. Typically, O3 atmospheres are generated by
means of an 03 generator that delivers a stable
concentration of O3. These- atmospheres are
assayed by some technique to determine their 03
concentration and then are used to calibrate the
analyzer. The Oa assay procedure may be based on
one of several different primary standards.
Kl Calibration Procedures
Until recently, Oa and total oxidant analyzers
have been calibrated using O3 atmospheres that
have been assayed using various Kl procedures.
Such procedures, however, have been severely
criticized lately for lack of accuracy and pre-
cision.9'15'24'101 Some specifics of these procedures
and associated criticisms are briefly reviewed
here.
The determination of the precise Oa atmos-
pheres by the Kl procedures consists of bringing
the air/O3 mixture into contact with buffered or
unbuffered Kl solutions of various concentrations.
In theory, absorption of one molecule of O3 in any
of these solutions should result in formation of one
molecule of la as shown by the following equation:
2H+ + 2f + 03 = I2 + 02 + H20
The I2 in the presence of a large excess of iodide ion
forms the intensely colored triiodide ion (I3). The
concentration of I3 is determined using a
spectrophotometer calibrated with standard
solutions of la/KI, standardized against a primary
standard such as arsenous oxide (As203). The most
commonly used procedures utilize a I- or 2-percent
NBKI solution, or a 2-percent unbuffered KI(UKI)
solution. When EPA first promulgated the
regulations (40 CFR Part 50) on April 30, 1971,104
the 1-percent NBKI procedure was considered to
be the best available for calibration of methods for
measuring photochemical oxidants. This
procedure is described in Appendix D of 40 CFR
Part 50.
Considerable evidence has accumulated to
indicate that the use of NBKI calibration
procedures can result in significant bias,
variability, or both. These problems were
documented in a joint EPA-NBS workshop held in
August 1974.19 Specific problem areas reported
included purity of reagents, time for maximum
color development, variability due to impinger type,
120
-------
and a positive bias when compared to other ozone
measurement methods such as UV photometry.
An intercomparison of three KI procedures was
carried out in California for the purpose of
determining the relative (accuracy) responses and
precisions of the three techniques. The three
techniques were the 1 -percent NBKI procedure (as
used in the EPA reference method) the 2-percent
NBKI procedure (as used by the California Air
Resources Board), and the 2-percent UK1
procedure (as used by the Los Angeles Air
Pollution Control District). The absolute levels of 03
were measured using UV photometry. The
comparative results are fully described in a report
by DeMore et al,z4 and should be consulted for
details. The major conclusions of the study are that
the EPA (1-percent NBKI) and California Air
Resources Board (2-percent NBKI) procedures give
higher values by about 15 to 25 percent compared
to the UV procedure. The Los Angeles County Air
Pollution District procedure (2-percent UKI) gives
values approximately 4 percent lower than the UV
photometry measurements, but with considerable
scatter in its values. Several other studies have
indicated discrepancies and considerable
variability between the various Kl procedures and
newly developed procedures based on GPT and
ultraviolet absorption (UV),9.15'39<41'51'53'73'79 All of
these and other studies have been reviewed
recently by Burton et al.14
Based on the results of the studies discussed or
cited above, it can be concluded that the Kl
procedures for measuring Oa, even when used in
the laboratory by skilled operators, suffer from
substantial systematic (accuracy-related) as well
as random (precision-related) errors. The factors
responsible for these errors, especially the random
errors, are not well understood. They are probably
related to differences in procedural detail among
the various operators. When these Kl procedures
are used in the field by unskilled operators, such
errors are certain to be considerably greater. The
most important implication of these problems with
the Kl procedures is discussed later in the section
on the relationships between ambient oxidant and
ozone data.
Recently Developed Calibration Procedures
There has, been a sustained effort to develop
other techniques for measuring absolute con-
centrations of ozone in calibration atmospheres.
These efforts have been successful, and it is all but
certain that the Kl calibration procedures will soon
be replaced by one or more new techniques.
0 n October 6,1976, EPA published in theFeder-^
al Register™3 a notice of its intent (1) to study and
evaluate several alternative calibration procedures
for reference methods for photochemical oxidants
and (2) to amend Appendix D of 40 CFR Part 50 to
revise the 1 -percent NBKI calibration procedure or
to replace it with one or more alternative
procedures. Four alternative calibration
procedures are currently under consideration: (1)
GPT with excess NO, (2) UV photometry, (3) GPT
with excess 03, and (4) boric acid Kl (BAKI). A brief
summary of each of these alternative procedures
follows.
The Os calibration procedure utilizing GPT with
excess NO is a relatively well developed procedure
and very similar to that specified by EPA for
calibration of N02 reference method analyzers.105
The procedure is based on the reaction of Oa with a
known quantity of NO and uses available NO
standards that can serve double duty in generating
NO2 standards as well. However, an NO analyzer,
which may not be readily available when
calibrating 03 analyzers, is required. In addition,
the procedure is somewhat complex and requires
accurate measurement of several gas flow rates.
The UV photometry procedure is based on the
absorption coefficient of O3 at 254 nm, which has
been well established by independent deter-
minations. The procedure is quite easy to carry out,
uses a physical measurement, and requires no
other gases or critical flow measurements.
However, the procedure as applied to the
calibration of O3 analyzers is relatively new and not
in general use in the air monitoring community and
has not been used extensively to measure Oa in the
sub-ppm range needed for ambient Oa analyzer
calibration.
The procedure utilizing GPT with excess Oa is
similar to GPT with excess NO but has the
advantage of not requiring an NO analyzer. If the
residence time of the Os and NO reactants in the
GPT reaction chamber is not carefully controlled,
errors can result from either incomplete reaction of
NO or reaction of the resulting NO? with residual
03. For this reason, accurate use of the procedure
is difficult, and further development may be
required before it is recommended for use under
field conditions.
Boric acid Kl is a modification of the NBKI
calibration procedure. This procedure utilizes a 1 -
percent Kl solution acidified with 0.1 molar boric
121
-------
acid. Preliminary results indicate that this
modification vastly improves both the variability
and accuracy of the Kl procedure.27
RELATIONSHIPS BETWEEN AMBIENT
OXIDANT AND OZONE DATA
The relationship, and more specifically the
direction and magnitude of differences, if any,
between chemiluminescence Oa (chemilum-Os)
data and Kl data taken in parallel from ambient
atmospheres are of interest because they relate to
the issue of the definition and justification of the
numerical air quality standard for photochemical
oxidants.26 This issue arises from uncertainty
regarding the specific species responsible for the
health effects attributed to photochemical
oxidants. The problem seems to be whether the
epidemiological evidence or the ozone-specific
toxicoiogical evidence should be used as the main
basis of the numerical air quality standard. Thus,
for example, if the epidemiological evidence alone
were selected as the main basis for the air
standard, then the responsibility for the observed
health effects would be placed on the entire
photochemical pollution complex rather than on
specific constituent(s), since epidemiological
evidence is not pollutant specific. In that case, the
smog constituent(s) measured and correlated with
the epidemiological evidence (i.e., oxidants
determined by Kl) would be viewed as a surrogate
species. This, in turn, introduces two re-
quirements. First, the numerical air quality
standard must be defined either in terms of the
same group of oxidants that was correlated with
the epidemiological evidence (i.e., oxidants
determined by the Kl method) or in terms of an
equivalent species (e.g., oxidants determined by
the chemilum-Oa method). An equivalent species
is one present at concentrations equal to those
determined by the Kl method for measurement of
oxidants. Second, the composition of the
photochemical oxidants mixture, or at least the
ratio of oxidants determined by the chemilum-Oa
method to oxidants determined by the Kl method
should be constant with location and time.
In contrast to the preceding case, if toxicoiogical
evidence is used alone or in conjunction with
epidemiological evidence to place the re-
sponsibility for the health effects on a specific
species, Os for example, then the numerical air
quality standard should be defined in terms of 03,
and the correlation or lack of correlation between
Oa and other smog constituents could be disre-
garded as being irrelevant.
The question of whether the air quality standard
should be defined in terms of oxidant-by-
chemilum-Og, oxidant-by-KI, or any other
measurable entity will not be examined here. The
comparison, however, of oxidant data obtained by
the Kl method and by the chemilum-Os is relevant
and will be discussed briefly.
As discussed in the previous section on Kl
calibration procedures, the experiences from the
laboratory application of Kl procedures have
demonstrated that such procedures, whether used
in standardizing ozone mixtures or in measuring
ambient oxidant, are substantially imprecise and
inaccurate because of the effects from factors that
are not completely understood. When the same
standard Os mixtures are used to calibrate a Kl
method and the chemilum-Oa method, then,
obviously, any difference in ambient meas-
urements obtained by the two methods reflects
errors other than those related to calibration. Such
errors include: (1) systematic errors related to the
different response specificities of the two
methods; (thus, for example, the Kl method should
give higher results because it responds to more
oxidants than the chemilum-Oa method), and
perhaps more important, (2) disagreement
between the two methods will reflect the
unknown-origin random errors discussed in the
preceding section. Because of the random nature
of these latter errors, and because of their
connection to the operator factor, the true
difference in results between two methods cannot
be determined from a single day's side-by-side
comparison of the two methods conducted by a
single operator. Comparisons by a variety of
operators using the same or equivalent calibration
procedures over a span of many days are
necessary if the comparison results are to be
credible.
Results from such multiday comparisons of Kl
procedures with other methods have been
reported in an averaged form by several
investigators. Such results have been reported in
the predecessor to this criteria document60 and
more recently by Ballard et al.,7 Stevens et al.,100
Clark et al.,18 and the California Air Resources
Board.16 These recent results show, in general,
small differences between Kl-oxidant data and
ozone data. Although these differences are usually
in the direction of higher oxidant than ozone, the
122
-------
imprecision of the calibration and operation of the
Kl methods makes drawing conclusions from such
studies difficult. Several studies are illustrated in
Figures 7-1 to 7.5.^,89,91,100
Results from side-by-side comparisons of the Kl
method and one Oa-specific method have been
reported also for individual days and will be briefly
reviewed. Davis and Jensen21 reported
measurements in Florida in which a coulometric
oxidant instrument (Mast) gave readings
consistently lower than those of a chemilu-
minescence ozone detector by a factor of three.
The coulometric instruments were operated using
factory calibrations, while the chemi-
luminescence monitors were calibrated according
to specified standards. The report contains no data
to demonstrate that the instrument would indicate
similar ozone levels if both were sampling the
same calibration atmosphere. Because of the
imprecision of the calibration procedures, it is
probable that a significant fraction of the
discrepancy was caused by calibration.
Furthermore, the differences between the two
instrument readings were on the order of 20 to 30
ppb. Since no measures were taken to remove or
correct for interferences from reducing agents, at
least part of the difference is probably due to
interference in the Mast instrument from S02.
These uncertainties raise pertinent questions
about the value of the Davis and Jensen21 report in
estimating the equivalency of the ozone and
oxidant methods. EPA measurements in St. Louis,
Mo., showed consistently higher Kl-oxidant values
relative to chemilum-Oa values.18 Okita and
Inugami69 reported their ozone measurements in
Musashino, Japan, to be in excellent agreement
with their Kl-oxidant measurements (Figure 7-2).
Carroll et al.,17 using smog chamber irradiated auto
exhaust mixtures, conducted and compared
parallel measurements of oxidant measured by Kl
and of ozone measured by chemiluminescence;
their comparison showed the Kl values to be higher
by 0 to 10 percent (Figure 7-3). Finally, Severs and
Neal64'89 reported the results of their comparison of
the Kl-oxidant and chemilum-Oa measurements of
the Houston atmosphere, which show the
chemilum-Oa values to be generally higher and,
occasionally, to be considerably higher (Figures 7-
4 and 7-5).
Thus with the exception of the Houston data
obtained by Severs and his coworkers,64'89 the Kl-
oxidant measurements appear, as a rule, to be
either roughly equal to or somewhat higher than
the chemilum-Oa measurements, a difference that
is directionally consistent with the difference in
response specificity between these methods. The
significance of the Severs results, although
difficult to assess, does not seem to be crucial.
E
a
a
O
Z
UJ
O
O
O
I
OZON E-CH EM ILUM • — • —
TOTAL OX-MAST
TOTAL OX-TECHNICON
Figure 7-1. Diurnal ozone-oxidant averages from September 4 through September 30, 1971.51
123
-------
First, it is difficult to explain the "irrational"
direction of the Kl-oxidant-versus-chemilum-
ozone disagreement reported by Severs. Second,
there are several other studies thatyielded rational
results. Finally,there is ample evidence attesting to
the poor precision of the Kl methods and,
specifically, to the presence of a substantial
operator error. In view of all these facts and
indications, one would be inclined to conclude,
presently at least, that the data analyzed by Severs
reflect a strong operator error in the oxidant meas-
urements. In a more recent study, Severs et al.90
offer the explanation that the differences in results
found in their method-comparison studies are
largely the result of sampling turbulent
atmospheres with instruments that use different
sampling intervals. This explanation, however, is
not consistent with the sustained method
2.0
I-
z
9
x
o
0.5
0.4
AVERAGE
INDIVIDUAL DAYS
I
11:00 NOON
|<—a.m—»>j-«—
1:00
2:00
— p.m. •
3:00
4:00
LOCAL STANDARD TIME
Figure 7-2. Ozone-oxidant ratios in Musashino, Japan.69
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
OZONE BY CHEMILUMINESCENCE, ppm
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
OZONE BY CHEMILUMINESCENCE, ppm
Figure 7-3. Comparison of the Kl, Mast, and chemiluminescence methods for measuring ozone in
irradiated exhaust mixtures.91
124
-------
disagreement over a period of several hours, as
shown in Figure 7-5.64 The major and well-
supported conclusion from the discussion in this
section on the data comparison of Kl oxidant
versus chemilum-O3 is that definition of airquality
and of air quality standards in terms of oxidant
measured by the commonly used Kl methods
would create problems because of the inaccuracy
and imprecision of such methods. These problems
would be nearly eliminated, however, if the air
quality and the air standard could be defined in
terms of ozone.
o
-------
PBzN to the relatively more stable methybenzoate.
The lower limit of sensitivity of this technique is 1
5,20,3e,«.<5,7<,94,95
5'20'38'"'45'59'74'9'
report from the Netherlands,5''8''5'9'74'' the
evidence indicates that PBzN does not occur in the
ambient air in measurable concentra-
tions 5>20>38'')')'')5>7')'94<95
Gas chromatography with either electron-
capture or FID appears presently to be the most
viable method for the analyses of the PAN family.
MEASUREMENT OF HYDROCARBONS
The types of hydrocarbons covered in this
discussion are those gas-phase hydrocarbons that
are likely to be found in urban atmospheres.
Measurement methodsforthese hydrocarbons fall
into two classes. Those that measure nonmethane
hydrocarbons (NMHC) and those that measure
individual hydrocarbons.
Measurement of Nonmethane Hydrocarbons
In typical polluted air samples, the principal
hydrocarbon component, methanejChU), is usually
more abundant than all other hydrocarbons
combined. Methane, however, is inert in most
photochemical reactions. Therefore, hydrocarbon
measurements are usually corrected for ChU by a
subtractive technique to better approximate the
reactive component of the total hydrocarbon
mixtures in the ambient air. Both CH4 and total
0.20 —
0.15
a
a
111*
O
0.10
0.05 —
8:00
TIME OF DAY
Figure 7-5. Comparative measurer),b,,19 of ambient oxWant/ozone by the potassium iodide (Kl),
chemiluminescence (Chem), and ultraviolet (UV) methods.64
126
-------
hydrocarbons are measured by continuous
monitoring analyzers that utilize the flame-
ionization detector (FID) as the sensing element.
Originally developed as a detector for gas
chromatography, the FID was later adapted for
total hydrocarbon analysis.4 In this technique, a
sensitive electrometer detects the increase in ion
intensity resulting from the introduction of air
containiny any hydrocarbon compounds into a
hydrogen flame.
The response of FID analyzers is related not only
to the concentration of the hydrocarbon being
measured, but also to the effective carbon number
of the hydrocarbon compound. The effective
carbon number varies depending on the number of
carbon atoms in the molecule and on the type of
compound {e.g., aliphatic, aromatic, olefinic,
acetylenic, etc.). Thus, without knowing which
hydrocarbon compound is being measured, the
reading cannot be related to the actual
concentration of the total hydrocarbons in the air
sampled.
For this reason, FID results are usually
expressed in terms of the calibration gas used (for
example, ppm C as ChU). Carbon atoms bound to
oxygen, nitrogen, or halogens give reduced
response or no response.96 There is no response to
nitrogen, carbon monoxide, carbon dioxide, or
water vapor. An oxygen effect does occur, but it
can be minimized by appropriate adjustment of
operating condition.
The response of the FID is rapid, and with careful
optimization of operating parameters and proper
calibration, it is sensitive to a fraction of a ppm C as
ChU, A number of studies have been carried out on
optimization of such operating variables as gas
flow rates, detector temperature, and oxygen
content of air, all of which can affect FID
response.
11,28,31
Early measurements of NMHC involved the use
of a carbon column pretreated with ChU3'70'82This
column was used to remove all other hydrocarbons
from the ambient sample. A total hydrocarbon
analyzer with a carbon column was operated either
parallel or a..ernatively with a total hydrocarbon
analyzer without a carbon column. Methane and
total hydrocarbons were measured, and by
difference, the NMHC value was computed. This
method was subject to much variability because of
the unpredictable behavior of carbon columns
when exposed to ambient conditions. However,
when the carbon column was relatively new and
the NMHC concentrations were high, the method
was reasonably accurate,
EPA REFERENCE METHOD
The EPA reference method for NMHC, which
was promulgated in 1971, involves gas
chromatographic separation of CH4 from total
hydrocarbons in an air sample followed by FID of
both CH4 and total hydrocarbons.106 The NMHC
value is obtained by subtraction of the two
measurements. In a 1974 survey, some 1 60 users
of this method were identified.76
Several studies have been conducted to
estimate the quality of ambient air data being
obtained with this method. One study76 involved
the analysis of synthetic hydrocarbon mixtures in
compressed air cylinders by 16 different users of
the reference method. The NMHC concentrations
tested were 0.23 and 2.90 ppm. The results of this
study are given in Table 7-4 and indicate the
percent error from the known concentrations.
TABLE 7-4. PERCENT DIFFERENCE FROM KNOWN
CONCENTRATIONS OF NONMETHANE
HYDROCARBONS OBTAINED BY SIXTEEN USERS
Known
concentration.
ppm
023
290
100
6
2
50-100
4
% difference
20-50
3
3
10-20
2
2
0-10
1
9
The results show that at the level of theNAAQS,
0.24 ppm carbon, most of the measurements were
in error by 50 to 100 percent. At 2.90 ppm, most of
the measurements were in error by only 0 to 20
percent. Thus the higher the concentration tested,
the better the accuracy.
The overriding causes of the inaccuracies were
the inability of the instrumentation to measure the
low levels of NMHC, the complexity of the
instrumentation, and the inability of the average
user to identify and correct problems.
Another study57 involved monitoring the same
ambient air with five different commercial FID
instruments operated by skilled professionals, The
results show that different analyzer pairs agreed
with 0,1 to 0.5 ppm carbon. Although these
differences represent only 1 to 5 percent of the full-
scale (0 to 10 ppm) range of the instruments and
are normal errors for ambient monitoring, they are
obviously very large relative to the NAAQS of 0.24
ppm carbon, (The 0- to 10-ppm range is necessary
to include all ambient values.) Thus even under
127
-------
optimum operating conditions, agreement
between instruments is poor at low
concentrations. This result is not surprising when
one realizes that the instrumentation is not
optimized for measurement at these low levels.
Several studies on the performance of ambient
hydrocarbon monitors have been published.4671 8197
In a recent EPA-sponsored study,35 the EPA ref-
erence method was subjected to a comprehen-
sive evaluation, including testing^ of six commer-
cial instruments. Results showed, in general, poor
performance of the commercial instruments, the
major problems being wide differences in re-
sponse to different NMHC species and discrep-
ancies apparently related to ambient water vapor
variation.
HEAT OF COMBUSTION METHOD
A different approach to measurement of
nonmethane hydrocarbons was developed in
1973.75 This technique utilizes the fact that CH4
has a higher heat of combustion than other
hydrocarbons. One portion of the air sample
passes through a catalyst bed where all
hydrocarbons, except CH4, are combusted. This
CH4-only stream passes to one FID. The other
portion of the sample passes directly to a second
FID for a total hydrocarbon measurement. By
simultaneous processing of the signals from the
two FID's and by subtraction of the CH4-only value
from the total hydrocarbon value, the NMHC value
can be obtained. This technique has several
advantages over the reference method. It is
simpler to use, it produces continuous NMHC
values, and the instrument system can maintain its
own zero. However, most of the shortcomings
attributed to the EPA reference method also apply
to this technique.
Measurement of Individual Hydrocarbons
GAS CHROMATOGRAPHY
Gas chromatography (GC) is the most effective
method for determining the concentration of
individual hydrocarbons that make up the
complicated hydrocarbon mixture such as is found
in auto exhaust and the ambient atmosphere. With
FID, the method is sensitive down to the ppb range.
GC methods in general involve the separation of a
complex mixture into individual compounds. This
separation is achieved using both capillary (liquid
phase coated on the inner surface of the column)
and analytical columns. The latter may contain
either a solid polymeric adsorbent (gas-solid
chromatography) or an inert support coated with a
liquid (gas-liquid chromatography!. An inert gas
(carrier gas) continuously flows through the GC
column. When a complex sample is injected onto
the GC column, the carrier gas carries the sample
through the column. Separation of the mixture
occurs as the result of the different propensities of
the respective components for interacting with the
column substrate. The net result is the separation
of a complex mixture into a number of individual
components. The degree of separation depends on
column length, carrier gas flow rate, temperature,
and a variety of other parameters. The column
support is selected from numerous available
materials, depending on the types of components
that make up the sample mixture,
GC techniques have been used to investigate the
hydrocarbon composition of both urban and rural
atmospheres.
2,47,50,78
In urban atmospheres.
hydrocarbons generally exist at levels high enough
to permit the collection of samples without
concentration. Samples can be collected in inert,
flexible bags and transported to a central
laboratory for GC analysis. Problems often arise
with this sampling technique because of
hydrocarbon losses resulting from adsorption on
the walls of the bags, contamination of bags, and
reactions that occur after the sample has been
collected. Since the hydrocarbon components in
rural ambient atmospheres are typically found in
ppb concentrations, cryogenic trapping techniques
for large air samples are required. Quantitative
measurements of the GC peaks can be made with
integrator and computer interfacing. Quantitative
evaluation of the individual peaks can be made
without standardization of each compound peak
analyzed. Computation of each component
concentration can be made by using an average
per-carbon factor determined by known
concentrations of the identified components,26
In recent years, the development of high-
resolution capillary and support-coated, open
tubular columns combined with wide-range
temperature programming from subambient to
elevated temperatures have provided better
separation of the complex ambient hydrocarbon
mixture. The complexity of the ambient air mixture,
however, still prevents complete resolution of the
air sample into all of its components. To determine
the purity of the GC peaks and to obtain
information for the identification of unknown
peaks, a combination of GC with either infrared (IR)
128
-------
or mass spectrometry (MS) can be used. These
approaches, however, require both large sample
concentration and highly trained personnel.
GC - CHEMILUMINESCENCE DETECTOR
A procedure has been developed that permits
detection of unsaturated hydrocarbons (e.g.,
ethylene, propylene, and terpenes) using
separation with a gas chromatograph and analysis
with a chemiluminescence detector.56'88 The
unsaturated compound of interest reacts with an
ozone stream to produce light in the region of 400
A. Chromatographic separation makes the
technique specific for each compound. Sensitivity
for this method is in the ppb range.88
INFRARED SPECTROMETRIC TECHNIQUES
Recent advances in long-path IR techniques
using Fourier-transform spectroscopy (FS) have
resulted in an increase in sensitivity (10 ppb) for
measurement of selected hydrocarbons such as
ethylene, propylene, methane, formaldehyde, and
PAN. Whereas previous IR techniques required a
concentration step to measure ambient levels of
hydrocarbons, the FS approach allows for ^direct
measurement at ow pollutant concentrations.33
The system has been used successfully in
atmospheric studhs as well as in characterization
of photochemice reactions.34 The size and
complexity of th; system, however, limits its
usefulness for nx tine air monitoring.
Measurement of Reactive Hydrocarbons
Hydrocarbons differ greatly in the ability to
produce photoch; mical smog. For example, CH4
and acetylene can be considered unreactive, and
ethylene and n-butane, reactive hydrocarbons.
However, n-butane concentrations several fold
higher than ethylene concentrations are required
to produce similar rates of smog formation. There
is accordingly a need for a measurement method
that gives an indication of total hydrocarbon
reactivity (i.e., a measurement that gives the sum
of the concentrations of each reactive hydrocarbon
species multiplied by a reactivity factor). Fontijnet
al.30 have recently developed a chemilumi-
nescence method that appears to satisfy this need.
This method is based on the difference between
the light emission intensities at 308.9 and 312.2
nm that result from the reaction between the
hydrocarbon sample and oxygen atoms under low
total pressure (1 torr) conditions. A limit of
sensitivity of 0.05 ppm ethylene-equivalent
hydrocarbon and a linear response to individual
hydrocarbons up to at least 3000 ppm have been
obtained. The response to hydrocarbon mixtures
appears to be additive, and interferences from CO,
C02, S02, CH4, CaH2 and NOX in concentrations
typical of those in auto exhaust were found to be
negligible. Such an instrument response obviously
measures the reactivity of the hydrocarbons with
oxygen atoms. Such reactivity, in turn, is a
measure—although a crude one—of the sample's
ability to produce oxidants.
Calibration
Calibration at regular intervals is, of course,
required for all analytical methods for measuring
air pollutants. Calibration techniques and the
required frequency of calibration vary for different
pollutants. The preferred procedure consists of
passing known concentrations of the pollutant in
air into the measuring system. This dynamic
procedure calibrates not only the detector
response, but also the inlet system to the detector.
Calibration gases in steel or aluminum cylinders
are commonly used to calibrate hydrocarbon an-
alyzers. Cylinder gases may be diluted with clean,
dry air to give the desired concentration range. Dry
air with a hydrocarbon concentration that does not
exceed 0.1 ppm is needed for dilution and for
providing zero gas for the instrument. Calibration
gases and zero gas are available commercially, and
some standard gases, such as propane in air, are
available from the National Bureau of Standards as
standard reference materials. All commercial
mixtures should be referenced against a standard
gas whenever possible. The air used for
instrument zeroing purposes should contain the
normal constituents of clean air— that is, Oa at 20
percent, N2 at 79 percent, C02 at 320 ppm, and the
normal levels of inert gases such as argon (Ar),
helium (He), neon (Ne), and nitrous oxide (N20).
Permeation tubes can also be used to generate
known concentrations of certain hydrocarbons
such as butane, which can be condensed at
moderate pressures. The gas trapped inside the
permeation tube will diffuse through the tube wall
at a rate dependent on the surface area and the
temperature of the tube.52'68 The permeation rate
of the tube can be determined gravimetrically by
holding the tube at a constant temperature
(+0.1 °C) and measuring weight loss with time. The
calculated permeation rate remains constant for
properly prepared tubes. Because the permeation
rates of different hydrocarbons are significantly
129
-------
different, only one hydrocarbon can be used in a
given permeation tube,
Another dynamic calibration technique involves
the addition of a measured a mount of pollutant to a
known, fixed volume of air in a large vessel and
then drawing the atmosphere into the analyzer.66
The addition may be made by syringe injection or
by the crushing of a weighted glass ampule
containing the pollutant. Rigid vessels or plastic
bags, which have the advantage of collapsing as
the sample is withdrawn, may also be used. Bags
must be made of an inert plastic to avoid changes
in concentration through absorption or reaction
with the walls of the bag. The bags must also be
checked for permeation rates of the compounds
being used, since permeation can result in loss of
sample or, in some cases, contamination of the
sample as a result of bag permeation by ambient
hydrocarbons. The above techniques are preferred
for more reactive hydrocarbons that do not have
sufficient stability to be stored in cylinders.
MEASUREMENT OF NITROGEN OXIDES
The NO, involved in photochemical oxidant
formation are NO and N02. Since oxidant control
regulations and practices to date have been based
on unilateral control of the hydrocarbon precursor,
interest in the NO, factor was only peripheral and
of research nature. The recent findings, however,
on the need to consider the NO, factor in
formulating oxidant control strategies have raised
questions regarding adequacy of the existing
analytical methods for ambient NO, measurement.
Since the primary needs insofar as the oxidant
problem is concerned are short-term (hourly) NO,
concentration data for urban atmospheres and
short-term (hourly) data for rural atmospheres,
analytical methods used for oxidant-related
applications must be capable of short-term
sampling and analysis and interference-free
response, and they must be sensitive. This
discussion will thus be limited to continuous
methods for measuring NO2 and NO.
Measurement of
CHEMILUMINESCENCE METHODS (EPA
REFERENCE METHODS)
On December 1, 1976, EPA promulgated a new
measurement principle and calibration procedures
for reference methods for measuring NO2.105The
measurement principle and calibration procedures
are described in Appendix F of 40 CFR Part 50. The
measurement principle is based on the gas-phase
reaction of 03 and NO. Atmospheric con-
centrations of N02 are measured indirectly by
photometrically measuring the light intensity (at
wavelengths greater than 600 nm), resulting from
the chemiluminescent reaction of NO with
29,54,99 NO2 is first quantitatively reduced to
NO
13,42,99
by means of a converter. NO, which
commonly exists in ambient air together with N02,
passes through the converter unchanged, causing
a resulting total NO, concentration equal to NO +
N02. A sample of the input air is also measured
without having passed through the converter. This
latter NO measurement is subtracted from the
former measurement (NO + NO2) to yield the final
NOZ measurement. The NO and NO + NO2
measurements may be made concurrently with
dual systems or cyclically with the same system.
Two calibration procedures are prescribed in
Appendix F of 40 CFR Part 50. One is a gas-phase
titration procedure referenced to an NO-tn-
nitrogen standard. The other calibration procedure
is referenced to an NO permeation device. Both
of these standards can be obtained from the
National Bureau of Standards as standard
reference materials. Chemiluminescence
N0/N0«/N02 analyzers will respond to other
nitrogen-containing compounds such as PAN,
which might be reduced to NO in the thermal
converter.107 Atmospheric concentrations of these
potential interferences are generally low relative
to NO2, and valid NO2 measurements may be
obtained.
GRIESS-SALTZMAN METHODS
Before the development of Chemiluminescence
analyzers, most NO2 data were collected using
methods based on variations of the Griess-
Saltzman83 methodology. These methods are
based on the specific reaction of the nitrite ion
(NOi) with diazotizing-couplmg reagents to form a
deeply colored azo dye. The NOs in the ambient air
is converted to nitrite ion on contact with an
absorbing solution. As applied to continuous
analyzers,84 the absorbing solution contains the
diazotizmg-coupling reagents, andtheabsorbance
of the azo dye is measured continuously with a
flowing photometer cell. The absorbance of the azo
dye solution is directly proportional to the
concentration of N02 absorbed.
For ordinary atmospheres, interferences have
been claimed to be of negligible importance. The
method is commonly calibrated with standard
nitrite solutions, assuming 0.72 mole of nitrite to
130
-------
be equivalent to 1 mole of N02. This latter
assumption has been at issue.86 In addition, there
are indications of an ozone interference problem,8
For all practical purposes, this method has been
replaced with the chemiluminescence method.
Measurement of NO
Before the advent of the chemiluminescence
method, NO measurements were made almost
exclusively by contacting the air sample with a
solid or liquid oxidant and measuring resulting N02
by conventional azo dye colorimetric pro-
cedures.48'49 Currently, the chemiluminescence
method is believed to be unquestionably superior
to all methods available. Thus, for example,
relative to the continuous Griess-Saltzman, the
.chemiluminescence method is more accurate
because it can be calibrated more accurately with
standard NO mixtures, and also because it is
subject to practically no interferences from other
vapors. The speed of response of chemilumi-
nescence-NO instruments is typically only a
few seconds, and their sensitivity is about 5
ppb. Overall, such instruments are practical,
reliable, and suitable for field use.
SUMMARY
The chemiluminescence method, based on the
gas-phase reaction of Oa with ethylene, was
adopted by EPA as the reference method and is
now being used extensively. Several techniques
using commercial analyzers have been designated
as reference methods under the EPA ambient air
monitoring reference and equivalent methods
regulations. Data indicate that the performance of
these analyzers is better than that specified by the
EPA regulations. Continuous analyzers based on
gas-solid chemiluminescence and UV photometry
have been developed, and techniques using such
analyzers have been designated as equivalent
methods.
The methods for measuring total oxidants are
based on the oxidation of Kl and the electro-
chemical or colorimetric detection of I2. Nitrogen
dioxide causes a positive interference, and S02
results in negative interference. Serious problems
arise, however, with methods for correcting for
S02 and N02 interferences.
Extensive studies in recent years have shown
the various Kl calibration procedures for oxidantor
03 analyzers to be deficient. These methods suffer
from systematic, accuracy-related errors as well as
from random, precision-related errors. EPA
intends to study and evaluate several new
calibration procedures and to amend Appendix D of
40 CFR Part 50 to revise the 1-percent NBKI
calibration procedure or replace it with one or more
of several alternatives under consideration. The
alternative procedures under consideration are: (1)
GPT with excess NO; (2) UV photometry; (3) GPT
with excess O3; and (4) boric acid Kl.
Analysis of the data indicates that oxidant-by-KI
measurements of oxidant levels are generally
comparable or slightly higher than levels of ozone
as measured by ozone-specific techniques.
However, most oxidant/Oa comparison studies
suffer from the imprecision of the oxidant methods
and from the lack of data regarding levels of
compounds known to interfere with the Kl method.
Commercial instruments for ambient total
NMHC monitoring have been tested extensively
and found to be inaccurate at levels approaching
the 0,24-ppm C air quality standard. At present,
ambient hydrocarbons can be accurately
measured only by sophisticated gas chroma-
tography. Routine methods are, however, under
development.
On December 1, 1976, EPA promulgated a new
measurement principle and calibration procedure
for measuring N02. The measurement principle is
based on the gas-phase chemilummescent
reaction of Oa and NO. Two calibration procedures
are prescribed. One is a GPT procedure referenced
to an NO-in-nitrogen standard. The other
procedure is referenced to an NOa permeation
device. Before the development of chemi-
luminescence analyses, most NO? data were
collected using methods based on variations of the
Griess-Saltzman methodology. For all practical
purposes, this method has been replaced with the
chemiluminescence method.
REFERENCES FOR CHAPTER 7
1. Air Pollution Control Office Air Quality Criteria for
Nitrogen Oxides. APCO Publication No AP-B4, U S
Environmental Protection Agency, Washington, D C ,
January 1971
2 Altshuller, A P,W A Lonneman, F D Sutterfield, and
S. L Kopczynski Hydrocarbon composition of the
atmosphere of the Los Angeles Basin — 1967 Environ
Sci. Technol. 51009-1016, 1971
3 Altshuller, A P,G C Ortman, B E Saltzman, and R E
Nehgan Continuous monitoring of methane and other
hydrocarbons in urban atmospheres J Air Pollut
Control Assoc ;687-91,1966
4. Andreatch, A J , and R Femland Continuous trace
hydrocarbon analysis by flame lonization Anal Chem
32 1021-1024, 1960
5 Appel, BRA new and more sensitive procedure for
131
-------
analysts of peroxybenzoyl nitrate J. Air Pollut. Control
Assoc 23 1042-1044, 1973.
6. Ballard, L F , J. B. Tommerdahl, C. E, Decker, T. M
Royal, and L K Matus. Field Evaluation of New Air
Pollution Monitoring Systems: St. Lou is Study, Phase 1.
Interim Report Contract No EPA 70-101, U.S.
Environmental Protection Agency, Research Triangle
Institute, Research Triangle Park, N.C., 1971.
7 Ballard, L F., J, B. Tommerdahl, C. E. Decker, T. M
Royal, and D R. Nifong, Field Evaluation of New Air
Pollution Monitoring Systems The Los Angeles Study.
APTD-0775, U S. Environmental Protection Agency,
Research Triangle Park, N C., April 1971
8 Baumgardner, R E , T A Clark, and R. K Stevens.
Comparison of instrumental methods to measure
nitrogen dioxide. Environ Sci Technol. 9:67-69, 1975.
9 Beard, M E,J H. Margeson, and E C. Ellis. Evaluation
of 1% Neutral Buffered Potassium Iodide Procedure for
Calibration of Ozone Monitors. EPA-600/4-77-005,
U S Environmental Protection Agency, Research
Triangle Park, N C., January 1977.
10 Becker, K H , U Schurath, and H Seitz Ozone-olefin
reactions in the gas phase. 1 Rate constants and
activation energies Int J Chem. Kinet 6:725-739,
1974
11 Bocek, P.J Novak, and J Janak Effect of pressure on
the performance of the flame lonization detector J
Chromatogr. 48 431-436, 1969.
12 Bowman, L D , and R. F.Horak A continuous ultraviolet
ozone photometer Air Qual. Instrum. 2'305-315, 1974
13 Breitenbach, L P, and M Shelef Development of a
method for the analysis of NOz and NH3 by NO-
measuring instruments. J, Air Pollut Control Assoc.
2.3.128-131, 1973,
14 Burton, C S , P,J Bekowies, R I Pollack, and P Connell.
Oxidant/Ozone Ambient Measurement Methods: An
Assessment and Evaluation EF76-111R, Systems
Applications, Inc , San Rafael, Calif., August 10, 1976
15 California Air Resources Board Differences in ARB and
LA-APCD oxidant readings Calif. Air Resour. Board
Bull 5 1, 1974
16 California Air Resources Board Ozone data Calif. Air
Qual Data 62-39, January-March 1974.
17 Carroll, H, B Dimitnades, and R. M. Ray. A Chemi-
lummescent Detector for Ozone Measurement Report
of Investigation Rl 7650 U S Department of the
Interior, Bureau of Mines, Washington, D.C. 1972.
1 8 Clark, T A,, R. E. Baumgardner, R. K Stevens, and K. J
Krost Evaluation of new ozone monitoring instruments
by measuring in non-urban atmospheres In.
Instrumentation for Monitoring Air Quality ASTM
Special Technical Publication 555, American Society for
Testing and Materials, Philadelphia, Pa , 1974. pp. 101-
111
19 Clements, J. B Summary Report Workshop on Ozone
Measurement by Potassium Iodide Method EPA-
650/4-75-007, U S Environmental Protection Agency,
Research Triangle Park, N.C., February 1975
20 Darley, E. F , K A Kettner, and E. R Stephens Analysis
of peroxyacetyl nitrates by gas chromatography with
electron capture detection Anal Chem 35589-591,
1963
21 Davis, D R , and R E Jensen Low level ozone and
weather systems In. Specialty Conference On
Ozone/Oxidants-lnteractions with the Total
Environment. Air Pollution Control Association,
Philadelphia, Pa., 1976 pp 242-251.
22. Decker, C.E., T.M. Royal, J.B, Tommerdahl, and LK
Matus. Field Evaluation of NewAir Pollution Monitoring
Systems: St. Louis Study, Phase II Interim Report. U S
Environmental Protection Agency Contract No EPA 70-
101, Research Triangle Institute, Research Triangle
Park, N.C., 1971
23. DeMore, W. B., and O Raper, Hartley band extinction
coefficients of ozone in the gas phase and in liquid
nitrogen, carbon monoxide, and argon. J. Phys, Chem.
68:412-414, 1964.
24. DeMore, W. B., J C. Romanovsky, M. Feldstein, W. J
Hamming, P K Mueller Interagency comparison of
iodometnc methods for ozone determination. In:
Calibration in Air Monitoring. ASTM Special "echnical
Publication 598, American Society for Testing and
Materials, Philadelphia, Pa , 1976. pp. 131-143.
25. Dietz, W. A. Response factors for gas chromatographic
analyses J Gas Chromatogr. 5:68-71, 1967
26. Dimitnades, B., and A. P Altshuller International
Conference on Oxidant Problems' Analysis of the
Evidence and Viewpoints Presented. Parti. Definition of
Key Issues. J. Air Pollut Control Assoc. 27 299-307,
1977.
27 Flamm, D. L Analysis of ozone at low concentrations
with boric acid buffered Kl Environ Sci. Technol.
11 978-983, 1977.
28 Folmer, 0 F , Jr., andD.J Haase. A statistical study of
gas chromatographic systems employing flame
ionization detectors. Anal Chim Acta. 48 63-78,1969.
29 Fontijn, A., A. J. Sabadell, and R. J Ronco.
Homogeneous chemilummescent measurement of
nitric oxide with ozone implications for selective
monitoring of gaseous pollutants Anal Chem 42:575-
579, 1970.
30 Fontijn, A, H N Volltrauer, and R Ellison. Chemi-
lummescent Reactive Hydrocarbon Analyzer for Mobile
Sources. EPA-650/2-75-069, U S Environmental
Protection Agency, Research Triangle Park, N.C , June
1975.
31 Grant, D W , and A. Clarke A systematic study of the
quantitative effects of instrument control on analytical
precision in flame ionization gas chromatography. Anal
Chem 43 1951-1957, 1971
32 Griggs, M Absorption coefficients of ozone in the
ultraviolet and visible regions J. Chem Phys 49 857-
859, 1968
33 Hanst, P L,A S. Lefohn, and B.W. Gay, Jr. Detection of
atmospheric pollutants at parts-per-billion levels by
infrared spectroscopy. Appl. Spectrosc 27188-198,
1973.
34. Hanst,P L,W E.Wilson,R K. Patterson, B.W Gay.Jr.,
L. W. Chaney, and C S. Burton A Spectroscopic Study
of California Smog. EPA-650/4-75-006, U S
Environmental Protection Agency, Research Triangle
Park, N C , February 1975
35 Harrison, J W , M L Timmons, R. B. Denyszyn, and C. E.
Decker Evaluation of the EPA Reference Method for
Measurement of Nonmethane Hydrocarbons. EPA-
600/4-77-033, U S Environmental Protection Agency,
Research Triangle Park, N C , June 1977
36 Hearn, A G The absorption of ozone m the ultra-violet
132
-------
and visible regions of the spectrum, Proc Phys, Soc.
(London) 78:932-940, 1961
37. Hersh, P., and R. Deunnger. Galvanic monitoring of
ozone in air Anal. Chem. 35.897-899, 1 963
38 Heuss, J M., andW, A. Glasson Hydrocarbon reactivity
and eye irritation Environ Sci. Technol 2,1109-1116,
1968.
39 Hodgeson, J. A. Ozone. Sampling analysis and methods
evaluation Presented at the 15th Conference on
Methods in Air Pollution Studies, Regional Workshop on
Quality Assurance, California State University, Long
Beach, Calif., January 14-15, 1976, and Sacramento,
Calif, January 20-21, 1976.
40 Hodgeson, J. A. Review of analytical methods for
atmospheric oxidant measurements Int. J Environ.
Anal. Chem. 2 113-132, 1972.
41 Hodgoson, J A , C B. Bennett, H. L. Kelly, and B. A.
Mitchell Ozone measurements by lodometry,
ultraviolet photometry and gas-phase titration
Presented at the American Society for Testing Materials
(ASTM) Symposium on Calibration in Air Monitoring,
University of Colorado, Boulder, Colo , August 5-7,
1975.
42 Hodgeson, J A., K. A. Rehme, B. E. Martin, and R. K.
Stevens Measurements for atmospheric oxides of
nitrogen and ammonia by cherniluminescence
Presented at 65th Annual Meeting, Air Pollution Control
Association, Miami Beach, Fla , June 18-22, 1972.
43. Inn, E. C Y., and Y. Tanaka Absorption coefficient of
ozone in the ultraviolet and visible regions. J. Opt. Sco
Am, 43870-873, 1953
44. Intersociety-Committee Tentative Method of Analysis
for Peroxyacetylnitrate (PAN) in the Atmosphere in
Methods of Sampling and Analysis. Amer. PubhcHealth
Assoc Washington, DC, 1972. pp. 215-219
45 Izumikawa, H., M, Hayafuku, K. Nakano, K. Asakino, and
T Ohira Continuous determination and quantification
of atmospheric PAN Tokyo-to Kogia Kenkyusho Nenpo
4.41-49, 1973
46. Kaplan, H. Plumbing problems—gases for 6800
hydrocarbon analyzers In Instrumentation for
Monitoring Air Quality ASTM Special Technical
Publication 555, American Society for Testing and
Materials, Philadelphia, Pa , 1974. pp. 85-87.
47 Kopczynski, S L,W A Lonneman, T Winfield, and R
Seila Gaseous pollutants in St. Louis and other cities J
Air Poltut Control Assoc 25251-255,1975
48. Kothny, E.L,W A Cook, B Dimitnades, E. F. Ferrand,
C. A Johnson, L. Levin, P. W McDaniel, and B. E
Saltzman Tentative method of analysis of nitric oxide
content of the atmosphere Health Lab. Sci. 371-74,
1972
49 Levaggi, D , E. L Kothnv, T Belsky, E. DeVera, and P. K.
Mueller Quantitative analysis of nitric oxide in
presence of nitrogen dioxide at atmospheric
concentrations. Environ. Sci Technol. S:348-350,
1974
50. Lonneman, W. A., S. L. Kopczynski, P. E. Parley, and F
D. Sutterfield. Hydrocarbon composition of urban air
pollution. Environ. Sci Technol. 8 229-236, 1974
51. Los Angeles County Air Pollution Control District and
California Air Resources Board A study of the effect of
atmospheric humidity on analytical oxidant
measurement methods Presented at the 15th
Conference on Methods in Air Pollution Studies
Regional Workshops on Quality Assurance, California
State University, Long Beach, Calif, January 14-15,
1976, and Sacramento, Calif, January 20-21, 1976
52. Lucero, D. J. Performance characteristics of permeation
tubes. Anal. Chem. 4J-1744-1749, 1971
53. MacPhee, R. D, J E Higuchi, and F K V Leh
Comparison of oxidant measurement methods,
ultraviolet photometry, and moisture effects In
Specialty Conference On. Ozone/Oxidants—
Interactions with the Total Environment Air Pollution
Control Association, Pittsburgh, Pa , 1 976. pp. 166-179
54. Martin, B E., J A Hodgeson, and R K Stevens
Detection of nitric oxide cherniluminescence at
atmospheric pressure Prepr. Pap Nail Meet Div
Environ Chem. Am. Chem. Soc ?2 182, 1972.
55 Mast, G. M , and H E Saunders Research and
development of the instrumentation of ozone sensing
ISA Trans. 7:325-328, 1962.
56 McClenny.W A , B. E Martin,R.E Baumgardner, Jr, R
K, Stevens, and A. E. O'Keeffe, Detection of vinyl
chloride and related compounds using a gas
chromatographic cherniluminescence technique.
Environ Sci, Technot. 70.810-813, 1976.
57 McElroy, F. F, and V L Thompson Hydrocarbon
Measurement Discrepancies Among Various Analyzers
Using Flarne-lomzation Detectors. EPA-600/4-75-010,
U S Environmental Protection Agency, Research
Triangle Park, N.C., September 1975
58. McKee, H.C , R.E.Childers, andV.B, Parr. Collaborative
Study of Reference Method for Measurement of
Photochemical Oxidants in the Atmosphere (Ozone-
Ethylene Cherrulummescent Method). EPA-650/4-75-
016, U.S. Environmental Protection Agency, Research
Triangle Park, N,C., February 1975
59 Meijer, G. M., and H. Nieboer Determination of
peroxybenzoyl nitrate (PBZN) in ambient air In. Ozon
und Begleitungssubstanzen im photochemischen
smog. VID Ber. (270):55-56, 1977.
60. National Air Pollution Control Administration. Air
Quality Criteria for Photochemical Oxidants NAPCA
Publication No. AP-63, U.S. Department of Health,
Education, and Welfare, Public Health Service,
Washington, D C , March 1970
61. National Air Pollution Control Administration Air
Quality Criteria for Hydrocarbons NAPCA Publication
No, AP-64, U S Department of Health, Education, and
Welfare, Public Health Service, Washington, D C ,
March 1970.
62. National Research Council. Analytical Studies for the
U.S. Environmental Protection Agency Vol IV
Environmental Monitoring National Academy of
Sciences, Washington, D C , 1 977.
63. National Research Council Ozone and Other
Photochemical Oxidants National Academy of
Sciences, Washington, D.C., 1977
64 Neal, R., and R. K. Severs Simultaneous
chemiluminescent and Kl oxidant measurements in
Houston, Texas, 1975. In: Specialty Conference on.
Ozone/Oxidants—Interactions with the Total
Environment, Air Pollution Control Association,
Pittsburgh, Pa., 1976 pp. 180-188.
65. Nederbragt, G W,, A Van der Horst, and J Van Duijn
133
-------
Rapid ozone determination near an accelerator Nature
(London) 206:87, 1965.
66. Nelson, G. 0. Controlled Test Atmospheres—Principles
and Techniques. Ann Arbor, Mich., Ann Arbor Science
Publishers, Inc. 1971
67 Office of Air Quality Planning and Standards. Standing
Air Monitoring Work Group. Air Monitoring Strategy for
Implementation Plans. EPA-450/2-77-010 U.S.
Environmental Protection Agency, Research Triangle
Park, N.C., June 1977.
68. O'Keeffe, A. E., andG. C.Ortman. Primary standards for
trace gas analysis. Anal Chem. 38 760-763, 1966.
69. Okita, T , and M. Inugami. Measurement of ozone by
chemiluminescent methods. Koshu Eisenn Kenkyu
Hokoku 20.207-211, 1971.
70. Ortman, G. C. Monitoring methane in atmosphere with
a flame lonization detector. Anal. Chem. 3S:644-656,
1966.
71 Ortman, G. C., and V. L. Thompson. Performance of
hydrocarbon monitoring instrumentation. In:
Instrumentation for Monitoring Air Quality. ASTM
Special Technical Publication 555, American Society for
Testing and Materials, Philadelphia, Pa., 1974. pp.74-
84
72. Ott, W. Development of criteria for citing air monitoring
stations. J. Air Pollut.Control Assoc 27 543-547,1977.
73. Paur, R. J., R. E. Baumgardner, W. A. McClenny, and R.
K. Stevens. Status of method forthe calibration of ozone
monitors. Prepr Pap. Natl Meet, Div. Environ. Chem. J
Am Chem. Soc /6.185-188, 1976
74. Pitts, J. N ,Jr, H. Fuhr, J S. Gaffney, and J.W. Peters.
Chemiluminescent reactions of peroxyacetylnitrate and
ozone with triethylamine. Possible atmospheric monitor
for peroxyacetylnitrate. Environ. Sci Techno). 7'550-
552, 1973.
75. Poll, A A., andT. L.Zinn. A new continuous monitor for
methane and non-methane hydrocarbons. Anal.
Instrum. / 7:135-142, 1973.
76. Reckner, L. R. Survey of Users of EPA-Reference
Method for Measurement of Non-Methane
Hydrocarbons in Ambient Air EPA-650/4-75-008, U.S
Environmental Protection Agency, Research Triangle
Park, N.C., December 1974.
77. Regener, V. H. Measurement of atmospheric ozone with
the chemiluminescent method J. Geophys. Res.
69.3795-3800, 1964.
78 Regener, V H On a sensitive method for the recording
of atmospheric ozone. J. Geophys Res. 65:3975-3977,
1960.
79. Rehme, K. A., B. E. Martin, and J A Hodgeson.
Tentative Method for Calibration of Nitric Oxide,
Nitrogen Dioxide, and Ozone Analyzers by Gas Phase
Titration EPA-R2-73-246, U.S. Environmental
Protection Agency, Research Triangle Park, N.C., March
1974.
80 Research Triangle Institute Investigation of Rural
Oxidant Levels as Related to Urban Hydrocarbon Control
Strategies. EPA-450/3-75-036, U.S. Environmental
Protection Agency, Research Triangle Park, N.C., March
1975.
81. Richardson, R. L. Performance of ambient hydrocarbon
monitors in the Texas continuous monitoring network.
Presented at the 67th Annual Meeting, Air Pollution
Control Association, Denver, Colo., June 9-13, 1974.
82 Saenz, 0., Jr., C A. Boldt, Jr , and D S Tarazi
Measurement of Non-Methane Hydrocarbons APTD-
0905, U S Environmental Protection Agency,
Cincinnati, Ohio, June 1971
83 Saltzman, B E. Colorimetnc microdetermmation of
nitrogen dioxide in the atmosphere Anal. Chem.
26 1949-1955, 1954
84. Saltzman, B E Modified nitrogendioxidereagentforair
analyzers. Anal Chem 32135,1960.
85. Saltzman, B. E., W. A. Cook, B Dimitnades, E.L Kothny,
L. Levin, D. W. McDaniel, and J. H. Smith Tentative
method for continuous monitoring of atmospheric
oxidant with amperometric instruments In Methods of
Air Sampling and Analysis American Public Health
Association, Washington, D C , 1972 pp 341-350.
86 Scarmgelli, F P, E Rosenberg, and K A Rehme.
Comparison of permeation devices and nitrite ion as
standards for the colonmetric determination of nitrogen
dioxide. Environ. Sci Technol 4 924-929, 1970
87 Scott, W. E ,E R Stephens,? L Hanst, andR.C.Doerr
Further developments in the chemistry of the
atmosphere. Proc. Amer Pet Inst Sect. 3 37.171-183,
1957
88. Seila, R L GC-chemilummescence method for the
analysis of ambient terpenes In: International
Conference on Photochemical Oxidant Pollution and Its
Control. Proceedings Vol I B Dimitnades, ed U S
Environmental Protection Agency, Research Triangle
Park, N.C .January 1977 pp 41-50.
89. Severs, R. K. Simultaneous total oxidant and
chemiluminescent ozone measurements in ambient air
J. Air Pollut Control Assoc 25 394-396, 1975
90 Severs, R K., R. M. Neal, and T D Downs Time as a
factor in atmospheric sampling. J Am Ind Hyg Assoc
3S.134-141, 1977
91. Smith, R G., R J. Bryan, M Feldstem, B. Levadie, F A
Miller, and E. R Stephens Tentative method of analysis
for peroxyacetylnitrate (PAN) in the atmosphere. In
Methods of Air Sampling and Analysis American Public
Health Association, Washington, D C , 1972 pp 215-
219
92. Stedman, D H, E E. Darby, F Stuhl, and H. Niki
Analysis of ozone and nitric oxide by a
chemiluminescent method in laboratory and
atmospheric studies of photochemical smog J Air
Pollut. Control Assoc. 22260-263, 1972
93. Stephens, E R,P L Hanst, R C.Doerr, andW E Scott
Reactions of nitrogen dioxide and organic compounds in
air. Ind Eng Chem 48 1498-1504, 1956
94. Stephens, E R , and M A. Price Analysis of an
important air pollutant peroxyacetylnitrate J Chem
Educ 50351-354, 1973
95 Stephens, E R,W. E Scott, P L Hanst, and R C Doerr.
Recent developments in the study of the organic
chemistry of the atmosphere J Air Poll. Control Assoc
6-159-164, 1956
96 Sternberg, J.C ,W S. Gallaway, and D T L Jones The
mechanism of response of flame lonization detectors.
In. Gaschromatography N Brenner,J E Callen.andM.
D Weiss, eds. Academic Press, New York, 1962 pp
231-267
97. Stevens, D. W. Obtaining reliable performance from
hydrocarbon analyzers. In. Instrumentation for
Monitoring Air Quality. ASTM Special Technical
134
-------
Publication 555, American Society for Testing and
Materials, Philadelphia, Pa , June 1974.
98 Stevens, R K,T A Clark, C.E. Decker, and L.F. Ballard.
Field performance characteristics of advanced monitors
for oxides of nitrogen, ozone, sulfur dioxide, carbon
monoxide, methane, and nonmethane hydrocarbons.
Presented at the 65th Annual Meeting, Air Pollution
Control Association, Miami, Florida, June 18-22, 1972.
99 Stevens, R K , and J. A Hodgeson. Applications of
chemiluminescent reactions to the measurement of air
pollutants. Anal Chem 45.443A-447A, 1973
100 Stevens, R. K., J. A. Hodgeson, L F Ballard, and C. E.
Decker Ratio of sulfur dioxide to total gaseous sulfur
compounds and ozone to total oxidants in the Los
Angeles atmosphere—an instrument evaluation study.
In Determination of Air Quality. G.MamantovandW. D.
Shults, eds Plenum Publishing Co., New York, 1972. pp
83-108
101 Tokiwa, Y , S Twiss, E R. de Vera, and P. K. Mueller
Atmospheric ozone determination by amperometry and
colonmetry. In. Determination of Air Quality. G
Mamantov and W D Shults, eds. Plenum Publishing
Co, New York, 1972 pp 109-130.
102 US Environmental Protection Agency Ambient air
monitoring reference and equivalent methods. Fed
Regist 40.7042-7070, February 18, 1975.
103 U.S Environmental Protection Agency. Measurement
of photochemical oxidants in the atmosphere.
Calibration of reference methods Fed Regist.
47-44049-44057, October 6, 1976.
104 US Environmental Protection Agency. National
primary and secondary air quality standards Fed.
Regist. 368186-8201, April 30, 1971.
105 US Environmental Protection Agency. Nitrogen dioxide
measurement principle and calibration procedure Fed.
Regist. 47 52686-52695, December 1, 1976.
106 US Environmental Protection Agency. Reference
method for determination of hydrocarbons corrected for
methane In, Code of Federal Regulations Title 40. Part
50 11 Appendix E. U S. Government Printing Office,
Washington, D.C , July 1, 1975 pp. 21-27.
107 Winer,A.M.,J W.Peters.J P.Smith.andJ N.Pitts.Jr
Response of commercial chemiluminescent NO-NO?
analyzers to other nitrogen-containing compounds.
Environ Sci Technol 81118-1121,1974.
108 Zafonte, L,W Long, and J N Pitts, Jr Evaluation of an
ozone photometer for ambient ozone measurements.
Anal Chem 461872-1874, 1974
135
-------
8. TOXICOLOGICAL APPRAISAL OF
PHOTOCHEMICAL OXIDANTS
INTRODUCTION
Photochemical oxidants such as ozone and
peroxyacylnitrates are gases that exert their toxic
effects by entering the body through inhalation. If
present in sufficient concentrations, they are
capable of causing death to various organisms. At
sublethal concentrations, they may alter, impair, or
otherwise interfere with physiological processes.
Alterations in pulmonary function and in
mechanical properties of the lungs are among the
effects that result from inhalation of these
compounds. Other effects include morphological
changes in the lungs, increased susceptibility to
infectious respiratory disease, and biochemical
alterations in the lungs and in other organs. Since
there is a scientific consensus that concentrations
above 1960 fjg/m3 (1 ppm) ozone are very
hazardous to health, generally only research
conducted at or below this concentration will be
described here.
The toxicological effects of ozone and two major
pollutant groups, oxidants (mixtures of substances
produced by photochemical reactions) and
peroxyacylnitrates, are discussed in this chapter.
Each compound or group is treated separately.
EFFECTS OF OZONE ON EXPERIMENTAL
ANIMALS
Respiratory Tract Transport and Absorption
NASOPHARYNGEAL UPTAKE
The extent to which air pollutants can directly
affect the lung depends largely on the amounts of
pollutant and reaction products that penetrate to
the lower airways. Nasopharyngeal removal of
ozone lessens the insult to the lung and must be
accounted for when estimating ozone con-
centrations reponsible for observed pulmonary
effects. Experimental estimates of nasopharyngeal
ozone removal determine the appropriate
boundary conditions when using convective/
diffusion equations to model ozone transport in the
lower airways.
Vaughn et al.200 exposed the isolated upper
airways of beagle dogs to ozone at a continuous
flow of 3.9 l/min and collected the gas just below
the larynx in a plastic (Mylar) bag. Concentrations
below 784 fjg/m3 (0.4 ppm) yielded essentially
100-percent uptake. However, Yokoyama and
Frank215 observed 60- to 70- percent uptake of
ozone by the upper airways in similar studies on
beagles where Teflon tubing was used. When
these investigators repeated the procedure of
Vaughn et al.,200 they found that some ozone was
lost because of adsorption on the bag wall.
Nasal uptake significantly exceeded (p<0.01)
oral uptake at both flow rates (3.5 and 35 l/min)
used in the Yokoyama and Frank215 studies. The
overall uptake coefficient for the nose increases
with air flow rate. This latter point was shown by
Aharonson et al.,1 who demonstrated the neces-
sity for expressing retention data on the basis of
the average uptake coefficient when comparing
uptake at different flow rates.
The decomposition of ozone in the nasopharynx
of acutely versus chronically exposed beagle dogs
was studied by Moorman et al.139 Dogs chronically
exposed (18 months) to 1960 to 5880A
-------
ppm). If one takes into account the sensitivity of the
Mast ozone meter used to measure the responses,
this latter difference would not represent a
significant difference in treatment unless the Mast
ozone meter had been modified. Moorman et al,139
did not state what modifications, if any, were
performed. Nevertheless, in view of the small
difference observed with exposure to 1960/^g/m3
(1.0 ppm), it is likely that chronic exposure to lower
levels would not result in tracheal ozone con-
centrations significantly greater than those
observed with acute exposure.
Nasopharyngeal removal of ozone in rabbits of
both sexes and in male guinea pigs has been
studied by Miller134 over a concentration range of
196 to3920pg/m3 (0.1 to 2.0 ppm). The tracheal
ozone concentration in these species was
markedly similar and was linearly related to the
chamber concentration of ozone drawn
undirectionally through the isolated upper
airways. Regression analyses showed that ozone
removal in the nasopharyngeal region is
approximately 50 percent in both species. No
difference in ozone removal was observed
between male and female rabbits.
TRACHEOBRONCHIAL UPTAKE
Lower airway removal of ozone from inspired air
was measured by Yokoyama and Frank216 in dogs
that were mechanically ventilated through a
tracheal cannula. Two ranges of ozone con-
centration were studied: (1) 1372 to 1666pg/m3
(0.7 to 0.85 ppm) and (2) 392 to 784 pg/m3 (0,2 to
0.4 ppm). The rate of uptake was found to vary
between 80 and 87 percent when the tidal volume
was kept constant and the respiratory pump was
operated at 20 to 30 cycles/min. This estimate of
uptake applies to the lung as a whole and does not
indicate the uptake of ozone by individual airway
generations.
A major goal of environmental toxicological
studies on animals involves the eventual
extrapolation of the results to man. In addressing
the likelihood that the animals studied mimic the
human response, estimation of the effective dose
delivered to the target organ is required. Also,
differences between man and experimental
animals in the ratio of effective dose to exposure
concentration must be considered. Extensive
lower airway morphometric data on human
beings,206 guinea pigs, rabbits, and rats,111 and the
technical inability of obtaining local lower airway
ozone uptake data make mathematical modeling
the method of choice forexamining ozone uptake
in the deep lung.
The mathematical model of lower airway ozone
uptake by McJilton et al.,'25 although widely cited,
has never been formally published. However, the
major features of the McJilton et al. model have
been described in some detail, and model results
for absorption of ozone in each generation of the
human tracheobronchial tree have been
presented.1SS Since nonreactivity of ozone with the
mucous layer was assumed, the model of McJilton
et al. is really more useful for estimating lower
airway uptake of water soluble and relatively water
insoluble gases that are nonreactive with the
mucous layer than it is for estimating the uptake of
ozone.
The problem of treating chemical reactions of
ozone with various components of mucus has been
included in the mathematical model of transport
and removal of ozone developed by Miller.134These
reactions were shown to be characterized by an
instantaneous reaction regime. Three cases,
which are a function of the lumen ozone
concentration, were used to characterize the
removal of ozone from an airway. As did the model
of McJilton et al.,125 Miller included the effects on
gas transport of convection, axial diffusion, and
radial diffusion, although the distance-time grid
mesh used was 5-fold smaller than that used by
McJilton et al. Also, Miller treated the effects of
both molecular diffusion and eddy diffusivity on
axial diffusion, rather than including only
molecular diffusion.
Miller et al.136 modeled tracheobronchial uptake
of ozone in the lungs of guinea pigs and rabbits, as
well as man. The predicted pulmonary ozone dose
curves obtained indicate that a general similarity
exists between these species in the shape of the
dose curves. Independent of the inhaled tracheal
concentration of ozone, the respiratory bron-
chioles are predicted to receive the maximum dose
of ozone, a result that is in good agreement with
experimental findings in various animal
species.50'190'191
The model predicts uptake of ozone by res-
piratory airway tissue for all tracheal con-
centrations studied (62.5 to 4000 fjg/m3, or 0.03 to
2.04 ppm). However, penetration of ozone to the
tissue in airways lined by mucus is not predicted
for inhaled tracheal concentrations less than 125
fjg/m3 (0.06 ppm) in man and 62.5 fig/m3 (0.03
ppm) in guinea pigs and rabbits. As the inhaled
tracheal concentration is increased from these low
137
-------
levels, the tissue deposition pattern of ozone in the
conducting airways becomes smoother, increases
in magnitude, and includes more airways. Alveolar
ozone doses decline sharply from the maximum
values predicted for the respiratory bronchioles.
As the tracheal ozone concentration decreases
from approximately 100A41,160AKj/m3(21 ppm) are required to reach the
LD5o. Mice treated with selected antioxidants can
be partially protected against mortality from high
dose ozone exposures. 132'159'174>175 Concentration
and length of exposure are notthe only factors that-
determine toxicity. Stokinger cited other factors
that influence the response to a specific level of
ozone.194 These include age, temperature,
exercise, dosage rate, respiratory infection, and
reducing agents.
In addition to these factors, Skillen189 confirmed
the observation of Fairchild et al.63 of a variation in
the effect of continuous ozone exposure to 11,800
/jg/m3 (6 ppm) on rats with varying thyroid status.
The average survival times of these rats were as
follows: More than 10 hr for those rats with
reduced thyroid function, 6.7 hr for rats with
unaltered thyroid activity, and 2.2 hr for rats with
stimulated thyroid function.
Pulmonary Effects
HOST DEFENSE MECHANISMS
The adverse effects of a number of air pollutants,
including ozone, on pulmonary defense
mechanisms against infectious disease have been
studied. This section will describe how low con-
centrations of ozone result in an enhancement of
mortality in animals exposed to bacterial aerosols.
This increased susceptibilty to pneumonia is
ascribed primarily to dysfunction of the primary
defense cells of the lungs, the alveolar
macrophages.
Interaction with Infectious Agents —The effects of
a number of pollutants on a host challenged with
infections microorganisms have been reviewed.367?
Briefly, mice are~exposed either to ozone or
filtered clean air. At various time periods, all of the
animals are challenged with an aerosol of viable
microorganisms, and the accumulated mortality
over a 15-day observation period in clean air is
reported. The concentrations of pollutants used
alone (i.e., without infectious agents) caused no
deaths.
Coffin et al.35 exposed mice to different con-
centrations of ozone ranging from 137 to 981
/Kj/m3 (0.07 to 0.5 ppm). The 3-hr exposures were
followed immediately by bacterial aerosol
exposure (Streptococcuspyogenes, Group C). With
the exception of the lowest ozone concentration,
137 /Kj/m3 (0.07 ppm), each concentration was
tested once (40 mice/concentration). It was
reported that > 157 /yg/m3 (0.08 ppm) ozone
resulted in a significant(p< 0.05) enhancement of
mortality of at least 23 percent. As the
concentration of ozone was increased up to 101 9
/yg/m3 (0.52 ppm), there was a trend toward
increased mortality.
Ehrlich etal.54 conducted studies usingthesame
microorganism as Coffin et al.,35 but with different
strains of mice. Gas concentrations were
monitored by the chemiluminescence method. A
single 3-hr exposure to 196A
-------
mortality of 21 percent was observed. When this
latter experimental regimen was used, exposure to
157 fjg/m3 (0.08 ppm) ozone resulted in a
significant (p < 0.05) increase in mortality of 5.4
percent.
The differences in results among these studies
may have been due to a variation in the sensitivity
of method of ozone monitoring, a difference in
mouse strain, changes in the pathogenicity of the
bacteria, or differences in sample size.
The interactions of other environmental stresses
(cold and other pollutants) using the infectivity
model have also been examined. Using S.
pyogenes as the infectious agent, Coffin and
Blommer34 showed that animals preexposed to
cold (6° to 9°C, or 43° to 48°F) for 3 hr had a
greater mortality after a 3-hr exposure to 1 372 to
1764 /jg/m3 (0.7 to 0.9 ppm) than did animals
exposed to ozone only.
The effects of exposures to mixtures of NO2 and
ozone were examined by Ehrlich et al.54 using the
infectivity model. Significant (p < 0.05) increases
in mortality occurred when mice were exposed to
concentrations at or above 3926/yg/m3 (2 ppm)
NC>2 and 98/ug/m3(0.05 ppm) ozone for 3 hr before
challenges with aerosols of Streptococcus.
Although not statistically tested, the effect of the
exposure to the combination appeared to be
additive (i.e., the increases in mortality were
approximately equivalent to the sum of those
induced by the exposure to each individual
pollutant. In addition, the animals exposed to the
mixture show a reduced rate of bacterial clearance
from the lung.
Gardner et al.80 examined the effects of
sequential exposure to 196/ug/m3(0.1 ppm) ozone
for 3 hr and 0.9 mg sulfuric acid (H2S04)/m3 (0.3
/urn) for 2 hr using CD-1 mice. The bacterial aerosol
was administered at the end of the pollutant
exposure. When the ozone proceeded the HzSCu
exposure, there was a significant increase in
mortality (p < 0.05), which was additive.
The results derived from studies utilizing the
infectivity model indicate its sensitivity for
detecting biological effects at low pollutant
concentrations and its response to modifications in
technique (i.e., using different mouse strains or
varying the time of bacterial challenge). The model
is supported by experimental evidence (discussed
in the following section and reviewed in reference
24) that shows that pollutants, albeit at different
concentrations, that cause an enhancement of
mortality in the infectivity system also cause
reductions in essential host defense systems, such
as pulmonary bactericidal capability, the
functioning of the alveolar macrophage, and the
cytological and biochemical integrity of the
alveolar macrophage.
Coffin and Blommer34 examined the mechanism
of action of the infection following pollutant
exposure. In these studies, the number of bacteria
in the lung and the presence of bacteria in the
blood was followed at various time periods. Mice
were exposed to 1 960/;g/m3(1 ppm) ozone for 3 hr
before a bacterial aerosol. Bacterial invasion of the
blood began 2 days after ozone exposure.
Additional research90 has shown that exposure to
ozone concentrations greater than 780/Lig/m3(0.4
ppm) results in lower deposition rates of inhaled
Streptococcus. Nevertheless, the bacteria that are
deposited multiply, so that shortly after exposure,
there are more bacteria in the ozone-exposed
lungs than in the lungs of control animals. Again, it
was hypothesized that this oxidant gas adversely
affected pulmonary defense systems other than
mucociliary clearance.
Goldstein et al.73'89 exposed mice to an aerosol of
32P-labeled Staphylococcus aureus either after a
17-hr exposure to ozone or before a 4-hr exposure.
Concentrations of ozone were 1180, 1370, 1570,
or 1960 fjg/m3 (0.6, 0.7, 0.8, or 1 ppm), or higher.
The mechanical clearance and bactericidal
capabilities of the lung were then measured 4 to 5
hr after bacterial" exposure. Exposure for 17 hr
before infection caused a significant (p < 0.05)
reduction in bactericidal activity beginning at 1 960
/jg/m3 (1 ppm). When mice were exposed to ozone
for 4 hr after being infected, there was a significant
decrease in bactericidal activity for each ozone
exposure level. The lowest level of measurable
effect was 1180 fjg/m3 (0.6 ppm). With increasing
ozone concentration, there was a progressive
decrease in bactericidal activity. The investigators
proposed that because mucociliary clearance was
unaffected by subsequent ozone exposure, the
bactericidal effect was due to dysfunction of
another type of pulmonary defense, most probably
the alveolar macrophage. Other workers203 found
that exposure to980/Lig/m3(0.5 ppm) of ozone, 16
hr/day for 7 months had no effect on the
mechanical clearance of polystyrene and iron
particles from rabbit lungs. In the guinea pigs
exposed to the same concentration for 1 6 hr/day
for 2 months, there was only a small reduction in
bacterial clearance.
Warshauer et al.203 demonstrated that a
139
-------
deficiency of vitamin E augmented the adverse
effect of ozone at 1374 //g/m3 (0.7 ppm) on murine
pulmonary bactericidal capacity only after a
prolonged exposure (7 days). Identical
hypovitaminosis failed to influence pulmonary
bactericidal activity in rats exposed for only 4 hr to
980 to 1960 //g/m3 (0.5 to 1.0 ppm) of ozone,
The effect of pollutant combinations on the
bactericidal function of the lung has also been
examined.84 Some mice received an intratracheal
injection of 10 mg of silica, and others received
latex. Seventy days later, the animals were
infected with 32P-labeled Staphylococcus aureus
and exposed to either 4 hr of 785 fjg/m3 (0.4 ppm)
ozone or to clean air. Both the silicotic and latex-
injected (control) mice had similarly reduced levels
(12 percent) of bactericidal activity after breathing
ozone (p < 0.05), when compared to their ozone-
exposed counterparts. Silicosis itself did not inhibit
the pulmonary bactericidal response. Goldstein et
al.91 conducted a similar study in which mice were
exposed to ozone/NOa combinations: From 196 to
790 //g/m3 (0.1 to 0.4 ppm) ozone, and from 3760
to 13,720 //g/m3 (2 to 7.3 ppm) NO2 either for 17 hr
before infection or for 4 hr after infection. The
lungs were capable of functioning well until the
higher pollutant level was reached. Physical
clearance was not affected at either level. The
demonstrated injuries were those that would be
expected from each individual oxidant.
Fairchild62 found that mice exposed for 1 hr to
1770 //g/m3 (0.9 ppm) ozone exhibited a 70-
percent increase in respiratory deposition of
vesicular stomatitis virus (p < 0.005) occurring in
the nasal cavity rather than in the lung. Also, the
minute ventilation was reduced (p < 0.05) by a
maximum of 30 percent at the end of the exposure.
Fifteen minutes later, the minute volume
increased but remained lower than that of the
controls.
An ozone concentration of 1177 //g/m3 (0.6
ppm) for 3 hr caused an inhibition of replication of
influenza virus (A2/Japan, 305/57) deposited in
the nasal cavities of mice. In contrast, a slight
increase in growth of vesicular stomatitis virus
was seen in the noses of mice exposed to 1767
//g/m3 (0.9 ppm) ozone for 3 hr.61
Laboratory-induced parasitic infection (Plas-
modium berghei) of mice was also exacerbated by
exposure to 1686 //g/m3 (0.86 ppm) ozone for 8
hr/day, 5 days/week for 6 months.182
Alveolar Macrophages — In most of the foregoing
studies, the authors recognized the possibility that
damage to the alveolar macrophages (AM) was
responsible primarily for the enhanced infectivity
and reduced bactericidal response. Recent
investigations have furthec advanced this
hypothesis. When rabbits were exposed to 980
//g/m3 (0.5 ppm) or 1313 //g/m3 (0.67 ppm) for 3 hr,
there was a significant reduction (p< 0.001) in the
number of bacteria phagocytized by AM.36'38
Because phagocytosis begins with particle
attachment to the cell membrane, increased
fragility of the AM could play a role. Rabbits
exposed to 980 //g/m3 (0.5 ppm) for 8 hr/day for 7
days exhibited a trend toward increased
membrane fragility of AM, although lipid
peroxidation was undetectable at this level.48
The AM within the lung reside within a liquid
layer. Gardner76 demonstrated that the protective
components of this layer are inactivated by a 2.5-hr
exposure to 196 //g/m3 (0.1 ppm) ozone. When
normal or ozone-exposed AM were placed in fluid
lavaged from exposed rabbits, they showed more
lysis (10 percent over controls). A similar effect
was seen when normal AM were placed in the
protective fluid that had been exposed in vitro to
ozone at 196 //g/m3 (0.1 ppm) or 1960//g/m3 (1
ppm) for 30 min.
Goldstein et al.85 studied the effect of a 2-hr
ozone exposure on the ability of AM to be
agglutinated by concanavalin A, a parameter
reflecting membrane organization. A decrease in
agglutination of rat AM was found after exposure
to 980 or 1960//g/m3 (0.5 or 1.0 ppm). A decrease
in concanavalin A agglutinability of trypsinized red
blood cells obtained from rats exposed for 2 hr to
1960 //g/m3 (1 ppm) was also noted. Hadley et al.94
investigated AM membrane receptors from rabbits
exposed to 980 //g/m3 (0.5 ppm) ozone for 3 hr.
Following ozone exposure, lectin treated AM have
increased rosette formation (p < 0.05) with rabbit
red blood cells. The authors hypothesized that the
ozone-induced response indicates alterations of
macrophage membrane receptors for the wheat
germ agglutinin that may lead to changes in the
recognitive ability of the cell.
Hurst et al.106 showed that the acid hydrolases of
AM lysosomes are significantly reduced after
rabbits are exposed for 3 hr to either 490 //g/m3
(0.25 ppm) or 980 //g/m3 (0.5 ppm) of ozone.
Greatest reductions (25 percent and 40 percent,
respectively) were seen in the lysozyme activity.
Acid phosphatase and /3-glucuronidase activities
were also affected. Since these enzymes are
involved in intracellular degradation of bacteria,
140
-------
their reduced activities could contribute to poor
macrophage functioning and, consequently, to
enhanced mortality from bacterial invasion.
Similar enzymatic reductions were observed in AM
exposed in tissue culture.105 The effects of ozone
on host defenses were also studied by means of
unilateral lung exposures of rabbits.3 Three hours
of ozone exposure was found to decrease cellular
viability, depress various intracellular enzymes,
and increase the number of pulmonary
polymorphonuclear leukocytes. The effects were
dose-related, beginning at 980 /Kj/m3 (0.5 ppm).
The responses were found to be specific to the lung
that breathed ozone rather than any generalized
systemic response.
In vitro techniques have also been employed to
study AM function and biochemistry after ozone
exposure. Weissbecker et al.207 established dose-
response curves for the effect of ozone on the
viability of AM, with effects being seen at 11 8
A
-------
Amino Acids and Proteins — The susceptibility of
aqueous solutions of ammo acids to oxidation in
vitro from ozone exposure was ranked by Mudd et
al.140 From high to low susceptibility, the ranking is
as follows: Cysteine, methionine, tryptophan,
tyrosine, histidine, cystine, and phenylalanine.
Other amino acids were unaffected. Sulfhydryl
compounds were found to the be most susceptible
to oxidation. In addition, when avidm was exposed
to ozone, its ability to bind to biotm was lost,
presumably because of oxidation of tryptophan
residues. In vitro ozone exposure 169 to formic acid
solutions of ammo acids was found to have the
following decreasing order of reactivity:
Tryptophan, methionine, cystine, tyrosine. Other
amino acids were not tested.
Protein synthesis of rat lung tissue is altered by
continuous exposure to 1570 fig/m3 (0.8 ppm)
ozone, as described by Mustafa et al.149 After 1 day
of exposure, there was no difference in the in vitro
incorporation of labeled amino acids into the lung
slices. However, 35- and 84-percent increases in
incorporation were observed after 2 and 3 days of
exposure, respectively. Throughout the remainder
of the study (up to 7 days), the level remained
elevated and unchanged. When in vivo amino acid
incorporation procedures were used, there was a
50-percent decrease in incorporation after 1 day of
exposure. On days 2 and 3, there were increases of
60 percent and 100 percent over control. A plateau
was reached on day 3 that extended to day 7. Under
the in vivo conditions, no radioactive incorporation
of amino acids occurred in the AM obtainable by
lavage. Similar studies conducted in the presence
of puromycin (an inhibitor of protein synthesis)
showed that the increase in ammo acid
incorporation observed in vitro and in vivo for the
ozone-exposed animals was reflected in an
increase in protein synthesis.
The influence of ozone on lung prolyl
hydroxylase (an enzyme thoughtto be rate-limiting
in collagen synthesis) and the product of its
reaction, hydroxyproline, was studied by Hussain
et al.108 R?*-; were exposed continuously to either
390, 980, or 1570 A
-------
leading to structural changes in lung tissue.
Last et al.118 studied in vitro mucus glycoprotein
secretion by tracheal explants from rats exposed to
390 or 1 570 /jg/m3 (0.2 or 0.8 ppm) ozone for 8 or
24 hr/day for periods of 2 to 90 days. Significant
ozone-concentration-related decreases in gly-
coprotein secretion were observed. The depression
of secretion at days 2 to 3 could be reversed by the
administration of indomethacin or hydrocortisone
during exposure.
Lipids — Many investigators postulate that lipid
peroxidation is an important toxic mechanism in
ozone injury. These ideas are presented in
reviews39i129that describe the propensity of ozone to
react with the ethylene groups of unsaturated fatty
acids. This reaction can lead to free radical
formation, which in the presence of molecular
oxygen leads to the peroxidation of the
unsaturated fatty acid by formation of fatty acid
lipoperoxides. Alterations found in erythrocytes
following ozone exposure are often cited as
evidence for these concepts. 82'86'87-132 (For a more
complete discussion of effects on erythrocytes, see
Chapter 9.)
Buell et al.22 suggested that the interaction of
ozone and water results in the fomation of atomic-
or radical-oxygen and molecular oxygen, whereas
Alder and Hill2 postulated the formation of a variety
of radicals. Although high-energy sources such as
X-irradiation undoubtedly decompose water into a
variety of free radicals, it has not been shown that
ozone can form more than one.
That free radicals may be involved in ozone
toxicity is suggested indirectly by the work of
Mustafa et al.151 The influence of ozone on
superoxide dismutase (SOD) activity of the lung
was investigated. SOD is an inducible enzyme that
catalyzes the dismutation of superoxide radical,
thereby preventing oxidant toxicity that may be
consequent to this radical. Rats were exposed to
390, 980, or 1570 /jg/m3 (0.2, 0.5, or 0.8 ppm)
ozone for 7 days. While it is not stated explicitly, it
can be assumed from the abstract that exposure
was continuous. SOD activity increased by 17
percent, 26 percent, and 38 percent after exposure
to 390, 980, and 1570/L/g/m3(0.2,0.5,or0.8ppm),
respectively. At the highest ozone concentration,
the location of the SOD activity was determined,
and it was found that exposure resulted in a 38-
percent increase in the cytosolic fraction and a 46-
percent increase in the mitochondrial fraction.
Using a different exposure mode, rats were
exposed stepwise to 1 570 pg/m3 (0.8 ppm) for 72
hr, to 2940 /jg/m3 (1.5 ppm) for 24 hr, and finally to
5890 /jQ/m3 (3 ppm) for 8 hr. Following treatment,
SOD activity increased (as ozone concentration
increased) by 15 percent, 25 percent, and 50
percent in the cytosol and 50 percent, 67 percent,
and 118 percent in the mitochondria, respectively.
The authors suggest that the increase in SOD
activity represents adaptive changes in the lung
that might reduce oxidant toxicity.
Pryor et al.171 and Pryor170 exposed pure samples
of polyunsaturated fatty acids (methyl linoleate
and methyl linolenate) toO to 2940/yg/m3(0to 1.5
ppm ozone) in vitro and followed the formation of
peroxides, conjugated dienes, and thiobarbituric-
acid-reactive material as a function of time. Ozone
shortened the induction period for auto-oxidation,
but did not significantly affect the rate of product
formation after the induction period. As the
concentration of ozone was raised from 0 to 2940
/jg/m3 (0 to 1 5 ppm) during the induction period,
there were greatly increased rates of peroxide
formation, a slightly increased rate of conjugated
diene formation, and no significant change in the
rate of production of thiobarbituric-acid-reactive
material. When vitamin E was added, the induction
period was lengthened, but the rate of product
formation during the auto-oxidation phase was
unchanged. The authors hypothesized that most of
the ozone-induced in vivo alterations occur during
the induction period of lipid auto-oxidation.
Roehm et al.
174,176
studied the oxidation of
polyunsaturated lipids in vitro by exposing samples
of pure methyl esters of fatty acids to 2940 /jg/m3
(1.5 ppm) ozone and 2820/L/g/m3(1.5 ppm)NO2. At
equivalent concentrations of these two oxidants,
ozone produced the greatest changes. Methyl
linoleate was completely reacted by ozone to yield
stable ozonides. These researchers hypothesized
that ozone acts by direct addition of ozone across
the fatty acid double bond, and this agrees with the
mechanism proposed by Criegee41 concerning the
primary mechanism for ozonolysis.
Other investigators have studied the effects of
ozonides in vitro and in vivo. Menzel et al.133
injected intradermally 10 picograms to 10 /jg of
fatty acid ozonides (from oleic, linoleic, linolenic,
and arachidonic acids) into animals and found
increased vascular permeability of the capillaries,
as measured by extravascularly located,
pontamine-blue-bound serum proteins. This
reaction was blocked by simultaneous anti-
histamme injection or by prior treatment with
compound 48/80 (a substance that can deplete
143
-------
histamine stores). It therefore appeared that
histamine was involved in the toxicity of the
ozonides (but histamine release could be
secondary and not primary). The same series of
experiments also showed that the ozonide caused
peroxidation of isolated microsomes and
mitochondria. There were also indications of
oxidation of erythrocyte membranes. Cortesi and
Privett40 also prepared a methyl linoleate ozonide
and administered it to rats by intravenous injection
or by oral dosage. This treatment was lethal as a
result of acute lung edema. Also, significant
changes occurred in the fatty acid composition of
serum and lung lipids.
To investigate some of the possible effects of
ozone on membrane lipids, Teige et al.196 exposed
phosphatidyl choline liposornes containing
trapped glucose and human erythrocytes to ozone
in vitro (3 //moles ozone/min), Phosphatidyl
choline is a phospholipid commonly found in cell
membranes. Ozone exposure caused a leakage of
glucose from the tiposomes Liposornes exposed to
ozone were more active than ozone alone in
causing lysis of red blood cells. Hydrogen peroxide
was found not to be responsible for this effect.
Chow and Tappel32 observed an increased
formation of malonaldehyde (a product of lipid
peroxidation) in the lungs of rats following
continuous exposure to 1370 yg/m3 (0.7 ppm) for
5 days or 1570 /jg/m3 (0.8 ppm) for 7 days. The
formation of malonaldehyde, along with the ozqne-
mduced increased activities of glutathione
peroxidase and glucose-6-phosphate dehy-
drogenase (G-6-PD), was partially inhibited as a
logarithmic function of dietary cr-tocopherol
(vitamin E) from 0 to 1 500 ^ug/kg. Mustafa et al.148
found no lipid peroxidation (as determined by assay
for thiobarbiturate-reactmg materials and con-
jugated dienes) in the lung homogenate of rats
exposed to 3920 //g/m3 (2 ppm) ozone for 8 hr,
Roehm et al.174'17S continuously exposed rats to
I960 fjg/m3 (1.0 ppm) ozone, which eventually
resulted in 100-percent mortality due to
pulmonary edema. Fifty percent of the a-
tocopherol-depleted group died (LT5o) after 8 2
days of continuous exposure, while the vitamin-
supplemented rats exhibited an LTso of 18.5 days.
When Fletcher and Tappel68 studied the
continuous effects on rats of ozone from exposure
to 1 570 jug/m3 (0.8 ppm) or higher concentrations
for 7 days, they too found that a-tocopherol
protected the animals from severe toxicity
(mortality) and weight change. At low levels, the
vitamin E protection against lung lipid peroxidation
was a reciprocal function of the logarithm of
dietary a-tocopherol. Rats maintained on vitamin-
E-supplemented diets and exposed to 1560 fjg/m3
(0.8 ppm) ozone continuously for 7 days were also
partially protected from alteration in 6-
phosphogluconate dehydrogenase (6-P-GD), 6-
phosphogluconate (6-PG), and malic enzyme,
Later investigations by Roehm et al.17S showed that
rats continuously exposed to 1960/ug/m3 (1 ppm)
ozone for 9 days showed a significantly shorter
LTso and greater symptoms of respiratory distress
when they were depleted of vitamin E. After 6
weeks of exposure to 980 (ig/m3 (0.5 ppm) ozone,
pulmonary edema and higher mortalities were
noted in animals with diets depleted in a-
tocopherol. Also, their lung weights were different
from those of the a-tocopherot-supplemented
groups. After 6 weeks of ozone treatment,
significant changes were seen in the fatty acid
composition of total lung lipid. The most extensive
change was an increase in arachidonic acid in both
diet groups (p < 0.01), but the increase was higher
in the vitamm-E-depleted rats (p < 0,05), This
effect occurred late in the exposure, and it was
associated with the onset of pulmonary edema and
death. Unlike the lung tissue lipids, the fatty acids
from endobronchial saline lavage were little
affected by ozone or vitamin E. In both exposures,
the onset of the edema was associated with
elevated levels of arachidonic acid and
docosahexenoic acid (22:6) (p < 0.05), These two
compounds are important constituents of cell
membranes, so membranes could be the site of
ozone damage.
Similar results were obtained by Menzel et al.132
who exposed rats continuously to ozone. After 3
days of exposure to 2078 ^/g/m3 (1.6 ppm), linoleic
and linotenic acid concentrations of lavage lipids
were reduced and remained reduced over 17 days
of exposure. Lung tissue lipids were also altered in
rats exposed for 9 days to 1960 fjg/m3 (1 ppm) and
maintained on diets either supplemented with or
deficient in vitamin E. The greatest change was an
increase in arachidonic acid that occurred to a
greater extent in the animals with vitamin E
deficiency. There were also decreases in linolenic
acid in both diet groups; however, supplemental
vitamin E partially retarded this change. Oleic,
stearic, and palmitic acids also were decreased,
but only in the vitamin-E-deficient group, Similar
measurements were made on rats exposed
continuously for 6 weeks to 980 //g/m3 (0.5 ppm)
144
-------
ozone. Under these conditions, the greatest
changes after ozone exposure were a decrease in
oleic acid (particularly in the vitamin-E-deficient
group), a decrease in linolenic acid in the vitamin-
E-deficient rats, an increase in palmitic acid in the
vitamin-E-deficient group, and an increase in the
arachidonic acid, particularly in the vitamin-E-
deficient animals. Sixty-five days of exposure to
980 fjg/m3 ozone also caused a slight decrease
(which appears not to be statistically significant) in
lung tissue sulfhydryl content. From day 1 8 to 65 of
a continuous exposure to an unspecified ozone
concentration (presumably 980/vg/m3, or0.5 ppm)
there was a decrease in serum-reduced gluta-
thione in both diet groups.
Menzel et al.131 also showed that a short
exposure of perfused rat lungs to 5880 /vg/m3 (3
ppm) decreased the enzymatic conversion of
arachidonic acid to prostaglandins in such a way
as to suggest an uncompetitive inhibition of
prostaglandin synthetase by ozone. If this
response occurs in vivo at lower concentrations of
ozone, it would be possible that this could be the
mechanism responsible for the ozone-induced
increase in arachidonic acid seen in the previously
discussed studies 132'175
Kyei-Aboagye et al.117 also studied pulmonary
lipids. Rabbits ex josed to 1960 /vg/m3 (1 ppm)
ozone for 4 hr sh wed a reduced incorporation of
3H-oleate into leci hin (p = 0.02). Pulmonary lavage
from these animo s revealed increased activity of
radiolabeled lecithins. The authors observed that
ozone may affect the lung by decreasing lecithin
formation while simultaneously stimulating the
release of surfac ant lecithins. The formation of
pulmonary phospholipid was investigated by Seto
et al.186 When rabbits were exposed to 980 and
1960 fjg/m3 (0.5 and 1 ppm) ozone for 1 to 2
weeks, the amount of 14C taken into sphing-
omyelin, lecithin, phosphatidyl inositol plus
phosphatidyl serine, and phosphatidyl ethanol-
amine was significantly reduced. The same
researchers studied these phenomena further.187
Under continuous exposure to 980 /vg/m3 (0.5
ppm) ozone for up to 2 weeks, rabbit lu ngs showed
decreased incorporation of 14C into lecithin until 3
hr post exposure, when this parameter returned to
normal. The reduction in phospholipid formation
was more pronounced after 2 weeks of exposure
rather than 1 week of exposure. An exposure to
1960 /vg/m3 (1 ppm) for 1 hr resulted in a
decreased synthesis of sphingomyelin, lecithin,
'phosphatidyl inositol, phosphatidyl serine, and
phosphatidyl ethanolamine.
Even though the lipid composition of pulmonary
wash fluid can be altered by ozone exposure, it
would appear that the surface-tension-lowering
properties of this fluid (which influence pulmonary
function) are relatively unaltered. Gardner et al.81
and Huber et al.103 reported that exposure of
rabbits to 19,600 /vg/m3 (1 0 ppm) and 9800/vg/m3
(5 ppm), respectively, did not destroy the surface
activity of pulmonary wash fluid. However,
Mendenhall and Stokinger127 exposed saline
washings obtained from the lungs of mice exposed
to 9800 to 15,700 /vg/m3 (5 to 8 ppm) ozone and
noted rapid increases in the film pressure (thev
force opposing surface tensions). The authors
suggested that if analogous changes were tooccur
in vivo, the consequence would be an increase in
the distensibility of the lungs, a situation proposed
to be conducive to the development of emphysema.
This effect was not confirmed in another study,8 in
which saline washings from the lungs of dogs
were exposed in vitro to similar concentrations of
ozone.
Sulfhydryl Compounds and Pyndine Nucleotides
— A number of intracellular compounds are active
in cellular redox reactions and as such can
constitute an antioxidant defense system. Many of
the effects of ozone on this system have been
reviewed. 136'149 Much of the research with ozone
has centered on the reduced pyridine nucleo-
tides, namely, reduced nicotinamide adenine
dinucleotide (NADH) and reduced nicotinamide
adenine dinucleotide phosphate (NADPH), and on
sulfhydryl compounds, particularly reduced
glutathione (GSH) and related enzymes.128 The
relationship of these compounds to antioxidant
activity has been described as follows:
<-G-6-P y NADPVy-^QSH y [0]
G-6-PD GSH GSH
HMP shunt reductase peroxidase
6-PG
>NADPH-
+
GSSG
H20
(G-6-P = glucose-6-phosphate; 6-PG = 6-
phosphogluconate; G-6-PD = glucose-6-
phosphate dehydrogenase; HMP shunt - hexose
monophosphate shunt; NADP+ = nicotinamide
adenine dinucleotide phosphate; NADPH =
reduced NADP; GSH = reduced glutathione; GSSG
= oxidized glutathione; [0] = oxidizing moiety [i.e.,
hydrogen peroxide, free radical, lipid peroxide];
145
-------
GSH peroxidase = glutathione peroxidase; and
GSH reductase = glutathione reductase.)
A number of researchers128'141'153 have shown
that ozonization of aqueous solutions of NADH and
NADPH results in their oxidation, Menzel128 found
that after oxidation (0.032 to 0.057 fjmole Oa/min
x 3 hr), NADH and NADPH were still biologically
active as coenzymes, However, Nasr et al.153
reported that ozone (33 ppm, 1000 ml/min, 50
min) destroyed NADPH but had no effect on
NADP+. In vitro exposure141 to 0.1 to 6 /umoles
ozone oxidized NADH, but it had no effect on
nicotinamide adenine dmucleotide (NAD). Similar
results were not found in vivo153 when rats were
exposed to 64,780 /ug/m3 (33 ppm) ozonefor 1 hr.
The NADPH/NADP+ ratio of the tracheal
epithelium was not significantly altered.
In vitro ozone exposure128 (0.032 to 0.057 fjmole
O3/min x 3 hr) was also able to oxidize aqueous
solutions of a number of biologically active
reducing substances (cysteine, GSH, and
thioglycolic acid). This treatment inactivated
papain and glyceraldehyde 3-phosphate dehy-
drogenase and oxidized GSH and GSSG to non-
biologically reducible forms.
The role of sulfur-containing compounds in
ozone toxicity is partially illustrated by the work of
Fairchild et al.64 Inhalation of a variety of sulfur
compounds (1-hexamethiol, methanethiol,
dimethyl disulfide, di-tert-butyl disulfide, benzene
thiol, and hydrogen sulfide) partially protected
mice from the lethal effects of a 4-hr exposure to
8000 to 12,000/ug/m3 (4.1 to 6.1 ppm) ozone. The
functional unit of the sulfur units appeared to be
-SHor SS-,or both, but not-S-r as dimethyl-sulfide
and thiophene were not effective.64
The content of nonprotein sulfhydryls (NPSH) (a
major source of cellular reducing substances),
GSH, and GSSG of lung homogenate from ozone-
exposed rats was determined by DeLucia et al,45
The NPSH level was not significantly affected by
exposure to 1570 /ug/m3 (0.8 ppm) for up to 24 hr,
to 2940 /ug/m3 (1.5 ppm) for up to 8 hr, or to 3920
/ug/m3 (2 ppm) for up to 4 hr. When the animals
were exposed to this higher concentration for 6
and 8 hr, however, there was a decrease (p<0.01)
in NPSH. GSH, which represents about 90 percent
of the lung NPSH, was significantly reduced at
7850 /ug/m3 (4 ppm) ozone for 6 hr, and GSSG
remained unchanged under these conditions.
Further experimentation suggested that mixed
disulfides were transiently formed that reacted
with other sulfhydryl groups of lung tissue.
Analysis also indicated that GSH appeared to be
the only NPSH bound to pulmonary tissue proteins
via a mixed disulfide linkage. In an earlier study,
DeLucia et al.44 found no changes in sulfhydryl
levels in lung homogenates of rats continuously
exposed (10 days) to 1 570 /ug/m3 (0.8 ppm) ozone.
However, there was a 20-percent decrease (p <
0.05) in cytochrome C reductase activity and a 32-
percent increase (p < 0.05) in G-6-PD activity.
The effects of dietary cr-tocopherol on enzyme
systems and lipid peroxidation within the lungs of
rats exposed continuously to 1370 /ug/m3 (0.7
ppm) for 5 days or 1570 /ug/m3 (0.8 ppm)for 7 days
were explored by Chow and Tappel.32 The
increased(p<0.05)formation of malonaldehydefa
product of lipid peroxidation) and the activities of
GSH peroxidase and G-6-PD (see preceding
section on lipids) were partially inhibited as a
logarithmic function of dietary cr-tocopherol (from
Oto 1500mg/kg). The increased (p< 0.05) activity
of GSH reductase was not affected by the level of
dietary vitamin E. In addition, the activity of GSH
peroxidase and concentration of malonaldehyde
were linearly related (p<0.001). The authors
suggest that this represents an apparent
compensation mechanism in that, with increased
levels of lipid peroxides, there is a corresponding
increase in the activity of GSH peroxidase, which
in turn increases lipid peroxide catabolism.
The effect of acute (4-hr) and chronic (4-hr/day,
up to 30 days) exposure of mice to ozone on lung
GSH and vitamin C content was examined by
Fukase et al.74'75 Immediately following the acute
exposure to 16,100 /ug/m3 (8.2 ppm) ozone, there
was a decrease (p
-------
Chow and Tappel31 studied the effects of a 7-day
continuous in vivo exposure to 1570 Aig/m3 (0.8
ppm) ozone on hexose monophosphate shunt and
glycolytic enzymes of lung homogenates of rats
maintained with and without supplemental dietary
or-tocopherol. The vitamin E had no effect on the
enzyme activities in air-control animals. Ozone
caused increases (p < 0.01) in G-6-PD and 6-
phosphogluconate dehydrogenase (6-P-GD)
activities in the lungs of animals from both diet
groups, although the increase was less (p < 0.05)
in those rats that received supplemental vitamin E.
Malic enzyme activity was also increased following
ozone exposure in the normal (p < 0.05) and the
supplemented (p < 0.1) diet groups, but the
increase was less (p < 0.05) in the supplemented
vitamin E group. Neither ozone nor or-tocopherol
supplementation altered the activity of soluble
malic dehydrogenase. The enzymes that regulate
glycolysis—phosphofructokinase and pyruvate
kinase—had increased activities following ozone
exposure and were not affected by vitamin E in the
levels administered. Ozone exposure resulted in
increases in aldolase activity (p< 0.1) in the
dietary-supplemented animals, but not in the
normal diet group (p < 0.1) compared to air
controls. The differences between the supple-
mented and normal diet groups receiving
ozone were not signif leant.Lactate dehydrogenase
(LDH) activity_increased in both diet groups fol-
lowing ozone exposure (p< 0.001), although the
increase wasT less (p<0.1) in the supplemental
diet group.
To determine the influence of the duration of
ozone exposure on some of these ozone-induced
responses, rats maintained on normal diets were
exposed continuously to 1470 Aig/m3 (0.75 ppm)
ozone.31 The activities of GSH peroxidase, GSH
reductase, G-6-PD, 6-P-GD, and pyruvate kinase
were decreased after 1 day of exposure.
Thereafter, they increased until day 10. By day 30,
all enzyme activities except for GSH peroxidase
and GSH reductase had decreased from their 10-
day values but were still increased over control
values. GSH reductase exhibited a very slight
decrease between days 10 and 30. GSH
peroxidase continued to increase over this time
period. The authors hypothesized that an
increased rate of glycolysis and hexose
monophosphate shunt activity is an adaptive
response to the increased need for reductive
detoxification of ozone-induced lipid peroxidation.
Dungworth etal.50 exposed monkeys to ozone for
8 hr/day for 7 days. Six rhesus monkeys received
1570 Aig/m3 (0.8 ppm), and six were exposed to
980 fjQ/m3 (0.5 ppm). Bonnet monkeys (2 to 4 per
exposure group) were exposed to either 980 /yg/m3
(0.5 ppm), 690 fjg/m3 (0.35 ppm), or 390 /ug/m3
(0.2 ppm). In the lungs of ozone-exposed monkeys,
there were increased activities of GSH peroxidase,
GSH reductase, G-6-PD, NADPH-cytochrome-C
reductase, succinate oxidase, acid phosphatase,
and /3-N-acetyl-glucosaminidase. Linear re-
gression analysis indicated that there was a
significant (p < 0.05) correlation between ozone
concentration and increase in enzyme activity in
both rhesus and Bonnet monkeys. Morphological
investigations of the same animals (see later
section on morphology studies) showed the
presence of a lesion at the level of the respiratory
bronchiole and increased numbers of Type 2 cells.
Chow et al.28 exposed rats to 390, 980, or 1 570
/yg/m3 (0.2, 0.5, or 0.8 ppm) ozone continuously (8
days) or intermittently (8 hr/day * 7 days) before
biochemical assays of lung homogenates. For the
continuous exposure to the two higher ozone
concentrations, activities of GSH peroxidase, GSH
reductase, and G-6-PD were increased (p < 0.05)
jver controls. At the lower ozone concentration,
there was an increase (p< 0.05) in the activities of
GSH peroxidase and GSH reductase. Linear
regression analysis indicated that there was a
linear increase in all 3 enzyme activities as the
concentration of ozone was increased (p < 0.001).
Similar results were obtained for the intermittent
exposure groups. No statistical comparisons of the
intermittent and continuous exposure groups were
made. The authors suggest that the increase in
enzyme activity is in response to the need for
reduction of ozone-induced lipid peroxides.
Similar measurements were made by Mustafa
and Lee150 of the lungs of rats and monkeys
exposed to various concentrations of ozone. Lung
homogenates of rats exposed continuously for 7
days to 1 570 Aig/m3 (0.8 ppm) were examined. On
the first day, the activities of microsomal NADPH-
cytochrome-C reductase and cytosolic G-6-PD
activities were slightly depressed, but by day 2 they
were elevated, reaching a plateau at days 3 to 4. By
day 7, NADPH-cytochrome-C reductase had
increasedby42percent(p<0.001), andG-6-PD by
67 percent (p < 0.001). Similar increases were
found in 02 consumption and mitochondrial-
succinate-cytochrome-C reductase (see the
following section on mitochondnal enzyme
activities).
147
-------
The influences of intermittent (8 hr/day) and
continuous exposure for 7 days were compared as
part of the same investigation by Mustafa and Lee.150
Exposures were to 390, 980, or 1570 pg/m3 (0.2,
0.5, or 0.8 ppm) ozone. Specific data were not
given for NADPH-cytochrome-C reductase or G-6-
PD; however, the authors stated that these enzyme
activities increased in a manner similar to
succinate oxidase activity (see the following
section on mitochondrial enzyme activities). Thus
it would seem that rats exhibited the greatest
significant increase, with effects being observed at
390 A
-------
challenge with 7644 /vg/m3 <3.9 ppm) ozone.
Allowing an 11 -day recovery period between the
two exposures resulted in findings similar to those
for the 8-day recovery. The authors suggest that
the tolerance observed might render the animals
more resistant to peroxidative, ozone-induced
damage.
Mitochondria! Enzyme Activities — Mustafa147
demonstrated alterations in lung mitochondria!
oxygen consumption in lung homogenates of rats
exposed to 5880 /vg/m3 (3 ppm) ozone for 4 hr and
then placed in air for several days. Oxidation of
succinate was decreased (p < 0.05) immediately
after exposure and remained so during the next 12
hr. The rate of oxidation reached control values by
24 hr and then increased and peaked (p < 0.001),
relative to control, between 72 and 96 hr. Oxygen
consumption then began to return to normal
values and was not significantly different from
control by day 21. These observations made with
succinate are similar to those made using 2-
oxyglutarate, glycerol-1-phosphate, and as-
corbate-Wurster's blue as substrates, except that
the magnitude of the changes was different.
Succmate-cytochrome-C reductase activity was
greater in the exposed lungs after recovery for 2 to
4 days. When the mitochondrial fraction of the
lungs was examined, generally similar changes in
Qz consumption were observed. The increase in
oxidation was found up to day 7 of recovery, but not
after days 14 and 21 of recovery. To elucidate the
reasons for the observed alterations in oxidative
activity, other experiments were performed. The
cytochrome content of mitochondria (on a per-
milligram-of-protein basis) was not significantly
different in the exposed animals when examined
after 2, 3, and 4 days of recovery. However, the
mitochondrial population of the exposed animals
increased and was significantly higher (p<0,01)
after 48 hr of recovery. After 7 days, the population
began to decline, Concurrent transmission
electron microscopic observations indicated that
there appeared to be no increase in number of
mitochondria within the cells of the exposed
animals' lungs, but the number of Type 2 cells,
which contain numerous mitochondria, had
increased by 50 to 100 percent after ozone
exposure, and succinate-Oa consumption of these
cells was unchanged (on a per-milligram-of-
protein basis).
In contrast to the inhibitory effects of acute
exposures to high concentrations of ozone, low-
level, chronic exposures result in an increase of
lung oxidative metabolism, Mustafa et al.144
studied mitochondrial oxygen consumption in rats
exposed conti nuously to 1 568 /vg/m3 (0.8 ppm) for
10 to 20 days. They found a 45-percent increase
(p<0,02)in oxygen utilization by the 20th day of the
experiment. There was also a 20- to 30-percent
increase in the oxidation of 2-oxyglutarate and
glycerol-1-phosphate. After 10 days of exposure,
there was a simultaneous increase (threefold)
(p<0.01) in the number of Type 2 alveolar cells that
are rich in mitochondria, so perhaps the increase in
02 comsumption was due to proliferation of these
cells. After 7 days of exposure to 392,980, or 1568
/vg/m3 (0.2, 0.5, or 0.8 ppm) ozone, there were
increases (p < 0.05) in Oz consumption of 1 7, 30,
and 40 percent, respectively.
The effects of various modes of ozone exposure
on 02 consumption of lung tissue were
investigated by Mustafa and Lee.150 In the first
experiment, rats were exposed to 1568^g/m3(0.8
ppm) continuously for 1 to 30 days. The rates of
oxygen consumption for oxidation of 2-
oxyglutarate, succinate, and glycerol-1 -phosphate
were slightly reduced in lung homogenate at day 1,
but began to increase thereafter. The increased
(p< 0,001) rate of Oz consumption reached a peak
at day 4 and plateaued over the remainder of the
exposure (up to 30 days). The oxidation of
succinate was increased more than the other twc
substrates. During a 7-day continuous exposure to
this ozone concentration (1568 pg/m3 or 0.8 ppm),
there was an initial decrease (day 1) and a
subsequent increase (day 2) in the activity of
mitochondrial-succinate-cytochrome-C reductase
that plateaued between days 3 and 7. By day 7, the
increased activity was 55 percent above control
(p< 0.001). Similar results were found for
microsomal NADPH-cytochrome-C reductase and
G-6-PD (see preceding section on , sulfhydryl
compounds and pyridine nucleotides).
As part of the same study,150 the effects of
continuous and intermittent(8 hr/day) exposure of
rats for 7 days were compared at the end of the
exposure. Exposures were to 392, 980, or 1568
/vg/m3 (0.2, 0.5, or 0.8 ppm) ozone. The increases
in 02 consumption (using succinate as a substrate)
and succinate-cytochrome-C reductase activity
were all significant (p < 0.05) and fairly pro-
portional to the ozone concentration. At the lower
ozone concentration, the increase in enzyme
activity was slightly greater in the intermittent
exposure group. However, the reverse occurred at
the two higher concentrations. The differences in
149
-------
the data from the two modes of exposure (i.e.,
intermittent, versus continuous) were not sig-
nificant. Similar concentration-related responses
occurred for NADPH-cytochrome-C reductase and
G-6-PD (see preceding section on sulfhydryl
compounds and pyridine nucleotides). Monkeys
were exposed intermittently to various
concentrations of ozone, and after 7 days, enzyme
activities of lung tissue were made. Rhesus
monkeys (n=24) showed elevations in succinate
oxidase activity of 12 percent (not statistically
significant) at 980 yug/m3 (0.5 ppm) and of 24
percent (p < 0.05) at 1568 yug/m3 (0.8 ppm).
Greater increases were observed in Bonnet
monkeys (n=13): 13 percent (not statistically
significant), 22 percent (p< 0.05), and 30 percent
(p < 0.02), for 392, 686, and 1568 yug/nn3 (0.02,
0.35, and 0.8 ppm), respectively. However, greater
increases were observed in the rats for the
corresponding ozone concentrations. Similar
species differences with respect to ozone
concentration were found for NADPH-
cytochrome-C reductase and G-6-PD. After 7 days
of exposure of the rats, there were significant
(p< 0.05) increases in NADPH-cytochrome-C re-
ductase and G-6-PD that were related to ozone
concentration. Though there was less increase in
the intermittent group, there was no statistically
significant difference between intermittent and
continuous exposure. Similar results were
observed for succinate oxidase and succinate-
cytochrome-C reductase (see preceding section on
sulfhydryl compounds and pyridine nucleotides).
As part of the same study, Mustafa and Lee 15°
also examined the effect of vitamin E on the effect
of ozone on succinate-dependent 02 consumption
of lung homogenate. Five weeks before ozone
exposure, rats were fed diets with either 66 ppm or
11 ppm of vitamin E. The authors state that 66 ppm
vitamin E is approximately twice the recommended
daily allowance, and that 11 ppm vitamin E is the
average concentration in American diets. The
animals were then exposed continuously for 7
days to either 196 or 392 yug/m3 (0.1 or 0.2 ppm)
ozone before assaysThe vitamin E-supplemented
group (66 ppm) exhibited an increase in succinate
oxidase activity of 3 percent (not significant) and
18 percent (p < 0.05) at 196 and 392 yug/m3 (0.1
and 0.2 ppm) ozone, respectively. Those animals
that received 11 ppm vitamin E had greater
increases in succinate oxidase activity, namely a
25-percent (p< 0.02) and a 38-percent (p <0.01)
increase for 196 and 392 Aig/m3 (0.1 and 0.2 ppm)
ozone, respectively. Vitamin E did not cause any
significant differences in succinate oxidase activity
of control animals. Mustafa and Lee 15° primarily
attribute the increases in these enzyme activities
to a concomitant increase of Type 2 cells (see the
later section on morphology studies), which are
rich in mitochondria and exhibit high metabolic
activity.
The effects of recovery from and re-exposure to
ozone on 02 consumption of lung was investigated
by Chow et al.29 Rats exposed continuously to 1568
yug/m3 (0.8 ppm) ozone for 3 days exhibited a 48-
percent increase (p < 0.001) in the rate of mito-
chondria! succinate oxidation. After the animals
recovered for 2 days in filtered air, this value began
to return to normal and reached control values 9
days after the initial exposure ceased. When the
rats were re-exposed to the same ozone
concentration (3 days, continuous) on days 6, 13,
and 27 of recovery, the rate of mitochondrial
succinate oxidation again increased to levels
similar to those observed for the intial ozone
exposure. Similar effects were observed for GSH
peroxidase, GSH reductase, and G-6-PD (see
preceding section on sulfhydryl compounds and
pyridine nucleotides). So again the data support
the inference that adaptation to these ozone-
induced responses did not occur.
Other Biochemical Alterations — In a review,
Mustafa et al. 149 present the res Jits of a study on
the effects of ozone on glucose metabolism in the
lung. Rats exposed to 1 568 fJQ/m3 (0.8 ppm) ozone
continuously for 4 days exhibited a 59 percent
(p< 0.02) increase in the rate of glucose consump-
tion. In addition, the rates of pyruvate and lactate
production both increased 43 percent (p < 0.02),
but the absolute amount of lactate produced (22.2
/ymoles/hr per lung) was greater than the amount
of pyruvate produced (3.7 jumojes/hr per lung).
This review149 also describes an effect of ozone
on monoamine oxidase (MAO), an enzyme that
catalyzes the metabolic degradation of biological
amines. Acute exposure (8 hr) to 3920 fjg/m3 (2
ppm) ozone caused a 30- to 40-percent (p < 0.05)
decrease in MAO activity of lung homogenate,
mitochondria, and microsomes. However, a 7-day
continuous exposure to 1568 fJQ/m3 (0.8 ppm)
resulted in a 20- to 35-percent (p < 0.05) increase
of MAO activity.
Other microsomal enzymes, namely mixed
function oxidases like cytochrome P4so (cyt. P45o)
150
-------
can be affected by ozone. Goldstein et al.88 found a
decrease (p < 0.01) in lung microsomal cyt. P45o
levels immediately or 1, 2, 3, or 5 days after rabbits
were exposed to 1960 yug/m3 (1 ppm) ozone for 90
min. The data for up to 45 days post exposure were
fitted to a parabolic curve with a significant
(p<0.001) multiple correlation coefficient. The
greatest depression of cyt. P45o occurred at 3,6
days after ozone exposure. Liver cyt. Pose levels
from the ozone-exposed rabbits were not
significantly different from control immediately or
1 day post exposure. In vitro exposure (1960
/yg/m3, or 1 ppm, for 90 min) of lung homogenate
resulted in a 52-percent decrease in microsomal
cyt. P4so levels. Lung microsomal fractions similar-
ly treated exhibited a 68-percent decrease in
cyt.P45o.
A cyt.-P45o-dependent mixed-function oxidase
(benzpyrene hydroxylase) was also examined for
ozone susceptibility. Palmer et al. made
measurements of both the lung parenchyma161
and tracheobronchial mucosa160 of Syrian golden
hamsters exposed to 1470 yug/m3 (0.75 ppm)
ozone for 3 hr. The lung tissue itself had a 33-
percent reduction in the activity of benzpyrene
hydroxylase, whereas the tracheobronchial
mucosa exhibited a 53-percent drop immediately
after exposure.The mucosal enzymatic activity had
recovered by 24 hr after ozone exposure.
Lysosomes (cellular organelles containing a
variety of enzymes) are also affected by ozone
exposure, as several researchers have shown. For
example, rats exposed to 1568 yug/m3 (0.8 ppm)
ozone continuously for 8 days showed significant
increases in lysozyme activity in lung homogenate
and plasma over the level of lysozyme activity in
rats exposed in the same manner to 0, 392, or 980
//g/m3 (0, 0.2, or 0.5 ppm) ozone. When animals
were exposed to the same ozone concentrations
intermittently (8 hr/day for 7 days), there were no
significant changes in pulmonary lysozyme
activity.28 In another study,46 rats were con-
tinuously exposed to 1372 /yg/m3 (0.7 ppm) ozone
for 5 days or 1568 /jg/m3 (0.8 ppm) for 7 days, and
the activities of various lysosomal hydrolases were
measured in whole lung homogenates and found
to increase significantly (p < 0.05). The authors
attributed this response to an infiltration of
phagocytic cells (which have a high concentration
of lysosomes) into the lung in response to ozone.
The higher enzymatic levels also possibly reflect
greater cell membrane lability. The authors
hypothesized that the increase in protease and
peptidase activities could lead to chronic
obstructive pulmonary disease.46 Afchough these
data are contrary to that of Hurst et al. 105'106 (see
the alveolar macrophage section), the authors
state that the difference was probably because
Hurst used isolated cells (alveolar macrophages),
whereas they used whole lung tissue. Damage to
lysosomal membranes was also proposed by
others after histo- and cyto-chemical studies of the
effects of 1 372 to 1568 yug/m3 (0.7 to 0.8 ppm)
ozone administered continuously for 7 days.25
These researchers observed increased acid
phosphatase activity and no changes in the /3-
glucuronidase activity in alveolar macrophages of
rats. The terminal airway epithelium and adjacent
structures also exhibited increased acid
phosphatase activity. This particular enzyme had a
nonvacuolar, as well as an intracellular distribu-
tion, indicating lysosomal membrane damage.
In another publication, the same authors26
examined similarly exposed rats for enzyme
activities. In areas of the bronchiolar epithelium
that were infiltrated by mononuclear cells, there
was a lower NADH- and NADPH-diaphorase
activity and an increased ATPase activity.
However, intra-alveolar septa in the centriacinar
region had increased diaphorase, LDH, ATPase
and cytochrome oxidase activities. These
increases were also seen in septa that were not
thickened or infiltrated with increased numbers of
cells. It was proposed that these enzymatic
changes are part of protective-adaptive
mechanisms that operate mainly in the
centriacinar region of the lung.
MORPHOLOGY STUDIES
Scheel et al.1B1 provided histopathologic
evidence of injury caused by a single acute
exposure to 1 960 or 6272 fjg/n\3 (1.0 or 3.2 ppm)
ozone for 4 hr in mice and by repeated intermittent
exposures of 1 5,680to88,200/yg/m3(8to45 ppm)
ozone for 1 hr in rabbits. Pulmonary edema was not
observed in mice sacrificed immediately after
exposure to 1960 /yg/m3 (1 ppm) ozone, but
moderately engorged capillaries containing an
excess of leukocytes were visible. Mice examined
20 hr after exposure showed mild edema and
migration of leukocytes into the alveolar spaces.
Superficial desquamation of the epithelium in the
bronchi and bronchioles was also observed.
Inhalation of 6272 jug/m3 (3.2 ppm) ozone
produced grossly visible edema. The perivascular
lymphatic vessels were distended and filled with
151
-------
edema fluid and precipitate. Hyperemia,
mobilization of leukocytes, and varying degrees of
extravasation of red cells into the tissues
accompanied the edema. Sheets of desquamated
bronchiolar epithelial cells were seen in the
lumina.
Pulmonary lesions observed in cats exposed to
510, 980, and 1960 //g/m3 (0.26, 0.5, and 1 ppm)
ozone for 4 to 6 hr indicated that desquamation of
the ciliated epithelium was apparently dose-
related.15'18 Other ultrastructural changes were
cytoplasmic vacuolizations of the ciliated celts that
occurred in airways predominantly 0,8 to 1.7 mm
in diameter. At the same sites, condensed
mitochondria with abnormal cristae appeared. In
the same reports,'5'15 swelling and desquamation
of Type 1 alveolar cells, swelling and breakage of
capillary endothelium, and erythrocyte lysis within
the interalveoiar capillaries were seen.
Mice (18- to 20-month-old males) were exposed
for 6 hr to 980 //g/m3 (0.5 ppm), 2160 //g/m3 (1.2
ppm), and higher levels of ozone and then were
given injections of tritiated thymidine that labeled
alveolar cells undergoing DMA synthesis,59
Inhibition of DNA synthesis occurred at all ozone
levels to a similar extent, thus representing
maximum inhibition. This depression was seen up
to 24 hr after the ozone-exposed mice were
transferred to clean air and did not return
completely to normal until 72 hr later.59
Evans et al.
57,58,60
studied the kinetics of alveolar
cell division in rats exposed continuously for 8 days
to 686 or 980 //g/m3 (0.35 or 0.5 ppm) ozone. Cell-
proliferation increased to a maximum at2 days and
began to decrease toward normal by the fourth
day. The principal dividing cell observed was the
Type 2 cell. This process is known to be a
mechanism for replacing damaged Type 1 cells.
After the fourth day, no further tissue damage or
increase in proliferation occurred, suggesting that
the tissue had become tolerant to that particular
concentration of ozone. To evaluate the extent of
this possible tolerance, after 4 days of exposure to
686 fjg/m3 (0.35 ppm), the ozone concentration
was elevated to 980 or 1372 //g/m3 (0.5 or 0.7
ppm) for up to 4 days. Under these conditions,
morphological damage increased, indicating that
tolerance did not occur. However, in rats exposed
initially for 4 days to 980 //g/m3 (0,5 ppm) and re-
exposed for 4 days to 1470 or 1960 //g/m3 (0.75 or
1 ppm), some tolerance was evident. Note that
studies with young and old mice exposed to ozone
have shown a decrease in cell proliferation.58'190
This result may be due to age, species, or the ozone
concentration used in the studies.
From the studies of Stephens et al,190 conducted
with rats, it also appears that the Type 1 cells are
the alveolar epithelial cells most sensitive to low
levels of ozone, 980 //g/m3 (0.5 ppm) for 2 hr. The
earliest change noted was a swelling of the
mitochondria of these cells, particularly those
located in the first two or three alveoli immediately
beyond the terminal bronchiolar epithelium. This
was followed by more severe alterations, which led
to leaving the basement lamina devoid of an
epithelial covering. Type 2 cells were resistant to
the ozone treatment, and after 4 to 6 hr of
exposure, showed signs of spreading over the
injured area.
Evans et al.5a examined renewal of the terminal
bronchiolar epithelium of rats at various time
periods after a 24-hr exposure to 1 372 //g/m3 (0.7
ppm). One hour after conclusion of the exposure,
almost all proliferating cells were nonciliated;
however, by 4 days, only approximately 76 per-
cent of the proliferating cells were nonciliated, the
remainder being ciliated. At 15 days, approxi-
mately the same relationship was observed. These
events as well as other observations led the
authors to the conclusion that nonciliated cells
divide after ozone exposure and form new ciliated
and nonciliated cells. In this manner, the
nonciliated cells were thought to act as progenitor
cells that participate in the recovery of the
damaged terminal bronchiotar epithelium.
Morphologic observations made by Dungworth
and coworkers,
49,50
indicate that the rat and
Bonnet monkey (Macaca radiata) are approx-
imately equal in susceptibility to short-term effects
of ozone. Mild but significant lesions were caused
in both species by exposure to 392//g/m3 {0.2 ppm)
for 8 hr/day for 7 days. The authors stated that
detectable morphological effects in the rat
occurred at levels as low as 196 //g/m3 (0.1 ppm).
In both species, the lesion occurred at the junction
of the small airways and the gaseous exchange
region. In rats, the prominent features were
accumulation of macrophages, replacement of
necrotic Type 1 epithelial cells with Type 2 cells,
and damage to ciliated and nonciliated Clara cells.
The principal site of damage was the alveolar duct.
In monkeys, the prominent ozone-induced injury
was limited to the small airways. At 392 //g/m3
(0.2 ppm), the lesion was observed at the proximal
portion of the respiratory bronchioles. As
concentrations of ozone were increased up to
152
-------
1568 )ug/m3 (0.8 ppm), the severity of the lesion
increased, and the damage extended distally to
involve the proximal portions of the alveolar duct.
Mellick and coworkers126 found similar but
more pronounced effects when rhesus monkeys (3
to 5 years of age) were exposed to 980 to 1568
fjg/rr\3 (0.5 and 0.8 ppm) ozone, 8 hr/day for 7
days. In these experiments, the respiratory
bronchioles were the most severely damaged, and
more distal parenchymal regions were unaffected.
Major effects were hyperplasia and hypertrophy of
the nonciliated bronchiolar epithelial cells and the
accumulation of macrophages intraluminally. In
mice, continuous exposure to 980/vg/m3(0.5 ppm)
ozone caused nodular hyperplasia of Clara cells
after 7 days of exposure.
Similar findings were reported by Schwartz183
and Schwartz et al.,184 who exposed rats to 392,
980, or 1 568 fjg/m3 (0.2,0.5 or 0.8 ppm) ozone for
8 or 24 hr/day for 1 week. Changes observed
within the proximal alveoli included infiltration of
inflammatory cells and swelling and necrosis of
Type 1 cells. In the terminal bronchiole, the
changes reported were shortened cilia, clustering
of basal bodies in ciliated cells suggesting cilio-
genesis, and reduction in height or loss of
cytoplasmic luminal projection of the-Clara cells.
Effects were-seen at_ozone concentrations as low
as 392 pg/m3 (0.2 ppm). A dose-dependent
pulmonary response to the three levels of ozone
was evident. No differences were observed in
morphologic characteristics of the lesions
between rats exposed continuously and those
exposed intermittently for 8 hr/day.
Using electron microscopy, Bils11 studied the
lungs of mice of different ages (4 days or 1 to 2
months) exposed to 1176 to 2548pg/m3 (0.6 to 1.3
ppm) ozone for 6 to 7 hr/day for 1 to 2 days and
noted swelling of the alveolar epithelial lining cells
without intra-alveolar edema. Swelling of
endothelial cells and occasional breaks in the
basement membrane were seen. The younger
mice exposed for 2 days were the most sensitive,
although these observations are contrary to the
demonstrations of the limited sensitivity of
newborn rats exposed from birth to NC>2.71'121
Brummer et al.21 used scanning electron
microscopy (SEM) to examine centriacinar regions
of the lung and to count the influx of cells within
the lumina in order to evaluate the cellular
response during ozone exposure. Continuous
exposure to 981 or 1570/yg/m3(0.5or0.8ppm)for
7 days resulted in an 18-fold increase in
inflammatory cells within the lumina of proximal
alveoli.
Sato et al.180 studied the effects of 588 /vg/m3
(0.3 ppm) ozone (3 hr/day for 16 days) on the
morphologic features of the conducting airways
and alveoli of rats using SEM. Observation of the
luminal surface of the bronchi, bronchioles, and
terminal bronchioles of the exposed animals
showed cilia that were swollen and adhered to one
another. Small, smooth-surfaced round bodies
were observed mainly around the tips of the cilia.
The ciliated areas of the lobar bronchi and proximal
bronchioles were covered with a pseudomem-
brane. Clara cells were more prominent and
greater in size in all rats exposed to ozone. The
surface of the alveolar ducts and alveolar walls
showed scattered areas of cytoplasmic swelling
and attachment of round bodies throughout the
parenchyma. Alveolar pores were distorted and
smaller in size than the control groups. The Type 1
cells, as others have shown, appeared to be most
vulnerable to ozone. These studies and others126
also showed that the vitamin-E-deficient rats were
more sensitive to ozone than rats fed a diet
supplemented with vitamin E.
Stokinger et al.195 reported that chronic
bronchitis, bronchiolitis, and emphysematous and
fibrotic changes develop in the lung tissues of
mice, rats, hamsters, and guinea pigs exposed 6
hr/day, 5 days/week for 14.5 months to a
concentration slightly above 1 960 pg/m3 (1 ppm)
ozone.
Rats exposed for 3 to 5 months to 1 568 /yg/m3
(0.8 ppm) ozone develop a disease that resembles
emphysema, and they finally die of respiratory
failure.192 Ozone results in a greater response of
fibroblasts in the lesion, thickening of the alveolar
septae, and an increase in number of alveolar
macrophages in the proximal alveoli.
Dogs were exposed to 1960 /-ig/m3 (1 ppm)
ozone for 8 hr daily for 18 months.72 The resulting
damage appeared roughly proportional to the time
and concentration of the ozone exposure. The
number of alveolar macrophages increased,
fibrous elements were deposited, and the small
airway lumina were reduced by a thickening of the
terminal airways and respiratory bronchiolar
walls. The peribronchiolar area contained
excessive numbers of lymphocytes and plasma
cells, and squamous metaplasia of the columnar
and cuboidal epithelium developed.
Another long-term experiment was undertaken
by Freeman et al.70 in which rats were
153
-------
continuously breathing ozone at either 1058 or
1725 /ug/m3 (0.54 or 0.88 ppm) for periods of up to
6 months. When the animals were exposed at the
lower ozone level for 2 hr, no immediate significant
histological changes were seen; but after 24 hr,
epithelial hypertrophy was found between the
distal portions of the terminal bronchioles and the
alveolar ducts. The peribronchiolar alveolar area
contained amorphous material resembling
necrotic cells. After 48 hr of exposure, the
hypertrophy of the epithelium of the alveolar ducts
was in sharp contrast to the controls, and
numerous mitotic figures were seen in the
terminal bronchiolar and proximal alveolar
epithelium. The interstitial tissue in this area
contained some cells with pyknotic nuclei.
Macrophages were prominent. Fibroblasts
appeared to be present beneath the ductal
epithelium, and the respiratory bronchiolar
epithelium had deposits of connective tissue
elements. Histochemical examination suggested a
change in the collagen. Additional rats were
studied at 6 days, 8 days, 3 weeks, and 6 months,
and progressive changes occurred in the airway
epithelium. Metaplasia of cuboidal epithelium was
seen, which continued to be replaced by
connective tissue. However, the terminal
bronchioles appeared more normal after 3 weeks
of ozone and remained so for the rest of the
exposure period. Even with this partial recovery,
the deposition of connective tissue interrupted the
continuity of the small airways with their proximal
alveoli
When the rats were exposed to 1725 /ug/m3
(0.88 ppm) ozone, epithelial injury was seen in 4
hr. The morphological alterations found at the
lower concentration were present in these rats
also, but they were more extensive and occurred
earlier. Furthermore, after 48 hr of ozone, the
bronchiolar epithelium was often metaplastic. By
day 3, there was an onset of early fibrosis, which
constricted some terminal airways. After 6 days,
adenoma-like structures containing large
numbers of macrophages were more prevalent.
Fibrosis had extended into the termmal
bronchioles. After 3 months of exposure to 1725
/ug/m3 (0.88 ppm) ozone, emphysema-like lesions
had developed. This study also included a group of
animals exposed to both 1692 /ug/m3 (0.9 ppm)
N02 and 1 764 /ug/m3 (0.9 ppm) ozone, and another
group that received 4900/ug/m3 (2.5 ppm) NO2 and
490 /ug/m3 (0.25 ppm) ozone. They, too, were
examined for pathological changes. However, the
authors found no synergism because ozone
appeared to be largely responsible for the
alterations seen at these concentrations.70
Investigations by others have supported
previously discussed data and, in some cases, have
added more information. P'an et al.162 exposed
rabbits to 784/ug/m3 (0.4 ppm) ozone for 6 hraday,
5 days a week for 10 months. The a uthors observed
both emphysematous and vascular lesions in the
lungs. The small pulmonary arteries were thicker
than in controls because of the increased size of
the tunica media vasorum, which in some
instances appeared to be a result of tissue edema
or muscular hyperplasia.
The possibility that daily ozone exposure might
increase the incidence of tumors has also been
studied. Stokinger193 reported that acceleration of
lung tumorigenesis (adenoma) in a strain of mice
susceptible to such tumors occurred from daily
ozone exposures to about 1960 /ug/m3 (1 ppm). At
15 months, a tumor incidence of 85 percent was
found in the ozone-exposed, as against 38 percent
in the control mice. The average number of tumors
per exposed mouse was 1.9 compared with 1.5 in
the controls. Experimental details were not given.
Kotin and coworkers114'116 studied the
experimental induction of pulmonary adenomas in
Strain A mice (which are prone to develop
pulmonary adenomas) and in C-57 black mice
exposed for 52 weeks to an atmosphere of
ozonized gasoline. The concentration of oxidants
varied from 1 to 3.8 ppm. The concentration of
other pollutants present was not reported. A
significantly greater incidence of tumors occurred
in the exposed than in the control mice. There was .
no significant difference in tumor incidence
between males and females. The respiratory
epithelium of the mice revealed significant
hyperplastic and metaplastic responses.
Nettesheim et al.157investigated the effect of o-
zonized gasoline on tumor production in hamsters.
The ozone concentration was 2353 /ug/m3 (1.2
ppm) in excess of that which reacted with the
gasoline. The resulting hydrocarbon mixture was
40 to 45 ppm methane equivalents. Animals were
exposed 6 hr/day, 5 days/week for life to ozonized
gasoline, to ferric oxide (Fe203), or to both. Some of
the groups also received injections of the
carcinogen, diethylnitrosamine (DEN). The
tumorigenicity of DEN was enhanced by Fe203,but
not by the ozonized gasoline. The ozonized
gasoline, alone or in combination with Fe2C>3, did
not appear to be carcinogenic. When the data for
154
-------
the groups of ozonized gasoline plus DEN and
ozonized gasoline plus FesOa were compared with
data for DEN and FesOa control groups, and were
analyzed under the assumption that the tumors of
the bronchi and lungs were lethal, there was a
reduction (p < 0.05) in tumor incidence in the
groups receiving ozonized gasoline.
PULMONARY FUNCTION
Scheel et al.181 exposed 75 rats to 3920 fjg/m3 (2
ppm) ozone for 3 hr and then measured their
oxygen uptake, tidal volume, and frequency of
breathing. The rats were examined at intervals
over a period of 960 hr after exposure. The lungs
were excised and weighed before and after drying.
It was found that the water content of the lungs
increased during the post-exposure period, the
increase reaching a maximum after 12 hr. These
authors expressed pulmonary edema in terms of
water in the lung per kilogram of body weight
instead of per unit of dry-lung weight. Since no
data are given on body weight, the changes in
lung-water content must be interpreted with
caution. A decrease in minute ventilation (the
volume of air breathed per minute), tidal volume
(volume per breath), and oxygen uptake occurred
immediately after exposure and reached minimum
recorded values 8 hr after exposure. At 20 hr after
exposure, all measurements had returned to their
initial values. The early fall in minute ventilation
and oxygen uptake could have reflected a decrease
in activity (metabolic rate) secondary to acute
injury of tissue and the development of pulmonary
edema. The simultaneous increase in minute
ventilation and decrease in oxygen consumption
after 40 hr suggests that delayed impairment may
have occurred.
Murphy et al.146 exposed guinea pigs to 666 to
2646 A
-------
pulmonary resistance. Greatest reductions in Cdyn
and DLco were observed at the higher ozone
concentration (1 960 fjQ/m3, or 1 ppm). There were
no changes in vital capacity. Alveolar stability was
also unchanged, as illustrated by the shapesof the
volume-pressure curves during deflation for
exposed and control animals. Pathological reports
on similarly treated animals indicated severe
injury to the bronchial and bronchiolar epithelium,
Type 1 alveolar cells, and mteralveolar capillaries.
This probably contributed to the depression of
diffusing capacity. No evidence of intra-alveolar
edema was seen. The authors suggest that acute
ozone exposure affected airway caliber more than
either the transfer of CO across the alveolar-
capillary membrane or alveolar surface forces.204
Bartlett et al.9 exposed young rats (3 to 4 weeks
old) continuously for 30 days to 392 fjg/m3 (0.2
ppm) ozone. No changes in respiratory frequency
occurred. Morphometric analysis indicated that
the numbers of alveoli were unchanged, but the
average lung volume increased by 16 percent
(p< 0.02) following ozone exposure.The increases
(p< 0.05) in mean cord length and alveolar surface
area of ozone-exposed animals reflect
overdistension of the lung. Overdistension is also
illustrated by the lower (p < 0.05) transpulmonary
pressure that occurred in both air- and saline-filled
lungs, which indicated that a decrease in tissue
elasticity rather than alterations of surface
tension, was responsible for the results. Using
light microscopy, no morphological changes were
evident. At higher concentrations of ozone,
investigators216 found no significant changes of
the pulmonary pressure/volume relationship or
the ratio of collagen to elastin in the lungs of rats
exposed to 880 //g/m3 (0.45 ppm) ozone for 6
hr/day for 6 to 7 weeks
Exposure to 1 960 fJQ/m3 (1 ppm) ozone for 3 hr,
using the unilateral lung technique, in which only
one lung is exposed to ozone, resulted in a loss of
elasticity in rabbit lungs immediately after
exposure (p < 0.05) and 3 days later (p < 0.05).213
By day 14, the lung had recovered, suggesting that
edema, present in the exposed lung on days 1
through 3 (p < 0.05), could have been responsible.
EDEMAGENESIS AND TOLERANCE
Data from many sources reviewed by Coffin and
Gardner37 indicate that brief exposure to ozone
may elicit edema and an acute inflammatory
response in the lungs of many species. It is striking
at higher levels of exposure and diminishes with
descending concentrations of the gas. Thus no
evidence of edema was noted in mice exposed to
1960 fJQ/m3 (1 ppm) ozone for 4 hr, whereas
exposure to this concentration for 2 hr resulted in
very slight edema when a gravimetric measure-
ment of lung water was employed.181'194 These
investigators16'194 detected evidence of edema in
rats after exposure to 3920 //g/m3 (2 ppm) ozone
for 3 hr. Exposures at the high level of 6272//g/m3
(3.2 ppm) ozone quickly produced gross evidence of
edema in mice.
It is generally believed that gross edema is
probably not elicited in any species exposed to
ambient concentrations of ozone. More refined
methods for the detection of possible edema were
developed by Alpert et al.,5 greatly increasing the
sensitivity of the measurement. They injected 132I-
serum albumin into rats and recovered pulmonary
fluid 6 hr later by lung lavage, which was then
tested for radioactivity. Radioactivity in the fluid
would be indicative of fluid flux across pulmonary
membrane barriers, and an increase over controls
would represent possible edema. After 6 hr of
exposure to 490 //g/m3 (0.25 ppm) ozone, there
was no significant effect; but to 980 fjg/m3 (0.5
ppm), there was increased recovery of 132I (two
times that of controls) (p< 0.001). Simultaneous
studies using lung wet weight/dry weight ratios
showed an effect only with 4900 //g/m3 (2.5 ppm)
ozone (p <0.01).
Another report102 indicates that a 5-hr exposure
to lower levels of ozone caused increased perme-
ability of the alveolar-capillary membrane in mice
as measured by recovery of radiolabeled albumin
in the lavage fluid. However, because the objective
of the study was to measure the ozone hazard in
UV isolation units, the mice were exposed to
varying concentrations of ozone (an average of 1 96
fjQ/m3, or 0.1 ppm) under conditions that could be
called uncontrolled relative to typical studies of
ozone toxicology.
Frank et al.69 exposed the right lung of rabbits to
4312 and 23,716 fjg/m3 (2.2 to 12.1 ppm) ozone
for 3 hr, the left lung having been collapsed before
exposure. When edema occurred in the right lung,
changes in surfactant (a substance that lowers
surface tension in the lung, helping to make
normal breathing possible) behavior were
observed in the left lung of some animals. No such
changes were observed in the absence of edema in
the right lung. These results suggest that ozone is
not only capable of inducing chemical changes in
exposed lungs, but also that the products of such
156
-------
changes are capable of producing deleterious
effects or compensatory responses in nonexposed
lungs.
Ozone has been found to induce tolerance,
which can be described as the acquired capacity of
a pretreated host to exhibit a lesser response to a
challenge than would be observed in a comparable
but nontreated host. Using the radiolabeled
albumin technique, Alpert et al.5 examined the role
of methylprednisoline (a corticosteroid with potent
anti-inflammatory properties) on edema and
tolerance to edema in rats. They found that it
increased susceptibility to edema production
following exposure to low levels of ozone.
However, when animals received the steroid
before exposure to 490 Aig/m3 (0.25 ppm) ozone,
they became tolerant to a subsequent ozone
challenge with 1 960pg/m3(1 ppm)48hr later(i.e.,
edema was not produced). Untreated rats did not
become tolerant.
Other work was undertaken to ascertain the
mechanisms and extensiveness of tolerance. In
one such study, the left lungs of Flemish Giant
rabbits were catheterized and exposed to 980
pg/m3 (0.5 ppm) ozone for 3 hr. The right lung
inspired clean air. This was followed by an 1 8-hr
latent period, after which, the whole animal was
challenged with 43,120 pg/m3 (22 ppm) for 3 hr.
The pre-exposed left lung exhibited tolerance to
pulmonary edema (p < 0.001), and the right lung
did not, indicating that tolerance is a localized,
rather than a systemic, effect."
Similar methods were used by Gardner et al.,78
who pre-exposed the left lung to 980 or 1960
fjg/m3 (0.5 or 1 ppm) ozone for 3 hr, and 18 hr later
exposed the whole animal to either 5880 pg/m3 (3
ppm) or 43,1 20 fjg/m3 (22 ppm) ozone for 3 hr.The
right and left lungs were then examined
separately. Again, it was found that pulmonary
edema was prevented only in the pre-exposed lung
(p< 0.001). Also, the 980 pg/m3(0.5 ppm) ozone
produced greater tolerance than did the 1960*
pg/m3 (1 ppm) ozone pre-exposure. The same
authors had demonstrated that ozone exposures
resulted in an influx of polymorphonuclear
leukocytes (PMN) into the lung, while the total
number of alveolar macrophages remained the
same. Therefore, the effect of ozone pre-exposure
on this parameter was studied. The tolerant pre-
exposed lung had the same total number of
alveolar macrophages as the nontolerant lung, as
well as a higher number of PMN This
demonstrated that the prior exposure induced
further chemotaxis of PMN but had no effect on
alveolar macrophages. Macrophage enzyme
activities (lysozyme and /i-glucuronidase) were
reduced in response to ozone, and pre-exposure
failed to influence the action of ozone on these
enzymes. These investigations indicated that the
extent of tolerance depends on the biological
endpoint and is more effective against pulmonary
edema than against other health effects.
Using the infectivity model, which measuresthe
accumulated mortality of mice exposed to
pollutants and aerosols of Streptococcus
pyogenes, Gardner and Graham77 investigated
ozone tolerance. Some animals received two 3-hr
exposures to the same concentration of ozone,
with a 24-hr interval between the exposures.
Other mice received only one 3-hr exposure to
ozone. The bacterial aerosol was administered
immediately after the last ozone exposure. At 196
/jg/m3(0.1 ppm), there was only a slight difference
in mortality between the mice receiving 2 and 1
exposures to ozone. As the concentration was
increased, up to 1960 Aig/m3 (1 ppm), there was
greater protection in those animals pre-exposed
(i.e., the mortality was less in those mice receiving
ozone twice, as compared to animals exposed
once). However, excess mortality occurred
irrespective of the number of ozone exposures,
indicating that complete tolerance was not
evident. The partial tolerance to infectivity that
occurred at higher ozone concentrations was
probably due to inhibition of edema, and not to
tolerance to an ozone effect on alveolar
macrophages. That ozone-induced damage to
alveolar macrophages cannot be completely
prevented by prior ozone exposure has been
demonstrated by Gardner et al.93
Evans et al.59'60 measured tolerance by studying
the kinetics of alveolar cell division in rats during a
period of exposure to an elevated ozone
concentration (980 or 1372 fjg/m3, or 0.5 or 0.7
ppm, up to 4 days) that followed initial exposure at
a lower concentration (686 pg/m3, or 0.35 ppm for
4 days). Tolerance in this case was the ability of
Type 1 cells to withstand re-exposure, and any
increase in the numbers of Type 2 cells would
indicate a lack of tolerance. These studies showed
that tolerance to the initial concentration of ozone
did not ensure total protection against re-exposure
to higher concentrations (see section on
morphology studies for more complete discussion).
Note that when mice were exposed to 1960
1 ppm) ozone for 1 hr and then X-irradiated
157
-------
with 800 R 10 days later, 60 percent of the animals
survived, whereas all previously non-ozone-
exposed mice died. This tolerance lasted 30 days or
more."
The mechanism for development of tolerance to
ozone-induced edema is not well understood, but
the thymus may play a role. In a study by Gregory et
al.,93 male mice thymectomized at birth were
unable to develop tolerance to a 2-hr exposure of
39,200//g/m3 (20 ppm) ozone when they had been
pre-exposed to 588 //g/m3 (0.3 ppm) of ozone for 1
hr {p < 0.01 ).93 Sham-operated animals exhibited
tolerance under the same oxidant conditions.
Contrary to these studies, Thompson198 found that
thymectomizing the animals had no effect on
tolerance.
Biochemical responses in lungs of ozone-
tolerant rats were studied by Chow27 and Chow et
al.29 (see section on sulfhydryl compounds and
pyridme nucleotides, and earlier section on
mitochondrial enzyme activities for complete
description). The lungs of tolerant animalsshowed
relatively less injury and maintained significantly
higher activities of glutathione peroxidase,
glutathione reductase, and glucose-6-phosphate
dehydrogenase, and higher levels of reduced
glutathione. The author speculates that this higher
activity in the lungs of tolerant animals may render
them more resistant to peroxidation damage
induced by ozone. However, with respect to the
mitochondrial oxidation of succinate, re-exposure
after a recovery period resulted in alterations in
enzymatic activity similar to those observed after
initial exposures.
Dixon and Mountain47 concluded that ede-
magenesis in mice exposed to ozone probably
involves the release of histamme and other
endogenous products such as slow-reacting
substances, but they found that lung histamine
content did not parallel tolerance and that
tolerance was not affected by histamine
antagonists or liberators. Their conclusion was
that histamine per se had no function in
development of tolerance to ozone.
Extrapulmo'.dry Effects
HEMATOLOGICAL AND SERUM CHEMISTRY
CHANGES
Some of the studies to be described here have
also been conducted with human blood and are
discussed in Chapter 9.
Brmkman et al.20 have shown that inhalation of
392 to 490 /jg/m3 (0 20 to 0.25 ppm) ozone over
periods of 30 to 60 min by mice, rats, rabbits, and
man increased the rate of sphering of red cells in
vitro, with the cells losing their characteristic
biconcave shape more rapidly when the diluted
blood is subsequently exposed to X-radiation.
These changes were greatest after 1 hr and
approached control values 6 hr after exposure
ended.
Menzel et al.132 investigated the effects of ozone
on erythrocyte hemolysis. Rats maintained on
diets either deficient in or supplemented with
vitamin E were continuously exposed to 980
/jg/m3 (0.5 ppm) ozone, and their red blood cells
were subjected to dialuric acid hemolysis to
estimate plasma tocopherol. After 23 days, the
erythrocytesfrom 60 percent of the rats depleted of
vitamin E and exposed to ozone showed hemolysis.
This response did not occur in the vitamin-E-
depleted, air-exposed control animals until 36
days. Those animals supplemented with vitamin E
did not show hemolysis, regardless of exposure.
These data are believed to illustrate that ozone
accelerates the depletion of tocopherol reserves,
Ross et al.177 examined erythrocytes of rabbits
exposed for 4 hr to 1960 or 5880 /ug/m3 (1 or 3
ppm) ozone. There was no significant effect,
immediate or delayed, on parameters that reflect
oxygen delivery by the red cell; i.e., oxyhemoglobin
affinity, heme-oxygen binding site interaction, and
2, 3-diphosphoglycerate concentration.
Other studies182 were conducted in which mice
breathed 1686 fjg/m3 (0.86 ppm) ozone for 8
hr/day, 5 days/week for 6 months, after which the
animals were infected w\ihPlasmodiumberghei(a
sporozoan, of the same family as the malaria
organisms, that parasitizes the red blood cells of
certain mammals). Up to the 6th week of exposure,
there was a significant increase in resistance of
the erythrocytes to acid hemolysis.182 Other
investigators observed increased resistance to
hemolysis in mice exposed to 1 960 /Jg/m3 (1 ppm)
ozone for 30 min.13e
Ozone may not only affect the unsaturated fatty
acids of cell membranes, but it may also damage
acetylcholinesterase (AChE), an enzyme bound to
red cell membranes. This was studied163 after in
vitro exposure of AChE derived from bovine
erythrocytes. The enzymatic activity was inhibited
after exposure to > 588 /yg/m3 (0.3 ppm). Eglite53
reported that whole blood cholinesterase of rats
decreased by the middle of the third month of a 93-
day continuous exposure to 110 /Jg/m3 (0.056
ppm), but returned to normal 12 days after
158
-------
exposure ended. The method of ozone measure-
ment was not specified.
Chow et al.30 found no statistically significant
changes in the level of GSH or the activities of GSH
peroxidase, GSH reductase, and G-6-PD in the
erythrocytes of rhesus monkeys or rats exposed to
0.980 /jg/m3 (0.5 ppm) ozone for 8 hr/day for 7
days. However, in the lungs of the rats, increases (p
< 0.05) of GSH and these enzyme activities were
observed (see earlier section on sulfhydryl
compounds and pyridine nucleotides).
Leukocytes also have been investigated. High
concentrations of ozone (> 2500 /xj/m3, or > 12.8
ppm for 2 hr) can increase the neutrophil:
lymphocyte ratio of circulating blood of rabbits.
However, exposure to 3920 or 2940 /ug/m3 (2 or
1.5 ppm) ozone for 5 hr did not change this ratio.17
P'an and Jegier164 observed serum protein
changes in rabbits exposed to 784 or 1 960//g/m3
(0.4 or 1.0 ppm) ozone for 6 hr/day, 5 days/week
for 10 months. Exposure to the lower ozone
concentration resulted in a small progressive
decrease of serum albumin (16 percent by day
210). Gamma globulin varied within ± 10 percent
of control values over the first 90 days, but
increased thereafter and was 45 percent above
control after 21 0 days of exposure to 780 //g/m3
(0.4 ppm) ozone. At the same ozone concentration,
there was a slight decrease (approximately 8 to 10
percent from controls) of a and /3 globulin over the
first 30 days of exposure; ft globulin then increased
to control levels for the remainder of the
experiment. However, the a globulin of the
exposed animals continued to increase, reaching
78 percent above controls at 210 days of exposure.
Similar types of changes were observed at the
1960 /ug/m3 (1 ppm) exposure for 190 days, at
which time a and y globulin had increased by 46
percent and 48 percent, respectively, and albumin
had decreased by 11 percent. No significant
changes were observed in total serum protein
concentrations of either of the exposure groups.
To determine whether ozone can cause pre-
emphysematous changes in the lungs of small
animals by destroying serum antitryptic factors,
P'an and Jegier165 undertook to examine the sera
of rabbits exposed to 784 /ug/m3 (0.4 ppm) ozone
for 6 hr/day, 5 days/week for 6 months. Except for
an increase in the serum trypsin inhibitor capacity
on the first day of ozone exposure, no significant
changes were found. A slow but steady rise in
serum protein esterase was found also after
rabbits were exposed to 784 /ug/m3 (0.4 ppm) for 6
hr/day, 5 days/week for 10 months.110 For 10
months, rabbits were exposed to 784 /yg/m3 (0.4
ppm) ozone, and the concentration of serum
trypsin protein esterase tripled by the end of
exposure. This rise may be related to the observed
thickening of small pulmonary arteries.166
Numerous other changes have been observed in
the blood of ozone-exposed animals. Veninga's202
work, in which rabbits were exposed for 60 min to
392 /ug/m3 (0.2 ppm) ozone, showed a small but
significant drop in total blood serotonin
immediately after termination of exposure. This
reduction was probably due to a loss of platelet-
bound amine because there was no measurable
free circulating plasma serotonin. Investigation of
plasma lysozyme28 activity showed that rats
continuously exposed to 1568 /ug/m3 (0.8 ppm)
ozone over 29 days had increased (p < 0.001)
enzyme activity by the third day of exposure. The
lysozyme activity remained elevated on day 1 Oand
day 29. In another study, mice exposed to 392
/ug/rn3 (0.2 ppm) ozone for 2 hr exhibited an
increase in serum glutamic-pyruvtc transaminase
(SGPT), indicating enhanced fat deposition in the
liver as well as an increase in hepatic ascorbic acid.
There was no change in blood catalase, a strong
reducing agent.
£01
CHROMOSOMES
There have been numerous studies designed to
provide information on the possible role of ozone
as a mutagenic agent, A number of investi-
gators4243 96'96'97185 have used the microorganism,
Escherichia cofi, as the indicator system for ozone-
induced specific modifications of the genetic
material. Fetner6b has demonstrated the capability
of ozone to produce chromosome breaks in the root
menstem cells of the broad bean, Vic/a faba. The
same author66 reports a significant delay in mitosis
when living neuroblasts of the grasshopper,
Chortophaga viridifaciata, are exposed to ozone.
No effect was detected until the embryos were
exposed to an ozone atmosphere present in a
closed system with a solution of 3500 to 4500 /ug
ozone/liter. These effects were reversible. The
effects of ozone-treated seawater on the oyster
has also indicated the presence of fragmented
nuclei, indicating chromosome breaks.122
Zelac et al.217 exposed female Chinese hamsters
to 470 /jg/m3 (0.24 ppm) ozone for 5 hr and then
examined circulating blood lymphocytes for
chromosome damage. Ozone-induced chromoso-
mal breaks (p < 0.05) and aberrations were still
159
-------
present 6 and 1 5 days post exposure to ozone. The
authors suggest the possibility of long-term effects
from these cellular alterations. Expanding these
studies, Zelac et al.218 calculated that the
exposure-adjusted break frequency was 1,67 x 10~3
breaks/cells (ppm/min). When they combined
ozone (392 /ug/m3, or 0.2 ppm, for 5 hr) with
radiation exposure (230 rads delivered in 5 hr), the
two agents simultaneously exhibited > 70 percent
of the total number of breaks anticipated,
assuming the actions of two agents were additive.
Gooch et al.92 studied the cytogenetic effects of
ozone in three systems: Chinese hamster bone
marrow, mouse peripheral leukocytes and mouse
primary spermatocytes. The hamsters were
exposed to 451 /ug/m3 (0.23 ppm) for 5 hr or
10,192 /ug/m3 (5.2 ppm) for 6 hr, The mice were
exposed to 294 /ug/m3 (0.1 5 ppm) and 41 2 /ug/m3
(0.21 ppm) for 5 hr, or 1940/ug/m3 (0.99 ppm)for2
hr. The data tended to disagree with those of Zelac
et al. 217>218 m that no appreciable increase in
frequency of ozone-induced chromosome-type
aberrations was observed. Other researchers 179
exposed cell cultures of embryonic chick
fibroblasts (total flask vol. of 1 60 ml) to 1 Oto 100/ul
ozone/ml for 30 min. These cells exhibited major
but nonspecific alterations in both mterphase and
rnitotic cells at the margins and top layers of the
cultures. Several hours post exposure, abnormal
bi- and poly-nucleated reconstructed cells were
present. A few cells had chromosome bridges in
anaphase and telophase together with nuclear
fragments that resembled X-ray-induced damage.
Fetner 67 exposed human cell cultures in vitro to
ozone (8.0 ppm by weight of ozone m 02 for 5 and
10 mm). This investigation demonstrated that
ozone is capable of producing chromatid breakage
in the KB human cell line.
A review article by Venmga,202 which includes
some of his own research, presents evidence for
the radiomimetic properties of ozone. C-57 black
mice were treated with 392/ug/m3 (0.2 ppm) ozone
for 7 hr/day, 5 days/week during gestation and
then for the first 3 weeks of life. Unlimited incisor
growth was found in 0.9 percent of the normal
newborn mice, whereas ozone-exposed animals
had an incidence of 5.4 percent. In the same
species, neonatal death rose from 9 to 34 percent
after exposure. In addition, blepharophimosis in
another strain of normal mice occurred m 0.6
percent of control animals and m 9.6 percentof the
ozone groups. Similar but less dramatic results
were seen in C-57 black mice m which the rate of
blepharophimosis rose from 4.5 to 9.2 percent. In
another study by the same researcher,201
binucleated lymphocytes doubled in number in
murine blood after exposure to 392 /ug/m3 (0.2
ppm) for 2 hr (p = 0.005). The action of ozone in
changing the adsorption spectra of nucleic acids33
and in altering the pyrimidine bases of E. coli
nucleic acids168 may be an indication of a
chromosomal effect.
CENTRAL NERVOUS SYSTEMS AND
BEHAVIORAL EFFECTS
Several studies have shown tne effects of ozone
on behavior patterns in animals.
Various responses of mice exposed for 30 min to
various concentrations of ozone from 1176 to
1 6,660 /ug/m3 (0.6 to 8.5 ppm) were studied.167 At
11 76 /ug/m3 (0.6 ppm), mice began attempting to
move away from the ozone. This avoidance
response was more definite at 2156 /ug/m3 (1.1
ppm) ozone, and it became more pronounced as
the ozone concentration increased. In addition, as
ozone concentration increased, activity decreased.
After a 20-rnin exposure to 2156 /jg/m3 (1.1 ppm),
the animals appeared confused. Other re-
searchers112'113 exposed rats to 98, 196, 392,980,
and 1960 /ug/m3 (0.05, 0.1, 0.2, 0.5, and 1 ppm)
ozone and found that the greater the concentration
of gas, the greater the depression in gross motor
activity (p < 0.05). Although not specified in the
paper, it would appear that a significant response
did not occur at the lowest concentration tested. In
addition, the pollutant-exposed animals had
prolonged periods of inactivity. The precise time of
the ozone exposures that were related to the
observations was not given, but it appeared to be
less than 45 min. In another series of
experiments,172 subhuman primates exhibited an
elevation of both simple and choice reaction time
in atmospheres of 980/ug/m3(0,5 ppm) ozone after
only a 30-min pre-exposure to this concentration.
When Fletcher and Tappel68 continuously exposed
rats to 1960 pg/m3 (1 ppm) for 7 days to ozone,
they found an 84-percent reduction in voluntary
activity (p < 0.001). Rats exposed to 11 0 pg/m3
(0.056 ppm) ozone for 93 days exhibited no change
in behavior or chronaxial muscle ratios.53
Xintans et al.210 found a depression of the
evoked response to flash in the specific visual
cortex and in the superior colliculus of rats in
reponse to exposures of 980 to 1 960/ug/m3 (0.5 to
1 ppm) ozone for 1 hr. He suggested that such
changes reflect a possible morphological or
160
-------
functional change in the central nervous system.
Bokina et al.19 reported on animals (species not
identified) that were initially exposed to (time not
given) 3920 to 5880 fjg/m3 (2 to 3 ppm) ozone and
then subsequently exposed (time not given) to 30
/jg/m3 (0,015 ppm). When this exposure was
combined with a rhythmic, light stimulus,
paroxysmal activity in the olfactory analyzer
structures was provoked. Evoked potential of the
optic cortex was also investigated19 during a 1.5-
month continuous exposure to 50 pg/m3 (0.026
ppm) ozone. The following observations were
made: A decrease in the amplitude of the primary
response of the evoked potential, a slow negative
wave, and a decrease in the duration of-the slow
negative wave. The authors suggest that this
represents a deterioration of the cortical
processes. The method of ozone monitoring was
not given.
EXTRAPULMONARY MORPHOLOGY
Atwall and Wilson,7 using light and electron
microscopy, found histological and ultrastructural
changes in the parathyroid glands of rabbits after
4- to 8-hr exposures to 1470/^g/m3 (0.75 ppm)
ozone. The major changes found in exposed
animals were Jarge numbers of secretion granules,
hyperplasia of the chief cells, proliferation and
hypertrophy of the rough endoplasmic reticulum,
free ribosomes and mitochondria, lipid bodies, and
an accumulaton of secretion granules of the chief
cells inside the varcular endothelium of the blood
stream. Alteration; were observed 18 and 22 hr
after exposure, bu by 66 hr, the cells had normal
appearances. The authors hypothesize that this
damage could lead to alteration of parathormone
production and storage.
Brinkman et a! 20 exposed adult mice to 390
/jg/m3 (0.2 ppm) ozone 5 hr/day for 3 weeks.
Structural changes in the cell membranes and
nuclei of myocardial muscle fibers were produced
that were reversible about 1 month following
exposure. The physiological implications of these
changes were not discussed.
MISCELLANEOUS EFFECTS
Gardner et al.78 exposed mice to 1960/Aj/m3 (1
ppm) ozone for 3 hr/day up to 7 days. Immediately
after exposure, the animals were injected with
sodium pentobarbitaL and the induction time for
sleep and the actual sleeping time were
determined. The induction time was not affected
by ozone; however, sleeping time was altered.
After the second ozone exposure, there was a 13-
min prolongation in sleeping time (p< 0.05),
whereas the increase was 9.2 mm (p < 0.05) after
3 days of ozone treatment. No significant changes
were evident after a single exposure or after four or
more successive daily exposures. After seven daily
exposures to 1960 pg/m3 (1 ppm) ozone, the
concentration was increased to 9800 /yg/m3 (5
ppm) for 3 hr. This increased the sleeping time by
59.5 min (p < 0.001) and was taken to indicate that
tolerance did not occur. The authors hypothesized
that these effects may reflect alterations in
cytochrome P4so(cyt. P^so), a microsomal enzyme (a
mixed function oxidase) that oxidizes pento-
barbital. Goldstein and Balchum83 investigatei'the
possible role of this enzyme in ozone-induced
toxicity. They found that injections of pheno-
barbital (an inducer of cyt. P45o) decreased (p =
0.05) the survival times of rats exposed to a lethal
ozone concentration (157,680 /jg/m3, or 8 ppm),
whereas treatment with allylisopropylaceta-mide
(which destroys cyt. Ptsa) increased the survival
time of ozone-exposed animals. Goldstein et al.88
also showed that there was a decrease (p< 0.01) in
lung microsomal cyt. ?45o levels immediately, or 1,
2, 3, or 5 days after rabbits were exposed to 1 960
/yg/m3 (1 ppm) ozone for 90 min. (See earlier
section on other biochemical alterations.) Liver cyt.
P450 levels of the ozone-exposed rabbits were not
significantly different from control immediately or
1 day post exposure.
The effects of ozone on pulmonary arterial
pressure were investigated in 31 dogs exposed to
1960 /jg/m3 (1 ppm) for 17 months for varying
hours each day.14 Three dogs developed
pulmonary arterial hypertension, and 9 dogs had
excessive systolic pressure. Because there was no
proportional relationship between pulmonary
arterial hypertension and oxidant exposure, it was
suggested that the results seen were due to
genetic susceptibility.
Roth and Tansy178 exposed rats to 490 /jg/m3
(0.25 ppm) ozone for 2 hr and found no differences
in gastric secreto-motor activities, although a
temporary effect was present at higher levels.
Trams et al.199 found no lipid peroxidation in
brain tissue of dogs chronically (8 to 24 hr/day)
exposed to 1960 pg/m3 (1ppm) ozone for 18
months. Monoamine oxidase activity was
increased (p < 0.05) in animals exposed for 8 to 16
hr/day, but it was decreased in those dogs exposed
for 24/hr day. Catechol-o-methyltransferase
activity was decreased (p < 0.05) for all periods of
161
-------
exposure. There were no significant changes in 5 -
nucleotidase, acetylcholinesterase, and ATPase
activities or in the levels of catecholamines.
The functional state of the adrenal cortex was
investigated by Eglite.53 Rats were exposed to 110
fjg/m3 (0.056 ppm) continuously for 93 days. After
6 weeks, the levelsof urinary 1 7-ketosteroidswere
elevated and remained elevated during the course
of the experiment. The ascorbic acid content of the
adrenal decreased during exposure, but returned
to normal levels 15 days after ozone exposure
ceased. The method of ozone measurement wa°
not specified.
Quantitative analyses of urinary constituents of
rats exposed to ozone ranging from 1 568 to 2940
/jQ/m3 (0.8 to 1.5 ppm) 5 days/week for 18 weeks
were made by Hathaway and Terrill.98 The urine of
the exposed animals had lower(p<0.05)titratable
acidity on days 91, 98, and 112, and a higher
(p<0.05) pH on days 98 and 11 2, but not day 91. On
various days (days 3 to 84), there were no
significant changes in urinary creatinine, creatine,
uric acid/creatinine, and amino acid nitro-
gen/creatinine. For the first 53 days of the study,
caging arrangements partially hampered the
animals' ability to obtain food, and this may have
influenced the results.
There are suggestive data that exposure to ozone
may accelerate the aging process. Bjorksten and
Andrews13 and Bjorksten12 have presented the
concept that aging may be due to irreversible
cross-linking between macromolecules, prin-
cipally proteins and nucleic acids. They included
aldehydes in their list of active cross-linking.
agents. Aldehydes are potential cross-linking
agents and may be produced in the lung by ozone
exposure.
Stokinger194 reported accelerated or premature
aging in rabbits after 1 year of weekly 1-hr
exposures to ozone (no concentration given). His
evidence included premature calcification of the
sternocostal cartilage, appearance and coarseness
of the pelage (hairy system of the body), severe
depletion of body fat, and general signs of
senescence, such as dull corneas and sagging
conjunctivae. -It has been suggested that the
radiomimetic properties of ozone are implicated in
the effects on aging.
Summary and Conclusions
Animal toxicological studies in conjunction with
human clinical and epidemiological studies serve
to provide information on the health effects of
ozone. Investigations with animals are valuable in
that they both support and extend the information
derived from other experimental approaches. For
example, with animals it is possible to use long-
term exposures, controlled pollutant mixtures, and
invasive measurement techniques that cannot be
employed with humans because of ethical
considerations. A major area of concern in animal
studies involves the validity of extrapolating the
results to man. Since many physiological
mechanisms are common to animals and man, it
can be hypothesized that if ozone causes a
particular health effect in several animal species, it
is likely to cause similar effects in exposed
humans. But at this time, it is not possible to
predict the concentration of ozone that would be
responsible.
Interpretation of animal study results regarding
the severity of the effect and its implication to
human health is also difficult. Exposure to ozone
results in an array of effects that vary from
relatively small changes observed at low
concentrations after short-term exposure, to gross
alterations (including death) with higher con-
centrations and longer exposure times. A
statistically significant change observed after
exposure to ozone may be of unknown
physiological significance. However, it could be
hypothesized that any detectable deviation from
normal is potentially undesirable.
With the foregoing as background, the results of
the animal studies can be summarized. The weight
of the data from numerous laboratories indicates
that exposure to ozone results in alterations of host
defense mechanisms, pulmonary morphology, and
pulmonary function; biochemical changes; and
effects on circulating blood cells and sera. In mice,
a 3-hr exposure to concentrations of ozone as low
as 157 fjg/m3 (0.08 ppm) results in a
concentration-related increase in susceptibility to
infectious respiratory disease, as measured by an
increase in mortality from laboratory-induced
pneumonia. Other studies employing this model
system indicated that the presence of other
pollutants (nitrogen dioxide and sulfuric acid)
would have additive effects with ozone.
Investigations of host defense mechanisms have
indicated that a 2.5- to 4-hr exposure of ozone
depresses pulmonary bactericidal activity (1176
/jg/m3, or 0.6 ppm) and causes increased lability of
functional and biochemical alterations of alveolar
macrophages (196 to 980 A
-------
Numerous investigations in a variety of animal
species have shown that exposure to 196 to 1568
/yg/m3 (0,1 to 0.8 ppm) ozone from a few hours to
several days alters the activity of several enzymes
and other biochemical constituents (proteins and
lipids) in the lungs. At the 196-//g ozone/m3 (0,1 -
ppm) concentration, a 7-day continuous exposure
of rats maintained on a vitamin E level typical of
that in the standard American diet caused an
increase in oxygen consumption of pulmonary
tissue. A similar effect was not noted in vitamin-E-
supplemented rats until the concentration was
increased to 392 /;g/m3 (0,2 ppm). Additional
research has also shown that vitamin E can protect
the host from the adverse effects of ozone. In a
comparative study, rats were affected (increased
02 consumption of pulmonary tissue) by a lower
concentration (392 /yg/m3, or 0.2 ppm) of ozone
than monkeys (686 /yg/m3, or 0.35 ppm) following
a 7-day intermittent exposure. The activities of
other enzymes (e.g., glucose-6-phosphate
dehydrogenase, glutathione reductase, and
glutathione peroxidase) and concentrations of
glutathione also increase following ozone
exposure to levels equal to or greater than 392
//g/m3 (0.2 ppm) for several days (typical
experiments were for 7 days, but in some studies,
effects were noted earlier). Since these enzyme
systems can detoxify oxidizing substances such as
ozone and its reaction products, these changes are
interpreted as representing an increased
protection against ozone. Thus the increased
enzymatic activity might represent a reaction to
potential toxicity, rather than a toxic response
itself.
Many of the biochemical effects of ozone have
been attributed to concomitant morphological
alterations of the lung. Concentrations of ozone
equal to or greater than 392 //g/m3 (0.2 ppm) can
lead to an increase in the number of Type 2 cells
that replace Type 1 cells in both rats and monkeys.
Since Type 2 cells are rich in mitochondria and
have higher metabolic activity, this situation would
lead to increased activity of several enzyme
systems in the lung. In monkeys, the predominant
injury is located at the respiratory bronchiole (8
hr/day for 7 days). With increasing concentrations
of ozone, the extent of the damage expands
distally. In rats, which do not have respiratory
bronchioles, the principal site of damage was the
alveolar duct. The lesion generally reaches a peak
in 3 to 5 days (after continuous or intermittent
exposure), and then progressively diminishes,
even during continuing ozone exposure. The
meaning of these changes is not immediately
apparent, but it has been suggested that the
alterations are part of an adaptive response.
Morphological alterations caused by ozone can be
quite extensive and involve areas of the lung other
than the area of the respiratory bronchiole and
alveolar duct. For example, a 3-month continuous
exposure to 1725 fjg/m3 (0.88 ppm) can produce
emphysema-like structures in rats. Intermittent
exposure (6 hr/day, 5 days/week for 10 months) to
984 /;g/m3 (0.4 ppm) can produce emphyse-
matous-like lesions in rabbits.
Alterations in pulmonary function have been
observed in animals following short-term
exposures equal to or greater than 510 //g/m3
(0.26 ppm) ozone. Pulmonary flow resistance (a
measure of airway diameter) and respiratory rates
increased, and tidal volume decreased as the
concentration increased. At a higher con-
centration (1960 /;g/m3, or 1 ppm), compliance (a
measure of lung distensibility) and carbon-
monoxide-diffusing capacity (a measure of
alveolar gas exchange) decreased. Young rats
continuously exposed (30 days) to 392 /yg/m3 (0.2
ppm) were also affected, as indicated by increased
lung volumes and other changes that were
interpreted as a decrease in lung tissue elasticity.
Sensitive measurements of small airway disease,
such as those made in humans, are very difficult to
perform in small animals because of the size and
voluntary maneuvers of the subject that are
required. Therefore it would be expected that many
of the slight-to-moderate morphological
alterations typically seen in animals would not be
detected by pulmonary function measurements.
The health effects of ozone are affected by other
environmental factors such as dietary levels of
vitamin E and the presence of other pollutants that
have already been mentioned. The potential of
ozone to modify the transformation of other
environmental contaminants or drugs is suggested
by the results of several studies that show that 1.5-
to 3-hr exposures to 1470 to 1960 /yg/m3 (0.75 to
1,0 ppm) ozone decrease the concentration of
mixed-function oxidases of pulmonary tissue. In
another experiment, a 2-day exposure (3 hr/day) to
1960 /yg/m3 (1 ppm) ozone was interpreted as
having a similar effect, perhaps on the liver. Lower
concentrations were not tested. Nonetheless, such
data would lead to the inference that by lowering
the concentration of mixed-function oxidases that
either detoxify or activate chemical compounds
163
-------
(depending on the compound), ozone would
influence the toxicity of other environmental
chemicals that are substrates for this enzyme
system. The net result depends on whether the
toxicity of the chemical would be enhanced or
decreased by mixed-function oxidase activity.
A considerable number of research studies have
addressed tolerance to ozone, a phenomenon by
which a prior dose of ozone will protect an animal
from a later higher (or equivalent) dose of ozone.
Though such work is interesting, its relevance to
the protection of health is difficult to assess, since
concentrations of ozone found to induce tolerance
also elicit alterations in biological endpoints. In
addition, tolerance is only highly efficient against
edema and results in partial or no protection
against some of the alterations in pulmonary bio-
chemistry, morphology, and host defense
mechanisms against infectious pulmonary
disease.
In some studies of host-defense mechanisms,
biochemistry, and morphology, concentration-
response relationships were investigated. As the
concentration of ozone was increased, the effect
also increased. Also, in most cases in which
different exposure modes were compared, there
were no significant differences between
intermittent (7 to 8 hr/day) and continuous
exposure. From such information, it can be
postulated that insufficient recovery occurs during
the clean-air period of the intermittent exposure
mode.
The persistence of the health effects caused by
ozone has only been measured in a few studies and
varies considerably. Some pulmonary biochemical
alterations exhibit a plateau or continue to
increase during exposure and then subside
immediately or a few days following cessation of
exposure Other enzyme activities remain elevated
a few weeks after exposure. The typical
morphological lesion (in which Type 2 cells replace
Type 1 cells) reverts toward normal during ozone
exposure. However, more severe pathological
effects such as emphysema would remain
throughout the life of the animal.
Though most ozone toxicological studies have
focused on the lung, systemic effects have also
been observed. Extrapulmonary effects might be a
result of neurohormonal mechanisms initiated by
ozone or of the toxic action of the reaction products
of ozone affecting blood components as they pass
through the capillaries surrounding the alveoli; or
these reaction products could possibly enter the
circulation, exerting effects at distant target sites.
Also conceivable is that ozone itself cou Id cross the
alveoli and affect blooa components, though such
an event is unlikely, since mathematical
predictions estimate that only a miniscule dose
would reach the level of the alveoli. Whatever the
reason(s), extrapulmonary effects have been noted
at concentrations as low as 392 /ng/m3 (0.2 ppm)
for a 30-min exposure.
Numerous alterations in erythrocytes and in the
sera have been observed following ozone
exposure. Increased lysis of red blood cells of
exposed (980 /xj/m3, or 0.5 ppm for 23 days,
continuous) vitamin-E-deficient-animals has been
reported. An enzyme bound to red cell membranes
is also affected by ozone. Serum protein changes
have also been shown to occur in animals exposed
to ozone levels between 392 and 1 568 fjg/m3 (0.2
and 0.8 ppm). Levels of albumin, globulins, and
serum enzymes were altered after a 6- to 10-
month intermittent exposure to 784 /yg/m3 (0.4
ppm). Exposure to 392 j/g/m3 (0.2 ppm) for 1 hr
reduced total blood serotonin, and a 3-week
intermittent exposure to this ozone level increased
serum glutamine pyruvic transaminase.
The influence of ozone on chromosomes has
also been investigated and is the subject of some
controversy. In one study, Chinese hamsters
exposed to 392 /^g/m3 (0.2 ppm) ozone for 5 hr
exhibited an increased number of chromosomal
breaks in circulating lymphocytes, which were
detectable up to 1 5 days following the end of the
ozone exposure. Another investigation using
approximately the same ozone exposure failed to
demonstrate any chromosomal aberrations in
Chinese hamster bone marrow, mouse peripheral
leukocytes, and mouse primary spermatocytes. A
number of investigations have shown that
bacteria, plant cells, oyster cells, and chick
fibroblasts exposed to ozone in vitro exhibited
chromosomal alterations. Because of these
conflicting data, no definitive conclusions can be
drawn at present regarding the effects of ozone on
chromosomes. Even if the work with circulating
lymphocytes were replicated in the future, the
biological significance of the findings (beyond that
of indicating systemic chromosomal damage from
an environmental agent) would be difficult to
assess Lymphocytes are considered to be a
terminal cell; but for proper immune functioning,
they must divide and form an additional line of
cells Consequently, immune functioning could be
affected by ozone. However, a potential danger to
164
-------
human health—that of chomosomal aberrations of
germ cells (i.e., ova or sperm)—has not been
directly observed. In spite of the conflicting results
of ozone with chromosomes, mutagenic effects
have been observed. The offspring of mice exposed
for 7 hr/day, 5 days/week to 392 //g/m3 (0.2 ppm)
during gestation and the first 3 weeks of life
exhibited increased neonatal deaths and
mutagenic effects such as increased incisor
growth and blepharophimosis (a narrowing of the
slit between the eyelids).
A number of other extrapulmonary effects have
also been reported. Structural changes have been
found in mouse heart muscle following a 3-week(5
hr/day) exposure to 392 //g/m3 (0.2 ppm) and in
the parathyroid glands of rabbits after a 4- to 8-hr
exposure to 1470 //g/m3 (0.75 ppm). Short-term (1
hr or less) exposures of greater than or equal to
980 //g/m3 (0.5 ppm) have caused alterations in
behavior and in responses of the central nervous
system (i.e., decreased activity and depression of
evoked reponse to flash).
Review of the literature of the health effects of
ozone provides strong evidence that in animals, a
large variety of pulmonary and extrapulmonary
alterations - are produced after a short-term
exposure to concentrations between 392 and 980
//g/m3 (0.2 aad 0.5 ppm). Pulmonary biochemical
alterations and an increased susceptibility to
infectious disease have been observed at levels
between 1 57 and 392 //g/m3 (0.08 and 0.02 ppm).
A prudent conclusion from the animal toxicological
data is that humans exposed to ozone may
experience similar effects, but it is not possible to
predict precisely the severity of the effects or the
concentrations at which they may occur in man.
EFFECTS OF PHOTOCHEMICAL OXIDANTS
ON EXPERIMENTAL ANIMALS
Experimental Data
Investigations have been conducted of the
potential biological actions of a complex photo-
chemical reaction mixture produced by irradiating
mixtures of air and auto exhaust under laboratory
conditions that simulated real driving patterns and
solar irradiation. 1°°.101.104.142.144 The effects of both
nonirradiated and irradiated exhaust mixtures
were studied. Clearly, irradiation of the air-exhaust
mixture led to the formation of photochemical
reaction products that were biologically more
active. The relative proportions of the suspect
biologically active chemical species varied with the
total concentration of exhaust gases in the
irradiated mixture and with the duration of
irradiation.
142,144
Single-inhalation exposure
studies lasting a few hours demonstrated that
irradiation of exhaust mixtures led to greater
effects on respiratory mechanics (increased flow
resistance and tidal volume and decreased
breathing frequency) in guinea pigs, greater
reduction in voluntary running activity of mice,
increased susceptibility to infection, and slightly
greater carboxyhemoglobin formation in rats,
compared with animals exposed to the same total
concentration of exhaust gases that were not
irradiated. The concentration of total oxidant as
expressed by ozone in these experiments ranged
between 588 and 1568//g/m3 (0.30 and0.80ppm)
in the irradiated-exhaust mixtures. Only a trace or
no oxidant was detected in the unirradiated
exhaust. The irritant aldehydes, formaldehyde and
acrolein, were also present in higher
concentrations (0.39 to 2.42 ppm and 0.09 to 0.2
ppm, respectively) in the irradiated atmospheres.
Formaldehyde concentrations were between 0.12
ppm and 0.38 ppm, and acrolein ranged from 0.02
to 0.07 ppm in the unirradiated exhaust chamber.
The effects that were noted were reversible within
a few hours when the animals returned to clean
air. The effects in animals exposed to the
irradiated-exhaust mixture are not necessarily
uniquely characteristic of ozone, but most of them
could have been produced by ozone.
Murphy et al.142'144 found that the nature of
changes in respiratory mechanics in guinea pigs
exposed to irradiated exhaust varied according to
the ratio of oxidant to aldehyde concentrations
(formaldehyde and acrolein were measured), and
this ratio in turn varied with the duration of
irradiation of the air-exhaust mixture. Thus when
the oxidant:aldehyde ratio was low, the guinea pig
respiration pattern resembled that reported for
animals exposed to irritants such as formaldehyde
and acrolein6'143 and was characterized by in-
creased pulmonary flow resistance, increasedtidal
volume, and decreased frequency of breathing.
Increasing the ratio resulted in a shift in the pattern
of respiration toward that produced by deep-lung
irritants (e.g., ozone and nitrogen dioxide146),
namely decreased tidal volume and increased
frequency, although the increased flow resistance
typical of the aldehyde effect persisted. This
interactive effect of an oxidant-aldehyde mixture
could be reproduced by a simple mixture of ozone
and acrolein.142
165
-------
TABLE 8-1. PULMONARY EFFECTS OF OZONE: HOST DEFENSE MECHANISMS
_— - -
Ozone,
^a/m3
1960
1960
1764
1568
1372-
1764
1372
1176
980
980
980
980
980
980
785
784
784
490
196
196
157
196
157
Ozone,
ppm
1
1
0&
0.8
07-
0.9
07
06
05
0.5
0.5
0.5
05
0.5
04
0.4
0.5
0.25
0.1
0.1
0.08
0.1
0.08
Length of
exposure
17 hr before
bacteria
3 hr
1-4 hr
11 days
3 hr
7 days
4 hr after
bacteria
3 hr
16 hr/day *
7 months
3 hr
2 hr
3 hr
3 hr
3 hr before
bacteria
4 hr
17 hr before
bacteria or 4 hr
after bacteria
3 hr
2.5 hr in vivo
or 30 mtn
in vitro
3 hr
3 hr
3 hr
Observed effect(s)
Decrease in bacterial pulmonary depo-
sition, decrease in bactericidal
activity
Bacteria observed in the blood
sooner and with increased frequency
Increased nasal, but not lung, deposi-
tion and growth of virus Decreased
minute ventilation
Decreased production of interferon bv
tracheal epithelial cells
Increased mortality of mice pre-exposed
over those animals not so treated
Aerosols of S pyogenes received by
all animals immediately after Oj
exposure
Deficiency of vitamin E further
reduced bactericidal activity after
7 days
No effect on bacterial deposition
and mucociliary clearance
Decreased bactericidal activity
Decreased enzyme activity
in alveolar macrophages, increase
in number of pulmonary
polymorphonuclear leukocytes
(appears to be linearly related to
dose)
No effect on clearance of polystyrene
and iron particles
Increased fragility of alveolar
macrophages
Decrease in agglutination of alveolar
macrophage, indicating membrane
alterations
Increased red blood cell rosette
formation by lectm-treated alveolar
macrophages
Reduction in phagocytosis of alveolar
macrophage
Lower deposition, but subsequently
a higher number of bacteria present
due to reproduction.
Bactericidal activity inhibited by O3,
but no role played by an induced
silicotic condition
Physical clearance not affected.
but bactericidal activity affected
No synergism noted with NO/3
Lysozyme, acid phosphatase, and &-
glucuronidase activities of alveolar
macrophages reduced
Lung protective factor partially
inactivated (Appears to be
dose-related )
Increased mortality when S pyogenes
aerosol challenge received imme-
diately after exposure Additive effect
from simultaneous exposure to
>3760 jjg/m3 (2 ppm) NO2 and
>98 jjg/m3 (0 05 ppm) 03
Significant increase in mortality of
mice exposed to aerosols of S.
pyogenes during ozone exposure
Increased mortality in mice chal-
lenged with aerosols of S pyogenes
immediately after exposure
Species
Mice
Mice
Mice
Mice
Mice
Rats
Mice
Rabbits
Rabbits
Rabbits
Rats
Rabbits
Rabbits
Mice
Mice
Mice
Rabbits
Rabbus
Mice
Mice
Mice
References
Goldstein et al 9°
Coffin and Gardner36
Fairchild61'62
Ibrahim et al 10!>
Coffin and Blommer3'1
Warshauer etal203
Goldstein et al a9'90
Alpert et al 3
Fnberg et al 73
Dowell et al 4a
Goldstein et al.es
Hadley et al M
Coffin and Gardner36
Coffin and Gardner36
Goldstein et al fi"
Goldstein et al 91
Hurst et al <06
Gardner76
Ehrhch et al 64
Miller et al 135
Coffin et al.36
166
-------
TABLE 8-2. PULMONARY EFFECTS OF OZONE: BIOCHEMISTRY
Ozone,
tig/m1
1960
1960
Ozone,
ppm
1
1
Length of
exposure
1 hr
Continuous
Observed effect(s)
Carbonyl compounds found
50% of vitamm-E-depleted group
Species
Rabbits
Rats
Reference
Bueli et al.23
Roehm et al.'74
Mustafa and Lee150
died in 8 2 days, 50% of vitamm-E-
supplemented group died in 1 8 5
days
1960 1 90 mm Decreased lung microsomal Rabbits Goldstein et a IB8
cytochrome Paso levels after 1,2,3, or
5 days The greater depression at 3 6
days following exposure.
1568 0.8 Continuous Decreased protein synthesis, day 1, Rats Mustafa et ai t49
7 days increased protein synthesis, days 2
and 3 (remained elevated and
unchanged)
1568 08 Continuous Increased rates of collagen and non- Rats Hussam et al.t07
7 days collagenous protein synthesis
1568 08 Continuous Increase in protein synthesis, non- Rats Chow et al29
3 days protein sulfhydryl content and activ-
ities of GSH peroxidase, GSH
reductase and G-6-PD Complete
recovery 9 days after exposure
ceased Increased rate of
mitochondnal succmate oxidation
returning to normal levels 9 days
after the original exposure ceased
1568 0.8 Continuous Increased rates of O? consumption. Rats
1-30 days reaching a peak at day 4 and
remaining at a plateau for the
remainder of the 30 days Most
increase in the oxidation of succmate
Also an initial decrease (day 1) and a
subsequent increase (day 2) in the
activity of succmate-cytochrome-C
reductase activity which plateaued
between days 3 and 7
1568 0.8 Continuous Increased rate of glucose consumption, Rats
4-7 days pyruvate and lactate production after
4 days of exposure Increased MAO
activity after 7 days of exposure
1568 0.8 7 days Increased activities of hexose mono- Rats
phosphate shunt and glycolytic
enzymes of lu ng
1568 08 10 days Decreased cytochrome-C-reductase Rats
1960 1 activity, increased G-6-PD activity,
no change in sulfhydryl levels
1470 075 3 hr Reduction in activity of benzpyrene
hydroxylase in both tracheobronchial
mucosa and lung parenchyma.
1470 0.75 Continuous Decreased activities of GSH peroxidase. Rats
30 days GSH reductase, G-6-PD, 6-P-GD, and
pyruvate kinase on day 1 Thereafter,
increase in most of these enzyme
activities until day 10, at which time,
beginning of a slight decrease At day
30, still elevated over control
1372 07 Continuous Increased formation of malonaldehyde Rats Chow and Tappel3
1568 08 7 days Activities of GSH peroxidase and
G-6-PD partially inhibited as a
logarithmic function of dietary
vitamin E Increased activity of GSH
reductase not affected by vitamin E
1372 07 Continuous Increase in lysosomal hydrolases in Rats Castieman et al35
1568 08 5 days homogenates and tissue sections Dillard et al46
7 days Increase in protease and peptidase
seen along with other enzymatic
changes Increased activity of acid
phosphatase in alveolar macro-
phages, terminal airway, epithelium,
and adjacent structures
Mustafa et al 149
Chow and Tappel31
Deluciaetal.44-'15
• Hamsters Palmer et al.160'161
Chow and Tappel3'
167
-------
TABLE 8-2. PULMONARY EFFECTS OF OZONE: BIOCHEMISTRY (cont'd).
980
1960
980
1960
980
1960
980
980
1568
980
1568
392
686
980
1568
392
980
1568
Ozone,
ppm
05
1
05
1
05
1
05
05
08
05
08
02
035
05
08
0.2
05
08
Length of
exposure
Continuous
9 days
4 hr
Continuous
1-2 weeks
Continuous
9 days
8 hr/day
7 days
8 hr/day
7 days
Continuous
7 days
8 hr/day
7 days
Continuous
(7-8 days)
or intermittent
(8 hr/day
for 7 days)
Observed effect(s)
Specses
Reference
Alterations of lung tissue lipids Rats
Greatest change an increase in
arachidonic acid, which occurred to a
greater extent in vitamm-E-deficient
rats. Decreases in Imolenic, oleic,
stearic, and palmitic acid
Release of surfactant lecithins and Rabbits
decreased lecithin formation
Vitamin E acts as an antioxtdant and Rats
protects against some ozone altera-
tions (mortality, high increases
m arachidonic acid content, and
hpid peroxidation).
Greater increases in the activities Rats,
of GSH peroxidase, GSH reductase, monkeys
and G-6-PD exhibited by rats Only
slight increases by monkeys in these
enzyme activities.
Increased succinate oxidase activity
Increases not significant at lower
concentrations
Increased prolyl hydroxylase activity Rats
No change at lower concentrations
At 1568^g/m3, partial return of
activity to normal after a 30-day
recovery, but still elevated Increased
hydroxyprolme after 3 days of
exposure to 1568Aig/m3, remaining
elevated after 28 days of recovery
Increased activities of GSH peroxidase, Mopkeys
GSH reductase, G-6-PD, NADPH-
cytochrome-C reductase, succinate
oxidase, acid phosphatase, and B-N-
acetyl-glucosaminidase A significant
correlation found between ozone
concentration and increased enzyme
activities
For the continuous exposure to the Rats
two higher concentrations, increased
activities GSH peroxidase, GSH
reductase, and G-6-PD At the lower
concentration (continuous), increased
activities of GSH peroxidase and GSH
reductase A linear increase in all
three enzyme activities as the con-
centration of ozone was increased
Increased Oz consumption (using
succmate-cytochrome-C reductase
activity fairly proportional to ozone
concentration Similar results
obtained for intermittent exposure
groups
Menzel et al.132
Kyei-Aboagye et al 1"
Setp et a I 1Be
Fletcher and Tappel e8
Roehm et al "^7S
Shakman188
Menzel et al '«
Chow et al 30
Monkeys Mustafa and Lee150
Hussam et al 10°
Dungworth et al.50
Mustafa and Lee160
Chow et al 28
Mustafa and Lee150
168
-------
TABLE 8-2 PULMONARY EFFECTS OF OZONE: BIOCHEMISTRY (cont'd).
Ozone,
ug/m
392
980
1568
392
980
1568
Ozone,
ppm
0.2
05
08
02
05
08
Length of
exposure
Continuous
7 days
Continuous
Intermittent
(8 hr/dav)
7 days
Observed effects) Species Reference
Increased activities of NADPH- Rats Mustafa and Lee150
cvtochrome-C reductase and G-6-PD
Activities of G-6-PD and NADPH- Rats Schwartz et al l8"
cytochrome-C reductase and
succmate oxidase increased in an
ozone-dose-dependent fashion
-392
1568
392
1568
392
980
1568
392
980
1568
196
392
02
08
0.2
0.8
0.2
05
08
0.2
t)5
08
0.1
02
Continuous
10-20 days
7 days
8 hr/dav or
continuous
2-90 days
Continuous
7 days
Continuous
(8 days)
intermittent
(8 hr/day,
7 days)
Continuous
7 days
following exposure No significant
differences found between the
intermittent and continuous exposure
groups
At the higher concentration Increase Rats
in the lung mitochondnal ozone
consumption in oxidation of 2-
oxyglutarate and glycerol-1-phos-
phate and the number of Type 2
alveolar cells that are rich in
mitochondria At the lower con-
centration Increase in 02
consumption.
Decreased glycoprotem secretion by Rats
tracheal explants
Increased superoxide dismutase Rats
activity as ozone concentration
increased
Increased lysozyme activities at 1568 Rats
^/g/m3 (continuous) No significant
changes following intermittent
exposure.
Significant increase in succmate Rats
oxidase activity at 196 fjg/m3 in
those animals maintained on 1 1 ppm
vitamin E Increased succmate
oxidase activity only at 392 ^g/m3 for
those rats receiving 66 ppm vitamin
Mustafa et a I
Last et al "8
Mustafa et al '5
Chow et al
Mustafa and Lee150
All the functional effects observed in short-
duration experiments with laboratory-produced
photochemical smog mixtures could have been
due to ozone alone if one considers total oxidant
concentraion of the mixture equivalent to ozone
concentration. One possible exception was the
increase in respiratory flow resistance in guinea
pigs, which is more characteristic of an irritant
aldehyde. The increase in respiratory frequency in
guinea pigs is most probably due to the oxidant (or
ozone) content of the mixture.
Comparison of the concentrations for equal
effectiveness in decreasing the spontaneous
running activity in mice also suggests that this
action of the mixture may be largely due to oxidant.
Likewise, the oxidant content of irradiated auto
exhaust appears to explain adequately the
increase in susceptibility to respiratory infection in
mice exposed to the mixture. These conclusions
must be qualified, since the possibility exists that
other chemical species, such as free hydroxyl
radicals, which were not measured, might also
have produced a similar effect. Nevertheless, it is
reasonable to conclude that many of the effects
produced by exposures to complex photochemical
oxidant mixtures are due to ozone.
Hueter et al.104 exposed animals to irradiated
automobile exhaust for periods of 6 weeks to 23
months. The concentrations were cycled to
simulate daily pollution concentrations in urban
cities. Daily peak concentrations of carbon
monoxide of 20, 50, 60, and 100 ppm were
169
-------
TABLE 8-3. PULMONARY EFFECTS OF OZONE: MORPHOLOGY
Ozone,
fjg/m3
1960-
510
1960
1568
1058
1725
1176
980
1568
980
980
1568
784
686-
980
588
392
392,
686
980,
1568
Ozone,
ppm
1-026
1
08
054
088
06
05
08
05
05
08
0.4
035-
05
03
02
02,
035
05,
08
Length of
exposure
4-6 hr
8 hr/day
Continuous
3-5 months
Continuous
6 months
6-7 hr/day
1-2 days
8 hr/day
7 days
6 hr
2,4,6,8 or
24 hr/day
for 7 days
6 hr/day
5 days/week
10 months
Continuous
8 days
3 hr/day
1 6 days
2 hr
8 hr/day
for 7 days
Observed effect(s)
Loss of ciliated epithelium, damage
to ciliated cells and mitochondria
in some airways, swelling and des-
quamation of Type 1 alveolar cells.
and swelling and breakage of inter-
alveolar capillaries Apparently
dose-related.
Damage roughly proportional to time
and dose of exposure, fibrous
elements deposited, lumina of small
airways reduced, metaplasia of
columnar and cuboidal epithelium.
Thickening of alveolar septa and
increase m number of alveolar
macrophages Emphysematous and
fibrotic changes.
Major site of injury at the junction
of the respiratory bronchiole and the
alveolar duct A multitude of altera-
tions observed at the lower
concentration. After 3 weeks,
terminal bronchioles seemed to
return to normal appearance In
general, at the higher concentrations.
pathological changes similar but
more extensive and occurred earlier
After 3 months, emphysema-like
structures seen.
Younger mice more sensitive than older
animals Swelling of epithelial
alveolar lining cells and endothehum
cells. Occasional breaks in basement
membrane
Lesions in centnacinar region, hyper-
plasia of nonciliated epithelial
cells and mtralummal accumulation
of macrophages. Replacement of
Type 1 with Type 2 cells
Lowered rates of DNA synthesis in
alveolar cells.
Accumulation of inflammatory cells in
centnacinar region of lung
Emphysematous and vascular-type
lesions observed in the lung, and
small pulmonary arteries thicker.
Type 1 cells replaced by Type 2
alveolar cells. No further tissue
damage after 4th day
Presence of hemispheric extrusions of
cilia, and small round bodies on
surface of airways. Changes in
Clara cells.
Type 1 cells replaced by Type 2
alveolar cells after 24 hr
Pulmonary lesions occurring in
respiratory bronchioles Hyperplasia
and hypertrophy of bronchiolar
epithelium. Increase in Type 2 cells
Damage to ciliated and Clara cells.
Rats and monkeys equally
susceptible
Species
Cats
Dogs
Rats
Rats
Mice
Rhesus
monkeys.
mice
Aging
mice
Rats
Rabbits
Rats
Rats
Rats
Rhesus
and
Bonnet
monkeys,
rats, mice
References
Boatman et al 1516
Freeman et al 72
Stephens et al 192
Freeman et al 70
Bils"
Mellick et al.126
Evans et al 69
Brummer et al 21
P'an et al.162
Evans et al 57
Sato et al.180
Stephens et al.'90
Evans et al.5860
Mellick et al 126
Dungworth49
Schwartz183
Schwartz et al.184
170
-------
established in four sets of exposure chambers.
There was considerable loss of ozone and nitrogen
dioxide on chamber walls, cages, and animal fur,
so the concentrations of chemically reactive gases
to which the animals were actually exposed
probably ranged from about 78 to 392 /yg/m3 (0.04
to 0.2 ppm) for ozone and about 280 to 940/yg/m3
(0.15 to 0.5 ppm) for nitrogen dioxide. Inter-
pretation of these studies is difficult because of
loss of contaminants on the chamber surface. No
significant treatment effects were observed when
pulmonary flow resistance, tidal volume,
respiratory frequency, and oxygen consumption
were measured in guinea pigs, mice, or rats at 16-
week intervals during the chronic exposures to
exhaust. Exhaust-exposed mice showed a
decrease in running activity for the first few weeks
of exposure but then recovered to attain control
levels. Decreases in mouse fertility rate and infant
survival rate occurred in the exhaust chambers.
This effect was confirmed in a second
experiment.119 Also, there was an increase in the
rate of spontaneous pulmonary infection in ex-
haust-exposed animals. There were no significant
effects of exhaust exposure on mortality,
histopathology, growth rate, or hematologic
indices.
Wayne and Chambers205 reviewed studies on
experimental animals exposed throughout their
lifetimes to ambient Los Angeles atmosphere.
Control animals were kept in rooms that were
ventilated with special filters that removed most of
the ambient air pollutants. The following maximal
peak concentrations were recorded (ranges of the
four exposure stations are given): 29 to 72 ppm CO,
50 to 1 21 pphm NO, 49 to 73 pphm N02, and 46 to
82 pphm oxidant. No clear evidence of chronic
injury from the ambient air pollution was observed.
However, there was suggestive evidence from
pulmonary function tests, electron-microscopic
examinations, and the incidence of pulmonary
adenomas that aged animals had been adversely
affected by ambient smog and that some reversible
changes in pulmonary function of guinea pigs
were noted during periods of peak air pollution.
Increased 1 7-ketosteroid excretion suggested that
breathing polluted ambient air was stressful for
guinea pigs. The reported effects were marginal,
and some of them may have been dueto variations
in temperature and humidity.
Beagles were exposed
120,200
to various pollutant
mixes for 1 6 hr/day, 7 days/week for 61 months,
and measurements of cardiovascular parameters
and pulmonary function were made periodically.
TABLE 8-4. PULMONARY EFFECTS OF OZONE: PULMONARY FUNCTION
Ozone,
jyg/m3
1960
1960
882
510
980
1960
980
392
666-
2646
Ozone,
ppm
1
1
045
026
05
1 0
05
02
034-
1 35
Length of
exposure
3 hr
6 hr/day
3-4 days
6 hr/day
6 days/week
6-7 weeks
2 hr
6 5 hr
2 hr
Continuous
30 days
2hr
Observed effect(s)
1-3 days post exposure, reduced
vital capacity, 7 days post exposure,
vital capacity reduced only slightly
A decrease in lung elasticity back to
near normal conditions 3 days post
exposure
Increased residual volume/total lung
capacity, functional residual
capacity/total lung capacity, and total
lung resistance; decreased chest wall
resistance No change in flow volume
curves
No change in pulmonary pressure-
volume relationship or ratio of
collagen to elastin
Increased pulmonary flow resistance
with increasing 03 levels. Vital
capacity not affected. Reduced
diffusion capacity shown by some
cats
Increased air current resistance
and frequency of respiration,
decreased tidal volume
Increase in lung volume and alveolar
dimensions and a reduction in lung
elasticity
Increased respiratory frequency and
decreased tidal volume
Species
Rabbits
" Rabbits
Rats
Cats
Guinea
pigs
Young
rats
Guinea
pigs
References
Yokoyama212'213
Yokoyama214
Yokoyama and
Ichikawa216
Watanabe et al 204
Yokoyama211
Bartlett et al.9
Murphy et al 146
171
-------
TABLE 8-5, PULMONARY EFFECTS OF OZONE: EDEMA AND TOLERANCE
Ozone,
490
588
588
Ozone,
Ozone, $jg/m3
ppm Length after
pre- of p?e- literu
exposure eMposure period
025
03
0.3
688-
9BO
980
980
980
03-
05
05
05
0.5
1470
1960
1960
6 hr
1 hr
3 hr
4 days
6 hr
3 hr
3 hr
ppm
aftef
latent
period
1966
39,200
588
980, 0 5,0 7,
1372, 1 0
1960
43,120
5880 or 3 and
43,120 22
075 3 days
7840
1 hr
1 hr
3920
exposure
alter
latent
period
Observed efforts)
Species
1 6 hr No tolerance to edema
unless pretreated with
methylpredmsoione,
20 2 hr Tolerance not developed
by thymectomized
animals, but developed by
sham-operated animals,
0.3 3 hr 20% lower mortality for
pre-exposed mice than
mice receiving only one
Oa dose. Partial toler-
ance probably due to
inhibition of edema-
genesis.
1,2,4 Lack of total protection
days indicated by increased
numbers of Type 2 celis
Edema as measured by re-
covery of 132I in pulmonary
lavage fluid.
22 3 hr Using unilateral lung
exposure technique,
tolerance to edema a local
effect and seen only in the
pre-exposed lung,
3 hr Using unilateral lung
exposure technique,
tolerance developed only
to pulmonary edema. No
tolerance to the_chemo-
taxis of polymorpho-
nuclaar leukocytes or
decreased lysosomal
hydrolase enzyme
activity,
40 8 hr A smaller decrease m acti-
vittttes of glutathione
peroxidase, glutathione
reductase, glucose-6-
phosphate dehydro-
genase and levels of re-
duce^d glutathione in
lungs of tolerant animals
as compound to
nontolerant animals.
Ail animals X-irradiated
to 800 R. 60% of Cb-pre-
exposed mice survived.
100% of controls died.
2 1 hr Tolerance to allergic
response to inhaled
acetylcholme.
Rats
Mice
Mice
Rats
Hats
Rabbits
Alpert et als
Gregory et al!
Coffin and
Gardner37
Evans et al"
Atpert et al5
Alpert et al,4
Rabbits Gardner et al,79
Hats
Chow27
Mice
Guinea
pigs
Hatton et al"
Matsumura
etal '»
172
-------
Treatment groups are shown in Table 8-10. There
were no specific abnormalities in cardiovascular
function attributable to air pollution exposure
during the exposure period. Pulmonary function
measurements of the beagles were made after
18,200 36,120 and 61120 months of exposure. After
18 months, there were no differences in CO
diffusing capacity(DLco), dynamic lung compliance
(Ctdyn), or total expiratory pulmonary resistance.
After 36 months of exposure, no statistically
significant effects were observed, although there
was an increased frequency of abnormal
measurements in those animals (Group 7}
receiving nitrogen oxides. By 61 months, more
changes were evident, but only those related to
oxidant exposure will be described here. Dogs
exposed to irradiated auto exhaust (Group 3) had a
higher (p < 0.05) mean single breath nitrogen
washout than control animals. When the data
were analyzed differently based on the number of
animals showing alterations, it was found that
more animals of groups 3 and 6 had higher (p <
0.0001) total expiratory resistances than the
comparable control animals in groups 1 and 4.
Pulmonary biochemical measurements, made
by Qrthoefer et al.158 on these dogs after a 21/2-year
TABLE 8-6. EXTRAPULMONARY EFFECTS OF OZONE: HEMATOLOGY AND SERUM CHEMISTRY
Ozone,
rt/m1
1960
1960
1686
1568
980
Ozone,
ppm
1
1
0.86
0.8
0.5
Length of
exposure
30 mm
30 mm
8 hr/dav
5 days/week
6 months
Continuous
29 days
Continuous
23 days
Observed effect(S)
Increased resistance to erythrocyte
hemolysis.
Increased resistance to erythrocyte
hemolysis.
Increased infestation and mortality
after infection with P berghei.
Increased acid resistance of
erythrocytes.
Increased lysozyme activity by day 3
Increased hemolysis of erythrocytes
of animals depleted of vitamin E N
980
784
784
784
784
392
392
392
392
294
451
1960
110
0.5
0.4
0,4
04
0.4
0.2
0-2
0.2
0.2
0.15
0.23
1.00
0056
8 hr/day
7 days
6 hr/day
,T 5 days/week
10 months
6 hr/day
5 days/week
6 months
6 hr/day
5 days/week
10 months
10 months
8 hr/day
5 days/week
3 weeks
60 min
2hr
5 hr
5 hr
5 hr
2 hr
93 days
Soecies
Mice
Mice
Mice
Rats
Rats
such change when rats received
vitamin E supplements.
No change in level of GSH or activities Monkeys,
of GSH peroxidase, GSH reductase, rats
or G-6-PD in erythrocytes.
Decreased serum albumin concen- Rabbits
tration; increased concentration of
or and y globulins, not much change in
0 globulin; no change in total serum
proteins.
No change in serum trypsin inhibitor Rabbits
capacity.
Increase in serum protein esterase. Rabbits
Increase in serurn protein esterase Rabbits
Increase in serum glutamic pyruvic Murine
transammase, and hepatic ascorbic
acid. No change in blood catalase
Small decrease m total blood Rabbits
serotinm.
Bmucleated lymphocytes in Murine
blood doubled in number
after exposure.
Circulating blood lymphocytes had
chromosome breaks up to 2 weeks
post Os. Additive effect with radiation
exposure.
No increase in chromosomal aber- Mice.
rations. hamsters
Decrease in whole blood Rats
eholinesterase, which returned to
normal 12 days after exposure
ceased.
Reference
Mizoguchi et al138
Chnstensen and Giese33
Schhpkoter and Bruch182
Chow et al.28
Menzel et a! 132
Chow et al30
P'an and Jegier164
P'an and Jegier166
Jegier'10
P'an and Jegier'66
Venmga201
Veninga202
Veninga202
Hamsters Zelac et al'
Gooch et al.92
Eglite53
173
-------
TABLE 8-7. EXTRAPULMONARY EFFECTS OF OZONE: CENTRAL NERVOUS SYSTEM AND BEHAVIOR
Ozone,
^g/m3
1960
1176
980-
1960
980
118
110
Ozone, Length of
ppm exposure
1 Continuous
7 days
0.6 30 mm
0.5-1 .1 hr
0.5 30 mm
006 —
0056 Continuous
93 days
Observed effect(s) Species
Reduction in voluntary activity
Avoidance response to Oa.
Depression of evoked response to
flash
Elevation of simple and choice re-
active time.
Depression of gross motor activity
as Oa concentration increased.
No behavior change No change in
chronaxial ratios in muscles
TABLE 8-8. EXTRAPULMONARY EFFECTS OF OZONE
Ozone,
fjg/m3
1470
392
Ozone, Length of
ppm exposure
0.75 4-8 hr
0.2 5 hr/day
3 weeks
Observed effect(s)
Morphological alterations of
parathyroid gland These changes
reverted to normal 66 hr after
exposure.
Structural changes in cell mem-
branes and nuclei of myocardial
muscle fibers that were reversible
about 1 month following exposure.
TABLE 8-9. EXTRAPULMONARY EFFECTS OF OZONE:
Ozone,
^g/m3
1960
1960
1960
1568
490
392
110
Ozone, Length of
ppm exposure
1 3 hr/day
up to 7
days
1 90 mm
1 7- to 24-
hr/day
18 months
0.8 5 days/week
18 weeks
0.25 2 hr
02 7 hr/day
5 -jys/week
during
gestation and
3 weeks
after birth
0 056 Continuous
93 days
Observed effect(s)
After 2 or 3 days of exposure, mice
slept longer after injections of
sodium pentobarbital.
No change in liver cytochrome PASO
level
In brain tissue, decreased COMT and
altered MAO activity, no lipid
peroxidation and no change in
catecholamme levels, 5-
nucleotidase, acetylcholmesterase.
or ATPase
Lower titratable acidity in urine
on days 91 , 98 and 1 1 2, and a higher
pH on days 98 and 1 1 2 No change in
urinary creatmme, creatine, uric
acid/creatmine or ammo acid
nitrogen/creatinme.
No differences m gastric secreto-
motor activity
Increased incidence of blepharo-
phimosis and unlimited incisor
growth
Increased levels of urinary 17-
ketosteroids that remained elevated
after exposure. Decreased ascorbic
acid content of adrenal gland, which
returned to normal levels 15 days
after exposure ceased.
Rats
Mice
Rats
Subhuman
primates
Rats
Rats
Reference
Fletcher et al.68
Peterson et al.'67
Xmtaras et al.210
Reynolds and Chaffee172
Konigsberg and Bachman"3
Eglite53
: MORPHOLOGY
Species
Rabbits
Mice
Reference
Atwal and Wilson7
Brmkman et al 20
MISCELLANEOUS
Species
Mice
Rabbits
Dogs
Rats
Rats
Mice
Rats
Reference
Gardner et al.78
Goldstein et al.88
Trams et al.'99
Hathaway and Terrill98
Roth and Tansy178
Veninga201
Eglite63
174
-------
recovery period, showed no significant changes in
collagen:protein ratios. However, prolyl
hydroxylase (thought to be the rate-limiting
enzyme in collagen synthesis) levels were elevated
(p < 0.05) most in the lungs of animals exposed to
irradiated exhaust (Group 3).158
Emik et al.56 exposed various species of
laboratory animals (mice, rats, and rabbits) to
ambient California air for 2.5 years. The average
ambient concentrations were 0.057 ppm oxidant,
1.7 ppm CO, 2.4 ppm hydrocarbons (as carbon),
0.019 ppm N02, 0.015 ppm NO, and 4.2 ppb PAN.
The following results were obtained: Reduced
pulmonary alkaline phosphatase in rats, reduced
serum glutamic oxaloacetic transaminase in
rabbits, increased pneumonitis in mice, increased
mortality in male mice but not in female mice,
reduced body weights in mice, and decreased
running activity of male mice. When aging guinea
pigs (24 months old) that lived in smog for 2 years
were allowed to recover for 6 weeks in clean,
filtered air and then exposed to 980 /ug/m3 (0.5
ppm) ozone for 10 min, they had a smaller increase
in lung resistance following ozone exposure than
animals that lived in clean air before the same
ozone exposure. The authors suggest that a
possible adaptation occurred. Pulmonary tumors
were found in the lungs of some mice, but there
was no significant induction of lung adenomas in
the two strains of mice exposed to ambient, smog-
containing air. They state further that the
exposure was probably near the threshold of
effect, and therefore measurement of differences
induced by smog was difficult.
The spontaneous activity of mice was reduced
when the mice were exposed to smog ozone +
gasoline vapor for 24 hr. Decreases(p<0.05)were
observed at 980 A/g/m3 (0.05 ppm) ozone. A
simultaneous measurement of 1.69 ppm oxidant
was made. As the concentrations of ozone and
oxidant were increased (3650 A/g/m3, or 1.86 ppm
ozone and 6.26 ppm oxidant), greater decreases in
spontaneous activity were observed.18 Emik and
Plata55 made a similar study of mice exposed
continuously either to filtered air or to ambient air
(oxidant from 0.062 to 0.239 ppm; N02 from 0.03
to 0.07 ppm; total hydrocarbons from 2.7 to 4.4
ppm) for 13 months. Animals exposed to the
ambient air had decreased running activity
compared to controls over the course of the study.
The differences in activity were strongly related to
oxidant for weekly intervals. High temperature and
age were related to decreased activity in both
groups.
Kotin and Thomas116 exposed mice of both sexes
continuously for 19 weeks to smog (formed by
reacting gasoline with ozone, average of 1.25 ppm
oxidant); natural urban atmosphere (fluctuated,
highest reading recorded was 0.4 ppm oxidant); or
clean, filtered air. Females held in the smog
chamber had a decrease in conception rate (p =
0.02). The significant difference in litter rate in the
smog-exposed animals was primarily attributed to
an effect on the females. The survival rate of the
newborns was also decreased (p < 0.01) when the
parents were exposed to smog. Those animals in
the smog chamber also had lower (p < 0.05)
average litter sizes.
TABLE 8-10. ATMOSPHERIC MEAN CONCENTRATIONS AND THEIR STANDARD DEVIATIONS ADMINISTERED
FROM 8 a.m. TO MIDNIGHT EACH DAY
Group
1
2
3
4
5
6
7
8
Atmosphere CO HC (as CH.)
Control air — —
Nonirradiated 1121+115 180±2.9
auto exhaust
Irradiated auto 1086±225 156+40
exhaust
SO2 + H2S04 — —
Nonirradiated 11 3.1 ±15 9 179+28
auto exhaust
+ SO2 * H2S04
Irradiated auto 1090±22.8 156+39
exhaust + SOa
* H2SO4
Nitrogen oxides — —
Nitrogen oxides — —
NO,
009 ±004
1 77±0 68
—
0.09±0 06
1 68±0 68
1 21±022
027±0 62
Pollutant. mg/m3
NO O.lasOJ
_ _
1 78±052 —
023±036 039±018
— —
1 86±50 54 —
0.23±0.36 039±016
031 ±008 —
2 05±0.26 —
SOz HjSO,
— —
— —
1 10±0 57 0.09±0.04
1 27±061 009±0.04
1 10+056 0 11 ±004
— —
— —
175
-------
Nakajima et al,152 studied histopathologic
changes in the lungs of mice that were exposed to
irradiated auto exhaust and oxidant-fortified
exhaust-gas mixtures for 2 to 3 hr/day, 5
days/week for a month. Histopathologic changes
resembling tracheitis and bronchial pneumonia
were observed in mice exposed to atmospheres
containing oxidant at 0.1 to 0.5 ppm. In those
exposed to atmospheres containing 0.1 to 0.15
ppm, the changes were minimal, the main finding
being irregular arrangement of the epithelial cells
of the relatively thick bronchioles.
Summary
Animal lexicological studies have investigated
the biological response of laboratory animals to
sample atmospheres of photochemical reaction
mixtures, Long-term exposure of various species
of animals to ambient California atmospheres have
produced changes in pulmonary functional
measurements in the guinea pig (980 /ug/m3, or
0.5 ppm) and a number of biochemical and
pathological effects in mice, rats, and rabbits.
Exposure to irradiated auto exhaust containing
oxidant levels ranging from 0.2 to 1.0 ppm also
produced numerous changes in experimental
animals. In mice, an increase in susceptibility to
infection and a decrease in spontaneous running
activity, infant survival rate, and fertility were
reported in exhaust-exposed animals. In guinea
pigs, the following changes were observed:
Increased tidal volume, increased minute volume,
increased flow resistance, and a decreased
frequency of breathing after short-term exposure
to irradiated auto exhaust. Dogs exhibited several
alterations in normal pulmonary function
capabilities.
Note that experimental exposure to irradiated
auto exhaust usually involves variable con-
centrations of carbon monoxide, hydrocarbons,
and nitrogen oxides, as well as oxidants. The
studies of the effects of oxidants on animals are
summarized in tabular form in Table 8-11.
TABLE 8-11. EFFECTS OF OXIDANTS ON ANIMALS
Pollutant
concentration
03 (980 ^g/m3,
0 5 ppm) and
gasoline vapor
(oxidant 1 69 ppm)
Oa and gasoline
(oxidant 1 25 ppm)
Irradiated auto
exhaust (oxidant,
01-05 ppm)
Oxidant, 0062-
0239 ppm, NO-i,
003-007 ppm,
hydrocarbons,
2 7-4 4 ppm
Oxidant, 0 057
ppm, CO, 1 7
ppm, hydro-
carbons, 2 4 ppm,
NO2, 0019 ppm,
NO, 0015 ppm,
PAN, 4 2 ppb
7 exposure groups
(see Table 8-10),
various
mixtures of CO,
hydrocarbons, N02,
NO, 03, S02, H2S04
and irradiated and
nonirradiated auto
exhaust
Length of
exposure
Observed effect(s|
Species
24 hr
Decreased spontaneous running Mice
activity
Continuous
19 weeks
2-3 hr/day
5 days/week for
1 month
Continuous
1 3 months
Decreased conception rate, litter Mice
rate, and newborn survival.
Tracheitis and bronchial pneumonia Mice
Decreased spontaneous running Mice
activity
Continuous Reduced pulmonary alkaline phos- Mice,
2 5 years phatase (rats), reduced serum rats,
glutamic oxaloacetic transammase rabbits
(rabbits), increased pneumonitts
(mice), increased mortality (male
mice), reduced body weights
(mice), decreased running activity
(male mice), no significant induction
of lung adenomas (mice)
16 hr/day, No changes in cardiovascular Beagles
7 days/weeks parameters attributable to
for 68 months pollutants, alterations in pulmonary
function
Boche and Quilhgan"3
Kotin and Thomas"6
Nakajima et al 152
Emik and Plata55
Emiket al66
Vaughan et al.200
Lewis et a I 12°
176
-------
EFFECTS OF PEROXYACETYLNITRATE ON
EXPERIMENTAL ANIMALS
Experimental Data
To determine the effects of PAN on acute
respiratory infections, Thomas et al.197 exposed
mice to PAN for 2 or 3 hr and then to aerosols of
Streptococcus pyogenes (Group C). The data
indicate that a 3-hr exposure to 26,000fjg/m3(5.2
ppm) PAN resulted in an increased (p < 0.05)
incidence of fatal pulmonary infection.
Campbell et al.24 exposed mice to PAN, 485,000
to 725,000 /jg/m3 (97 to 145 ppm) as measured at
the chamber outlet, for 2 hr at 80°F (27°C). The
studies demonstrated that most mice exposed to
550,000A
-------
16 Boatman, E S , S Sato, and R Frank Acute effects of
ozone on cats'lungs.ll. Structural Am Rev Respir Dis
1W 157-169, 1974
17 Bobb, G A, and E J Fairchild, II Neutrophil-to-
lymphocyte ratio as indicator of ozone exposure Toxicol
Appl Pharmacol 11 558-564, 1967
18 Boche, R D , and J J Quilligan.Jr Effects of synthetic
smog on spontaneous activity of mice Science
131 1733-1734, 1960
19 Bokma, A I., N D Eksler, A D Semenenko, and R V
Merkur'yeva. Investigation of the mechanism of action
of atmospheric pollutants on the central nervous system
and comparative evaluation of methods of study
Environ Health Persp 13 37-42, 1976
20 Brmkman, R , H B Lamberts, and T S Venmga
Radiomimetic toxicity of ozonized air Lancet 1 133-
136, 1964
21 Brummer, M E G., L W Schwartz, and N K McQuillen
A quantitative study of lung damage by scanning
electron microscopy Inflammatory cell response to high
ambient levels of ozone In Scanning Electron
Microscopy/1977 Vol II Proceedings of the Workshop
on Biomedical Application—SEM and General Organ
Systems ill Research Institute, Chicago, III, March
1977 pp 513-518.
22 Buell, G. C , E Jeung, and W Fennmger Chemical
changes in respiratory tissue following ozone exposure
Presented at the Third Technical Meeting, West Coast
Section, Air Pollution Control Association, Monterey,
Calif., September 19-26, 1963
23 Buell, G C, Y Tokiwa, and P K Mueller Potential
crosshnkmg agents in lung tissue Formation and
isolation after in vivo exposure to ozone Arch. Environ
Health 10 213-219, 1969
24 Campbell, K I , G L Clarke, L. 0 Emik, and R. L Plata
The atmospheric contaminant peroxyacetylnitrate
Acute inhalation toxicity in mice Arch Environ Health
15 739-744, 1967
25 Castleman, W L , D L Dungworth, andW S Tyler Cyto-
chemically detected alterations of lung acid
phosphatase reactivity following ozone exposure Lab
Invest 29 310-319, 1973
26. Castleman, W L., DL Dungworth, and WS Tyler
Histochemically detected enzymatic alterations in rat
lung exposed to ozone Exp Mol Pathol 19 402-421,
1973
27 Chow, C K Biochemical responses in lungs of ozone-
tolerant rats Nature (London) 260 721-722, 1976.
28 Chow, C. K , C J. Dillard, and A L Tappel Glutathione
peroxidase system and lysozyme in rats exposed to
ozone or nitrogen dioxide Environ Res 7311-319,
1974
29 Chow, C K , M Z. Hussam.C E Cross, D L Dungworth,
and M. G. Mustafa Effect of low levels of ozone on rat
lungs I Biochemical responses during recovery and
reexposure Exp Mol Pathol 25182-188,1976
30 Chow, C K, M G Mustafa, C E Cross, and B K
Tarkington. Effects of ozone exposure on the lungs and
the erythrocytes of rats and monkeys Relative
biochemical changes Environ Physiol Biochem
5 142-146, 1975
31 Chow, C K.,andA L Tappel Activitiesof pentoseshunt
and glycolytic enzymes in lungs of ozone-exposed rats
Arch Environ Health 27 205-108, 1 973
32. Chow, C. K., and A L Tappel. An enzymatic protective
mechanism against lipid peroxidation damage to lungs
of ozone-exposed rats Lipids 7 518-524, 1972
33 Christensen, E , and A C Giese Changes in adsorption
spectra of nucleic acids and their derivatives following
their exposure to ozone and ultraviolet irradiation Arch
Biochem Biophys 51 208-216, 1954
34 Coffin, D L , and E J Blommer Alteration of the
pathogenic role of streptococci group C in mice
conferred by previous exposure to ozone In
Aerobiology Academic Press, Inc., New York, 1 970. pp.
54-61
35 Coffin, D L, E J Blommer, D E Gardner, and R
Holzman Effect of air pollution on alteration of
susceptibility to pulmonary infections In Proceedings
of the 3rd Annual Conference on Atmospheric
Contaminants in Confined Spaces AMRL-TR-67-200,
Aerospace Medical Research Laboratories, Wright-
Patterson Air Force Base, Dayton, Ohio, 1968 pp 71-
80
36 Coffin, D L , and D E Gardner Interaction of biological
agents and chemical air pollutants Ann Occup Hyg
15 219-235, 1972
37 Coffin, D L , and D E Gardner The role of tolerance in
protection of the lung against secondary insults. In'
Erstes Internationales der Werksartz Symposium der
chemischen Industrie Medichem, Ludwigshafen, W
Ger , July 1972 pp 344-364
38 Coffin, D L , D E Gardner, R S Holzman, and F J
Wolock Influence of ozone on pulmonary cells Arch
Environ Health 16 633-636, 1968
39 Committee on the Challenges of Modern Society Air
Quality Criteria for Photochemical Oxidants and Related
Hydrocarbons. NATO/CCMS Publication No 29, North
Atlantic Treaty Organization, Brussels, Belgium,
February 1974
40 Cortesi, R , and 0 S Pnvett Toxicity of fatty ozonides
and peroxides Lipids 7 71 5-721, 1972
41 Criegee, R Course of ozonization of unsaturated
compounds Rec Chem Progr 78111-120,1957
42 Davis, I Microbiologic Studies with Ozone
Mutagenesis of ozone for Escherichia coll Document
No 61-60 US Air Force, School of Aerospace
Medicine, Aerospace Medical Center, Brooks Air Force
Base, Tex , June 1961
43 Davis, I Microbiologic Studies with Ozone
Mutagenesis of Ozone from Escherichia coll Report No
61-60, US Air Force, Aerospace Medical Center,
School of Aerospace Medicine, Brooks Air Force Base,
Tex, June 1961
44 DeLucia, A J , P M Hogue, M G Mustafa, and C E
Cross Ozone interaction with rodent lung Effect on
sulfhydryls and sulfhydryl-contammg enzyme activities
J Lab Clm Med SO 559-566, 1973
45 DeLucia, A J,M G Mustafa, M Z Hussam, and C E
Cross Ozone interaction with rodent lung III Oxidation
of reduced glutathione and formation of mixed
disulfides between protein and nonprotem sulfhydryls
J Clm Invest 55794-802, 1975
46 Dillard, C J , N Urnbarn, K Reddy, B Fletcher, S
Taylor, B DeLumen, S Langberg, and A L Tappel.
Increased lysosomal enzymes in lungs of ozone-
exposed rats Arch Environ Health 25 426-431, 1972
178
-------
47 Dixon, J R., and J. T. Mountain. Role of histamine and
related substances in development of tolerance to
edemagenic gases. Toxicol Appl. Pharmacol. 7 756-
766, 1965.
48. Dowell, A. R , L. A Lohrbauer, D. Hurst, and S. D. Lee.
Rabbit alveolar macrophage damage caused by in vivo
03 inhalation. Arch. Environ. Health27.121-127,1970.
49. Dungworth, D L. Short-term effects of ozone on lungs of
rats, mice, and monkeys Environ. Health Persp. 76'179,
1976.
50 Dungworth, D L., W. L. Castleman, C. K. Chow, P. W
Mellick, M. G. Mustafa, B. Tarkington, and W. S. Tyler.
Effect of ambient levels of ozone on monkeys. Fed. Proc
Fed. Am. Soc. Exp. Biol 34 1670-1674, 1975.
51. Dungworth, D. L., G L Clarke, and R L. Plata.
Pulmonary lesions produced in A-stram mice by long-
term exposure to peroxyacetylnitrate Am Rev Respir.
Dis 33565-574, 1969.
52 Easton, R. E, and S D. Murphy Experimental ozone
pre-exposure and histamine. Effect on the acute toxicity
and respiratory function effects of histamine in guinea
pigs Arch. Environ. Health 75 160-166, 1967.
53 Eglite, M E A contribution to the hygienic assessment
of atmospheric ozons. Hyg Sanit. 33 18-23, 1968.
54 Ehrhch, R., J. C Fmdlay, J. D. Fenters, and D. E.
Gardner Health effects of short-term inhalation of
nitrogen dioxide and ozone mixtures Environ Res
14 223-231, 1977
55 Emik, L 0 , and R L Plata Depression of running
activity in mice by exposure to polluted air Arch
Environ Health 18. 574-579, 1969
56 Emik, L 0., R L Plata, K. I. Campbell, and G. L. Clarke
Biological effects of urban air pollution Riverside
summary Arch. Environ. Health 23.335-342, 1971
57 Evans, M J , L V Johnson, R J Stephens, and G.
Freeman Cell renewal in the lungs of rats exposed to
low levels of ozone Exp. Mol. Pathol. 24:70-83, 1976.
58 Evans, M J , L V Johnson, R J. Stephens, and G.
Freeman Renewal of the terminal bronchiolar
epithelium in the rat following exposure to N02 or 03.
Lab Invest 35-246-257, 1976.
59 Evans, M J , W Mayr, R F.Bils.andC G Loosli Effects
of ozone on cell renewal in pulmonary alveoli of aging
mice Arch Environ Health 22 450-453, 1971
60. Evans, M J ,R J.Stephens, and G Freeman Renewal of
pulmonary epithelium following oxidant injury. In
Lung Cells in Disease A Bonhuys, ed Elsevier, New
York, 1976 pp 165-178.
61 Fairchild, G A Effects of ozone and sulfur dioxide on
virus growth in mice. Arch. Environ Health 32.28-33,
1977.
62 Fairchild, G A Ozone effect on respiratory deposition of
vesicular stomatitis virus aerosols. Amer Rev Respir
Dis 109 446-449, 1974.
63 Fairchild, E J,S L Graham, M. Hite, R Killens, and L
D Scheel Changes in thyroid I131 activity in ozone-
tolerant and ozone-susceptible rats Toxicol Appl.
Pharmacol 6607-613, 1964
64 Fairchild, E.J., II, S D Murphy, and H. E. Stokinger
Protection by sulfur compounds against the air
pollutants ozone and nitrogen dioxide Science
730861-862, 1959
65. Fetner, R H Chromosome breakage in Vicia faba by
ozone Nature (London) 181 504-505, 1958
66 Fetner, R.H Mitotic Inhibition Induced in Grasshopper
Neuroblasts by Exposure to Ozone SAM-TDM-63-39,
U.S Air Force, School of Aerospace Medicine,
Aerospace Medical Center, Brooks Air Force Base, Tex ,
June 1963
67. Fetner, R H. Ozone-induced chromosome breakage in
human cell culture. Nature (London) 194 793-794,
1962
68 Fletcher, B L., and A. L. Tappel Protective effects of
dietary alphatocopherol in rats exposed to toxic levelsof
ozone and nitrogen dioxide Environ. Res 6:165-175,
1973.
69 Frank.N. R ,J.D.Brain,andD E Sherry Unilateral lung
exposure to ozone. Fed Proc. Fed Am Soc Exp. Biol
27-279, 1968
70 Freeman, G , L T Juhos, N. J. Furiosi, R Mussenden, R
J Stephens, and M. J Evans Pathology of pulmonary
disease from exposure to interdependent ambient gases
(N02 and 03). Arch Environ Health 29 203-210, 1974
71. Freeman, G , L Juhos, N J Furiosi, W. Powell, and R
Mussenden Criteria from animals exposed to known
concentrations of nitrogen dioxide and ozone, with
potential used in epidemiology In Proceedings
International Symposium on Recent Advances in the
Assessment of Health Effects of Environmental
Pollution. Volume II EUR 5360, Commission of
European Communities, Luxembourg, 1975
72. Freeman, G , R.J. Stephens, D L. Coffin, and J F Stara.
Changes in dogs' lungs after long-term exposure to
ozone Arch Environ Health 29 209-216, 1973
73 Fnberg, L, B Holman, and R Rylander Animal lung
reactions after inhalation of lead and ozone Environ
Physiol Biochem 2170-178, 1972.
74. Fukase, 0 , K Isomura, and H. Watanabe Effect of
ozone on glutathione in vivo Taiki Osen Kenkyu 70.58-
62, 1975
75. Fukase, 0 , K. Isomura, and H Watanabe Effect of
ozone on vitamin C in vivo Taiki Osen Kenkyu 70.63-66,
1975
76 Gardner, D E Environmental Influences of Living
Alveolar Macrophages. Ph D thesis, University of
Cincinnati, Cincinnati, Ohio, 1971
77. Gardner, D E.andJ A Graham Increased pulmonary
disease mediated through altered bacterial defenses In
Pulmonary Macrophage and Epithelial Cells C L
Sanders, R P Schneider, G E. Dagle, and H. A Ragan,
eds ERDA Symposium Senes43, U S Energy Research
and Development Administration, Washington, D C ,
September 1977 pp 1-21
78. Gardner, D E,J W Illing, F J Miller, andD L. Coffin
The effect of ozone on pentobarbital sleeping time in
mice Res Commun Chem Pathol Pharmacol 9 689-
700, 1974
79 Gardner, D E.,T R Lewis, S M Alpert, D J Hurst, and
D L Coffin The role of tolerance in pulmonary defense
mechanisms Arch Environ Health 25 432-438, 1972
80. Gardner, D E,F J Miller, J W Illing, and J M Kirtz
Increased infectivity with exposure to ozone and sulfunc
acid Toxicol Lett 7 59-64, 1977
81. Gardner, D E , E. A Pfitzer, R T Christian, and D L
Coffin Loss of protective factor for alveolar
macrophages when exposed to ozone Arch. Intern
Med 727 1078-1084, 1971
179
-------
82 Goldstein, D B.andO J Balchum Effects of ozone on
lipid peroxidation in the red blood cell. Proc Soc Exp
Biol Med 126 356-358, 1967
83 Goldstein, B. D., and O J Balchum Modification of thf
response of rats to lethal levels of ozone by enzyme-
inducing agents Toxicol Appl Pharmacol 27.330-335,
1974
84 Goldstein, E , M C Eagle, and P. Hoeprich. Influence of
ozone on pulmonary defense mechanisms of silicotic
mice Arch Environ Health 24:444-448, 1972
85 Goldstein, B. D , S J. Hamburger, G. W. Falk, and M. A.
Amoruso Effect of ozone and nitrogen dioxide on the
agglutination of rat alveolar macrophages by
concanavalm A Life Sci 21.1 637-1 644, 1977.
86 Goldstein, B D.C.Lodi.C Collmson, andO J. Balchum
Ozone and lipid peroxidation Arch. Environ Health
18631-635, 1969.
87 Goldstein, B D,B Pearson, C. Lodi, R E. Buckley, and
0 J Balchum The effect of ozone on mouse blood in
vivo Arch Environ Health 16 648-650, 1968.
88 Goldstein, B D ,S Solomon, B S. Pasternack, and D R
Bickers. Decrease in rabbit lung microsomal
cytochrome P-450 levels following ozone exposure
Res Commun Chem Pathol. Pharmacol 10 759-762,
1975
89 Goldstein, E , W S Tyler, P D Hoeprich, and C. Eagle.
Adverse influence of ozone on pulmonary bactericidal
activity of munne lung Nature (London) 229-262-263,
1971
90 Goldstein, E., W S Tyler, D Hoeprich, and C. Eagle.
Ozone and the antibacterial defense mechanisms of the
murmelung Arch Intern Med 128 1099-1102, 1971.
91 Goldstein, E , M A Warshauer, W Lippert, and B.
Tarkington Oi and NOa exposure Munne pulmonary
defense mechanisms Arch Environ. Health 25-85-90,
1974
92 Gooch, P C , D A Creasia, and J G Brewen The
cytogenetic effect of ozone Inhalation and in vitro
exposures Environ Res 12.188-1 95, 1976
93 Gregory, A R , L A Ripperton, and B Miller Effect of
neonatal thymectomy on the development of ozone
tolerance in mice J Am. Ind. Hyg. Assoc 25.278-283,
1967
94 Hadley, J G , D E Gardner, D L Coffin, and D B.
Menzel Enhanced binding of autologous redcellstothe
macrophage plasma membrane as a sensitive indicator
of pollutant damage In Pulmonary Macrophage and
Epithelial Cells C L Sanders, R. P. Schneider, G E.
Dagle, and H A Ragan, eds ERDA Symposium Series
43, U S Research and Development Administration,
Washington, DC, 1977 pp 66-77
95 Hamelin, C , and Y S Chung Characterization of
mucoid mutants of Eschenchia co/i K-12 isolated after
exposure to ozone J Bactenol 722 19-24, 1975
96 Hamelin, C , and Y S Chung Rapid test for assay of
ozone sensitivity m Eschenchia coli Mol Gen Genet
145 191-194, 1976
97 Hamelin, C , and Y S Chung. Resistance a I'ozonechez
Eschenchia co/i II. Relations aveccertains mecanismes
de reparation de I'ADN Mol Gen Genet 129 177-1 84,
1974
98 Hathaway, J A, and R E Ternll Metabolic effects of
chronic ozone exposure on rats. J Am Ind Hyg Assoc.
23 392-395, 1962
99 Hatton, K , N Kato, M. Kmoshita, S Kmoshita, and T
Sunada Protective effect of ozone in mice against
whole-body X-irradiation. Nature (London) 735-1220,
1963
100 Hinners, R G. Engineering the chronic exposure of
animals to laboratory produced automobile exhaust. J
Air Pollut. Control Assoc. 12.527-530, 1962
101 Hinners, R. G Laboratory produced automobile exhaust
facility Biomed Sci. Instrum. 7 53-58, 1963.
102. Holt, P. G., and A. Nulsen. Ozone hazard in UV isolation
units. J Occup. Med 77.186-188,1975
103. Huber, G.L., R J Mason, M. LaForce, N.J. Spencer, D
E. Gardner, and D. L. Coffin Alterations in the lung
following the administration of ozone. Arch Intern.
Med. 728-81-87, 1971
104 Hueter, F. G , G L Contner, K A Busch, and R G
Hinners Biological effects of atmospheres
contaminated by auto exhaust. Arch. Environ Health
72 553-560, 1966
105 Hurst, D J , and D L Coffin Ozone effect on lysosomal
hydrolases of alveolar macrophages in vitro. Arch.
Intern Med 727.1059-1063,1971
106 Hurst, D J,D E Gardner, and D. L. Coffin. Effect of O3
on acid hydrolases of the pulmonary alveolar
macrophage Res J Reticuloendothel Soc 8 288-300,
1970
107 Hussam, M Z, C E Cross, M G Mustafa, and R S.
Bhatnagar Hydroxyprolme contents and prolyl
hydroxylase activities in lungs of rats exposed to low
levels of ozone Life Sci 75-897-904,1976
108 Hussam, M. Z., M G. Mustafa, C K Chow, and C. E
Cross. Ozone-induced increase of lung prolme
hydroxylase activity and hydroxy-prolme content Chest
69(Suppl 2).273-275, 1976.
109 Ibrahim, A L, Y C Zee, and J W Osebold The effects
of ozone on the respiratory epithelium and alveolar
macrophages of mice I Interferon production Proc
Soc Exp. Biol Med 752483-488,1976
110 Jegier.Z Ozone as an air pollutant. Can. J. PublicHealth
64.161-166, 1973
111. Khment, V Similarity and dimensional analysis,
evaluation of aerosol deposition in the lungs of
laboratory animals and man Folia Morphol. (Prague)
27 59-64, 1973
112 Konigsberg, A S. A study of the Measurementof Ozone
and Some Effects of Ozone and Nitrogen Oxides on the
Activity of Rats Ph D thesis, Syracuse University,
Syracuse, N.Y., 1969
113 Konigsberg, A S, and C Bachman. Ozonized
atmosphere and gross motor activity of rats. Int J
Biometeorol 74.261-266,1970
114 Kotin, P , and H L Falk II The experimental induction of
pulmonary tumors in stram-A mice after exposure to
atmosphere of ozonized gasoline Cancer (Philadelphia)
9:910-917, 1956
115 Kotin, P., H L Falk, and C J McCammon III The
experimental induction of pulmonary tumors and
changes in respiratory epithelium in C57BL mice
following exposure to ozonized gasoline Cancer
(Philadelphia) 77.473-481, 1958
116 Kotin, P, and M Thomas Effect of air contaminants on
reproduction and offspring survival m mice AMA Arch
Ind Health 76411-413, 1957
180
-------
117 Kyei-Aboagye, K., M Hazucha, I Wyszogrodski, D
Rubinstein, and M E. Avery The effect of ozone
exposure in vivo on the appearance of lung tissue lipids
in the endobronchial lavage of rabbits Biochem
Biophys Res Commun 54907-913,1973
118. Last, J. A, M Jennings, and C. E. Cross. Mucus
glycoprotem secretion by tracheal explants from rats
exposed to ozone Fed. Proc Fed Am Soc Exp Biol
36.413, 1977
119 Lewis, T. R., F G Hueter, and K. A Busch. Irradiated
automobile exhaust Its effects on the reproduction of
mice Arch Environ Health 75:26-35, 1967.
120. Lewis,! R., W. J Moorman, Y. Y. Yang, and J. F Stara
Long-term exposure to auto exhaust and other pollutant
mixtures. Effects on pulmonary function in the beagle
Arch Environ Health 29-102-106, 1974
121 Lunan, K D, P Short, D. Negi, and R J Stephens
Glucose-6-phosphate dehydrogenase response of
postnatal lungs to NOa and O3. In Pulmonary
Macrophage and Epithelial Cells C. L. Sanders, R. P.
Schneider, G E Dagle, and H A Ragan, eds ERDA
Symposium Series 43, U S Energy Research and
Development Administration, Washington, D.C., 1977
122 Maclean, S. A., A C Longwell, and W. J Blogoslawski.
Effects of ozone-treated seawater on the spawned,
fertilized, miotic, and cleaving eggs of the commercial
American oyster Mutat. Res 27.283-285, 1973
123 Matsumura, Y. The effects of ozone, nitrogen dioxide,
and sulfur dioxide on the experimentally induced
allergic respiratory disorder in guinea pigs. I The effect
on sensitization with albumin through the airway.
Amer. Rev Respir Dis 702.430-437,1970.
124. Matsumura, Y , K Mizuno, T Miyamoto, T Suzuki, and
Y Oshima The effects of ozone, nitrogen dioxide, and
sulfur dioxide on experimentally induced allergic
respiratory disorder in guinea pigs IV Effects on
respiratory sensitivity to inhaled acetylcholine. Amer
Rev Respir Dis. 705.262-267, 1972
125. McJilton, C. E., J Thielke, and R Frank. Ozone uptake
model for the respiratory system J Am. Ind. Hyg. Assoc
33 20, 1972
126 Mellick, P W , D L. Dungworth, L W Schwartz, and W
S. Tyler Short-term morphologic effects of high
ambient levels of ozone on lungs of rhesus monkeys.
Lab Invest 3682-90, 1977.
127 Mendenhall, R. M., and H E. Stokinger. Films from lung
washings as a mechanism model for lung injury by
ozone J Appl Physiol 7728-32, 1962.
128 Menzel, D B Oxidation of biologically active reducing
substances by ozone Arch Environ. Health 23 149-
153, 1971
129 Menzel, D B.Theroleoffreeradicalsin the toxicityof air
pollutants (nitrogen oxides and ozone). In Free Radicals
in Biology Volume II W A Pryor, ed. Academic Press,
Inc., New York, 1976 pp 181-202.
130 Menzel, D B Toxicity of ozone, oxygen, and radiation
Annu. Rev. Pharmacol 70-379-394, 1970.
131. Menzel, D. B., W G Anderson, M B Abou-Donia.
Ozone exposure modifies prostaglandm biosynthesis in
perfused rat lungs. Res. Commun, Chem. Pathol
Pharmacol 75-135-147, 1976.
132. Menzel, D B..J.N Roehm.andS D Lee VitammE.The
biological and environmental antioxidant J. Agric. Food
Chem 20481-486r 1972.
133 Menzel, D B., R. Slaughter, D Donovan, and A Bryant
Fatty acid ozonides as the toxic agents on ozone
inhalation Pharmacologist 75239, 1973.
134 Miller, F J A Mathematical Model of Transport and
Removal of Ozone in Mammalian Lungs. Ph D. thesis,
North Carolina State University, Raleigh, N C , 1977
135 Miller, F J, J W. Illmg, and D E Gardner Effect of
urban ozone levels on laboratory-induced respiratory
infections. Toxicol Lett. 2 163-169, 1978
136 Miller, F. J., D. B. Menzel, and D L. Coffin. Similarity
between man and laboratory animals in regional
pulmonary deposition of ozone. Environ. Res 7784-
104, 1978.
137 Mittler, S , D. Hendrick, M King, and A Gaynor Toxicity
of ozone 1. Acutetoxicity Ind Med Surg 25.301-306,
1956
138 Mizcguchi, I, M Osawa, Y Sato, K Makmo, and H.
Yagyu Studies on erythrocyte and photochemical smog
Part I Effects of air pollutants on erythrocyte resistance*
Taiki Osen Kenkyu 8 414, 1973
139 Moorman, W J , J. J Chmiel, J F Stara, and T R
Lewis. Comparative decomposition of ozone in the
nasopharynx of beagles. Acute vs chronic exposure
Arch Environ Health 26.1 53-155, 1973
140. Mudd, J. B,R Leavitt, A Ongun, and T T. McManus
Reaction of ozone with ammo acids and proteins
Atmos Environ 3-669-681, 1969
141. Mudd, J B., F. Leh, and T T. McManus Reaction of
ozone with nicotmamide and its derivatives. Arch
Biochem Biophys 767-408-419, 1974.
142. Murphy, S. D. A review of effects on animals of exposure
to auto exhaust and some of its components J Air
Pollut. Control Assoc 14 303-308, 1964
143 Murphy, S. D., D. A. Klmgshme, and C E Ulrich
Respiratory response of guinea pigs during acrolem
inhalation and its modification by drugs J Pharmacol
Exp. Ther. 747.79-83, 1963
144 Murphy, S D., J. K. Leng, C E Ulrich, and H V Davis
Effects on animals of exposure to auto exhaust Arch
Environ Health 7.60-70, 1963
145. Murphy.S D.andC E. Ulrich Multi-ammaltestsystem
for measuring effects of irritant gases and vapors on
respiratory function of guinea pigs J Am. Ind Hyg
Assoc. 25.28-36, 1964
146 Murphy, S. D , C. E Ulrich, S. H Frankowitz, and C
Xmtaris Altered function in animals inhaling low
concentrations of ozone and nitrogen dioxide J Am
Ind Hyg Assoc. 25246-253, 1964
147. Mustafa, M G Augmentation of mitochondnaloxidative
metabolism in lung tissue during recovery of animals
from acute ozone exposure Arch Biochem Biophys
765531-538, 1974.
148. Mustafa, M. G , A. J.DeLucia, G K York, C Arth, and C
E Cross Ozone interaction with rodent lung II. Effects
on oxygen consumption of mitochondria. J. Lab Clm.
Med. S2.357-365, 1973.
149. Mustafa, M.G ,A. B Hacker, J J Ospital.M Z Hussam,
and S. D. Lee Biochemical effects of environmental
oxidant pollutants in animal lungs. In' Biochemical
Effects of Environmental Pollutants S. D. Lee, ed Ann
Arbor Science Publishers, Ann Arbor, Mich., 1977 pp.
59-96
181
-------
150. Mustafa, MG, and S D Lee Pulmonary biochemical
alterations resulting from ozone exposure Ann Occup
Hyg. 19 17-26, 1976
151 Mustafa, M G, S M Macres, B K Tarkington, C K
Chow, and M Z Hussain Lung superoxide dismutase
(SOD). Stimulation by low-level ozone exposure Clm
Res. 23 138A, 1975
152 Nakajima, T , S Kusumoto, Y Tsubota, E Yonekawa, R
Yoshida, K Motomiya, K Ito, G. Ide, and H Otsu
Histopathological changes in the respiratory organs of
mice exposed to photochemical oxidants and
automobile exhaust gas Osaka-funtsu Koshu Eisei
Kenkyusho Kenkyu Hokoku Rodo Eisei Hen (10) 35-42,
1972
153. Nasr, A N.M , B D. Dinman, and IA Bernstein
Nonparticulate air pollutants II Effect of ozone
inhalation on nadide phosphate levels in tracheal
mucosa Arch Environ. Health 22 545-550, 1971
154 National Air Pollution Control Administration Air
Quality Criteria for Photochemical Oxidants NAPCA
Publication No. AP-63, US Department of Health,
Education, and Welfare, Public Health Service,
Washington, D C , March 1970
155 National Research Council Ozone and Other
Photochemical Oxidants National Academy of
Sciences, Washington, D C.1977 pp 306-312
156 National Research Council Toxicology In Ozone and
Other Photochemical Oxidants. National Academy of
Sciences, Washington, D C , 1977 pp 323-387
157 Nettesheim, P, D A Creasia, and T J Mitchell
Carcinogenic and cocarcmogenic effects of inhaled
synthetic smog and ferric oxide particles J Nat Cancer
Inst 55 159-166, 1975
158 Orthoefer, J G,R S Bhatnagar.A Rahman, Y Y Yang,
S D. Lee, and J F Stara Collagen and prolyl
hydroxylase levels in lungs of beagles exposed to air
pollutants. Environ Res 12 299-305, 1976
159. Pagnotto, L D, and S S. Epstein Protection by
antioxidants against ozone toxicity in mice Expenentia
25703, 1969.
160 Palmer, M S.R.W Exley, and D L Coffin Influence of
pollutant gases on benzpyrene hydroxylase activity
Arch Environ Health 25.439-442, 1972.
161. Palmer, M S,D H Swanson, and D L Coffin. Effect of
Oa on benzpyrene hydroxylase activity in the Syrian
golden hamster Cancer Res 31.730-733, 1971
162. P'an, A, J Beland, and Z Jegier Ozone-induced
arterial lesions. Arch Environ Health 24:229-232,
1972.
163 P'an, A. Y S , and Z. Jegier The effect of sulfur dioxide
and ozone on acetylchohnesterase Arch Environ
Health 21 498-501, 1970
164. P'an, A Y. S , and Z. Jegier Serum protein changes
during exposure to ozone J Am Ind Hyg Assoc.
37-329-334, 1976
165. P'an, A Y S , andZ Jegier The serum trypsin inhibitor
capacity during ozone exposure Arch Environ Health
23215-219, 1971
166 P'an, A. Y. S., and Z. Jegier Trypsin protein esterase in
relation to ozone-induced vascular damage Arch
Environ. Health 24 233-236, 1972
167 Peterson, D.C., and H. L Andrews The role of ozone in
radiation avoidance in the mouse Radiat Res 7S'331-
336, 1963.
168. Prat, R., C Nofre, and A Cier Effects of sodium
hypochlonte, ozone, and ionizing radiations on
pyrimidme constituents of Escherichia coli Ann Inst
Pasteur 114 595-607, 1968
169 Previero, A, A Signer, and S Bezzi Tryptophan
modification in polypeptide chains Nature (London)
204687-688, 1964
170 Pryor, W A Free radical reactions in biology Initiation
of hpid autoxidation by ozone and nitrogen dioxide
Environ Health Persp 16 180-181, 1976
171 Pryor, W A., J.P Stanley, E. Blair, and G B. Cullen
Autoxidation of polyunsaturated fatty acids Part I. Effect
of ozone on the autoxidation of neat methyl linoleate
and methyl Imolenate Arch Environ Health 37 201-
210, 1976
172 Reynolds, R W, and R R Chaffee Studies on the
combined effects of ozone and a hot environment on
reaction time in subhuman primates In Project Clean
Air Volume 2 California University, Research Project
S-6, Santa Barbara, Calif , 1970
173 Richmond, V L In vitro hydrolase and phagocytic
activity of alveolar macrophages J Lab Clin. Med.
S3.757-767, 1974
174 Roehm, J,J G Hadley, and D. B Menzel Antioxidants
vs. lung disease Arch Intern Med. 72S'88-93, 1971.
175. Roehm, J,J G. Hadley, and D B Menzel Influence of
vitamin E on the lung fatty acids of rats exposed to
ozone Arch Environ Health 24'237-242, 1 972
176 Roehm, J.N., J G Hadley, and D B Menzel. Oxidation
of unsaturated fatty acids by ozone and nitrogen dioxide
A common mechanism of action Arch Environ. Health
23.142-148, 1971.
177 Ross, B K,M P Hlastala, and R Frank On the lack of
effect of ozone on hemoglobin-oxygen affinity Arch
Environ Health (m press)
178 Roth, R. P., and M F Tansy Effects of gaseous air
pollutants on gastric secreto-motor activities in the rat
J Air Pollut Control Assoc 22706-709,1972
179. Sachsenmaier, W, W Siebs, and T Tan Effects of
ozone on mouse ascites tumor cells and on chick
fibroblasts in tissue culture Z Krebsforsch 67.113-
126, 1965
180. Sato, S , M Kawakami, S Maeda, and T Takishima
Scanning electron microscopy of the lungs of vita mm-E-
deficient rats exposed to low concentration of ozone
Am Rev Respir Dis. 773'809-821, 1976
181 Scheel, L D,O J Dobrogorski, J T Mountain, J L
Svirbely, and H E Stokmger Physiologic, biochemical,
immunologic, and pathologic changes following ozone
exposure J Appl Physiol 7467-80,1959
182 Schlipkoter, H, and J Bruch Functional and
morphological alterations caused by exposure to ozone.
Zentralbl Baktenol Parasitenkd. Infektionskr Hyg Abt.
1 Orig Reihe B 756486-499, 1973
183 Schwartz, L W Comparison of the effects of ozone and
oxygen on lungs of rats Environ Health Persp 76.179-
180, 1976
184 Schwartz, L. W., D. L Dungworth, M G Mustafa, B K.
Tarkington, and W S Tyler Pulmonary responseof rats
to ambient levels of ozone Effects of 7-day intermittent
or continuous exposure Lab Invest 34.565-578,1976
185 Scott, D B. M , and E C Lesher. Effect of ozone on
survival and permeability of Escherichia coli J
Bactenol 85 567-576, 1 963
182
-------
186 Selo, K, M Kawakami, K Sugita, M Shishido, M.
Yamaji, and T. Tsuda. Influence of inhalation of ozone
and nitrogen dioxide on the formation of phospholipids
on the lung. Igaku to Seibutsugaku 86 317-320, 1973.
187 Seto, K, M. Kawakami, M Takeshima, M Kon, K.
Sugita, and M. Shishido Influence of short period
exposure to ozone a nd nitrogen dioxide on the formation
of phospholipids in the lungs of rabbits Igaku to
Seibutsugaku 87 57-60, 1 973
188 Shakman, R Nutritional influences on the toxicity of
environmental pollutants Arch Environ Health
28 105-113, 1974.
189 Skillen, R G Progress Report—Research Contract No
50. State of California, Department of Public Health,
Berkeley, Calif., April 17, 1961
190 Stephens, R J, M F Sloan, M J Evans, and G.
Freeman Alveolar Type 1 cell response to exposure to
0.5 ppm O3 for short periods Exp Mol Pathol 20 11 -
23, 1974
191 Stephens, R J, M F Sloan, M J Evans, and G
Freeman Early response of lung to low levels of ozone.
Amer J Pathol 7431-58, 1973
192 Stephens, R J , M. F Sloan, and D G Groth Effects of
long-term, low-level exposure of NO? or Cb on rat lungs.
Environ Health Persp 16 178-179, 1976.
193 Stokinger, H E Effects of air pollution on animals. In. Air
Pollution Volume I A C Stern, ed. Academic Press,
Inc , New York, 1962 pp. 282-334
194 Stokinger, H E Ozone toxicology A Review of research
and industrial experience, 1954-1964 Arch Environ
Health 70.719-731, 1965
195 Stokinger, H E., W D Wagner, and O. L J Dobrogorski.
Ozone toxicity studies III. Chronic injury to lungs of
animals following exposure at a low level AMA Arch
Ind Health 76514-522, 1957.
196 Teige, B , T T McManus, and J. B. Mudd Reaction of
ozone with phosphatidylcholme liposomes and the lytic
effect of products on red blood cells Chem Phys Lipids
72 153-171, 1974
197 Thomas, G, J D Fenters, and R. Ehrlich Effect of
Exposure to PAN and Ozone on Susceptibility to Chronic
Bacterial Infection Quarterly Report IITRI Report
L6075-12, IIT Research Institute, Chicago, III , February
1977
198 Thompson, G E Experimental acute pulmonary edema-
in the rat Effect of Bordetella pertussis vaccine on
ozone mortality Arch Environ Health 23154-160,
1971
199 Trams, E G,C J Lauter, E A B Brown, and 0 Young
Cerebral cortical metabolism after chronic exposure to
03 Arch Environ Health 24 153-159, 1972
200 Vaughan.T R,Jr,L F Jennelle, andT R Lewis Long-
term exposure to low levels of air pollutants. Effects on
pulmonaryfunction inthebeagle Arch Environ Health
7945-50, 1969
201 Venmga, T S Ozone induced alterations in murme
blood and liver Presented at the 2nd International Clean
Air Congress, Washington, D C , December 6-11,1970
Paper MB-15E
202 Venmga, T Toxicity of ozone in comparison with
ionizing radiation Strahlentherapie 134 469-477,
1967
203 Warshauer, D , E Goldstein, P D Hoepnch, and W
Lippert Effect of vitamin E and ozone on the pulmonary
antibacterial defense mechanisms. J. Lab Clm. Med.
83 228-240, 1974
204. Watanabe, S., R Frank, and E Yokoyama. Acute effects
of ozone on lungs of cats. Amer. J Rev Respir Dis.
108 1141-1151, 1973
205. Wayne, L. G., and L A. Chambers. Biological effects of
urban air pollution V A study of effects of Los Angeles
atmosphere on laboratory rodents Arch. Environ.
Health 76.871-885, 1968.
206. Weibel, E R. Morphometry of the Human Lung.
Academic Press, Inc , New York, 1963.
207 Weissbecker, L., R. D Carpenter, P. C. Luchsinger, and
T. S Osdene In vitro alveolar macrophage viability.
Effect of gases Arch. Environ. Health 78-756-759,
1969.
208 Werthamen, S , P D Penha, and L.Amaral. Pulmonary
lesions induced by chronic exposure to ozone. Arch.
Environ Health 29 164-166, 1974.
209 White, A , P. Handler, and E. L Smith, eds. Principals of
Biochemistry Fifth Edition, McGraw-Hill Book Co., New
York, 1973. pp 409, 873.
210 Xmtaras, C ,B. L Johnson, C E. Clinch, R. E. Terrill, and
M F. Sobecki Application of the evoked response
technique in air pollution toxicology. Toxicol. Appl
Pharmacol S 77-87, 1966.
211 Yokoyama, E. A comparison of the effects of SOs, NOz,
and Os on pulmonary ventilation. Guinea pig exposure
experiments. Sangyo Igaku 77:563-568, 1969.
212. Yokoyama, E The effects of low concentrations of ozone
on the lung pressure of rabbits and rats Presented at
the 46th Annual Meeting, Japan Society of Industrial
Hygiene, April 1973. Paper 223.
213 Yokoyama, E The effects of Oa on lung function In:
Proceedings of the Science Lecture Meeting on Air
Pollution Caused by Photochemical Reaction Japan
Society of Air Pollution, Tokyo, March 1972 pp 123-
143.
214 Yokoyama, E On the maximal expiratory flow volume
curve of rabbit lung exposed to ozone Dihon Kyobu
Shikkan Gakkai Zasshi 72(Suppl)2, 1974.
215 Yokoyama, E , and R. Frank Respiratory uptake of ozone
in dogs Arch Environ. Health 25' 132-138, 1972.
216. Yokoyama, E., and I. Ichikawa. Study on the biological
effects of atmospheric pollutants (FY 1972-1975). In:
Research Report for Funds of the Environmental Agency
in 1974 Institute of Public Health, Department of
Industrial Health, Tokyo, 1974 pp 16-1 to 16-6.
217. Zelac, R. E., H. L. Cromroy, W E Bolch, Jr, B. G
Dunavant, andH A Bevis. Inhaled ozone as a mutagen.
I. Chromosome aberrations induced in Chinese hamster
lymphocytes. Environ Res 4-262-282, 1971.
218 Zelac, R. E , H. L Cromroy, W E. Bolch, Jr, B. G.
Dunavant, and H A. Bevis. Inhaled ozone asa mutagen.
II Effect on the frequency of chromosome aberrations
observed in irradiated Chinese hamsters Environ. Res.
4325-342, 1971
183
-------
9. CLINICAL APPRAISAL OF THE EFFECTS OF OXIDANTS
OCCUPATIONAL AND ACCIDENTAL EX-
POSURES TO OZONE
Schonbein,55 the discoverer of ozone, was
probably the first person occupationally exposed to
this gas. His description in 1851 of subjective
symptoms and basic characteristics of ozone was
very complete. Dadlez11 m 1928 found that 2940
A/g/m3 (1.5 ppm) ozone rendered the atmosphere
intolerable. He cites an investigation by D'Arsonval
who found that 780 /ug/m3 (0.4 ppm) ozone in the
atmosphere produced symptoms of discomfort and
irritation within about 30 min. after exposure
began. In 1931, Flury and Zernik14 described the
following symptoms as being characteristic of
increasing exposure to ozone: 920 fjg/m3 (0.47
ppm) causes distinct irritation of mucous
membranes, and 1840/^g/m3 (0.94 ppm) causes
sleepiness in 1 hr; at higher concentrations, ozone
causes increased pulse rate, sleepiness, and
prolonged headache. Ozone exposure may also
result in dyspnea (difficult breathing) and
pulmonary edema.
Kleinfeld et al.44 and Kleinfeld43 reported several
cases of severe ozone intoxication in welders using
a consumable electrode technique, which was
new at that time (1957). Three plants were
investigated, and in all cases, the ozone
concentration was monitored at the breathing
zone of the consumable electrode machine. In the
first plant, an ozone concentration of 490 /ug/m3
(0,25 ppm) was found. The workers had no
complaints, and clinical findings were
noncontributory. In the second plant, the ozone
concentration ranged from 590 to 1570 A/g/m3 (0.3
to 0.8 ppm). Two of the four welders complained of
chest constriction and throat irritation. Clinical
examination disclosed no abnormalities. In the
third welding plant, the ozone concentration was
17,990 A
-------
The following measurements of pulmonary
function were made: Vital capacity(VC),functional
residual capacity (FRC), maximal midexpiratory
flow rate (FEF 25 to 75), 0.75-second forced
expiratory volume (FEVors), and carbon monoxide
diffusing capacity (DLoo) at rest and at exercise. No
convincing evidence was found that functional
impairment developed in association with long-
term exposure to 390 to 590 /ug/ m3 (0.2 to0.3 ppm)
ozone in these seven smokers.
The available data on occupational exposures of
humans to ozone are summarized in Table
_. 10,11,14,38.43,44,61
TABLE 9-1. SUMMARY OF AVAILABLE DATA ON OCCUPATIONAL EXPOSURE OF HUMANS TO OZONE
10.11.14,38,43.44,61
Ozone,
cg/jn3
490
590-
1570
17,990
(peak
concen-
tration)
390
1570-
3330
Ozone,
ppm
0.25
03-
0.8
9.2
(peak
concen-
tration)
02
0,8-
1 7
Subjective complaints
None
Chest constriction
and throat irritation
in 2 to 4 subjects.
Severe headaches,
throat irritation,
and lassitude in
7 or 8 subjects.
Cough, choking,
dyspnea, and
substernal oppres-
sion in 3 of 8
subjects
Very severe head-
ache, dyspnea.
Substernal oppres-
sion in 1 of 8
subjects.
Dry mouth and throat.
irritation of nose
Clinical findings
attributed to ozone
Mone
None
By X-ray, molted
densities m both
lungs, clearing
after 9 days.
-Severe pulmonary
edema. By X-ray
penbronchial in-
filtration consis-
tent with penbron-
chial pneumonia.
None
None
Measurements
of pulmonary
function
None
None
None
None
None
None
None
Other comments
Negligible nickel
carbonyl and
oxides of nitro-
gen, Tnchloro-
ethylene de-
greaser located
50 ft from
welding area
Tests for
phosgene, negative
Concentration
of tnchloro-
Reference
Klemfeld
et al •"
Klemfeld
et al ""
Klemfeld
el al"
Klemfeld"
Challen
et al 10
Challen
et al.10
and eyes, detection
of disagreeable smell
in 11 of 14 subjects.
390- 0.2- Detection of irritating
590 0.3 odor, soreness of eyes
and dryness of mouth,
throat, and trachea in 1
of 7 subjects.
None VC decreased
in 3 of 7
subjects FRC
decreased in 2 of
7 subjects DLco
decreased in 1 of
7 subjects
ethylene up to 238
ppm found
All decreases in Young etal61
pulmonary func-
tion measurements
were small. All
subjects were
smokers
780
920
1,840
5,900
21,950
0.4
0.47
094
30
11 2
Discomfort and irri-
tation in about 30 mm
Distinct irritation of
mucous membranes
Coughing, irritation.
and exhaustion, within
V/i hr.
Sleepiness within
1 hr.
Profuse perspiration,
continual coughing;
decreased blood pres-
sure; weak, accelerated
pulse.
None
None
None
None
None
None
None
None
None
None
Dadlez"
(D'Arsonval)
Flury and
Zernik"1
Flury and
Zernik'4
Flury and
Zernik'4
Severe symptoms Kelly and
almost caused Gill3fl
subject to lose
consciousness
185
-------
CONTROLLED STUDIES OF HUMAN HEALTH
EFFECTS
Studies conducted before 1970 may be
summarized as follows.
Under experimental conditions, repeated
exposure of humans to ozone produced the
following effects:
1. No apparent effects on pulmonary function
were observed at concentrations up to 390
/jg/m3 (0.2 ppm)for3 hr/day, 6days/week
for 1 2 weeks.6
2. The threshold level at which nasal and
throat irritation will occur appears to be
about 590 /jg/m3 (0.3 ppm).61
3. Concentrations of 980 /jg/m3 (0.5 ppm)
have caused a 20-percent decrease in the
1-sec forced expiratory volume (FEVi)
observed after 8 weeks of intermittent
exposure (3 hr/day, 6 days/week); this
change returned to normal during the post-
exposure period of 6 weeks.6
In Table 9-2,
6,20,25,31 ,36,60
the results of
experimental appraisal of short-term exposures to
ozone are shown as follows.
1. Some subjects have shown small but
statistically significant increases in airway
resistance after 1 hrof (presumably resting)
exposure to 200 /Kj/m3 (0.10 ppm) ozone.20
However, the significance of this finding is
unclear, since increases in resistance were
small, and since the same subjects showed
smaller, nonsignificant increases in
resistance after exposure to 780 /jg/m3
(0.4 ppm) ozone than they had shown after
exposure to 200 /jg/m3 (0.1 ppm).
2. Concentrations of 200 to 780 /ug/m3 (0.1 to
0.4 ppm) for 1 hr have been shown to
increase airway resistance (Raw) slightly,
but adequate information for this
concentration range is lacking.20
3. Exposure to a concentration of 1960^g/m3
(1.0 ppm) for periods of 1 to 2 hr produced
changes in pulmonary function. These
were increased airway resistance,
decreased vital capacity, decreased carbon
monoxide diffusing capacity, and
decreased forced expiratory volume.20
4. One individual was unable to tolerate
concentrations of 1 960 to 5880/jg/m3 (1.0
to 3.0 ppm) over a period of about 2 hr.
Extreme fatigue and lack of coordination
were experienced.25
5. Concentrations of about 17,640/Kj/m3 (9.0
ppm) produced severe pulmonary edema
and possible acute bronchiolitis.36
Details of selected controlled human studies
reported after 1970 are discussed below.
Bates et al.3'4 measured significant changes in
lung function: A decrease (a) in maximal flow rate
at 50 percent of the vital capacity (FEF 50 percent)
and (b) in maximal transpulmonary pressure
(PstTLC); and an increase in total pulmonary
resistance (RL) in 10 normal male subjects aged 23
to 53 years (including two smokers) exposed to
pure ozone at 1470 /jg/m3 (0.75 ppm)for 2 hr. Two
of the three subjects who exercised intermittently
at twice the resting volume showed accentuated
effects. In a separate study by Hazucha et al.35 on
the effects of short-term exposure, significant
decreases in forced vital capacity (FVC), maximal
midexpiratory flow rates, FEF 25 to 75 percent,
FEVi, and FEF 50 percent, and increases in closing
capacity (CC) and residual volume (RV) were found
in 1 2 normal young males (including six smokers)
exposed to pure ozone at 1470 and 730 /jg/m3
(0.75 and 0.37 ppm) for 2 hr during alternating rest
and exercise periods. The higher concentration
affected smokers more than nonsmokers, whereas
at the lower concentration, the reverse wasfound.
In these two studies, most subjects complained of
cough, chesttightness, and substernal soreness. A
few also had pharyngitis, dyspnea, and wheezing.
In another study,2'32 all mean dynamic lung
function parameters in 10 nonsmokers and 10
smokers showed a progressive decrease with
continuation of exposure, returning close to pre-
exposure levels after 2 hr of recovery (Figure 9-I).32
Average minute ventilation changed only slightly
during the exposure, although the frequency of
breathing increased, and the tidal volume
decreased steadily with the duration of exposure
(Figure 9-2).
Bates and Hazucha2 and Hazucha and Bates33
reported a marked decrease in pulmonary function
among healthy subjects performing light exercise
while exposed to 730 /jg/m3 (0.37 ppm) of ozone
and 1000 /jg/m3 (0.37 ppm) of sulfur dioxide for 2
hr. Throat irritation, coughing, and chest pain were
also experienced. These effects, as well as the
decrease in lung function, occurred after exposure
to the mixture for periods as short as one-half to 1
hr and persisted for several hours after termination
of the exposure. Ozone concentrations were
measured by a Mast meter with CrO3 scrubber for
S02 removal, and S02 levels were monitored with
a conductimetric analyzer. To investigate further
186
-------
the possible interactive effects of S02 + 03 and the
mechanism of interaction, Bell et a!.5 initiated a
series of controlled human-exposure studies. The
studies of Bell et al, suggest less severe acute
toxicity of mixtures of (730 A/g/m3)0.37 ppm ozone
and (1000 A/g/m3) 0.37 ppm SO2 than Bates and
Hazucha had previously observed. There are
several possible explanations: (1) The Los Angeles
residents had undergone biological adaptation to
chronic pollutant exposure; (2) the Montreal
subjects used in the experiment were not as
reactive to the SOz + Oa mixture when tested in Los
Angeles as when tested in Montreal, possibly
because of significantly higher concentrations of
the sulfur compounds in the respirable aerosols
that can form in the Montreal chamber. The
background air in the Rancho Los Amigos studies
was highly purified, whereas the sulfur aerosol
concentration in the Montreal chamber during the
mixed-gas exposures was similar to 2-hr sulfate
aerosol concentrations during the worst pollution
episodes in urban areas. The effects observed in
the Montreal chamber probably more nearly
parallel the health effects that might result during
TABLE 9-2. SUMMARY OF DATA ON HUMAN EXPhRIMENTAL EXPOSURE TO OZONE BEFORE 19701
6,20,25,31,36,60
O/one.
9,800-
19,600
2,940-
3,920
390
980
1,180-
1,570
O/one, Length of No and sex Subjective
ppm exposure of subjects complaints
5-10 Not available 3 male Drowsiness,
headache
1 5-2 2 hr 1 male CNS depression.
lack of coordina-
tion, chest pain,
cough for 2 days,
tiredness for
2 weeks
02 3 hr/day 6 male None
6 days/week
for 1 2 weeks
05 3 hr/day 6 male No irritating
6 days/week symptoms, but
for 12 weeks could detect
ozone by smell
06-08 2 hr 10 male Substernal sore-
1 female ness and tracheal
irritation 6 to 12 hr
after exposure.
disappearing with-
in 12 to 24 hr in
10/1 1 subjects
Measurements of
pulmonary function
None
VC Decreased 13%,
returned to normal
in 22 hr FEV30.
Decreased 168%
after 22 hr MBC
Decreased very
slightly3
VC No change
FEVi o No change
VC Slight but not
significant de-
crease toward end
of 12 weeks
Returned to normal
within 6 weeks after
exposure
DLco Mean decrease
of 25% (11/11
subjects) VC Mean
decrease of 10%,
which was signifi-
cant (10/10
subjects) FEV07s *
40 Mean decrease
of 1 0%, which was
significant
Other
comments Reference
Measurement of Oa Jordan and
probably inaccurate Carlson36
Griswold
et al"
0 66 upper respir- Bennet6
atory infections/
person in 1 2 weeks.
Cf control group
had 095 in the
same period, 0 80
upper respiratory
infections/person
in 1 2 weeks
0 80 upper respir-
atory infections/
person in 1 2 weeks
Young etal 60
(10/10 subjects),
FEF 25-75 Mean
decrease of 15%,
which was not sig-
nificant Mixing
efficiency No
change (2/2 sub-
jects)
Airway resistance
Slight increase, but
within normal limits
Dynamic compli-
ance No change
(2/2 subjects)
Young etal G0
187
-------
TABLE 9-2,(cont'd). SUMMARY OF DATA ON HUMAN EXPERIMENTAL EXPOSURE TO OZONE BEFORE
•1 g -7^6,20 .25, 31 ,36 .60
Ozone,
//g/m3
Up to
7,800
Ozone,
ppm
Up to
40
Length of
exposure
1010 30
mm
No and sex
of subjects
Subjective
complaints
11 Headache, short-
ness of breath,
lasting more than
1 hr
200
780
01 1 hr
04 1 hr
4 male
4 male Odor
1,180
06 1 hr
4 male Odor
1,960
10 1 hr
4 male Throat irritation
and cough
Measurements of
pulmonary function
Other
comments
VC Mean decrease Only 5/11 subjects Hallett31
of 16.5% (4/8
subjects showed
decrease > 10%)
FEVto. Mean
decrease of 20%
(5/8 subjects
showed decrease
> 10%, FEF 25-75:
Mean decrease of
10.5% (5/6 subjects
showed a decrease)
MBC Mean
decrease of 12%
(5/8 subjects
showed a decrease)
DLco. Decrease of 20
to 50% in 7/11
subjects; increase
of 10% to 50% in
4/11 subjects.
Airway resistance
Mean increase of
3.3% at 0 hr after
exposure (1/4 sub-
jects showed an in
crease of 45%).
Airway resistance
Mean increase of
3.5% at 0 hr after
exposure; (1 /4
subjects showed an
increase of 60%),
mean increase of
12 5% 1 hr after
exposure).
Airway resistance
Mean increase of
5 8% at 0 hr after
exposure (1/4
subjects showed an
increase of 75%);
mean increase of
5% 1 hr after
exposure.
Airway resistance
Mean increase of
193% at 0 hr after
exposure (3/4
subjects showed an
an increase of >
20%); mean
increase of 5% 1 hr
after exposure
tolerated dose for
full 30 mm
Wide variations in
DLco.
One subject
had history of
asthma and ex-
perienced he-
moptysis 2 days
after 1 ppm.
Goldsmith
and Nadel20
Goldsmith
and Nadel20
flMBC - maximum breathing capacity
smog episodes in regions with high oxidant and
sulfur pollution than do the effects observed in the
Rancho Los Amigos chamber.
Studies by Hackney et al 26'2B'3D USed basically
the same rest or intermittent exercise protocol as
that developed in the Montreal laboratory. To
simulate summer exposure in the Los Angeles
southern coastal basin, the additional stress of
heat (31°C, or88°F at 35 percent relative humidity)
was included, and the same subjects were exposed
several times to either ozone or mixtures of ozone
with other pollutants. Careful attention was given
to environmental control, pollutant monitoring and
generation, and subject selection. Four male
188
-------
subjects, aged 36 to 49 years and judged by
subjective criteria to have normally reactive
airways (i.e., with no history of cough, chest
discomfort, or wheezing), completed this protocol.
As assessed by clinical response and measures of
respiratory, cardiac, and metabolic functional
change, no obvious effects were noted after
exposure for 4 to 5 hr to 980 fjg/m3 (0.5 ppm)
ozone, to a combination of 0.5 ppm ozone and 560
fjg/rr\3(Q.3 ppm) nitrogen dioxide, or to a mixture of
0.5 ppm ozone, 0.3 ppm nitrogen dioxide, and 35
fjg/m3 (30 ppm) carbon monoxide. A point of
interest is that in a group of unreactive subjects,
the same investigators observed substantial
decrements in lung function after only 2 hr of
exposure to 980 /vg/m3 (0.5 ppm) ozone and
intermittent light exercise. Another group of four,
aged 29 to 41 years, who had previously
experienced clinical bronchospasms and were
judged by subjective criteria to have hyper-reactive
airways, developed clinical discomfort and were
unable to complete the protocol. Exposed to ozone
at 980 A
-------
group of five subjects, aged 27 to 41 years, was
exposed to 730 /jg/m3 (0.37 ppm) ozone for 2 hr.
No important changes with exposure were found
in most physiologic measures. The experimental
design in all these studies does not permit
statistical analysis of the combined data; therefore,
no firm conclusions can be drawn concerning
dose-response relationships.
0-5 1-0 1-5 2-0 2-5 3-0 3-5 4-0
• EXPOSURE "-H ' RECOVERY
TIME, hours
Figure 9-2. Changes with time of minute ventilation, respiratory rate, and tidal volume in ozone-exposure (0.75)
and subsequent recovery period in exercising nonsmokers32 (Note: Heavy bars on the abscissa mark 15-min periods of
light exercise.)
190
-------
In studies by Kerr et al.,39 10 smokers and 10
nonsmokers, aged 21 to 60 years, were exposed to
ozone at 980 jug/m3 (0,5 ppm) for 6 hr in an
environmental chamber. During this period, they
engaged in two 1 5-min medium exercise sessions
(100 watts) on a bicycle ergometer. Subjects who
experienced typical symptoms were, in general,
the ones who developed objective evidence of
decreased pulmonary function. The most
significant changes from control values for the
group as a whole (20 subjects) after ozone
exposure were observed in several pulmonary
function tests—specific airway conductance
(SGaw), pulmonary resistance (RL), and FVC. No
significant change was observed with respect to
diffusing capacity (DLCo), static lung compliance
(Cst), or the various tests derived from the single-
breath nitrogen elimination rate. When the
smokers were considered as a separate group, no
significant decrease in pulmonary function was
observed. The group of nonsmokers considered
separately showed a significant decrease in
dynamic compliance as well as in those
parameters noted for all 20 subjects considered
together. Prolonged exposure of four subjects to
0.5 ppm for 10hr further decreased specific airway
conductance.
Folinsbee et al.17 tested the response of 28
subjects after ozone exposure to three stages of
ergometer exercise with loads adjusted to 45, 60
and 75 percent of maximal aerobic power. The
subjects were exposed to ozone at 730, 980, or
1470 jug/m3 (0.37, 0.50, or 0.75 ppm) for 2 hr, at
rest or while exercising intermittently. Subjects
had 15 min of rest alternated with 15 min of
exercise at a workload sufficient to increase
ventilation by a factor of 2.5. At submaximal
exercise, neither oxygen consumption nor minute
ventilation was significantly altered after ozone
exposure at any concentration. The primary
response was an alteration of exercise ventilatory
pattern. An increase in breathing rate (r=0.98) and
a decrease in tidal volume (r=0.91) correlated well
with the dose of ozone, calculated as the volume of
ozone inspired during exposure. It was concluded
that through its irritant properties, ozone modified
normal ventilatory response to exercise and that
this effect was dose-dependent.
In a second study, Folinsbee and associates16
assessed the effect of exposure to 1470 jug/m3
(0.75 ppm) ozone on the cardiac and pulmonary
function of maximally exercising, healthy young
adult males. During 2-hr exposures, subjects
alternately rested for 1 5 min and lightly exercised
for 15 min at a workload of about 50 watts. In
random order, each subject also underwent a 2-hr
control exposure to filtered air. At the conclusion of
all exposures, subjects exercised maximally until
exhausted. Values of the following parameters,
measured during maximal exercise, were
significantly lower after ozone exposure than after
filtered air exposure: Maximum work load attained,
heart rate, minute volume, tidal volume, and
oxygen uptake. Respiratory frequency and the ratio
of tidal volume to FVC during maximal exercise did
not differ significantly following ozone versus
filtered air exposures. The authors concluded that
the decrease in maximal exercise performance
after ozone exposure probably stemmed from
respiratory rather than cardiac factors. In their
minds, the most likely mechanism underlying the
observed results is the stimulation by ozone of
irritant receptors in the lung. Such stimulation in
turn produces restriction of inspiratory reserve
volume, reduction of tidal volume during maximum
exercise and a modest increase in airway
resistance.
A more specific study of airway irritability was
conducted by Golden et al.19 They showed that
after a 2-hr exposure to 1180 jug/m3 (0.6 ppm)
ozone, bronchial reactivity (as assessed by
response to histamine challenge) increased.
Seven days following the exposure, bronchial
sensitivity was still elevated, and in some subjects,
hyperirritability persisted for up to 3 weeks. The
authors believe that ozone damaged the airway
epithelium and sensitized bronchial irritant
receptors.
DeLucia and Adams'2 studied the effects of
graded exercise on lung function and blood
biochemistry in six men after 1 hr of exposure to
290 jug/m3 (0.15 ppm) and 590 fjg/m3 (0.30 ppm)
of ozone via a mouthpiece. The workloads were 25
percent, 45 percent, and 65 percent of each
individual's maximum oxygen uptake (Voz max).
Ventilation volume and Vo? were unaffected by
even the most severe exposure/exercise
protocols. However, most subjects demonstrated
signs of toxicity (symptoms such as congestion,
wheezing, and headache) during the most
stressful protocols. In addition, vital capacity,
forced expiratory volume at 1 sec, and
midmaximum flow rate decreased significantly
after inhalation for 1 hr of 590jug/m3(0,30ppm)at
65 percent Vo2 max. Discernible, though not
statistically significant, changes in respiratory
191
-------
pattern were also observed after exposure to 290
yug/m3 (0.1 5 ppm)ozone andexercise at65 percent
Vo2 max. This study emphasizes the importance of
actually quantifying the dose rather than merely
defining the exposure level, as well as quantifying
the differential effects that may result from mouth
breathing compared to nose breathing.
The combined effects of different levels of
exercise, heat stress, and ozone on lung function of
14 males were studied by Folinsbee et al.15
Subjects were exposed to 980 yug/m3 (0.5 ppm)
ozone for 2-hr periods under four environmental
conditions: (1) 25°C (77°F), 45 percent RH; (2)
31 °C (88°F), 35 percent RH; (3) 35°C (95°F), 40
percent RH; and (4) 40°C (104°F), 50 percent RH.
One30-min exercise period(40percent of Vo2 max)
was included in every 2-hr exposure run. The
results of this study show that besides well
established physical factors such as concentration
of ozone, duration of exposure, and intensity of
exercise, the extent of lung functional response
will depend also on timing of exercise into the
exposure period, lapse time between end of
exercise and testing period, relative humidity, and
ambient air temperature
Silverman et al.,66 using the Montreal exposure
protocol (intermittent exercise, 730 yug/m3 [0.37
ppm] or 1470 yug/m3 [0.75 ppm] ozone for 2 hr),
basically confirmed the results obtained previously
by Hackney et a|.26'2830 and Bates and Hazucha.2'19
Moreover, they showed a high linear correlation
between the effective dose of ozone (calculated as
concentration times the relative volume of
ventilation) and percent changes in lung function
measurements (r = 0.70 to 0.95). The reported data
indicate that for a given effective dose, exposure to
a high concentration for a short time is more
effective than a longer exposure to a lower
concentration.
Knelson et al.46 exposed 22 male volunteers,
aged 19 to 27, to ozone for up to 4 hr at 780yug/m3
(0 4 ppm). Subjects were sedentary during
exposure except for two 15-min exercise periods
on a bicycle ergometer at 700 kg-m/min exercise
that about doubled the heart rate and quadrupled
the minute ventilation. After 2 hr of ozone ex-
posure, there was a significant change (p< 0.05) in
FVC, midmaximal expiratory flow (MMEF), and
airway resistance (Raw). Several other measures
(FEVi, V50, and V25) were lower after 2 hr of
exposure, but the statistical significance was
borderline. However, after 4 hr of exposure, all
flow measures were significantly decreased,
compared with controls. After4hr, Raw increased,
FVC decreased further, and FENA decreased
significantly. Residual volume, functional residual
capacity, and total lung capacity (TLC) did not
change as a result of the ozone exposure. Ketcham
et al.,40 employing the same laboratory facilities as
the previous investigators, also demonstrated
significant deterioration of pulmonary function in
30 subjects. The average FEF 25 to 75 percent,
which is the most sensitive test, decreased by
almost 30 percent from the control value after 2 hr
of exposure to 1180 yug/m3 (0.6 ppm) ozone. All
other dynamic tests were also decreased
significantly. These changes were accompanied by
increased RV and FRC and slightly decreased TLC.
Kagawa and Toyama 37 have reported on the
results of limited studies involving four normal
male subjects exercising while exposed to ozone at
1760 yug/m3 (0.9 ppm) for 5 mm. A significant
decrease in SGaw was found during exposure and
after 5 min of recovery.
Von Niedmg et al.59 studied the effects of 196
yug/m3 (0.1 ppm) of ozone for 2 hr in 12 healthy
males aged 24 to 38 years. Arterial oxygen tension
(PaQj) decreased significantly by 7 mm Hg from a
mean of 84.6 + 1.3 mm Hg to 77.6 ± 1.9 (p<0.01)
and returned to the initial level at the end of the 1 -
hr post-exposure period (84.3 ± 1.8 mm Hg). The
alveolar-arterial gradient (AaDrjz) increased with
decreasing Pa^ (p<0.01). Airway resistance was
significantly higher at the end of the 2-hr exposure
than at the start. The subjects' airway resistance
had not returned to normal 1 hr after exposure.
Thoracic gas volume (TGV) did not change in the
course of the experiment. Limitations in this study
include the use of nonstandard measurement of
flow resistance (Rt) and the use of artenalized
capillary blood for PQ, measurement. Hence, until
confirmed, these studies must be interpreted with
caution.
Several investigators have reported apparent
adaptation to pulmonary effects of ozone on
repeated or chronic exposure. A discussion of
these studies, along with a recent controlled
exposure experiment, has been reported by
Hackney et al.29 Six men with respiratory hyper-
reactivity were exposed to 980 yug/m3 (0.5 ppm)
ozone for 2 hr each day during 4 successive days.
One subject showed little measurable response.
The other five subjects showed decrements in lung
function (generally on exposure days 1 to 3) that
were largely reversed on day 4. These results
suggest that some humans do not continue to
192
-------
experience the same decrements in lung function
after repeated exposures to ozone. It is not known
whether adaptation to other adverse effects of
ozone exposure occurs.
Hackney and associates27 also compared the
response to ozone exposure of four southern
Californians with that of four Canadians, whose
cumulative ambient ozone exposure appeared to
have been less than that of the Californians. In
Hackney's southern California laboratory, all eight
subjects underwent 2 hr of exposure to 730/ug/m3
(0.37 ppm) ozone and intermittent light exercise.
The investigators observed no statistically
significant changes in lung function. However, the
Canadians showed discernible post-exposure
decrements in most lung function parameters
measured, whereas the Californians did not. The
Canadians also showed statistically significant
changes \rtin //fro erythrocyte fragility, erythrocyte
acetylcholinesterase activity, and serum vitamin E
level. Among Californians, erythrocyte acetyl-
cholinesterase activity was the only blood
parameter showing post-exposure changes. To the
investigators, adaptation of southern Californians
to the effects of ozone exposure was the most
plausible explanation for the observed results.
However, selective migration of ozone-sensitive
pec Die away from southern California could not be
wholly ruled out.
I- a similar 4-day, 2-hr exposure protocol,
Me Tier et al.52 exposed six subjects to 730 ,ug/m3
(0.'7 ppm) ozone. Although conventional tests did
no* reveal any functional abnormalities, more
sensitive nitrogen-clearance measurements
shewed that in hyper-reactive subjects, the
spe.cific distribution of ventilation had been
alt. red.
Linn et al.46 studied the effects of ozone
exposure on symptomatology, respiratory
physiology, and blood biochemistry in 22
physician-diagnosed asthmatic subjects(20 males
and two females) living in the Los Angeles area.
Persons with marked respiratory disability were
excluded from study. Study participants varied in
age from 19 to 59 years and covered a clinical
range from slight wheezing to marked abnormality
in forced expiratory performance. All but six used
medication regulary.
Each subject was studied at the same time of day
on 3 successive days in the Rancho Los Amigos
exposure chamber. At the beginning of each day's
study, baseline lung function measurements were
obtained with subjects breathing purified air.
Following this, subjects exercised lightly and
rested over alternate 15-min periods for 2 hr.
Exercise work loads were chosen to double each
subject's resting minute ventilation (light
exercise). In all studies, temperature and relative
humidity were controlled to 31 ± 1 °C and 35 ± 4
percent, respectively. Subjects were instructed to
take oral medications on their regular schedules
but to refrain from using inhaled bronchodilators
during each 3-day study period.
On the first day, subjects were exposed only to
purified air. On the second day, they were exposed
to just enough ozone to permit detection of its odor
at the beginning of the exercis^ -rest period, and to
purified air for the rest of the period. On the third
day, they were exposed to 0.20 to 0.25 ppm ozone
throughout the period. At the end of each day's
study period, the lungfunctiontestsof thesubjects
were repeated, their symptoms were assessed,
and they were bled for measurement of
hematologic and biochemical parameters.
To quantitate the effect of the 3-day study
protocol on the parameters measured, the
investigators repeated the entire protocol, with no
ozone exposure, using a subsample of 14 persons
(1 2 men and the 2 women) from the original group
of subjects. On the average, subjects underwent
the 3-day control study 10 months after the
original study (range 1 to 23 months).
In all data analyses, the mean results of the
whole group or the subsample were compared
across different days of study. No formal attempt to
identify or characterize unusually sensitive
individuals was reported.
In the whole group of 22 subjects, the within-day
difference between pre- and post-exposure lung
function measurements differed significantly
between the oxidant-exposure day and any other
day for only one of nine measured parameters. This
was total lung capacity (TLC), whose mean value
increased by 0.06 liter on the odor-sham day and
decreased by 0.10 liter on the oxidant-exposure
day (p < 0.05). In the subsample of 14 subjects, no
statistically significant differences in lung function
between the ozone-exposure day and any other
day of the first 3-day protocol was observed. For
this group, between-day differences in lung
function were slightly larger during the second 3-
day (control) protocol than in the first.
For the whole group, ths mean symptom score
was higher (but not significantly so) on the ozone-
exposure day than on the other 2 days, and it was
virtually identical on the other 2 days. For the
193
-------
subsample in the first 3-day (exposure) protocol,
the mean symptom score was also highest by a
small margin on the ozone-exposure day.
However, the highest daily score in the subsample
occurred during the 3-day control protocol.
Ozone exposure was more strongly associated
with blood biochemical changes than with
symptoms or lung function changes. For the whole
group, mean values of the following parameters
differed signif ica ntly (p < 0.05) between the ozone-
exposure day and at least one of the other 2 days
(directions of difference on the ozone-exposure
day are given in parentheses): total hemoglobin
(lower), red cell fragility in HjC^ (higher), reduced
glutathione (higher than odor-sham days, lower
than clean-air days), acetylcholmesterase (lower),
glucose-6-phosphate dehydrogenase(higher), and
lactate dehydrogenase (higher). In the subsample
of subjects in the 3-day exposure protocol, the
same types of significant differences between
ozone-exposure days and other days were
observed, with the exception that no ozone-related
difference in reduced glutathione occurred. In the
3-day control protocol, the only biochemical
parameter showing significant differences
between days was acetylcholinesterase.
The results of this study suggest that, in a
reasonably representative sample of adult
asthmatics, short-term exposure to realistic ozone
levels (0.20 to 0.25 ppm) produces no measurable
adverse effect on lungf unction. On the other hand,
the results suggest a moderate tendency toward
increased symptom frequency in asthmatics at the
ozone level used, and they strongly suggest that
ozone at this level can alter certain biochemical
and hematologic parameters.
The changes that Linn et al.46 have associated
most confidently with ozone exposure are changes
of the least certain clinical significance. Since
most of the observed biochemical and hematologic
effects relate directly or indirectly to the oxygen-
carrying capacity of the blood, they might be
unusually important in groups, such as
asthmatics, with compromised pulmonary
function. Also, as the authors mention, these
findings raise the possibility that asthmatics may
react biochemically at lower ozone concentrations
than do normal persons. However, the observed
changes were small when compared to the normal
inter-individual variability of the parameters
measured, and they were considerably smaller
than would occur with obvious clinical disease.
Thus the question as to whether these changes
represent harmful effects in asthmatics remains
open.
As mentioned, Linn et al.46 reported only group
mean results Thus their findings do not touch on
the question of whether certain individual
asthmatics may be particularly sensitive or
resistant to ozone concentrations of 0.20 to 0.25
ppm Finally, their findings do not allow inference
as to what the effects of ozone might be in
childhood asthma.
The effects of sequential exposures separated by
1 day to 2 months, in combination with different
concentrations of ozone, have been investigated by
Hazucha et al.34 The 2-hr ozone exposure protocol
for 16 subjects, divided into four groups, was as
follows:
1. Group A. 780 /jg/m3 (0.4 ppm), 1- to 2-
month delay, 390 /jg/m3 (0.2 ppm), 1 -day
delay, 780 fjg/m3 (0.4 ppm) ozone;
2. Group B: 390 /ug/m3 (0,2 ppm), 1-day
delay, 780 /jg/m3 (0.4 ppm), 1 - to 2-month
delay, 780 /jg/m3 (0.4 ppm);
3. Group C; 1180 Aig/m3 (0.6 ppm), 1- to 2-
month delay, 390 jug/rn3 (0.2 pprn), 3-day
delay, 1180 /jg/m3 (0.6 ppm);
4. Group D: 11 80 /jg/m3 (0.6 ppm), 1 - to 2-
month delay, 780 /jg/m3 (0.4 ppm), 3-day
delay, 11 80 /jg/m3 (0.6 ppm).
From all four groups, only group D showed clear
signs of adaptation (i.e., response to the third
challenge exposure was less marked than a
response observed earlier after exposure to the
same concentration). Although the size of the
groups was small, the results obtained indicate
that the extent of changes induced by the
challenge exposure depend primarily on the ozone
concentration during pre-exposures and the time
interval between preconditioning and challenge
exposures.
In most reported studies,312'16'26'32'35'45'56 an
association between symptoms and changes in
lung function was usually found. Ozone-induced
defects in function were not usually found in the
absence of definite symptoms of ozone-induced
respiratory irritation.
The newer experimental studies described
above2'12'17'19'32'35'37'39'56 show that significant
adverse changes in lung function occur in humans
at ozone concentrations of 725 /jg/m3 (0.37 ppm)
and higher. Some limited studies show evidence of
human health effects of exposure to pure ozone at
concentrations as low as 490 /jg/m3 (0.25 ppm)
(see Figure 9-3).2832 The preliminmary findings of
194
-------
Von Nieding et al.59 suggest decrements in lung
function at 1 96 fjg/m3 (0.1 ppm) exposure for 2 hr.
Hematology
Although the toxic effects of oxygen on red blood
cells (RBC) have been known for a long time, it was
not until recently that injurious effects of a much
stronger oxidant, ozone, have been investigated.
Initial studies concerning ozone toxicity on RBC in
animals showed that ozone and oxygen share a
great number of pharmacological and toxicological
features. The striking similarity between the
toxicity of these two gases has attracted
considerable interest among both researchers and
clinicians and an increasing number of articles,
including a review,48 have been published on this
problem.
IN VITRO STUDIES
The RBC constituent most sensitive to ozone
damage seems to be the membrane. The increased
breakdown of unsaturated fatty acids in the red cell
membrane was observed after exposure to 780
fjg/m3 (0.4 ppm) ozone for 2 hr. The degree of
unsaturation of the acids was positively correlated
with the extent of lipid peroxidation.1 The same
intensity of exposure significantly reduced
acetylcholinesterase activity and decreased
neuraminic acid levels in cell membranes. Sub-
sequent incubation of the exposed cells in plasma
resulted in complement-mediated RBC membrane
damage as measured by acid, sucrose, and insulin
hemolysis tests. Since such a reaction is
characteristic of RBC in paroxysmal nocturnal
hemoglobinuria (PNH) disorder, Goldstein et al.
hypothesized that the ozone-induced changes in
the red cell membrane have produced PNH-like
RBC's.21 22 In addition, a recent study from the
same laboratory showed that exposure of RBC
under similar experimental conditions also caused
a loss of native protein (tryptophan) fluorescence in
the cell membrane.21 With regard to the above
studies, it can be argued that the concentration of
ozone used was unrealistically high, but serious
disturbances in cell chemistry caused by such
concentrations can be detected much more easily
than the subtle changes induced by lower levels.
Increased sophistication and sensitivity of various
tests have allowed considerable reduction of ozone
concentrations in hemotological studies. Recently,
Kindya and Chan 42 demonstrated that 3-rnm
exposure of 1.0- ml sample of erythrocyte mem-
brane fragments to 12 mmol ozone significantly
decreased ouabaine-sensitive adenosine
10.0
-30.0 - '
0.2 0.4 0.6 0.8
OZONE CONCENTRATION (BY NEUTRAL BUFFERED Kl METHOD), ppm
Figure 9-3. Dose-response curves for Los Angeles and Montreal subjects.3226
195
-------
triphosphatase (ATPase) activity. Since this
enzyme plays a major role in maintaining
intactness of erythrocytes, the investigators postu-
lated that such inactivation of the activated ATPase
in the cell membrane may lead to increased
osmotic fragility and sphering of RBC. It is of
interest to note that both cysteine and ascorbic
acid, when added to the red cell membrane
fragments before exposure, afforded protection to
the ATPase against ozone.
In recent years, a number of compounds have
been studied as to their antioxidant effects on
various enzyme systems. Goldstein and Levine23
demonstrated that p-aminobenzoic acid inhibited
the ozone-induced loss of activity of acety-
cholmesterase in red cell membranes. Menzel et
al.60 showed that addition of vitamin E (a-
tocopherol) or d-a-tocopheryl acetate prevented
formation of Heinz bodies formed in RBC by
ozomdes, which are potent indicators of lipid
peroxidation in cellular membranes.49
IN VIVO STUDIES
Brmkman and Lamberts7 were the first to report
that inhalation of ozone interfered with oxygen
exchange in the capillaries of skin. In a later study,
Brmkman et al.8 demonstrated potentiation of
erythrocyte sphering in subjects exposed to 490
/ug/m3 (0.25 ppm) ozone for periods of 30 to 60
mm. More than 10 years later, Buckley et al.9
employed a comprehensive battery of biochemical
tests to investigate the toxic effect of ozone to RBC
of subjects exposed to 980/7g/m3(0.5 ppm) ozone
for 23/4 hr. The observed changes were similar to
the effects reported previously in either animal or
human blood in vitro studies. They demonstrated a
significant increase in RBC fragility. Lactate
dehydrogenase (LDH) and glucose-6-phosphate
dehydrogenase (G-6-PD) enzyme activities were
increased, while acetylcholinesterase activity and
reduced glutathione (GSH) were depressed. Cell
membrane glutathione reductase (GSSRase)
showed no change in activities, whereas serum
GSSRase activities were decreased. The post-
exposure level of serum vitamin E was significantly
higher as was lipid peroxidation. Clearly these
results indicate that ozone or its by-products can
induce changes across the alveolo-capillary
barrier.
Although the presented data display a spectrum
of ozone interference with biochemical
mechanisms, the physiological significance of all
these studies remains to be established. To
evaluate clinical implications of the ozone-induced
cellular damage, quantitative studies at realistic
concentrations are needed.
Mutagenesis
The work described in the section on chrom-
osomes in Chapter 8 of this document prompted
investigators to evaluate the potential for ozone to
cause chromosomal aberrations. Merz et al.51
reported results from six men exposed to 980
/ug/m3 (0.5 ppm) ozone, two of them for 6 hr and
the other four for 10 hr. Blood was obtained from
the first two subjects before and immediately after
exposure. The other four subjects had blood drawn
before exposure, immediately after, and then at 2
and 6 weeks following exposure. The lymphocytes
were cultured and prepared for chromosomal
analysis following standard procedures. No true
chromosomal-type abnormalities were observed
except forone instance in one individual. However,
there were lesions that the authors interpreted as
achromatic, unrepaired single-strand breaks, and
chromatid deletions. Although such aberrations
were seen in some of the pre-exposure
preparations, the authors reported an increased
frequency of abnormalities 2 weeks post-
exposure, with reversion to lower frequency 6
weeks post-exposure.
Because of the observations of Merz et al.,51 EPA
conducted a study to evaluate the potential human
mutagenic effects of ozone in a more rigorous
fashion. Thirty nonsmoking, normal, young male
volunteers breathed ozone at 780/7g/m3(0.4ppm)
during 4 hr in a controlled-environment labor-
atory.58 The cytogenic results of this study have
been reported by McKenzie et al.47 Blood was
drawn from the subjects before exposure,
immediately following exposure, and at 3 days, 2
weeks, and 4 weeks after exposure. Lymphocytes
were cultured for 48 hr, and slides were prepared
in the usual way, with 100 metaphase spreads per
subject per treatment scored for chromosomal
aberrations. A total of 13,000 human lymphocytes
were cytogenetically analyzed in this study. Cells
with suspected aberrations were photographed,
destained, restained with banding procedure, and
photographed again to identify the specific
chromosomes and regions involved. The data from
this experiment consist essentially of five
measurements on each subject overthe period of 1
month. Thus the appropriate statistical analysis of
variance for reported measurements was done on
the raw data and on appropriately arcsin-
transformed observations. There was no
significant difference in any of the observations for
196
-------
the respective blood sampling times. The authors
conclude that there is no detectable human
cytogenetic effect related to exposure to ozone
under the conditions of this experiment.
These findings are in contrast to those reported
by other investigators. The in vitro and animal
studies discussed in the section on chromosomes
in Chapter 8 of this document used either
extremely high ozone doses or inadequate
controls. The study of Merz et al.51 used higher
ozone concentrations and much longer exposures
than that of McKenzie et al.47 Also, the small
number of subjects in the Merz et al. study did not
permit the rigorous statistical analysis of the
McKenzie et al. study. Clearly, additional
evaluation of the human mutagenic potential of
ozone is needed, and such research is now in
progress. Evidence now available, however, fails to
demonstrate any mutagenic effect of ozone in
humans when exposure schedules are used that
are representative of such exposures of the
population at large as might actually occur in urban
areas.
Controlled Studies of Human Health Effects of
Peroxyacetylhitrate
Experiments conducted on a group of male
college students averaging 21 years of age have
suggested that exposure to PAN results in
increased oxygen uptake during exercise
compared to uptake during breathing of clean air.57
This report, however, did not adequately describe
the experimental design or the statistical analysis.
Subjects were exposed to 1485 fig/m3 (0.3 ppm)
PAN by breathing through the mouth (nose clamps
were used) for 5 min while at rest. Then the
subjects were immediately engaged in 5 min of
exercise on a bicycle ergometer. Both air
containing PAN and air free of PAN were used.
Expiration velocity was reduced after exercise. The
changes could reflect an increase in the effort
needed in breathing as a result of the exercise or
an increase in airway resistance.
In a series of studies of the effects of PAN and
carbon monoxide on healthy young and middle-
aged males during treadmill work in an exposure
chamber, Raven and coworkers found no changes
in maximum aerobic power when the subjects
were exposed to .1337 fjg/m3 (0.27 ppm)
RAN 13,53,54,58,59,60 yy^ ^ subjectg Worked 3t 3
level requiring 35 percent of maximum oxygen
consumption for 3.5 out of 4 hr of continuous
exposure to 1188 /yg/m3 (0.24 ppm) PAN, no
significant changes were observed in the
metabolic, cardiovascular, or thermoregulatory
responses.18 A small (4 to 7 percent) but signi-
ficant,18 reduction in standing FVC was observed in
the younger group of subjects.18
SUMMARY
Convincing new information on the health
effects of oxidant exposure has emerged from
controlled studies on humans, from which ten-
tative dose-response curves have been con-
structed. The new data show statistically signi-
ficant reduced pulmonary function in healthy
smokers and nonsmokers at ozone concentrations
at and above 730 fjg/m3 (0.37 ppm) for 2-hr
exposures. However, a recent study by Von
Nieding et al.59 showed some effect at as low a
concentration of ozone as 0.1 ppm for 2 hr. Some
studies suggest that mixtures of sulfur dioxide and
ozone at a concentration of 0.37 ppm are more
active physiologically than would be expected from
the behavior of the gases acting separately. Wide
variation in response among different individuals
is a general finding in studies of oxidants, as well
as other pollutants. Undesirable health effects of
oxidant air pollution exposure are increased by
exercise and, as judged by informal surveys, many
people apparently limit strenuous exercise
voluntarily when oxidant pollution is high.
REFERENCES FOR CHAPTER 9
1. Balchum O. J., J. S O'Brien, and B D Goldstein Ozone
and unsaturated fatty acids Arch Environ Health22 32-
34, 1971
2 Bates, D. V , and M. Hazucha. The short-term effects of
ozone on the human lung. In Proceedings of the
Conference on Health Effects of Air Pollutants, prepared
for the U S Senate Committee on Public Works U S
Government Printing Office, Washington, D C , 1973 pp
507-540.
3 Bates, D. V , G M. Bell, C D Burnham, M Hazucha, J
Mantha, L D Pengelly, and F Silverman Short-term
effects of ozone on the lung J Appl Physiol 32 176-181,
1972
4 Bates, D. V , G M Bell, C D Burnham, M Hazucha, J
Mantha, L D. Pengelly, and F Silverman Problems in
studies of human exposure to air pollutants Can Med
Assoc J. 103 833-837, 1970
5 Bell, K A,W S Linn, M Hazucha, J D Hackney, and D
V Bates. Respiratory effects of exposure to ozone plus
sulfur dioxide in southern Californians and eastern
Canadians. J. Am. Ind Hyg Assoc. 38 696-706, 1977
6. Bennett, G Ozone contamination of high altitude aircraft
cabins Aerosp Med 33969-973, 1962
7. Brinkman, R., and H. B Lamberts Ozone as a possible
radiomimetic gas. Nature (London) 181 1202-1203,
1958
197
-------
8 Brinkman, R , H. B. Lamberts, and T. S. Venmg.
Radiomimetic toxicity of ozonized air Lancet 7 133-136,
1964.
9 Buckley, R D , J D Hackney, K Clark, and C Posm
Ozone and human blood Arch. Environ Health30 40-43,
1975.
10 Challen, P. J. R , D E Hickish, and J Bedford An
investigation of some health hazards in an inert-gas
tungsten-arc welding shop Br J Ind Med 75 276-282,
1958
11. Dadlez, M. J. Ozone formed by ultraviolet rays. Q. J
Pharm. 7 99-100, 1928
12 DeLucia, A. J., and W C Adams. Effects of 03 inhalation
during exercise on pulmonary function and blood
biochemistry J Appl Physiol. Respirat Environ
Exercise Physiol 4375-81, 1977
13 Drmkwater, B. L, P. B. Raven, S M Horvath.J A Glmer,
R O. Ruhlmg, N. W. Bolduan, and S Taguchi. Air
pollution, exercise, and health stress Arch Environ.
Health 28.177-181, 1974
14 Flury, F., and F Zernik Ozone In. Schadliche Gase,
Dampfe, Nebel, Rauch und Staubarten. Julius Springer,
Berlin, 1931 pp 115-116
15 Folmsbee, L J , S M Horvath.P B Raven, J. F Bedi.A. R
Morton, B. L. Drmkwater, N.W. Bolduan, and J A Glmer.
Influence of exercise and heat stress on pulmonary
function during ozone exposure J Appl. Physiol
Respirat. Environ. Exercise Physiol 43409-413, 1977
16 Folmsbee, L. J., F. Silverman, and R. J. Shephard.
Decrease of maximum work performance following
exposure. J Appl Physiol • Respirat Environ. Exercise
Physiol. 42531-536, 1977
17. Folmsbee, L. J , F. Silverman, and R J. Shephard
Exercise responses following ozone exposure J. Appl.
Physiol 38:996-1001, 1975.
18 Gliner, J.A., P. B. Raven, S M Horvath, B L. Drmkwater,
and J C. Sutton Man's physiologic response to long-
term work during thermal and pollutant stress. J. Appl
Physiol. 39.628-632, 1975
19 Golden, J A.J.A Nadel, and H A Boushey Bronchial
hyper-irritability in healthy subjects after exposure to
ozone. Am. Rev Respir Dis 118 287-294, 1 978
20. Goldsmith, J. R., and J. Nadel Experimental exposure of
human subjects to ozone J. Air Pollut Control Assoc
79329-330, 1969
21. Goldstein, B. D. Production of paroxysmal nocturnal
haemoglobinuna-hke red cells by reducing and oxidizing
agents Br. J. Haematol 26:49-58, 1974.
22. Goldstein, B. D., L. Y Lai, and R. Cuzzi-Spada
Potentiation of complement-dependent membrane
damage by ozone. Arch Environ Health 28 40-42,1974
23. Goldstein, B D , and M R. Levin p-Ammobenzoicaciaas
a protective agent in ozone toxicity. Arch Environ Health
24243-247, 1972
24. Goldstein, B. D., and E M. McDonagh. Effect of ozone on
cell membrane protein fluorescence. I. In vitro studies
utilizing the red cell membrane. Environ. Res. 9.179-186,
1975.
25 Griswold, S. M , L A Chambers, and H. L Motley Report
of a case of exposure to high ozone concentrations for two
hours. AMA Arch Ind Health 75108-110, 1957.
26. Hackney, J. D., W. S Linn, R D Buckley, E E Pedersen,
S. K Karuza, D C Law, and D. A. Fischer. Experimental
studies on human health effects of air pollutants. I.
Design considerations Arch Environ Health 30 373-
378, 1975
27 Hackney, J D , W S Linn, S K Karuza, R D Buckley, D
C Law, D V. Bates, M Hazucha, L D Pengelly, and F
Silverman Effects of ozone exposure in Canadians and
Southern Cahfornians Evidence for adaptation? Arch
Environ Health 32 110-116, 1977
28 Hackney, J D , W S Linn, D C Law, S K Karuza, H
Greenberg, R D Buckley, and E E Pedersen.
Experimental studies on human health effects of air
pollutants. III. Two-hour exposure to ozone alone and in
combination with other pollutant gases Arch Environ
Health 30385-390, 1975
29. Hackney, J.D.W S Lmn.J G Mohler, andC R Collier.
Adaptation to short-term respiratory effects of ozone in
men exposed repeatedly J Appl Physiol Respirat
Environ Exercise Physiol 43.82-85, 1977
30 Hackney, J D , W S Lmn.J G. Mohler, E. E Pedersen, P
Breisacher, and A Russo Experimental studies on
human health effects of air pollutants II Four-hour
exposure to ozone alone and in combination with other
pollutant gases Arch Environ Health 30 379-384,
1975
31 Hallett.W Y Effect of ozone and cigarette smoke on lung
function. Arch. Environ. Health 70295-302, 1965.
32. Hazucha, M. Effects of ozone and sulfur dioxide on
pulmonary function in man Ph D Thesis, McGill
University, Montreal, Canada, 1973
33. Hazucha, M , and D V Bates Combined effect of ozone
and sulfur dioxide on human pulmonary function Nature
25750-51, 1975.
34 Hazucha, M , C Parent, and D. V Bates Development of
ozone tolerance in man. In. International Conference on
Photochemical Oxidant Pollution and Its Control
Proceedings Vol. I. B Dimitnades, ed EPA-600/3-77-
001 a, U S Environmental Protection Agency, Research
Triangle Park, N C., January 1977 pp 527-541
35 Hazucha, M , F Silverman, C Parent, S Field, and D V
Bates. Pulmonary function in man after short-term
exposure to ozone. Arch Environ Health 27 183-188,
1973.
36 Jordan, E O , and A J Carlson Ozone Its bac-
teriological, physiologic and deodorizing action J Am
Med Assoc 67-1007-1012, 1913
37 Kagawa, J , and T Toyama Effects of ozone and brief
exercise on specific airway conductance in man Arch
Environ. Health 30 36-39, 1975
38 Kelly, F J , and W E Gill Ozone poisoning Arch
Environ. Health 70.517-519, 1965.
39. Kerr, H D,T J Kulle, M L Mclhany, and P Swidersky
Effects of ozone on pulmonary function in normal
subjects An environmental-chamber study Am Rev
Respir Dis 777763-773,1975
40 Ketcham, B , S Lassiter, E D Haak, Jr , and J H
Knelson Effects of ozone plus moderate exercise on
pulmonary function in healthy young men In.
International Conference on Photochemical Oxidant
Pollution and Its Control Proceedings Vol I B
Dimitnades, ed EPA-600/3-77-001a, U.S. Environ-
mental Protection Agency, Research Triangle Park, N.C ,
January 1977 pp 495-504
41 Ketcham, G , S Lassiter, E D Haak, Jr , and J. H.
Knelson. Ozone exposure impairs pulmonary function in
healthy young men Clm. Res. 25 37A, 1977
198
-------
42 Kmdya, R J , and P C Chan Effect of ozone on
erythrocyte membrane adenos'ne tnphosphatase
Biochim. Biophys. Ada 429 608-615, 1976
43 Klemfeld, M Acute pulmonary edema of chemical origin
Arch Environ Health 70942-946, 1965
44 Klemfeld, M., C Giel, and I R. Tabershaw Health
hazards associated with inert-gas-shielded metal arc
welding AMA Arch Ind Health 15 27-31, 1957.
45 Knelson, J H , M L Peterson, G. M. Goldstein, D E.
Gardner, and C G Hayes Health effects of oxidant
exposures A research progress report In Report on UC-
ARB Conference "Technical Basesfor Control Strategies
of Photochemical Oxidant. Current Status and Priorities
in Research " University of Caifornia, Statewide Air
Pollution Research Center, Riverside, Calif., May 1976.
pp 15-50
46 Linn, W S , R D Buckley, C. E. Spier, R. L. Blessey, M. P.
Jones, D A Fischer, and J D Hackney. Health effects of
ozone exposure in asthmatics Am Rev. Respir. Dis
117 835-843, 1978
47 McKenzie, W H , J. H Knelson, N. J. Rummo, and D. E.
House Cytogenetic effects of inhaled ozone in man
Mutat Res 48 95-102, 1977
48 Menzel, D B Toxicity of ozone, oxygen, and radiation.
Annu Rev Pharmacol 10 379-394, 1970
49 Menzel, D B , R. J Slaughter, A M Bryant, and H O.
Jauregui. Heinz bodies formed in erythrocytes by fatty
acid ozonides and ozone Arch. Environ Health 30 296-
301, 1975
50 Menzel, D B , R J Slaughter, A M Bryant, and H O
Jauregui Prevention of ozonide-mduced Heinz bodies in
human erythrocytes by vitamin E Arch Environ Health
30234-236, 1975
51 Merz, T, M A Bender, H D Kerr, and T J Kulle
Observations of aberrations in chromosomes of
lymphocytes from human subjects exposed to ozone at a
concentration of 0 5 ppm for 6 and 10 hr Mutat Res
37.299-302 1975
52 Mohler, J G,J P Butler, J D Hackney, C R Collier, D
C Law, and W S Linn Redistribution of specific
ventilation in humans exposed to ozone. A new nitrogen
model Am Rev Respir. Dis 7/3.227,1976
53 Raven, P B, B L Drmkwater, S M Horvath, R. O
Ruhlmg, J A Glmer, J C Sutton, and N W Bolduan
Age, smoking habits, heat stress, and their interactive
effects with carbon monoxide and peroxyacetyl nitrate on
man's aerobic power Int J Biometeorol 18 222-232,
1974
54 Raven, P B , B L Drmkwater, R O Ruhlmg, N W
Bolduan, S Taguchi, J A Glmer, and S M Horvath
Effect of carbon monoxide and peroxyacetyl nitrate on
man's maximum aerobic capacity J Appl Physiol
36288-293, 1974
55 Schonbem.C F On some secondary physiological effects
produced by atmospheric electricity Med Chir Trans
34205-220, 1851
56 Silverman, F, L J Folmsbee, J Barnard, and R J
Shephard Pulmonary function changes in ozone-
interaction of concentration and ventilation J Appl
Physiol 41 859-864, 1976
57 Smith, L E Peroxyacetyl nitrate inhalation
Cardiorespiratory effects Arch Environ Health 70 161-
164, 1965
58. Strong, A A, R Penley, and J H Knelson Human
Exposure System for Controlled Ozone Atmospheres.
EPA-600/1-77-048, U.S Environmental Protection
Agency, Research Triangle Park, N C , October 1977
59. Von Niedmg, G., M. Wagner, H. Lollgen, and H. Krekeler
Zur akuten Wirkung von Ozon auf die Lungenfunktion des
menschen In' Ozon und Begleitsubstanzen im
photochemichen Smog. VDI Ber. (270) 123-129, 1977.
60. Young, W. A., D B Shaw, and D. V. Bates. Effect of low
concentrations of ozone on pulmonary function in man. J.
Appl Physiol 79.765-768,1964.
61. Young, W. A., D. B. Shaw, and D V. Bates. Pulmonary
function in welders exposed to ozone. Arch. Environ.
Health 7337-340, 1963
199
-------
10. EPIDEMIOLOGIC APPRAISAL OF PHOTOCHEMICAL OXIDANTS
INTRODUCTION
A review is presented in this chapter of the
evidence gathered to date in the epidemiologic
studies of whether or not ozone and other photo-
chemical substances exert deleterious effects on
human health. Every effort has been made to
describe study designs and investigative methods
thoroughly so that readers with either a technical
or nontechnical background may judge both the
reliability of findings presented and the degree to
which observed effects may be attributed speci-
fically to photochemical air pollution,
This chapter consists of four sections. The first
deals with effects of short-term (generally less
than 1 week) exposures to photochemical air
pollutants. In many studies of short-term ex-
posures, pollutant concentrations have been
measured on the same day, or even in the same
hour, as the health indices to which they have been
compared. The second section deals with effects of
long-term exposures to photochemical air
pollutants. The health data for most of these
studies have been collected at one point in time.
Thus, even though certain health impairments
have been associated with long-term pollutant
exposures, it has generally not been possible to
state the duration of exposure that may have
produced those impairments.
The sections of this chapter dealing with short-
term and long-term exposures have been divided
into subsections in which studies of similar effects
are reviewed. Following each subsection is a
discussion that assesses the reliability of current
evidence and identifies the areas of uncertainty
requiring future research. Experimental studies,
which give some indication of physiological
mechanisms underlying findings of epidemiologic
studies, are discussed where practicable. The
degree to which observed epidemiologic findings
may be attributed specifically to photochemical air
pollution or to individual components of this broad
class is also assessed. Finally, an effort is made to
identify specific pollutant concentrations and the
duration of exposure necessary to promote the
observed effects.
Following the sections on short-term and long-
term exposures, a brief assessment of the attitudes
of laymen and physicians toward oxidant pollution
will be presented. Finally, the available
epidemiologic evidence relatedtooxidant pollution
will be summarized.
EFFECTS OF SHORT-TERM PHOTOCHEMI-
CAL OXIDANT EXPOSURES
To identify the acute health effects of
photochemical oxidant pollution, observations of
the same populations of communities are made
during periods of both high- and low-level
pollution. Changes in health-related indices are
compared with short-term pollutant concen-
trations measured as 24-hr averages, hourly
maxima, or instantaneous peak concentrations.
Daily Mortality in Relation to Variations in
Oxidant Levels
MORTALITY AMONG RESIDENTS AGED 65
YEARS AND OVER
Over a 3-year period from 1954 to 1957, the
California State Health Department conducted a
study of the relationship between daily
concentrations of photochemical oxidants and
daily mortality among residents of Los Angeles
County ages 65 years and over.67'8 During these
years, photochemical oxidant concentrations in
Los Angeles County were generally measured by
the potassium iodide (Kl) method. Ozone
concentrations were generally measured by visual
assessment of rubber cracking. Both methods
were calibrated with unbuffered 2-percent Kl
solution. The nu mber of deaths per day was related
to two indices—maximum daily temperature atthe
downtown weather bureau, and oxidant
concentrations from August through November
1954 and from July through November 1955.
200
-------
Data collected over these periods of time are wave on mortality. The figure reveals noconsistent
summarized in Table 10-1 and Figure 10-1. association between mortality and high oxidant
Examination of the table and figure reveals a concentrations in the absence of unusually high
pronounced effect of the September 1955 heat temperatures.
TABLE 10-1. AVERAGE NUMBER OF DEATHS PER DAY RESULTING FROM CARDIAC AND RESPIRATORY CAUSES
AMONG RESIDENTS OF LOS ANGELES COUNTY, AGED 65 AND OVER, AS RELATED TO OXIDANT CONCENTRATIONS
AND MAXIMUM DAILY TEMPERATURE BY MONTH, 1954-55
Temperature readings
Totals
Concentration,
ppm
August 1954
Low (00-0 24)
Medium (025-049)
High (0.50+)
September 1954
Low (00-0 24)
Medium (0.25-0.49)
High (0 50+)
October 1954
Low (00-0 24)
Medium (025-049)
High (Q 50+)
November- 1954
Low (00-0.24)
Medium (0.25-0.49)
High (0.50+)-
July 1955
Low (00-0 24)
Medium (0.25-0.49)
High (050+)
No readings
August 1955
Low (00-0 24)
Medium (0 25-049)
High (0 50+)
No readings
September 1955
Low (00-0 24)
Medium (0 25-0.49)
High (0 50+)
No readings
Mean
Low (00-0 24)
Medium (0 25-049)
High (0 50+)
No readings
Number
of days
8
22
1
1
25
4
9
14
8
16
14
4
11
2
14
5
16
4
6
5
13
7
5
48
115
26
25
10°-15°C 16°-20°C 21°
(50°-59°F) (60°-69°F) (70°
Average Average Average
number Number number Number number Number
of deaths of days of deaths of '•'ays of deaths of days
29 9
32.2
320
260
32 1
33 3
394
37 1
363
38.6
379
350
34.8
420
37 5
368
384
372
37 8
290
61.5
50 3
348
U335"
W35.5
U391
W37 9
U385
W400
U367
W370
4
7
1
1
5
1 31.0 6
1 430 10
3
1 39.0 5 360 4
4 368 5
1
5
1
10
1 330 3
4
1
4
1 39.0 7 333 19
5 39 9 36
6
14
-26°C
-79°F|
27°-31°C 32°-37°C
(80°-89°F) (90°-99°F)
Average
number Number
of deaths of days
282
30 1
320
260
296
38 2
360
367
420
348
280
344
42 0
38 7
27.3
308
320
33 2
31 6
326
35 7
360
4
12
16
4
2
2
5
6
5
3
6
1
4
4
16
4
4
1
3
1
20
60
15
8
•--38°C( -100°F)
Average Average Average
number Number number Number number
of deaths of days of deaths of days of deaths
31 5
31 8 3 390
351 4 360
333
47 5
395 1 370
360
386
41 8
37.3
35.2
420
345
35 5
384
372
385 2 365
300
43.7 1 520
280 3 407
1 41 0
367
379 9 41 0
35 3 3 40 7
36.5 3 388
1 420
5 988
2 850
1 420
5 988
2 850
Total days
Total deaths
Unweighted
Weighted
1 12
39.0
390
36.6
36 1
75
34.0
33.2
103
36.6
372
15
402
40.5
8
753
88.3
BU - unweighted, W = weighted
201
-------
Table 10-1 contains too few data to yield any
conclusive trends. However, examination of the
table's marginal means suggests that the effect of
temperature on death rates, at least during the
months considered, is smallest in the comfortable
temperature range of 21° to 26°C (70° to 79°F).
Such examination also suggests a positive
association between oxidant exposure and
mortality in this temperature range. Examination
in Table 10-1 of overall mean death rates by
pollution exposure category also suggests such an
association.
It must be emphasized that Table 10-1 does not
reveal a convincing relationship between oxidant
exposure and daily mortality. Indeed, the table
contains far too few data to warrant conclusions of
any kind, except perhaps that the effect of
unseasonable temperatures on mortality is large
enough to obscure any effect that oxidant pollution
may have. However, Table 10-1 also arouses
interest in the hypothesis that any effect that
oxidant might exert on mortality may be most
apparent at times when temperatures are
seasonal and comfortable. As yet, this hypothesis
has not been adequately tested.
MORTALITY AND HEAT WAVES
In the Los Angeles Basin, high temperatures and
elevated oxidant concentrations tend to occur
simultaneously. The question thus arises whether
oxidant exposure augments the effect of
temperature on mortality rates. Oechsli and
Buechley41 considered this question briefly in a
study of the effect on mortality of three Los
Angeles heat waves occurring in 1939, 1955, and
1963. Daily mortality during these heat waves was
compared with daily mortality occurring
immediately before and after each heat wave and
with mortality during the same season in 1947,
when no heat wave occurred.
Statistically significant increases in mortality
rates were observed during each heat wave,
particularly among elderly persons. However,
there was no apparent difference between the
effects of the 1939 and 1955 heat waves on
mortality rates. Quite probably, considerably less
photochemical oxidant pollution accompanied the
1939 heat wave than that of 1955. (Oxidant
concentrations were not routinely measured in
1939.) The comparison of the two heat waves thus
320
300
280
260
240
22°
200
18°
160
B 140
120
100
80
60
40
T I I T \ T I IliT I I I I I I I I I I I r
so oo
40 w
30 £
20 OC
uj
10 |
uj
0 I-
3 10 17 24 31
|-«-JULY »
7 14 21 28 7 14 21 28
AUG. »|< SEPT.
TIME, days
9 16 23 30 7 14 21 29
OCT.
Figure 10-1 . Comparison of deaths of persons aged 65 years and over, and maximum daily temperatures, Los Angeles
County, July 1 to November 30, 1955.678
202
-------
suggests that high photochemical oxidant
concentrations do not notably augment the effect
of high temperatures on mortality. On the other
hand, it must be remembered that Oechsii and
Buechley did not statistically determine whether a
relationship between mortality and oxidant
exposure existed.
MORTALITY OF NURSING HOME RESIDENTS
The California State Health Department also
made an effort to determine whether patients in
Los Angeles nursing homes, many of whom are
chronically ill, experienced increased mortality
during or just after days of high oxidant
concentrations.6'78 Deaths and transfer to
hospitals among residents of 16 Los Angeles
nursing homes having a total of 358 beds were
recorded for 1954. An unusually large number of
patients died following a particularly heavy
episode of smog during 1 week of the study period.
During the same week, however, the number of
residents transferred to hospitals did not appear to
be unusually high. Neither the number of deaths
nor the number of transfers to hospitals was
elevated during a high-smog period about a month
earlier than the one mentioned above.
Measurements of pollutant concentrations were
not reported.
A larger study of the nursing home population
was conducted from July through December 1955,
during which all nursing homes in Los Angeles
County containing 25 or more beds were surveyed.
Daily mortality rates were reported from 92 homes
with a total of 3826 beds. Daily mortality, the
corresponding maximum daily temperature, and
the occurrence of smog-alert days, with ozone
concentrations of 590/^g/m3 (0.30 ppm)or higher,
are shown in Figure 10-2. The heat wave in late
August and early September, during which several
smog-alert days occurred, showed a striking effect
on mortality. At no other time could a relation-
ship between daily mortality and smog-alert days
be discerned.
=DAYSOF OFFICIAL ALERTS,
OZONE > 0.30 ppm
1 5 10 15 202530 4 9 14 19 24 29 3 8 13182328 3 8 131823 28 2 7 12 172227 27 1217 222731
JULY »!•«•—AUG. 1*4-* SEPT. » i * OCT. »4*— NOV. H^ DEC- H
•+* SEPT. »1* OCT. -
TIME, days
Figure 10-2. Comparison of nursing home deaths, maximum daily temperature, and smog-alert days in Los Angeles
County. July through December 1955.67fl
203
-------
TWO-COMMUNITY STUDY
Massey et al.33 compared daily mortality in two
areas of Los Angeles County selected to be similar
in temperature but different in air pollution levels.
The investigators constructed two synthetic
communities, one of intermediate pollution and
one of high pollution, containing a combined
population of 944,391 persons. The pollutant
variables used in data analysis were the daily
maximum and mean oxidant levels, as measured
by the potassium iodide (Kl) method and sulfur
dioxide and carbon monoxide concentrations.
Intercommunity differences in temperature were
also considered. The mean number of daily deaths
in the intermediate pollution area was subtracted
from the mean number of deaths in the high
pollution area, and the differences were related by
multiple correlation and regression to differences
in pollution concentrations. No significant
correlations between differences in mortality and
differences in pollutant levels were observed.
Mean daily death rates per 1 00,000were3.1 Sand
3.06, respectively, in the intermediate- and high-
pollution communities.
MORTALITY RESULTING FROM CARDIAC AND
RESPIRATORY DISEASES
Hechter and Goldsmith22 analyzed the effect of
pollutant concentrationson average daily mortality
from cardiac and respiratory diseases in Los
Angeles County for the years 1 956 through 1 958.
During these years, photochemical oxidant
concentrations were measured in Los Angeles
County by the Kl method, using an unbuffered 2-
percent Kl solution as the calibration reagent. Daily
mortality, (averaged within each month of the
study) fluctuated between-1.0 and 1.3 per 100,000
population. These fluctuations were approximately
180 degrees (6 months) out of phase with
fluctuations in oxidant and temperaure values and
approximately in phase with maximum carbon
monoxide concentrations (Figure 1 0-3). To remove
the effect of season of year, the authors fitted
Fourier curves to the data. The residual variations
from these fitted curves for each of the variables
were presumed to be independent of season. The
relationship between pollution or temperature on
one day with the value of the same variable on the
preceding or following days was also accounted for
30
20
1-
<
I
o
30
u
25
20
UJ
15
30
£
£
of
o
o
2
Z 10
o
00
tc.
<
0
III 11 I 1111
III 11 111 in
YEAR
=
S1.
1.0
u tc.
uj nn in I
i Q °-8|LU-1
YEAR
Figure 10-3. Comparison of maximum concentrations of oxidant and carbon monoxide, maximum
temperature, and daily death rate for cardiac and respiratory causes, Los Angeles County, 1 956-58. 22
204
-------
in the analysis. When residuals from the fitted
curves were thus analyzed, no significant
correlations between pollutants and mortality for
cardiorespiratory diseases were found. Neither
were there significant correlations when a 1 -104-
day lag between exposure and mortality was
applied.
Mills36 correlated daily cardiac and respiratory
mortality in Los Angeles with corresponding daily
maximum oxidant concentrations for thefollowing
periods: 1947 through 1949; August 17, 1953,
through December 31, 1954; and January 1
through September30,1 955, excludingthesevere
heat wave of September 1 through 4. To correct for
possible seasonal effects on mortality, Mills
compared mean daily death rates on low-smog
days to mean rates on higher-smog days occurring
within the same month.
For 1947-49, mortality data were related to the
Stanford Research Institute's smog index, 56'57-5B
which was based on meteorologic conditions, not
on air pollution measurements. (Coefficients of
correlation between the smog index and maximum
ozone measurements, supplied by the U. S. Rubber
Co., ranged from 0.08 to 0.949, All were
statistically significant at a = 0.05.) For 1953-55,
mortality data appear to have been related to daily
maximum total oxidant measurements supplied by
Dr. A. J. Haagen-Smit of the California Institute of
Technology. These measurements were
presumably made with the phenolphthalein
method.
In all periods, Mills36 observed a consistent
positive relationship between daily mortality and
daily maximum ozone or oxidant concentration.
This relationship held for persons under age 65 as
well as for older persons. On the average, daily
death rates in 1947-49 were about 7 percent
higher when the smog index was six through nine
(corresponding to maximum ozone concentrations
of about 0.36 to 0.37 ppm) than when the smog
index was zero through five. In the same period,
daily death rates when the smog index was 10 or
above (corresponding to maximum ozone concen-
trations of at least 0.42 ppm) averaged 1 2 percent
higher than when the smog index was zero through
five. For 1953-55, the mean daily death rate when
maximum total oxidant exceeded 0.2 ppm aver-
aged about 5 percent higher then when maximum
total oxidant was below 0.2 ppm. Mills,36 like the
California Department of Health,6-7-8 performed an
analysis of death rates in Los Angeles nursing
homes during the second half of 1955. Unlike the
California Department of Health, Mills applied a
seasonal correction factor to death rates within
each month of this period. Mills' analysis suggest-
ed a continuous increase in daily death rates with
increasing daily maximum oxidant concentrations
above 0.2 ppm. Corrected daily death rates ranged
from 4.72 when daily maximum oxidant was below
0.2 ppm, to 6.23 when maximum oxidant was 0.8
ppm or above The correlation between seasonally
corrected death rate and maximum oxidant was
0.206 (p = 0.01).
In a separate report, Mills35 stated that more
than 334 deaths in Los Angeles during 1 954 were
associated with daily oxidant or ozone maxima
above 0.2 ppm.
The California Department of Health has
specifically questioned Mills' findings. In Clean Air
For California?7'8 the Department explains that the
Stanford Research Institute's smog index, by
which Mills classified mortality data for 1947-49,
was developed as a prediction device and not as a
measurement of air pollution. The Department
also states that the abandonment of the smog
index by the Institute indicated that it was not
satisfactory in either respect. They add that Mills'
manner of selecting smog-free days to be used for
comparison may not have been entirely
satisfactory.
The California Department of Health also applied
the essential details of Mills' analytical procedure
to Los Angeles County cardiac and respirator death
rates during August-November 1954.B7>S Results
of the California analysis are shown in Table 10-2.
Comparison of mortality rates on smoggy and
smog-free days within each of the months shown
reveals no effect of oxidant pollution on mortality.
Thus, as stated in references,6'7'8 and the Breslow
and Goldsmith's report,4 analyses performed in
California did not support Mills' conclusion that
334 deaths in Los Angeles during 1954 were
associated with oxidant levels above 0.2 ppm.
Interpretation of the results summarized in
Table 10-2 is clouded because references678 do
not contain an explanation of criteria used to dif-
ferentiate smoggy from smog-free days. Also, the
temperatures on the two types of days cannot be
ascertained from three references. However, that
Mills and the California Department of Health have
interpreted Los Angeles mortality data differently
is undeniable.
DISCUSSION OF MORTALITY STUDIES
As yet, no convincing association between daily
205
-------
mortality and daily oxidant concentrations in Los
Angeles has been shown. Mills has certainly
described a positive association between mortality
and oxidants. However, because the smog index
used for much of his analysis appearsto have been
inaccurate, and because investigators in California
have observed no oxidant-mortality association in
data used by Mills himself, the vigor with which
Mills' conclusions can be advanced must be
considerably restricted. However, careful ex-
amination of available studies raises the
interesting question as to whether oxidants might
affect mortality at times when temperatures are
comfortable and seasonal. To date, most efforts to
determine whether oxidants affect mortality have
focused on high-smog periods. During such
periods, temperatures also tend to be high and may
thus obscure or eliminate any effect of oxidants on
mortality. The data in Table 10-1, which are
apparently the only extant cross-tabulation of
mortality by pollution and temperature, suggest
that the effect of oxidant on mortality may be
strongest or most readily detectable, when
temperatures are seasonal. Indeed, it is plausible
that oxidants would noticeably affect mortality at
no other times.
Thus whether oxidants affect daily mortality has
not been conclusively resolved. Recent advances
in statistical analysis of mortality data, coupled
with the accumulation of such data over the two
decades since most oxidant-mortality studies were
performed in Los Angeles, should enable
investigators to clarify this issue considerably over
the next several years.
Hospital Admissions in Relation to Oxidant
Levels
ADMISSIONS TO LOS ANGELES COUNTY
HOSPITAL, 1954
The California State Health Department
examined several categories of admissions to Los
Angeles County Hospital from August through
November 1954.6'7'8 These categories were
asthma, other respiratory conditions, and cardiac
conditions. Total admissions, the number of
persons with acute conditions in all units, and the
number who ditsd were tabulated. A formal
statistical analysis does not appear in clean-air-
for-Ca/ifornia reports.67'6 However, visual
inspection of these data reveals no consistent
relationship between admissions and high smog
periods.
Brant and Hill2'3 studied 9- to 90-year-old
patients with respiratory or cardiovascular
diagnoses who were admitted to or discharged
from Los Angeles County Hospital between August
and December 1954. Patients were considered
only if they had resided for at least 3 years in an
area within 13 km (8 miles) of downtown Los
Angeles. Although the method of selecting
patients for the study was not described, the
selection process yielded 246 cardiovascular
admissions and 122 respiratory admissions. Daily
average total oxidant concentrations between 6
a.m. and 1 p.m., as measured about 1,6km(1 mile)
from the hospital, were the only indices of air
pollution employed in statistical analyses. Multiple
regression analyses were used to relate hospital
admissions to atmospheric and meteorological
variables. Calculations included regressions of
hospital admissions as late as 4 weeks after the
occurrence of a given set of environmental
measurements, which included total oxidant,
temperature, and relative humidity. Brant obtained
a positive partial correlation coefficient of 0.982
between the total oxidant level and the number of
respiratory and cardiovascular admissions
occurring 4 weeks later. However, the correlations
between total oxidants and admissions during the
same week and 2 weeks later were, respective-
ly, -0.986 and -0.870.
Although Brant's findings are interesting, they
must be regarded as inconclusive for four reasons.
First, it is difficult to rationalize why hospital
admissions should be more dependent on oxidant
TABLE 10-2. COMPARISON OF RESPIRATORY3 AND CARDIAC" DEATHS DURING SMOG AND SMOG-FREE PERIODS
OCCURRING IN LOS ANGELES COUNTY, AUGUST NOVEMBER 1954678
Smog periods
September 21-24
October 15-20
November 24-28
Total
deaths,
selected
causes
181
294
235
Daily
average
46
49
47
Smog-free periods
September 12-15
October 1-6
November 1 6-20
Total
deaths.
selected
causes
191
303
244
Daily
average
48
50
49
^Respiratory tuberculosis jOOl -008), pneumonia, ail forms (490-493) Sixth revision, mternationai hst numbers
"Diseases of the heart {410-443} Sixth revision, international list numbers
206
-------
levels 4 weeks before than on more recent levels.
Second, the partial correlation coefficients
between cardiovascular admissions and total
oxidant concentrations during boththe weekof the
admissions and 2 weeks before were very highly
negative. Third, as far as we can determine. Brant
did not examine the variables used in his statistical
models for autocorrelation. Fourth, no indices of
pollution other than oxidant concentrations were
considered.
HOSPITAL ADMISSIONS THROUGHOUT THE
CITY OF LOS ANGELES
In two studies, Sterling et al.59'60 assessed the
influence of air pollution exposures on persons
admitted to hospitals with more than 100 beds in
the city of Los Angeles. Only patients who were
members of Blue Cross were considered.
Discharge diagnoses were assigned to one of three
categories—highly relevant, relevant, and
irrelevant—which reflected the presumptive
susceptibility of the diagnoses to air pollution
exposure. Included in the "highly relevant"
category were allergic disorders, inflammations of
the eye, and acute upper and lower respiratory
infections. Included in the "relevant" category
were cardiovascular diseases and other
respiratory diseases. All other conditions were
included in the "irrelevent" category. The authors
corrected both number of admissions and pollution
exposure measurements for day-of-week effects.
Nine pollution indices, including oxidant, ozone,
carbon monoxide, sulfur dioxide, nitrogen dioxide,
nitric oxide, and paniculate matter, were
considered in Sterling's analysis. Air pollution
measurements were obtained from the Los
Angeles County Air Pollution Control District. For
each pollutant, the exposure index used in
statistical analyses was the average of the
maximum and minimum concentrations for each
day of the study.
In Sterling's first study,59 air pollution indices
were correlated with the number of hospital
admissions occurring on the same day in each
diagnostic category. In all, 223 days between
March and October 1961 were considered.
Correlation coefficients between admissions for
highly relevant diseases and all nine pollution
indices measured were all statistically significant
at a = 0.05. All but one correlation (that with sulfur
dioxide) were significant at a = 0.01. The
magnitude of these correlation coefficients ranged
from 0.1 6 (for sulfur dioxide)to 0.25 (for paniculate
matter). Correlations between relevant diseases
and pollution were significant for oxidant, ozone,
and sulfur dioxide. For all relevant diseases, the
highest correlations were observed with ozone
(0.27), sulfur dioxide (0.27), and oxidant (0.25). For
all pollutants except sulfur dioxide, correlations
between irrelevent diseases and pollution were
negative or not significant. Coefficients of
correlation between temperature or relative
humidity and admissions were not significant in all
disease categories.
In the second study performed by Sterling et
al.,60 length of stay in the hospital was correlated
with air pollution and meteorologic variables over
the same time period, as in the first study. The
absolute value of correlation coefficients was
usually less than 0.1 and never exceeded 0.169.
Despite their diminutive values, coefficients of
correlation between length of stay and all pollution
indices were statistically significant at a = 0.05 for
the irrelevent diseases and for all disease
categories combined. All correlations between
pollution indices and length of stay for irrelevant
diseases and all disease groupings combined were
positive, except that for ozone concentration. The
only pollutants consistently and significantly
positively correlated with length of stay for
relevant disease categories were sulfur dioxide,
nitrogen dioxide, and total oxides of nitrogen.
Coefficients of correlation between humidity and
temperature and length of stay in all disease
categories were negative and significant at a =
001. This finding with respect to temperature
contrasts with the findings of several of the
mortality studies cited above.
DISCUSSION OF STUDIES OF HOSPITAL
ADMISSIONS
Studies of the relationship of short-term oxidant
exposures to hospital admissions have yielded
mixed results Even studies in which a positive
association between exposures and admissions
has been observed must be considered
inconclusive. As mentioned above, it is difficult to
rationalize Brant's observation of a strong
association between pollutant levels in 1 week and
hospital admissions 4 weeks later, in the face of
highly negative associations in the intervening
weeks Also, in the light of Sterling's studies, one
must wonder whether Brant, if he had chosen to
calculate them, would not have obtained
correlation coefficients between admissions and
207
-------
pollutants other than oxidants that were
comparable to those calculated for oxidants.
Thus even if the association between short-term
exposures and hospital admissions observed in
Los Angeles proves to be valid, available evidence
in this field will not support the incrimmation of
certain pollutants and the exoneration of others.
Also, available evidence gives no indication of
pollutant concentrations at which hospital
admission rates might be expected to increase.
Furthermore, since no studies relating oxidant
concentrations to hospital admission rates have
been performed outside of Los Angeles, it is
impossible at present to say whether oxidants
influence admissions in other areas.
Aggravation of Existing Respiratory Diseases by
Oxidant Pollution
AGGRAVATION OF ASTHMA
As a result of reports by physicians that
asthmatic attacks are frequently associated with
high-smog periods, Schoettlm and Landau63
undertook a study to determine whether such a
phenomenon indeed takes place. Five physicians
selected 1 57 patients, 1 37 of whom participated in
the study All resided and worked in the Pasadena
area Fifty-four of the patients were younger than
age 1 5. Most had had asthma for at least 5 years.
Daily records of the time of onset and severity of
asthma attacks were maintained by the patients
between September 3 and December 9, 1956.
The Schoettlm and Landau study, as published,
does not mention pollution measurement methods
or averaging times used in data analysis. In Air
Quality Criteria for Photochemical Oxidants,39 it is
assumed that oxidant levels in this study were
daily instantaneous peak values as measured by
the phenolphthalem method. Errata to the
document were issued indicating that oxidant
levels in this study were, in fact, determined by the
Kl method and not the phenolphthalein
method.16'40'47
Since publication of Air Quality Criteria for
Photochemical Oxidants39 (AP-63) in 1970, a
considerable effort has been made to determine
with certainty the measurement methods and
averaging times used in the Schoettlm and Landau
study Consultations with the authors have
established that daily asthma attack rates were
correlated with daily maximum hourly average
oxidant levels, not with instantaneous peak levels
These consultations have also confirmed that
oxidant measurements were indeed obtained from
the Los Angeles County Air Pollution Control
District, and from no other source. Careful
examination of Air Pollution Control District
records has further established that the District
used only the potassium iodide method for
measuring oxidant levels in Pasadena over the
course of the Schoettlin and Landau study. Thus in
the discussion to follow, Schoettlin and Landau are
presumed to have used daily maximum hourly
average oxidant concentrations, as measured by
the potassium iodide method. In the discussion to
follow, we also assumed that oxidant levels
reported in this study correspond to daily
maximum hourly levels and not to instantaneous
peak levels, recognizing the error that might result
if readings had corresponded to the latter.
In the Schoettlin and Landau study, the greatest
number of asthma attacks occurred between
midnight and 6 a.m., whereas the maximum
oxidant levels were recorded between 10 a.m. and
4 p.m. Of the 3435 attacks reported, fewer than 5
percent were consciously associated with smog by
the patients. No severe attacks were consciously
associated with smog. One-third of the attacks
consciously associated with smog were reported
by a single patient. The correlation coefficient
between daily attacks and the concurrent
maximum oxidant reading was 0.37. The addition
of other variables to the analysis did not
significantly alter these results. Lagged
correlations between oxidant concentration and
the number of patients having attacks 6, 12, 18,
and 24 hr later were generally lower than
concurrent correlations. There was no significant
difference between the average number of
subjects having attacks on days above the median
maximum hourly oxidant level of 250/ug/m3 (0.13
ppm) ozone and the average number of patients
having attacks on days when levels were below the
median However, the mean number of patients
having attacks on days when daily maximum
hourly oxidant levels exceeded 490 /ug/m3 (0.25
ppm) was significantly higher than on days when
the daily maximum hourly oxidant level fell below
this level. The authors suggested that this finding
might indicate a threshold level for oxidants above
which a physiologic response could be observed.
Also, attack rates on days in which plant damage
occurred were significantly higher than on days
without plant damage. Specific oxidant levels
associated with plant damage were not reported in
this study, however The effect of oxidant on
208
-------
asthma attack rates was most pronounced among
persons who had lived in the area for 10 years or
more. The data were examined further to see
whether a small number of the subjects might be
responsible for the observed correlation. The
authors identified eight individuals (6 percent of
the total panel) whose attacks corresponded most
often to days on which plant damage occurred.
Seven of these eight people were females. No
other common characteristic could be discerned.
AGGRAVATION OF EMPHYSEMA AND CHRONIC
BRONCHITIS
Several studies have been conducted to
determine whether air pollution aggravates the
condition of subjects suffering from chronic
bronchitis and emphysema.
Motley et al.38 reported a study of the lung
function of 66 volunteers, 47 of whom had
pulmonary emphysema. Before testing, subjects
spent differing amounts of time in rooms from
which oxidants had been removed by activated
charcoal filters. Twenty-one subjects stayed in the
filtered rooms for periods of 2 to 4 hr, 20 subjects
for periods of 18 to 20 hr, and 25 subjects for
periods of 40 to 90 hr. The study was performed
over a 31/2-year period in the late 1950's. Air quality
mea'iurernents were provided by the Los Angeles
Cou ity Air Pollution Control District. On the days
of t iis study, maximum oxidant concentrations
(pre* umably outdoors) between 9 a.m. and 2 p.m.,
calculated as ozone, ranged from 390 to 1040
pg/m3 (0.20 to 0.53 ppm) at monitoring stations
sevtral kilometers from the chamber. Maximum
con< entrations of other pollutants fell into the
folk /ving ranges: Oxide of nitrogen, 0.2to 0.6 ppm;
sulfur dioxide, 0.05 to 0.25 ppm; and carbon
monoxide, 5 to 27 ppm. Air was classified as
smoggy when there was definite odor of ozone,
reduced visibility, eye irritation, and the prediction
of smog by the Los Angeles Air Pollution Control
District. The investigators measured vital capacity,
3-sec forced expiratory volume (FEV 3.0), and
maximal breathing capacity. Residual volume and
air distribution measurements were also recorded.
An improvement in lung function, particularly a
decrease in the residual lung volume, was
observed in emphysematous subjects who had
entered the filtered roomson smoggydaysand had
remained in them for 40 or more hr. No significant
changes in lung-volume measurements were
observed when normal subjects breathed filtered
air. No significant changes were observed when
emphysematous subjects entered the chamber on
nonsmoggy days. The smoking habits of subjects
are not discussed in the report, but variations in
these might have influenced the results,
particularly if smokers were not allowed to smoke
while in filtered rooms, The methods of analysis
used in this study did not allow the effects of
individual pollutants to be separated from one
another.
At Los Angeles County Hospital, Remmers and
Balchum45 utilized a room with an air-condition ing
system and a filter that could be used, at the
discretion of the investigators o remove photo-
chemical oxidants, ozone, nitrogen oxides, and a
portion of particulate matter from ambient air.
Studies were conducted between July 1964 and
February 1965. Analyses of total oxidant by the
buffered Kl method and of nitrogen dioxide by the
Saltzman method were performed five times daily.
Subjects performed lung-function tests one or
more times daily while they lived in the room. In
general, they spent 1 week in unfiltered air, a
second week in filtered air, and a third week in
unfiltered air. Throughout the study, air con-
ditioning was adjusted to maintain a room
temperature of 22° ± 1,4 °C (72° ± 2.5°F) and a
relative humidity of 50 ± 5 percent.
Among other tests, determinations of airway
resistance, carbon monoxide diffusing capacity,
capillary oxygen tension, and oxygen con-
sumption were performed while subjects were
resting. Several of these determinations were
repeated after subjects exercised.
Only four subjects were considered in the
Remmers and Balchum report. Allfour had chronic
lung disease. Two of them were considered to have
relatively advanced emphysema but the other two
were deemed to have only moderate pulmonary
impairment. Smoking habits of subjects were not
mentioned in the report of this study.
Beneficial effects of air filtration in studies of
resting patients were most pronounced in the pair
with only moderate pulmonary impairment. In this
pair, airway resistance was about 20 percent lower
during exposure to filtered airthan to unfiltered air.
A simultaneous increase in the speed of nitrogen
washout also occurred in this pair. These changes
were statistically significant at a = 0.10
The beneficial effects of ?ir filtration in studies of
exercising patients were slightly more pronounced
in patients with advanced emphysema than in
those with only moderate impairment. In all four
209
-------
patients, post-exercise oxygen consumption
decreased significantly during exposure to filtered
air. Capillary oxygen tension increased
simultaneously in all four patients. Changes in
oxygen tension were statistically significant only in
patients with advanced emphysema. During
studies of exercising patients, the mean oxidant
concentration (calculated as ozone) was about 255
jug/m3 (0,13 ppm) when air was not filtered, and
about 40 jug/m3 (0.02 ppm) when air was filtered.
For several reasons, the Remmers and Balchum
study must be regarded as inconclusive. First, and
probably most important, this report does not
consider variations in the smoking habits of
subjects. It has become quite clear that subjects
were not allowed to smoke while breathing filtered
air, but they were allowed to leave the study room
and smoke, if desired, during their 2 weeks of
exposure to ambient air.43 Variation in smoking
habits may thus have explained a major part of the
physiologic difference observed between subjects
exposed to filtered air and those exposed to
ambient air.
Second, the ambient air to which subjects were
exposed contained not only oxidants but other
substances as well. Thus, even if the observed
changes in lung function were attributable to air
pollution, they might also be attributable, at least in
part, to these other substances. Third, Remmers
and Balchum appear to have analyzed data for only
a very small number of subjects.
Though their report45 considered only four
patients, Remmers and Balchum collected data
from 15 patients, most of them with relatively
severe emphysema. Nine of these patients were
smokers, five were non-smokers, and one stopped
smoking when the study began. Patients generally
spent about 1 week in unfiltered air, then a second
week in filtered air, and finally a third week in
unfiltered air. Analysis of the study of all 15
patients was performed for inclusion in the
document, Air Quality Criteria for Photochemical
Oxidants.39 Results of that analysis are presented
in Table 10-3, which shows coefficients of
correlation between daily values of oxygen
consumption and airway resistance (both
measured under conditions of rest and exercise,
with corresponding daily oxidant values measured
in the morning and the afternoon). Each
correlation coefficient was converted to a "t"
statistic. The "t" statistics were then summed
within pollution - test - specific categories. The
investigators then derived a "z" statistic for each
category by dividing the su m of "t" statistics by the
sum's variance. The "z" statistics are presented in
the last line of Table 10-3.
Using these methods, the authors observed a
positive relationship, significant at a - 0.01,
between oxidant concentration and oxygen
consumption during rest and exercise They also
observed a generally positive relationship,
significant at a = 0.001, between oxidant con-
centration and airway resistance measured at rest.
As Table 10-3 shows, the relationship between
airway resistance and afternoon oxidant
concentrations was more consistently positive
than that between resistance and morning con-
centrations.
In a published report of the study just described,
Dry and Hexter62 mention that each subject
underwent a daily battery of 20 pulmonary
function tests. The authors also mention that
concentrations of three pollutants—oxidants,
nitrogen dioxide, and nitric oxide—were meas-
ured. Though it is not explicitly stated, a distinct
implication of the Ury and Hexter report is that
oxidant levels were more strongly correlated with
pulmonary function test results than were levels of
either nitrogen oxide or nitric oxide. Of the 20tests
performed, only airway resistance is considered in
this report. Thus the degree of correlation between
pollution exposure and results of other tests
cannot be determined, though the report implies
that the other tests were not as strongly correlated
with pollution exposure as was airway resistance.
Finally, Ury and Hexter performed separate
analyses for smokers and non-smokers. In
smokers, the relationship between airway
resistance and afternoon oxidant exposure, as
tested by "z" statistic derived as described above,
was significant at cr = 0.002. In nonsmokers, the
relationship was significant at a - 0.001
In the Ury and Hexter report, unlike that of
Remmers and Balchum, smokers are clearly
separated from nonsmokers. That summed "z"
statistics were more highly significant for
nonsmokers tends to corroborate the conclusion of
Remmers and Balchum, at least with respect to
airway resistance. However, these findings must
still be interpreted with some caution, since, as Ury
and Hexter note, correlations between airway
resistance and pollution exposure were significant
at a = 0.05 in only one-third of the individual
subjects.
The relative effects of smoking and pollution on
oxygen consumption are left largely unexplained.
210
-------
The following average correlations between
pollution exposure and oxygen consumption were
computed from Table 10-3;
Smokers
Resting patients,
a.m. oxidants 0.331
Resting patients,
p.m. oxidants 0.231
Exercising patients,
p.m. oxidants 0.292
Non-
smokers
0.034
0.175
0.055
That average correlations are consistently
higher for smokers than for nonsmokers suggests
that variations in smoking habits contributed at
least partly to observed changes in oxygen
consumption.
Rokaw and Massey50 conducted a preliminary
study of the effects of environmental variables on
pulmonary function in a group of 31 patients in a
chronic disease hospital in Los Angeles over a
period of 18 months. A group of normal subjects
was studied concurrently. All of the patients had
chronic, nontuberculous respriatory diseases
(predominantly pulmonary emphysema). Each
subject underwent a series of pulmonary function
tests four times weekly. In addition, function
residual capacity by the helium method was
determined monthly. Air pollution measurements
were obtained from a station about 0.4 km (0 25
mile) upwind from the hospital. Concentrations of
oxidant and oxidant precursor were measured by
the KI method. Concentrations of ozone were also
measured, but the method of measurement was
not reported. Statistical methods were employed to
detect associations between changes in
pulmonary function and air pollution levels. The
authors noted no consistent pattern of response to
episodes of high pollution exposure, though in 6 of
31 patients, coefficients of correlation between
daily pollution levels and corresponding
pulmonary function were high enough to be
termed "interesting." During the Rokaw and
Massey study,60 the mean odxidant concentration
was 120 /jQ/m3 (0 06 ppm), with a maximum of
820 /ug/m3 (0.42 ppm). These concentrations are
quite moderate for the Los Angeles Basin and may
well have been too low to promote any effect on
lung function.
TABLE 1O-3. CORRELATION OF MORNING AND EARLY AFTERNOON OXIDANT LEVELS
WITH OXYGEN CONSUMPTION AND AIRWAY RESISTANCE OF 15 PATIENTS
WITH CHRONIC RESPIRATORY DISEASE45
Patients
number
and
smoking
history8
102 S
[103 NS]
[104 NS]
106 S
[107 S]
[108 S]
110 S
[111 NS]
[112 S]
[113 NS]
[114 S]
[115 S]
[116 SS]
[117 S]
[118 NS]
Number of
observations
Oxygen
consump-
tion
11
14
14
17
17
18
14
15
15
12
13
14
14
15
15
,,z,,,
Airway
resis-
tance
17
14
14
17
17
18
16
17
15
12
13
14
15
15
15
Maximum
breathing
capacity
hters/mm
884
88 4
994
69 1
117 7
52 9
181 5
65 3
26.3
961
38 6
374
Observed correlation coefficients
J^xygen consumption0
Resting
a m oxidants p m n*idanis
0 282
0 123
-0313
0473
0473
0489
0 255
0434
0413
-0 107
-0114
0,423
0 288
29919
0405
0210
0007
0579
0 579
0448
-0 136
0092
0209
0222
-0 158
0 290
-0094
0 189
0345
27619
Exercising
p m oxidanis
0774
0251
0258
0521
0521
0409
-0 172
-0459
-0348
0088
-0 120
0 130
0751
0456
0 138
28379
Airway resist.
Resting
a m oxidants11 p
-0379
0717
0638
-0361
-0361
-0378
0431
0251
0339
0034
-0 161
0217
0715
3621"
a nee'7
m oxidants*"
-0313
0567
0641
0 146
0 146
0656
0 124
0 354
0433
0006
0.058
0557
0460
0609
0453
4976"
*5mokmy history S - smoker, MS = nonsmoker. SS '-'- Stopped smoking when study began Brackets indicate patients who were tested during the same period
kyhe firs! two correlations are with resting oxygen, which was measured around 11am, the third correlation is with exercise oxygen consumption, which was measured
around 3pm
f Airway resistance is measured while resting Values given are averages of four measurements made throughout the day
aa m oxidant was measured around 9 30 a m
e p m oxidant measured around 1 30 p m
i-values were found by converting the individual correlations to t values, us my the relationship t, - rsf(ni- 2} (1 ri21j 1 2 then sum mi ng the t values overall patients, the sums
have variance V(It,J - ?l{r\>-2$/{n,~4)] The z values shown are the ratio of 5{t,} '[J(n-2)- In,--1)1, which is approximately NlO.l}
^Significant at the 0 01 level
''Significant at the 0 001 ievel
211
-------
The analyses presented in Rokaw's report are
quite preliminary in nature. To the best of our
knowledge, a thorough analysis of Rokaw's data
has not yet appeared in the scientific literature
Shoettlin52 studied the effects of community air
pollution, occupational exposure to air pollution,
and smoking among Armed Forces veterans living
in the Domiciliary Unit and Chronic Disease Annex
of the Los Angeles Veterans Administration
Center Day-to-day variations in the physical
status of men with chronic respiratory disease
were studied in relation to changes in
environmental conditions in the coastal area of the
Los Angeles Basin. Two groups of men were
selected for study. One of them consisted of 200
men with clinically determined chronic respiratory
disease The second group consisted of 200
asymptomatic men, matched on age and smoking
history with the first group. Case-control pairs of
men were studied weekly by means of repeated
pulmonary function tests and responses to a
respiratory symptom questionnaire. The study was
performed between mid-August and mid-
December 1958.
An air pollution monitoring station was set up at
the site. Concentrations of total oxidant and
oxidant precursor were measured by the Kl
method
Multiple regression analysis showed no
statistically significant effect of air pollution on
respiratory function or symptoms. However, air
pollutant measurements consistently explained
more variation in symptom frequency and
objective findings in the diseased group than in the
control group. In this study, the statistical effect of
maximal oxidant precursorondependentvariables
was slightly stronger than the effects of other
pollution variables measured
Peters44 examined pulmonary function in
shipyard workers (welders, pipecoverers, and
pipefitters) exposed to a variety of pollutants in-
cluding metal fumes, asbestos, nitrogen oxides,
and ozone. The mean concentration of ozone to
which welders were exposed was 0.10 ppm(range
0.01 to 0 36 ppm); the mean concentration of
nitrogen dioxide was 0.04 ppm (range 0.01 to 0.08
ppm). The results of the study supported the
hypotheses (1) that chronic exposure to oxides of
nitrogen and ozone among welders favors devel-
opment of obstructive lung disease, perhaps in
association with elastic recoil and increased
residual volume, and (2) that chronic exposure to
asbestos among pipecoverers promotes the
development of restrictive lungdisease(decreased
total lung capacity). The pulmonary function of
these groups was consistently less robust than
that in pipefitters engaged in new ship con-
struction who had minimal or no exposure to
asbestos or welding fumes. These data suggest
that chronic exposure to ozone or nitrogen oxides
may contribute to the development of chronic lung
disease. However, since welders were exposed to
metal fumes and some asbestos as well as ozone
and nitrogen oxides, the data do not allow a
specific inference as to what the primary toxicant
might have been.
DISCUSSION OF STUDIES OF AGGRAVATION OF
EXISTING RESPIRATORY DISEASE
Available evidence tentatively suggests an
association between exposure to ambient
pollution in Los Angeles and both increased
frequency of asthma attacks and decrements in
pulmonary function among those with chronic
lung disease. Studies published to date suggest
that a maximum hourly oxidant concentration of
0.25 ppm may be sufficient to promote an increase
in the proportion of asthamatics having attacks.
It must be noted clearly, however, that no study
suggesting an association between oxidant
exposure and exacerbations of chronic respiratory
disease is without serious limitations. In studies
performed to date, variations in the smoking habits
of subjects have not been adequately considered.
Also, investigators have tended to consider oxidant
to the exclusion of other pollutants present in the
air. It is possible that pollutants other than oxidant,
acting singly or in combination with oxidant
components, may have accounted at least partly
for observed effects. Substantial further research
in which covariates are appropriately treated and
in which all pollutants present are considered
must clearly be performed before the acute effects
of photochemical oxidant pollution on persons
with existing respiratory disease are fully under-
stood.
Effects of Oxidants on the Promotion of
Symptoms and Illness in Healthy Populations
SYMPTOM REPORTING AMONG STUDENT
NURSES IN LOS ANGELES
Hammer et al 18 examined symptom reporting in
student nurses in relation to photochemical
oxidant exposure. Freshman students at two
nursing schools in Los Angeles were invited to
212
-------
participate in a prospective study of viral
respiratory disease To minimize bias, neither
faculty nor students were informed that effects of
air pollution were of major interest. Air quality was
measured at monitoring stations located 1.5 to 3.0
km (0.9 to 2 miles) from both hospitals. Oxidant
concentrations were measured by the Kl method.
Daily symptom diaries, which were collected each
week, included questions related to headache, eye
discomfort, cough, and chest discomfort.
Symptoms were graded mild, moderate, or severe.
Symptom reporting covered the period October
1961 through June 1964. A daily average of 61
students participated
Patterns of symptom reporting from both schools
were similar and were pooled in all analyses. All
three grades of symptoms were considered
positive reponses, although the occurrence of
moderate or severe symptoms was rare
Simultaneous daily measurements of oxidants,
carbon monoxide, nitric oxide, nitrogen dioxide,
sulfur dioxide, and maximum daily temperature
were available on more than 90 percent of the 868
days in this study.
Results of the study are presented inTable 10-4.
Simple headache frequency rose slightly at and
above oxidant concentrations of 0.25-0.29 ppm.
The frequency of simple eye discomfort, however,
increased as daily maximum hourly photochemical
oxidant levels exceeded 294 to 372 ,ug/m3 (0.1 5 to
0.19 ppm), Simple rates of cough remained
relatively constant until the maximum hourly
oxidant reached 588 to 764 A/g/m3 (0.30 to 0 39
ppm), at which level the rate began to increase
fairly steeply. The simple rate of chest discomfort
began to rise at an oxidant concentration of about
0.25 to 0.29 ppm
The authors also computed adjusted symptom
rates by excluding days in which student nurses
reported a fever. For all symptoms except
headache, the relationship of adjusted symptom
rates to oxidant exposures was similar to that of
simple rates. Temperature-fever adjusted head-
ache rates increased slightly at daily maximum
hourly oxidant levels of 015-0.19 ppm and in-
creased more steeply as oxidant levels reached
588 to 764 Aig/m3 (0.30 to 0.39 ppm).
Symptom frequencies in this study were more
closely related to photochemical oxidants than to
carbon monoxide, nitrogen dioxide, or daily
minimum temperature Oxidant concentrations at
which rates of eye discomfort were observed to
increase were comparable to those observed in
other published studies. Oxidant concentrations at
which cough and chest discomfort rates were
observed to increase were quite similar to ozone
concentrations observed to produce impairment of
pulmonary function and respiratory irritation in
Bates' experimental studies of humans.1 In these
experimental studies, healthy males performing
intermittent light exercise were exposed to 725
jjg/m3 (0 37 ppm) ozone for 2 hr. The tolerance to
the effects of photochemical oxidants of healthy
young adults like the student nurses studied by
Hammer et al. may be different from that of other
population segments like the aged, the very young,
the ill, and the pregnant.
The results of Hammer's study18 arouse interest as
to what the effects of photochemical oxidants on
these other population segments might be
TABLE 10-4. RELATIONSHIP OF AVERAGE DAILY SIMPLE AND ADJUSTED9 PERCENTAGE OF STUDENT NURSES
REPORT TO PHOTOCHEMICAL OXIDANT LEVELS FOR 868 DAYS, NOVEMBER 1961 THROUGH MAY 1964
IB
Daily maximum
hourly OMdant
level, ppm1*
005
010
o IE
020
025
030
0.40
Overall
£004
- 008
009
- 0.14
-019
-024
-029
-039
- 0.50
average
Number
of
days
229
184
35
176
144
63
25
' 9
3
Average number
of nurses
reporting daily
64
59
58
62
58
60
60
67
53
61
Average daily percent of symptoms
Headache
Simple
165
163
160
15 6
15.7
156
167
169
168
16.1
Adjusted
105
107
106
11 0
11.4
11 6
11 5
134
150
109
Eye discomfort
Simple
86
9.2
90
8.6
10.0
12 2
14 9
21 1
35 0
9.6
Adjusted
50
54
56
5.9
6 9
9.2
11 2
17 8
31 8
6.3
reported
Cough
Simple
12 7
134
13 3
11.9
12 3
11.6
124
152
19 3
12 6
Adjusted
9 1
9 9
102
9.4
9 7
9.1
96
11.7
169
9.5
Chest discomfort
Simple Adjusted
35 18
36
3 2
34
37
3 1
8
9
8
7
6
3.6 20
41 23
70 5.8
35 18
"Ail days on which the system was reported along wsth "feverish," "chilly," or "temperature" are excluded
bl ppm - 1960 jrtj/ma
213
-------
EFFECTS OF OXIDANTS ON NEW ILLNESS IN
COLLEGE STUDENTS
Durham12 reported a study of the short-term
effects of air pollution on the health of students at
seven California universities—five in the Los
Angeles area, and two in the San Francisco Bay
Area The Los Angeles area generally experiences
higher pollution concentrations than the Bay Area
All seven universities participating in Durham's
study were located within 8 km (5 miles) of a
weather station and a pollution monitoring station.
Nine indices of weather and eight indices of
pollution were considered. The method by which
oxidant concentrations were measured was not
reported, although presumably the method was Kl,
at least in the Los Angeles area
Health data were collected during the 1970-71
school year. Each time a student visited the
student health service with a health-related
complaint a sheet requesting demographic
information, smoking habits, and physician's
diagnosis was completed. Diagnoses were
assigned to 1 of 14 categories, including eye
irritation, respiratory and allergic diagnoses, and
gastroenteritis (a diagnosis not expected to be
related to pollution exposure). Only first visits with
individual illnesses were included in data analysis
About 22 percent of collected data sheets showed
no physician's diagnoses. In all, 11,659 sheets
with first-visit respiratory diagnoses were
received.
For each school, the author computed
coefficients of correlation between levels of
pollution and weather variables on 1 day and
proportions of new illness on the same day and on
days up through 7 days later
In general, correlations of pollution with
bronchitis were greatest when lagged 5 or 6 days;
correlations of pollution with combined respiratory
disease were greatest when lagged zero to 3 days;
and correlations with asthma, eye irritation,
headache, and hay fever were greatest when
lagged zero days or 1 day.
Between 20 and 25 percent of pollution/
respiratory illness correlations were statistically
significant at a - 0.05, whereas only 5 percent of
pollution/gastroenteritis correlations were
significant The illnesses most strongly associated
with pollution were, in descending order,
pharyngitis, bronchitis, tonsillitis, colds, and sore
throat. The pollutant variables most strongly
associated with illness in general were, in
descending order, peak oxidant, mean sulfur
dioxide, mean nitrogen dioxide, and mean nitric
oxide. Stronger associations between pollution
and illness were generally observed in the Los
Angeles area than in the Bay Area The authors
attributed this finding to the difference in pollution
exposure between the two areas, although
pollution measurements in the two areas were not
compared m the report
In a separate ana lysis, the days of the study were
separated into high- and low-pollution days for
each school in the Los Angeles area. High-
pollution days were defined as days on which
concentrations of at least half the measured
pollutants exceeded their means by at least 0.25
standard deviation. Days on which concentrations
of at least half the measured pollutants we.e less
than 0.25 standard deviation below their means
were assigned to the low-pollution category For
each school, the authors calculated ratios of upper
respiratory illness rates occurring on or following
high-pollution days to illness rates occurring on or
following low-pollution days. At the University of
California, Irvine, which consistently experienced
lower pollution levels than any other school in the
Los Angeles area, this ratio was very close to 1.0
for all lag periods from zero through 7 days
between the occurrence of pollution and the onset
of symptoms. At the University of Southern
California (USC), which experienced the highest
pollution levels of any school, this ratio was larger
than that observed at Irvine for all lag periods
except zero days. On the average, the ratio at USC
.exceeded the ratio at Irvine by 16.7 percent. The
largest mterschool difference in ratios, about 50
percent, occurred at a lag period of 6 days The
methods of analysis used in this study did not
permit inferences as to the lowest oxidant levels
necessary to promote increased illness rates.
OXIDANTS AND ELEMENTARY SCHOOL
ABSENTEEISM
Wayne and Wehrle64 examined the effects of
oxidant air pollution on respiratory illness among
children as measured by absentee rates from
elementary school. Absentee data were collected
from two elementary schools in Los Angeles
throughout the 1 962-63 school year. Reasons for
absences were initially obtained from parents and
were classified into various categories by teachers.
Air pollution measurements were supplied by the
Los Angeles Air Pollution Control District. Oxidant
concentrations were measured by the Kl method.
Measurements for school 1, located in east
214
-------
downtown Los Angeles, were made at a station
located within 3 km (2 miles) of that school. Air
quality data for school 2 were monitored at a
station within 3 km (2 miles) of the school during
the initial 3 months of study and at a station 6.5 km
(4.5 miles) from the school for the remainderof the
study. During this study, daily oxidant
concentration averaged from 10 a.m. to 3 p.m.
ranged from nearly zero to about 0.23 ppm. This
concentration exceeded 0.20 ppm on 6 (1.8
percent) of a possible 336 days. Oxidant was the
only pollutant considered in this study.
Absence rates in both schools were highest
during the winter season, when levels of oxidants
were generally lowest. No consistent association
was observed between weekly mean oxidant
levels and weekly absence rates for respiratory
disease. The highest absence rates were reported
on Mondays and Fridays at both schools; in
contrast, oxidant levels were consistently higher
during midweek periods. In a further analysis,
correlations were computed between uncorrected
daily absence rates and same-day average oxidant
concentrations between 10 a.m. and 3 p.m. In a
second analysis, absence rates corrected for day-
of-week effect were correlated with same-day 10
a.m. to 3 p.m. oxidant concentrations and with
corresponding concentrations occurring 1 day
before. Results from the two schools were kept
separate in all analyses. No analysis revealed a
positive association between oxidant con-
centration and absence rates.
EFFECT OF PHOTOCHEMICAL OXIDANTS ON
SYMPTOM RATES IN JAPANESE ELEMENTARY
SCHOOLCHILDREN
Shimizu55 discussed the effects of photo-
chemical smog exposures on symptom reporting
rates at two junior high schools in Osaka, Japan,
which will be designated schools A and B. At
school A, an unusually high number of students
complained of symptoms on 3 days in the fall of
1972 (September 13, 21, and 22). An unusually
high number of eighth graders at school B
complained of symptoms on October 3, 1972. At
each school, symptom questionnaires were
distributed on the day or the day after the unusual
complaint rates occurred. The questionnaires
inquired about the presence of nine symptoms,
including eye irritation, respiratory discomfort,
cough, dizziness, nausea, and numbness of the
extremities.
At school A, questionnaires appear to have been
distributed to all students. However, they were
collected only from students who reported the
presence of at least one symptom. Faced with
uncertainty as to the appropriate denominator, the
author selected the total enrollment of school A,
(1313 students) as the basis for calculation of
symptom rates.
Maximum concentrations of pollutants
measured at a station near school A (Station P) on
September 1 3 were as follows:
Pollutant: ppm
Oxidant (3 p.m.) 0.14
Nitric oxide (2 p.m.) 0.17
Nitrogen dioxide (2 p.m.) .0.1 2
Sulfur oxide (1 p.m.) 0.6
The maximum temperature, 29.4°C (85°F), was
recorded at 11 a.m. At another nearby station
(Station 0), a maximum oxidant concentration of
0.1 9 ppm was recorded between 2 and 3 p.m. Only
Station 0, which measured only oxidant, was
operating on September 21 and 22. At this station,
the maximum oxidant concentration on September
21 was 0.17 ppm, recorded at 1 p.m. On
September .22, the maximum oxidant
concentration was 0.11 ppm, recorded at 1 2 noon.
Aerometric measurement methods were not
reported.
At school A, a total of 263 students (20.0
percent) reported at least one symptom on
September 1 3. Corresponding totals on September
21 and 22 were, respectively, 133 (10.1 percent)
and 72 (5.5 percent).
It is of interest that on September 13, the rate of
symptom reporting in physical education classes
increased as concentrations of oxidants, NO, and
NC>2 increased. Physical education classes were
conducted outdoors. The average rate of symptom
reporting among students not having physical
education classes that day and students having
class in the morning was 7.5 percent. The highest
rate of symptom reporting, 50.9 percent, was
observed in the class which met between 1 and 2
p.m. Peak levels of NO and N02 were recorded
during this hour. As mentioned, the peak level of
sulfur oxide had been recorded at 1 p.m. In the
class that met 1 hr later, when peak levels of
oxidant were recorded, the rate of symptom
reporting dropped slightly to 45 percent. On
September 21 and 22, when both oxidant
concentrations and symptom rates were
somewhat lower than on September 1 3, symptom
215
-------
rates were distributed randomly among physical
education classes. Unfortunately, concentrations
of pollutants other than oxidant were not
measured on these two days.
The author noted that on September 1 3, not only
the rate of symptom reporting but also the
distribution of symptoms changed with physical
education class. Among students not having
physical education class or having it in the
morning, 94.6 percent of students with symptoms
reported eye irritation. Only 37.5 percent of these
students complained of sore throat, and 19.6
percent complained of coughing. In the physical
education class meeting between 1 and 2 p.m.,
however, only 50 percent of students with
symptoms reported eye irritation, while 81.0
percent reported sore throat, 60.7 percent reported
coughing, and 59.5 percent reported chest
discomfort. These findings suggested to the author
that the components of photochemical pollution
that promote eye irritation may be different from
those that promote respiratory symptoms.
At school A, the majority of students with
symptoms on September 21 and 22 complained of
eye irritation.
At school B, as mentioned, an unusually high
number of eighth-grade students complained of
symptoms on October 3, 1972. At two nearby
stations, maximum oxidant concentrations of 0.09
and 0.07 ppm were recorded on that day. No
mention of other pollutant concentrations was
made in Shimizu's report. Of 248 students
surveyed, 152(61.3 percent) reported some type of
symptom. The pattern of symptom reporting at this
school was somewhat different from either pattern
observed at school A. At school B, 53.5 percent of
males and 71.7 percent of the females surveyed
reported symptoms, whereasatschool Athere had
been little difference in symptom rates between
the sexes. In school B, higher percentages of
students with symptoms than in school A reported
such nervous-system-related symptoms as
headache, dizziness, nausea, and numbnessof the
extremities (which can result from hyper-
ventilation).These symptoms, like chest discomfort
occurred more frequently in females than in males.
In school B, only 21.7 percent of students with
symptoms reported eye irritation.
Shimizu's findings arouse interest astowhether
the components of photochemical air pollution that
promote nervous-system-related symptoms like
those described above are differentfrom eitherthe
components that promote eye irritation or those
that promote respiratory symptoms. As yet,
however, these findings are far from conclusive,
particularly since pollutants other than oxidants
were in the ambient air during this study.
Makino and Mizoguchi32 reported a study,
performed over the year July 1972 to June 1 973,
of the influence of air pollution exposures on rates
of symptom reporting in students. This study
included 854 students, of whom 110 were in
kindergarten, 327 were in elementary school, 335
were in junior high school, and 82 were in senior
~high school. All students attended schools in the
northern part of Tokyo. Each day they completed a
questionnaire inquiring about the presence of 17
symptoms, including eye irritation, irritation of the
upper respiratory tract, cough, phlegm production,
fatigue, headache, and fever. Daily rates of
symptom reporting were correlated with same-day
pollution concentrations. We believe that the
pollution concentrations used in statistical
analyses were daily maximum hourly
concentrations, but this is not clearly stated in the
report. Measured pollutants included oxidant,
nitric oxide, nitrogen dioxide, sulfur dioxide, and
micro-particulates. Measurement methods were
not reported. Daily maximum hourly daytime
temperature and minimum hourly daytime
humidity were also considered.
The data were analyzed in several different
ways. In the first analysis, rates of symptom
reporting on two days of high pollution, one in July
and one in August, were computed. On the first
day, maximum hourly concentrations of oxidant,
nitrogen dioxide, nitric oxide, and sulfur dioxide
were, respectively, as follows: 0.17, 0.10, 0.10,
and 0.03 ppm. On the second day, maximum
hourly concentrations of Ox, NOs, NO, and S02
were, respectively, as follows: 0.21, 0.07, 0.09,
and 0.14 ppm. On both days, rates of nearly all
symptoms were higher than corresponding
monthly average rates. The elevation in rates was
most marked for eye irritation, which an average of
10.7 percent of students reported on the two high-
pollution days, versus an average of 1.6 percent in
the months of July and August. Rates of sore
throats (average of 4.0 percent on high-pollution
days versus 0.65 percent in corresponding
months) and headache (2.9 vs. 1.0 percent) were
also elevated noticeably.
In the next analysis, individual environmental
measurements were correlated with same-day
symptom rates within each month of the study. Of
105 oxidant/symptom correlations, 37 (35
216
-------
percent) were significant at a - 0,05, Of these, 14
(13 percent) were significant at or = 0.001. Of 98
SOz symptom correlations, 22 (22 percent) were
significant at a - 0.05, and 10 (10 percent) were
significant at 0.001. Of 105 NO2/symptom
correlations, 22 {21 percent) were significant at a
= 0.05, and none were significant at a - 0.01. Of
116 temperature/symptom correlations, 19 (16
percent) were significant at or = 0.05. Only 5 (6
percent) of 88 NO/symptom correlations were
significant at a = 0.05. The directions of these
correlations were not presented.
When these types of correlations were
computed over the whole year of study, 6 (46
percent) of 13 oxidant/symptom correlations were
significant at a = 0.01. All coefficients were
positive. The association between oxidant
exposure and symptom rate was most marked for
dyspnea (for which the oxidant/symptom
correlation was 0.653) and for eye irritation,
lacrimation, and sore throat. For each of these
latter three symptoms, the oxidant/symptom
correlation was 0.621 (p < 0.001). No yearly
symptom/pollution correlations were significant
at or = 0.05 for S02, NOz, or NO. Seven of 13 (54
percent) .yearly temperature/symptom correla-
tions were significant ata=0.05. AH seven were
positive. Coefficients of correlation between
oxidant concentrations and symptoms were
generally larger than those between temperature
and symptoms.
In another analysis, pairs of environmental
variables were correlated with same-day rates of
symptom reporting. Correlations between oxi-
dant/NOz and symptom rates were significant
at a = 0.05 in 9 of 13 cases (69 percent).
Correlations between oxidant/SOz and symptom
rates were significant in 7 of 13 cases(54 percent),
as were correlations between oxidant/tempera-
ture and symptom rates. Environmental variable
pairs that did not contain oxidant were significantly
correlated with symptoms less often and less
strongly than those that contained oxidant.
In another analysis, the authors compared
symptom rates on days when the maximum hourly
oxidant level exceeded 0.10 ppm to rates on days
when it did not. Such comparisons are shown for
six symptoms in Figure 10-4. As the figure shows,
symptom rates, particularly rates of eye irritation
and sore throat, were quite consistently elevated
on the higher-oxidant days. The report does not
clearly state why the months of January through
April 1973 are missing from the line showing
symptom rates on higher-oxidant days in Figure
10-4. It is probably safe to surmise, however, that
since these are winter or early spring months, they
contained very few days on which the oxidant
concentrations exceeded 010 ppm,
Mizoguchi et al.37 investigated the effects of
short-term air pollution exposure on 51 5 students
in a junior high school in southeastern Tokyo. Over
a 2-month period between May and July 1974,
these students completed a daily questionnaire
inquiring about the presence of 1 7 symptoms of
the respiratory and other systems. Air pollution
measurements were made on the school grouVids.
Oxidants were measured each hour by the KI
method; ozone was measured by (presumably
ethylene)chemiluminescence; and nitric oxide and
nitrogen dioxide were measured by the Saltzman
method. Concentrations of eight other pollutants
were also measured. The maximum hourly oxidant
level observed during this study was 0.23 ppm. No
frequency distribution of pollutant concentrations
was presented.
In still another analysis, Makino and Mizoguchi
compared symptom rates on days when the
oxidant concentration exceeded 0.1 5 ppm to rates
on days when oxidant concentration did not
achieve 0.10 ppm. Results of this comparison are
shown in Figure 10-5. As expected from the results
of other analyses of these data, increases in
symptom rates on the higher-oxidant days were
most marked for eye irritation, sore throat,
headache, and coughing.
Three types of statistical analyses were
performed on data gathered from all students. In
the first type, a simple correlation analysis, daily
maximum hourly measurements of each pollutant,
were correlated with the proportion of students
reporting each symptom on the same day the
pollution measurements were made. Correlation
coefficients statistically significant at a = 0.001 (r>
0.5) were computed between oxidant
concentrations and 5 of 17 symptoms, including
eye irritation, shortness of breath, sore throat,
headache, and blurred vision. Positive coefficients
significant at a = 0.01 (r > 0.35) were computed
between oxidant concentrations and seven other
symptoms. Ozone was found to correlate
significantly at a = 0.001 with three symptoms (eye
irritation, shortness of breath, and sorethroat)and
at a = 0.01 with seven other symptoms. Sulfur
dioxide, like oxidant, was found to correlate
significantly with 1 2 of 17 symptoms at the 0.01
level. Suspended particulate matter correlated
217
-------
significantly with 13 symptoms. Daily minimum
relative humidity correlated negatively and
significantly with 15 symptoms. No correlations of
pollutants with other pollutants and with
meteorologic variables were presented.
In the second analysis, daily proportions of each
individual symptom were correlated multiply with
pairs of same-day measurements of environ-
mental variables (The report was unclear as to
whether all possible environmental-variable pairs
or only selected pairs were considered.) Six such
pairs were correlated significantly at a = 0.001 (r>
0.6) with both eye irritation and shortness of
breath. Oxidant was a member of all 12 of these
pairs Seven pairs, of which six contained oxidant,
were correlated significantly with sore throat. Five
pairs correlated significantly with hoarsness, but
only one of them contained oxidant. None of the
pairs of environmental variables that correlated
significantly with symptom rates contained ozone,
but whether ozone was included in any of the
tested pairs is not clear.
The third analysis was a principal components
analysis. Three environmental principal
components were derived. The first was loaded
heavily for oxidant, ozone, and sulfur dioxide; the
second was loaded heavily for most pollutants
except oxidant and ozone (the second also included
OXIDANT CONCENTRATION
- ^ 0.10 ppm
——— — < 0.10 ppm
SORE THROAT
LACRIMATION
-L-l I- 1 1 III I I
8 10 12 2 4 6 8 10 12 2 4 6 8 10 12 2 46
PERCEPTION OF
STRANGE ODOR
8 10 12 2 4 6 8 10 12 2 4 6 8 10 12 2 4 6
TIME, month and year
Figure 10-4. Monthly changes in symptom and complaint rates with daily maximum
hourly oxidant content above and below 0.10 ppm.32
218
-------
sulfur dioxide); and the third was loaded heavily for
temperature and carbon monoxide. Nearly all
symptoms, particularly such irritation symptoms
as eye irritation, sore throat, hoarsness, cough,
and phlegm, were associated strongly with the first
principal component. No symptoms were strongly
associated with either of the other two
components.
From the overall study sample, the authors
selected a group of 74 students with allergic
tendencies and a group of 47 with orthostatic
dysregulation (O.D.). In the report, the author
stated that O.D. is not the same entity as
orthostatic hypotension, though he did not define
the condition further, O.D, appears to be a
condition of unusually intense reactivity to stress.
Over the entire study period, the group with O.D.
reported noticeably but not significantly higher
rates of cough, headache, and sneezing than either
of the allergic group or the students as a whole.
Symptom rates in the allergic group were generally
higher than in the students as a whole, but not as
markedly as in the group with O.D. On a day when
maximum hourly oxidant achieved a concentration
of 0.23 ppm, rates of symptom reporting increased
in all groups. Increases in symptom rates were
largest in the O.D. group, particularly for eye
irritation, cough, and headache. Again, symptom
rates in the allergic group were intermediate
between the students as a whole and the group
with O.D, In this report, the author mentions
(though does not reference) a concurrent
investigation in which no effect of air pollution
exposure on pulmonary function was observed.
This finding is consistent with the hypothesis that
subjective symptoms may sometimes reflect
pollution exposures more accurately than
objective parameters.
The Japanese Environment Agency conducted a
large survey of student health in 1975. Results
were drawn from schools in seven prefectures
throughout the country. A preliminary analysis of
these results has been reported.21 The period
covered by this survey is not clearly stated, but it
SYMPTOMS/COMPLAINTS RATE, percent
SORE THROAT ••••^••l
HEADACHE •••••
COUGHS ••••
DYPSNEA ••••
LACRIMAT1ON •••
RHINORRHEA ••
PERCEPTION OF STRANGE ORDERS ••
DISCOMFORT ••
FEVERISHNESS ••
FATIGUE ••
NASAL CONGESTION •!
EXPECTORATION ••
DIZZINESS •
NAUSEA •
NUMBNESS 1
CONVULSIONS
| 1 1 i 1 1
ID
1
m
ID
3
_j
3
3
D
H
3
3
3
3
I J
6 5 4 3 2 1 0 0 0.1 0.2
Figure 10-5. Comparison of various symptom and complaint rates on days when
oxidant concentrations were over 0.1 5 ppm and below 0.10 ppm.32
219
-------
appears to have been about 180 days. The majority
of students included in the survey were in junior
high school, although students in elementary and
senior high school were also included.
Unfortunately, the report of the survey often
presents numbers of people reporting illness, but
not rates of illness. Rates cannot be computed
since the report does not present the total number
of persons included in the survey. Also, the report
does not present the methods used to generate
aerometric data, so its findings are quite difficult
to compare with findings made in the United States.
The primary survey tool was a symptom
questionnaire (in some prefectures self-
administered, and in others interviewer-ad-
ministered) that was completed by students who
complained of symptoms at school. It must be
remembered that this survey was based primarily
on students who voluntarily reported symptoms.
No systematic effort was made to determine either
the extent of nonresponse or the effects that
nonresponse bias may have exerted on the results.
Thus, as the authors acknowledge, the majority of
findings of this survey, though interesting and
suggestive, must be considered preliminary and
inconclusive.
The most common symptom reported was eye
irritation which, on the average, constituted about
36 percent of all symptoms reported. About 30
percent of all symptoms reported were general
symptoms, including numbness in the extremities,
headache, dizziness, nausea, and lethargy.
Irritation of the nose and throat constituted about
20 percent of all reported symptoms. Respiratory
symptoms, including cough, chest pain, and
phlegm production, also accounted for about 20
percent of all reported symptoms.
In Saitama prefecture, 10 schools containing
7440 students were included in the survey. On
days when the oxidant concentration never
exceeded 0.15 ppm, an average of only 0.052
percent of students reported smog-related
symptoms. On days with oxidant concentrations of
at least 0 15 ppm uut less than 0 25 ppm, this
percentage rose to 0.48 percent. On two days
when oxidant concentrations were at least 0.25
ppm, an average of 16.4 percent of students
reported symptoms. A similar (though less strictly
designed) survey conducted in Kanagawa
prefecture apparently yielded similar results. In
one elementary school, 140 of 1510 students (9
percent) complained of smog-related symptoms on
one afternoon when the following pollution
measurements were made at a nearby monitoring
station: oxidant, 0.253 ppm; nitrogen dioxide,
0.018 ppm; and oxides of sulfur, 0.042 ppm. (This
is the only instance in the report in which
measurements of pollutants other than oxidants
are presented.) Eleven students reporting feelings
of choking and numbness in the extremities were
taken by ambulance to a nearby hospital. Detailed
family and personal histories were obtained from
five hospitalized students. Of these, all had positive
personal histories of allergic conditions, and four
had positive family histories.
DISCUSSION OF OXIDANT EFFECTS ON
PROMOTION OF SYMPTOMS AND ILLNESS IN
HEALTHY POPULATIONS
A consistent positive association between
oxidant concentrations and rates of new illness
and symptoms has been observed both in the
United States and in Japan. The Hammer et al.
study of student nurses18 suggests that rates of
chest discomfort and cough begin to increase at
maximum hourly oxidant concentrations of 0.25 to
0.29 ppm and 0.30 to 0.39 ppm, respectively. As
mentioned, this finding isquiteconsistent with the
results of the human experimental studies by
Bates et al., in which decrements in pulmonary
function and respiratory irritation have occurred at
an ozone level of 0.37 ppm. Recent Japanese
studies suggest that cough, chest discomfort, and
other symptoms may occur in school children at
oxidant concentrations between 0.10 and 0.15
ppm.
The consistency between Hammer's epi-
demiologic observations and Bates' experimental
findings suggests that ozone alone may well
account for some of the symptoms associated with
oxidant exposures in epidemiologic studies.
However, it appears quite clear that ozone does not
account for all the symptoms observed in the field.
(Indeed, that ozone does not produce the eye
irritation so often associated with photochemical
oxidants has been quite well established.) Results
of the Japanese studies cited above (in which
correlations between total oxidant and symptom
reporting rates were larger and more consistently
significant than ozone/symptom correlations)
bring nonozone components of photochemical
pollution under suspicion.
As mentioned, Shimizu55 observed that the
majority of students reporting symptoms in the
morning of a high-pollution day complained of eye
irritation. By the afternoon of the same day, when
220
-------
ozone concentrations had risen above morning
levels, the majority of students with symptoms
complained of respiratory discomfort. This
observation, coupled with the findings of Bates and
Hammer, appears to favor the hypothesis that
certain oxidants other than ozone, though perhaps
including ozone precursors, produced the eye
irritation experienced in the morning, and that
ozone was an important contributor to the
respiratory irritation experienced in the afternoon.
On the basis of findings presented here, it
appears that lower levels of oxidant pollution affect
Japanese subjects more than American subjects.
One or both of the following factors may explain
this disparity. First, the components of Japanese
oxidant pollution may be different from
corresponding components in the United States.
For example, high sulfur oxide levels appear to
accompany high oxidant levels more frequently in
Japan than in the United States (e.g., Los Angeles).
Second, occurrences of elevated pollution appear
to receive considerably more publicity in Japan
than United States. This publicity may influence
symptom, reporting. For these reasons, the results
of Japanese and American studies should be
compared with caution, particularly in view of the
apparent discrepancy between the Japanese
studies and the Wayne and Wehrle study of
absenteeism.64
The Durham study12 and the Japanese studies
cited above arouse interest as to whether
photochemical oxidant pollution predisposes
young populationstothedevelopmentof increased
rates of chronic respiratory disease in adulthood.
Unfortunately, the relationship between acute
illness or symptoms in childhood and chronic
respiratory illness in adulthood is not at all clearly
characterized as yet. Until it has been
characterized, the question as to whether
transient irritative symptoms constitute a bona fide
impairment of health must remain open.
Impairment of Performance Associated with
Oxidant Pollution
ATHLETIC PERFORMANCE
Wayne et al.65 have studied the athletic
performance of student cross-country track
runners in 21 competitive meets at a high school in
Los Angeles County from 1959 to 1964.
Aerometric data were supplied by the Los Angeles
County Air Pollution Control District. Oxidant
measurements for the hour of the race, and for 1,
2, and 3 hr before the race were related to the
percentage of athletes who did not improve their
running times from the previous meet. Oxides of
nitrogen, carbon monoxide, particulates,
temperature, relative humidity, wind velocity, and
wind direction were also considered, but they bore
less relationship to performance than did oxidant.
Evidently, correlations between performance and
sulfur dioxide were not computed.
The authors observed a significant positive
relationship between hourly oxidant levels and the
percentage of team members whose performance
failed to improve from the previous home meet.
This relationship is shown in Figure 10-6. As the
figure shows, the proportion of runners failing to
improve their times rose as concentrations of
oxidant in the hour before the race increased over a
range of 60 to 590 /ug/m3 (0.03 to 0.30 ppm).
However, convincing linear relationships between
oxidant concentration and performance could not
be discerned within individual portions of this
concentration range. For both 1959-61 and 1962-
64, the coefficient of correlations between oxidant
concentration and percentage of runners failing to
improve their times was 0.945. As the interval
between the time of oxidant measurement and the
race increased, corresponding correlation
coefficients decreased.
The results of this study strongly suggest an
effect of some component of the pollution
measured as oxidant on team performance.
Speculating asto possible mechanisms underlying
the observed association, Wayne et al. cited the
possibility that some component of photochemical
smog might elevate the body's demand for oxygen
during exercise. They also felt that breathing
photochemical air pollutants might produce
increased airway resistance or might lead to
discomfort, which might in turn limit the runners'
motivation.
In an analysis performed at the University of
North Carolina, Herman23 also investigated the
effects of air pollution on performance of cross-
country runners. In his analysis, Herman used the
same data that Wayne et al.65 had used from the
cross-country seasons of 1 959-64 as well as data
from the seasons of 1966-68. The dependent
variable in Herman's analysis was the average
speed attained in each meet by each runner who
participated in all home meets in a given season,
corrected both for the runner's average speed over
the season and for the teams'average speed over
the season. This differed from the dependent
221
-------
variable used by Wayne, which was the
percentage of runners who had failed to improve
their times from the previous meet.
Herman constructed regression models that
incorporated all data from 1959-68 and that
employed the following as independent variables:
hourly air pollution measurements, the individual
running seasons, the number of days si nee the first
home meet of the season, the square of this
number of days, the maximum daily temperature,
and the square of this temperature. Stepwise
regressions in which pollution measurements
were entered into the model both before and after
temperature variables were performed. In both
cases, an inverse association between running
speed and oxidant measured in the hour before the
meet (significant at a - 0.0001) was observed.
Little association between running speed andtotal
suspended particulates, carbon monoxide, and
oxides of nitrogen was observed.
Folinsbee and his associates1415 have con-
ducted two experimental studies that suggest a
mechanism underlying Wayne and Wehrle's
epidemiologic observations. These studies are
described in detail in the preceding chapter. In the
first study, healthy young adults (20 males and 8
females, 18 nonsmokers and 10 smokers) were
assigned to one of six exposure groups. Two
groups of healthy young adults received a 2 hr
exposure to 0.37 ppm ozone; two other groups
were exposed to 0.50 ppm; and two others were
exposed to 0.75 ppm. At each exposure level, one
group rested throughout exposure, and the other
underwent alternating 15-min periods of rest and
exercise. Each of the sixgroups was also subjected
to a mock exposure while adhering to its experi-
mental protocol during actual exposure. Attheend
of all exposures and mock exposures, all subjects
underwent a submaximal exercise test.
Respiratory frequency significantly increased
over control values after intermittent exercise at all
exposure levels and with resting exposures to 0.75
ppm ozone. Tidal volume decreased significantly
after intermittent exercise at the 0.50- and 0.75-
ppm exposure levels. Forced vital capacity de-
creased significantly after intermittent exercise at
0.50 ppm ozone, and after both rest and exercise at
0.75 ppm. Maximum expiratory flow rate at 50
percent of vital capacity decreased significantly
after intermittent exercise at all three exposure
levels. Many subjects reported common symptoms
of ozone exposure, including sore throat and
80
LU
o
<
cc
o
o
cc
LU
a.
a
HI
ai
CC
o
LU
a
cc
HI
ca
HI
5
<
01
t-
60
40
20
_ —A— 1959-1961
— O-- 1962-1964
r= 0.945 FOR BOTH LINES
0 0.10 0.20 0.30
OXIDANT CONCENTRATION 1 hour BEFORE EVENT, ppm
Figure 10-6. Relationship between oxidant concentration in the hour before an
athletic event and percentage of the team failing to improve running time.65
222
-------
cough. Symptom severity increased with exposure
level and exercise. No significant changes in
minute volume or in oxygen uptake were observed
after the submaximal exercise tests employed in
this study.
The authors believed that because of the speed
with which they developed, the observed physio-
logic changes were due to reflex broncho-
constriction mediated by irritant receptors and
perhaps by Hering-Breuer (stretch) receptors in the
lung.
In the second study of Folinsbee et al.,14 healthy
young adult males were exposed to 0.75 ppm
ozone for 2 hr. During exposure, subjects
alternately rested for 1 5 min and lightly exercised
for 15 min. In random order, each subject also
underwent a 2-hr control exposure to filtered air.
At the conclusion of all exposures, subjects
exercised until exhausted.
Values of the following parameters, measured
during maximal exercise, were significantly lower
after ozone exposure than after filtered air
exposure: maximum work load attained, heart rate,
minute volume, tidal volume, and oxygen uptake.
The authors believed that the most likely
mechanism underlying the results observed was
that ozone produced stimulation of irritant
receptors in the lung, which in turn produces
restriction of inspiratory reserve volume, reduction
of rru ximal tidal volume, and a modest increase in
airw; y resistance.
In the light of Folinsbee's results, it is quite
conceivable that on days of high oxidant exposure,
ozone irritated the respiratory systems of cross-
cour- ry runners sufficiently to restrict inspiratory
capa :ity and to reduce tidal volume. These
char, jes might in turn have prevented the runners
from taking up enough oxygen to sustain the work
loads they might have sustained en lower-
exposure days. (In this case, the work load
sustainable by the runners would be inversely
proportional to running time.)
Folinsbee's work15 also suggests that a healthy
person exercising submaximally and exposed to
ozone levels near the highest ambient
concentrations ever achieved in Los Angeles is
able to maintain a normal rate of oxygen uptake,
even in the face of changes in respiratory pattern
and spirometric parameters. Thus, the decrements
in running time observed by Wayne in cross-
country runners might not have been seen, or
might not have been as evident, in shorter-
distance runners. Of course, neither the research
of Folinsbee nor of Wayne allows conclusions as to
the effects of ozone on oxygen uptake in the ill
when they are exercising submaximally.
Folinsbee's work has practical significance
beyond these interesting physiological
considerations. In epidemiologic studies,
decrements in health and performance have been
quite commonly associated with elevated
concentrations of photochemical oxidants as a
class. However, the extent to which such
decrements might be related specifically to ozone
exposure has usually been obscure. Though
Folinsbee's studies do not rigorously prove that
ozone exposure caused the Decreased running
times that Wayne observed, they do suggest a
plausible mechanism by which ozone alone might
have produced the decreased times.
AUTOMOBILE ACCIDENTS
To examine the possibility that oxidant pollution
may impair performance, Ury performed a study61
of the association of automobile accidents with
days of elevated hourly oxidant levels. Oxidant
levels supplied by the Los Angeles Air Pollution
Control District were measured at the District's
downtown station. Ury applied two nonpara-
metric analyses, a sign test and a Kendall's tau, to
data for each daylight hour of each weekday m the
3-month period from August through October for
both 1963 and 1965. Ninety sets of data covering 9
hr daily and 5 days weekly were available for
testing. The sign-test results of each set were
obtained by taking successive pairs of weeks (first
week compared with second; third week compared
with fourth, etc ) and scoring a plus (+) if the week
with the higher oxidant had more accidents for that
set, a minus (-) if it had fewer, and a tie if the
accident frequencies or the oxidant levels were
equal (Table 10-5).
All accidents involving driving under the
influence of alcohol, narcotics, mechanical
failures, and spillage on the roadway were
excluded from analysis, as were accidents oc-
curring in rain, drizzle, or fog. The author does not
state the fraction of total accidents included in
these categories, though it may well have been
quite large.
In both analyses, photochemical oxidant
concentrations correlated positively with the
frequency of motor vehicle accidents. The sign test
yielded a descriptive two-sided significance level
of 4 percent. Kendall's tau yielded a level of 5
223
-------
percent. In a further analysis, the authorcompared
accident rates on days when hourly oxidant levels
exceeded 0.15 ppm to rates on days when levels
did not achieve 0.10 ppm. Accident rates were
elevated on the higher-pollution days. A Wilcoxon
two-sample test yielded a two-sided significance
level of 0.07. In this study, oxidant was the only
pollutant whose relationship to auto accidents was
considered.
TABLE 10-5. SIGN-TEST DATA FOR TESTING THE
ASSOCIATION OF OXIDANT LEVELS WITH ACCIDENTS
IN LOS ANGELES, AUGUST THROUGH OCTOBER,
1963 and 196561
Weekday
Monday
Tuesday
Wednesday
Thursday
Friday
Total
Total
Grand total
Year
1963
1965
1963
1965
1963
1965
1963
1965
1963
1965
1963
1965
Plus'
22
18
26
23
20
26
26
31
27
26
121
124
245
Minus3
17
18
11
19
15
18
12
16
18
19
73
90
163
Tie"
6
12
7
9
8
10
11
5
8
7
40°
43C
83
aNumbers of accidents during pairs of hours at the same time of day on the same
day of week, one week apart, were compared When the number of accidents
during the hour with higher oxidant concentration was higher than during the
lower-oxidant hour, a +1 was assigned When the number of accidents in the
higher-oxidant hour was lower, a -1 was assigned Ties occurred when the
number of accidents during the 2 hr was identical
"Significantly positive at a - 0 001 excluding ties, a = 0 01 including ties
GSignificantlv positive at a - 0 05, including or excluding ties
Dry et al. also reported an extension63 of the
study cited above,61 in which the effects of both
oxidant and carbon monoxide concentrations on
motor vehicle accident frequency in Los Angeles
were investigated The investigators assessed the
effects of oxidants over the summers of 1963 and
1965 (the same two periods that had been
considered in the original study), and they
assessed the effects of carbon monoxide over the
same two periods and over the winter months of
1964-65 and 1965-66 as well. The same methods
of data reduction were used in both studies. In the
follow-up study, the sign test was used to correlate
the frequency of accidents with mean pollutant
concentrations during the hour of the accident and
during each of the 3 hr preceding it.
In Ury's second study,63 none of the 90 individual
sets of comparisons specific for day, hour, and year
yielded a statistically significant difference
between the number of positives and negatives
recorded. This is hardly surprising, since no set
contained more than six comparisons. However,
the degree of significance in the relationship
between accident frequency and the oxidant
concentration during the hour of the accident was
greaterthan in the original study.Thisdiscrepancy
between studies arose from a systematic error in
data coding, which was discovered between the
initial and followup studies. With respect to time of
day, the strongest relationships between oxidant
concentration and accident frequency were noted
at9a.m., 10a m.,and 12 noon. This finding raises
the question of whether accident frequency is
more strongly associated with an ozone precursor
than with ozone itself. With respect to day of week,
the strongest relationships were noted on Tuesday
and Thursday. No consistent relationship between
accident frequency and lagged oxidant
concentrations was observed. No relationship
between carbon monoxide concentration and
accident frequency was observed in any of the
periods considered.
Ury et al.63 present a discussion of variables that
might have confounded the observed association
between oxidant concentration and accident
frequency. It is quite unlikely that inaccuracies in
accident reporting or pollutant measurements
could have been large enough to influence study
results appreciably. Disparities between pollution
concentration at accident sites and monitoring
sites would have been fairly well absorbed by the
statistical methods employed. The study design
either negated or minimized effects of hour of day,
day of week, and weather. Finally, the lack of
association between carbon monoxide
concentration and accident frequency argues
against an effect of varying traffic density on the
results observed. However, the degree to which
carbon monoxide measurements actually reflected
traffic density must remain open to question, since
aerometric information was taken only from a
single station.
Possibly other pollutants, particularly nitrogen
and sulfur oxides, may have confounded the
observed association between oxidant
concentration and accident frequency. However, it
is unlikely that such other pollutants could have
accounted for the bulk of the observed effect, since
their concentrations tend to be lower in summer
than in winter.
Ury et al. do not speculate as to a mechanism
underlying the observed association between
224
-------
oxidant concentration and accident frequency.
Conceivably, reduced visual activity, which has
been observed by Lagerwerff26 in humans after a
3-hr inhalation of 392-980,ug/m3 (0.2 toO.5 ppm)
ozone, was partly responsible for the observed
results. Increased eye irritation9'"6'"8'"9'5'1 or
reduced visibility during smoggy periods may also
have been partly responsible.
VENTILATORY PERFORMANCE IN SOUTH-
ERN CALIFORNIA
The ventilatory performance (measured by the
Wright Peak Flow Meter) of two groups of third-
grade children living in the Los Angeles Basin was
assessed twice monthly for 11 months by
McMillan et al.34 One group of 50 children resided
in an area exposed to seasonally high photo-
chemical oxidant concentrations. The other group
of 28 children lived in a less polluted area.
Aerometnc data were obtained from the Los
Angeles County Air Pollution .Control District.
During the 11 months of the study, no correlations
between short-term changes in photochemical
oxidant pollution and ventilatory performance
were observed. Persistently higher ventilatory
performance results were obtained from the
children residing in the more polluted of the two
communities
The results of this study are difficult to interpret
because the two groups of children tested were
different in several respects. The children with
lower ventilatory performance reported upper
respiratory symptoms three times more frequently
than the children with higher performance. Also,
the majority of children in the less polluted
community were from a single ethnic group,
whereas children in the other community were
ethnically heterogeneous. As closely as we can
determine, no statistical adjustment of the data for
ethnic differences was made. Furthermore, as
noted by the authors, nitrogen dioxide levels were
consistently lowest during the winter in the city
with elevated oxidant exposures. It is conceivable
*hat the effects of nitrogen dioxide on the study
results may have offset those of oxidant. Finally,
differences in oxidant exposure in the two towns
appear not to have been substantial. During the
days of this study, the mean daily oxidant
concentration in the low-oxidant city was 0.088
ppm. The corresponding mean in the high-oxidant
city was 0.114 ppm.
PULMONARY FUNCTION IN ARIZONA
Lebowitz et al.27 reported a study of the
combined effects of air pollution and weather on
the ventilatory function of exercising children,
adolescents, and adults in Tucson, Arizona. The
study was conducted m the spring and summer of
one year.
Ventilatory function tests included the forced
vital capacity (FVC) and 1-sec forced expiratory
volume (FEV). Tests were administered to four
groups of persons living in Tucson. Two of these
groups consisted of elementary-school-age
children. Since no oxidant measurements during
studies of these groups are mentioned in the
report, results of these studies will not be
presented here.
The third group consisted of 1 0 white adults of
both sexes, aged 22 to 32 years, who walked or ran
5 km (3 miles) along major roads in Tucson during
the morning or afternoon of one Saturday in
summer. Members of this group appear to have
been tested before and after exercising.
The fourth group consisted of nine white
adolescents of both sexes, aged 1 6 to 18 years,
who participated in a 32 km (20-mile)walkthrough
major roads in Tucson one Saturday tn spring.
Members of this group were tested before and
after the walk.
Aerometric measurements were obtained from
the Pima County Air Pollution Control District and
the Arizona State Health Department. Stations
used in the studies of adults and adolescents were
at least 5 km (3 miles) away from testing sites.
Among adults who performed the 5-km (3-mile)
walk or run in the morning, no significant
differences between ventilatory function values
before and after exercise were observed During
the morning, the average temperature was 28°C
(83°F), the relative humidity was 43 percent, and
the concentrations of suspended particulate
matter and oxidants were, respectively, 1 04/jg/m3
and 0.01 ppm. Among adults who performed the
walk or run in the afternoon, post-exercise
decreases in ventilatory function of less than 5
percent were observed. These decreases were not
statistically significant, and they disappeared
within 30 mm after exercise. However, they were
evidently larger than changes in ventilatory
function among those who had exercised in the
morning. During the afternoon, the average
temperature was 34°C (94°F), the relative
humidity was 23 percent, and the concentrations
of suspended particulate matter and oxidant were,
225
-------
respectively, 89 fjg/m3 and 0.03 ppm On this day,
the suspended sulfate level was estimated to have
been less than 2.5 /jg/m3.
Among the nine adolescents who completed the
32-km (20-mile) walk in the spring, consistent and
statistically significant post-exercise decreases in
ventilatory function were observed. Average
values of FVC before and after exercise were,
respectively, 3.6 and 3.0 liters. Corresponding
average values of FEV, were 3.0 and 2.7 liters. On
the day of the 32-km (20-mile) walk, the
temperature ranged from 15°C to 31 °C (59°F to
88°F), the average humidity was 30.5 percent, the
average suspended particulate concentration was
133.7 /vg/m3, the average sulfate concentration
was 3.7 fig/m3, and hourly peak oxidant
concentrations increased with time from 0 01 to
0.1 2 ppm. The 32-km (20-mile) walk ended about 2
p.m., which is usually near the hour of daily
maximal oxidant concentration in Los Angeles.
Whether this is also true of Tucson is not stated in
the report.
Since oxidant levels appear to have been higher
during the 32-km (20-mile) adolescent walk than
during the 5-km (3-mile) adult walk or run, and
since statistically significant post-exercise
decreases in lung function were observed in the
adolescents but not in the adults, it is conceivable
that hourly maximum oxidant concentrations
somewhere between 0.03 and 0.12 ppm may
affect the ventilatory function of exercising
humans, if exercise is sufficiently strenuous or
sustained. However, the Lebowitz study does not
provide a rigorous test of this hypothesis, since
adolescents and adults underwent different
exercise regimens and since the aerometric
measurements employed were made at least 5 km
(3 miles) from the study site.
Lebowitz et al27 suggest five hypothetical
mechanisms by which pollution might produce
decreased ventilatory function: (a) an irritant effect
of pollution on the upper airways, (b) bronchial
constriction in-responseto irritation,(c) increase in
flow-resistance in the upper airways and possibly
in the lower airways, (d) decreases in lung
compliance and lung recoil, and (e) transient
edema of the alveoli.
Note that two factors other than oxidant
exposure may have influenced the outcome of the
studies of adolescents and adults. First, fairly
substantial levels of suspended particulate matter
were present along with oxidants on the days
when these studies were conducted. This fact
raises the possibility that particulates alone or in
combination with oxidants may have affected
ventilatory function. Second, the decline in
ventilatory function was greater after a 32-km (20-
mile) walk than after a 5-km (3-mile) walk or run.
The 32-km (20-mile) walk may well have been
more demanding for adolescents than the 5-km (3-
mile) walk or run for adults, even though the adults
were somewhat older. Thusthe results might have
been observed even in the absence of pollution
exposure These two factors considerably restrict
the vigor with which the hypotheses stated in the
previous paragraph can be advanced
EFFECTS OF OXIDANT ON PULMONARY
FUNCTION IN JAPAN
Kagawa and Toyama24 reported a study of the
effects of environmental factors on the pulmonary
function of 21 children, all aged 11 years, at an
elementary school in Toyko, Japan. The study
sample appears to have been quite evenly divided
by sex Pulmonary function tests were
administered once a week for 29 weeks between
June and December 1972. Tests were generally
performed between 1 and 3 p.m. Six physiologic
variables were measured, total airway resistance
(Raw), specific conductance (Gaw/Vtg), maximal
expiratory flow at 50 and 25 percent vital capacity
left to be expired (respectively, Vsoand V25), forced
vital capacity (FVC), and a pulmonary gas
distribution index (GDI) based on the single-breath
nitrogen elimination rate following a full
inspiration of pure oxygen.
Aerometric measurements were made on top of
the children's three-story school. The following
measurement methods were used; For oxidant,
neutral buffered potassium iodide a ndcoulometnc
methods; for ozone (03), ethylene chemilumines-
cence; for hydrocarbons (HC), hydrogen flame
lonization; for nitric oxide (NO) and nitrogen
dioxide (NOa), Saltzman method; for sulfur dioxide
(SO2), conductimetric method; and for suspended
particulate matter (SPM), a light scattering
method. Temperature and relative humidity were
also measured.
During this study, hourly average con-
centrations of oxidant ranged from about 0.03 to
about 0.17 ppm, and hourly average
concentrations of ozone ranged from about0.01 to
about 0.15 ppm. Over this period, the maximum
hourly concentrations of NO, NO2, and SOa were
about 0.08, 0.23, and 0.05 ppm, respectively. The
226
-------
maximum hourly average concentration of
paniculate matter was about 350/ug/m3.
Simple correlation coefficients were computed
between pulmonary function test results over the
29 weeks of study and mean pollutant
concentrations (1) between 1 and 3 p.m. on the day
of testing, (2) between 1 2 noon and 1 p.m. on the
same day, (3) between 1 p.m. on the day before
testing and 1 p.m. on the testing day. Correlations
between temperature and humidity (averaged over
the same periods) and pulmonary function test
results were also computed, A summary of
correlation coefficients significant at or = 0.05 is
presented in Table 10-6. As the table shows,
pulmonary function test results were significantly
correlated with temperature far more oftenthan
with any other environmental factor. Temperature
was also correlated positively and significantly
with ozone, and negatively and significantly with
NO and suspended particulate matter (SPM),
Significant correlations between temperature and
Raw, ^50, ^25, and GDI were consistently positive.
The degree and consistency of correlation of
temperature with VSo and Vas were particularly
striking. Significant correlations of temperature
with Gaw/Vtg and with FVC were consistently
negative.
The degree of correlation between pollutant
concentrations and pulmonary function test
results was generally greater with concentrations
during testing (1 to 3 p.m.) than during either of the
two previous periods. All significant correlations
between pollutant concentrations and test results
bore the sign that would be expected if pollution
TABLE 10-6. NUMBER OF SUBJECTS FOR WHOM CORRELATION COEFFICIENTS BETWEEN ENVIRONMENTAL
AND RESPIRATORY FUNCTION MEASUREMENTS ARE SIGNIFICANT AT P < 0.05, JAPAN, 1372*2"
Physiologic variables
Environmental
variables
Qxidant
Oa
Hydro-
carbon
NO
NO2
SO2
SPM
Tempera-
lure
Relative
humidity
D°
1d
24"
D
1
24
D
1
24
D
1
24
D
1
24
D
1
24
D
1
24
D
1
24
D
1
24
Raw
1(0 53f
1(049)
2{0 65-0 63)
5(0 80-0 45)
2(0 66-0 63)
3(0.62—045)
HO 59}
2(0 53-0 48)
0
0
0
1 (0 69)
0
0
0
2f047)
1(054)
0
0
0
1(058)
12(072-046)
13(072-044)
14(073-048)
0
0
2(0 53-0 45)
Gaw/
0
1 (-0 56)
0
5(-0 66-
2(~0 68-
3(-0 59-
Vtg
-0 48)
-0 52)
-0 45)
2(-0 51—044)
K-057)
0
0
0
1(-Q70)
0
0
0
1(-0 50)
2(-0 53-
0
0
0
0
9(-0 76-
10(-0 75-
9(-0 79~
0
0
3(-0 54-
V max
a?
50% FVC
0
0
K-071)
1(-0 62)
0
0
1 (-0 50)
2(~0 74-
0
-056)
8I-Q.56— 047)
4(-0.56~
4(-0 50-
-049)
-044)
2(-0 51--044)
3(-0 56--0 44)
-045)
-052)
-044)
-045)
-046)
0
0
K-047)
0
K-045)
0
0
19(0 82-0
1 9(0 83-0
1 9{0 90-0
0
1(046)
1 5(0 68-0
56)
55)
58)
44)
V max
at
25% FVC
0
0
1 (-0 60)
K-066)
0
0
0
1{-059)
0
101-060—047)
7(-071~-045)
7( --064 —047)
K-058)
2(-068 -047)
0
K-076)
K-075)
0
1(-O44)
1( O46)
0
17(0 81-046)
1 3(0 81 -0 44)
19(0 82-058)
1 (0 50)
1(055)
12(0 74-046)
0
0
0
2|0
0
0
4(0
1(0
0
0
0
0
0
0
0
6(0
1(0
0
0
0
0
3(0
4(0
4(0
0
Gas
distribution
index
56-0 48)
76-0 56)
55)
55-0 47)
52)
56-045)
58-0 44)
61-046)
1(044)
3(0
55-048)
FVC
K-045)
K-044)
0
3{~0 69-
4(-0 62-
2(-0 44)
4(-0 63-
-048)
046)
-046)
5(-0 70--0 44)
0
0
0
0
0
0
0
0
0
3(-0 59-
0
0
0
5(-0 75-
-044)
-047)
6( -077 —047)
5< -075 —055)
0
0
4(-0 74--0 51 )
aTotai oJ 21 children
^D, During measyrernonf of respiratory funcuan
cNumbers >.n paientheses are the range of their correlation coefficients
"One hour before measurement of respirarruy funefion
*?4. Average af hourSy values during 24 hf before measurement of respiratory function
227
-------
indeed exerted a deleterious effect on pulmonary
function. The pollutants most frequently correlated
significantly with test results were (in descending
order) NO, ozone, HC, and S02. A point of interest is
that oxidant and N02 were only infrequently
correlated with test results. In about 25 percent of
subjects, the ozone concentration during testing
was significantly correlated with Raw, Gaw/Vtg,
and FVC, Concentrations of S02 were not
significantly correlated with pulmonary function
parameters as frequently as were concentrations
of ozone. However, SOz concentrations tended to
correlate significantly with the same parameters
as ozone Nitric oxide and NO?, unlike ozone and
S02, tended to correlate significantly with Vso and
V2s. On the basis of these findings, the authors
suggested that ozone and SO? may exert effects
primarily in the upper airways, while the nitrogen
oxides may exert effects primarily in the lower
airways.
Kagawa and Toyama24 also presented the
individual significant correlations between
pulmonary and pollution concentrations between
1 and 3 p.m. on the testing day. Such correlations
were unusually frequent in three of the 21
students.
In further analysis, the authors statistically
corrected for the effects of temperature by
computing partial correlations between pollution
concentrations (presumably between 1 andSp.m.)
and pulmonary function test results. This analysis
greatly reduced the number of significant"
correlations between pollution and pulmonary
function. However, significant partial correlations
of ozone concentration with Raw were observed in
three subjects. Significant partial correlations of
ozone with Gaw/Vtg and with V5Q or V25 were
observed in one subject each. The pattern of
significant partial correlation between S02 and
pulmonary function remained very similar to that
between ozone and pulmonary function. After the
partial correlation analyses, only one significant
correlation of nitrogen oxides with V5o or V25
remained.
Kagawa and his associates continued to collect
pulmonary function data from the students
described above. Results obtained from November
1972 to October 1973 have been presented by
Kagawa et al in a separate report.25 In Kagawa's
original study, temperature was observed to exert
a consistent effect on pulmonary function To deal
with this effect, the investigators divided the study
into a low-temperature season lasting from
November 1972 to March 1973, and a high-
temperature season lasting from April to October
1973. Data were collected on 19 days in the low-
temperature season and on 30 days in the high-
temperature season. Data from 19 students (10
males and nine females) were analyzed.
As in the first study, simple correlations between
levels of environmental factors and pulmonary
function test results were computed. The period
over which levels of environmental factors were
averaged is not clearly stated in the second study.
However, we believe that these levels were
averaged over the 2 hr between 1 2 noon and 2p.m.
on the day of pulmonary function testing. In the
analysis of the second study, the investigators
considered the environmental and pulmonary
function variables that had yielded an appreciable
number of significant correlations in the first
study. The environmental variables considered
were temperature, ozone, NO, NOa, [NO + N02],
SOg, and SPM. Oxidant was not considered. The
pulmonary function variables considered were
Raw, Gaw/Vtg, Vso, and VZ5.
During this study, hourly averaged ozone
concentrations ranged from about 0.01 to about
0.30 ppm. The maximum hourly average
concentrations of NO, NOs, and S02 were about
0.18, 0.30, and 0.16 pprn, respectively. The
maximum hourly average concentration of
suspended particulate matter was about 450
In the high- and low-temperature season-
specific analyses, temperature was consistently
positively correlated with Raw, Vso and VZ5, and
negatively correlated with Gaw/Vtg. However,
when data for one full year were analyzed,
temperature was negatively associated with Raw
in 17 of 19 subjects. Thus it appeared that the
effect of temperature on Raw might be quite
heavily dependent of the selection of the study
period.
The results of this second study were somewhat
similar to those of the first with respect to ozone.
Concentrations of ozone were generally positively
correlated with Raw and negatively correlated with
Gaw/Vtg m both the high- and low-temperature
seasons. Interestingly, this was more consistently
true during the low-temperature season than
during the high-temperature season, when
measured ozone concentrations were highest
During the low-temperature season, 8 of the 19
ozone-Raw correlations and 7 of 1 9 ozone-
Gaw/Vtg correlations were significant at or= 0.05.
228
-------
When partial correlation analysis, which
statistically removed the effects of temperature,
was performed, five ozone-Raw correlation
coefficients retained statistical significance. Thus
it appeared that the apparent effect of ozone on
Raw during the low-temperature season would not
have been due to temperature alone. In the high-
temperature season, only 1 of 38 correlations of
ozone with Raw or Gaw/Vtg was significant.
Simple correlations suggested little detrimental
effect of ozone on V5o or V25 in either season.
However, after a partial correlation analysis that
incorporated temperature, the number of
significant negative correlations between ozone
and V50 increased from one to five. This result is
difficult to square with results of Kagawa's first
study,24 in which correction for temperature
considerably reduced the number of significant
correlations between pollution and pulmonary
function.
As in Kagawa's first study, concentrations of NO
were significantly correlated (negatively) with V5o
and V2s considerably more consistently than with
Raw or Gaw/Vtg. However, concentrations of N02
were significantly correlated (positively) with Raw
and (negatively) with Gaw/Vtg more consistently
than with -V50 or V25. Significant correlations of
N02 with Raw and Gaw/Vtg occurred only during
the high-temperature season. Few significant
correlations between SCb concentrations and
pulmonary function parameters were observed.
During the low-temperature season, con-
centrations of SPM were generally correlated
negatively with Raw and positively with Gaw/Vtg.
During the high-temperature season, the direction
of these correlations was generally reversed. In
both seasons, SPM concentrations were generally
negatively correlated with both Vsoand Vas. These
results suggest quite strongly that the calculated
effects of pollution on pulmonary function, as with
temperature, may depend quite heavily on the
selection of study period.
Among the 19 students in Kagawa's second
study, 5 showed significant correlations both
between pulmonary function and at least three
environmental factors, and between
environmental factors and measures of airway
resistance (Raw or Gaw/Vtg) as well as maximal
flow ( Vso or V25). Thus the second study, like the
first, suggests than some segments of the healthy
population may be more susceptible to pollution
exposure than others.
Three interesting considerations emerge from
the results of Kagawa's studies. First, the results
disclose an association between ozone (but not
oxidant) and decrements in pulmonary function.
Second, even within the healthy population,
certain groups may be more sensitive than others
to the effects of air pollution exposure. The results
of several studies already described in this chapter
are consistent with the same hypothesis. One
might hypothesize further that such sensitive
groups may be at unusually high risk of developing
chronic illness after years or decades of elevated
pollution exposure. No studies have as yet
specifically addressed this hypothesis. Third,
statistical techniques currently available do not
appear fully equal to the task of separating the
health effects of pollutants from temperature or
from each other. Available statistical techniques
tend to focus the attention on individual
environmental factors as separate entities. In
reality, however, the effects of an individual
pollutant may be dependent both on the presence
of other pollutants and combinations of climatic
factors. Statistical separation of the effects of
individual components of existing environmental
conditions may therefore yield a misleading
description of the true situation. It is hoped that
future advances in statistical methodology will
enable investigators to characterize interactions of
environmental factors more completely than is
possible at present.
DISCUSSION OF THE EFFECTS OF OXIDANTS ON
PERFORMANCE
The studies reported by Wayne et al.65 and by
Dry61 suggest that photochemical oxidant pollution
can impair performance of tasks as different as
competitive running and automobile operation. As
mentioned in the Wayne eta I. study, the proportion
of runners failing to improve their times increased
quite consistently (r = 0.945) with increasingly
hourly oxidant exposures over a range of 60 to 590
/jQ/m3 (0.03 to 0.30 ppm). However, within
individual portions of the oxidant concentration
range, unequivocal linear relationship between
oxidant exposure and performance could not be
discerned.
Findings of Folinsbee's experimental studies of
ozone exposures14'15 suggest that on high-oxidant
days, air pollutants may have restricted the
mechanical ventilatory function of the runners
observed by Wayne et al. to the point where these
229
-------
runnerc were no longer able to take up enough
oxygen to support optimum performance at
maximum exertion. Folinsbee's findings further
suggest that ozone alone may have accounted for
the results observed by Wayne et al. This
suggestion remains open to question, since
Folinsbee's exposure concentration of ozone, 0.75
ppm, was considerably higher than most oxidant
concentrations observed in the Wayne et al. study.
Studies of ambient oxidant exposure and
pulmonary function have yielded mixed results.
McMillan's study34 showed no association
between either short-term or long-term oxidant
exposure and impairment of pulmonary function in
children. However, this study is difficult to
interpret, since pulmonary function tests were
performed at rather wide intervals (twice per
month) and since the children in the high- and low-
pollution communities were ethnically dissimilar.
The findings of Lebowitz27 and Kagawa et al.24'25
in contrast to those of McMillan,34 suggest that
oxidant pollution may contribute to decrements in
pulmonary function. The results of Lebowitz, like
those of Folinsbee, are consistent with the
hypothesis that the degree to which oxidants affect
pulmonary function is positively related to the level
of exercise undergone by the subjects. Clearly,
though, the Lebowitz study must be interpreted
cautiously, since its design did not allow for
separation of the effects of environmental and
meteorologic factors.
The results of Kagawa et al,24'25 suggest that
short-term exposure to oxidant air pollution may
affect the large airways more than the small ones.
A point of interest is that Kagawa observed
stronger correlations of pulmonary function with
ozone than with oxidant. Conceivably, the
mechanism underlying Kagawa's observations is
similar to that underlying Folinsbee's. In any case,
the results of Kagawa et al., like those of Wayne et
al., enhance confidence that ambient ozone alone
exerts effects on humans, whatever the effects of
other photochemical substances may be.
Whether transient decrements in pulmonary
function constitute a clean-cut hazard to health is
not yet known. Recent evidence suggests that
childhood infection, often accompanied by
impairment of pulmonary function, may
predispose subjects to chronic respiratory disease
later in life.28 However, the degree to which
pollution-associated impairment promotes such
an outcome has not yet been determined.
Eye Irritation In Relation to Variations in Oxidant
Levels
PANEL STUDIES
No symptom has been associated with ambient
photochemical pollution more frequently or
consistently than eye irritation. In an effort to
determine the types and concentrations of
pollutants responsible for eye irritation,
investigators have studied a variety of individuals
m the Los Angeles area. Studies conducted m
1954, 1955, and 1956, reported by Renzetti and
Gobran46 of the Air Pollution Foundation, San
Marino, California, were among the first studies in
this field.
The first part of the Renzetti and Gobran study
was conducted from August through November
1954.46 In this study, several panels of observers
were asked to report eye irritation on Tuesdays and
Fridays. Later, panelists were asked to report only
on those days for which eye-irritating levels of
pollution had been predicted. This latter study
design may have introduced bias into observed
results. In general, the observers were office and
factory workers. One of the panels consisted of a
group of scientists of the California Institute of
Technology. The eye-irritation data were
compared with instantaneous values of oxidant
concentrations as measured by potassium iodide
recorders.
Data from the 1954 study46 are summarized in
Figure 10-7 and in Tables 10-7 and 10-8. These
data strongly suggest that the degree of eye
irritation increases as the oxidant level increases
over a range from nearly zero to about 0.45 ppm.
From Figure 10-7, it is impossible to discern a
discrete threshold oxidant concentration below
which no eye irritation occurs. In Table 10-8, data
from a panel of scientists are summarized. The
table suggests that oxidant was not the only
environmental variable statistically associated
with eye irritation in Los Angeles in 1954,
although oxidant explained a higher proportion of
the variation in eye irritation than any of the other
variables measured. Data from the panel of
scientists yielded the greatest number of
significant correlations between eye irritation and
environmental variables other than oxidant.
The second part of the study46 was conducted
from August through November 1955. In Figure
10-8, a regression line relating maximum eye
230
-------
irritation in a panel of scientists to maximum
oxidant concentration is presented. The
relationship between eye irritation and oxidant
concentration is qualitatively similar to that
observed in 1954 in that eye irritation increased as
the oxidant level increased, and in that a discrete
oxidant threshold concentration could not be
clearly discerned.
From the data provided by the Air Pollution
Foundation studies,46 linear mathematical
relationships between maximum oxidant values
and mean maximum eye-irritation values were
derived. The data from these studiesdemonstrated
increasing eye irritation with increasing
concentrations of oxidant pollution over the range
of instantaneous values from 100 to 880 ^g/m3
(0.05 to 0.45 ppm), although no clear threshold
level for this effect was apparent (Figure 10-7).
Other studies on eye irritation have been
performed, including one in which a panel of
employees of the Los Angeles Air Pollution Control
District was queried during the period 1955-58,17
A group of environmental sanitation workers in the
San Francisco Bay Area was also studied during
this same period. These panels demonstrated a
tendency to experience increasing occurrence of
eye irritation with increasing oxidant levels. As in
all such studies, there were some individuals who
reported eye irritation even when there was no
oxidant present.
TABLE 10-7. CORRELATION OF EYE IRRITATION WITH
SIMULTANEOUS OXIDANT CONCENTRATIONS,
IN ORDER OF DECREASING EYE IRRITATION SCORE,
FOR A NUMBER OF STATIONS IN THE
LOS ANGELES AREA, 1954"
Average
eye
irnta-
tion
score
26.2
220
21 9
21 3
182
13.0
188
Average
oxidant
concen-
tration,
ppm
013
0.10
0.21
0.11
0.15
0.17
0 14
Variation in eye
irritation score
explained by
ojadaru concen-
tration (r'(
0.88
0.68
0.76
006
056
0.65
0.18
Station
5
8
4La
2
3
4E°
11
Number
of daily
observa-
tions
25
29
24
30
67
66
344
aL - panel of laymen m Pasadena, California. E-pane! of scientists atthe California
Institute of Technology in Pasadena
A large study of student nurses, conducted by
Hammer eta I.,18 has been described previously. As
mentioned, rates of eye irritation in this study
increased with daily maximum hourly oxidant
levels in and above the range of 294 to 372 /jg/m3
(0.15 to 0.19 ppm),
EVALUATION OF FILTERS FOR REMOVING IRRI-
TANTS FROM POLLUTED AIR
A study was conducted by Richardson and
Middleton48 49 to evaluate the sensory
effectiveness of air filter media for removing eye
irritants from polluted air. Eye irritation in two
F
z
CD
DC
40
30
LU 20
EC
o
z
O
IE
tc
10
• — • — STATION 3
STATION 4Ea
STATION 4La
STATIONS
STATIONS
ALL STATIONS
10 20 30 40
OXIDANT CONCENTRATIONS, pphm
50
"E"=panel of scientists at the California Institute of Technology.
L=panel of laymen in Pasadena, California. For both "E" and "U" panels,
aerometric measurements from station 4 were used.
Figure 1 0-7. Regression curves relating eye irritation scores and simultaneous oxidant concentrations from a number
of stations in the Los Angelas area, 1954.46
231
-------
groups of 20 female telephone company em-
ployees, similar with respect to age and job
characteristics and employed in identical adjacent
rooms, was evaluated over 123 workdays from
May to November 1 956. Active and dummy filter
units were switched periodically between the two
rooms so that the groups were alternately exposed
to test and control conditions. The sensory
response of the subjects was measured daily at 11
a.m. by means of a questionnaire; simultaneous
measurements of oxidants, particulate matter, and
nitrogen dioxide were obtained within each of the
two rooms and immediately outside the building.
The differences in eye irritation between the
activated-carbon filtered and nonfiltered test
situations were in all cases highly significant
(Table 10-9). A statistically significant correlation
between eye irritation and oxidant concentrations
occurred in the nonfiltered room (Table 10-10). The
scatter diagram of results (Figure 10-9) suggests
that the severity of eye irritation begins to increase
above a simultaneous oxidant concentration of
about 200/vg/m3 (0.10ppm) as measured by the Kl
method.
Nitrogen dioxide concentrations were reduced
by the activated carbon filters during their early
use but, after a period of time, nitrogen dioxide
concentrations in the filtered atmosphere in-
creased. No significant correlations between eye
irritation and nitrogen dioxide levels were
observed, nor were significant correlations found
between eye irritation and concentrations of
particulate matter.
TABLE 10-8. CORRELATION BETWEEN EYE IRRITATION
AND SIMULTANEOUS ENVIRONMENTAL
MEASUREMENTS, AS JUDGED BY A PANEL OF
SCIENTISTS. 195446
Environ-
mental
variables
Oxidant
NO,
CO
Hydro-
carbons
Visibility
Parti-
•culates
Alde-
hydes
Variance in eye
irritation score
explained by
environmental
variable (r2)
065
007
0.53
039
017
053
048
Average value
of variable
0 17 ppm
0 20 ppm
0 27 ppm
0 17 ppm
1 2 miles
21 1 Coh
units
0 19 ppm
Average eye
irritation
score
1 30
13 1
142
140
13.3
137
140
Number of
observations
66
51
47
53
56
26
18
PHOTOCHEMICAL OXIDANT AND EYE
IRRITATION IN LOCATIONS OTHER THAN
CALIFORNIA
Oxidant measurements at levels possibly
associated with eye irritation have been reported
from a number of cities other than Los Angeles.
Circumstantial evidence of increased eye irritation
has been reported in Washington, D.C., Denver,
New York City, and St. Louis. An epidemiologic
study of eye irritation and other health indiceswas
carried out by Cassell et al.9 on a population living
on the lower East Side in Manhattan. In this study,
families reported the presence or absence of
symptoms, including eye irritation, each week. In
October 1963 (a period of increased pollution) the
40
ir
°= 5 20
u <
> z
" uio
DC
o
I
I
I
I
I
I
I
I
10
80
90
20 30 40 50 60 70
MAXIMUM OXIDANT CONCENTRATION, pphm
Figure 10-8. Variation of mean maximum eye irritation, as judged by a panel of scientists, with maximum oxidant
concentrations, Pasadena, August-November 1955.
232
-------
frequency of new reports of eye irritation increased
from about 2 to nearly 5 percent of the population.
However, Cassell's study design did not allow the
effects of individual pollutants to be separated
from each other.
In the Makino and Mizoguchi study of Japanese
students, cited above,32 the daily frequency of eye
irritation regressed on corresponding daily
(presumably maximum hourly) oxidant con-
centration for each of the 12 months of study.
Regression lines that yielded significant
associations between irritation rate are presented
"in Figure 10-10 As the figure shows, the rate of
eye irritation rose most quickly with increasing
oxidant levels during July. The data for that month
predicted that 10 percent of students would
complain of eye irritation when the oxidant level
was about 0.18 ppm, and that 15 percent of
students would similarly complain when the level
was about 0.26 ppm. On the other hand, data for
December predicted that only about 3.5 percent of
students would complain of eye irritation when the
oxidant level was about 0.26 ppm. Statistically
significant associations between oxidant
concentration and eye irritation rates occurred
more frequently in summer and autumn than in
winter and spring.
Shimizu et al.54 reported a combined
epidemiologic, clinical, and toxicologic study of the
effects of photochemical smog on the eyes of
humans and rabbits. In the epidemiologic portion
of the study, all 515 students at a junior high
school in Tokyo were requested to complete a
symptom questionnaire each day over the 61 -day
period between May 20 and July 19, 1974. The
questionnaire inquired about the presence of 17
symptoms, including eye irritation, eye pain,
hyperemia of the eye, lacrimation, blurred vision,
and a variety of respiratory and constitutional
symptoms.
TABLE 10-9. EFFECT OF FILTER ON SENSORY IRRITATION AND CHEMICAL MEASUREMENTS49
Test condition
Eye
irritation
index3
Oxidants,
pphmb
N02,
pphm
0.032c-Activated carbon filter
Mean, nonfiltered room
Mean, filtered room
Difference between means
Probability that the difference could have
occurred by chance
0016 Activated carbon filter
Mean, nonfiltered room
Mean, filtered room
Difference between means
Probability that the difference could have
occurred by chance
0 075 Activated carbon filter
Mean, nonfiltered room
Mean, filtered room
Difference between means
Probability that the difference could have
occurred by chance
0 0030 Activated carbon filter
Mean, nonfiltered room
Mean, filtered room
Difference between means
Probability that the difference could have
occurred by chance
Paniculate filter
Mean, nonfiltered room
Mean, filtered room
Difference between means
Probability that the difference could have
occurred by chance
1 99
1 01
098
<001
1.95
1 41
1 54
<001
545
2.35
3 1O
<005
235
1 19
1 16
<001
2.13
1 91
0.22
9.8
049
94
«001
84
1 8
6.7
«0.01
13.9
4.9
90
<001
73
36
37
<0.01
5.7
34
2.3
<001
1 5
041
1 1
«001
34
1 6
1 8
«0.01
2 7
57
3 0
47
49
02
63
5.5
03
<0.02
"An eye irritation index of 3 corresponds to barely noticeable irritation, an index of 7 corresponds to moderate irritation, an index of 11 corresponds to severe irritation
"Measured by the Kl method
cRefers to detention time in seconds
dDifference not significant
233
-------
During the 61 days of the epidemiologic study,
the weather appears to have been humid and
rather cool, with maximum temperatures rarely
exceeding 30°C. The oxidant level exceeded 0.1 5
ppm on only 5 days, and the highest oxidant level
recorded during the study was 0 23 ppm. Though
averaging times were not specified in the report,
these levels appear to have been maximum
instantaneous levels. Levels of SO2 as well as
oxidant were presented in the report. The methods
used to measure oxidant and SO2 concentrations
were not mentioned.
Symptom rates averaged over the 5 days when
oxidant concentrations exceeded 0.15 ppm were
compared to corresponding rates averaged over all
days of the study. The percentages of students
reporting eye symptoms on the high-oxidant days
and on all study days, respectively, were as
follows: eye irritation, 8.0 and 2.4 percent; eye
pain, 3 3 and 1.9 percent; blurred vision, 1.4 and
1.1 percent; lacrimation, 1.4 and 0.5 percent; and
hyperemia of the eyes, 0.6 and 0.5 percent,
Differences between high-oxidant days and all
study days in eye irritation and lacrimation were
statistically significant at a = 0.05. Incidentally, a
significantly higher mean percentage of students
also reported sore throat (2,6 versus 1.1 percent)
and dyspnea (2,5 versus 1.9 percent) on the high-
oxidant days than on all study days These findings
are quite consistent with the findings of Makmo
and Mizoguchi32 and Mizoguchi et al.37
TABLE 10-10. PEARSON PRODUCT MOMENT CORRELATION OF COEFFICIENTS BETWEEN EYE IRRITATION AND
ENVIRONMENTAL FACTORS IN A NONFILTERED ROOM49
Item
Irritation
Irritation
Irritation
Irritation
Irritation
Irritation
Temperature
Temperature
Oxidants
concentration,
by phenolphtalem method
NOa concentrations
Correlation
vs
vs
vs
vs
vs
vs
vs
vs
vs
vs
Oxidants concentration, by phenoiphtalem method
Oxidants concentrations, by Kl method
NOz concentrations
Paniculate
Temperature
Relative humidity
Relative humidity
Oxidanis concentration, by Kl method
Oxidants concentration, by Kl method
Oxidants concentration, by Kl method
081
031
005
0 15
049
-024
-038
029
088
-0.15
18
16
t 12
o
X
111
o
z
LLI
S
"MODERATE" IRRITATION
. "BARELY NOTICEABLE" IRRITATION
35
40
05 10 15 20 25 30
OXIDANT CONCENTRATION, pphm
Figure 10-9. Mean index of eye irritation versus oxidant concentration,49
234
-------
Data presented in the Shimizu et al. report54
suggest that the highest concentrations of both
oxidant and SOa tended to occur on the same days.
Concentrations of S02 and oxidant appear to have
been of generally similar magnitudes. Thus we do
not believe it is justifiable to attribute solely to
oxidant the differences between high- and low-
oxidant days in symptom rates observed in this
study. It is conceivable that the presence of SOa or
other sulfur oxide compounds may have been a
necessary factor in the production of these
symptoms.
In the clinical portion of the Shimizu et al.
study,54 the eyes on 41 seventh graders and 41
eighth graders were examined on three sets of 2
consecutive days—one set in May, one in July, and
one in October 1974. On the first day of each set,
either the seventh or eighth graders were tested.
On the next day, the remaining grade was tested.
The second day of the 2-day set in July happened to
be one of the epidemiologic study days on which
the oxidant concentration exceeded 0.15 ppm. On
each of the 6 days of the clinical study, subjects
were given a physical examination of the eye, the
volume and pH of tear fluid were measured, and
the amount and activity of tear lysozyme were
determined. On the high-oxidant day in July, four
(10.0 percent) of 40 students reported symptoms of
eye irritation. (This figure is consistent with the 8.0
percent of all students reporting eye irritation on
days with oxidant levels exceeding 0.15 ppm.) On
the other 5 days of clinical examinations, an
average of 1.2 students, or about 3 percent of
students examined, reported such symptoms. On
the second day of the July clinical examinations,
the mean pH of the eighth-graders' tear fluid was
significantly lower (about 0.1 pH unit lower) than it
had been in the seventh graders tested the day
before. On the other two sets of days, when
between-days differences in oxidant con-
centrations were smaller than in July, no
significant between-days difference in lacrimal pH
was observed. We believe that this finding, though
suggestive, must be interpreted cautiously,
because different subjects were tested on the 2
days of each set of days. The authors observed no
significant pollution-related differences in lacrimal
volume or in the volume or activity of tear lysozyme
secreted. However, a slight shift in the pattern of
lysozyme activity was noted on the high-oxidant
day. The significance of this shift is as yet
uncertain.
The toxicologic portion of the Shimizu et al.
study54 had two facets. In the first, the eyes of four
men and two women aged 20 to 40 years were
exposed to 0.5 percent (5000 ppm) acrolein gas. In
most subjects, smarting of the eyes developed
within 3 min. In others, however, no unusual
sensation was noted.
30
20
a
x
o
10
SIGNIFICANT AT:
• •• 0.1%
• • 1 %
• 5 %
*The regression lines for September and October
are nearly identical.
5 10 15
SYMPTOM AND COMPLAINT RATE, percent
20
Figure 10-10. Monthly regression lines between oxidant levels and eye irritation (symptom no. 1).32
235
-------
In the second facet of the Shimizu et al.
toxicologic study,54 six rabbits were exposed to 10
ppm acrolein gas for up to 2 hr. One eye of each of
two rabbits was kept closed throughout the
exposure. After 1 hr of exposure, 3 of 10 eyes had
developed spotty, fluorescent stainable, white
corneal lesions. After 2 hr of exposure, this type of
corneal lesion covered approximately the entire
area of the pupil in all the experimental eyes. The
authors did not describe the reversibility of these
lesions. This type of lesion did not develop in the
two closed eyes.
DISCUSSION OF PHOTOCHEMICAL OXIDANTS
AND EYE IRRITATION
In the Los Angeles area and in Japan, strong and
consistent associations between photochemical
oxidant exposures and both the frequency and
severity of eye irritation have been observed. In
most studies performed to date, the rate of subjects
reporting eye irritation has been higher at oxidant
concentrations above 0.15 ppm than at lower
concentrations. Hammer observed the frequency
of eye irritation to increase at maximum hourly
oxidant concentrations in the range of 294 to 372
fjg/m3 (0.15 to 0.19 ppm). Recent Japanese
studies, especially thatof Makmoand Mizoguchi,32
suggest that eye irritation may begin to occur even
at (presumably maximum hourly) oxidant
concentrations of 0.10 ppm or below.
As yet, the components of photochemical
pollution specifically responsible for eye irritation
have not been identified in epidemiologic studies.
Such identification must await refinements in
available aerometric technology. Experimental
studies indicate that ozone, which is the principal
contributor to total oxidant concentration, is not
itself an eye irritant at ambient levels. However,
the possibility remains that ozone in the ambient
atmosphere may interact with other substances
(which need not necessarily be a part of the
photochemical pollutant complex) to produce eye
irritation.
Constituents of the photochemical oxidant
complex that have been shown to be eye irritants
include peroxyacetylnitrate, peroxybenzoylmtrate,
acrolein, and formaldehyde. The broad range of
actual and potential eye irritants in photochemical
smog render it possible, even likely, that different
pollutants or blends of pollutants account for eye
irritation in different places, or in the same place at
different times. This hypothesis can be tested only
by further epidemiologic study of pollution-related
eye irritation in a variety of locations.
The strong and consistent association observed
between short-term photochemical oxidant
exposure and eye irritation arouses interest as to
what the effects of such long-term exposures on
ocular structure and function might be. This
concern is heightened by the demonstration by
Shimizu et al.54 of opaque corneal lesions in rabbits
resulting from as little as 1 hr of exposure to
acrolein. Little epidemiologic effort has as yet been
devoted to detecting the ocular effects of chronic
ambient oxidant exposures. It is quite conceivable
that such effects have not yet become apparent,
since no city in the United States has experienced
consistently elevated oxidant concentrations for
more than 30 years or so. Overall conclusions as to
the impact of photochemical oxidant exposures on
the human eye must await a thorough, systematic
investigation of chronic exposures.
EFFECTS OF CHRONIC PHOTOCHEMICAL
OXIDANT EXPOSURES
Introduction
Knowledge of the effects of long-term photo-
chemical oxidant exposure is still distressingly
limited. In most studies to date that have assessed
such effects, health data have been collected at
one point in time in areas whose long-term
pollution exposures are expected to have differed.
Studies of this design are termed cross-sectional.
In carefully planned cross-sectional studies, study
populations are selected to be as similar as
possible with respect to all characteristics except
pollution exposure, lest nonpollution variables
interfere with comparisons made among
populations. In environmental epidemiology,
nonpollution variables of particular importance
include cigarette smoking habits, occupational
exposure to respiratory irritants, socioeconomic
status, and climate.
For two reasons, cross-sectional studies of the
effects of long-term pollution exposures must be
interpreted cautiously. First, it is virtually never
possible to study populations similar in all
characteristics except their pollution exposure.
Indeed, there is little question that not all the
variables complicating comparisons of populations
with different pollution exposures have as yet even
been identified, let alone controlled infield studies.
Second, even when health effects conform to
pollution exposure gradients, the investigator,
since he has made health measurements at only
236
-------
one point in time, is usually unsure as to the
duration or concentration of Exposure that may
have been necessary to produce the observed
effects.
Existing studies must be replicated, or as nearly
replicated as possible, before the effects of chronic
oxidant exposures can be confidently described.
Full characterization of such effects will require
carefully designed and patiently conducted
prospective studies. As will be seen, these
conditions have not yet been fulfilled.
Mortality in Areas of High- and Low-Oxidant
Pollution
LUNG CANCER MORTALITY
It is known that active organic carcinogens are
found in polluted atmospheres51 and that ozone, in
concentrations above those found in ambient air,
has radiomimetic properties,13 However it is not
yet established whether ambient concentrations of
organic carcinogens or ozone are sufficient to
promote the development of cancer in humans.
Buell et al.6 reported a prospective study of lung
cancer among 69,160 members of the California
Division of the American Legion. This number
represented about 50 percent of all eligible
individuals. Since military personnel must meet
certain standards of health, the American Legion
may under-represent the population with respect
to the prevalence of chronic disease. However, it
may over-represent the population with respect to
cigarette smoking. The cooperating subjects
reported by postal questionnaire their residence,
occupation, and smoking histories. Identifying data
for each individual were maintained on a roster
against which the death certificates were checked
for the 5-year period 1958-62. It thus became
possible to carry out a reasonably economical
longitudinal study comparing the major
metropolitan areas of California with respect to
mortality due to lung cancer and other conditions.
A total of 336,571 man-years of observation was
included in the report.
As shown in Table 10-11, long-term residents of
Los Angeles County had slightly lower age- and
smoking-adjusted lung cancer ratesthan residents
of the San Francisco Bay Area counties and San
Diego County. These urban groups had higher
rates than the population residing in the rest of the
state. The relative risk of lung cancer for heavy
smokers (more than one pack a day) was greater in
Los Angeles County than in other areas of
California (Table 10-1 2). For nonsmokers, the rate
in the two metropolitan groups was substantially
greater than in the rest of the state. The San
Francisco/San Diego rates, however, were higher
than those in Los Angeles.
For three reasons, the Buell et al. study must be
considered inconclusive. First, the rate of
questionnaire return, though impossible to
determine precisely, probably did not exceed 60
percent. Thus it is quite conceivable that un-
detected selective factors may have influenced the
published results. Second, if chronic oxidant
exposures produce lung cancer, the latent period
between the first exposure and tumor
development may have exceeded the maximum
observation period possible in this study, This
maximum period was only about 15 years,
between the late 1940's, when Los Angeles first
experienced elevated photochemical pollution
levels, and 1962, Third, Buell et al. did not relate
mortality rates to actual pollution measurements.
CHRONIC RESPIRATORY DISEASE MORTALITY
in
Mahoney reported a preliminary study
which respiratory disease mortality rates in Los
TABLE 10-11. TOTAL LUNG CANCER MORTALITY IN AN AMERICAN LEGION STUDY POPULATION,
CALIFORNIA, 1958-62*
San Francisco
Item
Age-adjustedb
Age- and smokmg-
adjusted
Residency.'
At least 10 years
Less than 10 years
Unknown
Los Angeles
County
Mortality
rate8
959
954
96.6
76.7
1234
Bay Area
and
San Diego Counties
Totai
deaths
n/a
n/a
79
27
12
Mortality
rate8
1045
102.0
1063
69.1
215.3
Tola!
deaths
n/a
n/a
58
13
10
All
Calif
Mortality
rate'
75 3
75 5
199
685
652
other
counties
Total
deaths
n/a
n/a
69
30
6
"Deaths per 100,000 man-years
"Age-adjusted by the direct method to the tota! study population
cAge- and smokmg-adjusled
237
-------
Angeles during 1961 were related to wind-flow
patterns in the city. In the Los Angeles Basin, the
prevailing wind blows in a southwesterly direction
from the Pacific Ocean. Mahoney constructed five
concentric zones within the Basin (Zones 1
through 5), each about 10 km (6 miles) wide and
each inland and downwind from the last. He then
computed respiratory death rates separately for
each zone. Death rates were computed by dividing
the number of deaths in 1 961 by the population in
the zone as determined by the 1 960 census. Zonal
populations ranged from 36,312 in Zone 5, the
most downwind and inland zone, to 574,512 in
Zone 3. All death rates were adjusted, by the
indirect method, for age, sex, and income level. All
persons in each zone were assumed to have the
median income for that zone. Only whites were
considered in the analysis. Variables such as
smoking, migration within the city, and variation
among zones in population density were not
considered.
Adjusted respiratory death rates per 100,000 in
Zones 1 through 5, respectively, were as follows:
53, 51, 58, 66, and 111. This finding is consistent
with the hypothesis that photochemical air
pollution influences mortality rates, since pollution
levels are generally higher in leeward areas of Los
Angeles than in windward areas. However, as the
author mentions, other factors such as
temperature and humidity are also intimately
associated with wind-flow pattern. The relative
contribution to mortality rates of these factors,
pollution, and demographic and behavioral
variables not considered could not be determined
from Mahoney's report. Also, no pollution
measurements were presented to document the
degree to which pollution levels increased with
increasing distance from the ocean. Thus as the
author himself observed, this report must be
considered inconclusive, suggestive as it is of an
effect of pollution on mortality.
Rates of mortality resulting from chronic res-
piratory diseases other than lung cancer were
briefly considered in the Buell et al. study.5 These
rates were somewhat higher in Los Angeles than
in San Francisco and San Diego Counties among
persons residing for 10 or more years in their
respective counties, but the rates were highest in
the other less urbanized counties (Table 10-13).
Interpretation of these observations is clouded by
several factors: (1) The possibility that oxidant
pollution may induce chronic respiratory disease
only after a long latent period, (2) the lack of actual
air pollution measurements, and (3) the lack of
control for possible differences in socioeconomic
level among the areas compared. Winkelstem et
al.66'67 have observed a strong effect of
socioeconomic level on chronic respiratory disease
mortality. (In these reports, interarea differences in
socioeconomic level may have been minimized by
the fact that only American Legion members were
studied.)
DISCUSSION OF CHRONIC PHOTOCHEMICAL
POLLUTION EXPOSURES AND MORTALITY
Buell et al.5 observed no consistent association
between long-term oxidant exposure and lung
cancer mortality in California. Mahoney31 reported
higher total respiratory disease mortality rates in
inland, downwind sections of Los Angeles than in
coastal, upwind sections. In Los Angeles, oxidant
concentrations tend to be higher inland than in
coastal areas. Socioeconomic, demographic, and
behavioral variables were not fully controlled in
either study. Neither were mortality rates related
TABLE 10-12. LUNG CANCER DEATHS AND RELATIVE RISKS PER 100,000 MAN-YEARS IN AN
AMERICAN LEGION STUDY POPULATION, BY EXTENT OF CIGARETTE SMOKING AND RESIDENCE.
CALIFORNIA, 1958-625
Los Angeles
Daily cigarette smoking.
lifetime
hr jry"
None
Less than one pack
About one pack
More than one pack
County
Rate
28 1
636
1260
241 3
Relative
risk
2.5
57
11 3
21 5
S F Bay Area and
San Diego
Rate
43 9
77.1
1345
2260
Counties
Relative
risk
39
69
120
202
AM other
countie
Rate
11 2
6.10
1249
1375
is"
Relative
risk
1.0
54
11 2
123
Ratio
More than one pack
none
36
5 1
12.3
aAge-adjusted by the direct method to the total study population
°Nonsmokers m all other counties taken as unit risk
238
-------
to actual pollution measurements. Thus both of
them must be considered as inconclusive.
In view of the long latent periods known to be
involved in the development of many cancers and
chronic diseases, it may be that effects of
photochemical oxidant exposures on mortality
rates and other health indices have not yet become
apparent. Considerable further study, in which the
effects of socioeconomic and demographic
variables are carefully considered, is required
before the true relationship of photochemical
pollution to mortality rates can be conclusively
defined.
General Morbidity in Areas of High- and Low-
Oxidant Pollution
STATE OF CALIFORNIA HEALTH SURVEY
In 1954, weekly surveys of new illness and
injury were conducted throughout California by
the State Department of Public Health.6'78
Sampling units consisted of homes selected to be
representative of all homes in the state. Weekly
rates of illness and injury in Los Angeles County
were compared to those in the rest of the state
during the 17 weeks from August 2 through
November 29, 1954. Combined weekly incidence
of colds, asthma attacks, hay fever, and other
respiratory conditions in persons of all ages are
presented in Figure 10-11. As the figure shows,
there was little difference between Los Angeles
County and the rest of California in average rates of
these illnesses over the study period. The high
peaks of incidence in October are quite likely
reflections of the recent beginning of theacademic
year, though this is not stated in the report.
In Figure 10-1 2, weekly incidences of all illness
and injury in persons aged 65 and over are
presented. As the figure shows, the average
weekly incidence of illness and injury was about
10.4 percent in Los Angeles County and about 7.0
percent in the rest of the state. This finding
suggests that long-term exposures to
photochemical oxidants may promote increased
susceptibility to illness in the elderly. However, the
finding must be considered inconclusive, since the
investigators did not adjust for differences
between Los Angeles and the rest of California in
such factors as population density, ethnic charac-
teristics, and socioeconomic level.
In Figures 10-11 and 10-12, three high-smog
periods in Los Angeles County are noted. Criteria
for selection of these periods were not presented in
the report. Aerometric data collected during this
study in Los Angeles City and Pasadena suggest
that during these high-smog periods, maximum
concentrations of suspended particulates,
airborne lead, and carbon monoxide were
generally unusually high. However, maximum
concentrations of oxidant and nitrogen dioxide do
not appear to have been unusally high during these
periods. In any event, weekly incidences of illness
in Los Angeles County do not generally appear to
have risen appreciably during these periods.
In May and June 1956, another adult health
survey was undertaken throughout the state of
California.19'20 One goal of this survey was to
determine the prevalence of various conditions in
different areas of the state, A probability sample of
3545 households, selected to be representative of
all households in the state in the ratio of 1 to 1155,
was selected. Populations in service camps and
other institutions were not sampled. A ques-
tionnaire was administered by personal interview
to one adult in each sample household. After
weighting for the number of adults in each house-
hold, the total study sample consisted of 6939
persons. In this phase of the survey, care was
taken not to lead subjects toward attributing their
illnesses to air pollution.
Study results were distributed by area of the
respondent's residence. In Table 10-14 are pre-
sented the proportions of respondents reporting
TABLE 10-13. TOTAL CHRONIC RESPIRATORY DISEASE MORTALITY IN AN AMERICAN LEGION STUDY POPULATION,
CALIFORNIA, 1958-625
Residency
10 years
Less than 10 years
Unknown
Total
Los Angeles
County
Mortality
rale"
33.4
41 2
139 1
46 7
Total
deaths
31
14
12
57
San Franscisci
Area and San
Counties
Mortality
rate"
283
45 6
59 8
340
u Bay
Diego
Total
deaths
15
8
3
26
All other
counties
Mortality
rate"
456
41 3
397
444
Total
deaths
40
17
4
61
"Age- and smoking-adjusted by the direct method to (he total study population
"Per 100,000 man-years
239
-------
various respiratory conditions in three counties in
the Los Angeles/San Diego area, nine counties in
the San Francisco Bay Area, and the remainder of
the state. As the table shows, the proportion of
respondents reporting cough, nose complaints,
and throat complaints was higher in the Los
Angeles/San Diego area than elsewhere in the
state. The proportion of respondents reporting hay
fever was lowest in the Los Angeles/San Diego
area. Differences among areas in the proportion of
respondents reporting the other conditions listed
were slight.
75
CO
° 50
oc
UJ
Q.
§
o_
oc
UJ
Q.
CO
UJ
I-
< 25
oc
I I I I I I I
LOS ANGELES COUNTY
CALIFORNIA LESS
LOS ANGELES COUNTY
=HIGH SMOG PERIODS
I I I I I I
I I I I I I I I I I I
8
h*-
15 22
-AUG.
29
12 19
— SEPT.-
26
10 17
OCT.
24
31
WEEK ENDING
Figure10-11. Combined weekly incidence rates of selected conditions (colds, hay fever, asthma, and other respiratory
conditions) in persons of all ages, Los Angeles County and the remainder of California, August 7 to November 28,
1954.678
50
40
CO
1
oc
£30
§
cc
£20
01
10
1IIIIII
T 1 I
i I r
LOS ANGELES COUNTY
• CALIFORNIA LESS
LOS ANGELES COUNTY
=HIGH SMOG PERIODS
I I
8 15 22 29
AUG. »
5 12 19 26 3
•*—SEPT. H-«-
10 17 24 31
OCT. - *
7 14 21 28
NOV.
WEEK ENDING
Figure 10-12. Weekly incidence rates of illness and injury for persons aged 65 and over in
Los Angeles County and the remainder of California, August 2 to November 28,1954.678
240
-------
Another goal of this survey was to determine the
degree to which subjects attributed exacerbations
of respiratory conditions to air pollution. Figure 10-
1 3 presents the proportions of respondents in Los
Angeles County and the San Francisco Bay Area
attributing exacerbation of respiratory conditions
to air pollution and other factors. As the figure
shows, the proportion of respondents attributing
exacerbation of each condition to air pollution was
higher in Los Angelesthan in San Francisco. For all
conditions but bronchitis, intercity differences
were quite marked. Clearly, it cannot be concluded
that air pollution in Los Angeles caused all the
exacerbations that respondents attributed to it.
However, the findings presented in Figures 10-1 3
at least indicate that Los Angeles residents
considered local air pollution a hazard to health.
A panel consisting of all persons at least 30
years of age who appeared likely to have chronic
respiratory disease was selected from the 1956
general population survey described above.
Specifically, the panel was selected on the basis of
having reported chronic or repeated attacks of
bronchitis, asthma, or coughing in 1956. In
selecting the panel, the aid of a chronic respiratory
disease specialist and of other medical consultants
was enlistf d. The initial panel consisted of 1070
persons. The panel was interviewed on four
occasions- -twice in 1957, and once each in 1958
and 1959 Of the 1070 persons in the original
panel, 524 were interviewed all four times.
Results of the panel study were reported by
Hausknecr t.19 Of the general population sample of
3545 adul's interviewed in 1956 (representing a
weighted sample of 6939 persons), 41 percent
lived in three counties in the Los Angeles/San
Diego area, 27 percent lived in nine San Francisco
Bay Area counties, and 32 percent lived in the
remainder of the state. Of the 524 panelists
completing all four interviews in 1957-59, 48
percent lived in the Los Angeles/San Diego area,
25 percent lived in the San Francisco Bay Area,
and 27 percent lived in the remainder of the state.
Since all persons throughout the state having
reported evidence of chronic respiratory disease
were selected for the panel, this finding suggests
that rates of chronic respiratory disease were
disproportionately high in the Los Angeles/San
Diego area. Though it is not clearly b.ated in the
report, proportional distributions of bronchitis,
asthma, and cough appear to have been very
similar in the Los Angeles/San Diego area and in
the rest of the state, excluding the Bay Area. In the
Bay Area, the proportions of panelists with
bronchitis, asthma, and coughing appear to have
been, respectively, lower than, similar to, and
higher than corresponding proportions in panelists
elsewhere in the state. The logical consequence of
these finding is that, among the general
population, rates of bronchitis and asthma in the
Los Angeles/San Diego area appear to have been
higher than elsewhere in the state. Rates of cough
in the Los Angeles/San Diego area seem to have
been higher than in the rest of the state (excluding
the Bay Area), and they may or may not have been
higher than in the Bay Area.
Three factors substantially cioud the
interpretation of Hausknecht's findings. First,
many rates necessary to confirm logical inferences
are not presented in the report. Second, over half of
the original chronic respiratory disease panel did
not complete all four interviews. Illness rates
among these nonrespondents are not presented.
Third, statistical adjustments are not made for
TABLE 10-14. SELECTED RESPIRATORY CONDITIONS REPORTED BY GENERAL POPULATION SAMPLE.
CALIFORNIA, MAY 195619
Los Angeles, Orange,
and
reported
Bronchitis
Asthma
Cough
Sinus
Hayfever
Nose complaints
Throat complaints
Number of persons
interviewed
Call
Frequency
309
188
1341
1202
695
751
848
6939°
forma
Percent"
4
3
19
17
10
11
12
100
San Diego
Frequency
156
104
746
576
265
445
505
3450
Counties
Percent
5
3
22
17
8
13
15
100
San Francisco Bay
Area
Frequency
71
45
323
302
221
186
192
1846
Counties3
Percent
4
2
17
16
12
10
10
100
Rest
Frequency
82
39
272
324
209
120
151
1643
of state
Percent
5
2
17
20
13
7
9
100
3San Francisco, Alameda, Contra Costa, San Mateo, Santa Clara, Mann, Napa, Solano, and Sonoma Counties
bPercentages will not add to 100 because of reports of multiple conditions or erf none of the conditions listed
cThe number of persons interviewed personally was 3545 After weighting for the number of adults in sample households, this figure represented a total sample of 6939
241
-------
possible interarea differences in the distribution of
important covariates (most notably, cigarette
smoking habits).
CHRONIC RESPIRATORY DISEASE SURVEY OF
TELEPHONE WORKERS
Deane et a!.11 and Goldsmith and Deane17 used
standardized repsiratory illness survey techniques
to compare respiratory symptom rates in outdoor
telephone company workers in Los Angeles and
San Francisco. No aerometric data were presented
m these reports. In the older group (aged 50 to 59
years), respiratory symptoms were more frequent
in the population of Los Angeles than in that of San
Francisco. Persistent cough and phlegm were
reported by 31.4 percent of the group aged 50to 59
years in Los Angeles, compared with 1 4.3 percent
in San Francisco. This difference could not be
explained by differences in smoking habits, since
in both age groups, proportions of smokers were
higher in San Franciscothanin Los Angeles.There
were no substantial intercity differences in the
AIR POLLUTION
CONDITION
ASTHMA
NOSE
COMPLAINTS
THROAT
COMPLAINTS
SINUS TROUBLE
HAY FEVER
BRONCHITIS
51
50
10
44
41
30
10
22
11
OTHER FACTORS,a'b
49
94
50
90
56
95
59
91
70
90
78
89
n
n
LOS ANGELES COUNTY
SAN FRANCISCO BAY AREA
Includes the "don't know" responses.
Includes specific foods, overeating, working too hard,
not getting enough sleep, emotional upsets, smoking,
and the presence of other conditions.
WORSENING EFFECTS, percent
Figure 10-13. Percentage of respondents in Los Angeles County and San Francisco Bay Area
attributing exacerbation of respiratory conditions to air pollution and other factors, 1956. °
242
-------
results of pulmonary function tests. In the younger
group (aged 40 to 49 years), the frequencies of
respiratory symptoms were quite similar in the two
cities,
A point of interest is that rates of persistent
cough and persistent phlegm in San Francisco
were higher in the younger age group than in the
older. The proportion of San Francisco smokers
was also higher in the younger age group, which
may partly explain this finding. However, this is
probably not the full explanation, since the
younger group also had a higher proportion of
smokers in Los Angeles, where symptom rates
increased with age. The authors suggested that a
lost-to-study factor might have been operating in
the older group in San Francisco. Deane and
Goldsmith drew no conclusion as to the effect of
chronic oxidant exposures on respiratory symptom
rates.
Questions regarding eye irritation were also
asked by Deane and Goldsmith. The group aged 40
to 49 years in San Francisco reported eye irritation
about 10 percent of the time, and the group aged
50 to 59 reported it about 4 percent of the time The
corresponding figures in Los Angeles were 30 and
29 percent, respectively. Perhaps more important
is the fact that more than 50 percent of those in
both age groups in San Francisco had never
experienced eye irritation, whereas the
corresponding figure for Los Angeles was less
than 10 percent.
OXIDANTS AND EPIDEMIC INFLUENZA
Pearlman et al.42 studied whether chronic
exposure to photochemical oxidants affects
susceptibility to influenza infection. These
investigators performed a retrospective study
during 1968 and 1969 among 3500 elementary
school children residing in five southern California
communities. Seven to eight hundred second-,
fourth-, and sixth-grade students from upper-
middle-class socioeconomic areas in each city
participated. Selection of these communities was
based on historical differences in oxidant
exposure, although no difference in exposure was
present before or during the epidemic, Pasadena
and Riverside were chosen as the high-oxidant
cities, Garden Grove as an intermediate city, and
San Diego and Santa Barbara as low-oxidant
cities. Seasonal peak mean daily maximum hourly
oxidant levels recorded in each city from August to
October during 1964-67 were, respectively, 451,
368, 235, 1 57, and 1 76 fjg/m3 (0.23, 0 19, 0.1 2,
0.08, and 0.09 ppm). Concentrations of pollutants
other than oxidant in the study cities were not
presented in the report.
The incidence and duration of influenza-like
illnesses between November 1968 and January
1969 were determined by administration of a
questionnaire to the parents of the subjects during
March 1969. Parents were asked whether their
children had had influenza. If so, the length of
illness and presence or absence of fever, coryza,
and myalgia were also ascertained. Occurrence of
an influenza-like illness was defined as a febrile
illness accompanied by coryza and/or myalgia.
During April 1969, finger-prick blood specimens
were obtained from these children for subsequent
titration by hemagglutination inhibition (HI)
against Az/Hong Kong influenza and by
complement fixation against type A soluble
antigen. Complement fixation established a
hemoglobin HI liter of at least 1:32 as indicative of
recent infection with A2/Hong Kong influenza.
In the questionnaire survey, 72 8 percent of the
children reported illness, 35 4 percent reported
febrile illness, and 21.5 percent reported
influenza-like illness. Intercity morbidity did not
consistently reflect differences in chronic oxidant
exposure. A high- and a low-exposure city
(Pasadena and San Diego) reported the greatest
degree of febrile illness. Pasadena and Santa
Barbara, high- and low-exposure cities,
respectively, reported the most influenza-like
illness. Adjustment for age and sex differences by
probit analysis did not reveal any significant
differences in morbidity rates among cities.
Overall, 24.6 percent of children tested were
found to have HI liters greater than or equal to
1:32. The percentage of children with significant
titers rose from 10.3 percent among those with no
illness, to 28.9 percent among those with any
illness, to 39.5 percent in those with febrile illness,
and to 45.6 percent among those with an
influenza-like illness The frequency of positive
titers among febrile and influenza-like illness
categories in the two high-exposure cities was
lower than that in the two low-exposure cities. This
difference was statistically significant. Although
the two high-pollution cities studied, Pasadena
and Riverside, reported higher subclmical illness
rates (elevated titers without illness) than the low
pollution cities, the differences were not
statistically significant.
Thus Pearlman's study did not demonstrate a
positive association between chronic oxidant
243
-------
exposure and influenza-like illness In animal
studies, however, increased morbidity has
resulted from acute oxidant exposures.
Confirmation of Pearlman's results must await
studies in cities exhibiting clear-cut differences in
both acute and chronic oxidant concentrations.
Chronic Oxidant Exposure and Pulmonary
Function
SURVEY OF INSURANCE COMPANY WORKERS
Linn et al.29 reported a study in which pulmonary
function and respiratory symptom rates in indoor
insurance company workers in Los Angeles were
compared to those in workers for the same
company in San Francisco. Respiratory symptom
questionnaires based on the standard
questionnaire of the National Heart and Lung
Institute were administered to all subjects.
Pulmonary function tests included the forced vital
capacity (FVC), the 1 -sec forced expiratory volume
(FEVi), maximal expiratory flow at 50 percent and
25 percent of vital capacity ( V50 and V25), closing
volume, and slope of the single-breath N2 alveolar
plateau (AN2) Questionnaires were administered
in April 1973, and pulmonary function tests were
performed in August 1973 In one analysis, sex-
specific intercity comparisons of pulmonary
function test results were adjusted for age and
height. In another analysis, city, smoking, and city-
smoking interaction were employed as
independent variables. In all, 441 workers in San
Francisco and 206 workers in Los Angeles
participated in this study.
The proportions of participating males and
females were very similar in the two cities, as were
the mean heights and weights of men and women.
However, the study sample in San Francisco was
somewhat older than that in Los Angeles. In San
Francisco, 57 percentof the sample wasat least40
years of age, in Los Angeles, the corresponding
proportion was 48 percent. Also, higher
proportions of both men and women were current
smokers in San Francisco (45 percent and 30
percent, respectively) than in Los Angeles (35
percent and 19 percent, respectively). Further-
more, the Los Angeles sample contained a higher
proportion of Latin Americans than that in San
Francisco, whereas the San Francisco sample
contained a higher proportion of Asiatics than that
in Los Angeles.
Air pollutant concentrations in the two cities had
been measured from 1 969 to 1 972 at central-city
monitoring stations. Over this period, the median
oxidant concentration in Los Angeles, expressed
according to the post-1974 California standard
ultraviolet calibration method, was 0.07 ppm, and
the concentration exceeded on 1 0 percent of days
was 0.15 ppm. Corresponding concentrations in
San Francisco were 0.02 and O.C3 ppm,
respectively. Concentrations of nitrogen dioxide,
carbon monoxide, and total suspended
particulates were also higher in Los Angeles than
in San Francisco. Median concentrations of sulfur
dioxide appearto have-been very similar inthetwo
cities. Temperature was generally higher and
rainfall lower in Los Angeles than in San
Francisco, although relative humidity in summer
appears to have been slightly higher in Los
Angeles.
Sex-specific pulmonary function test results
were very similar in the two cities for all tests
performed except the AN2 in women, which was
higher in San Francisco than in Los Angeles to a
degree approaching statistical significance (p =
0.06). (Elevations in the AN2, in the absence of
other perturbations in pulmonary function, are
thought to reflect early obstruction of small
airways.) Cigarette smoking was found to have a
significant deleterious effect on several indices
of pulmonary function in males, and one index
(AN2) in females.
Nonpersistent coughing during bad weather at
times of day other than the first thing in the
morning was reported by 30 percent of women in
Los Angeles and by 14 percent of women in San
Francisco Nonpersistent phlegm production
under the same circumstances was reported by 23
percent of women in Los Angeles and by 11
percent of women in San Francisco. Respondents
had been instructed that high-smog episodes
could be considered periods of bad weather.
Among women, intercity differences in rates of
nonpersistent cough and phlegm were significant
at a = 0.01
The results of this study did not reveal a
consistent association between chronic
photochemical oxidant exposure and impairment
of respiratory health, though they suggested an
association between oxidant exposure and non-
persistent respiratory symptoms in women. In an
interesting discussion, the authors gave several
alternative explanations for their findings. First,
the effects of acute oxidant exposures may be
wholly reversible. Second, oxidant concentrations
in Los Angeles may have been offset by other
244
-------
adverse environmental factors such as cold and
dampness in San Francisco. Third, oxidant
exposures in Los Angeles may actually promote
the development of chronic lung disease, but the
difference in exposure between Los Angeles and
San Francisco may not have been great enough to
show such an effect.
STUDY OF SEVENTH DAY ADVENTISTS
Cohen et al.10 examined the effects of oxidant air
pollution on pulmonary function in nonsmoking
adults. Seventh Day Adventists were chosen for
study, since few of them smoke, and they
constituted a readily accessible population that
was interested in health. The San Gabriel Valley in
California was selected as the high-pollution area,
and the San Diego area was chosen as a low-
pollution area.
The study population in each area consisted of
white English-speaking adults aged 45 to
64 who had not smoked in the last 20 years and
who had a lifetime history of smoking less than one
pack per year. Data were collected in January
1970. Each participant was interviewed by a
physician and underwent tests of ventilatory
function, including the 1-sec expiratory volume
(FEV,) and the forced vital capacity (FVC). Study
participants were also requested to complete a
respiratory questionnaire that was a modified
combination of the British Medical Research
Council respiratory questionnaire and the London
School of Hygiene cardiovascular questionnaire.
Aerometric data were obtained from the National
Air Pollution Control Administration, which made
measurements during September 1969 and
January 1970. Daily oxidant measurements for
January 1970 were also obtained from the Los
Angeles County and San Diego County Air
Pollution Control Districts. In addition, historical
records from the Air Pollution Control Districts of
these two counties were reviewed to obtain an
estimate of chronic exposures.
Average daily maximum hourly oxidant levels in
the San Gabriel Valley and in San Diegoduring the
month of study were 235 and 137 pig/m3 (0.12
ppm and 0.07 ppm), respectively. Measurements
of total suspended particulates, respirable
particulates, and sulfur dioxide were essentially
equivalent in both areas during the time of study.
From 1963 through 1967, both areas experienced
similar arithmetic mean oxidant concentrations
(0.047 ppm in the San Gabriel Valley and 0.038
ppm in San Diego). However, over the same period,
the mean of daily maximum hourly oxidant
concentrations in the San Gabriel Valley (0.144
ppm) was nearly twice as high as the cor-
responding mean of 0.074 ppm in San Diego. Also,
the proportion of days in this period on which daily
maximum hourly oxidant concentrations in the San
Gabriel Valley exceeded 0.1 5 ppm (44.8 percent of
days) was about seven times higher than the
corresponding proportion of days (6.1 percent) in
San Diego. Average concentrations of other
pollutants (carbon monoxide, nitric oxide, nitrogen
dioxide, hydrocarbons, sulfur dioxide, and
particulates) during 1 963-67 were generally high.-
er in the San Gabriel Valley than in San Diego.
The prevalence of respiratory symptoms did not
differ significantly between the two areas.
Prevalence rates generally were considerably
lower than those observed in most other studies.
Similarly, no significant differences were found in
measurements of pulmonary function between
areas, and all measurements were within ex-
pected normal values.
Thus Cohen et al.10 demonstrated no significant
difference in symptom rates or pulmonary function
between the San Gabriel Valley and San Diego.
Interpretation of this study is clouded by the
similarity of annual average exposure levels
between the two areas. Clarification of the
question of whether ambient oxidant exposure
affects pulmonary function and respiratory
symptom rates must await studies in areas having
unequivocally different acute and chronic
exposure levels.
DISCUSSION OF CHRONIC OXIDANT
EXPOSURES AND PULMONARY FUNCTION
As yet, no association between chronic ambient
oxidant exposure and pulmonary function has
been observed. However, as with other indices of
chronic exposure, very few studies investigating
such a relationship have yet been performed.
Furthermore, those that have been performed are
limited in their import, because oxidant exposures
for the test population cannot be known with
certainty and because complicating effects of other
personal and community environmental factors
cannot be reliably evaluated.
We have been unable to find any studies of long-
term exposure effects on the pulmonary function
of those whose health is impaired. Also, no
prospective studies have yet been undertaken to
determine whether chronic exposures promote
declines in the pulmonary function of once-healthy
245
-------
individuals. Finally, effects of long-term oxidant
exposure on pulmonary function, as on other
indices of chronic exposure, may only now be
starting to become evident.
ATTITUDES OF LAYMEN AND PHYSICIANS
TOWARD OXIDANT AIR POLLUTION
State of California General Health Survey
At the end of each interview in the general
health survey undertaken in California in 1956 and
previously described,20 direct questions con-
cerning the effects of air pollution were asked.
Seventy-five percent of the whole surveyed
population in Los Angeles County was bothered by
air pollution, in contrast to 24 percent in the San
Francisco Bay Area and 22 percent in the rest of
the state. Corresponding proportions among the
working population were, respectively, 80, 29, and
27 percent (Table 10-15). Thirty-two percent of
native-born Californians were bothered by air
pollution, and 39 percent of those moving to
California were so bothered. Of those who were
bothered by air pollution, 17 percent in Los
Angeles considered moving because of it, in
contrast to 4 percent in San Francisco and 12
percent in the rest of the state. Of the same total
number, 9 percent in Los Angeles considered
changing jobs because of air pollution, in contrast
to 3 percent in both San Francisco and the rest of
the state. About 20 percent of the state residents
who had moved out of a polluted area reported that
pollution had some influence on their decision to
move; 4 percent gave air pollution as their sole
reason for moving. Among those who moved from
California communities because of air pollution,
75 percent had moved out of Los Angeles County, 8
percent had moved out of the San Francisco Bay
Area, and 17 percent had moved out of other areas
of the state. Air pollution was given as the reason
for 13 percent of the moves from Los Angeles
County since 1947. In other areas of the state, the
proportion of moves attributed to air pollution was
1 percent or less.
Eye irritation was the most frequently reported
effect of air pollution. In some instances, this
symptom was accompanied by nasal irritation
(Table 10-1 6). In metropolitan areas, 80 percent of
respondents bothered by air pollution complained
of aye irritation. In nonmetropolitan areas, the
corresponding proportion was 29 percent.
Before air pollution was specifically mentioned
by interviewers, respondents were asked several
questions about satisfaction with the communities
in which they lived. About 21 percent of all
respondents in Los Angeles County expressed
dissatisfaction with the communities in which they
lived, as compared to 18 percent in both the Bay
area and the rest of the state. A far greater
proportion of dissatisfied Los Angeles residents
(32 percent) voluntarily attributed their
dissatisfaction to air pollution than did residents of
the San Francisco Bay Area (1 percent) or those in
the rest of the state (6 percent).
TABLE 10-15. PERCENT OF SURVEY RESPONSE OF
GENERAL AND WORKING POPULATION BOTHERED
BY AIR POLLUTION, BY MAJOR GEOGRAPHIC
AREAS IN CALIFORNIA, MAY 195620
Responses
General popula-
tion sample
% Not bothered by
air pollution3
% Bothered by air
pollution
Either at home
or work
Both at home
and at work
At home only
At work only
Total % at home
Total % at work
Working popula-
tion sample
% Not bothered by
air pollution
% Bothered by air
pollution
Either at home
or work
Both at home
and at work
At home only
At work only
Total % at home
Total % at work
California
6393
55
45
14
24
7
38
21
3732
51
49
25
12
12
37
38
Los
Angeles
County
2892
24
75
27
39
8
66
35
1577
20
80
49
16
15
66
65
San
Francisco
Bay Area
1846
76
24
4
14
6
18
10
1028
71
29
7
11
11
18
18
Rest
of
state
2210
78
22
4
13
5
17
9
1127
73
27
9
8
10
17
18
aPercents are rounded independently
Survey of Los Angeles Physicians
A survey of Los Angeles physicians was
conducted jointly by the Los Angeles County
Medical Association and Tuberculosis and Health
Association in December I960.30 A sample
representing about one in 16 of those physicians
registered to practice in the county during 1958
was drawn, resulting in a sample of 526 from a
total of 9228 physicians. A pretested question-
naire was mailed with a letter signed by the
246
-------
chairman of the air pollution subcommittee. A
follow-up was also mailed, and telephone calls
were made to the offices of those physicians who
had not responded.
Of the questionnaires mailed, 350 were re-
turned. Of those, 307 (58 percent of the original
sample of 526) were completed and tabulated. The
words "air pollution" did not appear in the
questionnaire, although bias could have been
introduced by the fact that the chairman of the air
pollution subcommittee attached a letter to the
questionnaire. Seventy-seven percent of the
physicians who returned completed ques-
tionnaires believed that air pollution adversely
affected the health of their patients. Two-thirds of
the responding physicians felt that air pollution
was a factor adversely affecting chronic
respiratory disease. One-third of the physicians
had advised one or more of their patients to leave
the Los Angeles area for health reasons; air
pollution was a factor mentioned in two-thirds of
these instances. By extrapolation from the sample,
assuming it to be representative, it was estimated
that physicians had advised more than 10,000
patients to move. It was reported that
approximately 25 percent of the patients had done
so. Nearly one-third of the physicians had
themselves considered moving from the Los
Angeles area because of air pollution. Among
other environmental factors mentioned as
deleterious to health were overcrowding and
traffic congestion, but these factors were not
mentioned nearly as often as air pollution.
Discussion of Attitudes of Respondents and
Physicians Toward Oxidant Air Pollution
The studies summarized in this section (10.4)
were subjective in nature and did not yield
scientifically confirmed information about rates of
illness or functional impairment associated with
oxidant pollution. However, the results presented
here are interesting in that they demonstrate
considerable concern, on the part of laymen and
physicians alike, about the effects of such
pollution.
SUMMARY OF EPIDEMIOLOGIC APPRAISAL
OF PHOTOCHEMICAL OXIDANTS
In this summary, a brief discussion of findings in
the preceding three sections of this chapter will be
presented (Effects of Short-Term Photochemical
Oxidant Exposures, Effects of Chronic Photo-
chemical Oxidant Exposures, and Attitudes of
Respondents and Physicians Toward Oxidant Air
Pollution). The discussion of each section will
begin with the health indices that can be most
confidently associated with ambient oxidant
exposures and will work toward those indices for
which little association with such exposures has
been shown. Directions for future research will
also be identified.
TABLE 10-16. AIR POLLUTION EFFECTS REPORTED IN GENERAL POPULATION SURVEY, BY TYPE OF COMMUNITY
AND BY MAJOR GEOGRAPHIC AREAS IN CALIFORNIA, MAY 195620
Total
California
Item
General population sample
Respondents bothered
by air pollution
% Bothered by air pollution
Air pollution effects cited
Eyes, effects
Eye irritation
Eye and nasal irritation
Eye irritation and annoying
Eye, nasal irritation and annoying
Nasal irritation, eye not mentioned
Nasal irritation
Nasal irritation and annoying
Annoying only
Other effects only
No effects reported
Total %
At
home
6939
2616
38
75
44
23
5
3
10
8
2
5
5
5
100
At
work
3732
1410
37
76
46
24
3
3
9
8
1
7
2
6
100
Los
Angeles
County
At
home
2392
1904
66
89
54
26
6
3
5
4
1
2
2
2
100
At
work
1577
1012
64
% Persons
88
53
27
4
4
4
3
1
3
1
4
100
San
Francisco
Bay Area
At
home
1846
326
18
bothered
38
17
15
4
2
22
19
3
17
5
18
100
At
work
1028
190
18
39
18
14
3
4
23
19
4
26
5
7
100
Rest
of
Sti
At
home
2201
386
17
41
22
14
3
1
22
18
4
10
17
10
100
ite
At
work
1127
208
19
51
30
19
2
21
21
8
4
16
100
247
-------
Effects of Short-Term Photochemical Oxidant
Exposures
The health index most frequently and con-
sistently associated with short-term exposures to
ambient photochemical oxidants is eye irritation,
All available evidence suggests that ozone alone is
not an eye irritant at ambient concentrations.
However, the possibility that ozone may interact
with other substances to produce eye irritants
remains open. In American studies, daily
maximum hourly oxidant concentrations above
which the rates and severity of eye irritation have
been observed to increase have ranged from about
200 /jg/m3 (0.1 ppm) to about 294 /jg/m3 (0.15
ppm). Recent Japanese studies raise the possibility
that even lower oxidant concentrations may
promote eye irritation under certain conditions.
The consistency of the association between short-
term exposures and eye irritation arouses interest
as to what the ocular effects of long-term
exposures might be. Until careful studies of long-
term exposure effects on the eye have been
performed, the question of whether transient
instances of eye irritation constitute bona fide
impairments of health must remain open.
Short-term photochemical oxidant exposures
have been quite consistently associated with
decrements in human performance. Wayne etal.65
have observed the proportion of high school cross-
country runners failing to improve their running
times to increase as hourly oxidant concentrations
increased from about 59 to 590 /jg/m3 (0.03 to
0.30 ppm). However, in the range of 0.03 to 0.1 0
ppm oxidant, no consistent linear relationship
between oxidant concentration and performance
could be detected. Folinsbee's experimental
studies suggest that on high-oxidant days, the
irritant effect of pollutants may have restricted the
runners' mechanical lung function sufficiently to
prevent them from taking up enough oxygen to
support the performance level of which they were
potentially capable. The studies of Folinsbeeet
al.14'15 suggest further that ozone alone may have
been responsible for this effect.
Dry et al.63 have reported a statistically
significant positive association between hourly
oxidant levels and automobile accident frequency
in Los Angeles. The study design employed by Dry
did not allow dose-response relationships to be
developed. Reduced visual acuity, reduced
visibility, and eye irritation may all be contributing
factors to Ury's findings.
The observations of Lebowitz et al.27 suggest
that the degree to which photochemical oxidants
affect pulmonary function may depend quite
heavily on the subject's level of exercise. This
observation is quite consistent with the
experimental findings of Bates1 and Folinsbee et
al.1415 It must be noted that too few pollution
measurements were reported by Lebowitz to
support a dose-response relationship. Also, the
Lebowitz et al. study design did not permit oxidant
effects to be separated from the effects of other
pollutants and of meteorologic'factors.
Kagawa and Toyama24'25 observed that
impairment of pulmonary function in Japanese
schoolchildren was more strongly associated with
exposure to ozone than to total oxidants. This
observation enhances confidence that ozone alone
may be responsible for decrements in pulmonary
function, no matter what the effects of other
pollutants may be. Kagawa also observed larger
correlations of ozone concentration with indices
thought to reflect large airway function (airway
resistance and specific conductance) than with
indices thought to reflect small airway function
(instantaneous flow at 25 and 50 percent of vital
capacity). The Kagawa and Toyama study design
did not allow dose-response relationships to be
inferred.
Several investigators have noted associations
between short-term oxidant concentrations and
the frequency of respiratory and other symptoms in
healthy people. In their large study of student
nurses, Hammer et al.18 observed simple
frequencies of cough and chest discomfort and the
adjusted frequency of headache to increase with
daily maximum hourly oxidant concentrations in
and above the range of 588 to 764 jug/m3 (0.30 to
0.39 ppm). Japanese investigators have observed
the frequencies of several symptoms in
schoolchildren, including sore throat, headache,
cough, and dyspnea, to be higher on days when
maximum hourly oxidant concentration equalled
or exceeded 0.15 ppm than on days when
corresponding concentrations were below 0.10
ppm. In Japanese studies, symptom frequencies
have generally been more strongly associated with
total oxidant concentrations than with ozone
concentrations. Results of Japanese studies also
suggest that people with allergic tendencies and
orthostatic dysregulation are more susceptible to
short-term photochemical oxidant exposure than
are other segments of the population
Japanese studies have generally shown
oxidant- or ozone-associated effects at lower
248
-------
measured oxidant or ozone concentrations than
American studies. The consistent difference
between Japanese and American findings raises
several questions, including whether the
components of oxidant pollution in Japan are
different from those in Los Angeles, or, indeed,
from those in any U.S. location. It is possible, for
instance, that oxidants in Japan are accompanied
by higher sulfur oxide levels than in the United
States. In any case, the Japanese results cited m
this chapter underscore the point that
epidemiologic results gathered in one area can be
generalized to other areas only with the greatest
caution. The Japanese results also demonstrate
the need to gather comprehensive data in U.S.
locations other than Los Angeles.
As yet, it is not possible to judge confidently
whether transient symptoms, irritation, or
decrements in pulmonaryfunctionconstitutebona
fide impairments of health. As with eye irritation,
confident judgement must await studies of the
cumulative effect of many of these minor insults
occurring over long periods of time.
Short-term photochemical oxidant, exposures
have also be_en associated with aggravation of
existing disease, though not as reliably as with the
production of symptoms and minor illness in the
healthy. Schoettlin and Landau53 observed a
significantly higher rate of asthma attacks on days
when the (presumably maximum hourly) oxidant
concentration exceeded 0.25 ppm than on days
when it did not. The results of Schoettlin and
Landau also suggested that a portion of the
asthmatic population might be particularly
susceptible to oxidants.
Motley et al.38, Remmers and Balchum,45 and
Ury and Hexter62 have all observed a beneficial
effect of air filtration on the lung function of
patients with chronic respiratory disease. In
studies reported by these investigators, changes in
pulmonary function appear to have been more
strongly correlated with changes in oxidant
concentrations than with changes in nitric oxide or
nitrogen oxide concentrations. However, these
studies do not support confident estimates of dose-
response relationships. Also, these studies are
open to fairly serious methodologic questions.
Studies of the relationship between short-term
oxidant exposures and hospital admissions have
yielded mixed results. Although Brant and Hill2'3
and Sterling et al.59'60 have observed positive
associations between exposures and admissions,
their observations must as yet be considered
inconclusive for at least three reasons. First, these
investigators have not ruled out an association
between nonoxidant pollutants and admissions
that might be just as strong as that between
oxidant and admissions. Second, the Brant and Hill
observation of high positive correlations between
oxidant concentrations and admissions 4 weeks
later is difficult to rationalize pathophysiologically,
particularly in view of the negative correlations
occurring m the intervening time. Third, co-
efficients of correlation between oxidant
concentration and same-day admissions have not
exceeded the relatively small value of 0.27 (r2 -
0.073).
As yet, no convincing association has been
shown between short-term oxidant exposures and
rates of mortality resulting from any cause. The
positive association observed by Mills36 may yet
prove to be valid. However, the vigor with which his
conclusions can be advanced is considerably
limited by the reanalysis by Breslow and
Goldsmith4 of a considerable portion of the data
used by Mills, in which no association between
oxidants and mortality was observed.
Interpretation of mortality studies to date has
been hampered by limitations in statistical
methodology. It has not yet proven possible to
separate fully the effects on mortality of oxidants,
other pollutants, and meteorologic factors, most
notably temperature. Future improvements in
statistical methodology may allow these factors to
be separated more clearly than they can be at
present. Future biomedical research should also
enhance knowledge of the combined effects of
pollution and meteorologic factors. Only with such
advances can the true relationships of oxidants to
mortality be determined,
Effects of Chronic Photochemical Oxidant
Exposures
The effects on human health of long-term
photochemical oxidant exposures have not been
characterized nearly as completely as those of
short-term exposures. When impairments in
health have been associated with long-term
exposures, it has not yet proven possible to
determine the level or duration of exposure
necessary to promote the impairments, since most
studies of long-term exposure effects have been
cross-sectional in design, not longitudinal. Also,
studies of long-term exposure effects, whether or
not they have revealed associations between
exposure and health impairment, have often been
249
-------
rather limited with respect to methods of data
collection or analysis. Furthermore, since no
American city has experienced elevated oxidant
levels over more than about 30 years, consistent
effects of long-term exposures conceivably have
gone undetected before the present time, no
matter how carefully studies were designed and
conducted.
Several studies suggest a tenuous association
between long-term oxidant exposures and chronic
respiratory morbidity. Deane and Goldsmith11 have
observed higher rates of persistent cough and
phlegm in Los Angeles telephone workers aged 50
through 59 years than in comparable workers in
San Francisco. Hausknecht has reported that a
disproportionate number of chronic respiratory
disease patients, randomly selected from the
whole state of California, lived in the Los Angeles
area. Linn has observed higher rates of non-
persistent cough and phlegm in women living in
Los Angeles than in San Francisco, a finding that
would reflect an effect of oxidant on chronic or
acute respiratory morbidity However, the findings
of these studies are not clear with respect to the
existence of an association between oxidant ex-
posure and chronic respiratory disease. Nor do the
findings allow confident inference as to the level or
duration of oxidant exposure necessary to promote
increased rates of chronic respiratory illness.
Buell et al.5 observed no consistent association
between long-term oxidant exposure and lung
cancer mortality in California. Mahoney31 has
reported higher total respiratory disease mortality
rates in inland, downwind sections of Los Angeles
than in coastal, upwind sections. However,
limitations in statistical control for important
covanates and a lack of actual pollution
measurements render the results of both studies
inconclusive.
In view of the long post-exposure latent periods
known to be involved in the development of cancer
and other chronic diseases, it is quite conceivable
that an influence of chronic oxidant exposure on
mortality could go undetected until now or
sometime in the future Studies conducted
carefully over the next decade or so may well
provide the most useful information concerning
the relationship of chronic oxidant exposure to
mortality.
As yet, no association between chronic oxidant
exposure and acute illness has been shown. As far
as we are aware, the Pearlman et al study42 is the
only one available in this field
Linn et al.29 and Cohen et al.10 have reported
epidemiologic studies of chronic oxidant
exposures and pulmonary function. In neither
study has an association between such exposure
and pulmonary function been observed. However,
interpretation of these results is limited by
imperfect matching of high- and low-exposure
populations, incomplete knowledge of actual
exposure levels, and the relative absence of
individuals likely to be at high risk in the study
populations (e.g., those with clear-cut chronic
respiratory disease).
Attitudes of Laymen and Physicians Toward
Oxidant Air Pollution
Studies of community attitudes toward oxidant
pollution do not yield scientifically confirmed
information regarding the effect of oxidant
pollution on illness rates and physiologic function.
However, these results are interesting in that they
demonstrate considerable community concern
about the effect that oxidant pollution may be
exerting on the public health
REFERENCES FOR CHAPTER 10
1 Bates, D V , and M Hazucha The short-term effects of
ozone on the human lung In Proceedings of the
Conference on Health Effects of Air Pollutants, prepared
for the U S Senate, Committee on Public Works
Government Printing Office, Washington, D C , 1 973 pp
507-540
2 Brant, J W. A Human cardiovascular diseases and
atmospheric air pollution in Los Angeles, California Int
J Air Water Pollut 9219-231,1965
3 Brant, J W A, and S R G Hill Human respiratory
diseases and atmospheric air pollution in Los Angeles,
California Int J Air Water Pollut S 259-277, 1 964
4 Breslow, L, and J Goldsmith Health effects of air
pollution Am J Public Health 48 913-917, 1958
5 Buell, P , J E Dunn, Jr , and L Breslow Cancer of the
lung and Los Angeles-type air pollution Prospective
study Cancer (Philadelphia) 202139-2147, 1967
6 California State Department of Public Health Clean Air
for California Initial Report of the Air Pollution Study
Project State of California, Department of Public Health,
San Francisco, Calif , March 1955
7 California State Department of Public Health Clean Air
for California Second Report State of California,
Department of Public Health, Berkeley, Calif, March
1956
8 California State Department of Public Health Report III A
Progress Report of California's Fight Against Air
Pollution State of California, Department of Public
Health, Berkeley, Calif, February 1957
9 Cassell, E J,J R McCarroll, W Ingram, and D Wolter
Health and the urban environment Air pollution and
family illness III Two acute air pollution episodes in New
250
-------
York City. Health effects Arch Environ Health 10 367-
369, 1965
10 Cohen, C A, A R Hudson, J L. Clausen, and J H
Knelson Respiratory symptoms splrometry, and oxidant
air pollution in nonsmoking adults Amer Rev Respir
Dis 105 251-261, 1972
1 1 Deane, M., J R Goldsmith, and D Tuma Respiratory
conditions in outside workers Report on outside plant
telephone workers in San Francisco and Los Angeles
Arch Enrivon Health 10 323-331, 1965
12 Durham, W H Air pollution and student health Arch
Environ Health 28 241 -254, 1974
13 Fetner, R H Ozone-induced chromosome breakage in
human cell cultures Nature (London) 194 793-794,
1962
14 Folmsbee, L J, F Silverman, and R J. Shepard
Decrease of maximum work performance following
ozone exposure J Appl Physiol Respirat Environ
Exercise Physiol 42 531-536, 1977
1 5 Folmsbee, L J , F Silverman, and R J Shepard Exercise
responses following ozone exposure J Appl. Physiol
58996-1001, 1975
16 Goldsmith, J R Personal communication to Delbert S
Barth, Director, Bureau of Criteria and Standards,
National Air Pollution Control Administration, July 20,
1970
17 Goldsmith, J R , and M Deane Outdoor workers in the
United States and Europe Milbank Mem Fund Q 43
107-116, 1965
18 Hammer, D I, V Hasselbald, B Ponnoy, and P F
Wehrle The Los Angeles student nurse study Daily
symptom reporting and photochemical oxidants Arch
Environ Health 28 255-260, 1 974
19 Hausknecht, R Experiences of a respiratory disease
panel selected from a representative sample of the adult
population Am Rev Respir Dis. 86 858-866, 1962
20 Hausknecht, R , and L Breslow Air Pollution Effects
Reported by California Residents from the California
Health Survey State of California, Department of Public
Health, Berkeley, Calif, 1960
21 Air Quality Bureau Health Hazards of Photochemical Air
Pollution (The Results of a Survey of Health Hazards of
Photochemical Air Pollution in 1975) Environment
Agency, Tokyo, Japan, March 1976
22 Hechter, H H.andJ R Goldsmith Air pollution anddaily
mortality Am J Med Sci 241 581-588, 1961
23 Herman, D R The Effect of Oxidant Air Pollution on
Athletic Performance M S thesis, University of North
Carolina, Chapel Hill, N C , 1972
24 Kagawa, J , and T Toyama Photochemical air pollution
Its effects on respiratory function of elementary school
children Arch Environ Health 30 1 1 7-1 22, 1975
25 Kagawa, J , T Toyama, and M Nakaza Pulmonary
function tests in children exposed to air pollution In
Clinical Implications of Air Pollution Research A J
Fmkel and W C Duel, eds Publishing Sciences Group,
Inc , Acton, Mass , 1 976 pp 305-320
26 Lagerwerff, J M Prolonged ozone inhalation and its
effects on visual parameters Aerosp Med 54479-486,
1963
27 Lebowitz, M D,P Bendheim, G Cristea, D Markowitz, J
Misiaszek, M Staniec, and D VanWyck The effect of air
pollution and weather on lung function in exercising
children and adolescents Am Rev Respir Dis 703.262-
273, 1974
28 Leeder, S R Role of infection in the cause and course of
chronic bronchitis and emphysema J Infect. Dis. 757
731-742, 1975
29 Linn, W S,J D. Hackney, E E Pedersen, P Breisacher,
J. V Patterson, C A Mulry, and J F Coyle Respiratory
function and symptoms in urban office workers in
relation to oxidant air pollution exposure Am Rev
Respir Dis 774477-483,1976
30 Los Angeles County Medical Association Physicians
Environmental Health Survey A Poll of Medical Opinion
County of Los Angeles, Medical and Tuberculosis and
Health Associations, Los Angeles, Calif , May 1961
31 Mahoney, L E , Jr Wind flow and respiratory mortality in
Los Angeles. Arch Environ. Health 22 344-347, 1971
32 Makino, K , and I. Mizoguchi Symptoms caused by
photochemical smog Nihon Koshu Eisei Zasshi 22 421-
430, 1975
33 Massey, F J,Jr,E Landau, and M Deane Air Pollution
and mortality in two areas of Los Angeles County
Presented at the Joint Meeting of the American
Statistical Association and the Biometric Society (ENAR),
New York, December 27, 1 961
34 McMillan, R S, D H Wiseman, B Hanes, and P F
Wehrle Effects of oxidant air pollution on peakexpiratory
flow rates in Los Angeles schoolchildren Arch Environ
Health 18 94-99, 1969
35 Mills, C A Do smogs threaten community health?
Cincinnati J Med 38259,1957
36 Mills, C A Respiratory andcardiacdeathsin LosAngeles
smogs Am J Med Sci 255379-386,1957
37 Mizoguchi, I , K Makino,S Kudou.andR Mikami Onthe
Relationship of Subjective Symptoms to Photochemical
Oxidant In International Conference on Photochemical
Oxidant Pollution and Its Control Proceedings Vol I B
Dimitriades, ed EPA-600/3-77-001 a, US
Environmental Protection Agency, Research Triangle
Park, N C , January 1977 pp 477-494
38 Motley, H L, R H Smart, and C I Leftwich Effect of
polluted Los Angeles air (SMOG) on lung volume
measurements J Am Med Assoc 777 1469-1477,
1959
39 National Air Pollution Control Administration Air Quality
Criteria for Photochemical Oxidants NAPCA Publication
No AP-63, U S Department of Health, Education, and
Welfare, Public Health Service, Washington, D C , March
1970
40 National Air Pollution Control Administration Errata for
Air Quality Criteria for Photochemical Oxidants NAPCA
Publication No AP-63, U S Department of Health,
Education, and Welfare, Public Health Service,
Washington, D C , March 1970 p 2
41 Oechsli, F W , and R W Buechley Excess mortality
associated with three Los Angeles hot spells Environ
Res 5 277-284, 1970 (1b)
42 Pearlman, M E,J F Fmklea, C M Shy, J Van Bruggen,
and V A Newill Chronic oxidant exposure and epidemic
influenza Environ Res 4129-140,1971
43 Personal communication from 0 J Balchum to R E
Carroll 1967
44 Peters, J M., R L H Murphy, B G Ferns, W A Burgess,
M V Ranadive, and H P Pendergrass Pulmonary
251
-------
function in shipyard welders An epidemiologic study
Arch Environ Health 26 28-31 1973
45 Remmers, J E.andO J Balchum Effects of Los Angeles
urban air pollution upon respiratory function of
emphysematous patients Presented at 58th Annual
Meeting, Air Pollution Control Association, Toronto,
Canada, June 21, 1965
46 Renzetti, N A, and V Gobran Studies of Eye Irritation
Due to Los Angeles Smog 1954-1956 Air Pollution
Foundation, San Marino, Calif , July 1, 1957
47 Renzetti, N A , and H G Marcus, eds An Aerometric
Survey of the Los Angeles Basin August-November,
1954 Report No 9, Air Pollution Foundation, Los
Angeles, Calif , July 1, 1955
48 Richardson, N A , and W C Middleton Evaluation of
filters for removing irritants from polluted air Heat
Piping Air Cond 30147-154,1958
49 Richardson, N A, and W C Middleton Evaluation of
Filters for Removing Irritants from Polluted Air Report
No 57-43, University of California, Department of
Engineering, Los Angeles, Calif , June 1957
50 Rokaw, S N , and F Massey Air pollution and chronic
respiratory disease Am Rev Respir Dis 86 703-704,
1962
51 Sawicki, E Airborne carcinogens and allied compounds
Arch Environ Health 7446-53, 1967
52 Schoettlm, C E The health effect of air pollution on
elderly males Am Rev Respir Dis 86 878-897, 1962
53 Schoettlm, C E, and E Landau Air pollution and
asthmatic attacks in the Los Angeles area Public Health
Rep 76 545-548, 1961
54 Shimizu, K , M Harada, and M Miyata Effects of
photochemical smog on the human eye Rinsho Ganka
30407-418, 1976
55 Shimizu, T Classification of subjective symptoms of
junior high school students affected by photochemical air
pollution Taiki Osen Kenkyu 3 734-741, 1975
56 Stanford Research Institute The Smog Problem in Los
Angeles County An Interim Report on Smog Research.
Western Oil and Gas Association, Committee on Smoke
and Fumes, Los Angeles, Calif, 1 948
57 Stanford Research Institute The Smog Problem in Los
Angeles County Second Interim Report on Studies to
Determine the Nature and Sources of the Smog Western
Oil and Gas Association, Committee on Smoke and
Fumes, Los Angeles, Calif, 1949
58 Stanford Research Institute The Smog Problem in Los
Angeles County Third Interim Report on Studies to
Determine the Nature and Sources of the Smog Western
Oil and Gas Association, Committee on Smoke and
Fumes, Los Angeles, Calif , 1951
59 Sterling, T D.J J Phair, S V Pollack, D A Schumsky,
and I DeGroot Urban morbidity and air pollution A first
report Arch Environ Health 13 158-170, 1966
60 Sterling, T D, S V Pollack, and J H Phair Urban
hospital morbidity and air pollution A second report
Arch Environ Health 15 362-374, 1967
61 Ury, H K Photochemical air pollution and automobile
accidents in Los Angeles Arch Environ Health 1 7334-
342, 1968
62 Ury, H K , and A C Hexter Relating photochemical
pollution to human physiological reactions under
controlled conditions Arch Environ Health 75 473-480,
1969
63 Ury, H K., N M. Perkins, and J R Goldsmith Motor
vehicle accidents and vehicular pollution in Los Angeles
Arch Environ Health 25 314-322, 1972
64 Wayne, W S., and P F Wehrle Oxidant air pollution and
school absenteeism Arch Environ Health 73 31 5-322,
1969
65 Wayne, W S,P F Wehrle, and R E Carroll Oxidant air
pollution and athletic performance J Am Med Assoc
733901-904, 1967
66 Wmkelstem, W., Jr , S Kantor, E W Davis, C S Manen,
and W E Mosher The relationship of air pollution and
economic status to total mortality and selected
respiratory system mortality in men I Suspended
particulates Arch Environ Health 14 162-171, 1967
67 Wmkelstem, W , Jr ,S Kantor, E W Davis, C S Manen,
and W E Mosher The relationship of air pollution and
economic status to total mortality and selected
respiratory system mortality in men II Oxides of sulfur
Arch Environ Health 76401-405, March 1968
252
-------
11. EFFECTS OF PHOTOCHEMICAL OXIDANTS ON
VEGETATION AND CERTAIN MICROORGANISMS
INTRODUCTION
Injury to terrestrial vegetation was one of the
earliest manifestations of photochemical oxidant
• I .- 2,8,9,21,22,24,100,144,160,199,200,201,203,212,216,
217,2lS°20,238,263,266.273,274,300,314 |nvestigatjons by
Middleton et a!.203 in 1944 described smog-
induced injury to leafy vegetables, ornamentals,
and field crops in a small area of Los Angeles
County. By 1950, such injury was observed over a
large portion of southern California and the San
Francisco Bay Area.199 Plant injury caused by
photochemical oxidants now occurs commonly in
most, if not all, of the metropolitan areas of the
United States. With increasing frequency,
photochemical oxidants are associated with injury
to vegetation in rural areas far removed from urban
complexes. -
Analysis of the photochemical oxidant air
pollution complex has resulted in the isolation of
three specific phytotoxic components: ozone,
nitrogen dioxide, and peroxyacylnitrates. The
peroxyacylnitrates are the most phytotoxic of the
known photochemical oxidants. A homologous
series of compounds, they include
peroxyacetylnitrate (PAN), peroxypropionylnitrate
(PPN), peroxybutyrylmtrate (PEN), peroxyiso-
butyrylnitrate (P1SOBN), and peroxybenzoylnitrate
(PBzN). The degree of phytotoxicity of these
compounds increases with the increase in their
molecular weight: PAN149.197'~199'201'202>204'212'216.
217,219,220,221,238,255,266,300,301
The limited knowledge relating to the direct
effects of photochemical oxidants on nonvascular
plants and microorganisms is also discussed in
this chapter. Although some mosses and lichens
may be useful as biological indicators of sulfur
compounds, there is no evidence indicating that
these nonvascular plants are highly sensitive to
photochemical oxidants.
VASCULAR PLANT RESPONSE TO
PHOTOCHEMICAL OXIDANTS
The effects of photochemical oxidants on
vascular plants can be envisioned as occurring at
several response levels, from the molecular to the
organismal (Figure 11-1), depending on the
concentration of pollutant, length of exposures,
and elapsed time between the exposure and the
observation of the effects. The earliest effects
include an increase in cell membrane
permeability, a decrease in carbon dioxide fixation,
253
-------
and a stimulation of stress-induced ethylene
production.113'283'294 These subtle cellular changes
are followed by both inactivation and/or activation
of specific enzymes, alteration in metabolite pools,
and modified metabolite translocation.283'286'295
Biochemical changes in individual plants are
ultimately expressed in visible foliar injury,
premature senescence and increased leaf
abscission, reduced plant vigor and growth, and
death. In the final analysis, biochemical
modifications on an individual level are manifested
by changes in plant communities and, ultimately,
in whole ecosystems.217'283 These changes
occurring in stressed systems can ultimately be
measured in socioeconomic impacts. The
sequence of topics in this chapter describing
photochemical oxidant effects on vascular plants is
based on the logical hierarchical ordering of plant
response depicted in Figure 11-1. The complexities
of the entire subject are apparent in the sections on
dose response and factors affecting plant
response.
Physiological Processes
In vascular plants, foliage is the primary receptor
of photochemical pollutants. For photochemical
oxidants to produce an effect in plants, they must
come into contact with the leaf, pass through the
stomata, and dissolve in the aqueous layer coating
the cell walls.274'283 Since stomata are the principal
entry sites for ozone and PAN into plant leaves,
stomatal closure presents a physical barrier to the
entrance of oxidants and effectively protects the
plant from injury.185 Several studies suggest that
oxidants may cause stomatal closure.144'160'180'249
In one study, stomatal closure was associated with
a genetic trait in onion wherein the stomata of
sensitive plants did not close in response to
ozone.78 Dean65 related the differences in ozone
sensitivity between two tobacco cultivars to
differences in stomatal density. Evans and Ting82
found that maximum sensitivity of bean primary
leaves was not associated with differences in
stomatal number or leaf resistance to gas transfer.
The effect of ozone and PAN on stomatal opening
depends on both environmental and genetic
factors.
Ozone can modify amino acids, proteins,
unsaturated fatty acids, and sulfhydryl residues113
located in cellular membranes. The initial effect of
oxidants is to increase the leakage of water and
ions, such as K+ (potassium ion), from cells. In
addition, ozone can induce the production of
stress-induced ethylene.1'50>283'294 The amount of
ethylene produced in response to ozone stress is
proportional to ozone concentration or duration of
CHANGES IN PLANT COMMU-
NITIES AND
ECOSYSTEMS
REDUCED PLANT GROWTH
REDUCED PLANT YIELD
ALTERED PRODUCT QUALITY
LOSS OF PLANT VIGOR
ALTERED ENZYME ACTIVITIES
ALTERED METABOLIC POOLS
ALTERED TRANSLOCATION
REDUCED PHOTOSYNTHESIS
INCREASED MEMBRANE PERMEABILITY
PRODUCTION OF STRESS ETHYLENE
BIOCHEMICAL AND CELLULAR
CHANGES
Figure 11-1. Sequence of ozone-induced responses.282
254
-------
exposure.294 Plants sensitive to ozone produce
more stress ethylene than less sensitive plants.
Ethylene production is induced before the
appearance of visible injury and frequently occurs
without visible injury.
Todd296 and Todd and Probst298 also measured
the effects of ozone at 7840 /yg/m3 (4 ppm) for 40
min on photosynthesis and found that the
development of symptoms was associated with an
inhibition of carbon dioxide fixation. This effect
was also confirmed by Macdowall,177who reported
that the inhibition of photosynthesis was greater
than that which could be accounted for by
chlorophyll destruction alone. Hill and Littlefield130
associated decreased net photosynthesis caused
by ozone (at 0.60 ppm for 1 hr)with both decreased
stomatal opening and decreased transpiration.
These studies have generally shown that net
photosynthesis can be reduced without the
association of visible injury.
The photosynthetic rate is an important indicator
of plant vigor. In.ponderosa pine, a reduction in this
rate may occur at a low dose of ozone without
causing visible symptoms. Miller et al.210 found
that a daily 9-hr exposure to ozone at 294/yg/m3
(0.1 5 ppm) reduced apparent photosynthetic rates
by 10 percent after 30 days, without typical ozone
symptoms. Botkin et al.23 found that a threshold
ozone dose for suppression of net photosynthesis
in eastern white pine was a 4-hr exposure to 980
/yg/m3 (0.50 ppm).
Ozone, in addition to inhibiting photosynthesis,
alters the way in which the products of
photosynthesis are distributed within plants.
Results of studies in which alfalfa and ponderosa
pine were exposed to low levels of ozone for a
growing season indicated that sugars were
retained in the foliage and were not translocated to
the roots.218'295 The resulting reduction of
carbohydrates in the roots could reduce nitrogen
fixation and plant growth.93'260
The vital physiological process of nitrogen
fixation may be a key factor in the impact of
photochemical oxidants on the plant
community/ecosystem complex. Ozone is known
to suppress nodulation in the roots of
soybean18'246'284 and ladino clover.2'156 Blum and
Tingey18 found no direct effect of ozone on
Rhizobium or on nodule formation. They attributed
the decrease in nodule number and nitrogen
fixation to the reduction in the available energy in
the root tissues. Studies of the chronic exposure (7
hr per day) of alfalfa to 98 /vg/m3 (0.05 ppm) 03 by
Tmgey283 and Neely et al.218 show that the total
amount of fixed nitrogen was depressed by 40
percent. Weber307 found that the effect of ozone
and a mixture of ozone and sulfur dioxide on
soybean resulted in suppression of nitrogen
fixation by 50 percent. Suppression of atmospheric
nitrogen fixation by root nodules could affect total
biomass and agricultural yield, especially in areas
of high oxidant pollution and low soil nitrogen.
Kochhar156 also reported an inhibition of plant
growth and nodulation of Trifolium repens (clover)
when plants were treated with root exudates taken
from fescue grass that had been exposed to ozone.
If these responses are of widespread occurrence,
the competitive ability of plant species could be
modified with a resulting change in plant diversity
and a possible decrease in productivity.
The acute responses of plants to ozone and PAN
result from disruption of normal cell structure and
processes. The initial response results in cellular
disorganization. Water and salts are lost from the
cell interior; then plasmolysis results and cell
death normally occurs. However, depending on
dose and environmental conditions, membrane
permeability may be restored and cell recovery
occurs. The extent of recovery depends on the
severity of the external stresses and on the ability
of the cells to repair themselves. As indicated
previously, the biochemical and physiological
effects of ozone on plants are better understood
than those of PAN.
Visible Symptoms
Visible symptoms are commonly used in
characterizing the response of vegetation to a
variety of stresses, including air pollution.301
Different types of stress induce similar symptoms,
thus causing difficulty in diagnosis. The visible
symptoms resulting from oxidant injury to plants
can be classified as belonging to two general
categories—acute and chronic injury. Acute injury
is usually manifested by cell destruction. Necrotic
symptoms of acute injury are sometimes
characteristic of a given oxidant. These patterns, at
the least, demonstrate the effect of a chemical
toxicant. Acute injury usually appears within 24 hr
after exposure and is associated with short
exposures (hours) to a specific oxidant or pollutant
mixtures at relatively high concentrations. Chronic
injury, whether mild or severe, is usually
associated with long-term or multiple exposures to
low concentrations of oxidants. Disruption of
normal cellular activity occurs, leading to chlorosis
255
-------
or other pigment or color change. Cell death may
eventually result. Chronic oxidant injury patterns
may be confused with symptoms resulting from
normal senescence and biotic pathogens,
including insects, nutritional disorders, or other
environmental stresses. These patterns may
appear as premature leaf senescence.
Bobrov,19'20'21'22 Bystrom et al.,37 and Glater et
al.96 conducted the earliest studies of microscopic
changes and developmental patterns resulting
from simulated smog and ambient pollution. These
workers first suggested that the initial effects were
due to cell membrane injury. They described
developmental patterns and showed that leaves
near maximal expansion were most sensitive and
that sensitivity was related to stomatal function,
volume of intercellular spaces, and the extent of
suberization of mesophyll cells. Other workers
have since substantiated their studies.192'276'281'286
Although ozone and PAN are considered to be
the two primary phytotoxic oxidants in the
photochemical complex, the response of plants to
simulated polluted atmospheric conditions
suggests the existence of other ambient phytotoxic
oxidants.96 The symptoms associated with many of
these reactant mixtures are closely related to those
caused by ozone and PAN.115'118 In some tests,
however, the reactant mixtures used would not
have produced either ozone or PAN. In other cases,
the pattern of injury on sensitive test plants or age
of tissue affected suggested the presence of one or
more pollutants other than ozone or PAN Plant
injury symptoms observed in the field often
resemble those reported from controlled
exposures to ozone or PAN, but the response
pattern in some cases is sufficiently different that
accurate fie Id diagnosis is difficult. Brennanet a I.30
correlated development of oxidant symptoms with
aldehyde concentrations in New Jersey and
suggested that aldehyde may be a major phytotoxic
component of the photochemical oxidant complex.
The symptoms observed were probably not in
response to the aldehydes, but rather to a
compound or group of compounds present under
the same conditions as the aldehyde 132 The
concentration of the compound or group of
compounds was probably directly related to the
concentration of the aldehydes
OZONE
The four general types of lesions found on plant
leaves as a result of ozone injury are described in
detail in Recognition of Air Pollution Injury to
Vegetation: A Pictorial Atlas.'29 These are
pigmentation (stippling), bleaching, foliar
chlorosis, and bifacial necrosis. All except the last
type of symptom occur primarily on the upper
surface of the leaf.
1. Pigmentation (stippling). The palisade1
mesophyll cells of injured leaves take on
brown, black, red, or purple colorations.
The injury is evident primarily on the upper
leaf surface. On many deciduous trees,
shrubs, and some herbaceous plants, the
accumulation of pigments in the dying cells
results in small dot-like lesions.
2. Bleaching (fleck). Bleachedor unpigmented
lesions usually occur on upper leaf
surfaces following collapse of the palisade
cells and sometimes of the epidermal cells.
Individual lesions are usually small and
irregular in shape but may coalesce and
become large, frequently resulting in
slightly sunken areas on the upper leaf
surface in certain herbaceous species.
3. Foliar chlorosis. Chlorotic areas usually
occur on the upper surfaces of leaves, often
coalescing to present a mottled
appearance. Pine needles and grasses that
do not possess differentiated mesophyll
tissue may develop the chlorotic mottling
on either leaf surface.
4. Bifacial necrosis. Bifacial necrosis results
when the tissues connecting the upper and
lower leaf surface are killed. The coloration
of the tissue in the lesions may range from
ivory to orange or brownish-red. Upper and
lower surfaces often are drawn together,
forming a thin, papery lesion.
The two classic symptoms (fleck and stippling)
are more widely associated with the response of
dicotyledonous plants to ozone than are other
symptoms. Many plants (e.g., pinto bean,
cucumber, tomato, soybean, and sycamore) may
have the entire upper surface covered with a
bleached appearance as a result of ozone
exposure, with no observable injury on the lower
surface. On closer examination, the bleached area
is seen to be made up of many small groups of
palisade cells that are dead and contain no
pigment. In other plants or under different
conditions, the palisade cells may accumulatedark
alkaloid pigments (stipple) coincidentally with cell
death After exposure to higher ozone
concentrations or after longer periods of exposure,
injury extends to the spongy cells, producing
256
-------
bifacial necrosis. Plants exposed to a high
concentration of ozone or to a high concentration
of ambient pollution during a pollution episode
usually develop dark, watersoaked areas in the leaf.
within a few hours. Leaves may show partial
recovery, or these areas may form light-tan bifacial
necrotic lesions within 24 to 48 hr. Individual
lesions may be small, but groups of them can
extend and affect a considerable portion of the leaf.
In monocotyledonous plants (grasses or cereals)
and in some nonmonocotyledonous plants there is
no division of mesophyll tissue, and injury usually
appears as a bifacial fleck.21
The foregoing discussion is not descriptive of
•effects noted in coniferous trees such as white
pine. Two classic oxidant (ozone) syndromes of
pine, one in the eastern and the other in the
western United States, have been described.
Ozone is probably the cause of emergence tipburn
in white pine (white pine needle dieback).15 The
injury is characterized as a tip dieback of newly
elongating needles and occurs throughout the
range of eastern white pine. Affected trees are
found at random in a stand, and symptoms develop
in discrete episodes in successive years.217
Primar-' roots of affected trees often die after
repeat d needle injury. Costonis and Sinclair48
reported- silvery or chlorotic flecks, chlorotic
mottlir g, and tip necrosis of needles as a result of
ozone exposure.
Ohio-otic decline, a needle injury of ponderosa
pine, v ras first noticed in 1953 and was related to
oxidant air pollution by 1961,234 Chlorotic decline
was characterized by a progressive reduction in
termir al and diameter growth of the tree; only the
currer t season's needles were retained. Yellow
mottling and a reduction in the number and size of
the remaining needles were noted. Eventually the
tree died. The chlorotic decline was not associated
with stresses other than ozone.235
PAN
The injury symptoms for the various peroxyacyl
compounds are similar and cannot be
distinguished. The characteristic symptoms are
glazing, silvering, or bronzing of the lower surface
of the leaves of plants such as spinach, garden
beets, romaine lettuce, and chard. Symptoms on
endive and turnip appear as a bleaching of the
lower leaf surface that often develops into light tan
necrotic areas. In some instances, upper leaf
surfaces may also be affected.216'217'238'272
Microscopic examination of affected areas
reveals a collapse of spongy mesophyll cells with a
subsequent development of large intercellular air
pockets, especially near the stomata. The air
pockets give the leaves the glazed appearance.
217,238
212,
Young expanding leaves are normally more
sensitive to PAN than mature leaves. Leaves of
plants such as grasses, tobacco, and petunia,
which do not mature uniformly, may show leaf
banding where injury is related to a sensitive
region of the expanding leaf. Successive exposure
to PAN may result in the formation of multiple
bands.20'238
PAN symptoms are slowtodeve'~>pand may take
up to 72 hr for full development. A complete
description of PAN injury to plants was written by
Taylor and MacLean.272
Growth and Yield
Photochemical oxidants, including ozone and
PAN, suppress growth and yield in many plant
species. Although growth and yield effects are
normally associated with visible injury, oxidants
can inhibit growth with little or no injury. The basis
for this variability is not understood, but probably it
is associated with the genetic composition of
plants, environmental condition, and pollutant
combination and dose.
Attempts have been made to associate visible
injury with other indices of plant response. Todd
and Arnold297 compared visible injury with
biomass production and chlorophyll content of
pinto bean exposed to a synthetic oxidant (ozone
plus hexene). They noted a logarithmic
relationship between injury and the two response
parameters and suggested that leaf injury was not
a reliable index for estimating growth losses. A
group of Canadian workers3'7 used physiological
measurements exclusively to monitor the effects
of acute ozone (oxidant) exposures. Because of the
subjective nature of injury measurements and the
often apparent poor correlation with growth
responses, growth or physiological parameters are
often evaluated directly.
AMBIENT AIR STUDIES
Since the first observations of oxidant injury by
Middleton et al.,203 there has been a continuing
flow of reports mentioning the effects of oxidants
on vegetation. The majority of these reports
associate ozone or oxidant injury symptoms with
episodes of given levels of oxidant or ozone. Engle
et al.80 in Wisconsin, Laurence et al.167 and Kohut
et al.159 in Minnesota, Gardner91 in South Dakota,
Weaver and Jackson306 and Haas101 in Ontario,
257
-------
Rich et al.247 in Connecticut, Daines et al.54 in New
Jersey, Brasher et al.25 in Delaware, Skellyetal.259
in Virginia, Reinert et al.243 in Ohio, Tingey and
Hill289 in Utah, and Oshima et al.,232 Cobb and
Stark,43 and Miller and Millecan209 in California all
noted plant effects from ambient oxidant. Haas101
found that the rate of growth during oxidant
exposure influenced the symptom severity, that
the stage of growth determined the dose required
to produce injury effects, and that the physical
conditions within a field could cause variable plant
responses to oxidant.
Cobb and Stark43 and Miller and Millecan209
reported the severe effects of ambient oxidant on
ponderosa pine in the San Bernardino Mountains.
The results of quantitative studies are reported in
Chapter 12. Skelly et al.259 associated post-
emergence tipburn of eastern white pine with high
levels of ambient oxidant.
Greenhouses with filtered and nonfiltered air or
field chamber studies have been used to determine
the effects of ambient pollutants on growth and
yield of selected crops. Reductions in yield and
biomass were some of the observed effects noted.
Summary data from selected studies are shown in
Table 11-1. Thompson and Taylor280 summarized
several years of detailed field chamber studies,
during which lemon and orange trees were
exposed to many pollutant combinations. The
inhibiting effects of the pollutant on fruit size and
number and on total yields were noted in these
studies. The results of these studies provided the
basis for the projected 50-percent reductions in
citrus yields from the Los Angeles Basin as a result
of photochemical oxidants. Thompson and Katz279
showed that some of the effects noted on the citrus
trees were probably related to concentrations of
PAN in the ambient air. Thompson and
coworkers,277'278 using charcoal-filtered and
unfiltered chambers to determine effects of
ambient oxidants on grapes, found nominal yield
effects in the first year and 50 to 60 percent
reduction during the next 2 years. The year to year
differences were attributed to the effects of
oxidants on floral initiation the year before the first
year study. A similar response may be expectedfor
plants that initiate floral structures in the season
previous to experimental treatments. A wax
emulsion spray on the leaves gave a 20-percent
increase in yield over unsprayed plants,
suggesting that selected protectants may be useful
in protecting grapes grown in areas of high oxidant
concentration.
TABLE 11-1. EFFECTS OF OXIDANTS (OZONE) IN AMBIENT AIR ON GROWTH, YIELD, AND
FOLIAR INJURY IN SELECTED PLANTS3
Plant
species
Lemon
Orange
Oxidant concen-
tration, ppm
>0 10
>0 10
Duration of
exposure
Over growing season
148 hr/month avg from March-
Oct , 254 hr/month avg from
Plant response
((eduction from control
listed as °o)
32, yield
52, yield (leaf drop and other effects)
54, yield
(other reductions found)
Reference
280
280
Grape, cultivar >0 25
Corn, sweet
Bean, white
020-035
>008
Tobacco, cultivar 002-003
Bel W3
Tobacco, cultivar >0.05
Bel W3
Cotton, cultivar Ambient
Acula
Potato, 4 >0 05
cultivarsb
Potato, cultivar 0 15
Haig
July-Sept
Often over May-September
growing season
Hourly maximum for 3 to 4 days
before injury
9 hr
6 to 8 hr
Often over growing
season
Over growing season
326 to 533 hr (2 years)
3 consecutive days
12, yield (first year) 277
61, yield (second year) 278
(increased sugar content)
47, yield (third year) 275
67, injury (10 cultivars, 5 unmarketable) 38
1 8, injury (13 cultivars)
1, injury (1 1 cultivars)
(Bronze color, necrotic stipple, premature 306
abscission)
(Minimal injury) 11 9
22 (fresh wt), top 108
27 (fresh wt), root
7-20, lint + seed (3 locations, 1 972) 33
5-29, lint + seed (3 locations, 1973)
34-50, yield (2 years for 2 cultivars) 1 22
20-26, yield ( 1 year for 2 cultivars)
95, injury (leaf area covered) 25
'Table taken from Ref 217
bGreenhouse studies
258
-------
Other experimental studies noted the effects of
ambient air levels of oxidant on cotton, potato, and
tobacco yields. In the San Joaquin Valley,
California, a 5- to 29-percent reduction in lint and
seed was noted over a 2-year period when cotton
was grown in charcoal-filtered vs unftltered field
chambers,33 Heggestad122 reported the results of a
3-year study comparing the growth of four potato
cultivars in greenhouses with charcoal-filtered
and unfiltered ambient air. Yields of sensitive
cultivars were suppressed as much as 50 percent
by ambient oxidants. Heagle etal.,108 using open-
top field chambers, reported preliminary results
indicating a reduction in leaf yields of a sensitive
tobacco cultivar when plants grown in ambient air
and nonfiltered chambers were compared with
plants grown in chambers with charcoal filters,
The results of the research discussed in this
section indicate that oxidants in the ambient air
decrease plant growth and yield. Injury symptoms
resulting from oxidant exposure have been noted
on a nationwide basis. The most severe effects are
found in California in the San Bernardino
Mountains and the Los Angeles Basin,
CONTROLLED CHAMBER STUDIES
To define the effects of oxidants on plant growth
and yields more accurately, many workers have
used controlled additions of ozone to determine the
effects of acute or chronic exposures on a variety of
growth parameters. Although the results of the
effects of ambient oxidants on vegetation may not
be directly comparable with the results obtained in
the controlled exposure studies discussed in the
following sections, reduced yields and similar
oxidant symptoms are common results. Most of the
results discussed are from exposures m
greenhouses and controlled environmental
chambers, but several are from studies using
exposure chambers placed over field plantings
Short-Term Exposures — Adedipe and
associates3 reported reduced biomass and floral
production in four bedding plants and reduced
biomass in two radish cultivars7 from acute ozone
exposures (Table 11 -2). They reported no effects
on marigold, celosia, impatiens, and salvia
cultivars, even at the high concentration of 0.80
ppm for a 2 hrexposure. Research involving radish
has included the effects of exposure temperatures
on the growth response to ozone.7 One cultivar
reacted to ozone in the same way regardless of
temperature variation, but the response of the
second cultivar was influenced by temperature.
Tmgey et al.285 exposed radish at 7, 14, or 21 days
of age and all combinations of these ages to ozone
at 785 pg/'m3 (0.40 ppm) for 1.5 hr. For a single
exposure, the greatest effect on root growth was
noted on the 14-day-old cultivar; for double
exposures, the greatest root growth effect
occurred with the 7- and 14-day-old cultivars.
However, the greatest reduction of root growth
occurred with triple exposures of 7-, 14-, and 21 -
day-old cultivars. The reductions in root growth
from the multiple ozone exposures were equal to
the additive effects of three single exposures. It
was concluded that root growth reductions
resulted from the preferential use of
photosynthate for foliar growth. Inhibition in the
rate of root growth remained during the second
week after exposure. Tmgey and Blum284 reported
that when soybean plants were exposed to 1468
#g/m3 (0.75 ppm) of ozone for 1 hr, root growth
was consistently reduced more than top growth
and, in addition, there was a reduction in nodule
weight and number. Evans81 exposed the middle
leaflet of the first trifoliate leaf of pinto bean to
various ozone concentrations. The results
indicated a differential growth response in various
leaf positions. A reduction in leaflet expansion was
noted after a 12-hr treatment with 98pg/m3 (0.05
ppm) ozone. The foregoing discussion represents
only a few of the representative examples
regarding the effects of acute ozone exposure on
plant growth.
Long-Term Exposures — Experimental long-term
exposures of various crops as well as ornamental
and native plants to ozone have resulted in a
reduction in growth and/or yield. Harward and
Treshow105 exposed 14 species representative of
the aspen plant community to ambient air
containing ozone concentrations of 98 to 137
fiQ/m3 (0.05 to 0.07 ppm) and to ozone at 290 and
588 ,ug/m3 (0 15 and 0 30 ppm) for 3 hr/day, 5
days/week, andtocharcoal-filteredairthroughout
the growing season, (nail species, foliar injury was
seen at the highest pollution concentration (Table
11-3). There was considerable variability in the
responses, and only six species formed seed,
however, tn most cases, growth was reduced and
most species were sensitive Price and Treshow239
found major biomass reductions in six grass and
two tree species exposed for 4 hr/day to ozone at
290 to 647 /jg/'m3 (0,15 to 0.33 ppm) over a
growing season. They also found reduced
reproduction and a loss of some reproductive
components. These effects on growth and
259
-------
reproduction could result in subtle shifts in
community composition after several years of
ozone exposure. Taylor et al.271 reported a 52-
percent reduction in the fresh weight of avocado
seedlings exposed to a synthetic smog (ozone plus
hexene) for 280 hr. Tingey et al.293 found that the
growth of two soybean cultivars (Hood and Dare)
was inhibited by intermittent exposure to ozone at
196 /jg/m3 (0.10 ppm)for 3 weeks. A decrease in
both root and top growth occurred. Similar
results287 were noted with radish, except that
growth was inhibited at 98 /ug/m3 (0.05 ppm). The
reduced growth noted in the aforementioned
studies occurred even though there were very few
visible symptoms of plant injury.
A 30-percent reduction in wheat yield occurred
when at anthesis, wheat was exposed to ozone at
392 /xg/m3 (0.2 ppm) 4 hr/day for 7 days.156
Oshima et al.231 reported a reduction in tomato
yield at 686 /xg/m3 (0.35 ppm) over an exposure
period of 97 hr. Significant injury occurred at both
392 and 686 /xg/m3 (0.20 and 0.35 ppm) ozone.
Plants exposed to the low ozone concentration
tolerated a considerable amount of defoliation and
a significant decrease in biomass without a cor-
responding reduction in yield. A significant re-
duction in yield, however, was recorded from
plants in the higher (686 /xg/m3, or 0.35 ppm)
ozone exposure where a greater decrease in bio-
mass was observed. This yield reduction was
caused by a lower fruit set and subsequently fewer
harvested fruit. Oshima suggested an injury
tolerance threshold for tomato below which no
reduction in yield would occur In an eariler study,
Oshima229 reported a decrease in kernel weight of
sweet corn exposed to ozone for 4 percent of the
growing period at concentrations of 392 or 686
(0 20 or 0.35 ppm). The reduced weight was
TABLE 11-2. EFFECTS OF ACUTE EXPOSURE ON GROWTH AND YIELD OF SELECTED PLANTS3
Plant
species
Begonia, cultivar
White Tausendschon
Petunia, cultivar
Capri
Coleus, cultivar
Scarlet Rainbow
Snapdragon, cultivar
Floral Carpet, mixture
Radish, cultivar
Cavalier, Cherry Belle
Radish
Cucumber, cultivar
Ohio Mosaic
Potato, cultivar
Norland
Tomato, cuitivar
Fireball
Tomato, cultivar
Fireball
Onion, cultivar
Spartan Era
Tobacco, cultivar
Bel W3
Ozone concen -
tration, ppm
0 10
020
040
080
0 10
020
040
080
0 10
020
040
080
0 10
020
040
080
025
040
1 00
1.00
1 00
1 00
050
1 00
050
1 00
020
1 00
1 00
030
Exposure
time, hr
2
2
2
2
2
2
2
2
2
2'
2
2
2
2
2
2
3
1 5(1 )c
1 5(2)c
1 5(3)c
1
4
4
4(3)c
1
1
1
1
24
1
4
2
Plant response
(reduction from control
listed as %)"
5, avg of 3 growth responses shoot wt, flower wt,
flower no
10, avg of same responses
19, avg of same responses
38, avg of same responses
9, avg of same responses
1 1 , avg of same responses
21, avg of same responses
31, avg of same responses
2, avg of same responses
17, avg of same responses
24, avg of same responses
39, avg of same responses
0, avg of same responses
6, avg of same responses
8, avg of same responses
16, avg of same responses
36, top dry wt (Cavalier)
38, root dry wt (Cherry Belle)
37, root dry wt
63, root dry wt
75, root dry wt
19, top dry wt (1 % injury)
37, top dry wt (18% injury)
0, tuber dry wt (no injury)
30, tuber dry wt (injury severe)
15, plant dry wt (grown in moist soil)
20, plant dry wt (grown in moist soil)
15, increase in plant dry wt (grown in dry soil)
25, increase in plant dry wt (grown in dry soil)
0, effect
19, plant dry wt (no injury)
49, plant dry wt
48, chlorophyll content
Reference
3
3
3
3
7
285
226
226
153
153
226
4
aTaken from Ref 21 7
"Unless otherwise noted
'Number of exposures in parentheses
260
-------
associated with a shriveled-ear condition (kernels)
that might be related to the effects of ozone on
pollen development.
Craker49 reported a reduction in the weight of
petunia flowers after plants were exposed to ozone
for 53 days at 98 to 137 /ug/m3 (0.05 to 0.07 ppm);
however, an increase in the weight of petunia
flower82 was found when exposure to three
different concentrations of ozone was for 7 days.
Carnations (80 plants) continuously exposed for 38
days to 98 to 196 //g/m3 (0.05 to 0.10 pprn)
produced a single deformed flower, while the
controls had 24 normal flowers.84 Poinsettia bract
area was decreased by 39 percent after a 50-day
exposure (6 hr/day) to ozone at 196 to 235/ug/m3
(0,10 to 0.12 ppm).52
Heagle and associates110 found a reduction in
yield of sweet corn and soybean109 after exposure
to ozone at 196 /ug/m3 (0.10 ppm) for 6 hr/day
administered over much of the growing season.
These exposures were carried out in field
chambers set up over soybean and corn plots in the
field, They suggest that a threshold for measurable
effects on these crops would lie between ozone
(oxidant) "concentrations of 98 and 196 /ug/m3
(0.05 and 0,10 ppmjfor 6 hr/day if exposures were
continued throughout most of the growing season,
These concentration values resemble growing-
season averages found in the eastern United
States.
In one of the few examples of chronic ozone
studies involving forest tree species, Wilhour and
Neely310 exposed eight species of conifers to 200
/ug/m3 (0.10 ppm) of Oa for 6 hr/day, 7 days/week,
for 18 continuous weeks. Significant growth
reduction (given as percent reduction from control)
was observed in root length (12 percent), dry
weight (DW) of stem (21 percent), and DW of root
(26 percent) of ponderosa pine (Pinus ponderosa
Laws), and in western white pine (P. monticola
Dougl.) in DW of foliage (13 percent) and DW of
stem (9 percent). Biweekly harvests of ponderosa
pine (exposed as above) throughout a 20-week
study period showed highly significant growth
reductions in root DW (26 percent). A trend of
increased growth differences associated with
increased length of exposure to ozone was evident.
Bennett et al.12 found experimental evidence for
stimulation of growth at low concentrations of
ozone. They exposed bean (cultivar Pure Gold
Wax), barley (cultivar Brock), and smartweed to
ozone at 59 /ug/m3 (0.03 ppm) and found instances
of significant growth increases. The concept needs
further_study in light of the current concern over
"normal" background ozone concentrations,
A number of significant investigations have
been conducted that demonstrate the response of
vegetation to long-term exposure of relatively low
ozone concentrations. The results of these chronic
ozone studies are summarized in Table 11 -4. The
results of such experiments suggest that overall
effects on agriculture production could be
extensive, depending on the sensitivity of the
cultivars used.
To illustrate the relationship between ozone
concentration, duration of exposure, and effects on
foliage, growth, or yield, the data from Table 11-4
have been graphed (Figure 11 -2). Each point on the
graph represents a combination of concentration
and time at which a significant reduction in
growth, yield, or photosynthesis occurred. The
dashed line is an indicator of the level below which
TABLE 11-3. EFFECTS OF OZONE ON SELECTED UNDERSTORY SPECIES FROM AN ASPEN COMMUNITY"
Plant
species
Chenopodtum album L
C fremontii L
Descuratnta sp
Geranium fremonm, L Torr
ex A Gray
Lepiclium virginieum L
Madia glomerata
Polygonum av/culare L
P. douglasn Greene
Plant response
Foliar injury, % of control
Ambient
air"
0
10
0
7
50
40
30
5
290 09 'm'
(01 5 ppm)
35
35
15
50
100
100
95
35
58B pg/nv'
(0 30 ppmi
40
90
55
90
100
100
95
95
at different ozone concentration"*
Plant wl, % of control
Ambient
air1
103
71
179C
93
121
112
271
87
290 fig/m^
(0 15 ppm)
71£
98
77
94
72
38C
186C
29°
588 09/m-
10 30 ppm|
83
32
56C
53°
31°
13C
29o
5'
Seed wi. % of control
Ambient
air"
87
102
94
79
127
93
290 //g/m3
\Q 1 5 ppm|
99
96
69
74
43
84
588 £/g/ mj
(0 30 ppm|
87
94
47*
18
50
3C
'Table from Re! 217
"Exposures were 3 hr/day. 5 days/week through growing season
^Significant eflect
"Ambient an = 98 to 13? /Jg/m3 (005 to 0 07 ppm)
261
-------
TABLE 11-4. EFFECTS OF LONG-TERM, CONTROLLED OZONE EXPOSURES ON GROWTH, YIELD,
AND FOLIAR INJURY TO SELECTED PLANTS8
Plant
species
Lemna, duckweed
Carnation
Geranium
Petunia
Pomsettia
Radish
Beet, garden
Bean, cultivar
Pinto
Bean, cultivar
Pinto
Bean, cultivar
Pinto
Bean, cultivar
Pinto
Tomato
Corn, sweet.
cultivar Golden
Jubilee
Wheat, cultivar
Arthur 71
Soybean
Soybean
Alfalfa
Grass, brome
AlfaUac
Fig i'-2
Nos
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Ozone concentration.
f/g/mMppm)
1 96 (0 1 0)
98-177 (005-009)
137-196(007-010)
98-137 (005-007)
196-235(010-0 12)
98(005)
98 (0 05)
392 (0 20)
255 (0 13)
290 (0 1 5)
490 (0 25)
686 (0 35)
290 (0 15)
290(0 15)
290(0 15)
290 (0 15)
440(0225)
440 (0 225)
588 (0 30)
588 (0 30)
392 (0 20)
686 (0 35)
392 (0 20)
686(0.35)
392 (0 20)
98 (005)
196(010)
196(0 10)
290(0 15)
390 (0 20)
290-647
(0 15-033) (varied)
1 96 (0 1 0)
Exposure time, hr
5/day, 1 4 days
24/day. 90 days
9 5/day, 90 days
24/day, 53 days
6/day, 5 days/week.
10 weeks
8/day, 5 days/week,
5 weeks
8/day, 5 days/week
(mixture of O3 and SO2
for same periods)
3/day, 38 days
8/day, 28 days
2/day, 63 days
2/day, 63 days
2/day, 63 days
2/day, 14 days
3/day, 14 days
4/day, 14 days
6/day, 14 days
2/day, 14 days
4/day, 14 days
1/day, 14 days
3/day, 14 days
2 5/day, 3 days/week
14 weeks
2 5/day, 3 days/week,
1 4 weeks
3/day, 3 days/week
till harvest
3/day, 3 days/week
till harvest
4/day, 7 days
(anthesis)
8/day, 5 days/week
3 weeks
8/day, 5 days/week
(mixture of Oa and SOj
for same periods)
8/day, 5 days/week
3 weeks
2/day, 21 days
2/day, 21 days
2/day, 21 days
4/day, 5 days/week
growing season
6/day, 70 days
Plant response, percent
reduction frnm control
100, flowering, 36, flowering
(1 wk after exposure completed)
50, frond doubling rate
50, flowering (reduced vegetative
growth)
50, flowering (shorter flower
lasting time, reduced vegetative
growth!
30, flower fresh wt
39, bract size
54, root fresh wt
20, leaf fresh wt
63, root fresh wt
22, leaf fresh wt
50, top dry wt
79, top fresh wt
73, root fresh wt
70, height
33, plant dry wt, 46, pod fresh wt
95, plant dry wt, 99, pod fresh wt
97, plant dry wt, 100, pod fresh wt
8, leaf dry wt
8, leaf dry wt
23, leaf dry wt (Data available on
whole plants, roots.
leaves, injury, and
49, leaf dry wt 3 levels of soil
moisture stress)
44, leaf dry wt
68, leaf dry wt (Data available on
whole plants, roots,
40, leaf dry wt leaves, injury, and
3 levels of soil
76, leaf dry wt moisture stress)
1, yield, 32, top dry wt, 11,
root dry wt
45, yield, 72, top dry wt, 59,
root dry wt
1 3, kernel dry wl, 20, top dry wt,
24, root dry wt
20, kernel dry wt, 48, top dry wt.
54, root dry wt
30, yield
13, foliar injury
16, foliar injury
20, root dry wt
21, top dry wt
9, root dry wt
16, top dry wt
26, top dry wt
39, top dry wt
83, biomass
4, top dry wt, harvest 1
Reference
86
83
83
49
52
287
223
182
139
175
175
231
229
156
293
138
239
218
Alfalfa0
32 98(005)
7/day, 68 days
20, top dry wt, harvest 2
50, top dry wt, harvest 3
30, top dry wt, harvest 1
50, top dry wt, harvest 2
218
262
-------
TABLE 11-4. EFFECTS OF LONG-TERM, CONTROLLED OZONE EXPOSURES ON GROWTH, YIELD,
AND FOLIAR INJURY TO SELECTED PLANTS' (cont'dl.
Plant
species
Alfalfa
Pine, eastern
white
Pine, ponderosa
Pine, ponderosa
Poplar, yellow
Maple, silver
Ash, white
Sycamore
Maple, sugar
Corn, sweet,
cultivar Golden
Midget0
Pine, ponderosa0
Pine, western
white0
Soybean, cultivar
Darec
Poplar, hybrid
Fig 11-2
Nos"
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
Ozone concentration.
pg/mMppm!
98 (0 05)
196 (0.10)
290(015)
290 !0 1 5)
290(0.15)
290(0 15)
588 (0 30)
588 (0 30)
588 (0.30)
880 (0 45)
588(030)
588 (0.30)
588 (0 30)
588 (0 30)
588 (0 30)
98 (0.05)
1 96 (0 1 0)
1 96 (0 1 0)
1 96 <0 1 0)
98(005)
196(0 10)
290 (0 15)
Exposure time hr
8/day, 5 days/week
1 2 weeks
4/day, 5 days/week
4 weeks (mixture of Oa
and SOj for same periods)
9/day, 10 days
9/day, 20 days
9/day, 30 days
9/day, 60 days
9/day, 10 days
9/day, 20 days
9/day, 30 days
9/day, 30 days
8/day, 5 days/week
13 weeks
8/day, 5 days/week.
1 3 weeks
8/day, 5 days/week
13 weeks
8/day, 5 days/week
13 weeks
8/day, 5 days/week
13 weeks
6/day, 64 days
6/day, 64 days
6/day, 126 days
6/day, 126 days
6/day, 133 days
6/day, 133 days
8/day, 5 days/week
6 weeks
Plant response, percent
reduction from control
1 8, top dry wt
3, needle mottle
(over 2-3 days of exposure)
16, needle mottle
4, photosynthesis
25, photosynthesis
25, photosynthesis
34, photosynthesis
12, photosynthesis
50, photosynthesis
72, photosynthesis
85, photosynthesis
82, leaf drop, O, height
50, leaf drop, 78, height
66, leaf drop; O, height
O, leaf drop; 22, height
28, leaf drop; 64, height
9, kernel dry wt, 14, injury
(12, avg 4 yield responses)
45, 25, 35 for same responses
12, root length
21, stem dry wt, 26, root dry wt
13, foliage dry wt
9, stem dry wt
3, seed yield, 22, plant fresh wt.
1 9, injury, defoliation, no reduction
in growth or yield
55, 65, 37 for same responses
50, shoot dry wt, 56, leaf dry wt.
47, root dry wt
Reference
290
69
210
210
147
147
147
147
147
110
310
310
109
148
aModified from Ref 217
""Numbers in this column are keyed to numbers in Fig 11-2
^Studies conducted under field conditions, eKCEptthat plants were enclosed to ensure controlled pollutant doses Pi ants grown under conditions making them more sensitive
significant effects probably do not occur. In
general, as the length of exposure increases up to
15 days, the average ozone concentration must be
decreased to prevent growth effects. Beyond 1 5
days, if the average concentration of ozone
exceeds 98 /ug/m3 (0.05 ppm), significant growth
and yield effects could occur.
Quality — A limited number of studies have
investigated the effects of ozone on crop quality.
Thompson and his co-workers z77-278 compared the
quality of grape juice made from grapes harvested
from plants grown for two seasons in either
ambient air oxidants or in carbon-filtered air.
Sugar content of the grapes grown in ambient air
was reduced 13 to 17 percent, but the
concentration of organic acids showed no
detectable trends. Frey89 reported a 5-percent
decrease in lipids and a 21-percent increase in
amino acids in seeds from soybeans exposed to
ozone. Exposure of alfalfa to 98 jug/m3 (0.05 ppm)
or to 196 fjg/m3 (0.10 ppm) ozone increased the
levels of protein and amino acids, had no effect on
carbohydrates or lignin, and decreased ft carotene
per unit weight of tissue.218 The protein increase
per unit weight of tissue is of no practical
significance, because the ozone exposure
produced a reduction in growth that resulted in a
decrease in total production of protein.
DOSE-RESPONSE RELATIONSHIPS
An understanding of dose-response
relationships is important for a basic
understanding of the mechanism of oxidant effects
on plants. Ideally this relationship could be
263
-------
expressed by a set of standard equations that
would relate plant response to pollutant
concentration and duration of exposure and would
also incorporate the effects of all other factors that
control the responses of plants. Development of
such equations requires a data base sufficient to
relate a given dose (concentration of pollutant
times duration of exposure) of oxidant (e.g., ozone
or PAN) to some meaningful plant effect. Such
equations are not yet available; however,
discussions of the relationship of time,
concentration, and response have been
published.
117,174
Limiting Values — The concept of limiting values
may be used to define the boundary between doses
of a pollutant that are likely to cause measurable
injury to an organism and those that are not likely
to cause injury.145 Jacobson146 used this concept to
estimate limiting values for the exposure of trees,
shrubs, and agricultural crops to ozone and PAN. In
estimating limiting values, several factors were
taken into consideration. These were: (1) Of the
known effects on vegetation, which were the most
I.Oi-
important? (2) What kinds of data should be
included—data from the field, where ambient
pollutant concentrations were monitored, or from
the results of experimental fumigations with
controlled dosages of the pollutant? (3) Which
model should be used to estimate limiting values?
Alteration of biochemical or physiological
processes, reductions in growth or yield, or foliar
lesions are effects on vegetation that can be
considered. However, most of the published data
discussing dose-response relationships use visible
foliar injury as the index of response.
Consequently, Jacobson145 in estimating limiting
values, used the dose-response data for visible
foliar injury on the assumption that prevention of
foliar injury would also prevent other adverse
effects. In addition, indenving limiting values, only
data from experimental investigations in which
both concentration and duration of exposures
were monitored or controlled were used. More
than 200 studies were surveyed to obtain the data.
The concentration-time model developed by
Jacobson for summarizing the dose-response
relationships for effects of oxidants on vegetation
a
a
0.1
Z
LU
U
z
o
U
LU
O
N
O
0.01
47 •
21D 11D
44* 19CK18 «45 «46 «48~52
\24 16D017 031 7DD20
\ 15 30 59
40* O»O14fD •
• 8
,10
C
\12, 13 41
\ 1C»26£D29
\
\
V-. 6 — 57
• 38»«»34 •-
53
Z5»
5« 33
54 ••55,56 ••58
3«
4*
EXPOSURE, hr/day
* < 1.99
O 2 TO 3.99
O 4 TO 5.99
• ^ B
NOS. = REF. NOS. ON TABLE 11-4
8 10 20 40 60 80 100
EXPOSURE PERIOD, days
200
400
Figure 11-2. Relation between ozone concentration, foliar injury, and a reduction in plant growth or yield (see Table
11-4).
264
-------
is similar to the one used by McCune189 for
fluorides and by Larsen and Heck166 for ozone and
SO2. In the model, foliar injury is expressed as a
function of the concentration of ozone or PAN and
the logarithm of the duration of exposure.
Limiting Values for the Effects of Ozone — A
single coordinate point for each published study
was used in developing the limiting values in
Figures 11 -3 and 11 -4. The single point represents
the lowest concentration and duration of exposure
in the published study that resulted in foliar
symptoms. The data points used to develop the
limiting values that will protect vegetation against
oxidant-induced foliar injury are not necessarily
threshold values. Since these points are based on
available published research data (more than 100
studies), the shortcomings that were present in the
experiments will be reflected in the derived points
(e.g., shortcomings in calibration and
measurement techniques and the use of
subjective methods to determine foliar injury).
Woody Trees and Shrubs, The limiting values for
the effects of ozone on trees and shrubs (Figure 11 -
3) are based on a single coordinate corresponding
to the lowest concentration of ozone and duration
of exposure that produced foliar injury on woody
plants. The data used to establish the coordinates
l 1 o * _l- 13,14,17,47,56,57,60,126,131,152,158,
came from 18 studies.
161,168,171,211,250,253,311
3 depicts the range of uncertainty and refers to the
range, as given in the reports, of the lowest doses
of ozone observed to cause foliar injury. Doses of
ozone above and to the right of the shaded area are
likely to cause foliar injury to susceptible woody
plant species. There is little likelihoodthat doses of
ozone below and to the left of the shaded area will
cause foliar injury.
Agricultural Crops. More data (over 100 studies) on
dose-response relationships for agricultural crops
are available than for any other class of vegetation.
A wide range of uncertainty exists for the limiting
values as given in Figure 11-4 because of the
variability in plant response. In addition, the
limiting values (shaded area. Figure 1 1 -4) overlap
the range of ozone concentrations producing 5
percent injury to crops as reported by Heck and
Tingey121 (Figure 1 1 -5, Tables 1 1 -5 and 1 1 -6), and
the response threshold for ozone effects on plants,
published by the National Research Council of
Canada,238 corresponds closely to the lower curve
in Figure 1 1 -4. Doses of ozone above and to the
right of the shaded area in Figure 1 1 -4 will prob-
ably cause foliar injury in susceptible agricultural
0.4
a
a
UJ
O
O
o
UJ
z
o
N
O
0.2
T
T
I I I I
RANGE OF UNCERTAINTY FOR
SUSCEPTIBLE SPECIES
LIMITING VALUES FOR
TREES AND SHRUBS
I
I I I I
800
400
0.1
0.5 1
DURATION OF EXPOSURE, hours
z
UJ
O
o
o
UJ
O
N
O
Figure 11-3. Limiting values for foliar injury to trees and
shrubs by ozone.145
265
-------
crops Doses below and to the left of the shaded
area are unlikely to cause foliar injury. The dose-
response data used in establishing the limiting
values in Figure 11-4 were obtained from many
sources98'117174'238 (see also author citations, Table
11-2) The data available at present are not
sufficient to determine if the curves in Figure 11-4
become parallel to the horizontal axis or, if at
durations of 8 hr or more, they approach the axis
asymptotically. The inaccuracies in measurements
make the interpretation of results of repeated or
long-duration exposures difficult, and limiting
values of ozone concentrations below 100/^g/m3
(005 ppm) are not useful.145
Limiting Values for the Effects of PAN — The
limiting values for PAN are represented by the solid
line m Figure 11 -6. Appreciable nsk of foliar injury
to susceptible species of vegetation will occur
above and to the right of the limiting values A low
risk of foliarmjury exists below and to the left of the
limiting values. The limited information base for
the effects of PAN makes it impossible to estimate
a range of uncertainty at present.
CONCLUSIONS
Based on the data available, the limiting values
for the effects of ozone on trees and shrubs range
from 400 to 1000 jug/m3 (0.2 to 0.5 ppm) for a
duration of 1 hr, 200 to 500 pg/m3 (0.1 to 0.25
ppm) for 2 hr, and 120 to 340pg/m3(0.06 to0.153
ppm) for 4 hr. For agricultural crops, the ranges for
limiting values are 400 to BOOpg/m3 (0.2 to 0.41
ppm)for0.5hr, 200to500^g/m3(0.1 to0.25ppm)
for 1 hr, and75to180A
-------
making an economic evaluation of the response of
plants to air pollutants, including ozone and other
oxidants. They used the distinction between injury
and damage proposed by German workers.98 Injury
is defined as any indentifiable and measurable
response of a plant to air pollution; damage is any
measurable adverse effect on the desired or
intended use of the plant. Hence, before an effect
on a plant can be evaluated in terms of economics,
the plant must have been altered either
quantitatively or qualitatively in such a way as to
reduce its use value. In this context, visible
symptoms or transient changes in physiologic
responses may not result in an economic loss. For
example, leaf necrosis in soybean is injury; to be
classified as damage, the injury must affect bean
yield. In contrast, an oxidant episode that bronzes
the leaves of romaine lettuce may not affect
biomass, but it will affect use and may result in
complete economic loss. Similarly, ornamental
plants, whose major use value depends on
appearance, may be both injured and damaged.
Emergence tipburn of eastern white pine15 and
other physiogenic diseases of white pine are
associated with air pollution throughout the
natural range of the species. Insular stands
containing ponderosa pine are frequently injured
in the Southwest. The significance of the
physiogenic diseases is not understood because
only sensitive genotypes are affected. The disease
occurs randomly in the forest, and its most obvious
effect is the gradual elimination of genotypes that
may have otherwise superior silvicultural
characteristics. This could be a serious loss to
future tree improvement efforts, and similar
effects may be occurring in other forest species.
Economic considerations have not been addressed
in any reasonable way.
Oshima and his coworkers
227,228,230
have
developed a methodology for evaluating and
reporting economic crop losses that involves
continuous air monitoring, chamber exposures,
and monitoring plant species. Oshima has
attempted to weld these parameters into a
comprehensive method of determining yield
reductions. Several species have been
investigated, but the procedure has not been fully
clarified. No economic analyses of the yield
reductions were presented. This type of approach
should be investigated further.
Any attempt to assess oxidant damage to
agricultural crops requires judgement by a
i.o- -
E '
a
°- 0.8
Z
° 0.7
0.5--
£E °-6 +
z
LU
u
8 0.4
LU
O
N
O
0.3- •
0.2- -
0.1 - •
TIME
0.5
1.0
2.0
4.0
8.0
CONCENTRATION, ppm
SENSITIVE
0.431 ± 0.044
0.218 ± 0.023
0.111 ± 0.020
0.058 ± 0.022
0.031 ± 0.023
INTERMEDIATE
0.637 ± 0.043
0.347 ± 0.020
0.202 ± 0.017
0.129 ± 0.019
0.093 ±0.021
RESISTANT
0.772 ± 0.070
0.494 ± 0.028
0.355 ± 0.023
0.286 ± 0.026
0.251 ± 0.034
012345678
TIME, hours
Figure 11-5. Ozone concentrations versus duration of exposure required to produce a 5% response in three differ-
ent plant susceptibility groupings. The curves were generated by developing 95% confidence limits around the
equations for "all plants" in each susceptibility grouping from Table 11-5. Curves: a=sensitive plants, b=intermediate
plants, c=resistant plants.217
267
-------
competent investigator. Landau and Brandt165
suggest that the accuracy of crop surveys is
directly related to the number of subjective
decisions required in data collection. The first
surveys were conducted in California203'204 and
were used, with a general survey of conditions on
the east coast, to develop some of the early
economic loss estimates of $8 to $10 million in
California and $18 million on the east coast.217
These estimates were raised to $500 million on the
basis of increased awareness of pollution effects
and increased recognition of additional sensitive
species. However, all early estimates were
subjective, with no substantial backup data.
Since 1969, a number of states have instituted
an intensive training program for county
agricultural agents and have made detailed reports
of crop injury and damage87'164'165205'214'236 (Table
11-7). The first such report came from
Pennsylvania164 in 1969 and gave an estimate of a
$9.6 million loss to agronomic commodities from
oxidant pollutants. This survey included direct and
some indirect costs. Similar surveys have been
conducted in New England,214 New Jersey,236'237
and California.205 These surveys considered yield
reductions on the basis of injury and made no
direct assessments of growth or yield, although
subjective estimates of damage were obtained.
Pell and Brennan237 presented a well developed
thesis on the rationalefor differences in estimating
losses to agriculture in New Jersey between 1 971
and 1972, This subject was discussed in relation to
the overall problem of,assessing agricultural
losses.
A study was initiated in 196911 by Stanford
Research Institute (SRI) to develop an empirical
model for assessing damage to vegetation. This
program made useof laboratoryandfield data from
controlled exposures of various crops and of
chemical data from simulated reaction chambers
so that estimates of ozone and other oxidants could
be made on the basis of concentrations of primary
pollutants. Hydrocarbon was chosen as the basic
pollutant from which to develop the model for
TABLE 11-5. CONCENTRATION, TIME, AND RESPONSE EQUATIONS FOR THREE SUSCEPTIBILITY GROUPS AND FOR
SELECTED PLANTS OR PLANT TYPES WITH RESPECT TO OZONE8
Plants
Sensitive
All plants
Grasses
Legumes
Tomato
Oat
Bean
Tobacco
Intermediate
All plants
Vegetables
Grasses
Legumes
Perennial
Clover
Wheat
Tobacco
Resistant
All plants
Legumes
Grasses
Vegetables
Woody plants
Cucumber
Chrysanthemum
-00152
-00565
00452
-00823
-00427
-0 0090
00245
00244
-00079
00107
001 16
00748
-0 0099
-0 0036
00631
0 1689
00890
0 1906
0 1979
02312
0 1505
0 2060
C~AO . A,, • A,
•>- 0 00401
+ 0 00481
+ 000361
+ 000431
+ 000511
+ 0 00301
+ 0 00341
+ 000651
+ 0 00641
+ 0 00591
+ 000741
+ 0 00701
+ 000711
* 000811
+ OOO87I
+• 0 00951
t 001081
+ 001171
+ 001261
+ 000611
+ 001411
+ 000521
if
+ 0213/T
+ 0291/T
+ 0172/T.
+ 0 243/T
+ 0 273/T
+ 0 164/T
* 0 137/T
+ 0 290/T
0 263/T
0292/T
0329/T
0237/T
0268/T
0 302/T
0 152/T
+• 0 278/T
0 304/T
0 263/T
0 107/T
0208/T
0106/T
0256/T
K1'
057
074
046
050
076
058
052
074
0 79
082
0 81
077
095
088
078
051
082
055
0 70
045
083
040
Threshold
cuncentration,
t he 4 hr
022
026
024
0 18
026
017
0 18
035
029
033
0.38
035
029
034
026
050
045
051
038
047
033
049
006
004
0 11
None
005
005
008
0 13
009
Oil
0 13
0 17
009
0 11
0 14
027
022
0 31
029
031
0 25
030
Mean values1*
ppm"
Bhr
003
001
009
None
002
003
0 06
009
006
009
009
0 14
006
0 08
0 13
025
0 18
0 20
0 20
0.30
023
027
No data
pomts
471
71
100
20
30
62
197
373
25
6B
104
27
24
15
59
291
36
13
16
46
18
45
Cone iCi
ppm
029
037
034
031
037
030
023
037
041
039
040
0.36
0 28
047
0 28
045
030
045
0 55
039
0 41
039
Time (T),
hr
1 74
1 66
1 42
1 50
1 66
1 23
1 90
1 67
1 29
1 61
1 59
1 91
2 13
1 25
1 99
1 55
1 89
1 47
1 50
250
1 41
2 17
Response (I
45 4
509
40 1
56 5
40 2
472
389
27 0
335
31 0
250
22 9
230
289
157
106
122
6 5
178
78
133
12 6
Dose,
ppm 'hr
0503
0608
0480
0491
0611
0370
0448
0625
0532
0625
0 642
0687
0595
0508
0551
0696
0722
0655
0819
0905
0 581
0847
^Equations were developed from exposures limited iri lime (0 5 lo B hr, except far 2 to 12 hr points in the sensitive group) and denote acute responses of the plants
Concentrations range from Q Ob to U 99 U 0} ppm and responses from 0 to 99 0 00}% of control Reference 217
CC is ozone concentration m ppm I is percent injury T is time m hours, and Ac, At, and A? are constants {partial regression coefficients! thai are specific for poilman! plant
species or group of species and environmental conditions used
^Multiple correlation coefficienr squared, which represents the percent variation explained by the model
aFor 5% response in 1 - 4- and B~hr periods
'Prom the computer analysis
268
-------
prediction of expected oxidant values and,
therefore, effects on various crops. From the
oxidant value, injury and damage for specific crops
were calculated for over 100 statistical reporting
areas in the United States. The report used many
subjective assumptions and was related primarily
to visible injury symptoms. On the basis of the SRI
model, the 1969 estimated loss to vegetation from
oxidants in the United States was approximately
$1 25 million. According to the National Research
Council,217 if the increase in crop values are
considered and if it is assumed that oxidant
concentrations have not been significantly
reduced during the last few years, the loss in 1974
from oxidants could approach $300 million as
based on the SRI report.
A summary of estimates derived from various
surveys and assessment techniques is shown in
Table 11 -7. These values are suggestive at best. As
with all values developed for agricultural losses,
these have been directed at losses to the
producer—not the consumer.
Whereas a reduction in production may actually
increase the aggregate farm income and produce
serious income distribution problems, the
consequent reduction in marketable surplus would
cause a significant rise in the cost to consumers
0.24
because of the inelastic consumer demand for
most agricultural crops. Thus at the consumer
level, losses based on farm prices are not
appropriate and are likely to be conservative.
Because of percentage markups and fixed
wholesale and retail marketing costs, the cost to
the consumer of agricultural losses could be twice
as great as that observed at the farm level.
TABLE 11-6. OZONE CONCENTRATIONS FOR SHORT-
TERM EXPOSURES THAT PRODUCE 5 OR 20 PERCENT
INJURY TO VEGETATION GROWN UNDER
SENSITIVE CONDITIONS"
Ozone concentrations that may produce 5% or 20% injury, ppm"
Exposure
time, hr Sensittve plants intermediate plants Resistant plants
05
1 0
20
40
80
035
(045
015
(020
009
(0 12
004
(010
002
(006
-050
-060)
-025
-035)
-0 15
- 0 25)
-009
-015)
- 004
-012)
055
(065
025
(035
015
(025
010
(015
007
(0 15
- 070
- 085)
- 040
- 055)
-025
-035)
-015
-030)
-012
- 025)
>0 70 (0 85)
>0 40 (0 55)
>0 30 (0 40)
>0 25 (0 35)
>0 20 (0 30)
"Data developed from analysis of acute responses shown in Table 11 ~5andF
-------
TABLE 11 -7. ESTIMATES OF ECONOMIC LOSSES TO CROPS AND VEGETATION IN THE UNITED STATES
ATTRIBUTABLE TO OXIDANT AIR POLLUTION3217
Area
United States
California
Pennsylvania
New Jersey
New England
and year
1963
1963
1970
1963
1969
1970
1971
1972
1971
Estimated
loss, S10'
S 65,000
121,400
33,700
17,500
6,300
9,600
60
960
60
1,100
Comments
First approximation for commercial crops (SRI)
Revised SRI report to include ornamentals
Revised SRI report
Does not include ornamentals or indirect costs
Revised SRI report
Pa survey, includes indirect costs
Pa survey, as above
N J survey of a limited number of crops, based on
visible injury
N J as above
Mass survey, primarily crops and ornamentals
Reference
11
11
11
205
11
164
163
87
236
214
aSome of variation reflects methodology, other reflects differences in plant susceptibility and pollution over years
Factors Affecting Plant Response
The response of plants to ozone, PAN, or any
environmental stress is conditioned by complex,
interacting internal and external factors. These
include climatic and edaphic factors, interactions
among atmospheric pollutants, genetic variability
(both between and within species) of plants, the
growth and physiologic age of susceptible plant
tissue, and interaction of the plants with a variety
of pathogens. A conceptualization of the
interacting factors involved in air pollution effects
on vegetation is shown m Figure 11 -7.
GENETIC FACTORS
Resistance to ozone and PAN varies among
species of a given genus and among cultivars
within a given species. Variations in response are
functions of genetic variability and environmental
stresses as they affect morphological,
physiological, and biochemical characteristics. In
native populations and in breeding experiments,
both ozone and PAN may act as stresses
influencing selective pressure. Variations in
species response to ozone and PAN are well
documented.126'129'150'168'270'302'312
A number of papers discussing the variation in
response of cultivars within species have been
published.123"125'250 The research of Brennan et
al.31 indicated *liat cultivars of oats and potato
responded differently to ozone. Heck 114
summarized the variation in response of cultivars
within several plant species, while Reinert241
compiled a compendium of papers that discussed
the response of cultivars of horticultural crops to
ozone, PAN, and other pollutants. Variations in the
responses of cultivars within three species have
been studied intensively to determine their
sensitivity to ozone. They are: petunia,25'40'85
tomato,41'42 and tobacco.34'66'97'192^195
Additional studies discussing cultivar responses
to ozone are available for bean,58 begonia,3
morning glory,215 chrysanthemum,27'155
poinsettia,183 spinach,184 lettuce,244 radish,244
turfgrasses,26 forage legumes,32 alfalfa,141
safflower,143 soybean,142'208'291 small grains (oat,
rye, wheat, barley),242 and English holly.28 Petunia
has been studied with regard to both auto exhaust
and PAN,85 and the effects of PAN alone on both
petunia and chrysanthemum have been
studied.71'313 Most of these studies reported the
results of acute exposures, which may or may not
be related to resultsfrom chronic exposures. Foliar
injury was generally used as a measure of
response. It is not known whether rankings
according to foliar injury relate to loss in economic
yield.
Hanson103 published a list of 1 60 woody species
that were sensitive or tolerant to oxidants, based
on observations madeatthe Los Angeles State and
County Arboretum. A number of recent
investigations have considered susceptibility of
tree species to Ozone.l4.l7.47'55-57.6°'l26.l47.2°6.31° A
limited number of studies discuss selection of
resistant individuals within native tree
species.14'60'210'253
Two studies have explored tiie mechanism of
genetic resistance to ozone. Engle77 and Engleand
Gableman79 found that ozone sensitivity in onion is
probably controlled by a single gene pair, with
dominance of the resistant gene. In resistant
plants, the membrane of the guard cells was more
sensitive to ozone; when exposed to ozone, it lost
its differential permeability, thus causing stomatal
closure. This did not occur in the guard cells of the
sensitive cultivar, and thus the stomata remained
open. Taylor267 crossed the sensitive Bel W3
tobacco with a resistant line and found that the Fi
generation was of intermediate sensitivity. The F2
270
-------
generation segregated into 40 percent resistant,
10 percent sensitive, and 50 percent intermediate.
He suggested that sensistivity was controlled by at
least two genes. Resistance mechanisms to air
pollutants are poorly understood, and further
studies are needed.
A study using Arabidopsis thaliana was
conducted to determine the potential mutagenic
effects of ozone.36 This plant completes its life
cycle in about 35 days. Many generations may
therefore be grown within a relatively short period
of time. Plants were exposed to acute doses of
ozone for 6 hr/day,3 days/week, throughout the
life cycle. Seeds were collected from control and
exposed plants and planted for seven generations.
Seed production and biomasswere reduced within
a specific generation, but none of the factors
studied were transmitted to subsequent
generations. The results showed that for this
particular species, the concentrations of ozone
used produced no mutagenic effects.
The aforementioned studies suggest that a
spectrum of genotypes exists that varies in
susceptibility to oxidants, and that environmental
factors are important in conditioning a plant's
susceptibility to pollutants. The degree to which
environmental modification can control the
response of sensitive genotypes to ozone may be
less pronounced if specific biochemical or
physiological mechanisms involved in plant
response to pollutants have weak interactions with
the environment.
PHYSIOLOGICAL AGE
The physiological age of a plant is an important
factor in modifying plant sensitivity. Young, rapidly
expanding leaves are most sensitive to PAN
injury.20'238 Ozone, in contrast, affects more
POLLUTANT
CONCENTRATION
NUMBER OF.
EXPOSURES
DOSE
CLIMATIC FACTORS
EDAPHIC FACTORS
.DURATION OF
EACH EXPOSURE
PLANT
RECEPTOR
•BIOTIC FACTORS
I
GENETIC
COMPOSITION
STAGE
OF
DEVELOPMENT
T
MECHANISM OF ACTION
I
EFFECTS
Figure 11-7. Conceptual model of factors involved in air pollution effects on vegetation (modified from van Haul and
Stratmann),302
271
-------
mature tissue. Maximum sensitivity isfound when
cotton leaves are approximately 70 percent
expanded,281 red maple 90 percent,299and soybean
60 to 80 percent.286 In tobacco, maximum
sensitivity occurs just after full leaf expansion.176
Davis55 and Davis and Wood60found that the age
of Virginia pine needles influenced their response
to a 4-hr exposure to ozone at 490 ^g/m3 (0.25
ppm). Generally, cotyledons were more sensitive
than primary needles, which in turn were more
sensitive than secondary needles. Secondary
needles of seedlings were approximately as
sensitive as those of 3-year-old trees. The
cotyledons and secondary needles became
resistant after 16 and 18 weeks of age,
respectively, whereas the primary needles
remained sensitive beyond 18 weeks. Berry13
reported that Virginia, shortleaf, loblolly, and slash
pines at 2 to 6 weeks from seed were most
sensitive when exposed to ozone at 490 jug/m3
(0.25 ppm) for 2-hr.
Hanson et al.104 reported increased tolerance in
petunia cultivars to ozone as they approached the
flowering stage. This was true for both ozone-
sensitive and tolerant cultivars, with the latter
being the more strongly influenced. They
suggested that bud development resulted in the
production of diffusible substances that moved
down the plant and acted as a protectant.
Generally, studies have shown that plants are
most sensitive to ozone at a physiologic age
associated with full leaf expansion. Sensitivity is
also associated with functional stomata,
intercellular spaces, and rate of cutm formation on
cell walls 20 Plants generally are more sensitive to
oxidants during stages of rapid growth and less
sensitive as leaves mature. When oxidant episodes
occur throughout the growing season, the older
leaves, weakened during their stage of maximal
physiologic growth, senesce prematurely.
CLIMATIC CONDITIONS
Plant response to ozone and PAN may be altered
by climatic conditions before, during, and after
exposure.114 Plants may be sensitized by a given
set of conditions after 1 to 5 days.
Conditions before and during exposure are
critical, and those after exposure may be
important. The responses of plants to ozone and
PAN under varied climatic conditions were studied
primarily under laboratory and greenhouse
conditions. Field observations have often
substantiated these results Most studies have
involved individual environmental factors and one
or two response measures, including usually the
evaluation of injury. Several investigations have
dealt with the interactions of environmental
factors. Information sufficient to make
generalizations regarding plant response to
oxidants exists, but there is much uncertainty
because of the small number of species studied
and the lack of information on the interactions of
the environmental factors evaluated.
Light Quality — The quality of light affects the
growth and development of plants and plays a role
in determining the response of plants such as pinto
bean to PAN72'73 Injury to pinto bean from PAN
was maximal when exposed to 420 and 480 nm
and less than half at 640 nm. The response was
apparently associated with changes in carotenoid
pigments. Shinohara et al.268 reported the effect of
lightquality on the response of tobacco(H-mutant)
exposed to 785 /jg/m3 (0.40 ppm) ozone for 30 mm.
The least injury occurred in far red light, followed
by blue, green, and white, with the greatest effect
in the red spectrum.
Photoperiod — Physiological control over some
aspects of plant development is exerted by a
specific light period based on a 24-hr cycle.
Research has shown that plants are more sensitive
to ambient oxidants and ozone when grown under
an 8-hr photopenod compared to either a 12- or
16-hr photoperiod,119149'176
Light Intensity — The intensity of light affects
many physiological processes within plants and is
known to affect the response of plants to oxidants.
Studies have been reported on the effects of light
intensity before, during, and after exposure to
oxidants.
Dugger et al.73 found a direct correlation
between the sensitivity of pinto bean to PAN and
increasing light intensity. Pinto bean required light
before, during, and after a PAN exposure for injury
to occur.72'73 This was not true of response to
ozone, although some light period was necessary.
Generally, plants are more sensitive to ozone
when grown undera short photoperiod at low-light
intensities. This was demonstrated with pinto
bean74119 and tobacco.74'119'258
Temperature — The response of plants to ozone
varies with temperature. The effect of temperature
on plant sensistivity to ozone varies with plant
species, and thus no constant pattern of response
is discernible. Some plants (radish7'225) were more
sensitive to ozone if grown under cool conditions,
272
-------
and others (snap bean,5 soybean,75 Bel Wa
tobacco,193'257 Virginia pine,55'59 and white ash309
were more sensitive if grown under warm
conditions. Pinto bean73 was sensitive to ozone
regardless of temperature. Macdowell176 found
that a low day or high nighttemperature increased
the susceptibility of White Gold tobacco to ozone.
Juhren et al.149 usedeightcombinationsof dayand
night growth temperatures with Poa annua and
then exposed these plants to ambient oxidants for
a day. The sensitivity varied with plant age and was
greatest at the 26° to 20°C day/night
temperatures. Both high and low temperatures
during growth could, in certain plants, produce
physiological changes that are associated with
stress resistance and thus protect the plant from
air pollutant stress. Difficulties arise in
interpreting the results of environmental studies
conducted under greenhouse conditions.
Relative Humidity — Davis55 and Davis and
Wood59 found an increase in sensitivity of Virginia
pine to ozone at high humidity during exposure, but
they reported no effects from post-exposure
changes in humidity. Otto and Daines233 found
similar results for pinto bean and tobacco exposed
over a wide humidity range and at several ozone
concentrations. They did not study growth or post-
exposure conditions. The studies of Dunning and
Heck74 rhowed a significant increase in the
response of pinto bean, but not of tobacco, when
humidity was increased during ozone exposure.
Pinto be, n grown at 60-percent relative humidity
was mote sensitive to ozone when compared to
plants iirown at 80-percent humidity. The
response- of tobacco to ozone was unaffected by
humiditv differences during exposure; however, its
sensitivity was conditioned by the humidity before
exposure. Sensitivity of tobacco decreased with an
increase in humidity. Response of pinto bean was
unaffected by humidity when grown under a light
intensity of 4000 ft-c (43,060 Ix), but it was
increased with an increased humidity (80 percent)
and a decrease in light intensity levels of 2000 ft-c
(21,530 Ix) during growth.74 Table 11-8
summarizes some of these results.
Carbon Dioxide — Stomatal opening is decreased
by high carbon dioxide levels and may affect plant
sensitivity to oxidants. Heck and Dunning119
reported a decrease in sensitivity of tobacco to
ozone if the tobacco was exposed to elevated
carbon dioxide concentrations of 500 ppm
immediately before and during exposure to ozone
(22 percent injury with added carbon dioxide, and
66 percent injury without).
METEOROLOGICAL PARAMETERS
Canadian workers using an ambient oxidant
dose have correlated variations in meteorological
parameters with plant injury.
179,213
A correlation
was discovered when an empirical relationship
involving evapo-transpiration (the coefficient of
evaporation) was developed and used to modify the
dose information This empirical relationship has
been used on a limited basis to predict damaging
oxidant concentrations from monitored
meteorologic conditions.
Linzon172'173 also noted plant injury that was
associated with changes in meteorological
conditions. He reported needle blight symptoms on
white pine subsequent to several days of wet
weather followed by a continuous sunny period.
Symptoms were observed several timesdurmg the
1957-64 growing seasons at Chalk River, Ontario.
However, the time of occurrence did not correlate
well with peak oxidant concentrations.
Skelly et al.269 associated post-emergence acute
tipburn on white pine growing at three different
sites along the Blue Ridge Parkway with high
oxidant during the summers of 1975 and 1976.
The two episodes were associated with
meteorological conditions that resulted in the
transport of oxidant or oxidant precursors from
many miles away. The episode of July 1975
resulted because of a low-pressure system
(Hurricane Amy) that was off the Atlantic Coast in
the vicinity of New York and a high-pressure
system from Canada that became stationary over
the Great Lakes. The circulation of air around a
low-pressure system is counter clockwise, and it is
clockwise around a high-pressure system. This
pattern forced air from the northeast south into
Virginia. The movement of the high-pressure
system into Virginia permitted further oxidant
synthesis. The June 5-1 2, 1 976, episode resulted
because of winds from the northeast on June 5 and
6 and a stationary high-pressure system that
continued through June 1 2. As a result of the air
movement, oxidant levels in July 1975 and again
in June 1 976 exceeded 1 60 /ug/m3 (0.08 ppm). In
July 1 975, needle injury was associated with a 1 -
day episode, and from June 5-1 3, 1 976, the three
sites exceeded 160/ug/m3(0.08 ppm)for 110,186,
and 122 consecutive hours, respectively.
A widely held belief is that vegetation growing in
the humid eastern United States would be severely
273
-------
injured if oxidant concentrations reached the daily
peak concentrations of 390 to 780 //g/m3 (0.20 to
0.40 ppm) commonly experienced in the less
humid sections of southern California, An air
pollution episode that occurred on July 27-30,
1970, in the Washington, D.C., area (as well as
those decribed above by Skelly et al.259) is
indicative of what may happen. During this 4-day
period, the peak oxidant concentrations ranged
from 220 to 400 //g/m3 (0.14 to 0.22 ppm) and
were accompanied by a low concentration of sulfur
dioxide (0.04 ppm). Oxidant injury was observed on
31 tree, 15 shrub, and 1 8 herbaceous species in an
area of about 187 km2 (72 miles2). Increased
emissions of the precursors associated with
oxidant fomation could result in repeated
occurrences of acute injury or even chronic injury
to eastern vegetation.5
Such factors as windspeed and atmospheric
pressure appear to play little or no role in affecting
plant sensitivity to oxidant pollutants. Air
movement would be expected to play a role under
ambient conditions because of its known effect on
air boundary layers of leaves, but it probably has
little effect in chamber work unless wind velocities
are greater than 1.6 km/hr (1 mph).29'112'128
EDAPHIC CONDITION
Khatamian et al.153 reported that tomato plants
grown under water stress, which did not in itself
cause a reduction in growth, were more tolerant of
ozone injury. The effects of soil moisture on the
response of selected plants to ozone are presented
in Table 11-9. Markowski and Grzesiak136 found
that bean and barley grown under drought
conditions were protected from ozone injury.
Several field studies67'268304 showed that there
was a close positive correlation between soil
moisture and oxidant injury to several cultivars of
tobacco. Field observations generally suggest that
TABLE 11 -8. RESPONSE OF PLANTS TO OZONE, AS CONDITIONED BY HUMIDITY DURING GROWTH AND EXPOSURE
Plant
species
Pine, Virginia
Bean, cultivar
Pinto, and
Tobacco, cultivar
Bel Wa, averaged
Bean, cultivar
Pinto
Tobacco,
cultivar Bel W3
Tobacco,
cultivar Bel W3
Bean, cultivar
Pinto
Ash, while
Q/one
concentration.
p»pm, hr
025, 4
025, 4
040, 1
040, 1
Notes "
Control conditions.
3-year seedlings
Juvenile
Control conditions,
PP. 8 hr
Control conditions,
PP, 8 hr, 75% GH
Control conditions.
PP, 8 hr, 2,000
ft-c (21,529 Ix)
control conditions,
PP, 8hr
Growth DT
exposure1'
Exposure
Growth
Exposure
Growth
45% EH
90% EH
Exposure
Growth
Exposure
<60)
4
50
1
(45)
36
73
41
(60)
66
52
Response,
% injury (% RH)e
(85)
25
58
35
(60) (75)
39 41
67 81
53 70
(80)
78
67
Reference
59
(90) 233
31
80
81
74
040, 1
030, 1 5
0 20, 1 5
0 25, 4
Control conditions. Growth 42 36
PP, 8 hr, (2,000 ft-c
(21,529 Ix) control Exposure 33 36
conditions, PP, 8 hr
Control conditions, (26) (51)
31°C Exposure 9 39
Control conditions, Exposure 0 0
31°C
(95)
50
55
Control conditions,
1 -yr seedlings
(60) (80)
Growth 33 46
Exposure 38 41
Post-exposure 36 41
74
233
233
309
aPP= photopenod. GH - hunmdny durmg growlh, EH = humidity during exposure, h-c - loot candle
Time of humidity treatment
cHumidity values are in parentheses
274
-------
sensitive plants may become resistant under
drought conditions. Rich and Turner249 found rapid
stomatal closure in pinto bean during a 30-min
exposure to ozone at 392 to 490 jug/m3 (0.20 to
0.25 ppm) if plants were grown under a soil
moisture stress; however, closure was slower
under optimal water availability. They also
reported rapid stomatal closure in pinto bean and
Bel W3 tobacco if the plants were conditioned and
exposed to ozone at a relative humidity of 37
percent; however, a similar reponse was notfound
at a relative humidity of 73 percent. This evidence
suggests that ozone may induce more rapid
closure of stomata when plants are already under
some type of water stress.
Starkey265 found that the foliage of bean that
was sensitive to subacute concentrations of PAN
was also injured by drought following PAN
exposure. Plants resistant to PAN were not injured
similarly by drought following PAN exposure.
Salinity and low soil moisture have been shown
to increase the resistance of plants to ambient
oxidants.138'139'175'222223 Salinity, however,
suppresses plant growth so that the protective
effect is offset. The protective effect against ozone
is enhanced with increasing salinity.34
The interaction of soil fertility and the response
of plants to oxidant is not well understood.
Nitrogen nutrition has received some attention.
Menser and Street,196 in an ambient oxidant study,
reported an increase in tobacco fleck with
increasing applications of soil nitrogen. This
response was also found in spinach exposed to
ozone.34 Leone et al.170 reported that
concentrations of nitrogen optimal for growth
resulted in tobacco being most sensitive to ozone,
whereas either higher or lower nitrogen
applications increased resistance. The opposite
was reported for White Gold tobacco, in which the
optimal nitrogen concent ration for growth gave the
greatest resistance to ozone.178 Ormrod et al.225
found that concentrations of nitrogen had no effect
on growth of radish exposed to ozone. It is apparent
that the nitrogen concentrations and other
experimental conditions used in these studies
were not sufficiently critical for an evaluation of
nitrogen-oxidant interactions.
Several studies have explored the importance of
phosphorus in modifying the sensitivity of plants to
oxidant. Ripaldi and Brennan252 reported an
increase in phosphoruswithinleavesof pintobean
after exposure to ozone. Leone and Brennan189
reported increases in ozone injury and in
phosphorus content of tomato leaves with
increasing applications of phosphorus. Brewer et
al.34 found an interaction between potassium and
phosphorus in the response of spinach to ozone.
They reported that at low phosphorus content, an
increase in potassium tends to increase injury, but
at high concentrations of phosphorus, the increase
in potassium tends to inhibit injury. Dunning et
al.75 found that pinto bean and soybean were more
TABLE 11-9. EFFECTS OF SOIL MOISTURE ON RESPONSE OF SELECTED PLANTS TO OXIDANT STRESS2"
Response, percent reduLUori f
PSanl species
Tobacco, cullrvar
Catterton
Tomato, cultivar
Fireball
Beet, garden
Bean, cultivar
Pinto
Oxidant concentration
ppm. hr
Ambient oxidant
1 00, 1 5
1 00, 1 0
0 50, 1 0
1 00, 1 0
Control
0 20, 3 (daily for 38 days)
Control
0 15, 2/day (63 days)
0 25, 2/day (63 days)
Control
0 15, 2/day (63 days)
0 25, 2/day (63 days)
ro'Ti control
Moisture condii'Ons'1
Type of response
% injury
% red in chlorophyll
% red in chlorophyll
% red in chlorophyll
% red leaf dry wt
% red in dry wt of
storage root from
nonsaline control
% red in shoot dry
wt from nonsaline
control
% reduction in root
dry wt from nonsaline
control
h.yh
Irrigated
29
90% turgid
54
67
36
48
40 kPa
0
40
-40 kPa
0
27
93
0
25
91
to
Normal
11
80%jturgjd
10
24
+ 3
+40
^440 kPa
24
52
-200 kPa
18
42
91
25
28
89
low Refer)
19
15
15
15
-840 kPa 22
68
69
-400 kPa 1 3
78
87
88
65
78
79
aSpeoai soil moisture conditions are underlined kPa = kilopascals ^bars * 10"-')
275
-------
sensitive to ozone at low potassium
concentrations. Adedipe et al.6 reported that low
sulfur concentrations increased the response of
Blue Lake snapbeans to ozone. Mcllvenn et al.190
reported increased injury with increasing
concentrations of zinc in the soil. Soil nutrition
probably plays an important role in the response of
plants to pollutants only in cases of nutrient
imbalances, although total nutrient salts could
affect response under some conditions. In soils
with balanced fertilizer regimes, plants with
similar genotypes would probably respond fairly
uniformly to oxidant stress. The responses of
plants to ozone stress when grown under various
nutrient regimes are summarized in Table 11-10.
POLLUTANT INTERACTIONS
Oxidants are components of a complex mixture
of gases in the atmosphere, many of which may be
phytotoxic. Except for ambient air studies and
studies simulating photochemical oxidants, there
have been few investigations of the effects of
pollutant combinations on vegetation.
Ozone and sulfur dioxide mixtures are of special
interest, owing to their widespread co-occurrence
and to the possibility of a greater than additive
TABLE 1110. EFFECTS OF VARIOUS NUTRIENTS ON RESPONSE OF SELECTED PLANTS
TO OZONE (OXIDANT) STRESS217
Plant species
Tobacco, cultivar
White Gold
Tobacco, cultwar
Catterton
Tomato, cultivar
Rutgers
Spinach, cultivar
Viroflay
Radish, cultivar
Cavalier
Tomato, cultivar
Rutgers
Spinach, cultivar
Viroflay
Radish, cultivar
Cavalier
Bean, cultivar
Blue Lake
Soybean, cultivar
Dare
Bean, cultivar
Pinto
0 xida nt
concentration
ppm, hr
035, 48
Ambient
oxidant
0 18, 4
025,
9/day
3 days
0 25,4
0 15, 3
025, 3
030, 3
025,
9/day
3 days
0 25, 4
050,
2 day
2 days
8 doses
(acute)
8 doses
(acute)
Response, % reduction from control
Type of response
% injury
% injury
% injury
% injury
% growth reduction
% injury
% injury
% injury
% injury
% growth reduction
% chlorophyll
reduction
% injury
% injury
Element
Nitrogen
Nitrogen
Nitrogen
Nitrogen
Nitrogen
Phosphorus
Phosphorus
Phosphorus
Phosphorus
5 ppm K
40 ppm K
Phosphorus
Sulfur
Potassium
Potassium
Low
01
21
60 Ib/acre
29
28 mg/liter
20
10 ppm
3
60 mg/liter
50
1 5 ppm
0
20
20
20 ppm
14
25
30 mg/liter
35
1 3 mg/liter
55
105 mg/liter
40
26
nutritional levels"
to
05 10
16 4
90
20
280 560
40 40
30 ppm 90 ppm
18 22
300 mg/hter
55
155 ppm 62 ppm
0 40
40 60
60 60
1 50 ppm
13
9
150 mg/liter
43
32 mg/liter
11
710 mg/liter
23
18
High Heterence
5 Ob 176
22
120 196
11
1120 170
20
34
225
170
34
225
• 5
75
75
''Nutritional values are underlined and show a number of different units
"Relative to a full-strength nutrient solution
276
-------
effect on vegetation.191 Most studies of the effects
of ozone on vegetation have been directed
primarily to understanding the responses of
various plant species or cultivars to a single air
pollutant and the manner in which these
responses are modified by environmental factors.
However, in the field, vegetation is always exposed
simultaneously to several air pollutants. When the
effect produced by exposure to a combination of
pollutants results in a greater effect than exposure
to a single pollutant, the effects are synergistic.
Antagonism is the opposite reaction. The effect
level resulting from exposure to the combined
pollutants is less than the sum of the injury
produced by exposure to the individual pollutants.
If the effect level resulting from exposure to the
combined pollutants is not significantly different
from the sum of the levels of the effects from
individual exposures, no interaction has occurred
and the foliar response to the combined pollutants
is additive. Menser and Heggestad in 1966191 first
reported that exposure of the sensitive Bel W3
tobacco to mixtures of sulfur dioxide (1 31 0/ug/m3
or0.50ppm)and ozone (59/ug/m3 or0.03ppm)for
2 or 4 hr caused foliar injury of 23 to 48 percent,
whereas the same concentrationsof the individual
gases produced no injury. The synergistic
interaction between ozone and sulfur dioxide
depicted in the above study stimulated plant
scientists to begin research on pollutant
combinations. Middleton et al.,202 working with
ratios of sulfur dioxide to ozone of 4:1 to 6:1 ,did not
observe an increase in injury, although they found
that at 4:1, ozone appeared to interfere with the
expected sulfur dioxide injury. Macdowall and
Cole178 reported that the two-gas combination
lowered the threshold concentrations for injury of
tobacco (cultivar White Gold) by sulfur dioxide but
not the threshold for ozone injury. Injury from
ozone alone was first noted at a dose of 98 /ug/m3
(0.05 ppm) for 1 hr. From S02 alone, injury was
noted at 260 /ug/m3 (0 1 ppm). And when the two
gases were combined, injury was noted at the end
of 1.5 hr. Macdowall et al.179 defined thethreshold
in terms of dose when they reported the threshold
at 20 pphm-hr (392/ug-hr/m3, or 0.20 pphm-hr).
This has not appeared to be true in some other
reports.191'292 Symptoms observed when sulfur
dioxide was below the threshold for S02 injury for
the specific plant are similar to those reported for
ozone.
Tingey et al.292exposed 11 species of plants to
different ratios of sulfur dioxide and ozone
mixtures. Plants were exposed to either 98 or 1 96
fjg/m3 (0.05 or 0.1 ppm) ozone and to 260, 660, or
1310/ug/m3 (0.1, 0.25, or 0.5 ppm) sulfur dioxide
for 4 hr. There was no general trend observed in
the manner in which the ratios of pollutant
concentrations caused foliar injury. Additive,
greater than additive, and less than additive
responses were noted (Table 11-11). Menser et
al.,
194,195
Grosso et al.,97 and Hodges et al.134
determined the response of several Nicotiana
species and various N. tabaccum cultivars to sulfur
dioxide and ozone mixtures. They found that ozone^
and sulfur dioxide acted synergistically and
produced ozone-type symptoms on all cultivars of
Maryland tobacco. When plants were fumigated for
4 hr with 60 to 70/ug/m3 (0.03 to 0.035 ppm) ozone
alone or with 115 to 130 /ug/m3 (0.045 to 0.05
ppm) with S02 alone, no injury was observed.
However, when the gases were combined and the
plants were exposed for the same length of time,
the result was leaf injury ranging from 5 to 1 5
percent. Jacobson and Colavito146 found that
sulfur dioxide at 79 /jg/m3 (0.04 ppm) decreased
the sensitivity of bean to ozone and increased that
of tobacco during a 4-hr exposure.
TABLE 11-11. SUMMARY EFFECTS OF SULFUR DIOXIDE
AND OZONE MIXTURES ON FOLIAR INJURY292
Plant species
Alfalfa
Broccoli
Cabbage
Radish
Tomato
Tobacco, Bel W3
Response at concentration ratio, SO:
050/005 050/010 010/010
+ +
+ 0 +
0 + 0
0 + +
00-
+ + 0
!/O3, ppm"
025/0 10
f
0
0
0
+
a+ - greater than additive, 0 = additive, - = less than additive
Differential susceptibility of individual clones of
eastern white pines to ozone and sulfur dioxide
was shown by Berry and Heggestad16 and
Costonis.46 When Dochinger et al.70 determined
that chlorotic dwarf could be caused by an
interaction of ozone and sulfur dioxide, they used a
chlorotic-dwarf-susceptible clone to eliminate the
genotype variable. Houston140 tested the response
of tolerant and susceptible clones of eastern white
pine (on the basis of symptom expression under
ambient conditions) to ozone or sulfur dioxide.
Injury caused by sulfur dioxide or sulfur dioxide
with ozone correlated well with the earlier field
responses, but ozone alone did not produce a
consistent response. Dochinger et al.70 and
Houston140 found that the SCh-Oa mixture
277
-------
increased the amount of injury, but Costoms47
reported less injury from the mixture than from
sulfur dioxide alone. Both Costonis and Houston
reported effects from sulfur dioxide and its mixture
with ozone at concentrations of both gases weil
below those reported by others. It is possible that
they used ultra-sensitive clonal materials
Whatever the reason, this work needs verification.
Applegate and Durrani10 also reported injury to
peanut at sulfur dioxide/ozone concentrations and
ozone concentrations well below those reported
for other plants. Their work also requires
substantiation. In the latter two cases, the
concentrations reported are close to the detection
limits of the gas-measuring instruments used.
Tingey et al.287 found an additive inhibition of top
growth of radish and a less-than-additive
inhibition of root growth after exposure to sulfur
dioxide-ozone mixtures. Tingey and Reinert290 and
Tingey et al.293 exposed soybean, tobacco, and
alfalfa to mixtures of sulfur dioxide and ozone and
reported a greater-than-additive inhibition of root
growth of soybean, an additive inhibition for
tobacco, and a less-than-additive inhibition for
alfalfa. Heagle et at.109 reported a greater-than-
additive effect on growth and yield in soybean
grown under field conditions in a mixture of these
gases, but the differences between the mixture
and the ozone treatments were not significant.
Weber307 reported a reduction in the growth of
leaves, stems, and roots of soybeans and an
increase in leaf abscission as influenced by a
mixture of ozone and sulfur dioxide. These
changes were similar to those caused by ozone
alone.
Combinations of other pollutants with ozone,
PAN, or both may be important, but they have
received little study. Matsushima187 reported
additive foliar effects on pinto bean and tomato
from a mixture of sulfur dioxide and PAN and a
less-than-additive effect on tomato from mixtures
of ozone and nitrogen dioxide. Fujiwara90 reported
a greater-than-additive effect on pea from a
mixture of ozone and nitrogen dioxide. Kress161
and Kohut167 studied the response of hybrid poplar
to ozone-PAN mixtures. Kress used sequential
exposures and found a greater-than-additive effect
after most exposures. After others, he reported
mixed responses. Kohut used simultaneous
exposure and found all three types of responses m
replications of the study. The reasons for these
variations are unclear.
Fujiwara90 and Reinert et al.242 have reviewed
the subject of pollutant interaction. Reinert etal.242
tabulated some of the studies dealing with
pollutant interaction. This information is presented
in Tables 11-12 and 11-13.
Studies of pollutant interactions are preliminary.
It is still not possible to define adequately the
potential impact of pollutant combinations on the
production of quality food, feed, and fiber. Plant
species are known to respond differently to
combinations of pollutants, and the responses can
be additive, greater than additive, or less than
additive. The variation in cultivar or species
response or the variation in the response of plants
grown or exposed under a variety of enviromental
stresses is still not understood.
POLLUTANT-PARASITE INTERACTIONS
Infection by biotic pathogens has been shown to
be a factor in the response of vegetation to oxida nts
(primarily ozone). Such responses have been
studied from several perspectives since Yarwood
and Middleton314 accidentally found that rust-
infected bean leaves were less sensitive to
photochemical oxidants. A number of investigators
have studied the protection from ozone injury
afforded to plants with active infections. Others
have noted that ozone injury increases the
sensitivity of plants to infection. Some
investigators have studied the effects of ozone on
pathogens, and several have found no interacting
effects. Other studies have involved bacteria,
fungi, viruses, insects, and nematodes. Stark et
al.264 and Miller et al.207 in California reported that
oxidant (ozone) injury to ponderosa pine
predisposed the trees to subsequent invasion by
pine bark beetles. The beetles increase the rate of
tree decline and may be the final cause of tree
mortality43 (Chapter 1 2). It is possible that oxidant
stress in other parts of the country contributes to
insect infestation in the forest areas (e.g., the Blue
Ridge Parkway, Va.259). Weber308 has shown that
ozone and mixtures of ozone with sulfur dioxide
(0.25 ppm, 4 hr/day) can cause a decrease in the
populations of several species of nematodes
parasitizing roots of soybean or leaves of begonia.
These types of interactions may be of significance
in areas of the country with significant oxidant
pollution problems. In general, it is felt that the
interacting effects resulting in modification of
pathogens are due to changes in the host
physiology and not to direct effects on the
pathogen. Heagle106 has reviewed pollutant-
parasite interactions and has also discussed the
278
-------
direct effects of pollutants on pathogens
themselves.
OTHER FACTORS
Heagle and Heck111 found that Bel W3 tobacco
was predisposed to oxidant injury by previous
exposure to ambient pollutants. Macdowall176
reported the same results when the first and
second exposures were to low oxidant levels.
Antagonism was noted when both exposures were
to high doses.
Although there has been considerable interest
in understanding how various factors affect the
response of plants to pesticides, especially
herbicides, very little has been done with the
effects of interaction between pesticides and air
pollutants. Hodgson and associates135"137 first
showed an effect of ozone on the metabolism of
herbicides. They found that ozone inhibited the
dealkylation of atrazine in corn and altered the
pathway of diphenamid metabolism in tomato. The
changes could be beneficial if oxidants increase
pesticide degradation, or harmful if oxidants stop
the biologic breakdown at a toxic intermediate or
retard the degradation process. Carney et al.39
reported that the herbicide pebulate in
combination with ozone caused a greater-than-
additive effect on White Gold tobacco and that
TABLE 11-12. FOLIAR RESPONSE OF SELECTED PLANTS TO SULFUR DIOXIDE AND OZONE MIXTURES26
Plant
species
Bean, garden
Bean, lima
Broccoli
Cabbage
Tomato
Radish
Alfalfa
Eastern white pine
Tobacco, cultivar
Bel W3
Tobacco, cultivar
Bel W3
Tobacco, Md
(6 cultivars)
Concentration
ratio, S02/03,
ppm
1 70/0 19
025/005
0.50/005
1 00/0 1 0
0 10/0 10
0 50/0 10
0 50/0.10
0 025/0 05
025/003
0.50/010
0.50/0.10
Exposure
duration,
hr
05
4
4
4
4
4
4
6
4
4
2
Foliar
injury,
24
0
17
28
10
50
60
26
41
88
20
Response to
mixture3 Reference
+ 187
0 292
+ 292
0 292
292
+ 292
+ 292
+ 140
+ 191
+ 292
+ 195
greater than additive, 0 - additive, - - less than additive
TABLE 11-13. GROWTH RESPONSE OF SELECTED PLANTS TO SULFUR DIOXIDE AND OZONE MIXTURES26
Plant
species
Radish
Radish
Alfalfa
Soybean
Soybean
Tobacco
Concentration
ratio, SOz/Os,
ppm
005/005
045/0.45
005/005
0 05/0 05
0.10/010
005/005
Exposure
duration,
hr
8/day, 5
days/week,
5 weeks
4
8/day, 5
days/week,
12 weeks
7/day, 5
days/week,
3 weeks
7/day, 5
days/week
until harvest
7/day, 5
days/week,
4 weeks
Plant response,
% reduction
from control3
10TDW
55 ROW
16 TOW
70 ROW
18 TDW
24 ROW
24 RFW
72TFW
63 seed wt
32 ROW
49 ROW
Response
to
mixture^
0
0
0
-
+
0
0
0
0
Reference
287
242
242
293
109
242
aTDW = top dry wt, ROW = root dry weight, RFW = root fresh weight, TFW = top fresh weight
b+ = greater than additive, 0 - additive, - = less than additive ,
279
-------
chloramben did the same with Delhi 34 tobacco.
They reported a less-than-additive response of
both tobacco cultivars to the combination of
benefin and ozone. These and other herbicides
acted independently of ozone exposure on tomato
and white bean. Ordin et al.224 found that Avena
coteopti/e growth was less inhibited by PAN when
2,4-D was used in amounts giving optimal growth.
These interactions with herbicides need additional
investigation to determine whether the responses
noted are of general importance. Research needs
to be directed at possible interactions between
atmospheric pesticides (vapors or fine particles)
and oxidant air pollutants.
Another interaction has recently been reported
between cadmium applied to soil and ozone
exposure of cress,53 If cadmium potentiates the
ozone response of cress, possibly other heavy
metals can cause a similar interactive response.
RESPONSES OF MOSSES, FERNS, AND
MICROORGANISMS
Mosses, ferns, and lichens have not been
extensively studied to determine ozone and
oxidant effects. Comeau and LeBIanc45 found that
a 4-hr exposure of Funaria hygrometrica to ozone
at 490 to 1960 /yg/m3 (0.25 to 1.00 ppm)
stimulated the regenerative capacity of the moss
leaves. This did not occur with 6- and 8-hr
exposures. Glater95 reported oxidant injury to
several species of fern growing in the Los Angeles
area. Initially, tan colored lesions appeared near
the smaller veins, but in no special pattern. Later,
the entire leaf became necrotic. Symptom
development and sensitivity of leaves were
different from those noted in more highly evolved
plants. All leaves appeared to be equally sensitive,
except for the growing tip and the youngest
uncoiling leaves Occasionally, a young plant was
killed
Lichenologists have long used the presence and
abundance of lichen species to map the biologic
impact of large urban and industrial areas.88
Though early workers considered changes in the
presence and abundance of these organisms to be
related to such factors as temperature and
humidity, more recently, most researchers have
tended to relate changes in lichen population more
to industrial air pollution than to other
environmental changes There is strong indication
that both the presence and the abundance of
certain lichen species are correlated with sulfur
dioxide concentrations in urban and industrial
areas. However, little is known about the direct
effects of ozone or other oxidants on lichen
morphology or physiology. Ozone or some of the
other oxidants may have an adverse ecological
impact on lichens
Ozone in high concentrations has been used in a
variety of applications for the control and
suppression of fungi and bacteria. These
applications have included food protection,
drinking-water purification," and treatment of
sewage,154'188 The germicidal effectiveness
depends on concentration, relative'humidity, and
the specific organism. In many cases, even a
concentration of 5880 to 9800/ug/m3 (3 to 5 ppm)
was not sufficient to kill some bacteria. Burleson et
al.36 showed mactivation of several viruses and
bacteria after ozone exposure, and a greater
inactivation with simultaneous sonication.
Zobnma and Morkovma315 related the tolerance of
a carotmoid strain of Mycobacterium carothenum
with the presence of the pigment.
Large doses of ozone may inhibit growth and
sporulation of fungi on fruit, although most fungi
tested were resistant to ozone, Spalding262
reported that ozone acted as a surface biocide.
Above 980 £ig/m3 (0.5 ppm), ozone inhibited
surface growth of fungi on strawberry and peach.
Ridley and Sims251 extended the shelf-life of
strawberry and peach by exposing them to ozone,
but stated no concentrations, Ozone at 1960 to
3720 /ug/m3 (1 to 2 ppm) for 1 to 2 hr/day
controlled the surface growth of fungi and
sporulation on apple, reduced offensive odors, and
decreased the ripening rate.261 Watson305 found
that ozone at 785 to 3920 fjg/m3 (0.4 to 2.O ppm)
acted as surface fungicide on fruit. Fungal growth
within the fruit was not affected. Sporulation and
some control of decay by Penicilltum digitatum and
P. italicum were noted in open storage boxes of
lemon and orange exposed to ozone at 1960^ig/m3
(1 ppm).51 Ozone was more effective than a
fungicide dip in controlling Botryt/s bud rot of
gladiolus, but no concentrations were given.181 In
general, researchers have suggested that rather
large dosages of ozone are required to protect
storage fruits from fungal infection. These
concentrations may be so high as to preclude the
use of ozone in storage facilities. The ability of
ozone to reduce spore germination in fungi
apparently depends on species, spore morphology,
moisture, and substrate.106 Single-celled spores
and those with thin cell walls are most sensitive
Wet spores are more sensitive than dry spores.
280
-------
Hibben127 found ozone toxic to moist fungus
spores of some species, even at concentrations of
200 /vg/m3 (0.1 ppm). Exposure to 980 and 1960
fjg/m3 (0.5 and 1.0 ppm) reduced or prevented
germination of spores of all species tested. Ozone
at 200/vg/m3(0.1 ppm) for 4 hr oral 1960/vg/m3
(1.0 ppm, for 2 hr stopped apical cell division of
conidiophores of Alternaria solani and caused
collapse of the apical cell wall.248
Heagle106 has reviewed the effects of ozone on
fungus growth, sporulation, and germination.
Ozone may inhibit colony growth on artificial
media, but it rarely causes death, even at high
concentrations Differences in species
susceptibility are known. In several fungi,
exposure to ozone at 196 or 785 /vg/m3 (0.10 or
0.40 ppm) for 4 hr caused a 10- to 25-fold increase
in sporulation.106 Heagle107 also reported the
effects on three obligate parasitic fungi of low
exposures to ozone. Germination of spores was not
affected in any of these studies. Kuss162 grew 30
representative fungi on agar and found that spore
production in some species was increased after
exposure to ozone.
Rabotnova et al.240 exposed two species of yeast
to ozone: Candida lipolytica was sensitive, and C.
auilliermondii was resistant. The biocidal activity
of ozone was determined under various cultural
conditions with air streams of about 294,000 to
10,780,000 /vg/m3 (1 50 or 5500 ppm) (v/v for 10
to 30 min). Ozone was an effective biocide under
most conditions. Its effectiveness increased with
decreasing pH of the culture medium. Kanoh151
found that exposure to ozone at 58,800/vg/m3 (30
ppm) for 30 min increased oxygen uptake in slime
mold homogenate from Physarum polycephalum.
Ozone also increased succmoxidase activity and
inhibited part of glycolysis. Effects of ozone on
Euglena gracilis that were reported by de Koning
and Jegier61'64 included reduction of net
photosynthesis, increase in respiration, and
effects on pyridine nucleotide reduction and
phosphorylation. They also reported that the
reduction of net photosynthesis was a logarithmic
function of ozone concentration in 1 -hr exposures.
These investigators also found a 5-percent
reduction in oxygen evolution after a 1 -hr exposure
to 980 /vg/m3 (0.5 ppm) ozone bubbled into 5 ml of
solution and an additive effect with a mixture of
sulfur dioxide and ozone.63 Verkroost303 carried out
a detailed study of the effects of ozone on
Scenedesmus obtusiusculus, Chod. with special
concern over the effects on photosynthesis and
respiration. A major weakness in this study was
the use of an air stream containing ozone at
58,800 fjg/m3 (150 ppm). The report suggested
that the primary site of ozone action is the
membrane structure, which produces changes in
photosynthesis and respiration.
Giese and Christiansen94 found that protozoa in
hanging drop suspensions exposed to
approximately 8 percent ozone in ozonized water
were killed within a period of 4min(Co/p/d/um)toa
maximum of 64 min (Tillina).
Names102 reported that ozone at 4 ppm retarded
the growth of Escherichia coli, whereas 1 9,600
/vg/m3 (10 ppm) prevented growth. Scott and-
Lesher254 found that approximately 2 x 107
molecules of ozone per bacterium killed 50 percent
of the cells of E. coli and that the primary effect
was on the cell membrane. Elford and van den
Ende76 reported that ozone at 390/vg/m3(0.2ppm)
had a lethal effect on some bacteria deposited from
aerosol mists on various surfaces. Relative
humidity is an important factor, particularly when
ozone concentration is low. They found little death
at a humidity below 45 percent at ozone
concentrations of 1 960/vg/m3 (1 ppm), as opposed
to a 90-percent kill in 30 minutes at 40 /vg/m3
(0.025 ppm) ozone with a humidity of around 70
percent. A 5-min exposure of Bacillus cere us to
ozone at 0.12 mg/hter was the minimal lethal
dose, whereas 0.10 mg/liter was effective for B.
megaterium and E. coli.62 Spores of the Bacillus sp.
were killed by ozone at 2.29 mg/liter. In most of the
research studying the effects of ozone on lower
organisms, ozone has been used as a germicide or
as part of an attempt to understand the
interactions of pollutants and pathogens on higher
plant response.
SUMMARY
Oxidant injury to vegetation was first identified
in 1944 in the Los Angeles Basin. Our
understanding of oxidant effects and of the
widespread nature of their occurrence has
increased steadily since then. Although the major
phytotoxic components of the photochemical
oxidant complex are ozone and peroxyacetyl nitrate
(PAN), some data suggest that other
phytotoxicants are also present. The
peroxyacylnitrates are the most phytotoxic of the
known photochemical oxidants; however, the
ubiquitous nature of ozone and its association with
widespread injury to vegetation make it the most
281
-------
important phytotoxic component of the
photochemical oxidant complex
The effects of photochemical oxidants on
vascular plants can be envisioned as occurring at
several levels, ranging from the subcellular to the
organismal, depending on the concentration and
duration of exposure to the pollutant, and on the
elapsed time after exposure that effects are
observed.
The earliest effect is an increase m cell
membrane permeability. Cellular and biochemical
changes are ultimately expressed on the
organismic level in visible foliar injury, increased
leaf drop, reduced plant vigor, reduced plant
growth, and death. In the final analysis,
biochemical modifications on an individual plant
level are manifested by changes in plant
communities and then in whole ecosystems.
Leaf stomata are the principal plant entry sites
for ozone and PAN. Oxidants affect such
physiologic processes as photosynthesis,
respiration, transpiration, stomatal opening,
metabolic pools, biochemical pathways, and
enzyme systems.
Visible injury is identifiable as pigmented,
chlorotic, or necrotic foliar patterns Metabolic
cellular disturbances can occur without visible
injury and may be reversible However, most of the
growth effects reported until recently were
associated with visible injury.
Classic ozone injury is demonstrated by the
upper-surface leaf fleck on tobacco and the leaf
stipple of grape. Many plants show an upper-
surface response with no associated injury to the
lower surface of leaves However, in
monocotyledonous plants such as grasses or
cereals, and some nonmonocotyledonous plants,
the mesophyll tissue is not divided, and bifacial
necrotic spotting (flecking) is a common symptom
of ozone injury.
Coniferous trees exhibit different symptoms.
Ozone is probably the cause of emergence tipburn
in eastern white pine (white pine needle dieback)
and chlorotic decline, a needle injury of ponderosa
pine.
Classic PAN mjury appears as a glaze followed
by bronzing of the lower leaf surface of many
plants Complete collapse of leaf tissue can occur if
concentrations are sufficiently high. Early leaf
senescence and abscission usually follow the
chronic response symptoms. Chronic injury
patterns generally are not characteristic and may
be confused with symptoms caused by biotic
diseases, insect infestation, nutritional disorders,
or other environmental stresses.
A great deal of research has been done to define
the effects of oxidants on plant growth and yield
more accurately Studies comparing the growth of
plants in filtered and nonfiltered field chambers
using oxidants in the ambient air have reported up
to 50-percent decreases in citrus yield (orange and
lemon); 10 to 15 percent supression in grape yield
in the first year, and 50 to 60 percent reduction
over the following 2 years; and a 5 to 29 percent
decrease in yield of cotton lint and seed m
California Losses of 50 percent in some sensitive
potato, tobacco, and soybean cultivars have been
reported in the eastern United States. It is apparent
that oxidants in the ambient air reduce yields of
many sensitive plant cultivars.
Experimental chambers with controlled
environments have been used to study both short-
term and long-term effects of exposure to ozone
(Tables 11 -2 and 11 -4). Multiple acute exposures
(785 fjg/m3, or 0.40 ppm) of radishes for 1.5 hr
resulted m reductions in root growth. The
reductions in root growth from the multiple ozone
exposures were equal to the additive effects of
three single exposures. When soybean plants
were exposed to 1468 /jg/m3 (0.75 pprn) ozone for
1 hr, root growth was consistently reduced more
than top growth. There was also a reduction in
nodule weight and number The greater reduction
of root growth than top growth is related to the
transport of photosynthate. Ozone also affects the
process of nitrogen fixation in clover, soybean, and
pinto bean through reduction in nodule number,
but neither nodule size nor efficiency of nitrogen
fixation is influenced. The effect of ozone on the
nitrogen fixation process in legumes, if
widespread, could have a major impact on plant
communities and affect fertilizer requirements.
There are indications that the effect of ozone on
nodulation may be related to carbohydrate supply
in host plants.
Experimental long-term exposures of a variety of
crops as well as ornamental and native plants to
ozone have resulted in a reduction in growth
and/or yield. Throughout a growing season, 14
species representative of the aspen plant commu-
nity were exposed to ambient air (98 to 1 37 /ug/m3
ozone, or 0.05 and 0.07 ppm) and to ozone (290 and
588^g/m3 or 0.15 and 0.30 ppm) for 3 hr/day, 5
days/week, and to charcoal-filtered air. Foliar
injury to all species resulted at the highest pollution
concentration. The growth of two soybean cultivars
282
-------
(Hood and Dare) was inhibited by intermittent
exposure to ozone at 196/ug/m3 {0.10 ppm) for 3
weeks. Growth of both root and top was decreased.
Similar results were noted with radish, except that
a lower concentration (98 /^g/m3 or 0.05 ppm)
inhibited growth. In the aforementioned studies,
the reduced growth occurred even though there
were very few visible symptoms of plant injury.
A reduction of 30 percent in the yield of wheat
occurred when, atanthesis, wheat was exposed to
ozoneat 392 ftg/m3(Q,2 ppm), 4hr/day for 7 days.
A significant reduction m the yield of tomato was
noted when plants were experimentally treated
with ozone at 686 (Ug/m3 (0,35 ppm). Lower fruit
set and fewer harvested fruit caused the reduction.
Chronic exposures to ozone at 98 to 290 ^g/m3
(0.05 to 0.15 ppm) for 4 to 6 hr/day produced
reductions in yield in soybean and corn grown
under field conditions. The threshold for meas-
urable effects for ozone appears to be between 98
and 196 /xg/m3 (0.05 to 0.10 pprn) for sensitive
plant cultivars. This is well within the range of
ozone levels monitored in the eastern United
States Growth or flowering effects were reported
for carnation, geranium, radish, and pinto bean
grown in greenhouse chambers exposed to ozone
at 98 to 294 ^g/m3 (0 05 to 0.15 ppm) for 2 to 24
hr/day
The two most critical factors in determining plant
response to air pollution are duration of exposure
and concentration of pollutants. These two factors
describe exposure dose. In determining the re-
sponse of vegetation to oxidants, concentration is
more important than time. Any given dose pre-
sented to a plant in a short period of time has a
greater effect than the same dose applied over a
longer period
The concept of limning values was used by
Jacobson to define the boundary between doses of
a pollutant that are likely to cause measurable
foliar injury to vegetation and those that are not.
Foliar injury was used as the index of plant re-
sponse, The ranges for limiting values for effects of
ozone are.
1. Trees and shrubs.
400 to 1000 ^g/m3 (0.2 to 0.51 ppm) for 1
hr
200 to 500 /jg/m3 (0.1 to 0.251 ppm) for 2
hr
1 20 to 340 /jg/m3 (0.06 to 0.17 ppm) for 4
hr
2. Agricultural crops:
400 to 800 jug/m3 (0.2 to 0.41 ppm) for 0.5
hr
200 to 500 pg/m3 (0.1 to 0.251 ppm) for 1
hr
75 to 180 jug/rn3 {0.04 to 0.09 ppm) for 4 hr
The range of limiting values for PAN is:
1000 /jg/m3 (0.2 ppmj for 0.5 hr
500/jg/m3 {0.1 ppm) for 1 hr
175 /jg/m3 (0.035 pprn) for 4 hr
Doses of ozone or PAN greater than the upper
limiting values are likely to cause foliar injury.
The data points used to determine the limiting
values listed above are not necessarily threshold
values but were based on available published
research data. More than 200 studies were
surveyed. Any limitations that were present in the
experimental techniques used in the studies are
expressed in the data points. Another constraint of
the PAN data is the limited number of studies.
Plant sensitivity to ozone and PAN is conditioned
by many factors. Genetic diversity in sensitivity to
ozone between species and cultivars within a
species is well documented. Variants in sensitivity
to ozone within a natural species are well known
for several pine species, including white, loblolly,
and ponderosa. Plant sensitivity to oxidants can be
changed by both climatic and edaphic factors. A
change in environmental conditions can initiate a
change in sensitivity at once, but it will be 3 to 5
days before the response of the plant is totally
modified. Plants generally are more sensitive to
ozone when grown under short photopenods,
medium light conditions, medium temperature,
high humidity, and high soil moisture. Injury from
PAN may increase with increasing light intensity,
Conditions during exposure and growth affect the
response of plants to oxidants in similar ways. In
general, environmental conditions optimum for
plant growth tend to increase the sensitivity to
ozone. At the time of exposure, factors that
increase water stress tend to make plants more
tolerant to ozone. Soil moisture is probably the
most important environmental factor that affects
plant response to oxidants during the normal
growing season. Physiologic age affects the
response of the leaf to oxidants, Young leaf tissue
is most sensitive to PAN, whereas newly
expanding and maturing tissue is most sensitive to
ozone. Light is required for plant tissue to respond
to PAN; a similar light requirement is not needed
for plants to respond to ozone,
The majority of effects observed, such as
suppression of root growth, mineral uptake, and
nitrogen fixation, apparently result from a
283
-------
suppression in photosynthesis and modifications
in photosynthate distribution. This suppression in
metabolic reserves ultimately slows plant growth
and renders the plant more sensitive to other
stresses. Physiological changes can provide a
sensitive means of monitoring the health and vigor
of the plant with or without visual injury. Ozone
affects pollen germination in some species and
may affect yield through incomplete pollination of
flowers. Investigations with Arabidopsis thaliana
showed no mutagenic effects from ozone over
seven generations.
Mixtures of pollutants can cause effects below
the levels caused by either gas alone; however,
there is some disagreement concerning ozone
interactions with other gases. Ratios of gas
mixtures, intermittent exposures, sequential
exposures to pollutants, and predisposition by one
pollutant to the effects of a second pollutant may
be important in nature, but insufficient knowledge
is available to elucidate the effects.
The response of plants to oxidants may be
conditioned by the presence or absence of biotic
pathogens. Depending on the plant and the
pathogen, oxidants may cause more or less injury
to a given species. Oxidant injury to ponderosa
pine predisposes the trees to later invasion by bark
beetles. Ozone and ozone-sulfur dioxide mixtures
can decrease the population of some plant-
parasitic nematodes. Variable plant responses
have been noted when herbicides were used in the
presence of high oxidant concentrations.
Little research on the effects of oxidants on
ferns, nonvascular green plants, and
microorganisms has been reported. Lichens and
mosses are responsive to acid gases, but there is
no definite evidence that they respond to oxidants.
Ferns may be especially sensitive, but their injury
response is different from that of higher vascular
plants. Growth and sporulation of fungi on
surfaces are usually, but not always, affected.
Ozone from 0.1 to several milligrams per liter of
solution is required to kill many microorganisms in
liquid media. Most work with microorganisms has
been done to study the effectiveness of ozone as a
biocide in the storage of vegetation or treatment of
water or sewage supplies.
REFERENCES FOR CHAPTER 11
1 Abeles, A L , and F B Abeies Biochemical pathway of
stress-induced ethylene Plant Physiol 5O 496-498,
1972
2 Adams, D F The effects of air pollution on plant life A
review AMA Arch Ind Health 14 1 99-245, 1956
3, Adedipe, N. 0., R E Barrett, and D P Ormrod
Phytotoxicity and growth responses of ornamental
bedding plants to ozone and sulfur dioxide J Am Soc
Hort Sci 97341-345, 1972
4 Adedipe, N. O., R, A. Fletcher, and D. P Ormrod. Ozone
lesions in relation to senescence of attached and
detached leaves of tobacco Atmos Environ 7357-361,
1973.
5 Adedipe, N. 0 , G. Hofstra, and D P Ormrod Effects of
sulfur nutrition on phytotoxicity and growth responses
of bean plants to ozone Can J Bot 501789-1793,
1972.
6. Adedipe, N. O., H. Khatamian, and D P Ormrod
Stomatal regulation of ozone phytotoxicity in tomato Z
Pflanzenphysioi 68 323-328, 1973
7. Adedipe, N 0., and D P Ormrod Ozone-induced
growth suppression in radish plants in relation to pre-
and post-fumigation temperatures. Z Pflanzen Physiol
7/-281-287, 1974.
8. Air Pollution Control Board. Technical and
Administrative Report on Air Pollution Control in Los
Angeles County, 1949-1950 County of Los Angeles, Air
Pollution Control District, Los Angeles, Catif, 1950
9 Air Pollution Control Board Second Technical and
Administrative Report on Air Pollution Control in Los
Angeles County, 1950-1951 County of Los Angeles, Air
Pollution Control District, Los Angeles, Calif . 1951
10 Applegate, H. G., and L C Durrani Synergisticactionof
ozone-sulfur dioxide on peanuts Environ Sci Technol
3759-760, 1969
11 Benedict, H.M , C J. Miller, and R E Olson Economic
Impact of Air Pollutants on Plants in the United States
APTD-0953, U S. Environmental Protection Agency,
Research Triangle Park, N C , November 1971
12 Bennett, J. P, H M Resh, and V C Runeckles
Apparent stimulations of plant growth by air pollutants
Can J Bot. 5235-42, 1974
13 Berry, C R Age of pine seedlings with primary needles
affects sensitivity to ozone and sulfur dioxide
Phytopathology 64 207-209, 1974
14 Berry, C R. Relative sensitivity at red, jack, and white
pine seedlings to ozone and sulfur dioxide
Phytopathology 61 231-232, 1971
15. Berry, C R White pine emergence tipburn, a
physiogenic disturbance US For Serv Southeast For
Exp. Stn. Stn Pap (130)1-8,1961
16 Berry, C. R, and H E Heggestad Air Pollution
Detectives in Yearbook o1 Agriculture 1968 Science
for Better Living U S Government Printing Office,
Washington. D C , 1968 pp 142-146
17 Berry, C R , and L A. Ripperton. Ozone as possible
cause of white pine emergence tipburn Phytopathology
53552-557, 1963
18 Blum, U., and D T Tmgey A study ot the potential ways
by which ozone could reduce root growth and
nodulation of soybean Atmos Environ 11 737-739,
1977
19 Bobrov, R A. Cork formation in table beet leaves {Beta
vulgans) in response to smog In Proceedings of the
Third National Air Pollution Symposium National Air
Pollution Symposium, Los Angeles, Calif, 1955 pp
199-206
20 Bobrov, R. A The anatomical effects of air pollution on
plants In Proceedings of the Second National Air
284
-------
Pollution Symposium National Air Pollution
Symposium, Los Angeles, Calif , 1952 pp 129-134
21 Bobrov, R. A The effect of smog on the anatomy of oats
leaves. Phytopathology 42 558-56*3, 1952
22 Bobrov, R. A The leaf structure of Poa annua with
observations of its smog sensitivity in Los Angeles
County Am J. Bot 42.467-474, 1955
23 Botkm, D. B ,W.H. Smith, R W. Carlson, andT L Smith
Effects of ozone on white pine saplings Variation in
inhibition and recovery of net photosynthesis. Environ
Pollut 3.273-289, 1972.
24 Brandt, C S.andW W Heck Effects of air pollutants on
vegetation In Air Pollution Volume I. Air Pollution and
Its Effects Second Edition. A C. Stern, ed Academic
Press, Inc., New York, 1968 pp 401-443.
25 Brasher, E P , D J. Fieldhouse, and M Sasser Ozone
injury in potato variety trials Plant Dis. Rep 57 542-
544, 1973
26 Brennan, E., and P M. Halisky Response of turfgrass
cultivars to ozone and sulfur dioxide in the atmosphere
Phytopathology 60 1544-1546, 1970.
27 Brennan, E , and I A Leone Chrysanthemum response
to sulfur dioxide and ozone Plant Dis. Rep 56 85-87,
1972
28 Brennan, E , and I A Leone. The response of English
holly selections to ozone and sulfur dioxide. Holly Lett
376-7, 1970
29 Brennan, E , and I A Leone The response of plants to
sulfur dioxide or ozone-polluted air supplied at varying
flow rates Phytopathology 58 1661 -1664, 1968
30 Brennan, E. G , I A. Leone, and R. H Dames
Atmospheric aldehydes related to petunia leaf damage
Science 143 818-820, 1964
31 Brennan, E, I A Leone, and R. H. Dames The
importance of variety in ozone plant damage. Plant Dis.
Rep 45923-924, 1964
32. Brennan, E , I. A. Leone, and P M Halisky Response of
forage legumes to ozone fumigations Phytopathology
59-1458-1459, 1969
33. Brewer, R. F , and G. Ferry Effects of air pollution on
cotton in the San Joaquin Valley. Calif Agnc. 28:6-7.
1974
34 Brewer, R. F., F. B. Guillemet, and R. K. Creveling.
Influence of NPK fertilization on incidence and severity
of oxidant injury to mangels and spinach. Soil Sci.
32298-301, 1961
35 Bruton.V.C Environmental influences on the growth of
Arabidopsisthaliana PhD thesis, North Carolina State
University, Raleigh, N C , 1974.
36 Burleson, G R , T M. Murray, and M. Pollard
Inactivation of viruses and bacteria by ozone, with and
without sonication. Appl. Microbiol. 29:340-344, 1975
37 Bystrom, B G , R B Glater, F. M Scott, and E S. C
Bowler Leaf surface of Beta vulgans—electron
microscope study Bot. Gaz (Chicago) 129 133-138,
1968
38 Cameron, J W , H Johnson, Jr, O C Taylor, and H W
Otto Differential susceptibility of sweet corn hybrids to
field injury by air pollution Hort. Sci. 5'21 7-219, 1970.
39 Carney, A W,G R Stephenson, D. P. Ormrod.andG C
Ashton Ozone-herbicide interactions in crop plants
Weed Sci 2/.508-511, 1973.
40 Cathey, H M , and H E. Heggestad. Reduction of ozone
damage to Petunia hybnda Vilm. by use of growth-
regulating chemicals and tolerant cultivars J Am Soc
Hort Sci 37695-700, 1972
41 Clayberg, C D Evaluation of tomato varieties for
resistance to ozone Conn Agnc Exp Stn New Haven
Circ (246)1-11, 1972
42 Clayberg, C D Screening tomatoes for ozone
resistance Hort Sci 6396-397, 1971
43 Cobb, F W.Jr.andR W Stark Declmeand mortalityof
smog-injured ponderosa pine J For 68 147-149,
1970
44 Cole, G A Air pollution with relation to agronomic
crops III Vegetation survey methods in air pollution
studies Agron J 50553-555, 1958
45 Comeau, G , and F LeBlanc Influence de I'ozone et de
I'anhydnde sulfureaux sur la regeneration des feuilles
de Funana hygrometnca HEDW Nat Can (Que ) 98
347-358, 1971
46 Costonis, A C Acute foliar injury of eastern white pine
induced by sulfur dioxide and ozone. Phytopathology
60994-999, 1970
47 Costonis, A. C Injury to eastern white pine by sulfur
dioxide and ozone alone and in mixture Eur J For
Pathol 3 50-55, 1973
48 Costonis, A C, and W A Sinclair Relationships of
atmospheric ozone to needle blight of eastern white
pine Phytopathology 59 1566-1574, 1969
49 Craker, L E Decline and recovery of petunia flower
development from ozone stress Hort Sci 7484,1972
50 Craker, L E Ethylene production from ozone-injured
plants Environ Pollut / 299-304, 1971.
51 Craker, L E Influence of ozone on RNA and protein
content of Lemna minor L Environ Pollut 3319-323,
1972
52 Craker, L E , and W A Feder Development of the
inflorescence in petunia, geranium, and pomsettia
under ozone stress Hort. Sci 7.59-60,1972
53 Czuba, M , and D P Ormrod Effects of cadmium and
zinc on ozone-induced phytotoxicity in cress and lettuce
Can J. Bot 52.645-649, 1974
54 Dames, R H,l A Leone, and E Brennan Air pollution
as it affects agriculture in New Jersey N J Agnc Exp
Stn Bull (794) 1-14, 1960
55 Davis, D. D The Influence of Ozone on Conifers Ph D
thesis, Pennsylvania State University, University Park,
Pa., 1970
56 Davis, D D.andJ B Coppolmo. Relationship between
age and ozone sensitivity of current needles of
ponderosa pine Plant Dis. Rep 58 660-663, 1974
57 Davis, D D, and J B Coppolmo Relative ozone
susceptibility of selected woody ornamentals Hort Sci
9.537-539, 1974.
58 Davis, D D , and L Kress. The relative susceptibility of
ten bean varieties to ozone Plant Dis Rep 58 14-16,
1974
59 Davis, D D, and F A Wood The influence of
environmental factors on the sensitivity of Virginia pine
to ozone Phytopathology 63 371-376, 1973
60 Davis, D D , and F A Wood The relative susceptibility
of eighteen coniferous species to ozone Phytopathology
62 14-19, 1972
61 de Konmg, H W , and Z Jegier A study of the effects of
ozone and sulfur dioxide on the photosynthesis and
285
-------
respiration of Euglena graci/is Atmos Environ 2321-
326, 1968
62 de Komng, H W , and Z Jegier Effect of ozone on
pyndine nucleotide reduction and phosphorylation of
Euglena gracilis Arch Environ Health 18 913-916,
1969
63 de Komng, H W.andZ Jegier Effects of sulfur dioxide
and ozone on Euglena gracilis Atmos Environ 4357-
361,1970
64 de Komng, H W , and Z Jegier Quantitative relation
between ozone concentration and reduction of
photosynthesis of Euglena gracilis Atmos Environ
2615-616, 1968
65 Dean, C E Stomata density and size as related to ozone-
induced weatherfteckin tobacco Crop Sci 12 547-548,
1972
66 Dean, C E Weather fleck on Florida shadegrown
tobacco. Tob Sci 741-45,1963
67 Dean, C E , and D R Davis Ozone and soil moisture in
relation to the occurrence of weather fleck on Florida
cigar-wrapper tobacco in 1966 Plant Dis Rep 51 72-
75, 1967.
68 Dochmger, L S The impact of air pollution on eastern
white pine The chlorotic dwarf disease J Air Pollut
Control Assoc 18 81 4-81 6, 1 968
69. Dochmger, L S.andC E Seliskar Air pollution and the
chlorotic dwarf disease of eastern white-pine For Sci
16 46-55, 1970
70 Dochmger, L S,F W Bender.F L Fox.andW.W Heck
Chlorotic dwarf of eastern white pine caused by an
ozone and sulphur dioxide interaction. Nature (London)
225476, 1970
71 Drummond, D B The effect of peroxvacetyl nitrate on
petunia (Petunia hybrida Vilm ) CAES Publication No
260-72, Center for Air Environment Studies,
Pennsylvania State University, University Park, Pa ,
June 1972
72 Dugger, W M , Jr , O C Taylor, W H Klein, and W
Shropshire, Jr Action spectrum of peroxyacetyl nitrate
damage to bean plants Nature (London) 198 75-76,
1963.
73 Dugger, W.M. Jr,O C Taylor, C R Thompson, and E
Cardiff The effect of light on predisposing plants to
ozone and PAN damage J Air Pollut Control Assoc
73-423-428, 1963
74 Dunning, J A.andW W Heck Response of pinto bean
and tobacco to ozone as conditioned by light intensity
and/or humidity Environ Sci Technol. 7824-826,
1973
75. Dunning, J. A , W W Heck, and D T Tmgey Foliar
sensitivity of pinto bean and soybean to ozone as
affected by temperature, potassium nutrition and ozone
dose Water Air Soil Pollut 3 305-313, 1974
76 Elford, W. J _.nd J van den Ende An investigation of
the merits of ozone as an aerial disinfectant J Hyg
42240-265, 1942
77 Engle, R L Reaction of Onion, Allium cepa L , and Pmto
Bean, Phaseolus vulgans L , to Ozone. Ph D thesis,
University of Wisconsin, Madison, Wis., 1966
78 Engle, R L , and W H Gabelman Inheritance and
mechanism for resistance to ozone damage in onion,
Allium cepa L. Proc Am Soc Hort Sci 89 423-430,
1966
79 Engle, R L , and W H Gabelman The effects of low
levels of ozone on pinto beans, Phaseolus vulgaris L
Proc Am Soc Hort Sci 91 304-309, 1967
80 Engle, R L,W H Gabelman, and R R Romanowski, Jr
Tipburn, an ozone incited response in onion, Allium
cepa L Proc Am Soc Hort Sci 86 468-474, 1965
81 Evans, L S Bean leaf growth response to moderate
ozone levels Environ Pollut 417-26,1973
82 Evans, L S , and I P Ting Ozone sensitivity of leaves
Relationship to leaf water content, gas transfer
resistance, and anatomical characteristics Am. J Bot
61 592-597, 1974
83 Feder, W A Plant response to chronic exposure of low
levels of oxidant-type air pollution Environ Pollut 1 73-
79, 1970
84 Feder, W A , and F J Campbell Influence of low levels
of ozone on flowering of carnations Phytopathology
58 1038-1039, 1968.
85 Feder, W A,F L Fox, W W Heck, and F J Campbell
Varietal responses of petunia to several air pollutants
Plant Dis Rep 53.506-510, 1969
86 Feder, W A , and F Sullivan Ozone. Depression of
frond multiplication and floral production in duckweed
Science 165 1373-1374, 1969
87 Feliciano, A 1971 Survey and Assessment of Air
Pollution Damage to Vegetation in New Jersey EPA-
R5-72-010, U S Environmental Protection Agency,
Research Triangle Park, N C , October 1972.
88 Ferry, B W., M S Baddeley.andD. L.Hawksworth, eds
Air Pollut ion and Lichens Athlone Press, London, 1973.
pp 39-88, 109-123.
89 Frey, J E The Effects of Ozone on the Synthesis and
Storage of Nutrients in Soybean Seeds Ph D thesis.
University of Northern Colorado, Greeley, Colo., 1972
90 Fujiwara, T Damage to plants caused by complex
pollution of the atmosphere Shokubutsu Boeki.27,233-
236, 1973
91. Gardner, W S Ozone injury to tobacco plants in South
Dakota Plant Dis Rep 57106-110,1973
92 Gentile, A G,W A Feder, R E Young, and Z Santner
Susceptibility of Lycopersicon spp to ozone injury J
Am Soc Hort Sci 9694-96, 1971.
93. Gibson, A H Limitation to dmitrogen fixation by
legumes In Nitrogen Fixation W. E Nexton and C J
Nyman, eds Washington State Univ Press, Pullmab,
Wash 1976 pp 400-428
94 Giese, A C , and E Christensen. Effects of ozone on
organisms Physiol Zool. 27-101-1 15, 1954
95 Glater.R A B Smog damage to ferns mtheLos Angeles
area Phytopathology 45.696-698, 1956
96 Glater, R B., R A Solberg, and F M Scott A
developmental study of the leaves of Nicotians
g/utinosa as related to their smog-sensitivity Am. J
Bot 49954-970, 1962
97 Grosso, J J , H A Menser, Jr, G. H Hodges, and H. H
McKmney Effects of air pollutants on Nicotiana
cultivars and species used for virus studies
Phytopathology 61:945-950, 1971.
98. Gudenan, R., H van Haul, and H. Stratmann Probleme
der Erfassung und Beurteilung von Wirkungen
gasformiger Luftverunremigungen auf die Vegetation
Z. Pflanzenkr Pflanzenpathol Pflanzenschutz 67.257-
264, 1960
99 Gumarch, P Three years of ozone sterilization of water
286
-------
in Pans. In Ozone Chemistry and Technology, Adv 119
Chem. Ser. 27,416-429, 1957.
100. Haagen-Smit, A J Chemistry and physiology of Los
Angeles smog. Ind Eng. Chem. 44 1342-1346, 1952
101. Haas, J. H Relation of crop maturity and physiology to 120,
air pollution incited bronzing of Phaseolus vulgans
Phytopathology 60:407-410, 1970.
102 Haines, R. B. Effect of pure ozone on bacteria In Report
of the Food Investigation Board for the Year 1935 His
Majesty's Stationery Office, London, 1936. pp. 30-31,
103 Hanson, G P. Relative air pollution sensitivity of some 121
Los Angeles Arboretum plants Lasca Leaves 22 86-89,
1972
104 Hanson, G, P, L. Thorne, and D H Addis Ozone
sensitivity of Petunia hybnda Vi/m. as related to
physiological age J Am. Soc, Hort. Sci. 700 188-190, 122
1975
105. Harward, M , and M. Treshow Impact of ozone on the
growth and reproduction of understory plants in the 123.
aspen zone of western U.S A. Environ. Conserv. 2 17-
23, 1975
106 Heagle, A S Interactions between air pollutants and
plant parasites. Annu, Rev Phytopathol 11 365-388,
1973 124.
107 Heagle, A. S Response of three obligate parasites to
ozone Environ Pollut 3,91-95, 1975.
108 Heagle, A. S , D. E Body, and W W Heck An open-top 125
field chamber to assess the impact of air pollution on
plants. J Environ, dual 2365-368,1973
109 Heagle, A S ,D E Body.andG E Neely Injury and yield 126
responses of soybean to chronic doses of ozone and
sulfur dioxide in the field Phytopathology 64 132-1 36, 1 27
1974.
110. Heagle, A. S., D. E, Body, and E K Pounds, Effect of 128,
ozone on yield of sweet corn Phytopathology 62 683-
687, 1972
111 Heagle, A S , and W.W Heck. Predisposition of tobacco 129
to oxidant air pollution injury by previous exposure to
oxidants Environ Pollut 7247-252, 1974
112 Heagle, A S , W W Heck, and D. Body Ozone injury to
plants as influenced by air velocity during exposure
Phytopathology 6/1209-1212, 1971 130.
113 Heath, R L Ozone. In Response of Plants to Air
Pollution. J B MuddandT T Kozlowski, eds Academic
Press, Inc, N Y, 1975 pp 23-55 131
114 Heck, W. W Factors influencing expression of oxidant
damage to plants Annu Rev. Phytopathol 6165-188,
1968 132
115 Heck, W. W Plant injury induced by photochemical
reaction products of propylene-nitrogen dioxide
mixtures J Air Pollut Control Assoc, 74.255-261, 133.
1964.
116 Heck, W. W The use of plants as indicators of air
nollution. Air Water Pollut W 99-11 1, 1966
117 Heck, W W.andC S Brandt Effects on vegetation 134
Native, crops, forests In Air Pollution Volume II The
Effects of Air Pollution Third edition A C Stern, ed
Academic Press, Inc , N Y., 1978 pp 157-229 135.
118 Heck, W.W, R H Dames, and 1 J Hmdawi Other
phytotoxic pollutants In Recognition of Air Pollution
Injury to Vegetation A Pictorial Atlas J S Jacobson
and A C Hill, eds Informative Report No. 1, Air 136
Pollution Control Association, Pittsburgh, Pa , 1970 pp
F1-F24
Heck, W W., andJ.A Dunning The effects of ozone on
tobacco and pinto bean as conditioned by several
ecological factors. J. Air Pollut Control Assoc 77.112-
114, 1967
Heck, W W , F. L Fox, C S Brandt, and J A. Dunning
Tobacco, A Sensitive Monitor for Photochemical Air
Pollution NAPCA Publication No AP-55, US.
Department of Health, Education and Welfare, Public
Health Service, National Air Pollution Control
Administration, Cincinnati, Ohio, June 1969
Heck, W W., and D. T. Tingey Ozone Time-
concentration model to predict acute foliar injury In,
Proceedings of the Second International Clean Air
Congress H M Englund and W T Beery, eds
Academic Press, Inc., N.Y., 1971 pp 249-255.
Heggestad, H E. Photochemical air pollution injury to
potatoes in the Atlantic coastal states Am Potato J
50315-328, 1973.
Heggestad, H E,F R Burleson.J T Middleton, and E
F, Darley Leaf injury on tobacco varieties resulting from
ozone, ozonated hexene-1 and ambient air of
metropolitan areas Int J Air Water Pollut S.I-10,
1964
Heggestad, H E , and H A Menser Leaf spot-sensitive
tobacco strain Bel W-3, a biological indicator of the air
pollutant ozone Phytopathology 52.735, 1962.
Heggestad, H E , and J T Middleton Ozone in high
concentrations as cause of tobacco leaf injury Science
729 208-210, 1959.
Hibben, C R Ozone toxicity to sugar maple.
Phytopathology 53.1423-1428, 1969
Hibben, C R Sensitivity of fungal spores to sulphur
dioxide and ozone. Phytopathology 56.880-881, 1966
Hill, A.C, A special purpose plant environmental
chamber for air pollution studies J Air Pollut Control
Assoc 77743-748, 1967.
Hill, A C , H. E. Heggestad, and S N Linzon. Ozone In
Recognition of Air Pollution Injury to Vegetation. A
Pictorial Atlas J S Jacobson and A. C. Hill, eds
Informative Report No 1, Air Pollution Control
Association, Pittsburgh, Pa , 1970 pp B1-B22
Hill, A C , and N Lmlefield Ozone Effect on apparent
photosynthesis, rate of transpiration, and stomatal
closure in plants Environ Sci Technol 3 52-56, 1969
Hill, A C , M R. Pack, M Treshow. R J Downs, and L G
Transtrum Plant injury induced by ozone
Phytopathology 57 356-363, 1961
Hmdawi, I J.,andA P Altshuller Plantdamage caused
by irradiation of aldehydes Science 746.540-542,
1964
Hmdawi, I J., J. A. Dunning, and C. S. Brandt.
Morphological and microscopical changes in tobacco,
bean and petunia leaves exposed to irradiated auto
exhaust Phytopathology 55 27-30, 1965
Hodges, G H , H A Menser, Jr , and W B Ogden
Susceptibility of Wisconsin Havana tobacco cultivars to
air pollutants Agron J 63107-111,1971
Hodgson, R H Alteration of triaztne metabolism by
ozone Paper No 28, presented at the Weed Science
Society of America, Montreal, Quebec, February 3,
1970
Hodgson, R H, K E Dusbabek, and B L Hoffer.
Diphenamid metabolism in tomato Time course of an
ozone fumigation effect Weed Set 22 205-210, 1974
287
-------
137 Hodgson, R H, D S Frear, H R Swanson, and L A
Regan. Alteration of diphenamid metabolism in tomato
by ozone Weed Sci 27 542-549, 1973 157
138 Hoffman, G J,E V Mass, and S L Rawlins Salinity-
ozone interactive effects on alfalfa yield and water
relations J Environ Qual 4326-331,1975
139 Hoffman, G J,E V Mass, and S L Rawlins Salinity- 158
ozone interactive effects on yield and water relations of
pinto bean J Environ Qual 2 148-152, 1973
140 Houston, D A Response of selected Pinus strobus L
clones to fumigations with sulfur dioxide and ozone 159
Can J For Res 4 65-68, 1974
141 Howell, R K,T E Devme.andC H Hanson Resistance
of selected alfalfa strains to ozone Crop Sci 77.114- 160
115, 1971
142 Howell, R K , and D F Kremer Ozone mjury.to soybean
cotyledonary leaves J Environ Qual 794-97,1972. 161
143 Howell, R K , and C A Thomas Relative tolerances of
twelve safflower cultivars to ozone Plant Dis Rep 56
195-197, 1972
144 Hull, H H , and F W Went Life processes of plants as 162
affected by air pollution In Proceedings of the Second
National Air Pollution Symposium National Air
Pollution Symposium, Pasadena, Calif, 1952 pp 122- 163
128
145 Jacobson, J S The effects of photochemical oxidants
on vegetation In Ozone und Begleitsubstanzen im
photochemischenSmog VDI Ber (270)163-173,1977 164
146 Jacobson, J S and L J Colavito The combined effect of
sulfur dioxide and ozone on bean and tobacco plants.
Environ Exp Bot 76277-285,1976
147 Jensen, K F Response of nine forest tree species to
chronic ozone fumigation Plant Dis Rep 57914-917,
1973.
148 Jensen, K F , and L S Dochmger Responses of hybrid 165.
poplar cuttings to chronic and acute levels of ozone
Environ Pollut 6289-295, 1974
149 Juhren, M, W Noble, and F W Went The 166
standardization of Poa annua as an indicator of smog
concentrations 1 Effects of temperature, photopenod,
and light intensity during growth of test plants Plant
Physiol 32 576-586, 1957 167
150 Kadota, M , and K Ohta Ozone sensitivity of Japanese
plant species in summer, with special reference to a
tentative sensitivity grade list for applying to field survey 1 68
on ozone injury Taiki Osen Kenkyu 7 19-26 , 1972
151 Kanoh, F Ozone reaction on slime mold Tokyo Toritsu
Eisei Kenyusho Kenkyu Nempo 24 337-342, 1 972 1 69
152 Karnosky, D F Threshold levels for foliar injury to
Populus Iremuloides by sulfur dioxide and ozone Can
J For Res 6166-169,1976 170
153 Khatamian, H , N 0 Adedipe, and D P Ormrod Soil-
plant-water aspects of ozone phytoxicity in tomato
plants Plant Soil 38 531-541, 1973
154 Kmman, R N Water and wastewater disinfection with 171
ozone. A critical review Crit Rev Environ Control
5 141-152, 1975
155 Klmgaman, G L, and C B Link Reduction in air 172
pollution injury to foliage of Chrysanthemum
morifo/ium Ramala using tolerant cultivars and
chemical protectants J Am Soc Hort Sci 700173-
175, 1975 173
156 Kochhar, M Phytotoxic and Competitive Effects of Tall
Fescue on Ladino Clover as Modified by Ozone and/or
Rhizoctonia solani. Ph.D thesis, North Carolina State
University, Raleigh, N C., 1974.
Kohut, R J Response of Hybrid Poplar to Simultaneous
Exposure to Ozone and PAN CAES Publication No 288-
72, Pennsylvania State University, Center for Air
Environment Studies, University Park, Pa , 1972.
Kohut, R. J The Interaction of 03 and PAN on Hybrid
Poplar and Pinto Bean CAES Publication No 428-76,
Pennsylvania State University, Center for
Environmental Studies, University Park, Pa., 1975.
Kohut, R J.S.V Krupa, andF. Russo An open-topfield
chamber study to evaluate the effects of pollutants on
soybean yields Proc. Am Phytopathol Soc 425, 1977.
Koritz, H G ,andF W Went The physiological action of
smog on plants I Initial growth and transpiration
studies Plant Physiol 28 50-62, 1953
Kress, L W Response of Hybrid Poplar to Sequential
Exposure to Ozone and PAN CAES.Publication No 259-
272, Pennsylvania State University, Center for Air
Environment Studies, University Park, Pa , 1972
Kuss, F R The Effect of Ozone on Fungus Sporulation
M S thesis, University of New Hampshire, Durham,
N.H , 1950
Lacasse, N L Assessment of air pollution damage to
vegetation in Pennsylvania CAES Publication No 209-
71, Pennsylvania State University, Center for Air
Environment Studies, University Park, Pa 1971
Lacasse, N L , and T C Weidensaul A cooperative
extension-based system of assessing air pollution
damage to vegetation. Organization, results and
recommendations for future surveys In Proceedingsof
the Second International Clean Air Congress. H. M
Englund and W T Beery, eds Academic Press, Inc ,
N.Y., 1971, pp. 132-136.
Landau, E., and C S Brandt The use of surveys to
estimate air pollution damage to agriculture Environ
Res 354-61, 1970
Larsen, R I , and W W Heck An air quality data
analysis system for interrelating effects, standards, and
needed source reductions Part 3 Vegetation injury J
Air Pollut Control Assoc 26325-333,1976
Laurence, J. A , F A Wood, and S V Krupa Possible
transport of ozone and ozone precursors Proc Am
Phytopathol Soc. 4.31, 1977
Ledbetter, M C,P W Zimmerman, and A E Hitchcock
The histopathological effect of ozone on plant foliage
Contrib Boyce Thompson Inst 20275-282, 1959
Leone, I A.andE Brennan Ozone toxicity in tomato as
modified by phosphorus nutrition Phytopathology
60 1521-1524, 1970
Leone, I. A , E Brennan, and R H Dames Effect of
nitrogen nutrition on the response of tobacco to ozone in
the atmosphere J Air Pollut Control Assoc 76 191-
196, 1966
Linzon, S N Ozone damage and semimature-tissue
needle blight of eastern white pine Can J Bot
452047-2060, 1967
Linzon, S N Semimature-tissue Needle Blight of
Eastern White Pine and Local Weather Information
Report O-X-1, Canada Department of Forestry, Ontario
Region, Sault Ste Marie, Ont , Canada, August 1965
Linzon, S N , The development of foliar symptoms and
the possible cause and origin of white pine needle
blight Can J Bot 35153-161, 1960
288
-------
174 Lmzon, S N , W W. Heck, anfl F. D. H. Macdowall
Effects of photochemical oxidants on vegetation. In
Photochemical Air Pollution Formation, Transport and
Effects Environmental Secretariat Publication No.
NRCC 14096, National Research Council of Canada,
Associate Committee on Scientific Criteria for
Environmental Quality, Ottawa, Canada, 1975
175 Maas, E. V., G J Hoffman, S. L Rawlins, andG Ogata
Salinity-ozone interactions on pinto bean Integrated
response to ozone concentration and duration J
Environ Dual. 2,400-404, 1973.
176 Macdowall, F D H Predisposition of tobacco to ozone
damage Can J. Plant Sci 45 1-12, 1965
177 Macdowall, F D H Stages of ozone damage to
respiration ol tobacco leaves Can J Bot 43:419-427,
1965
178 Macdowall, F D H , and A F W Cote Threshold and
synergistic damage to tobacco by ozone and sulfur
dioxide Atmos Environ §553-559, 1971
179 Macdowall, F D H.E I Mukammal, and AF W. Cole
Direct correlation of air-polluting ozone and tobacco
weather fleck Can J Plant Set 44410-417, 1964
180 MacKmght, M L The Effect of Ozone on Stomatal
Aperture and Transpiration M A thesis, University of
Utah, Salt Lake City, Utah, 1968
181 Magie, R O Controlling gladiolus botrytis bud rot with
ozone gas Proc Fla StateHort Soc 73 373-375,1960
182 Manning,W J,W A Feder, P M.Papta.andl Perkins
Influence of foliar ozone injury on root development and
root surface fungi of pinto bean plants Environ Pollut
1 305-312, 1971
183 Manning,W J ,W A Feder, and I. Perkins Responseof
pomsettia cult jars to several concentrations of ozone
Plant D«s Rep 57,774-775, 1i73
184 Manning, W J , W A Feder, and i Perkins Sensitivity
of spinach.cul'ivars to ozone Plant Dis Rep 56832-
833, 1972
185 Mansfield. T / The role of stomata in determining the
responses of (. lants to air pollutants Comment Plant
Sci 2 11-20, «973
186 Markowski, A and S Grzesiak Influence of sulphur
dioxide and o one on vegetation of bean and barley
plants under .Itfferent soil moisture conditions. Bull
Acad Pol Sci Ser Sci Biol 22.875-887, 1975
187 Matsusnima, J On composite harm to plants by
sulfurous acid gas and tmidant Sangyo Kogai 7 218-
224, 1971
188 McCarthy, J J , and C H Smith. A review of ozone and
its application to domestic wastewater treatment J
Am Water Works Assoc 55718-725,1974
189 McCune, D C On the Establishment of Air Quality
Criteria with Reference to the Effects of Atmospheric
Fluorine on Vegetation Air Quality Monograph No 69-
3, American Petroleum Institute, N Y , 1969
190 Mcllvenn, W D, R A Spotts, and D D. Davis The
influence of soil zinc on nodulation, mycorrhizae, and
ozone-sensitivity of pinto bean Phytopathology 65 645-
647, 1975
191 Menser, H A, and H E Heggestad Ozone and sulfur
dioxide synergisni Injury to tobacco plants. Science
153 424-425, 1966
192 Menser, H A, H E Heggestad, and O E Street.
Response of plants to air pollutants 11 Effects of ozone
concentration and leaf maturity on injury to Nicotiana
tobacum Phytopathology 53 1 304-1 308, 1963
193 Menser, H A,H E Heggestad, 0 E Street, and R N
Jeffrey Responseof plants to air pollutants I Effectsof
ozone on tobacco plants preconditioned by light and
temperature Plant Physiol 38 605-609, 1963
194 Menser, H A, Jr. and G H Hodges Effects of air
pollutants on burleytobaccocultivars Agron J 62 265-
269, 1970
195 Menser, H A, Jr, G H Hodges, and C G McKee
Effects of air pollution on Maryland (Type 32) tobacco J
Environ Qua! 2.253-258, 1973
196 Menser, H A, Jr , and 0 E Street Effects of air
pollution, nitrogen levels, supplemental irrigation, and
plant spacing on weather fleck and leaf tosses of
Maryland tobacco Tobacco 155 192-196, 1962
197 Middleton, J T Photochemical air pollution damage.to
leaves Annu Rev Plant Physiol 12 431 -448, 1961
198 Middleton, J T Responseof plants to airpoliution J Air
Pollut Control Assoc 6 1-4, 1956
1 99 Middleton, J. T., A S Crafts, R F Brewer, and O C
Taylor Plant damage by air pollution Calif Agnc 109-
12, 1956
200 Middleton, J T,E F Parley, and R F Brewer Damage
to vegetation from polluted atmospheres J Air Pollut
Control Assoc 59-15, 1958
201 Middleton, J T., J B Kendnck, Jr , and E F Parley Air
pollution injury to crops Calif Agnc 711-12, 1953
202 Middleton, J T , J B Kendrick, Jr, and E F Parley Air-
borne oxidants as plant-damaging agents In
Proceedings of the Third National Air Pollution
Symposium National Air Pollution Symposium, Los
Angeles, Calif , 1955 pp 191-198
203 Middlelon, J T,J B Kendrick, Jr., and H W Schwalm
Injury to herbaceous plants by smog or air pollution
Plant Dis Rep 34 254-252, 1950
204 Middleton, J T,andA 0 Paulus The identification and
distribution of air pollutants through plant reponse
AMAArch Ind Health 14 526-532, 1956
205 Millecan, A A A Survey and Assessment of Air
Pollution Damage to California Vegetation in 1970
APTD-0694, U S Environmental Protection Agency, Air
Pollution Control Office, Research Trrangle Park, N C ,
June 1971
206 Miller, P R Susceptibility to ozone of selected western
(US) conifers For Abstr 30(4)2095,1975
207 Miller, P R , F W Cobb, Jr, and E Zavarm
Photochemical oxidant injury and bark beetle
(Coleoptera Scolytidae) infestation of ponderosa pine
III Effect of injury upon oleoresm composition, phloem
carbohydrates, and phloem pH Hilgardia 39 135-140.
1968
208. Miller, V L , R K Howell, and B E Caldwell Relative
sensitivity of soybean genotypes to ozone and sulfur
dioxide J Environ Qual 335-37,1974
209 Miller, P R , and A A Millecan Extent of oxidant air
pollution damage to some pines and other conifers in
California Plant Dis Rep 55555-559, 1971
210 Miller, P R,J R Parmater, Jr, B H Flick and C W
Martinez Ozone dosage response of ponderosa pine
seedlings J Air Poliul Control Assoc 19 435-438,
1969
211 Miller, P R,andS Rich Ozone damage on apples Plant
Dis Rep 52730-731, 1968
289
-------
212 Mudd, J B, and T T Kozlowski, eds Responses of
Plants to Air Pollution Academic Press, Inc , N Y , 1 975
213 Mukammal, E I Ozone as a cause of tobacco injury
Agric Meteorol 2145-165,1965 229
214 Naegele, J A, W A Feder, and C J Brandt
Assessment of Air Pollution Damage to Vegetation in 230
NewEngland July1971 -July1972 FmalReport EPA-
R5-72-009, U S Environmental Protection Agency,
Research Triangle Park, N.C., August 1 972
215 Nakamura, H , and S Matsunaka Indicator plantsforair
pollutants (1) Susceptibility of morning glory to 231
photochemical oxidants Varietal difference and effect
of environmental factors Nippon SakumotsuGakkai KIJI
43 517-522, 1974 232
216 National Air Pollution Control Administration Effectsof
photochemical oxidants on vegetation and certain
microorganisms In Air Quality Criteria for 233
Photochemical Oxidants NAPCA Publication No AP-
63, U S. Department of Health, Education, and Welfare,
Public Health Service, Washington, D C , March 1970 234
pp 6-1 - 6-23
217 National Research Council Ozone and Other
Photochemical Oxidants Washington, D C , National 235
Academy of Sciences, 1977 pp 437-585
218 Neely, G E,D T Tmgey, and R G Wilhour Effectsof
ozone ana sulfur dioxide singly and in combination on 236
yield, quality and N-fixation of alfalfa In International
Conference on Photochemical Oxidant Pollution and Its
Control. Proceedings Volume II B Dimitnades, ed
EPA-600/3-77-001b, US Environmental Protection 237
Agency, Research Triangle Park, N C , January 1977
pp 663-673
219 Noble, W Air pollution effects Pattern of damage
produced on vegetation by smog J Agric Food Chem 238
3 330-332, 1955
220 Noble, W Smog damage toplants Lasca Leaves 15 2-4,
1965
221 Noble, W M , and L A Wright Air pollution with
relation to agronomic crops II A bio-assay approach to 239
the study of air pollution Agron J 50551-553,1958
222 Oerth, J J Effect of salinity on susceptibility of
sunflower plants to smog Soil Sci 87 249-251, 1959
223 Ogata, G , and E V Maas Interactive effects of salinity
and ozone on growth and yield of garden beet J 240
Environ Qual 2518-520,1973
224 Ordm, L,M J Garber, andJ I Kmdmger Effect of 2,4-
dichlorophenoxy acetic acid on growth and on /3-glucan
synthetases of peroxyacetyl nitrate pretreated Avena
coleoptile sections Physiol Plant 2617-23, 1972 241
225 Ormrod, D P , N O Adedipe, and G Hofstra Ozone
effects on growth of radish plants as influenced by
nitrogen and phosphorus nutrition and by temperature 242
Plant Soil 38 437-439, 1973
226 Ormrod, D P , N 0 Adedipe, and G Hofstra Responses
of cucumber, onion, and potato cultivars to ozone Can
J Plant Sci 51 283-288, 1971 243
227 Oshima, R J Development of a System for Evaluating
and Reporting Economic Crop Losses Caused by Air
Pollution in California I Quality Study State of
California, Department of Food and Agriculture, 244
Sacramento, Calif , July 1973.
228 Oshima, R J Development of a System for Evaluating
and Reporting Economic Crop Losses Caused by Air 245
Pollution in California II Yield Study IIA Prototype
Ozone Dosage-Crop Loss Conversion Function State of
California, Department of Food and Agriculture,
Sacramento, Calif , June 1 974
Oshima, R J Effect of ozone on a commercial sweet
corn variety Plant Dis Rep 57719-723,1973
Oshima, R.J., M P Poe, P K Braegelmann, D W
Baldwin, and V Van Way Ozone dosage-crop loss
function for alfalfa A standardized method for
assessing crop losses from air pollutants J Air Pollut
Control Assoc 26861-865, 1976
Oshima, R J , 0 C Taylor, P K Braegelmann, and D W
Baldwin Effect of ozone on a commercial variety of
tomato J Environ Qual 4463-464, 1975
Oshima, R,0 C Taylor, and E A Cardiff Possible new
toxicant indicated in severe air pollution episode m
South Coast Basin Calif Agric 28 1 2-1 3, 1 974
Otto, H W, and R H Dames Plant injury by air
pollutants Influence of humidity on stomatal and plant
response to ozone Science 7631209-1210, 1969
Parmeter, J R , Jr , R V Bega, and T Neff Achlorotic
decline of ponderosa pme in southern California Plant
Dis Rep 46 269-273, 1962
Parmeter, J R , Jr , and P R Miller Studies relating to
the cause of decline and death of ponderosa pine m
southern California Plant Dis Rep 52 707-711, 1968
Pell, E J 1 972 Survey and Assessment of Air Pollution
Damage to Vegetation m New Jersey EPA-R5-73-022,
U S Environmental Protection Agency, Research
Triangle Park, N C , June 1973
Pell, E J , and E Brennan Economic impact of air
pollution on vegetation in New Jersey and an
interpretation of its annual variability Environ Pollut
823-33, 1975
Photochemical Air Pollution Formation, Transport, and
Effects Environmental Secretariat Publication No
NRCC 14096, National Research Council of Canada,
Associate Committee on Scientific Criteria for
Environmental Quality, Ottawa, Canada, 1975
Price, H , and M Treshow Effects of ozone on the
growth and reproduction of grasses In Proceedings of
the International Clean Air Conference Clean Air
Conference, National Science Centre, Victoria,
Australia, 1972 pp 275-280
Rabotnova, I L, V S Somov, T S Bobkova, I V
Zlochevskaya, L N Checkunova, I F Knyazeva, and S I
Belen 'Kn Toxic action of ozone on yeasts as a function
of some components and pH of the medium Vestn
Mosk Umv Biol Pochvoved 2647-51,1971
Reinert, R A Monitoring, detecting, and effects of air
pollutants on horticultural crops Sensitivity of genera
and species Hort Sci 10 495-500, 1975
Reinert, R A, A S Heagle, and W W Heck Plant
responses to pollutant combinations In Responses of
Plants to Air Pollution B MuddandT T Kozlowski, eds
Academic Press, Inc , N Y , 1975 pp 159-177
Reinert, R A, A S Heagle, J R Miller, and W R
Geckeler Field studies of air pollution injury to
vegetation m Cincinnati, Ohio Plant Dis Rep 548-11,
1970
Reinert, R A, D T Tmgey, and H B Carter Ozone
induced foliar injury in lettuce and radish cultivars J
Am Soc Hort Sci 97711-714,1972
Reinert, R A,D T Tmgey, W W Heck, and C Wickhff
Tobacco growth influenced by concentrations of sulfur
290
-------
dioxide and ozone Agron Abstr 61 34,1969 263
246 Remert, R A,D T Tmgey, and C E Koons The early
growth of soybean as influenced by ozone stress Agron
Abstr 63 148, 1971
247 Rich, S , G S Taylor, and H Tomlmson Crop damaging
periods of ambient ozone in Connecticut Plant Dis Rep 264
53969-973, 1969
248 Rich, S,, and H Tomlmson Effects of ozone on
conidiophores and conidia of Alternana solant
Phytopathology 58 444-446, 1968
249 Rich, S , and N C Turner Importance of moisture on 265
slomatal behavior of plants subjected to ozone J Air
Pollut Control Assoc 22 718-721, 1972
250 Richards, B L, J T Middleton, and W B Hewitt Air
pollution with relation to agronomic crops V Oxtdant
stipple of grape Agron J 50559-561, 1958 266
251 Ridley, J D , and E T Sims, Jr Preliminary
Investigations on the Use of Ozone to Extend the Shelf-
Life and Maintain the Market Quality of Peaches and 267
Strawberries South Carolina Agricultural Experiment
Station Research Series No 70, Clemson University, 268
Clemson, S C , January 1966
252 Ripaldi, C P, and E Brennan Effect of ozone on 269
phosphorus levels of pinto bean stems and foliage
Phytopathology 63 206, 1 973 270
253 Santamour, F S , Jr Air pollution studies on Plantanus
and American elm seedlings Plant Dis Rep 53 482-
484, 1969 271
254 Scotl, D B M , and E C Lesher Effect of ozone on
survival and permeability of Eschenchia coli J
Bactenol 85 567-576, 1963
255 Seidman, G, I J Hmdawi, and W W Heck
Environmental conditions affecting the use of plants as 272
indicators of air pollution J Air Pollut Control Assoc
15 168-170, 1965
256 Shannon, J G , and C L Mulchi Ozone damage to
wheat varieties at anthesis Crop Sci J4 335-337,
1S74
257 Shinohara, T , S Kuroda, T Kitamura,and K Kunisawa 273
Studies on weather fleck on tobacco leaves VI Effects
of temperature on the fleck injury induced by air-borne
oxidants Okayama Tabako Shikenjo Hokoku 33 51 -54,
1973 274
258 Shinohara, T , Y Yamamoto, H Kitano, andM Fukuda
Interactions of light and ozone injury in tobacco Nippon
Sakumotsu Gakkai Ki|i 43 433-438, 1 974
259 Skelly, J M, C F Croghan and E M Hayes Oxidant
levels in remote mountainous areas of southwestern 275
Virginia and their effects on native white pine IPinas
strobus L } In International Conference on
Photochemical Oxidant Pollution and its Control
Proceedings Volume II B Dimitnades, ed EPA-600/3- 276
77-001b, US Environmental Protection Agency,
Research Triangle Park, N C , January 1977 pp 61 1 -
620
260 Smith, D , and J P Silva Use of carbohydrate and 277
nitrogen root reserves in the regrowth of alfalfa from
greenhouse experiments under light and dark
conditions Crop Sci 3464-467,1969 278
261 Smock, R M , and R D Watson Ozone in apple storage
Refng Eng 42 97-101, 1941
262 Spaldmg, D H Appearance and decay of strawberries, 279
peaches, and lettuce treated with ozone U S Dept
Agnc Mark Res Rep (756)1-11, 1966
Stanford Research Institute The Smog Problem in Los
Angeles County A Report by Stanford Research
Institute on the Nature and Causes of Smog Western
Oil and Gas Association, Los Angeles, Calif January
1954
Stark, R W , P R Miller, F W Cobb, Jr , D L Wood, and
J R Parmeter, Jr Photochemical oxidant injury and
bark beetle (Coleoptera Scolytidae) infestation of
ponderosa pine I Incidence of bark beetle infestation in
injured trees Hilgardia 39 121-126, 1968
Starkey, T D The Influence of Peroxyacetyl Nitrate on
Bean (Phaseolus vulgans L ) Subjected to Post-
Exposure Water Stress CAES Publication No 400-75,
Pennsylvania State University, Center for Air
Environment Studies, University Park, Pa , 1975
Stephens, E R , E F Parley, O C Taylor, and W E
Scotl Photochemical reaction products in air pollution
Int J Air Water Poltut 479-100,1961
Taylor, G S Ozone injury on Bel W-3 tobacco controlled
by at least two genes Phytopathology 53 1069, 1968
Taylor, G S,H G DeRoo.andP E Waggoner Moisture
and fleck of tobacco Tob Sci 462-68, 1960
Taylor, 0 C Effects of oxidanlair pollutants J Occupat
Med r0(2) 485-492, September 1968
Taylor, 0 C Importance of peroxyacetyl nitrate(PAN)as
a phytotoxic air pollutant J Air Pollut Control Assoc
J9 347-351, 1969
Taylor, 0 C,E A Cardiff, J D Mersereau, and J T
Middleton Effect of air-borne reaction products of
ozone and 1-n-hexene vapor (synthetic smog) on growth
of avocado seedlings Proc Am Soc Hort Sci 77 320-
325, 1958
Taylor, 0 C , and D C MacLean Nitrogen oxides and
the peroxyacyl nitrates In Recognition of Air Pollution
Injury to Vegetation A Pictorial Atlas J S Jacobson
and A C Hill, eds Informative Report No 1, Air
Pollution Control Association, Pittsburgh, Pa , 1 970 pp
E1-E14
Thomas, M D Effects of air pollution on plants Air
Pollution World Health Organization Monograph Series
No 46, Columbia University Press, New York, 1961 pp
233-278
Thomas, M D , R H Hendncks, and G R Hill Some
impurities in the air and their effects on plants In Air
Pollution Proceedings of the United States Technical
Conference on Air Pollution McGraw-Hill, Inc , N Y ,
1952 pp 41-47
Thompson, C R Personal communications Cited in
Ozone and Other Photochemical Oxidants National
Academy of Sciences, Washington, D C , 1977 pp 464-
465
Thompson, W W , W M Dugger, Jr, and R L Palmer
Effects of ozone on the fine structure of the palisade
parenchyma cells of bean leaves Can J Bot 44 1677-
1682, 1966
Thompson, C R , E Hensel, and G Kats Effects of
photochemical air pollutants on Zinfandel grapes Hort
Sci 4 222-224, 1969
Thompson, C R, and G Kats Antioxidants reduce
grape yield reductions from photochemical smog Calif
Agnc 24 12-13, 1970
Thompson, C R . and G Kats Effects of ambient
concentrations of peroxyacetyl nitrate on navel orange
trees Environ Sci Technol 335-38, 1975
291
-------
280 Thompson, C R, and 0 C Taylor Effects of air
pollutants on growth, leaf drop, fruit drop, and yield of
citrus trees Environ Sci Technol 3934-940, 1969
281 Ting, I P.,andW M Dugger.Jr Factors affecting ozone
sensitivity and susceptibility of cotton plants J Air
Pollut Control Assoc 18 810-81 3, 1968
282 Tingey, D T Letter toJ H B Garner,Criteria and Special
Studies Office, HERL, U S Environmental Protection
Agency, Research Triangle Park, N C , October 1 976
283 Tingey, D T Ozone induced alterations in plant growth
and metabolism In International Conference on
Photochemical Oxidant Pollution and Its Control
Proceedings Volume II B Dimitnades.ed EPA-600/3-
77-001 b, US Environmental Protection Agency,
Research Triangle Park, N C , January 1977 pp 601-
609
284 Tingey, D T , and U Blum Effects of ozone on soybean
nodules J Environ Qual 2341-342, 1973
285 Tingey, D T , J A Dunning, and G M Jividen Radish
root growth reduced by acute ozone exposures In
Proceedings of the Third International Clean Air
Congress Verems Deutscher Ingenieur, Kommission
Remhaltungder Luft, Dusseldorf, 1973 pp A154-A156
28& Tingey, D T,R C Fites.andC Wickliff Foliar sensitivity
of soybeans to ozone as related to several leaf
parameters Environ Pollut 4183-192,1973
287 Tingey, D T ,W W Heck, and R A Reinert Effect of low
concentrations of ozone and sulfur dioxide on foliage,
growth and yield of radish J Am Soc Hort Sci
96369-371 1971
288 Tingey, D T,W W Heck, and R A Reinert Reduction of
radish yields by low concentrations of ozone and sulfur
dioxide Hort Sci 4 189, 1969
289 Tingey, D T, and A C Hill The occurrence of
photochemical phytotoxicanls in the Sail Lake Valley
Proc Utah Acad Sci Arts Lett 44387-395,1967
290 Tingey, D T , and R A Reinert The effect of ozone and
sulfur dioxide singly and in combination on plant
growth Environ Pollut 9117-126,1975
291 Tingey, D T,R A Reinert, and H B Carter Soybean
culttvars. Acute foliar response to ozone Crop Sci
72 268-270, 1962
292 Tingey, D T , R A Reinert, J A Dunning, and W W
Heck Foliar injury responses of eleven plant species to
ozone/sulfur dioxide mixtures Atmos Environ 7201-
208, 1973
293 Tingey. D T, R A Reinert, C Wickliff. and W W Heck
Chronic ozone or sulfur dioxide exposures, or both,
affect the early vegetative growth of soybeans Can, J
Plant Sci 53875-879, 1973
294 Tingey, D T, C Standley, and R W Field Stress
ethylene envolution A measure of ozone effects on
plants Atmos Environ 70969-974,1976
295 Tingey, D T , R G Wilhour. and C Standley The effect
of chronic ozone exposures on the metabolite con tent of
ponderosa pine seedlings For Sci 22234-241, 1976
296 Todd, G W Effect ol ozone and ozonated 1-hexene on
respiration and photosynthesis of leaves Plant Physiol
33416-420, 1958
297 Todd, G W,andW N Arnold AnevaSuation of methods
used to determine injury to plant leaves by air
pollutants Bot Gaz (Chicago) 723 1 51-154, 1961
298 Todd, G W.andB Probst Changes in transpiration and
photosynthetic rates of various leaves during treatment
with ozonated hexene or ozone gas Physiol Plant
16 57-65, 1963
299 Townsend, A M , and L S Dochmger Relationship of
seed source and developmental stage to the ozone
tolerance oi Acer rubrum seedlings Atmos Environ
5957-964, 1974
300 Treshow, M Environment and Plant Response
McGraw-Hill, !nc New York, 1970
301 Treshow, M Evaluation of vegetation injury as an air
pollution criteria J Air Pollut Control Assoc tB 266-
269, 1965
302 van Haul, H , and H Slratmann Experimental
investigations of the effect of nitrogen dioxide on plants
Schrtftenr LIB Landesanst Irnmissions
Bodenutzungssehutz Landes Nordheim Westfalen
{F} 50-70, 1967
303 Verkroost,M,The effect of oione on photosynthesis and
respiration of Scenedesmus obtusiusculus Chod with a
general discussion of effects of air pollutants in plants
Meded Landouwhogesch Wagenmgen 741-78,
1974
304 Walker, E K , and L S Vickery Influence of sprinkler
irrigation on the incidence of weather fleck on flue-
cured tobacco in Ontario Can J Plant Sci 41 281-287,
1961
305 Watson, R D Ozone as a Fungicide Ph D thesis,
Cornell University, Ithaca,N Y , 1942
306 Weaver, G M , and H 0 Jackson Relationship
between bronzing in white beans and phytotoxic levels
of atmospheric ozone in Ontario Can, J Plant Sci
48 561 -568, 1968
307 Weber, 0 E Interactions of Plant-Parasitic Nematodes
and Ozone and Sulfur Dioxide with Soybean and
Begonia Ph D thesis. North Carolina State University,
Raleigh, N C , 1976
308 Weber, D 3. The effects of ozone on plant-parasitic
nematodes and certain plant-microorganism
interaction In International Conference on
Photochemical Oxidant Pollution and Its Control
Proceedings Volume II B Dimitnades, ed EPA-600/3-
77-GOIb, US Environmental Protection Agency.
Research Triangle Park, N C , January 1977 pp 621-
631
309 Wilhour, R G The Influence of Ozone on White Ash
CAES Publication No 188-71, Pennsylvania State
University, Center for Air Environment Studies,
University Park, Pa , 1971
310 Wilhour, R G, and G E Neely Growth response of
conifer seedlings to low ozone concentrations In
International Conference on Photochemical Oxidant
Pollution and Its Control Proceedings Volume II B
Dimitnades, ed EPA-600/3-77-001 b. US
Environmental Protection Agency, Research Triangle
Park, N C , January 1977 pp 635-645
311 Wood, F A.andJ B Coppolmo The influence of ozone
on deciduous forest tree species Mitt Forst
Bundesversuchsanst, Wien 97 233-253, 1 972
312 Wood, F A,andJ B Coppolmo The influence of ozone
on selected woody ornamentals Phytopathology
61 133, 1971
313 Wood, FA, and D B Drummond Response of eight
cultivars of chrysanthemum to peroxyacetyl nitrate
Phytopathology 64 897-898, 1974
292
-------
314 Yarwood, C E , and J T Middleton Smog injury and
rust infection Plant Physiol 29393-395,1954
315 Zobnina, V P , and E A Morkovina Effect of ozone on
survival of carotmoid strain of Mycobactenum
carothenum and its white mutant obtained with
nitrosoguanidme Microbiology (Engl Transl ) 40 79-
81, 1971
293
-------
12. ECOSYSTEMS
INTRODUCTION
The preceding chapters have dealt with the
formation and chemistry of ozone and other
oxidants and with the effects of oxidant/ozone on
human health and on experimental plants and
animals. In this chapter, the effects of oxidant-
pollutant stress on simple and complex
communities are discussed to illustrate that
plants, animals, man, and the environment are
interconnected and that the relationships between
them are profoundly complex.
Plants, animals, and microorganisms usually do
not live alone, but exist as populations. Populations
live together and interact as communities.
Commmunities, because of the interactions of
their populations and of the individuals that
constitute them, respond to pollutant stress
differently from individuals. Man is an integral part
of these communities, and as such, he is directly
involved in the complex ecological interactions
that occur within the communities and the
ecosystem of which he is a part.
The stresses placed on the communities and the
ecosystems in which they exist can be far-
reaching, since the changes that occur may be
irreversible. For example, it has been suggested
that the arid lands of India are the result of
defoliation and elimination of vegetation that
induced local climatic changes not conducive to
the reestablishment of the original vegetation.10
An ecosystem (e.g., the planet Earth, a forest, a
pond, an old field, or a fallen log) is a major
ecological unit made up of living (biotic) and
physical (abiotic) components through which the
cycling of energy and nutrients occurs (Table 12-
1). A structural relationship exists among the
various components. The biotic units are linked
together by functional interdependence, and the
abiotic units make up all of the physical factors and
chemical substances that interact with the biotic
units. The processes occurring within the biotic
and abiotic units and the interactions among them
are subject to environmental influences.9
TABLE 12-1. COMPOSITION OF ECOSYSTEMS
Component
Description
Biotic (biological)
Individuals
Producers
Consumers
Decomposers
Populations
Communities
Abiotic (physical)
Energy
Water
Atmosphere
Fire
Topography
Geological
substratum
Plants, animals (man), and micro-
organisms These are eitherproducers,
consumers, or decomposers
Green plants
Herbivores, carnivores
Macroorgamsms (mites, earthworms,
millipedes, and slugs) and micro-
organisms (bacteria and fungi)
Groups of similar and related
organisms
Interacting populations linked together
by their responses to a common
environment
Radiation, light, temperature, and heat
flow.
Liquid, ice, etc
Gases and wind
Combustion
Surface features
Soil, a complex system Nutrients.
Ecosystems tend to change with time, and
sequential changes in the types of populations
within a community are usually recognizable.
Adaptation, adjustment, and evolution are
constantly taking place as the biotic component,
the populations, and the communities of living
organisms interact with the abiotic component in
the environment. The interaction and exchange of
energy and nutrients over time result in sequential
or, in some cases, cyclic or telescoped changes in
populations and communities. The sequential
replacement of one population by another in a
continual series, going from pioneer (first and less
diversified) populationsto so-called climax(mature
and more diversified) communities, is termed
"succession,"9 Climax communities are
structurally complex and more or less stable, and
they are held in a steady state through the
operation of a particular combination of biotic and
abiotic factors. Man is often a factor, as for
294
-------
example, when his grazing cattle maintain a
pasture.9 The disturbance or destruction of a
climax community or ecosystem results in its being
returned to a simpler stage.71'72 Existing studies
indicate that changes occurring within
ecosystems, in response to pollution or other
disturbances, follow definite patterns that are
similar even in different ecosystems. It is therefore
possible to predict broadly the basic biotic
responses to the disturbance of an eco-
system.22'39'70"72 These responses to disturbances
are:
1. Reduction in standing crop,
2. Inhibition of growth or reduction in
productivity.
3. Differential kill (removal of sensitive
organisms at the species and subspecies
level),
4. Food chain disruption.
5. Successional setback,
6. Changes in nutrient cycling rates.
Organisms vary in their ability to withstand
environmental changes. The ability of a population
to withstand injury from polluted air, weather
extremes, herbicides, or other disturbances
depends on its range of tolerance—that is, the
range of variation within which it can survive and
function. The range of tolerance (also termed the
"law of tolerance," Figure 12-1), differs because of
inter- and intraspecific variations in susceptibility
to injury by polluted air or other disturbances.50
The highly specialized populations (those with a
narrow range of tolerance or adaptability) are
either reduced in number or eliminated.
Diversity and structure are most changed by
pollution as a result of the elimination of sensitive
species of flora and fauna and the selective
removal of the larger overstory plants in favor of
plants of small stature.71'72 The result is a shift from
the complex forest community to the less complex,
hardy shrub and herb communities. The opening of
the forest canopy changes the environmental
stresses on the forest floor, causing differential
survival and, consequently, changed gene
frequencies in subcanopy species.
Associated with the reduction in diversity and
structure is a shortening of food chains, a
reduction in the total nutrient inventory, and a
return to a simpler and less stable Successional
stage.71'72 Jn addition, the pollutants act as
predisposing agents, and an increase in the activity
of insect pests and certain diseases occurs,"8'71 It
TOLERANCE RANGE
ZONE OF
INTOLERANCE
ZONE OF
INTOLERANCE
ZONE OF
PHYSIOLOGICAL
STRESS
ZONE OF
PHYSIOLOGICAL
STRESS
RANGE OF OPTIMUM
ORGANISMS
INFREQUENT
ORGANISMS
INFREQUENT
ORGANISMS
ABSENT
ORGANISMS
ABSENT
GREATEST
ABUNDANCE
Figure 12-1. Law of tolerance.50
295
-------
should be emphasized that ecosystems are usually
subjected to a nu mber of stresses at the same time,
not just to a single perturbation.
Ecosystems are usually evaluated by modern
man solely on the basis of their economic value to
him. This economic value, in turn, depends on the
extent to which man can manipulate the
ecosystem for his own purpose. This single-
purpose point of view makes it difficult to explain
the many benefits of a natural ecosystem to man's
welfare in terms of the conventional cost-benefit
analysis. Gosselink et al.24 have, however, placed a
value on a tidal marsh by assigning monetary
values to the multiple contributions to man's
welfare such as fish nurseries, food suppliers, and
waste treatment functions of the marsh. They
estimate the total social values to range from
$50,000 to $80,000/acre.
Westman67 also evaluated the benefits of
natural ecosystems by estimating the monetary
costs associated with the loss of the free services
(absorption of air pollution, regulation of global
climate and radiation balance, and soil binding)
provided by the ecosystems. Westman estimated
that the oxidant damage to the San Bernardino
National Forest could result in a cost of $27 million
per year(1 973 dollars} for sediment removal alone
because of erosion as long as the forest remained
in the early stages of succession.
Estimates of the dollar values of such items as
clean air and water, untamed wildlife, .and
wilderness (once regarded as priceless) are
attempts to rationalize the activities of
civilization.87 When environmental decisions are
made based on the cost of damaged ecosystems, it
is usually assumed that the path chosen will be
that which is most socially beneficial, as indicated
by costs compared to benefits. As Westman67
points out, certain corollaries accompany the
assumption that decisions maximizing the benefit.
cost ratios simultaneously optimize social equity
and utility. These corollaries arep
(1) The human species has the exclusive right to use and
manipulate nature for its own purposes (2) Monetary
units are socially acceptable as means to equate the value
of natural resources destroyed and those developed. (3)
The value of services lost during the interval before the
replacement or substitution of the usurped resource has
occurred is included in the cost of the damaged resource
(4) The amount of compensation in monetary units
accurately reflects the full value of the loss to each loser in
the transaction (5) The value of the item to future
generations has been judged and included in an accurate
way in the total value (6) The benefits of development
accrue to the same sectors of society, and in the same
proportions, as the sectors on whom the costs are levied,
or acceptable compensation has been transferred Each of
these assumptions, and others not listed, can and have
been challenged 67
In the case of corollary 4, for example, the losses
incurred include species other than man, butthese
losses are seldom, if ever, compensated. Also, the
public at large is not usually consulted to
determine whether the dollar compensation is
adequate and acceptable. Frequently, there is no
direct compensation. Corollary 5 can never be
fulfilled because it is impossible to determine
accurately the value to future generations.
Evaluating the contribution of functioning
natural ecosystems to human welfare is a very
complex task and usually involves weighing both
economic and human social values. However,
because natural ecosystems are life support
systems, their values should not be quantified in
economic terms.
With the passage of time, man has destroyed
many of the naturally occurring ecosystems of
which he was a part and has replaced them with
simplified ecosystems wholly dependent on his
care and protection and requiring a large input of
energy.50
Man favors the simple, unstable and synthetic
ecosystems, because when they are extensively
managed and subsidized by the use of fossil fuels,
they are highly productive. Productivity is the rate
at which energy is stored by the photosynthetic
activity of green plants.9'50 Young and suc-
cessional ecosystems are more productive
because they add biomass (accumulate energy)
each year. An agricultural ecosystem
(agroecosystem) is an example of such a simplified
ecosystem. Urban ecosystems, with their
cultivated trees, shrubs, grasses, and flowers, are
also examples of ecosystems simplified by man.
The structural simplicity of these ecosystems (in
many cases monocultures) makes them less
resistant to environmental and disease stresses or
to perturbations such as oxidant pollution. The
greater diversity of a matureecosystem retardsthe
disruption of normal structure and function caused
by any type of perturbation. For example, a forest
ecosystem in which the communities are
composed of many species would show less
immediate damage by stress than successional
stages having only a few species. Even greater
damage would be anticipated in an agroe-
cosystem, which may be considered the simplest
of successional stages since often only a single
producer species is present.
296
-------
GENERAL RESPONSES OF NATURAL AND
AGROECOSYSTEMS TO STRESS BY
OXIDANTS
Agroecosystems
Photochemical oxidant air pollutants have been
recognized as a type of chronic pollution problem
only during the last 20 to 25 years. The problem
was first recognized in southern California, where
it perhaps has had its greatest impact. Citrus
groves and vineyards in the inland valleys of
southern California are prime examples of stress to
agroecosystems caused by chronic exposure to
oxidants.
Studies were initiated in 1960 on lemon and
navel orange trees58 and in 1968 on wine grapes to
determine the economic losses caused by oxidant
pollutants. Both studies were conducted under
field conditions for several years. The lemon and
orange tree studies provided only a small amount
of data that may be interpreted in an agroe-
cosystem context because they were highly
oriented toward the primary producer com-
ponents,58"60 and no effort was made to examine
the effects of oxidants on consumer components
and decomposer organisms. The studies indicated
that oxidant stress reduced water use and
photosynthesis, increased leaf and fruit drop, and
resulted in a severe reduction in the yield of
marketable fruit. All of these effects occurred
without the development of plainly visible leaf
symptoms. None of the above studies managed to
distinguish the different effects of ozone and
peroxyacetylnitrate (PAN) or its homologues in the
photochemical oxidant mixtures; however, infer-
ences drawn from these data suggest probable
impact on the various components of these
agroecosystems, particularly on the levels of
consumer and decomposer populations.
Accelerated leaf drop may influence the devel-
opment of insect pests such as aphids, scale
insects, and red citrus mites. Higher con-
centrations of ammo acids or free sugars in injured
leaves before abscission could result in an
increase in pest populations or a diminution if
leaves fall too rapidly. Leaf and fruit drop would
provide an increased substrate for populations of
decomposer organisms at the soil surface.
Natural Ecosystems
Natural ecosystems in California and in the
Appalachian Mountains in the eastern United
States have also suffered from oxidant pollution.
THE APPALACHIAN MOUNTAINS
In the eastern United States, a disease called
emergence tipburn, found in eastern white pine,
was related to ozone by Berry and Ripperton.5 The
disease is characterized by bands of necrosis
initiated in the semimature tissue of elongating
needles and spreading to the needle tip.
Under forest conditions, the affected trees occur
randomly in the stand, and the same trees are
injured repeatedly in a single season or in
successive years.4 Only sensitive genotypes are
affected, and these are gradually eliminated.
Eastern white pine either forms pure stands or
occurs in mixtures with other species m
abandoned fields. Under these conditions, it is an
important pioneer tree.66 In established stands, it is
a major component of 4 forest types and an
associate in 14 other types and inhabits a range
extending over 7 million acres, from the Lake
States to the Appalachian Mountains.65 Berry4 in
1961 reported that post-emergence tipburn occurs
throughout the natural range of this species. There
is also evidence of a slow decline in tree vigor
caused by the deterioration of feeder rootlets.
Skelly et al.49 in 1975 and 1976 also noted post-
emergence acute tipburn on white pine growing at
three different sites along the Blue Ridge Parkway.
High concentrations of ozone in the forested
areas of the eastern United States cause severe
injury to eastern white pine3'49'65 and other forest
species. Berry3 reported levels of oxidant as high as
230 /yg/m3 (0.12 ppm) at a mountamtop
monitoring station in June 1 962, An air monitoring
network operated by Virginia Polytechnic Institute
and State University at three locations (the Blue
Ridge Mountains, the Shenandoah Valley, and the
southern Appalachians) recorded total oxidant
peaks as high as 250 /yg/m3 (0.13 ppm) during
early July 1975, along with 43 hr of exposure at
concentrations of 160 //g/m3 (0.08 ppm) or higher
for 9 days.22 In June 1976, oxidant levels exceeded
160 fjg/m3 (0.08 ppm). After these episodes,
significant increases in oxidant injury were
observed (particularly in the Blue Ridge
Mountains) involving three categories of eastern
white pine: Those previously without symptoms,
those with chlorotic mottle, and those exhibiting
chlorotic dwarf symptoms. Such incidences
suggest the need for more comprehensive studies
of oxidant (and sulfur dioxide) effects in forests of
the eastern United States.
The oxidant peaks mentioned above were
related to oxidant transport from urban areas to the
297
-------
northeast (see Chapter 11, section on mete-
orological parameters).
Studies by Davis and Wood 15 suggest that other
conifer species, in particular, Virginia pine and jack
pine, may be more sensitive to ozone than eastern
white pine. In addition, there is a synergistic
interaction between low concentrations of ozone
and sulfur dioxide that is the cause of the chlorotic
dwarf disease of eastern white pine.16 A study by
Ellertsen et al.19 showed 10 percent mortality
between 1956 and 1965 in dominant and co-
dominant eastern white pines near an
industrialized area that included several hundred
square miles on the Cumberland Plateau. The
combined effects of ozone and sulfur dioxide were
considered responsible for the tree decline.
Because eastern white pine represented only 5
percent of the total wood volume available for
harvest, the economic impact was slight. There
was no effort to interpret pollutant effects in an
ecosystem context. In the long run, the broader
question to be addressed regards the effects of
pollutant stress on all major ecosystem
components, primarily on producers, consumers,
and decomposers. An analysis of the multiple
effects of oxidants on eastern forest ecosystems
will present a much more adequate picture of the
effects pollution is having on these forests.
The studies to delineate the effects of oxidant
episodes in the Appalachian Mountains are
continuing.
CALIFORNIA
In southern California, the coastal chaparral
ecosystem, dominated by chamise and manzanita
or by woodland species(mcludingthe liveoaksand
big cone spruce), and the coniferous forest
ecosystems have received severe exposure. The
desert ecosystems in the vicinity of mountain
passes connecting the coastal and desert regions
have also undoubtedly been exposed.
Injury has been well documented only in the
mixed-conifer forest ecosystem of the San
Bernardino Mountains. Early symptoms of injury in
conifer species were reported in a number of
California national forests in 1970.36 In the
southern Sierra Nevada, Forest Service surveys
made in 1 97469 have detected increased injury in
ponderosa pine at many locations in the Sequoia
National Forest, Sequoia National Park, and Kings
Canyon National Park since 1970. Specific stands
of mixed conifer forest on the western slopes of the
southern Sierra Nevada now appear to be affected
by oxidants from the San Joaquin Valley.35 The
potential loss of timber growth in this area alone is
a very serious prospect.
A project is now underway55'56 to determine in
detail the effects of oxidant air pollutants on a
mixed conifer forest ecosystem in southern
California. The planning documents and early
results from this study as described in the NAS
document, Ozone and Other Photochemical
Oxidants,2* constitute the major source of
information found in the remainderof thischapter.
ORIGIN OF INJURIOUS CONCENTRATIONS
OF OZONE AND OXIDANTS
Advection from Urban Centers to Remote Areas
in Southern California
Descriptions of the vertical and horizontal
distributions of photochemical smog in the Los
Angeles Basin (southern coastal air basin) during
typical summer days have been provided by
Blumenthal et al.,6 Edinger,17 Edinger et al.,18 and
Miller et al.34 Important observations to be drawn
from their reports are the interactions of basin and
mountain topography and local meteorology in
determining the transport and concentrations of
oxidant air pollutants in relation to elevational
zones of vegetation.
The marine temperature inversion layer that
frequently forms above the heavily urbanized Los
Angeles metropolitan area often extends inland as
far as 144 km (90 miles), depending on season and
time of day. Surface heating of air under the
inversion increases with distance eastward in the
basin and often disrupts the inversion by mid-
morning at its eastern edge. The northern rim of
the basin is formed by the San Gabriel and San
Bernardino Mountains, interrupted only by the
Cajon Pass about 88 km (55 miles) inland (see
Figure 12-2). The marine temperature inversion
layer encounters the mountain slopes, usually
below 1200 m (4000 ft). In the morning, the
temperature inversion often remains intact at this
juncture, and air pollutants are confined beneath
it. Studies by Edinger17 and Edinger et al.18 have
described how the heated mountain slopes act to
vent oxidant air pollutants over the crest of the
mountains and cause the injection of pollutants
into the stable inversion layer horizontally away
from the slope. Oxidant concentrations within the
inversion are not uniform, but occur in multiple
layers and strong vertical gradients. In some cases,
the inversion may serve as a reservoir for oxidants,
298
-------
principally ozone, which may arrive at downwind
locations along the mountain slopes relatively
undiluted because of a lack of vertical mixing
within the inversion layeranda lack of contact with
ozone-destroying material generated at the
ground. The important result of the trapping of
oxidant in these layers is its prolonged contact with
high terrain at night.
The most important effect of the interaction of
pollutant and inversion layer at the heated
mountain slope is the vertical venting of oxidants
over the mountain crest by upslope flow. A
gradient of oxidant concentrations or doses is
established across distinct vegetation zones
ordered along an elevational gradient, i.e., the
chaparral and conifer forest ecosystems that
occupy the slopes and mountain terrain beyond the
crest, respectively. The altitudinal sequence of
these ecosystems is illustrated in Figure 12-3.
According to Horton,26 the chaparral zone is
subdivided, from lower to higher elevations, into
three subzones called the chamise, manzanita,
and woodland chaparral; the mixed conifer forest
occupies the mountain crest.
The daytime changes in oxidant concentrations
at several stations (A, B, C, and D in Figure 12-2)
along the southern slope and at the crest of the San
Bernardino Mountains is illustrated in Figure 12-4.
In the late afternoon, the highest concentrations.in
this profile were at 900 to 1200 m (3000 to 4000 ft)
and adjacent to the mountain slope. The oxidant
was not always confined belowthe inversion layer,
but it was present in the inversion layer and above
the mountain crest.18 An instrumented aircraft
measured concentrations of oxidant ranging from
100 to 200 /ug/m3 (0.05 to 0.11 ppm) as high as
2432 m (8000 ft) approximately 1033 m (3400 ft)
above the ridge crest during several flights at noon
and at 4 p.m. At these times, downwind diffusion of
oxidant beyond the crest, on the basis of
measurements taken on aircraft flights
approximately 150 m (500 ft) above the terrain,
suggested only slight dilution in the first 10 km (6
miles) north of the crest.18 Observations on other
days34 suggested a 50- to 60-percent reduction in
oxidant 10 km (6 miles) north of the crest. There
was no ground station at this point for comparison
with the aircraft measurements at either time.
( V _" Soledod
•• .'• ' ^A/> *:*""
r- 's'<~^ ">">
r~ "V'> T^-^-' '^ CW^>rCs—. C
f % AN ./i.N.JS-A^l^r
Son
>
V
-7
O ^ - 7-
*3H? ,-i _
rv /-" V'-X / •*
•V-/ \ ' S-V.
V,ctor,,ll.o > -; . >'_ .
• '"- r ^: ' •r'
<;
.~s - ^
, ..^ . T^.^ , - N-..- f "~—-» ^esoer.o"
I SAN /^J^A^R I Ej/,
?»^,f-? '^'^-'i"
^ &PT H%M.,^
San ( /
Fernando Volley
x^ s
. ' '-x'
Riolto-Miro -,
^PRedlands' ' X.t *
" .. ' ' '•' • . t' "-'"••-
' "~'x^-'-J
Feet above seo level
ABOVE 9,000
5,000-9,000
3,000-5,000
1,000-3,000
BELOW 1,000
Figure 12-2. Major topographic features of the Los Angeles Basin with inland valleys and mountains. Station locations: A.
Highland; B, City Creek; C, Mud Flat; D, Rim Forest. Aircraft flight paths for the study area are also shown.18
299
-------
Total oxidant, temperature, and vapor-pressure
gradient were measured continuously during 16
days in July and August at the mountain slope and
crest stations (A, B, C, and D, in Figure 12-2),
Figure 1 2-5 shows that the time of the daily peak
oxidant concentration was progressively later at
stations of higher elevation. Temperatures and
vapor-pressure gradients were also progessively
lower at higher elevations at the time of oxidant
peak. The duration of oxidant concentrations ex-
ceeding 200Afg/m3(0.10ppm) was 9, 13, 9 and 8
hr/day going from lower to higher stations. The
longer duration at City Creek (elevation 817 m, or
2680 ft) probably coincides with the zone where
the inversion layer most often contacts the
mountain slope. The oxidant concentrations rarely
decreased below 98 /L/g/m3 (0.05 ppm) at night on
the slope of the mountain crest, whereas they
usually decayed to near zero at the basin station
(Highland).
The vegetation zones along the slope and at the
crest are subjected to oxidant exposure differently.
Even though the lower chaparral receives longer
exposure, the peak concentrations coincide more
closely with the maximal evaporative stressforthe
day. There is some support for the hypothesis that
plant stomata would be closed during this period
and that pollutant uptake would thus be lower.
There is, in fact, very little visible injury to the
species in this zone; however, chaparral is more
sclerophyllous and therefore less likely to show
visible injury. In contrast, the daily oxidant peak
occurs well after the maximum temperature and
vapor-pressure gradients have occurred in the
conifer forest at the mountain crest, where oxidant
injury to plant speciesis severe. These suggestions
of possible microclimatic control of the sensitivity
of these native species to ozone form a working
hypothesis that needs further investigation.
2745
2135
H 1525
1U
x
915
305
0
CHAMISE
CONIFEROUS
FOREST
CITY CREEK
DEEP
CREEK
REDLANDS
HIGHLAND
I I
90
70
50
30
10
0
LU
X
12 16 20 24
DISTANCE, kilometers
28
32
36
Figure 12-3. Altitudinal sequence of ecosystems in the San Bernardino Mountains.34
300
-------
San Joaquin Valley and Adjacent Sierra Nevada
Mountains
Field surveys36'69 have confirmed oxidant injury
to ponderosa pine and associated species at
numerous locations in the Sierra Nevada foothills
east and southeast of Fresno. Oxidant meas-
urements at ground stations and by instrumented
aircraft show late-afternoon peaks of transported
oxidant on the western slopes of the Sierras.
Limited measurements by instrumented aircraft
suggest the development of a layer of oxidant
approaching the forested mountain slopes
between 610 and 1829 m (2000 and 6000 ft)
during the late afternoon.35 A very weak inversion
or isothermal layer may serve as a reservoir of
oxidant, which is advected to the mountain slope in
the southern coastal air basin, as suggested by
Edinger.17
Considerable concern has been registered about
air quality in the Lake Tahoe Basin, where local
development may cause adverse oxidant con-
centrations.
Seasonal and Daily Variations of Injurious
Concentrations—Synoptic Weather Patterns
Associated with Episodes of High Pollution
McCutchan and Schroeder30 classified 145 days
during May through September 1970 with respect
to pollution and meteorology on the basis of data
collected on the southern slopes of the Sari
Bernardino Mountains. Five general weather
patterns were described, along with the associated
synoptic patterns at the surface and at 500
millibars (mb) (Table 12-2) An analysis of eight
meteorologic variables was used to classify days
during May through September into the five
general categories. Of 123 days classified by
stepwise discriminant analysis, 10,13,44, 23, and
24 were placed in classes 1 through 5,
respectively; the remaining 9 days could not be
June 19,1970
Coast
El
0-0.10 ppm
0.10-0.20
0.20-0.30
0.30-0.40
ABOVE 0.40
Rialto-Miro
Redlands
TEMPERATURE INVERSION BASE
TERRAIN PROFILE
Figure 12-4. Daytime changes in oxidant concentrations along a west-to-east transect in the southern coastal air
basin, including the slopes of the San Bernardino Mountains (see Figure 12-2). Reprinted with permission from Edinger
et al.18
301
-------
o.
a
I
X
O
O
0.3S
0.30
0,25
0.20
0.15
0.10
0.05
i— 95
90
— ,85
UJ
DC
< 80
ct
a.
75
70
65
9 12
TIME, hours
15
18
24
18 o
<
oc
a
UJ
EC
12
HI
cc
a.
K
O
Q-
21
Figure 12-5,. The relationship of time of occurrence of the daily peak oxidant concentration to temperature and vapor-
pressure gradients in an elevationat sequence (see Figures 12-1 and 12-2) on the slopes of the San Bernardino Mountains,
July-August 1969,M
0.40
0.05
Figure 12-5 (continued). The relationship of time of occurrence of the daily peak oxidant concentration to temperature and
vapor-pressure gradients in an alevational sequence (see Figures 12-1 and 12-2) on the slopes of the San Bernardino
Mountains, July-August 1969 "
302
-------
0.35
X
E
— 18 O
cc
o
ill
CC
01
cc
a.
CC
o
a.
9 12 15
TIME, hours
18 21
Figure 12-5 (continued). The relationship of time of occurrence of the daily peak oxidant concentration to temperature and
vapor-pressure gradients in an elevational sequence (see Figures 12-1 and 12-2) on the slopes of the San Bernardino
Mountains, July-August 1969,34
Q
Z
I
DC
O
6
HI
0.20
Q 0.15
9 0.10
X
o
< 0.05
o
DC
O
ISO
160
140
HI
OC
EC
HI
a.
in
BO
75
70
65
VPG
I
15
12
D E
$ E
in .
EH
a. Z
cr •"
9 12 15
TIME, hours
18
21
Figure 12-5 (continued). The relationship of time of occurrence of the daily peak oxidant concentration to temperature and
vapor-pressure gradients in an elevational sequence (see Figures 12-1 and 12-2) on the slopes of the San Bernardino
Mountains, July-August 1969.34
303
-------
placed clearly in any of the five classes. The oxidant
concentrations associated with these classes were
Class 1, low; Class 2, low and high; Class 3, high;
Class 4, moderately high; and Class 5, low. In this
sample, the sum of days in classes 2 and 3 (57
days) suggests that about 46 percent of the sample
days had high concentrations of oxidant air
pollutants.
Class 2 and 3 days are both characterized by high
pressure (a 500-mb ridge over the area).
Descriptions of the synoptic weather patterns
(contained in the U.S. Forest Service California Fire
Weather Severity, 10-Day Summaries) for May
through August 1 972-74 were compared with
episodes of severe oxidant concentration at Rim
Forest and nearby Sky Forest. For the purposes of
this comparison, synoptic patterns were reviewed
for all days having a maximum hourly
concentration of 646 /;g/m3 (0.33 ppm) or higher.
In all cases, the persistent 500-mb high pressure
over the area was the most common synoptic
feature. For 8 qualifying days in 1972, the mean of
the maximum hourly concentrations was 724
fjg/m3 (0.37 ppm); for 16 days in 1973, it was 803
//g/m3 (0.41 ppm); and for 46 days in 1 974, it was
744 ^g/m3 (0.38 ppm). The highest hourly
concentration obtained at that time was 1176
//g/m3 (0.60 ppm) on June 28, 1974.
During episodes of severe oxidant pollution, the
weather is generally very hot (85° to 100° F, or
about 29° to 38°C). The relative humidity may be
either low or moderately high on class 2 and 3
days, depending on the behavior of the marine
layer. The small difference in the means of maxi-
mum hourly concentrations on high-pollution days
in 1972 through 1974 suggests that heavy primary
pollutant emission continues in spite of current
control strategies.
Annual Trends of Total Oxidant Concentrations
at a San Bernardino Station and the Nearby City
of San Bernardino
Since 1968, total oxidant concentrations have
been measured continuously with a Mast ozone
meter (calibrated by the California Air Resources
Board method) from May through September at
Rim Forest/Sky Forest.34Thefall,winter, andearly
spring months have generally been omitted until
recently, because synoptic patterns are usually not
conducive to oxidant accumulation and transport.
For example, average maximum hourly oxidant
concentrations from October through March 1973
and 1974 stayed below 196 /vg/m3 (0.10 ppm);
those for March were 1 96 to 294 /;g/m3 (0.10 to
0.15 ppm).3" The main data collection period
coincides with the growing season and thus
permits a reasonable estimate of the total annual
dose of oxidant air pollutant received by
vegetation.
In documentation of oxidant concentration
trends during the 1 968 to mid-1 976 period at Rim
Forest/Sky Forest, the accumulated dose is
expressed as the sum of all hourly values (in
//g/m3), including concentrations for each month
separately, June through September (the main
part of the growing season) (Figure 12-6), These
Class
TABLE 12-2. DESCRIPTIONS OF METEOROLOGIC PATTERNS FOR FIVE CLASSES OF
SPRING AND SUMMER DAYS IN SOUTHERN CALIFORNIA9
General weather
Hot, dry continental air
throughout the day
(Santa Ana)
Relatively dry forenoon,
modified marine air in
afternoon, very hot
(heat wave)
Moist, modified marine air,
hot in afternoon
Moist, modified marine air,
warrn in afternoon
Cool, moist, deep marine
air throughout the day
Associated synoptic pattern
Oxidant
concentration
Large high pressure over
Great Basin
High pressure over Great
Basin and thermal trough
over desert
Strong northerly winds over Low
area with trough east of
the area
Subtropical closed high over Low and high
area
Thermal trough over desert Ridge over area
Thermal trough over desert Trough over area
High
Moderately high
Synoptic low over desert
Deep trough or closed low Low
over area
Modified 1iom Relerence 3fi
304
-------
doses exclude background concentrations (those
less than or equal to 59 /vg/m3, orO.03 ppm).23The
percentage of valid data recovered, as well as the
total possible hours for which data could be
obtained for each month, June through Sep-
tember, is also indicated. The absence of some
data, ranging up to 17.8 percent in 1970, but
averaging 8.3 percent during the 7 years,
represents a margin of error that cannot at present
be adjusted with any certainty. The total number of
hours with concentrations of 160 /yg/m3 (0.08
ppm) or more during June through September was
never less than 1300 hours during each of the first
7 years. Recent predictions suggest that oxidant
doses will either increase annually or oscillate
around the mean of the present high levels unless
dramatic improvements occur.33
Data from the downtown San Bernardino station
operated by the County Air Pollution Control
District (APCD) are available back to 1963. The
colorimetric potassium iodide method used to
measure total oxidants was calibrated according to
the method of the California Air Resources Board.
A positive correction factor of 1.22 was used to
adjust mountain data for the decreased air
prer.sure at the higher elevation.
The data from the Rim Forest/Sky Forest station
were compared with published data from the San
Bernardino County APCD. The number of hours
with concentrations exceeding 392 /yg/m3 (0.20
ppm) during July, August, and September 1969
through 1974 were compared (Figure 12-7). For
1963 through 1968, data are shown only from the
San Bernardino APCD. A large part of the year-to-
year differences at the same station and between
stations can be attributed to differences in synoptic
and mesoscale (local) meteorologic patterns. For
example, the increases in 1 972,1 973, and 1974 at
Rim Forest/Sky Forest are as ociated with 6, 16,
and 46 days, respectively, when a persistent 500-
mb ridge occurred overtheSouthwest, particularly
southern California. The difference between
stations in the same year is probably influenced
most by inversion height. Lower inversions
partially restrain transport upslope to shorter
periods each day. Higher inversions would have
the opposite effect and also allow the greater air
volume below to dilute the oxidants. The index for
comparison chosen in Figure 12-7 (i.e., hours >
392 /yg/m3) would be sensitiveto inversion height.
The 3-year moving averages for each station tend
to remove some of the variation resulting from
PERCENT OF TOTAL HOURS OF DATA:
1970
1972 1973
AUGUST
1974 1975 1976
f!-!-!-!l SEPTEMBER
Figure 12-6. Monthly summation of total oxidant concentration data at Rim Forest/Sky Forest, San Bernardino Mountains,
California. June through September, 1968-75."
305
-------
meteorology. In terms of numbers of hours with
oxidant equal to or exceeding 392 /ug/m3, the
moving average between 1970 and 1973 at Rim
Forest/Sky Forest has increased from 1 75 to 290
hr. The increased oxidant concentrations at both
these stations are the reverse of those in upwind,
urban Los Angeles County, where increased
emmission of NOX tends to shift the chemical
equilibrium to the left, toward the ozone
precursors. The most recent data13 firmly indicate
that oxidant concentrations will either increase
annually or continue to oscillate around the mean
of present high concentrations in the foreseeable
future at these distant locations.
Robinson's tropospheric ozone cycle13 des-
cribing rural upwind, urban, and rural downwind
variations in concentration can be easily
demonstrated in the Los Angeles and connected
inland basins. The effects on both natural
ecosystems and agroecosystems also can be
demonstrated. For example, oxidant meas-
urements were made during August 1972 at six
oxidant stations (Costa Mesa, La Habra, Corona,
Riverside, City Creek, and Rim Forest) extending in
a line northeastward from the coast to the
mountains. These measurements show low doses
at the coast, increasing to a maximum on the
chaparral-covered mountains and decreasing
beyond the mountain crest (Figure 12-8). In Figure
12-8, the oxidant dose is indicated by the dashed
line. A crude estimate of the relative economic
value of ecosystems encountered along this
transect is expressed in a very general way by the
solid line and the nomenclature on the abscissa.
Finally, the relative complexity of the ecosystems
involved is shown below the abscissa. This
conceptualization emphasizes the enormously
greater dosage received by the natural ecosystems
during this month. This pattern of dosages is very
typical of June, July, and August, but the offshore
flow typified by the Santa Ana winds may reverse
the situation in September or October. Susceptible
crops growing on the coastal plain may be
seriously damaged.
In general, the permanent vegetation con-
stituting natural ecosystems receives much
greater chronic exposure than the short-lived,
higher-value vegetation constituting the agroe-
cosystems of the coastal plain. The latter can be
subjected to injurious doses, but in intermittent,
short-term fumigations. Each situation has
measurable economic and aesthetic effects, but on
different time scales. The simple agroecosystem
has little resiliency to pollutant stress; losses are
immediate and may be catastrophic. The complex
natural ecosystem is initially more resistant to
0 400
P
OC
1-
tu
Z 300
O
u
Z
£-
5 "Si
0 5 200
-J rg
?8
O u.
1- O
I
t 100
3
00
oc
O
Z
n
_ SAN BERNARDINO APCD (DOWNTOWN) [7
— RIM FOREST/SKY FOREST Fl
99 _
~" JULY, AUGUST, SEPTEMBER PERCENT DATA AVAILABLE RF/SF
98
— 3 -year MOVING AVERAGES:
APCD • •
—
RF/SF • •
—
— mm
""" j^
_ /'
ra. __•, n ^"'
— »--- * " n "~*
ELL y J_
_
78 93 91 96 .-i
/
f
'
•
•
./
- — * **•*
jjf
^^ T~*t
^
"W
n
**
I
•n
P
•.
.*.
•
'•m
*
>:
!*,
'I*
•^
«
l--
'.*
|«|
(*(
1 «
!;
—
_.
—
^H
—
i-—
—
•—
^^
— «
—
— 400
— 300
- 200
_ 100
1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974
Figure 12-7. Number of hours of total oxidant, July through September, greater than or equal to 392/ug/m3 (0.20 ppm) at the
downtown San Bernardino County Air Pollution Control District Station, 1963-74, and Rim Forest/Sky Forest, 1968-74.33
306
-------
pollutant stress, but the longer chronic exposures
cause the disruption of both structure and function
in the system. The damage may not be reversible.
INVESTIGATING THE EFFECTS OF OXIDANT
STRESS ON ECOSYSTEMS
Ecosystem Modeling
Ecosystems are very complex and extremely
difficult to understand. An approach to
understanding these complexities is to describe
them in mathematical terms. However, before
mathematical shorthand can be used, the
components to be studied must be defined and
their interactions and relationships with one
another identified. To understand the stresses
placed on ecosystems by perturbations such as
ozone-oxidant air pollution, those elements that
could be particularly vulnerable to pollutants must
also be identified. The mathematical and computer
capability for modeling ecosystems exists, but
ecological science does not yet possess sufficient
understanding of all of the interactions at the
ecosystem level 2 Therefore, simulation models of
forest ecosystems are extremely scarce.68
Modeling of the coniferous forest biome is
discussed by Overton,41 and modeling of the
eastern forest biome is treated by O'Neill.40 The
usual modeling approach is to select specific
processes such as energy flow and nutrient
cycling. These processes are a result of the
structural interrelationships existing among the
various components (soil, water, nutrients,
producers, consumers, and decomposers).48
Energy and nutrients may enter or leave an
ecosystem by a variety of pathways. These
processes may be described mathematically, and a
model may be developed that is an approximation
of the real world. This model may then be used to
make predictions about the ecosystem. Computer
programs for solving the equations explaining
interactions of the ecosystem components already
exist. Once the model is running smoothly on the
computer, it is possible to impose various
experimental conditions (e.g., clear cutting or
selective cutting) and observe the effects on
specific processes such as energy or nutrientflow.
Sensitivity analysis can then be performed to
determine which transfer coefficients are most
sensitive to stress and how much stress is required
to cause deterioration of the ecosystem.
O
o
o
U
UJ
CITY CREEK
NOTE: OXIDANT DOSES, WHICH ARE REAL
DATA, DEFINE THE POLLUTANT CONCEN-
TRATION GRADIENT FROM COAST TO INTERIOR.
-•30
CORONA
_L
-.20
COSTA MESA . .10
~~>___ LA HABRA
J_
X
a
a
CM"
a>
r-
a
o
O
O
Q
x
o
ui CLIMAX WOODY PERENNIAL
O VEGETATION SHRUBS GRASSES
a AND WEEDS
VEGETATIVE COMPOSITION SIMPLE AGROECOSYSTEMS
IRRIGATED
PASTURE
LOW
VALUE
CROPS
HIGH
VALUE
CROPS
COASTAL
LANDS
• COMPLEX NATURAL ECOSYSTEMS
Figure 12-8. Hypothetical relationships of oxidant dose, natural ecosystems, agroecosystems, and the relative values of their
vegetation components along a transect of southern coastal air basin of California.38
307
-------
No models dealing with oxidant stress on
communities or ecosystems exist; however,
models do exist that deal in general terms
with disturbance,29 sensitivity,82027'44 and
stability82027'44 at community and ecosystem
levels. Forest succession has been described by
Schugart et al. 47 and Botkin and Miller.7 The area
occupied per biotic unit over time is very important
for defining and predicting long-term successional
changes in plant communities as a result of
pollutant stress. In view of the trend of increasing
oxidant concentrations downwind from urban
areas,6 a much more precise understanding of
chronic effects on both natural ecosystems and
agroecosystems is required Descriptive and
predictive models dealing with the responses of
biologic systems can provide the best input to the
decision models needed for standard setting, land
use planning, and resource management.
Modeling the Effects of Oxidant Stress on a
Western Mixed Conifer Forest Ecosystem
The recent history of the mixed conifer forest of
the San Bernardino Mountains has been
analyzed.55 The analysis was conducted by nearly a
dozen coinvestigators representing many
disciplines.56 This analysis included an initial
inventory of ecosystem components and
processes, as indicated in Figures 12-9 through
12-13. The inventory emphasizes ponderosa and
Jeffrey pines, the most dominant species in the
climax community, and is stratified according to
organizational level- Organism or tree (Figures 1 2-
9 and 1 2-10), community or stand (Figure 1 2-11
and 1 2-1 2), and, finally, a time-oriented ana lysis of
plant succession in the aggregation of relatively
distinct communities or stands that collectively
make up the mixed conifer forest ecosystem
(Figure 12-13). These figures present in
diagrammatic form the complexity of ecosystem
interactions at each level of biological
organization. Only the interactions considered to
be the most important ones guiding the course of
plant succession in the several forest plant
communities now constituting the mixed conifer
forest have been selected by project investigators
for immediate investigation.57 In the case of each
interaction, it is important to decide the time
frequency at which the selected variables will be
measured (i.e., hourly, weekly, or annually)
Because this project is the most active effort to
date directed toward understanding the effects of
ozone and other oxidants on a natural ecosystem,
most of the remainder of this chapter is a
discussion of its findings.
Effects on Primary Producers (Green Plants)
The ozone dose responses of both herbaceous
and woody plants and the specific effects on the
photosynthetic activity (principally in controlled
short-term exposures) are discussed in Chapter
11. The aim of this section is to evaluate the effects
of the chronic exposure of vegetation in natural
ecosystems to total oxidants(more than 90 percent
ozone) under field conditions or simulated field
conditions. The effects of chronic exposure on
agroecosystems are also discussed to a limited
extent in Chapter 11.
Smith51 has suggested three classes of
ecosystem response to air pollutants: Those in
which vegetation and soils serve only as a 3ink for
pollutants, with no visible vegetation injury; those
in which some species, or sensitive individuals
within species, are injured and are moresubjectto
other stresses; and those in which high dosages
cause acute morbidity or mortality of some species.
Although these classes are convenient for
discussion, they are not clearly separated in space
in most situations. More often, these classes of
response occur along a gradient of decreasing
pollutant exposure, as in the San Bernardino
Mountains.
EXTENT AND INTENSITY OF INJURY TO
OVERSTORY TREES IN THE SAN BERNARDINO
NATIONAL FOREST
Wert66 used aerial photography to determine the
extent of oxidant injury to ponderosa and Jeffrey
pines in diameter classes larger than 30 cm. Injury
was categorized as heavy, moderate, light, or
negligible, and it generally decreased with
distance from the source area. Of the 64,380 ha
(160,950 acres) of ponderosa-Jeffrey type within
the forest boundaries, 18,492 ha (46,230 acres)
had heavy damage, 21,568 ha (53,920 acres) had
moderate damage, and 24,320 ha (60,800 acres)
had light or negligible damage. An estimated
1,298,000 trees were affected. Of these, 82
percent were moderately affected, 15 percent
were severely affected, and 3 percent had died.
The term "ponderosa-Jeffrey type" is a general
term that includes a mosaic of five subtypes
described by McBride on the basis of species
dominance.38 These subtypes are: Ponderosa pine
308
-------
forest, ponderosa pine/white fir forest, ponderosa
pine/Jeffrey pine forest, Jeffrey pine forest, and
Jeffrey pine/white fir forest. The injury by oxidant
air pollutants is most intense in the Jeffrey pine
types. In the field plots of these various types, as a
result of oxidant air pollution, the average area
covered by shrubs is only 3.8 percent in the
ponderosa types, but 26 percent in the Jeffrey pine
types.38
Later studies32'55'56 have expressed the amount
of injury to ponderosa and Jeffrey pines and
associated tree species in permanent study plots
as a numerical score. The range of scores is
subdivided into seven categories, ranging from
dead to no visible symptoms.32
In 1 974, all tree species at 19 permanent study
plots were scored individually by binocular
inspection. The data can be obtained from conifers
early in the fall, but the most important deciduous
species, black oak, was evaluated during the period
from August 28 to 31 to prevent confusion of
oxidant-injury symptoms with natural autumn
senescence of leaves. The injury to black oak on
August 31, 1974, at several representative study
sites, and the June-August accumulated dose at
nearby monitoring stations are shown in relation
to the topographic projection of the San
Bernardino Mountains in Figure 12-14.28 Lower
scores mean greater injury. The darkened portion"
of the bar representing oak injury is for leaf
OXIDANT
AIR
POLLUTANTS
SMALL
MAMMALS
SOIL
MICRO-
ARTHROPODS
CONE
INSECTS
7 8 *g 10 11 12 13 14
BARK
BEETLES
NATURAL
ENEMIES OF
BARK BEETLES
\ '\ IS/ ' PATHOGENS
25 V/26
22 /WOODPECKERS
AND OTHER
HOLE NESTING
BIRDS
WON- » 27
PATHOGENIC
FUNGI
Figure 12-9. Organism-level interactions in a mixed conifer forest.56
309
-------
1. Competition for food supply
2. Direct effect of ozone on small-mammal physiology
3. Climatic control of oxidant concentration in forest
4. Direct effect of ozone in soil microarthropods
5. Effect of precipitation and temperature on soil moisture and soil temperature
6. Insect damage to developing cones
7. Damage by small mammals to developing cones and importance of cone crop as animal
food supply
8. Direct effect of ozone on physiology of cavity-nesting birds
9. Role of fruiting bodies of nonpathogenic fungi in nutrition of small animals
10. Direct effect of ozone on tree physiology
11. Role of fruiting bodies of pathogens in nutrition of small animals
12. Direct effect of ozone on nonpathogenic fungi
13. Effect of temperature and evaporative stress on tree growth
14. Direct effect of ozone on pathogenic fungi
15. Interaction of nonpathogenic fungi and soil microarthropods
16. Interaction of pathogenic fungi and soil microarthropods
17. Effect of soil microarthropods on litter reduction
18. Impact of predation and parasitism on bark beetles
19. Predation of bark beetles by woodpeckers
20. Effect of bark beetles on tree mortality and vigor and effect of phloem thickness
and moisture on bark beetles
21. Effect of phloem moisture and thickness on natural enemies of bark beetles
22. Influence of woodpeckers on rate of parasitism
23. Effect of nonpathogenic fungi on tree growth
24. Effect of pathogens on tree vigor and mortality
25. Effect of soil moisture and soil temperature on tree growth
26. Effect of soil moisture and temperature on occurrence of nonpathogenic fungi
27. Interaction of pathogenic and nonpathogenic fungi
28. Effect of soil moisture and temperature on occurrence of pathogenic fungi
Figure 12-9 (continued). Organism-level interactions in a mixed conifer forest. Types of interactions.
chlorotic mottle and interveinal necrosis. A score
of 8 means no injury. The remaining portion of the
score is the sum of scores for leaf complement, leaf
size, and twig mortality (not shown separately).
These data suggestthat oak shows no visible injury
where the accumulated June-August dose does
not exceed about2.5 x105/yg/m3timeshoursfrom
about Snow Valley eastward.
The distribution of ponderosa and Jeffrey pines
into various injury classes with respect to the
distance of the study site along the gradient of
oxidant dose (June-September) is illustrated above
the topographic projection in Figure 12-15.28 It is
important to realize that the 1974 distribution into
injury classes is also a product of earlier years,
when the oxidant concentrations were not always
as high as in 1974 (see Figure 12-6). The trend
toward greater numbers in the very slight injury
(29 to 35) and no visible injury (36 and greater)
categories is quite evident in the eastern plots
receiving lower doses—for example, Holcomb
Valley. The assumption has been made that
ponderosa and Jeffrey pine respond similarly to
oxidant. Ponderosa pine is replaced by Jeffrey pine
in the natural stands east of Camp O'ongo, and
they intermix at Barton Flats. The validity of this
assumption can be partially verified by examining
the distributions of the two species into injury
classes at Barton Flats(Figure 1 2-1 5).28Thesedata
indicate reasonable similarity at a common site;
but the influence of other environmental variables
that change continuously along the oxidant
gradient (such as soil moisture availablity,
temperature, and humidity) must be examined
more intensively to understand the degree to
which they influence oxidant susceptibility (see
Figures 12-9 throughl 2-1 3).
OXIDANT DAMAGE TO CONIFER SPECIES IN THE
19 MAJOR STUDY PLOTS FROM 1973 TO 1974
The first evaluation of oxidant injury to all tree
species in the new study plots was completed in
September and October 1973. The second
evaluation in 1 974 offered the first opportunity to
compare increases in tree injury and mortality. In
Table 12-3, the plots are arranged in the order of
severe injury (first) to no visible injury (last),
according to the 1974 average injury score for all
ponderosa or Jeffrey pines in each plot. Of the 1 2
plots categorized as severe and moderate, all but 6
showed injury significantly increased (lower
scores), with a probability of 0.05. Among the six
310
-------
remaining plots where injury was classified as and two decreases (one significant and one
slight, very slight, or not visible, there were three insignificant), all at a probability of 0,05.
significant increases, one insignificant increase.
F OXIDANT 1
Figure 12-10. Tree-level interactions in a mixed conifer forest ecosystem (a = direct influence of external factors on trees; b and
c = influence of tree condition on external factors).56
311
-------
A, Relationship between net photosynthesis, foliage age, and foliage retention
B. Effect of plant moisture stress on net photosynthesis
C. Relationship between mineral uptake and net photosynthesis
D. Relationship between net photosynthesis and carbohydrate storage
E. Relationship between water uptake and mineral nutrition
F. Relationship between carbohydrate storage and bark characteristics
G, Relationship between carbohydrate storage and wood production
H. Relationship between carbohydrate storage and cone production
I. Relationship between cone crop and seed crop
6. Cone insect damage to developing cones
7a. Small mammal damage to developing cones
7b. Small-mammal predation of seeds
lOa. Effect of ozone on photosynthesis and respiration
lOb. Effect of ozone on needle retention
13. Effect of temperature, light intensity, and evaporative stress on photosynthesis
and respiration
20a. Effect of bark beetles on water uptake and transpiration, and relationships of
tree moisture stress to bark beetle attack
20b. Effect of bark beetles on tree carbohydrate concentration and relationships of
carbohydrate concentration to bark beetle population in trees
20c. Relationship between bark beetle characteristics and bark beetle attack and
population
21, Relationship between bark characteristics and natural enemies of bark beetles
23a. Relationship between root characteristics and mycorrhizal-forming non-
pathogenic fungi
23b. Effect of nonpathogenic fungi on mineral uptake
24a. Effect of pathogens on tree carbohydrate concentration and relationship of
carbohydrate concentrations to pathogen attack
24b. Effect of pathogens on water uptake and transpiration
24c. Relationship between stem and root characteristics and pathogen attack
25a. Effect of soil moisture and temperature on water uptake and transpiration
25b, Effect of soil mineral concentration and temperature on mineral nutrition of tree
Figure 12-10. Traa-level interactions in a mixed-conifer forest ecosystem.
Tree mortality among ponderosa and Jeffrey
pines was about the same m 1973 and 1 974. The
greatest mortality was in the permanent study
plots listed as receiving moderate injury. Possibly
the populations in these plots still retain greater
numbers of the more susceptible genotypes. In
previous years, tree mortality rates for ponderosa
or Jeffrey pines in the stands suffer ing moderate to
severe injury were 8 and 1 0 percent from 1 968 to
1972,55 8 percent from 1969 to 1971,32 and 24
percent from 1966 to 1969.12 The death of
weakened trees is usually due to the pine bark
beetle.63 Mortality has not been observed in tree
species other than ponderosa and Jeffrey pine.
Two 2-ha (5-acre) control plots in the vicinity of
Barton Flats in the San Bernardino National Forest
provide data for the longest observed period of tree
decline, 1952 to 1 972.65 These plots are located in
the Jeffrey pines, where trees with a diameter
breast height (dbh) of 30.5 cm (12 in.) or greater
were measured and their vigor described by
judging the risk or probability that they would be
susceptible to attack and kill by bark beetles
(Dendroctonus sp). Risk classes 1 and 2 indicate
low-risk trees that would definitely be preserved if
trees were being marked for a timber sale. Classes
3 and 4 are high-risk trees that would be marked
for removal in a timber sale. In Table 12-4, the
changes m merchantable volume in board feet (bd
ft) in all four classes are recorded for two control
plots in 1952, 1963, and 1972. The increase in
volume of trees in the high-risk category since
1952 is astounding, whereas the decreases in
volume of low-risk trees and total plot volume are
very large. The decrease in total volume is due to
one-by-one removal of bark-beetle-killed trees in
the plots as indicated by the increase in number of
snags and stumps; it is also possibly due to
suppressed radial growth.
Information available from the study of the
ponderosa pine subtype suggests that the biomass
production of overstory species is diminished in
proportion to the oxidant dose received. Parmeter
et al.43 observed decreases in the height of
312
-------
ponderosa pine that showed injury symptoms.
Injured trees did not respond with greater growth
during years with more favorable soil moisture
content; but uninjured trees (often side by side
with injured trees) did increase in height during
these years. In another study38 of two populations
(one ranging in age from 52 to 71 years and the
other from 20 to 39 years) of dominant ponderosa
pine stands, the influence of the Los Angelessmog
was noted. The width of growth rings before 1 940
was compared with those after 1940. After the
influence of precipitation on growth was
evaluated, a difference of 0.20 cm in average
annual growth was attributable to oxidant air
pollution injury.
OXIDANT
AIR
POLLUTANT
SMALL
MAMMALS
SOIL
MICRO-
ARTHROPODS
CONE
INSECTS
BARK
BEETLES
10 9 12
NATURAL
ENEMIES
OF BARK
BEETLES
WOOD-
PECKERS
AND OTHER
HOLE NESTING
BIRDS
IMON-
PATHOGENIC
FUNGI
Figure 12-11. Community-level interactions in a mixed-conifer forest ecosystem.5
313
-------
1, Competition between woodpeckers and small
mammals
2, Climate control of oxidant concentration in different
forest communities
3. Effect of precipitation and temperature on soil
moisture and soil temperature in different forest
communities
4, Predation of bark beetles by woodpeckers in different
forest communities
5, Effect of cone crop abundance on cone insect
populations in different forest communities
6. Effect of cone crop abundance on smatl-mammal
populations in different forest communities
7, Fruiting bodies of nonpathogenic fungi as food for
small mammals in different forest communities
8. Smog-caused mortality and morbidity in different
forest communities
9. Fruiting bodies of pathogens as food for small
mammals in different forest communities
10. Effect of temperature and evaporative stress on
species composition in different forest communities
11. Relationship between soil characteristics and
population density of burrowing small mammals in
different communities
12. Relationship between plants and microarthropod
population
13. Relationship between soil characteristics and
microarthropod population
14. Bark beetle mortality caused by natural enemies in
different forest communities
15. Effect of bark beetles on tree mortality and vigor in
different forest communities
16. Relationship between soil characteristics and forest
community composition and growth
17. Relationship between soil characteristics and species
distribution and behavior of nonpathogenic fungi
18. Relationship between soil characteristics and species
distribution and behavior of pathogens
19. Influence of forest community type on populations
of natural enemies of bark beetles
20. Woodpecker distribution and density in different forest
communities
21. Effect of pathogens on tree vigor and mortality in
different forest communities
22. Relationship between nonpathogenic fungi and
forest community composition and growth
Figure 12-11. Community-level interactions in a mixed- conifer forest ecosystem.5
The growth suppression of ponderosa pine
saplings has been demonstrated experimentally by
Thompson.58 Trees grown m a plastic-covered
greenhouse and provided with charcoal-filtered air
(after an initial lag in the growth of shoots and
branches in the first years of the experiment)
showed dramatically increased growth when
compared to trees grown in ambient air outside or
trees grown in a greenhouse with unfiltered air
(Figure 12-16)28 Understory ponderosa pines
appear to be more susceptible to oxidant than
larger-sized trees 12 The probable effect of tree
mortality on stand composition, based on the
present species and size-class composition, is
shown inTable 1 2-5. The incense cedar and sugar
pine are the most tolerant to oxidant; however, the
sugar pine is present only in very low numbers. The
accelerated mortality of ponderosa pines has
particular significance m this ecosystem because it
is the dominant member of the climax community
The direct effects of ozone on plant species
constituting the shrub layer m the conifer forest
are not yet sufficiently understood to permit any
conclusions to be drawn. Initial field observations
suggest that at least 10 species are obviously
injured by total oxidant m areas receiving high
dosages These areas have been subject to
exposure for at least 20 years, so some species or
subspecies of annual plants may have been
completely eliminated.
In a series of experiments, Treshow and
Stewart64 fumigated native plants growing under
natural field conditions. The ozone concentrations
required to injurethemore prevalentspecies mthe
grassland-oak and aspen-conifer communities
during 2-hr exposures to 290, 490, 588, and 785
fjg/m3 (0 1 5,0,25,0 30, and 0 40 ppm) ozone were
determined. Seventy common plant species
indigenous to the plant communities were
fumigated. Of these, 5 species were injured at 290
/ug/m3 (0 1 5 ppm), and 20 species at 490 /ug/m3
(0.25 ppm). These species included grasses,
perennial forbs, and trees, and they were
distributed over all vegetation zones (seeTable 1 2-
6). The sensitivity of aspen in the aspen-conifer
community was considered of great significance.
Not only is it dominant, but its presence is vital to
the integrity of the community. The shade of the
aspen is required by the seedlings of the white fir
(Abies concolor) and some other species of plants
for growth. In the aspen forest of the Wasatch
Mountains, where this study was conducted,64 the
tree canopy acted as an effective filter. When
ozone concentrations of 1770 /jg/m3 (0.9 ppm)
were measured in the canopy and m clearings, the
concentrations under the canopy were only 196
314
-------
/jg/m3{0.1 ppm). It was also noted that though one
exposure to 290 /^g/m3 (0.15 ppm) appeared to
produce no effect, repeated exposures could injure
both growth and reproduction. The implications for
possible imbalances in community stability are
readily apparent if these plant communities
received dosages similar to those typical in the San
Bernardino Mountains.
Figure 12-12. Stand-level interactions in a mixed conifer forest ecosystem (a = direct influence of external factors on trees in
stand; b and c influence of stand condition on nonpathogenic fungi)?6
315
-------
A. Effect of plant competition on abundance of cone and seeds
B. Effect of plant competition on characteristics of living trees
C, Effect of plant competition on characteristics of shrubs
D. Effect of plant competition on characteristics of herbs
E. Effect of plant competition on tree mortality
5. Abundance of cones and seeds and predation by cone insects
6a. Abundance of cones and small-mammal predation of cones
6b. Characteristics of forest stands and small-mammal populations
6c. Influence of shrub and herb layer vegetation on small-mammal populations
ga. Smog-caused mortality and morbidity in different tree species with forest
stands
8b. Influence of stand conditions on concentration of oxidants
Ida. Influence of temperature and evaporative stress on stand structure and
composition
12b. Influence of stand condition on temperature and evaporative stress
15a, Influence of bark-beetle caused tree mortality on stand condition
15b. Influence of stand condition on bark beetle population
16a. Influence of soil moisture and temperature on stand characteristics
16b. Influence of stand condition on soil characteristics
19, Influence of stand condition on population dynamics of natural enemies
of bark beetles
20, Relationship between smog occurrence and woodpecker population
21 a. Influence of pathogen-caused mortality on stand condition
21b. Influence of stand condition on pathogen population
22a. Influence of mycorrhiza fungi
22b. Influence of snags land downed trees) on nonpathogenic fungi
22c. Influence of stand condition on nonpathogenic fungi
Figure 12-12 (continued). Stand-level interactions in a mixed conifer forest ecosystem."
EFFECTS ON REPRODUCTION
The effect of ozone injury on herbaceous plant
reproduction has been mentioned earlier in this
chapter and in Chapter 11. Seed production by
annuals is influenced mainly by the environmental
conditions of the cur rent year, but perennial woody
plants, particularly conifers, are erratic seed
producers. Factors affecting cone production
include age, vigor, seasonal temperature, and soil
moisture 2)
Decrease in tree vigor may decrease or totally
eliminate cone production or result in a decrease in
the size, weight, and germination of the seed
produced.45'46
Fruit and seeds make up the largest part of the
diet of most of the common small mammals on the
study sites, particularly the deer mouse fPero-
myscus sp.), harvest mouse (Reithrodontomys),
chipmunk (Eutamias), ground squirrel
{Callospermophilusj, and western gray squirrel
(Sciurus gr/seus anthonyi). The gray squirrel is an
excellent example of the interactions within this
forest and of the potential effects of oxidant air
pollution. It is abundant throughout the mixed-
conifer-type forest, depending specifically on the
pines and oaks for most of its food, cover, and nest
sites. This squirrel eats or stores a major portion of
the pine and oak seed crops each year. On some
yellow pine trees in the study plots, gray squirrels
cut more than 2000 cones per tree. During periods
of low seed production because of diminished tree
vigor, squirrels converge on the few remaining
vigorous ponderosa pines and consume about half
the seed crop before it matures and reaches the
ground. In the areas severely affected by ozone
(oxidant), squirrels return to the same trees year
after year. After the seed reaches the ground, other
small vertebrates, such as mice, seek it out. The
habit of preferential seed use by small vertebrates,
when considered additively with ozone injury,
could seriously reduce the regeneration potential
of ponderosa pine.
Thus tree squirrels are a major souce of loss of
seeds from the ponderosa, Jeffrey, and sugar
pines, and of black oak acorns. Vertebrates, then,
can have a major effect on the reproduction of
these species, particularly because the gray
squirrel is only one of numerous species in this
forest that feed on conifer seeds and acorns
Removal of the dominant species in a specific
community can have a profound effect on re-
production of other species. Mention has already
been made of the effect of the disappearance of the
aspens from the aspen-conifer communities. The
316
-------
r OXIDANT 1
SOIL
CHARACTER-
ISTICS
1. Oxidant modification of community structure
2. Climate modification of community structure
3. Fire modification of community structure
4. Soil characteristics and their influence on community
structure
6. Insect modification of community structure
6. Mammal modification of community structure
7. Pathogen modification of community structure
8. People impact on community structure
9. Logging modification of community structure
10. Predictive capability through integration of submodels
for items 1-9
Figure 12-13. Community-succession interactions in a mixed conifer forest ecosystem.56
317
-------
( ) AIR MONITORING
v STATION
Figure 12-14. Topographic projection, San Bernardino Mountains, with comparison of oxidant injury to black oaks at major
study sites, August 31, 1974. with accumulated total oxidant dose for June-August measured at nearby monitoring
stations.28
O- AIR MONITORING STATION
A, dead, 0; B, very severe, 1-8;C, severe, 9-14; D, moderate, 15-21; E, slight, 22-28; F, very slight, 29-35; G, no visible damage, 36+
Figure 12-15. Topographic projection, San Bernardino Mountains, showing how ponderosa pines (PP) and Jeffrey pines (JP)
in major study sites are distributed in six injury classes according to seasonal dose of total oxidant.28
318
-------
reproduction of white fir is reduced or eliminated,
since the shade required by white fir seedlings is
no longer present.64 Other shade-tolerant species
are also affected.
Changes in such physical factors as light, tem-
perature (particularly maximums and minimums),
relative humidity, and wind speed for forest
communities subject to structural alteration by
mortality of susceptible species could change the
suitability of some sites for growth, reproduction;
and reestabfishment of survivor species. Some of
the possible secondary effects are only speculative
until more data are gathered.32
EFFECTS ON CONSUMER POPULATIONS
Vertebrate Populations
The effects of oxidant air pollutants on verte-
brates can be separated into direct and indirect
categories. Direct effects include clinical and
pathological alteration of tissues as a result of
TABLE 12-3, CHANGES IN OXIDANT INJURY SCORES AND MORTALITY RATES OF PONDEROSA AND JEFFREY
PINES AT 18 MAJOR STUDY PLOTS, 1973-7438
Schneider Creek
Camp O'ongo
Sky Forest
University Conf Center
Breezy Potnt
Camp Paivika
Dogwood A
Tunnel Two Ridge
Camp Angeles
Barton Flats
Barton Flats
Snow Valley 2
Green Valley Creek
Camp Oceola
Bluff Lake
N E Green Valley
Heart Bar
Sand Canyon
Holcomb Valley
JP
PP
PP
PP
PP
PP
PP
PP
PP
PP
JP
JP
JP
JP
JP
JP
JP
JP
JP
Tree
density/
1973
28
90
144
309
236
217
168
122
112
200
124
129
43
192
186
120
130
56
193
Av«r«
sc
1973
12.4
15.1
133
15.5
163
17 1
19,9
19 5
25,8
21 4
21 0
221
21 8
|« injury
ore"
1S74
11 7
129"
137
15.6
160
16 4
16.5"
167"
168b
187"
19 T
19.7"
20 5b
Morli
rate,
1973
00
0.0
08
00
2 7
00
1 2
00
00
37
3.6
0.0
0,0
lllfV
%
1974
0.0
0,0
0.8
1 5
2.7
31
0,0
1 4
1,5
37
36
1.0
1 5
Accumulated
mortality. %
1974 Injury description
0 0 Severe
0,0
1 6
1 5 Moderate
5.5
31
1 2
1.4
1.5
7,4
73
1.0
1.5
21 7
29 4
33 1
440
41.3
46 4
22 gb
31 3"
32 1
39.2°
473"
477
08
0.0
0 0
0,0
00
00
72
00
0.0
09
00
00
80
0.0
00
09
0.0
00
Slight
Very slight
No visible symptoms
"Number of trees per hectare
^Difference significant at probability of DOS ^comparisons valid r
ily between
at a single plotl
TABLE 12-4. CHANGES OF TIMBER VOLUME AND PERCENTAGE OFTOTAL JEFFREY PINES" AT BARTON FLATS IN THE
SAN BERNARDINO NATIONAL FOREST38
1952
Risk classes
Timber
volume . bd ft
1963
1972
Percentage Timber Percentage Timbar
of trees volume, bd ft of trees volume, bd ft
Percentage
of trees
Control plot 1 (JCA Camp, Highway 38)
Total, all classes
Risks 1 and 2
Risk 3
Risk 4
Snags and current stumpsb
Total, all classes
Risks 1 and 2
Risk 3
Risk 4
Snags and current stumps"
73,040
58,520
6,740
7,780
1
120,130
110,830
5,990
3,310
3
100
87
7
5
1
100
93
3
2
2
63,530
38,700
14,630
10,200
1 1
Control plot 2 (Camp
112,660
98,080
10,170
4,410
13
100
73
13
7
7
Oceofa
100
82
6
6
6
52,730
23,780
14,140
14,810
13
Road)
112,930
45,670
37,420
29,840
18
100
55
16
20
8
100
32
30
28
10
**|n four insect rtsk classes 01 two control plots excluded from sanitation salvage Logging lieiweffln 1952 and 1972 "i!
AccuFT'uiartion during lO-v&ar pa nod Data obtained from the Supervisor's OHice, San Bernardino National Forest
319
-------
exposure to ambient air. Indirect effects result
from alterations in numbers or distribution of the
plant and animal population exposed to ambient
air. For example, if air pollution eliminates or
decreases the quantity of a susceptible plant
species, the food chain of the consumers that feed
on it may break down. The result could be a simpler
and less stable ecosystem, with fewer numbers
and species of plants and animals.
The clinical and pathological effects of oxidant
air pollutants on domesticated vertebrates have
been studied in the laboratory. No major refer-
ences to studies of these effects on free-ranging
native species seem to exist. The possible
interactions of vertebrates, determined by
extrapolation from laboratory studies, are shown
in Figures 12-11 through 12-13 for the San
Bernardino Mountain ecosystem.
When changes occur in one part of an
ecosystem, the intimate nature of the inter-
relationships results in changes in many other
parts. Any factor that causes change in one
component of a system potentially affects all
subsystems of that ecosystem. The most important
indirect effects of oxidant air pollutants on
vertebrates are those resulting in changes in the
habitat. Foremost among these effects are those
on the vegetation and the successional patterns of
a
z
ID
ANNUAL GROWTH OF BRANCHES, UPPER HALF OF SAPLINGS
32 -
30 -
28
26
24
22
20 •<
18
16 *
14 '
12 '
10 J
8 -i
6 J
4 -
TERMINAL
SHOOT GROWTH
1- AMBIENT AIR OUTSIDE
2 - AMBIENT AIR HOUSE
3 - FILTERED AIR HOUSE
T
4-
1
~i
l
T
rf
123 123
1969 1970
r
1
_
2 3
1971
T
r
1 2 3
1972
1
T
2 3
1973
32-
30
28
26 "
o 24
I 22-
0 20
g 18
-i 16
14-
12-
10-
8-
6-
4-
FIRST
1
2
3
-AMBIENT AIR
ORDER BRANCH GROWTH
OUTSIDE
-AMBIENT AIR HOUSE
- FILTERED AIR
T •
T
1 Tr
. \
123 12
HOUSE
3
r
1969 1970
1
'
\
2 3
1971
• r
123 1
i
p
2 3
1972 1973
Figure 12-16. Annual growth of the terminal shoot (upper)
and first-order branches (lower) in upper half of ponderosa
pine saplings maintained in filtered or unfiltered (ambient)
air greenhouses, or in outside ambient air (1968-73).2a
320
-------
the plant community. Because of the high degree
of interrelationship and interaction of the
vegetation, the fauna, and the inorganic matrix of
an ecosystem, effects of air pollution on the
vegetation potentially can result in changes
throughout the ecosystem. Damage to vegetation
is probably the most important effect of chronic,
low-concentration air pollution on wildlife.
Ponderosa pine, Jeffrey pine, and black oak are all
susceptible to injury, and these are the most
important trees within the forest as providers of
food and habitat for wildlife. A similar selectivity by
species doubtless occurs within the shrub and
herb layers of the vegetation. The long-term effects
will be (1) reduced production of fruits and seeds
and (2) elimination of the sensitive plant species
and a consequent reduction in the diversity of the
vegetation. These effects will in turn lead to a
reduction in the abundance and diversity of
vertebrate fauna (in the numbers of pme-seed-
eating squirrels, for example).
Likewise, Woodwell's prediction71 of en-
hancement of the activity of insect pests and some
disease agents (which has been demonstrated in
this forest) could lead to an increase in vertebrate
species that feed on invertebrates or utilize dead
p'ants for cover. Birds would be the most likely to
increase, as would (to a smaller extent) such small
rrammals as deer mice, which are partially
ir sectivorous.
1 ABLE 12-5. TREE SPECIES AND SIZE COMPOSITION IN
£ STUDY AREA8 AFFECTED BY OXIDANT AIR POLLUTION3
ree size and class
Ponderosa
pine
Incense
cedar
White
fir
Sugar
pine
Understory,
number/acre
Seedlings (up to
3 00 ft tall) 1057 2381 1043 302
Saplings (more than
3 01 ft tall, less
than 3 99 in dbh) 3 33 57 10
Poles (4 00 to 11 99
in dbh) 21 12 38 3
As % of total 22.2 48.6 22.8 6 3
Overstory.
number/acre
Standing (12 00 to
23 99 in dbh)
Veteran (24 in dbh
and larger)
As % of total
>8
12
496
9
5
22.7
8
4
19.7
3
2
8.0
The indirect effects of ozone through modi-
fication of the availability of food for insects,
particularly in a conifer forest ecosystem, has
received some investigation.52'53 The weakening of
ponderosa pines by chronic exposure to ozone
makes them more vulnerable to successful
infestation by pine bark beetles (Dendroctonus
brevicomis and D. ponderosae). Figure 12-17
shows a positive relationship between degree of
tree injury and frequency of bark beetle infestation
and the relative frequency of attack by the two
species of bark beetles.52 The relationship (see
Figures 12-10 through 1 2-T3) between tree health
and brood productivity and the population
dynamics of bark beetles and their insect
associates in infested trees is under
investigation.3856 Pine bark beetles have been a
constant threat to ponderosa pines in the San
Bernardino Mountains for many years, since
before the inception of oxidant air pollution
injury.54 Bark beetles are a key element in
accelerating the modification of stand structure.
"Trees within a 575-acre study area, San Bernardino National Forest, Calif
Plant Parasites and Symbionts
Ozone injury to needles of eastern white pine
increased infection by Lophodermium pinastri and
Aureobasidium pullulans.^ These results suggest
that leaf tissue of many species may become
susceptible to fungi that are normally saprophytes
but become parasites when circumstances permit.
Another way that oxidant air pollution could
affect this ecosystem is through an alteration in
forest moisture. Elimination of vegetation cover
allows the exposed soil to dry more rapidly, which
would affect soil-burrowing and soil-inhabiting
vertebrates. Also, the lower moisture content
could reduce or inhibit fruiting-body formation of
fleshy fungi. These fleshy fungi are an important
food source for tree squirrels and make up a third
or more of their diet in some seasons. A reduction
in this food source would doubtless result in an
even greater utilization of conifer seeds and
acorns, thus reducing further the reproductive
capability of these trees and eventually limiting
future food supplies for the squirrel population.
Beneficial mycorrhizal fungi infect the small
feeder roots of trees and other plants. The resulting
relationship is symbiotic and involves an intimate
exchange of minerals and essential metabolites.
The host tree benefitsthrough increased efficiency
of nutrient uptakef rom the soil. Any interruption or
imbalance of the exchange of materials between
the host root tissue and the fungus mantle
321
-------
surrounding it can have deleterious effects on the
fungus and the host. Such stresses as air pollution
injury to the host undoubtedly disrupt this
balance.25 The feeder rootlet system of ponderosa
pines in the San Bernardino Mountains and those
of eastern white pine have shown marked
deterioration involving a decrease in numbers of
mycorrhizal rootlets and their replacement by
saprophytic fungi that decay the small rootlets.43
Root-infecting fungi such as Armiltaria me/tea
and Formes annosus are generally more virulent
pathogens when they encounter trees already
weakened by other stresses. This observation has
been made mostly in Europe, where sulfur dioxide
was the principal pollutant.
EFFECTS ON DECOMPOSERS
Although some of the solar energy fixed by
producer plants is released by the respiration of
these plants and of animals, much of it is stored in
dead organic matter until released by decomposer
organisms at rates that vary greatly with place,
season, and kind of orga nic matter. Generally, one-
third or more of the energy and carbon fixed
annually in the forests is contributed to the forest
floor as litter (mostly leaves),43 Because litter is
generally related to the quantity of photosynthetic
tissue in the ecosystem, it is a useful index of
ecosystem productivity.
TABLE 12-6. INJURY THRESHOLDS FOR 2-HOUR EXPOSURES TO OZONE
Specses
naury threshold
ippm ozone
far 2 hrj
Grassland-oak community species
Trees and shrubs
Acer grandidentatum Nutt
Acer negundo L
Artemesia tridentata Nutt,
Mahonta repens G. Don
Potentille fruiicosa L,
Quercus gambelu Nutt.
Toxicodendnn radicans (L.) Kuntze
Perennial forbs
Achi/lea miltefalium L.
Ambrosia psilostachya DC
Caiochortus nuttaltti Torr
Cirsium arvense (L.) Scop.
Conium maculatum L
Hedysarum borea/e Nutt
Hetianihus anuus L
Medocagp satova L.
fiumex crispus L.
Unica gracilts Ait
Vicia amencana Muhl
Grasses
Bromus braaeformis Fish, St May.
Bromus tectorum L
Poa pratensis L,
Aspen and conifer community species.
Trees and shrubs'
Ab/es concotor (Gord 81 Glend.) Lindl
Amelanchier alnifotia Nutt,
Pachystima myrsimtes (Pursh) Raf,
Populvs tremuloides Michx.
Ribes hudsomanum Richards,
Rosa woodsii Lindl
Sambucus me/snocarpa A, Gray
Symphoncarpos vaceimonlers Rydb.
Perennial forbs
Actaeu arguta Nutt.
Agastache urt/c/folia (Benth.) Kuntze
njury threshold
Ippm ozone
for 2 hr!
Perennial forbs
A/ltum acummaturn Hook 25
Angelica pinrtata S Wats. under 25
over 40 Aster engelmanni {Eat.} A Gray 15
over ,25 Carex siccata Dewey 30
40 Cichor turn imybus L. 25
over .40 Cirsium arvense (L) Scop under 40
.30 Epilobtum angustifolium L 30
.25 Epilobium watsom Barbey 30
over 30 Enogonum heracltoides Nutt 30
Fragana ovalis (Lehm.) Rydb 30
Gent/ana amare/ta L over 15
over 30 Geranium fremontu Torr, under 40
over 40 Geranium nchardsonn Fisch & Traut 15
over 40 Juncus sp over 25
40 Lathyrus lanzwenu Kell over ,25
over 25 Lathyrus paucifiorus Fern 25
.15 Mertensia ariionica Greene 30
over ,30 Mimulus guttatus DC over 25
25 Mimulus moschatus Dougl under 40
25 Mitel/a stenopetala Piper over ,30
3O Osmorhiza ocadernalis Torr 25
over 40 Phacelia heterophytla Pursh under .25
Polemontum fo/tosissimum A Gray 30
Rudbeckia occidentalis Null 30
.30 Saxifraga arguta D Don under 30
15 Senecio serra Hook. 15
25 Taraxacum ofhcinale Wiggers over 25
Tha/ictrum lendlen Engelm over 25
Veronica anagallts-aquattca L 25
Victa amencana Muh! over 25
Viola adunca Sm over 30
.25
.20 Annual forbs
over .30 Chenopodium fremontu Wats. under .25
15 Callomia /means Nutt under .25
30 Descuramia catifornica (Gray) 0 E Schuls 25
over 30 Gahum bifoltum Wats. over 30
over .25 Gayophytum racemosum T & G 30
.30 Polygonum douglasu Greene over .25
Grasses.
25 Agropyron caninum (L.) Beauv over 25
20 Bromus cannatus Hook & Arn under 25
322
-------
One of the predicted effects of pollutants on
ecosystems suggested by Woodwell71 is a
reduction in the standing crop of organic matter,
which would lead to a reduction in nutrient
elements held within the living system. The
evidence discussed earlier definitely shows that
primary production is much lower in an ozone-
stressed conifer forest ecosystem. This result
would be anticipated in all natural ecosystems or
agroecosystems under similar stress.
The reservoir of energy and mineral nutrients
represented by litter is a very important resource in
natural ecosystems with closed nutrient cycles.
The growth of new green plant tissue depends on
the slow release of nutrients by decomposer
organisms. In agroecosystems geared for high
production, litter is often removed or burned, and
fertilizer is added to the soil; the nutrient cycle is
open and subsidized.
In a conifer forest, litter production and
decomposition release about 80 percent of the
total minerals in the biomass of the stand; the
remainder is retained in the living parts of the
tree.31 Standing dead material is not considered
litter.
In terrestrial ecosystems, most decomposers
occupy the mantle of litter on the surface layers of
the soil, where they supply the necessary recycling
mechanisms to convert dead plant or animal
material into humus and eventually into minerals,
gases, and water. Small animals, arthropods,
fungi, and bacteria exist as a complex in intricate
food chains in which they feed on dead material
and on one another as well, ultimately releasing
the mineral nutrients needed by the producer
populations. Without the decomposers, some
essential elements (such as calcium, phosphorus,
and magnesium) would concentrate in the litter
until the supply in the soil wasdepleted. Growth of
green plants would then be seriously limited.
It is not known whether ozone, PAN, or other
oxidants could have any direct influence on
decomposer organisms other than fungi (see
Chapter 11, section on responses of mosses, ferns,
and microorganisms) in the litter layer. But there
does appear to be a rapid flux of ozone to soil
surfaces. The ozone flux to some kinds of surfaces
constituting ecosystems (e.g., vegetation, soil, and
water) has been determined by Aldaz,1 who
expressed the flux as molecules/cm2-sec* 1 O.The
relative fluxes into differentsurfaces, assuming an
ozone concentration of 40 pg/m3, were: Fresh
water, 0.5; snow, 0.9; grass, 1.1; sandordry grass,
5; and juniper bush, 10, Furthermore, it was found
that bare soil destroyed about 75 percent more
ozone when dry than when moist. The
determination of ozone flux to surfaces may be a
far more realistic measure of dose to living
organisms than atmospheric concentration of
ozone, according to Munn.37
In summary, it is anticipated that decreasing
litter production by green plants experiencing
pollutant stress would result in a similar reduction
in the inventory of nutrient elements held within
(A
HI
HI
CC
o
a.
ui
to
2
z
500
400
300
200
100
TOTAL NUMBER OF TREES
| TREES KILLED BY BARK
BEETLES
a
UJ
-J
_J
ui
X
o
cc
UJ
CD
D
60 r
50
40
30
20
10
0, ponderosae ONLY
D, brevicomis ONLY
ATTACKED BY BOTH
SPECIES
HEALTHY
LIGHT MODERATE SEVERE
INJURY INJURY INJURY
HEALTHY LIGHT MODERATE SEVERE
INJURY INJURY INJURY
SMOG INJURY RATING
SMOG INJURY RATING
Figure 12-17. Relationship of degree of oxidant injury in ponderosa pines with bark beetle attack (left) and number of trees
killed by western pine beetle, mountain pine beetle, or both species (right).46
323
-------
the system, owing to the interruption of nutrient
cycling pathways and mechanisms of nutrient
conservation.71 72
Overall, it can be seen that subtle and simple
initial changes may radiate and magnify
throughout all trophic levels of the ecosystem.
Restoration of the system may be impossible,
SUMMARY
Plants, animals, and microorganisms usually do
not live alone but exist as populations Populations
live together and interact as communities.
Communities, because of the interactions of their
populations and of the individuals that constitute
them, respond to pollutant stress differently from
individuals. Man is an integral part of these
communities, and as such, he is directly involved in
the complex ecological interactions that occur
within the communities and the ecosystem of
which he is a part.
The stresses placed on the communities and the
ecosystems in which they exist can be far-
reaching, since the changes that occur may be
irreversible. For example, it has been suggested
that the and lands of India are the result of
defoliation and elimination of vegetation that
induced local climatic changes not conducive to
the reestabhshment of the original vegetation
An ecosystem (e.g , the planet Earth, a forest, a
pond, an old field, or a fallen log) is a major
ecological unit made up of living (biotic) and
physical (abiotic) components through which the
cycling of energy and nutrients occurs. A
structured relationship exists among the various
components. The biotic units are linked together by
functional interdependence, and the abiotic units
make up all of the physical factors and chemical
substances that interact with the biotic units. The
processes occurring within the biotic and abiotic
units and the interactions among them can be
influenced by the environment
Ecosystems tend to change with time. Adap-
tation, adjustment, and evolution are constantly
taking place as the biotic component, the
populations, and the communities of living
organisms interact with the abiotic component in
the environment. Recognizable sequential
changes occur With time, populations and
communities may replace one another This
sequential change, termed "succession," may
result in climax communities The latter, which are
structurally complex and more or less stable, are
held in a steady state through the operation of a
particular combination of biotic and abiotic factors.
The disturbance or destruction of a climax
community or ecosystem results in a return to a
simpler stage. Existing studies indicate that
changes occurring within ecosystems in response
to pollution or other disturbances follow definite
patterns that are similar even, in different
ecosystems. The basic biotic responses to the
disturbance of an ecosystem can thus be broadly
predicted.
Diversity and structure are most changed by
pollution as a result of the elimination of sensitive
species of flora and fauna and the selective
removal of the larger overstory plants in favor of
plants of small stature. The result is a shift from the
complex forest community to the less complex,
hardy shrub and herb communities. The opening of
the forest canopy changes the environmental
stresses on the forest floor, causing differential
survival and, consequently, changed gene
frequencies in subcanopy species. Associated with
the reduction in diversity and structure is a
shortening of food chains, a reduction in the total
nutrient inventory, and a return to a simpler and
less stable successional stage.
It should be emphasized that ecosystems are
usually subjected to a number of stresses at the
same time, not just to a single perturbation {e.g.,
oxidant).
The effects of oxidants on the San Bernardino
Forest graphically demonstrate the changes that
occur in natural ecosystems, as discussed earlier.
The San Bernardino Forest has been undergoing
oxidant stress as a result of long-range transport
from Los Angeles, 224 km (140 miles) away, since
the early 1 940's. Losses of ponderosa and Jeffrey
pines, the overstory vegetation, have increased
dramatically as pollutant levels have risen. Black
oak has also suffered oxidant injury. An alteration
in the composition of both plant and animal
populations has resulted because of the death of
the ponderosa and Jeffrey pines.
The interaction of pollutant and inversion layers
at the heated mountain slope results in the vertical
venting of oxidants over the mountain crest by up-
slope flow.
Oxidant concentrations ranging from 100 to 200
yug/m3 (0.05 to 0.10 ppm) at altitudes as high as
2432 m (8000 ft), approximately 1033 m (3400 ft)
above the mountain crest, were measured by
aircraft
The total oxidant concentrations have been
measured continuously from May through
-------
September since 1968 at the Rim Forest/Sky
Forest monitoring station. During each of thefirst?
years of monitoring, between June and
September, the total number of hours with
concentrations of 160^g/m3(0.08 ppm)or more of
ozone was never less than 1300 hours. The
number of hours wherein the total oxidant
concentration was 390/yg/m3(0.20 ppm) or higher
increased from fewer than 100 in 1969 to nearly
400 in 1974. It is not uncommon to observe
momentary oxidant peaks as high as 11 80 fjg/m3
(0.60 ppm). The duration of oxidant concentrations
exceeding 200 pg/m3 (0.10 ppm) was 9, 13, 9, and
8. hr/day going from the lower to the higher
altitude stations.
The most recent data firmly indicate that oxidant
concentrations will either increase annually or
oscillate around the mean of present high
concentrations in the foreseeable future.
The transport of the urban plume from the coast
northeastward to the mountains can be readily
demonstrated. Because of this transport, the
permanent vegetation constituting natural
ecosystems receives a large chronic exposure, and
the short-lived, higher-value vegetation con-
stituting the agroecosystem of the Los Angeles
coastal plajn is subjected to injurious doses in
intermittent, short-term fumigations. Each sit-
utation has measurable economic and aesthetic
effects, but on different time scales. The single-
species argricultural ecologic system (the
agroecosystem) has little resilience to pollutant
stress; losses are sometimes immediate and
occasionally catastrophic. The complex natural
ecosystem is initially more resistant to pollutant
stress, but the longer chronic exposures cause
disruption of both structure and function in the
system that may be irreversible.
Oxidant injury to the mixed conifer stands of the
San Bernardino Mountains beginning in the early
1940's, as indicated above, is well advanced. A
similar problem is developing in the forests of the
southern Sierra Nevada. Both areas show direct
and indirect effects on all subsystems of the forest
ecosystem—producers, consumers, and de-
composers.
In summary.
1. Ozone injury has limited the growth and
caused the death of ponderosa and Jeffrey
pines. An estimated 1,298,000 trees have
been affected. Decrease in cone production
has resulted in a decrease in reproduction.
Black oak has also suffered injury from
ozone.
2. Reduction in fruits and seeds that make up
the diet of most of the common small
mammals influences the populations of
these organisms.
3. Essential processes, such as recycling of
nutrients, may be disrupted, causing a
limitation in the growth of vegetation.
4. Death of the predominant vegetation has
caused an alteration in the species
composition and a change in the wildlife
habitat.
The San Bernardino Mountain study illustrates
the complexity of the problems caused by
environmental pollution. The changes that have
occurred in this mountain ecosystem as a result of
oxidant transport have already influenced the
importance and value of this natural resource to
the residents of southern California.67
The injury to the eastern white pine in the
Appalachian Mountains resulting from oxidant
transport from the urban northeast has begun a
similar sequential change that could degrade this
important recreational area. Total oxidant peaks as
high as 220 fjg/m3 (0.11 ppm) were recorded for
July 1 975. Concentrations exceeding 1 60 fjg/m3
(0.08 ppm) were measured in June 1976. These
episodes resulted in significant increases in
oxidant injury to three categories of eastern white
pine in the Blue Ridge Mountains.
Evaluating the contribution of functioning
natural ecosystems to human welfare is a complex
task and usually involves weighing both economic
and social values. However, because natural
ecosystems are life support systems, their value
should not be quantified in economic terms.
With the passage of time, man has destroyed
many of the naturally occurring ecosystems of
which he was a part and has replaced them with
simplified ecosystems wholly dependent on his
care and protection and requiring a large input of
energy.50
Man favors the simple, unstable and synthetic
ecosystems, because when they are extensively
managed and subsidized by the use of fossil fuels,
they are highly productive. An agricultural
ecosystem (agroecosystem) is an example of such
a simplified ecosystem. The effects of oxidants on
agroecosystems have been under study for more
than 20 years. The study of effects on natural
ecosystems is much more recent.
Plants grown in agroecosystems are largely
325
-------
annuals and can be replaced when they are
susceptible to pollutant stress. Natural ecosystems
remain in place year after year. Manmade
pollutants are undoing relationships developed
within these ecosystems over millions of years.
REFERENCES FOR CHAPTER 12
1 Aldaz, L Flux measurement of atmospheric ozone over
land and water. In Studies of Atmospheric Ozone
Section 111 AFCRL-69-0138, U S Air Force, Cambridge
Research Laboratories, Bedford, Mass , March 1969.
2. Auerbach, S.I ,R.L, Burgess, and R V O'Neill.The biome
programs' Evaluating an experiment Science 195 902-
904, 1977.
3 Berry, C. R Differences in concentrations of surface
oxidant between valley and mountain-top conditions m
the southern Appalachians J Atr Pollui Control Assoc
7 4/238-239, 1964
4 Berry, C R White pine emergence tipburn, a physi-
ogenic disturbance U.S For Serv Southeast For Exp.
Stn. Stn Pap (130).1-8, 1961
5 Berry, C. R , and L A Ripperton Ozone, a possible cause
of white pine emergence tipburn Phytopathology
53.552-557, 1963
6. Blumenthal, D. L , W H White, R. L Peace, and T B
Smith Determination of the Feasibility of the Long-range
Transport of OzoneorOzonePrecursors EPA-450/3-74-
061, U.S. Environmental Protection Agency, Research
Triangle Park, N.C., November 1974.
7 Botkin, D B , and R. S. Miller Complex ecosystems.
Models and predictions Am. Sci 62.448-453, 1974
8 Botkin, D B, and M J Sobel. The complexity of
ecosystem stability In, Ecosystem Analysis and
Prediction. S. A. Levin, ed Society for Industrial and
Applied Mathematics, Philadelphia, Pa , 1976 pp 144-
150
9. Boughey, A S Fundamental Ecology Intex Educational
Publishers, Scranton, Pa,, 1971 pp 1-50
10. Bryson, R. A , and W. M Wendland Climate effects of
atmospheric pollution In. Global Effects of
Environmental Pollution S F Singer, ed Sprmger-
Verlag, N.Y, 1970 pp 130-138
11. California Air Resources Board. Air Quality in the Tahoe
Basin, Summer 1 973 State of California, Air Resources
Board, Sacramento, Calif , March 1974.
12. Cobb, F W,Jr., andR W Stark. Decline and mortality of
smog-injured ponderosa pine J For. 68 147-149, 1970
13. Corn, M ,R.W, Dunlap, L. A. Goldmuntz, and L. H.Rogers.
Photochemical oxidants' Sources, sinks, and strategies.
J. Air Pollut. Control Assoc. 25:16-18, 1975.
14. Costonis, AC The Relationships of Ozone,
Lophoderm/um prnastn andPullularia to Needle Blight of
Eastern White Pine. Ph D. thesis, Cornell University,
Ithaca, N.Y , 1968.
15 Davis, D. D , and F A. Wood The relative susceptibility of
eighteen coniferous species to ozone Phytopathology
52:14-19, 1972
16 Dochmger, L S,F W Bender, F L Fox, and W W Heck
Chlorotic dwarf of eastern white pine caused by an ozone
and sulphur dioxide interaction, Nature (London)
225:476, 1970
17 Edmger,J G. Vertical distribution of photochemical smog
in Los Angeles basin Environ Sci Technol 7.247-252,
1973
18 Edmger, J. G , M H McCutchan, P R Miller, B C Ryan,
M J Schroeder, and J V. Behar Penetration and
duration of oxidant air pollution m the south coast air
basin of California J Air Pollut Control Assoc 22882-
886, 1972
19 Ellertsen, B W,C J Powell, and C L Massey Report on
study of diseased white pine in Tennessee Mitt Forst
Bundesversuchsanst. Wien 37 195-208, 1972
20 Estberg, G N , and B C Patton The relation between
sensitivity and persistence under small perturbations In.
Ecosystem Analysis and Prediction S A Levin, ed
Society for Industrial and Applied Mathematics,
Philadelphia, Pa , 1976 pp 151-154
21 Powells, H. A., and G H Schubert. Natural reproduction
in certain cutover pine-fir stands of California J For
43192-196, 1951
22 Garrett, A Compositional changes of ecosystem during
chronic gamma irradiation In Symposium on
Radioecology CONF-670503, D J Nelson and F C
Evans, eds US Atomic Energy Commission, Oak Ridge,
Tenn 1969 pp.99-109
23 Gloria, H R., G Bradburn.R F Reimsch,J.N Pitts, Jr,J.
V Behar, and L Zafonte Airborne survey of major air
basins in California J Air Pollut, Control Assoc 24 645-
652, 1974
24. Gosselmk.J G,E P Odum.andR M Pope TheValueof
the Tidal Marsh LSU-SG-74-03, Louisiana State
University, Center for Wetland Resources, Baton Rouge,
La , 1974
25. Hacskaylo, E The Torrey symposium on current aspects
of fungal development IV Dependence of mycorrhizal
fungi on hosts Bull TorreyBot Club 700 217-223, 1973.
26 Norton, J S Vegetation types of the San Bernardino
Mountains Forest Service Technical Paper No 44, U S
Department of Agriculture, 1960 29 pp.
27. Innis, G Stability, sensitivity, resilience, persistence.
What is of interest? In' Ecosystem Analysis and
Prediction. S A Levin, ed Society for Industrial and
Applied Mathematics, Philadelphia, Pa , 1976. pp 131-
140.
28 Kickert, R. N , P R Miller, 0 C Taylor, J R. McBnde, J.
Barbien, R Arkley, F Cobb, Jr, D Dahlsten, W. W
Wilcox, J Wenz, J. R Parmeter, Jr, R F. Luck, and M
White. Photochemical Air Pollutant Effects on Mixed
Conifer Forest Ecosystems A Progress Report. EPA-
600/3-77-058, U S. Environmental Protection Agency,
Corvalhs.Ore , May 1977
29. Levin, S. A., and R, T. Paine The role of disturbance in
models of community structure In.Ecosystem Analysis
and Prediction. S A Levin, ed. Society for Industrial and
Applied Mathematics, Philadephia, Pa., 1976 pp. 56-76
30 McCutchan, M. H , and M.J. Schroeder Classification of
meteorological patterns in southern California by
discriminant analysis J Appl Meteorol 72,571 -577,
1973
31 Millar, C. S Decomposition of coniferous leaf litter In
Biology of Plant Litter Decomposition Volume 1 C H
Dickinson and G. J. F Pugh, eds. Academic Press, Inc.,
N.Y., 1974, pp 105-128
32. Miller, P L Oxidant-mduced community change in a
mixed conifer forest. Adv, Chem, Ser. (122)101-117,
326
-------
1973
33 Miller, P R, and M J Elderman, eds Photochemical
Oxidant Air Pollutant Effects on a Mixed Conifer Forest
Ecosystem A Progress Report 1976 EPA-600/3-77-
104, US Environmental Protection Agency, Corvalhs,
Ore., September 1977
34 Miller, P R,M H McCutchan, and B C Ryan. Influence
of climate and topography on oxidant air pollution
concentrations that damage conifer forests in southern
California Mitt Forst Bundesversuchsanst Wien
97585-607, 1972
35. Miller, P R ,M H McCutchan, and H P Milligan Oxidant
air pollution in the central valley. Sierra Nevada foothills,
and Mineral King Valley of California Atmos Environ.
6 623-633, 1972.
36 Miller, P R , and A A Millecan Extent of oxidant air
pollution damage to some pines and other conifers in
California Plant Dis Rep 55555-559,1971
37. Munn, R E Biometeorological Methods Academic
Press, Inc. N Y 1970
38 National Research Council Ozone and other
Photochemical Oxidants National Academy of Sciences,
Washington, D C, 1977 pp 586-642
39 Odum, E P Summary In Ecological Effects of Nuclear
War G M Woodwell, ed BNL 917 (C-43), Brookhaven
National Laboratory, Upton, NY , 1965 pp 69-72
40 O'Neill, R V Modeling in the eastern deciduous forest
biome In Systems Analysis and Simulation Ecology
Volume III B C Patten, ed Academic Press, N Y , 1975
pp 49-72.
41 Overton, W S The ecosystem modeling approach in the
coniferous forest biome In Systems Analysis and
Simulation Ecology Volumelll BC Patten,ed Academic
Press, N Y , 1975 pp 117-138
42 Ovington, J D Dry-matter production by P/nus sylvestr/s
L Ann Bot (London) 21 287-314, 1957
43 Parmeter, J R , Jr , R V Bega, and T Neff A chlorotic
decline of ponderosa pine in southern California Plant
Dis Rep 46269-273, 1962
44 Patten, B C The relation between sensitivity and stability
In Ecosystem Analysis and Prediction S A Levin, ed
Society for Industrial and Applied Mathematics,
Philadelphia, Pa, 1976 pp 141-143
45 Pelz, E Untersuchungen uber die Fruktification
rauchgeschadigter Fichtenbestande Arch Forstwes.
12 1066-1077, 1963
46 Scheffer, T C, and G G Hedgecock Injury to
northwestern trees by sulfur dioxide from smelters U S
Dept Agric Tech Bull (1117)1-49,1955
47 Shugart, H H , Jr, T R Crow, and J M Hett Forest
succession models A rationale and methodology for
modeling forest succession over large regions For Sci
19 203-21 2, 1973
48 Sinclair, W A Polluted air Potent new selective force in
forests J For 67305-309,1969
49 Skelly, J M , C F Croghan, and E M Hayes Oxidant
levels in remote mountainous areas of southwestern
Virginia and their effects on native white pine (P/nus
strobus L ). In International Conference on
Photochemical Oxidant Pollution and Its Control
Proceedings Volume II B Dimitnades, ed EPA-600/3-
77-001 b, US Environmental Protection Agency,
Research Triangle Park, N.C , January 1977 pp 611-
620
50 Smith, R. L. Ecologyand Field Biology 2nd Ed Harperand
Row, N Y , 1 974 pp 30-85, 1 70
51 Smith, W H , Air pollution—effects on the structure and
function of the temperate forest ecosystem Environ
Pollut 6111-129, 1974
52 Stark, R W , and F W Cobb, Jr Smog injury, root
diseases, and bark beetle damage in ponderosa pine
Calif Agric 23 13-15, 1969
53 Stark, R W , P R Miller, F W Cobb, Jr , D L Wood, and
J R Parmeter, Jr Photochemical oxidant injury and bark
beetle (Coleoptera Scolytidae) infestation of ponderosa
pine I Incidence of bark beetle infestation in injured
trees Hilgardia 39 121 -1 26, 1968
54 Taylor, 0 C Oxidant Air Pollutant Effects on a Western
Coniferous Forest Ecosystem Task B Report Historical
Background and Proposed Systems Study of the San
Bernardino Mountain Area EPA-R3-73-043a, U S
Environmental Protection Agency, Washington, D C ,
1973
55 Taylor, 0 C Oxidant Air Pollutant Effects on a Western
Coniferous Forest Ecosystem Task C Study Site
Selection and On-site Collection of Background
Information EPA-R3-73-043B, U S Environmental
Protection Agency, Corvalhs, Ore , February 1973
56 Taylor, 0 C Oxidant Air Pollutant Effects on a Western
Coniferous Forest Ecosystem Task D University of
California, Statewide Air Pollution Research Center,
Riverside, Calif , July 1974
57 Taylor, 0 C , and P R Miller Modeling and oxidant air
pollutant impact on a forest ecosystem Calif AirEnviron
4 1-3, 1973
58 Thompson, C R Effects of air pollutants on lemons and
navel oranges Calif Agric 222-3, 1968
59 Thompson, C R , and G Kats Antioxidants reduce grape
yield reductions from photochemical smog Calif Agric
2412-13, 1970
60 Thompson, C. R., G Kats, and E Hensel Effects of
ambient levels of ozone on navel oranges Environ Sci
Technol 6:1014-1016,1972.
61 Thompson, C.R., and 0 C Taylor Effects of air pollutants
on growth, leaf drop, fruit drop, and yield of citrus trees
Environ Sci Technol 3934-940, 1969
62 Thompson, C R , 0 C Taylor, and B L Richards Effects
of photochemical smog on lemons and navel oranges
Calif Agric. 24 10-11, 1970
63 Thompson C R, 0 C Taylor, M D Thomas, and J 0 Ivie
Effects of air pollutants on apparent photosynthesis and
water use by citrus trees Environ Sci Technol / 644-
650, 1967
64 Treshow, M , and D Stewart Ozone sensitivity of plants
in natural communities Biol Conserv 5 209-21 4, 1973
65 US Forest Service Silvicultural systems for the major
forest types of the United States U S Department of
Agriculture, Agric Handb (445)1-114, 1973
66 Wert, S L A system for using remote sensing techniques
to detect and evaluate air pollution effects on forest
stands In. Proceedings of the 6th International
Symposium on Remote Sensing of the Environment
Universityof Michigan, Ann Arbor, Mich 1969 pp 1169-
1178
67 Westman, W E How much are nature's services worth?
Science 197 960-964, 1977
68 Wiegert, R G Simulation models of ecosystems Annu
Rev Ecol Syst 6311-338, 1975
327
-------
69 Williams, W T ,M Brady, and S C Willison Airpollution
damage to the forests of the Sierra Nevada Mountains of
California. J Air Pollut Control Assoc 27 230-234,
1977
70 Woodwell, G M Effects of ionizing radiation on
terrestrial ecosystems Science 138 572-577, 1962
71 Woodwell, G M Effects of pollution on the structure and
physiology of ecosystems Science 168 429-433, 1970
72 Woodwell, G M The ecological effects of radiation Sci
Am 208 40-49, 1963
328
-------
13. EFFECTS OF OZONE ON MATERIALS
INTRODUCTION
Ozone is a major causative factor in the overall
deterioration of many different types of organic
materials. In fact, certain polymeric materials are
more sensitive to ozone attack than are humans or
animals. The magnitude of ozone damage,
however, is difficult to assess because it is only one
of the many photochemical oxidants in the
atmosphere that contribute to the weathering of
materials. Nevertheless, researchers have shown
that ozone does accelerate the deterioriation of
several classes of material, including elastomers
(rubber), textile dyes and fibers, and certain types
of paints. Also, a National Academy of Sciences
report28 goes into detail to show that the damage
observed in most laboratory controlled
experiments is caused by ozone rather than some
other unmeasured oxidizing species.
That same document postulates that in urban
atmospheres, other oxidizing species may be more
damaging than ozone. It was assumed that the
relative reaction rates of various oxidizing species
with trans-2-bulene (a gas) are the same for
reactions with solid surfaces. This assumption did
not consider that at ambient levels, reactions with
surfaces will most likely be diffusion-controlled
rather than activation-energy-controlled. Thistype
of rate-controlling mechanism depends on the
concentration gradient at the surface, which is
roughly proportional to the ambient concentration.
Ozone is the most abundant of the active oxidants.
MECHANISMS OF OZONE ATTACK
Ozone is so chemically active that, when
concentrated, handling it becomes a problem
because of its- effect on ordinary materials. In
general, materials of organic origin are affected
deleteriously by concentrated ozone.9 Bailey1 has
thoroughly reviewed the literature from 1939 to
1957 for reactions of ozone with a host of organic
compounds, and he describes the reaction
mechanisms in detail. Although it is incorrect to
assume that all of these reactions would occur at
atmospheric concentrations of ozone, it is possible
that many of these reaction mechanisms would be
operant.
Many polymers are sensitive to atmospheric
concentrations of ozone,22 resulting m the
occurrence of chain-scissioning or cross-linking or
both. These reactions are related to the prevalence
of double bonds in the polymer structure. Chain-
scissioning results in a reduction in average
molecular weight and in decreased tensile
strength. Cross-linking increases the rigidity of
some polymers, thus increasing brittleness and
reducing elasticity.
Effects On Elastomers
OZONE CRACKING
The cracking of stressed natural rubber exposed
to outdoor environments has been a troublesome
materials problem for many years. At first, most
rubber technologists believed that sunlight was
the major cause of this phenomenon. However,
several investigators, mainly Williams39 and Van
Rossem et al.,38 concluded that cracking was due
to atmospheric ozone rather than sunlight, but the
relative importance of ozone was not immediately
recognized. Not until the 1940's did the research of
Norton,30 and especially that of Newton,29 and
Crabtree et al.,10'11 suggest that free radicals,
produced by the catalyzed photolysis of volatile
peroxides present in smog, may also contribute to
rubber cracking.
Cracking is related to the chemical structure of
rubber compounds and their elastic properties.
Vulnerable elastomers are those containing many
double bonds. An example of a chemical
mechanism may be written:28
329
-------
03 + C = C •
I I
H H
c— o — o"
H
A
o o
—c — c —
I I
H H
+ O = C -
(13-1)
The zwitterion and aldehydic group may combine
to form a more brittle oxidized structure:
C+—0 0" + 0=C - NC
I I /\ /
H H O O
H
(13-2)
The most vulnerable compounds are natural
rubber and synthetic polymers of styrene-
butadiene, polybutadiene, and polyisoprene.
These elastomers account for about 75 percent of
the annual production value in the United States.
Except for natural rubber, the major use for
these elastomers is in tires. Butyl, halogenated
butyl, polychloroprene, vinyl-modified nitrile-
butadiene, and carboxylated nitrile elastomers
have some ozone resistance but required special
formulation for optimum performance. Synthetic
elastomers with saturated chemical structures
such as silicones, ethylenepropylenes,
chlorosulfated polyethylenes, polyacrylates, and
fluorocarbons have inherent ozone resistance.
These latter special-application materials,
however, are relatively expensive and account for
only a small fraction of the market on a weight
basis.
Tensile stress is necessary to produce ozone
cracking in sensitive elastomers. The resulting
cracks develop at right angles to the direction of
stress. Elastomers in a relaxed state can be
exposed for long periods of time to relatively high
concentrations of ozone without developing
cracks. Crabtree and Malm12 reported that cracks
will develop in sensitive elastomers when strained
as little as 2 to 3 percent and exposed to
atmospheres containing only 20/vg/m3(0.01 ppm)
ozone. As strain is increased to 100 percent plus,
the number of cracks increases from a few deep
cuts to numerous shallow ones. At still higher
strain, small cracks are so numerous that the
surface simply presents a frosted appearance, and
damage from deep cracks is negligible.
Bradley and Haagen-Smit7 summarized the
factors that influence the action of ozone on
elastomers. These include (1) the nature of the
elastomeric compound, (2) degree of stress, (3)
concentration of ozone, (4) time of exposure, (5)
rate of the ozone contacting the elastomer, and (6)
temperature. Furthermore, cracking may be
divided into two main phases: crack formation and
crack growth. Scientists have proposed several
theories to explain ozone cracking, but the exact
mechanism is still not clear. Braden and Gent6
explained the strong dependence of crack growth
on nominal tensile stress in terms of the Griffith
theory of fracture mechanism. Fracture mechanics
alone only partially explain observed relationships
between number and size of cracks as a function of
strain. Crack growth rate is also ozone-
concentration dependent, which suggests that the
rate of diffusion of ozone to the root of a crack is the
rate-controlling factor.
More recently, Devries and Simonson13 have
applied the Griffith theory on a molecular scale.
The energy required for crack growth is extracted
from the elastic strain energy in the elastomer and
is expressed in terms of the energy to rupture
bonds and the number of bonds broken. They
indirectly determined the cracking rates
experimentally by measuring the number of
resulting free radicals with a highly sensitive
electron spin resonance (ESR) technique. Their
observations were consistent with both theory and
the results of previous damage studies.
ANTIOZONANTS
Fisher17 and Mueller and Stickney27 discussed
the use of antiozonant additives to protect
elastomers from ozone degradation. First
developed in 1955, antiozonants are organic
compounds, generally secondary aromatic amines
and phenols. They are incorporated into
elastomeric formulations during mixing and tend
to migrate to the surfaces of parts after
vulcanizing, where they form a film that chemically
reacts with atmospheric ozone before it contacts
the elastomer. These protective films are effective
even when elastomers are stretched or flexed.
Although widely used in such elastomeric
products as tires, conveyer belts, automotive parts,
and cable insulation, antiozonants have
330
-------
limitations. They are expensive and add to the cost
of finished products. For example, the addition of
1.5 percent antiozonants to automobile tires
increases the cost by as much as $0.50 per tire.
Furthermore, during product usage, antiozonants
provide only temporary protection against ozone
damage because they react with ozone and are
inactivated. Contact with oils, gasoline, and other
chemicals may also remove them by extraction.
The amount of antiozonant necessary for
effective control increases with the anticipated
concentration of atmospheric ozone. Although
undocumented, several sources have told EPA
scientists that tire manufacturers roughly double
the amount of antiozonant added to rubber used in
tires sold in California. Despite preventive
measures, however, ozone cracking of sensitive
elastomeric products still remains a major
problem,
DOSE-RESPONSE DATA
In reviewing dose-response data, one must keep
in mind the many factors that affect the rate of
attack by ozone on vulnerable elastomers.
Furthermore, methods of reporting effects data
vary from one research to another and thus further
complicate comparison. For example, an effect
may be expressed as time required to initiate either
microcracks or visible cracks, or it may be
expressed as crackdepth or crackgrowth rate.
Much of the documented research on ozone-
elastomer effects has been conducted at ozone
concentrations considerably higher than those
normally found in polluted environments. This has
been intentional in order to accelerate reaction
rates. However, some investigators have carried
out research at lower ozone concentrations.
Bradley and Haagen-Smit7 evaluated a natural
rubber formulation (Table 13-1) for susceptibility to
ozone cracking. Strips were strained
approximately 100 percent by bending. When
exposed in a 13~mm-diameter glass tube to 40,000
mg/m3 (2 percent or 20,000 ppm) ozone in air at a
flow rate of 1 5 l/min, these specimens cracked
instantaneously and broke completely within 1
sec; however, when they were exposed to various
lower concentrations of ozone, different time
periods were required for cracks to develop (Table
1 3-2). (Ozone levels were measured by the neutral
Kl method.) Based on these data, the initiation of
cracks is controlled by the dose of ozone
(concentration * time). The mean and estimated
standard deviation of the dose for this particular
experiment is a product of 1.32 ± 0.03 ppm *
minutes.
TABLE 13-1. FORMULATION OF HIGHLY
OZONE— SENSITIVE NATURAL RUBBER8
Ingredients
Rubber
Tire reclaim
SRF Black
Stenc acid
Pine tar
Zinc oxide
Mercaptobenzothiazole
Diphenylguamdme
Sulfur
Parts by weight
100
125
33
1.5
84
4 7
08
0 1
5
% by weight
35.91
44.88
11.85
054
3.02
1,69
0.29
0.03
1.79
"Cured 40 mm at 310,050 N/m> (45 psi) steam pressure
TABLE 13-2. EFFECT OF OZONE ON NATURAL RUBBER"
Ozone concentration
ppm
Time -to first Sign of crack at
4* magnification, mm
40
500
900
0.02
025
045
65
5
3
"See Table 13-1 for formulation, strained 100 percent
Meyer and Sommer25 exposed thin
polybutadiene specimens, under constant load, to
room air for which the average concentrations of
ozone had been measured by the neutral Kl
method. Specimens exposed in the summer
months to average ozone concentrations of about
96 ^g/m3 (0.048 ppm) failed by breaking into two
separate parts after 150 to 250 hr. In the fall, at
average ozone concentrations of 84 jug/m3 (0,042
ppm), specimens failed between 400 and 500 hr. In
the winter, at average ozone concentrations of 48
/jg/m3 (0.024 ppm), failures occurred between 500
and 700 hr. These data show the strong
dependence of. cracking rate on average
concentrations of ozone. It takes much longer for
complete failures to occur than it does to initiate
cracks. The mean and estimated standard
deviation of the dose required to ca use 100 percent
complete failures in these specimens were 996 +
270 ppm * minutes.
Edwards and Storey16 determined the effects of
ozone on elastomeric formulations (Table 13-3)
containing various amounts of antiozonants and
either a styrene-butadiene rubber (SBR) polymer
that had been polymerized at elevated
temperatures ("hot" SBR) or one polymerized at
room temperaluresJ^co!d"SBR).The basic poly-
mers were used in tire sidewall formulations. The
investigators exposed test specimens under strain
(100 percent) to ozone concentrations (measured
by the neutral Kl method) of 490 ± 100 pg/m3
331
-------
(0.25 ± 0.05 ppm) at 49°C (120°F). Table 13-4
shows the calculated rate of cracking for the
various formulations. The results provided
evidence that the hot SBR polymer is inherently
superior in resistance to attack by ozone, and
increasing amounts of antiozonants produced
significant reductions in the rate of cracking and
thus increased the dose necessary to initiate
visible cracks. The function is exponential:
Dose (ppm x minutes) = a ebz (1 3-3)
where z is percent antiozonant. Regressions on
these data show that b is essentially the same for
the two formulations (1.5) and that the a values are
17.3 and 9.8, respectively, for the hot and cold
SBR. The dose would have to be three orders of
magnitude greater for cracks to grow to 2.54 mm.
To prevent failure of 5-mm-thick sidewalls
exposed under 100 percent strain at 49°C (1 20°F)
to 100 /jQ/m3 ozone for 3 years (781,840 ppm x
minutes), an estimated 0.56 percent and 0.94
percent of antiozonant in the hot and cold SBR
would be needed, respectively.
TABLE 13-3. TIRE SIDEWALL FORMULATION16
Ingredients
Polymer (hot or cold SBR")
Ocosot 2 XH
FEF black
SRF black
Zinc oxide
Steric acid
Antiozonant (Santoflex AW)
Crystex
Parts by weight
100
10
30
10
3
2
Variable
2
aSBR is sty re ne-butadiene rubber polymerized either at higher room temperature
TABLE 13-4. EFFECTS OF OZONE ON SIDEWALL
FORMULATIONS CONTAINING VARIOUS
ANTIOZONANT CONCENTRATIONS
Polymer
Hot SBR
Cold SBR
Antiozonant
concentration.
0
032
063
1 25
0
032
063
1 25
Rate of cracking
ym/hr
234
1 75
089
033
401
2 16
1 45
061
in /ht
092
069
035
013
1 58
085
057
024
Time to first
sign of crack,
mma
65
87
170
460
38
71
105
250
a Added to enable comparison with data i
sign of crack was assumed to be 2 54 /jm
depth, visible at 4* magnification
paper by Edwards et al 1b
Table 13-2 First
x 10 " m ) crack
Data are not found in the
Hofmann and Miller21 found that the behavior of
rubber exposed to ozone under laboratory
conditions correlated well with the service
behavior of tires in localities where atmospheric
ozone concentrations were high. The relative
susceptibility of different formulations of white
sidewall rubber remained the same, whether
exposed under laboratory conditions to as much as
1000/Kj/m3 (0.5 ppm) ozone, or in the ambient air
of the Los Angeles areas. The rate of cracking is
thus a function of ozone concentration.
In a factorial-design, controlled-environment
experiment, Haynie et al.19 exposed several
classes of materials to realistic levels of pollutants
and climatic factors. White sidewall specimens
from a top-quality steel-belted radial tire were
exposed (strained at 10 and 20 percent) for up to
1000 hr. The level of ozone was a statistically
significant factor in the rate of cracking of the
white sidewall rubber. The average results with
respect to strain and ozone level are given in Table
13-5.
TABLE 13-5. CRACKING RATES OF WHITE SIDEWALL
TIRE SPECIMENS19
Ozone cone ,
^g/m3 (ppm)
1 60 (0 08)
1000(05)
Strain, %
10
20
10
20
Mean cracking rate ± S D ,e
mm/yr
1 0 36 % 7 76
1 1 70 % 7 22
1 9 80 % 9 64
24.09 % 6 24
aS D = estimated standard deviation oi the mean
Cracking rates are not directly proportional to
ozone concentrations for these two levels. A
coefficient for estimating crack depth based on the
lower concentration is 34.5 ±48 mm/ppm-year.
This indicates it would take 3 years to crack a 5-mm
sidewall exposed to 100 //g/m3 (0.05 ppm). This
does not include the time requiredforthe resulting
exposed tire cord to fail.
For this particular premium tire, one would
expect the tread to wear out before sidewall failure
from ozone damage occurred. However, the casing
would have questionable value for retreading.
ECONOMICS
As the tread wear on passenger car tires
improves, more or better antiozonants will have to
be added to sidewall rubber formulations to
prevent cracking from becoming the limiting factor
in tire life. Thus part of the cost of premium tires
will be the result of increased protection against
atmospheric ozone.
Mueller and Stickney,27 on the basis of a rubber
industry survey, assessed the economic impact of
332
-------
air pollution on elastomeric products in the United
States. They estimated the loss at the consumer
level to be $500 million annually. About $170
million of this total was associated with various
means of protection, including antiozonants and
special polymers. The remaining $330 million was
associated with premature failure of rubber
products. (See Reference 28 for a more complete
description of technique and costs.)The respective
annual per capita costs in 1970 are $0.85 and
$1.65. If more were spent on preventive measures,
there would be less cost for failures. The
relationship between dose to initiate cracks and
antiozonant level suggests that for a fixed dose in
1970, failure costs (Y) would be related to
preventive costs (X) by the relationship
Y = 5.9e"15X (13-4)
The total cost is the sum of X and Y. By
differentiating with respect to (X) and setting it
equal to zero, the optimum total cost in 1970 is
calculated to be $2.1 2. This value is $0.38 less per
person than was estimated.
By reducing ambient levels of ozone, it is
possible to establish some new optimization of
preventive and failure costs. Assuming that a
population-weighted national annual average for
ozone was 65 pg/m3 in 1970, the optimum per
capita cost as a function of average ozone
concentration in the ambient is:
Cost = 0.66 (In O3 1)
(13-5)
when 03 concentration is fjg/m3 annual average.
Because it is unlikely that consumers will
optimize present prices with anticipated use life,
these costs serve as a minimum estimate. A cost
function with a 25-percent error range is
Cost = $0.88(1 + 0.25)x(|nO3 1)0 3-6)
This relationship suggests that if the national
annual average concentration of ozone were
lowered from 65/Mg/m3to50/ug/m3therecouldbe
a savings in elastomer damage and damage
prevention costs of from 37 to 61 million dollars.
Comparably, lowering the annual average from
100 fjg/m3 to 50 fjg/m3 for 6.9 million people in
the Los Angeles Standard Metropolitan Statistical
Area (SMSA) could save from 2.0 to 3.3 million
dollars.
Effects on Textiles
Both fibers and dyes can be damaged by ozone;
however the fading of dyes is a more significant
factor than the loss of fiber strength.
DYE FADING
Although atmospheric nitrogen oxides were
identified during the early 1 900's as an important
cause of color fading in certain textile dyes, mainly
disperse dyes on acetate fabrics, it was not until
the mid-1950's that the research of Salvin and
Walker32 showed the effects of atmospheric ozone.
This discovery came about as a result of field
testing acetate fabrics containing newly
synthesized blue disperse dyes. Previous
laboratory exposures of these dyes had shown
them to be highly resistant to fading by nitrogen
oxides. Field testing was carried out to determine ifv
the performance of these new dyes was as good in
actual use as in laboratory tests.
The researchers exposed fabrics dyed with one
of the new blue dyes (as a color component) in
homes located in areas known to have either low or
relatively high levels of nitrogen oxides. At the low-
nitrogen-oxides sites, they found that fading for
many of the fabrics equaled, and in some cases
exceeded, the fading observed at the higher-
nitrogen-oxides site. Much of the observed fading
was characterized by a bleached, washed-out
appearance ratherthanthefamiliar reddening that
nitrogen oxides produce in most sensitive dyes.
This anomalous behavior suggested that some
other oxidizing agent was responsible and was
present at all exposure locations.
To describe this fading phenomenon, the
investigators coined a new term, "0-fading."
Follow-up laboratory investigations and general
observations showed atmospheric ozone to be
responsible. Ozone concentrations of 200 fjg/m3
(0.10 ppm) produced marked fading in most of the
blue disperse dyes and in some of the reds and
yellows.
The discovery of the O-fading was useful in
explaining much of the anomalous fading of
certain dyed fabrics observed during subsequent
lightfastness testing and service exposure trials
reported by Schmitt.34'35 Other types of dyes
besides the disperse dyes showed abnormal
fading. Furthermore, the results of the service
trials emphasized the importance of relative
humidity, suggesting that the increased moisture
content of fibers such as cotton promoted and
accelerated the absorption and reaction of
pollutants with vulnerable dyes. They also alerted
the textile industry to the possibility of potential
consumer complaints. Such complaints began to
appear in the early 1 960's and concerned mainly
333
-------
two distinctly different textile materials—
polyester-cotton/permanent press fabrics, and
nylon carpet,
Salvm31 reported that the fading of polyester-
cotton/permanent press fabrics was first noted on
the folds and edges of slacks stored in warehouses
or on the stock shelves of retail outlets. Some
garments faded after storage in warehouses for as
few as 10 days. Incidents occurred in various
locations including California, Texas, and
Tennessee. Because of the volume of garments
involved, some producers suffered heavy
economic losses. Fading was marked by a loss in
blue along with a slight increase in red, suggesting
that ozone and, to a lesser degree, nitrogen oxides
may have been the active fading agents, since
sunlight was obviously not a factor. This
conclusion did not seem reasonable, however,
because the cotton was dyed with vat dyes and the
polyester with disperse dyes, and both dye-fiber
systems are generally resistant to fading by
common air pollutants. Furthermore, fading
occurred only on fabrics that had been made into
finished products and then cured to set the
permanent press resin Fading had not been
observed during the temporary storage of dyed
fabrics either before or after treatment with
permanent press resins (before curing).
However, after a thorough investigation,
including extensive laboratory tests, researchers
found ozone to be the major fading agent, with
nitrogen oxides also capable of causing fading, but
to a lesser extent. The fading mechanism, which is
unique and complex, takes place as a result of the
curing operation and involves the disperse dyes on
the polyester fibers rather than the vat dyes on the
cotton. During curing, some disperse dyes partially
migrate to the permanent press finish, which is a
combination of reactant resin, catalysts, softeners,
and nonionic wetting agents. The disperse dyes
migrate to the solubilizing agents (nonionic
surfactants and softeners) in the finish and are
then in a medium in which fading by air
contaminants can easily occur. Softeners are an
especially good medium for absorbing gases
The choice of catalyst used in the finish plays an
important role, as the migration of disperse dyes
increases significantly when magnesium chloride
is used as a catalyst rather than zinc nitrate.
Magnesium chloride is capable of forming
complexes with certain anthraqumone disperse
dyes (blues and reds), and these complexes are
soluble in the resin finish Therefore, as Dorset15
pointed out, to eliminate this fading problem,
textile processors must carefully select materials
that make up the permanent press finish and/or
replace vulnerable dyes with those that resist
migration.
According toSalvin,31 the fading of nylon carpets
originated mainly in the warm humid areas from
Texas to Florida and, as a result, became known as
"Gulf Coast fading." In one case, nylon carpeting
in a Texas apartment complex showed visible
fading only 30 days after installation. Afewsimilar
incidents were also noted along the east coast and
in the Los Angeles area. Fading occurred on
carpets manufactured from both nylon 77 and
nylon 6 fibers. Disperse dyes were used because
many possess easy leveling properties necessary
to avoid dye streaks. Fading took place largely on
those carpets dyed with Disperse Blue 3 as one of
the color components. Avocado, a tertiary-dyed,
dull green shade, was a particularly sensitive color.
The fading of this color was characterized primarily
by loss in blue, which caused the green color to
turn gradually to a dull orange shade.
Investigators eliminated sunlight as a primary
cause of fading, since the phenomenon occurred in
rooms where light intensity was low. Exposure of
carpet samples to standard test methods for ozone
also failed to duplicate the color change. Since
fading complaints occurred in humid
environments, and since previous lightfastness
testing and service trials had established that
conditions of high relative humidity promoted
fading of certain dyes by ozone, the investigators
next exposed carpet samples to ozone in the
presence of high relative humidity (85 to 90
percent). Under these conditions, pronounced
fading took place on those samples containing
Disperse Blue 3, and the fading was similar to the
color changes observed in homes along the Gulf
Coast. Later experiments established that the
relative humidity must be somewhat above 65
percent for pronounced ozone fading to occur. The
fading problem may be prevented by using dyes
that are more ozone resistant and by using nylon
fibers that have been modified (by dry-heat
texturing) to decrease the accessibility and
diffusion rate of ozone.
Complaints were also received from consumers
about faded drapery and upholstery fabrics.
Salvm31 reported that the combination of ozone
and high humidity causes pronounced fading of
certain dyes on cotton and rayon fabrics.
334
-------
Beloin3'4 investigated the effects of air pollution
on various dyed textiles by conducting both field
and controlled-environment laboratory studies.
The field study consisted of exposing a wide range
of dyed fabric in light-tight cabinets located at four
urban and four rural (control) sites that were
nearby and had similar characteristics except for
air pollution. These sites represented a cross
section of various types of pollution (urban sites)
and climates. The study was carried out over a 2-
year period, with eight consecutive 3-month
seasonal exposures.
Color change data, along with air pollution and
weather measurements, were statistically
analyzed in an effort to identify the factors that
caused fading. Beloin found that about two-thirds
of the fabrics studied showed appreciable fading.
Most of these fabrics faded significantly more at
urban sites than at the rural sites, andthe amount
of fading varied among metropolitan areas and
among seasons. Air pollution was assumed to
account for most of the environmental differences
between the urban and the corresponding rural
control sites. Analysis suggested that the
pollutants measured (NO?, ozone, and SO?)
appeared to be those most responsible for fading.
Generally it was impossible to separate the effects
of individual pollutants because they were
confounded with the effects of other pollutants.
The controlled-environment laboratory study
was designed to assess the effects of common air
pollutants, temperature, and relative humidity on
the colorfastness of 30 dyed fabrics selected from
those exposed during the field study. Fabric
samples were exposed to two concentrations of
ozone: 100 fig/m3 (0.05 ppm) and 1000 #g/m3
(0.50 ppm). Ozone concentrations were
continuously measured by coulometric (Mast)
analyzers. As might be expected, under similar
exposure conditions, high ozone levels produced
significant fading in more fabric samples than did
low levels. Low levels, nevertheless, produced
visible fading in about one-third of the sensitive
fabrics—an important finding since the low levels
were similar to those frequently found in
metropolitan areas. The study also demonstrated
that high relative humidity (90 percent) and, to a
lesser extent, high temperature (32°C or 90°F) are
significant factors in promoting and accelerating
ozone-induced fading, thus confirming what
investigators observed during previous service
trials, The amount of ozone fading as a function of
time was dependent on such factors as color and
type of dye, fiber substrate, and environmental
conditions.
Haynie et a!,19 and Upham et a!,37 exposed three
moderate-to-high-volume-usage drapery fabrics
to combinations of controlled-environment
conditions to determine both direct and synergistic
effects of gaseous pollutants. There were no
statistically significant direct or synergistic effects
of ozone on the fading of any of the dyes after
exposures of 1000 hr to ozone up to 1000 fjg/m3.
Relative humidity was a significant fading factor
for all three fabrics, and nitrogen dioxide faded one
of the fabrics. Fading as a function of time was
consistent with the theoretical relationship:
AE = AEm{1-e^al) (13-7)
where AEm represents the maximum fade or
complete dye destruction, t is time, and a is a
constant containing the effects of environmental
variables.
An unacceptable level of fading is subjective and
varies considerably with individuals. A percentage
of useful life lost depends only on the ozone level,
because other factors, including subjective
unacceptable Levels of fading, cancel out. This
relationship is:
Percentage life lost =
100 (13-8)
where ao is the exponential coefficient for clean
air, and ai is the value with some level of pollution.
These functions should be applicable to ozone
fading as well as NOz fading.
Haylock and Rush18 studied ozone fading of
anthraquinone dyes on nylon fibers in controlled
environments. Olive I and Olive II dyes on nylon 6
carpets yarn was exposed to 400 fjg/m3 and 1 800
vg/m3 ozone at 40°C (104°F) and 90 percent
relative humidity. The same materials were
exposed to 400 ^g/m3 ozone at 40°C(1O4°F) and
70 percent and 80 percent relative humidity. C.I.
Disperse Blue 3 and C.I. DisperseBlue7on nylon 6
were exposed to 400 pg/m3, 1000 j/g/m3, and
2000 ^g/rn3 ozone at 40°C(104°F) and 90 percent
relative humidity. They were also exposed at 50°C
(122DF) and 30 percent relative humidity to 400
^/g/m3 ozone.
The fading curves were highly consistent with
the theoretical relationship of equation 13-7.
Regression equations on the Haylock and Rush18
data can account for 99 percent of the variability.
With these data, coefficients were calculated for
three of the dyes that were significantly affected by
335
-------
different ozone levels. Linear regressions of these
coefficients with respect to ozone were then
calculated. For the exposures at 40°C (104°F) and
90 percent relative humidity, the results are as
follows:
Olive I:
a-0.381 +3.65 x 10"5 03 (13-9)
Olive II:
a = 0.00107 + 1.18 x 10~503 (13-10)
C.I. Disperse Blue 7:
a = 0.021 2 + 5.35 x 10~603 (13-11)
where a is reciprocal hours, and 03 is in /7g/m3.
Assuming that the relative effect of ozone with
respect to other fading mechanisms remains the
same at normal temperatures and relative
humidities, these relationships convert to the
following.
Olive I:
Percent life lost =
Olive II:
Percent life lost -
0.096 03
1 +0.00096 03 (13-12)
1.1 03
1 +0.011 03 (13-13)
C.I. Disperse Blue 7:
Percent life lost=
°-025
1 +0.00025 03 (13-14)
These functions curve only slightly when ozone is
below 100 fjg/m3. Thus the costs of early
replacements can be assumed to be directly
proportional to ozone levels.
FIBERS
Because cellulose fibers are vulnerable to
oxidation, atmospheric ozone is a potential cause
of degradation. With this in mind, Bogaty et al.5
carried out experiments to study the possible role
of ozone in the deterioration of cotton text lies. They
exposed samples of duck and print cloth to air
containing between 40 and 1 20 £
-------
humidity by placing a pan of water on the chamber
floor. At 3-day intervals, the cotton samples were
removed from the chamber. Half of them were
machine-washed, and the other half were soaked
in water for 1 min. All samples were passed
through a hand wringer to remove excess water
before they were returned to the chamber for
further exposure. Control samples were kept in a
light-tight box maintained at 21 °C (70°F) and 65
percent relative humidity, and they were given the
same washing and soaking treatment.
After an exposure period of 60 days, which
included 20 washing and soaking treatments, the
change in strength of the control fabrics was not
significant. By comparison, the fabrics exposed to
ozone changed significantly; the loss in strength
for the washed fabrics was 18 percent, and for the
soaked fabrics, 9 percent. The investigators
attributed these losses to ozone.
Since Morris26 found no degradation under
exposure conditions similar to those used by Kerr,
the washing and soaking treatment would appear
to affect in some way the sensitivity of the fabrics
to ozone. Nevertheless, when one attempts to
equate Kerr's findings with actual conditions
encountered by ^consumers, the degradation
seems minimal in view of the fact that average
levels of ozone under field conditions are less than
10 percent of the levels used in the laboratory
exposure.
In laboratory studies, Zeronian et al.40
simultaneously exposed modacrylic (Dynel),
acrylic (Orion), nylon 66, and polyester (Dacron)
fabrics to artificial sunlight (xenon arc) and
charcoal-filtered clean air contaminated with 400
/yg/m3 (0.2 ppm) ozone (analytical method not
given) at 48°C (118°F) and 39 percent relative
humidity. During exposure, the fabric samples
were sprayed with water for 18 min every 2 hr.
Ozone damage was measured by comparing these
samples with fabrics exposed to the same
environmental conditions but without ozone. After
exposure for 7 days, the investigators found that
ozone did not affect the modacrylic and polyester
fibers; but it did seem to affect the acrylic and nylon
fibers slightly.
ECONOMICS
Barrett and Waddell2 used data from an
unpublished contract report by Salvin to estimate
the national cost of ozone fading at approximately
$80 million/year. These costs were associated
with the fading of acetates/triacetates, nylon
carpets,and permanent press fabrics. Salvin
produced his estimates from costs of preventive
measures as well as loss-of-use life based on an
industry survey. Preventive measures included
more expensive dyes and manufacturing
procedures, research, and quality control testing.
The estimated totals for preventive and failure
costs were $43 and $37 million, respectively. On a
per capita basis, these respective annual costs are
$0.22 and $0.18.
Because there was a lack of dose-response data,
Barrett and WaddelP did not attempt to establish
an economic damage function based on these
estimates. One can be postulated assuming an
analogy with elastomer economic damage costs.
For elastomers, an estimate of the amount of
damage saved by preventive measures is 5 times
the cost of those measures An analogous
relationship between damage costs (Y) and
preventive cost (X) is:
Y - 1.28e-
(13-15)
It follows, using the same procedures as with
elastomers, that a reasonable economic damage
function with 50 percent error is:
Cost = $0.22 (1 ± 0.5) (1 n 03 - 0.74) (1 3-1 6)
over the range of ozone concentrations greater
than 5.7 ^g/m3.
This relationship suggests that from $6.2 to
$18.6 million could be saved annually by lowering
the national annual average ozone concentration
from 65 to 50 //g/m3.
Effects on Paints
DAMAGE FUNCTION
Campbell et al.8 have conducted laboratory
research and field studies indicating that ozone
may damage paint. The technique used to assess
damage, erosion measurements (weight loss
converted to thickness loss), proved to be the most
meaningful. Erosion, or the gradual weathering of
the surface of paint film, is the normal mechanism
of failure that limits the service life of well
formulated and properly applied exterior coatings.
The researchers selected five exterior coatings
from four commercially important paint types:
house paints (oil and acrylic latex), coil coating*
finishes (urea-alkyd), automotive refmishes
(nitrocellulose/acrylic), and industrial
maintenance coatings (alkyd).
"Coil coating is a factory-applied coating to a cod of sheet metal
337
-------
Panels coated with the different exterior paints
were exposed to controlled-environment
conditions, including clean air (control), clean air
containing 200 /yg/m3 (0.1 ppm) ozone, and clean
air containing 2000/jg/m3(1.0 ppm) ozone. Ozone
concentrations were measured using the buffered
Kl method. The exposure chamber operated on a 2
hr light-dew cycle consisting of 1 hr of xenon light,
70 percent relative humidity, and a black panel
temperature of 66°C (151°F), followed by 1 hr of
darkness at 100 percent relative humidity of 49°C
(120°F), during which time moisture condensed on
the coated panels. One group of panels was
exposed to simulated sunlight while a like group of
panels was shaded during the exposures.
Erosion measurements were made after
exposure periods of 0, 400, 700, and 1000 hours.
Erosion rates for each paint type and exposure
condition were then calculated from the erosion
results. Zero-hour results, however, were
excluded from these calculations because erosion
rates during the 0- to 400-hr period were greater
than during the remaining exposure (400 to 1000
hr). The researchers concluded that this was
primarily caused by removal of water soluble
materials and that materials subsequently eroded
consisted mainly of binder and pigment.
Generally, exposures to 2000 /jg/m3 (1 ppm)
ozone produced statistically significant increases
in erosion rates compared to clean air (zero
pollution) conditions. Erosion rate increases,
however, varied considerably among paint types.
Oil-based house paint experienced the largest
erosion rate increase; industrial maintenance
paint, a moderate increase; and latex and coil
coatings and automotive refmishes the smallest
increases. As would be expected, unshaded panels
eroded more (degradation by sunlight)than shaded
panels. Because of considerable data variability,
exposures to 200 /yg/m3 (0.1 ppm) ozone generally
did not produce statistically significant erosion rate
increases over clean air exposures.
Table 13-6 gives regression coefficients for the
effects of ozone bas^d on the data of Campbell et
al.8 Because the rate of ozone damage is controlled
by the rate of ozone transport to the paint rather
than activation energy for reaction, these
coefficients should be applicable to lower ambient
temperatures.
All of the coefficients are positive, and half are
statistically significant. The largest coefficients
were determined on data with the greatest
variance. Consequently, they are not statistically
significant, although the effects they represent
could be real.
Field exposures were conducted in Leeds,
N.Dak., Valparaiso, Ind., Chicago, III., and Los
Angeles, Calif. These locations, respectively,
represent clean, moderate-S02, high-S02, and
high-ozone environments.
Panels of the five different coatings were
mounted at an angle of 85° from the horizontal to
represent normal exposure conditions of house
siding. Half of the panels were mounted facing
south, and half facing north. Panels were
evaluated after 0, 3, 7, and 14 months of exposure.
As with the laboratory data, the zero-month results
were not used in calculating erosion rates.
TABLE 13-6. PAINT EROSION RATE COEFFICIENTS
FOR THE EFFECTS OF OZONE IN
LABORATORY-CONTROLLED ENVIRONMENTS36
Coating
Mean and estimated S 0 of erosion rate
Coefficients for ozone, ^tm/yr per ^g/m3
Shaded Unshaded
000216+000043" 000325 ± 0 00026a
0.00099 ± 0 00074 0.00214+0.00007"
Automotive
Latex
Industrial
maintenance 0.01073 ±001473 001098 + 0.01774
Coil coating 0 00129 ± 0 00207 0.00752+000176
Oil house paint 0.00812+0.00378° 0.02726 + 0.01894
"Coefficient significantly greater than zero at 0 95 probability level
The erosion rates for Leeds and Los Angeles are
presented in Table 13-7. With the exception of the
north-facing automotive coating, all of the erosion
rates are significantly higher in Los Angeles than
in Leeds. The latex, coil coating, and oil house paint
contain fillers that are susceptible to SO2 attack.
The automotive and industrial maintenance
coatings do not contain fillers. Thus, many of the
differences in erosion rates between Leeds and
Los Angeles for the filler-containing paints could
be caused by differences in S02 levels. The
observed differences in the automobile and
industrial maintenance coatings could be
attributed primarily to higher ozone levels in Los
Angeles. For example, using the laboratory-
obtained coefficients, ozone differences of 85 to
111 fjg/m3 can account for the erosion rate
difference between Leeds and Los Angeles. Ozone
levels were not measured, but an average
difference of 40 to 60 /jg/m3 ozone is consistent
with reported air pollution data. Also, average
differences in temperature and other pollutants
would be expected to contribute to the higher
erosion rates in Los Angeles.
338
-------
TABLE 13-7. PAINT EROSION RATES AT FIELD EXPOSURE SITES
Mean and estimated S D
Panel facing north
Coaung
Automotive
Latex
Industrial maint
Coil coating
Oil house paint
Leeds
0 305 ± 0.076
0 305 ± 0.046
0.914 ± 0 198
0610 ±0.091
0.305 ±0.1 37
Los Angeles
0.305 ± 0 046
0914±0137
1 829 ±0213
3.088 + 0351
3.962 ± 0945
of erosion rates, /jm/year
Pane! facing south
Leeds
0.30510.107
0.305 ± 0.076
1 219 ±0.366
0610 ± 0 122
0610 + 0.213
Los Angeles
0 610 +0031
0914 ±0.351
2 438 ± 0.320
3658 +0381
4.877 ± 0.793
In laboratory-controlled environments, Spence
et al.36 studied the direct and synergistic effects of
gaseous pollutants on commonly used exterior
paints. Useful data were obtained on a vinyl coil
coating, an acrylic coil coating, and an oil-base
house paint. The oil-base house paint contained a
calcium carbonate filler that was strongly attacked
by SO2 and moisture. The magnitude and
variability in this effect masked any possible
observed effect of ozone. Statistically significant
effects of ozone were observed for the vinyl coil
coating and the acrylic coil coating. There was a
positive interaction effect between ozone and
relative humidity on the vinyl coil coating, and a
positive direct ozone effect on the erosion rate of
the acrylic coil coating. Both of these coatings are
very durable, which is why they were selected for
factory application to aluminum siding material.
Coatings as thin as20/ym should last more than 20
years.
The vinyl coil coating was not affected by ozone
at an input relative humidity of 50 percent, but it
was at 90 percent. At the high relative humidity,
the multiple regression coefficient for ozone was
0.00166 ^m/year x (/yg/m3). The average erosion
rate in clean air was 1.3 /ym/year.
A linear regression for the acrylic coil coating
data gives.
Erosion rate = 0.159 + 0.000714 O3 (13-17)
where erosion rate is /ym/year and 03 is /yg/m3.
Although the ozone effect on this coating is
statistically significant, it has no practical
significance because the erosion rate is so slow. At
an average annual Os level of 100 /yg/m3, this
regression predicts that a 20-/ym-thick coating
would last over 80 years.
ECONOMICS
Most of the laboratory-observed effects of ozone
on paint have little practical significance because
their relative contributions to reducing useful life
at ambient concentrations are so small. Ozone
damage to some industrial maintenance paints
and vinyl coil coatings could possibly produce
economic losses.
Estimates of damage functions for the two
affected coatings can be obtained by extrapolating
the laboratory data to expected ambient
conditions. The result for the industrial
maintenance paint is:
Erosion rate = 1 + 0.01 03 ± 0.3 (13-18)
where erosion rate is mm/year, Os is/yg/m3, and
±0.3 is the estimated standard deviation.
Similarly, the result for the vinyl coil coating is:
Erosion rate = 1 + 0.001 03 ± 0.4 (13-1 9)
If erosion rate controls paint life, a percent life
lost as a function of ozone concentration can be
calculated. With estimates of the existing amounts
of exposed painted surfaces and repainting costs,
economic loss functions can be postulated.
Paint industry statistics are reported annually by
Charles H. Kline and Co., Inc.24 Estimated
shipments of all industrial maintenance coatings
in 1974 were 55 million gal. Alkyds that do not
contain fillers probably account for 25 percent of
those sales. Assuming a coverage of 300 ftVgal at
$0.40/ft2 and half exterior exposure, the annual
per capita repainting costs are estimated at $3.83.
Considerable labor is involved in surface
preparation and application of industrial
maintenance paints.14
Coil-coated aluminum and glavanized steel have
experienced a growth rate in shipments of
approximately 13 percent from 1962 to 1974.24
Seventy percent of these shipments are used in
exterior applications. It is estimated from these
production figures that 290 ftVperson of all coil
coatings are exposed to exterior ambient
conditions. If 40 percent are vinyl coatings and
repainting costs are $0.25/ft2, the per capita
repainting costs for exposed susceptible coatings
are $29. At a 23-year life expectancy, the annual
cost is $1.26.
The fraction of paint life lost as a function of
ozone concentration times the annual per capita
339
-------
repainting costs of affected paints gives an annual
economic loss function. For industrial
maintenance paints:
Annual per capita costs = 0-0383 03 (13-20)
1 +0.01 03
And for the vinyl coatings:
Annual per capita costs = 0.00126 O3 (13-21)
1 + 0.001 O3
Industrial maintenance costs are indirectly paid
by the public in higher prices for products and
services. These costs are considered a nontaxable
cost of doing business on which profit margins and
markups are partially based. These difficult-to-
quantify factors were not included in the per capita
cost estimates as a function of ozone level. Losttax
revenues and markups could add considerably to
the cost of ozone damage. Preventive maintenance
is a nonproductive consumption of energy, labor,
and materials that could be redirected to improve
quality of life.
Effects on Other Materials
In a review of the effects of photochemical smog
on materials, Sanderson33 included possible
effects on plastic tract recorders and asphalt.
Because the effects were observed in laboratory
tests at extremely high ozone levels (200,000
/ug/m3 and 104,000 /ug/m3, respectively), they
should have no practical significance at ambient
levels.
Haynie and Upham20 reported a possible
beneficial effect of photochemical oxidants on the
corrosion behavior of steel. Their study used a
nonlinear multiple regression technique to
evaluate corrosion data as a function of field
environmental conditions. Laboratory studies19 did
not show any statistically significant effect of
ozone on steel corrosion. Some unmeasured factor
that was covariant with photochemical oxidants
could have caused the beneficial effect that was
observed in the field studies.
SUMMARY
Although many organic materials have been
shown to be susceptible to ozone attack, only
damage to certain paints, elastomers, and dyes
represents significant economic loss. Measures to
prevent ozone damage to elastomers and dyes are
a major cost. In contrast, the paint industry has not
recognized the economic significance of possible
ozone damage, and costs to prevent such attack
are not identifiable.
Figure 13-1 graphically presents a summary of
the estimated total annual per capita cost of ozone
damage and preventive measures as a function of
annual average concentration. The low and high
estimates are based on data error and lack of
confidence in many of the assumptions that were
necessarily made.
This figure predicts that considerable savings
from reduced damage to materials will be realized
as the annual average ozone levels are decreased.
This economic benefit is in addition to improved
health, agriculture, visibility, and ecology, andthus
it adds to the total quality of life.
Ozone materials damage as an individual
problem can be solved more cost effectively by
developing, identifying, and buying ozone-
resistant materials. This approach has the distinct
advantage of being applicable below natural
background levels of ozone because, in general,
there are no threshold limits to material damage.
100
ANNUAL AVERAGE OZONE CONCENTRATION,
Figure 13-1. Effect of annual average ozone concentration
on added costs resulting from damage to materials and
preventive measures.
REFERENCES FOR CHAPTER 13
1 Bailey, P S The reactions of ozone with organic
compounds Chem Rev 58 925-1010, 1958
2 Barrett, L. B. and T E Waddell Cost of Air Pollution
Damage—A Status Report AP-85, U S Environmental
Protection Agency, Research Triangle Park, N C ,
February 1973 pp 23-25
3 Beloin, N J Fading of dyed fabrics by air pollutants A
chamber study Text Chem Color 5128-133, 1973
340
-------
4. Belom, N J. A field study. Fading of dyed fabrics by air
pollution Text Chem. Color. 4 77-82, 1972.
5 Bogaty, H , K S Campbell,andW. D.Appel.Theoxidation
of celluloseby ozone in small concentrations Text. Res J
2281-83, 1952.
6 Braden, M , and A N. Gent The attack of ozone on
stretched rubber vulcanizates I Therateofcutgrowth.il.
Conditions for cut growth. J Appl Polym Sci 3 90-93,
100-106, 1960
7 Bradley, C. E., and A J. Haagen-Smit The application of
rubber in the quant native deter mi nation of ozone Rubber
Chem Technol 24750-755, 1951.
8. Campbell, G G,G G Schurr, D. E Slawikowski, and J
W Spence Assessing airpollutiondamagetocoatings. J.
Paint Technol 4659-71, 1974
9 Cloyd, D R , and W. J Murphy. Technology Survey
Handling Hazardous Materials NASASP-5032, National
Aeronautics and Space Administration, Washington,
D C , September 1965
10 Crabtree, J , and B S Biggs. Cracking of stressed rubber
by free radicals J Polym. Sci ; 7.280-281, 1953.
11 Crabtree, J, and A R. Kemp Accelerated ozone
weathering test for rubber Ind Eng. Chem Anal Ed.
18 769-774, 1946
12 Crabtree, J , and F S Malm Deterioration of rubberfrom
use and with age In Engineering Uses of Rubber. A T
McPherson and A. Klemm, eds Remhold Publishing
Corp, NY, 1956. pp 140-170
13 Devries, K L, and E R Simonson Section 2 Ozone and
elastomers In Ozone Chemistry Technology. A Review
of the Literature. 1961-1974. J. S. Murphy and J. R Orr,
eds. Franklin Institute Press, Philadelphia, Pa , 1975 pp
257-294
14 Dillon, C P The economic advantage of deferred topcoat
zinc-base painting systems. Mater. Perform. 13 29-31,
1975
15 Dorset, B C M Pollution and fading fabrics. Text Manuf
3927-29,31, 1972.
16 Edwards, D C., andE B Storey. A quantitative ozone test
for small specimens Chem. Can. 77:34-38, 1959.
17. Fisher, H L Antioxidation and antiozonation. In'
Chemistry of Natural and Synthetic Rubbers. Remhold
Publishing Corp, N Y, 1957 pp 49-55.
18 Haylock, J.C , and J.L Rush Studies on the ozone fading
of anthraqumone dyes on nylon fibers Tex Res J 461-
8, 1976
19 Haynie, F H , J W. Spence, and J. B Upham. Effects of
Gaseous Pollutants on Materials—A Chamber Study
EPA-600/3-76-015, U.S. Environmental Protection
Agency, Research Triangle Park, N C., February 1976.
20. Haynie, F. H., and J. B. Upham. Effects of atmospheric
pollutants on corrosion behavior of steels Mater Prot
Perform. 7018-21, 1971
21 Hofmann.C M.andR L Miller. Resistanceof passenger
tires to atmospheric exposure J. Mater. 4'31-43, 1969
22 Jellmek, H. H. G. Fundamental degradation processes
relevant to outdoor exposure of polymers In
Weatherability of Plastic Materials. Appl Polym Symp
(4)41-59, 1967
23 Kerr, N , M A Morris, and S H. Zgronian The effect of
ozone and laundering on a vat-dyed cotton fabric Am.
Dyest. Rep. 58.34-36, 1969.
24 The Kline Guide to the Paint Industry 1975. Charles H.
Kline and Co , Inc., Fairfield, N J., 1975.
25. Meyer, D A, and J G Sommer The Development of
Weather and Aging Resistant Pneumatic Tires and
Mechanical Rubber Goods Final Technical Report
Dayton Rubber Co , Dayton, Ohio, June 1957
26 Morris, M. A Effect of weathering on cotton fabrics Calif
Agnc Exp Stn Bull (823).1-29, 1966
27 Mueller,W. J.,andP. B Stickney A Survey and Economic
Assessment of the Effects of Air Pollution on Elastomers
NAPCA Contract No CPA-22-69-146, Battelle Memorial
Institute, Columbus, Ohio, June 1970
28 National Research Council Ozone and Other
Photochemical Oxidants National Academy of Sciences,
Washington, D C., 1977
29 Newton, R G Mechanism of exposure-cracking of
rubber (With a review of the influence of ozone ) J.
Rubber Res 7427-62, 1945
30 Norton, F J Action of ozone on rubbt. Rubber Age(N Y )
4787-90, 1940
31 Salvm, V S. Ozone fading of dyes Text Chem Color
7.245-251, 1969
32 Salvm, V. S ,andR A Walker Service fading of disperse
dyestuffs by chemical agents other than the oxides of
nitrogen Text Res J. 24 571-585, 1955
33. Sanderson, H P The effect of photochemical smog on
materials In Photochemical Air Pollution Formation,
Transport and Effects Environmental Secretariat
Publication No. NRCC 14096, National Research Council
of Canada, Associate Committee on Scientific Criteria for
Environmental Quality, Ottawa, Canada, 1975 pp 71-
87.
34 Schmitt, C H. A. Daylight fastness testing by the Langley
System. Am Dyest Rep 57:664-675,1962
35 Schmitt, C H. A. Lightfastness of dyestuffs on textiles
Am Dyest Rep 43974-980, 1960
36 Spence, J W., F. H Haynie and J B Upham Effects of
gaseous pollutants on paints A chamber study. J Paint
Technol 47 57-63, 1975
37. Upham, J B , F H. Haynie, and J. W Spence. Fading of
selected drapery fabrics by air pollutants J. Air Pollut
Control Assoc 26790-792, 1976.
38. Van Rossem, A , and H W Talem The appearance of
atmospheric cracks in stretched rubber Kautschuk 7 79-
86, 115-117, 1931.
39. Williams, I Oxidation of rubber exposed to light Ind Eng
Chem 18 367-369, 1926
40 Zeronian, S. H , K W Alger, andS.T Omaye Reaction of
fabrics made from synthetic fibers to air contaminated
with nitrogen dioxide, ozone, or sulfur dioxide In'
Proceedings of the Second International Clean Air
Congress H M EnglundandW T Berry, eds Academic
Press, Inc , N Y , 1971 pp 468-476
341
U S GOVERNMENT PRINTING OFFICE: 1979 659-345
------- |